To the Editor:             

Antipsychotics all share the property of blocking dopamine receptors in the brain.1 Almost all, including the “atypical antipsychotics, ” have been associated with extrapyramidal side effects.1,2 These drugs have been almost exclusively prescribed by psychiatrists who are aware of the risks involved. They have been trained in recognizing tardive dyskinesia (TD), and counsel their patients about this risk. They document their examination for TD with the Abnormal Involuntary Movement Scale as a standard routine.

When counseling for TD risk and routine evaluations are not performed, doctors are open to liability suits. Metoclopramide is widely used to treat gastric reflux disorders, gastroparesis, and nausea, and can cause the same extrapyramidal adverse effects, including TD, as caused by antipsychotics. It now carries a “black box” warning. Since it often produces dramatic improvement in gastrointestinal symptoms, is well tolerated, and is inexpensive, it is often continued for years, increasing the risk of TD. Since most doctors are unaware of the risk, patients are rarely told about the potential for TD and chart notes documenting an evaluation for movement disorders are rare. Potential suitors are recruited in television advertisements by personal injury law firms.

Aripiprazole, the most recently approved atypical antipsychotic has recently been implicated as a cause of TD in patients never exposed to other dopamine receptor-blocking agents3,4 and at the low doses recommended for treating depression.4 We are concerned that the recent Food and Drug Administration approval of aripiprazole for use in treating depression may result in an increase in the number of cases of TD. Primary care physicians (PCPs) often treat depression without consultation with psychiatrists. The addition of aripiprazole into the depression arena, assisted by mass market advertising, especially on television, may lull unsuspecting non-psychiatrists into believing that this drug, and possibly similar drugs in the future, are as risk-free in terms of TD as the selective serotonin reuptake inhibitors. The medical-legal issues that PCPs are currently facing with metoclopramide law suits may occur again in a few years with aripiprazole.

There are few data concerning the incidence of TD in neuroleptic-naïve patients starting aripiprazole. Although the atypical antipsychotics are believed to be “relatively” free of extrapyramidal side effects, the results of the Clinical Antipsychotic Trials in Intervention Effectiveness studies5 have questioned this belief. In a recent review6 of studies using comparable doses of antipsychotics, the annualized incidence of TD was estimated to be 3.9% for atypical antipsychotics compared with 5.5% for conventional antipsychotics—only a modest improvement.

Aripiprazole may produce TD, which may be permanent. It should therefore be used only when necessary. We believe that it should only be used for refractory cases by doctors knowledgeable about antipsychotics. The dose should be as low as possible and attempts to wean off the drug should be made once the patient has achieved a stable improvement. Alternatives include switching to other antidepressants, including the tricyclics; combined therapy with drugs of different chemical families; adjunctive exercise; and psychotherapy. Other alternatives such as lithium or thyroid augmentation should be considered by psychiatrists. TD should be specifically evaluated and commented upon at each office visit so that the drug may be stopped if TD begins.


Joseph H. Friedman, MD, and Daniel Tarsy, MD

Dr. Friedman is professor and chief of the Division of Movement Disorders in the Department of Neurology at Warren Alpert Medical School of Brown University/Butler Hospital in Providence, Rhode Island. Dr. Tarsy is professor in the Department of Neurology at Harvard Medical School/Beth Israel Deaconess Medical Center in Boston, Massachusetts.

Disclosures: Dr. Friedman is a consultant to Acadia and EMD serono; on the speakers’ bureaus of GlaxoSmithKline and Ingelheim-Boehringer; and receives research support from Cephalon, EMD Serono, Epivax, the National Institutes of Health, and The Michael J. Fox Foundation. Dr. Tarsy is a consultant to Esai Neuroscience; receives grant support from Allergan and Schwarz Pharma; and receives foundation support from The Michael J. Fox Foundation and the National Parkinson Foundation.


1. Tarsy D, Baldessarini RJ. Epidemiology of tardive dyskinesia: is risk declining with modern antipsychotics? Mov Disord. 2006;21(5):589-598.
2. Kane JM. Tardive dysinesia rates with atypical antipsychotics in adults: prevalence and incidence. J Clin Psychiatry. 2004;65(suppl 9):16-20.
3. Lungu C, Aia PG, Shih LC et al. Tardive dyskinesia due to aripiprazole: report of 2 cases. J Clin Psychopharmacol. 2009;29(2):185-186.
4. Friedman JH. Aripiprazole induced tardive dystonia in a neuroleptic naïve patient. J Clin Psychiatry. In press.
5. Miller DD, Caroff SN, Davis SM, et al. Extrapyramidal side effects of antipsychotics in a randomised trial. Br J Psychiatry. 2008;193(4):279-288.
6. Correll CU, Schenk EM. Tardive dyskinesia and new antipsychotics. Curr Opin Psychiatry. 2008;21(2):151-156.



Needs Assessment:
Suicide remains an important public health problem, and suicide prevention is a major challenge. Primary care physicians treat numerous patients with known risk factors for suicide. Early detection, along with vigorous treatment of affective and substance use disorders—the two most common disorders found in people who commit suicide—is considered the most effective way to prevent suicide. However, advances in suicide prevention are hampered by lack of sensitive measures to evaluate the effect of specific interventions. In this article, evidence that toxicologic monitoring of suicide may provide valid ex vivo markers of treatment rate and substance abuse is discussed in the context of known risk factors and the recent black box warnings against antidepressants.

Learning Objectives:

• Describe risk factors for suicide.
• Explain why the effects of most interventions on suicide prevention are unknown.
• Understand the value of postmortem toxicologic investigations in individual suicides.
• Identify the potential of ex vivo markers of treatment rate and substance abuse in suicides.

Target Audience: Primary care physicians and psychiatrists.

CME Accreditation Statement: This activity has been planned and implemented in accordance with the Essentials and Standards of the Accreditation Council for Continuing Medical Education (ACCME) through the joint sponsorship of the Mount Sinai School of Medicine and MBL Communications, Inc. The Mount Sinai School of Medicine is accredited by the ACCME to provide continuing medical education for physicians.

Credit Designation: The Mount Sinai School of Medicine designates this educational activity for a maximum of 3 AMA PRA Category 1 Credit(s)TM. Physicians should only claim credit commensurate with the extent of their participation in the activity.

Faculty Disclosure Policy Statement: It is the policy of the Mount Sinai School of Medicine to ensure objectivity, balance, independence, transparency, and scientific rigor in all CME-sponsored educational activities. All faculty participating in the planning or implementation of a sponsored activity are expected to disclose to the audience any relevant financial relationships and to assist in resolving any conflict of interest that may arise from the relationship. Presenters must also make a meaningful disclosure to the audience of their discussions of unlabeled or unapproved drugs or devices. This information will be available as part of the course material.

This activity has been peer-reviewed and approved by Eric Hollander, MD, chair and professor of psychiatry at the Mount Sinai School of Medicine, and Norman Sussman, MD, editor of Primary Psychiatry and professor of psychiatry at New York University School of Medicine. Review Date: October 24, 2007.

Drs. Hollander and Sussman report no affiliation with or financial interest in any organization that may pose a conflict of interest.

To receive credit for this activity: Read this article and the two CME-designated accompanying articles, reflect on the information presented, and then complete the CME posttest and evaluation. To obtain credits, you should score 70% or better. Early submission of this posttest is encouraged: please submit this posttest by November 1, 2009 to be eligible for credit. Release date: November 1, 2007. Termination date: November 30, 2009. The estimated time to complete all three articles and the posttest is 3 hours.


Dr. Dhossche is professor in the Department of Psychiatry and Human Behavior at the University of Mississippi Medical Center in Jackson.

Disclosure: Dr. Dhossche receives grant support from the American Foundation for Suicide Prevention.

Please direct all correspondence to: Dirk M. Dhossche, MD, PhD, Professor of Psychiatry, Department of Psychiatry and Human Behavior, University of Mississippi Medical Center, 2500 N State St, Jackson, MS 39216; Tel: 601-984-5805; Fax: 601-984-6965; E-mail:



Is surveillance of postmortem toxicology a useful way of evaluating the prior rate of psychiatric treatment and substance abuse in people who commit suicide? The question is an important public health issue given the lack of sensitive measures of suicide prevention and current controversy about the role of psychotherapeutic medications. This article reviews the literature on the toxicology of suicide. Although toxicologic investigations are considered an integral part of medical-forensic investigations in individual cases of suicide, very few states require comprehensive toxicology on all suicide victims. Instead, subjective determinations for toxicologic testing based on local policy and individual coroner or medical examiner preference are common practices. Systematic routine toxicologic monitoring in all suicides may serve as ex vivo markers of treatment rate and substance abuse patterns in suicides. The potential of such markers are noted in studies that evaluate risk factors of suicide, including the effects on suicide rates of changing prescription rates of psychotropic medications.

“Anatomy should be recalled from the dead.” —Andreas Vesalius (1514-1564)1



An important limitation of suicide research concerns the lack of access to the chief informant, ie, the deceased victim. Honoring Vesalius’ historic reappraisal of direct observation of anatomic structures over theory and speculation, the main tenet of this article concerns systematic toxicologic investigations of suicide, one of few objective pieces of information on the deceased victim, and its potential benefits both in individual cases and in epidemiologic research.

Ideally, toxicologic analysis should be conducted in every suspected suicide and other types of unnatural death as an integral component in the investigation, requiring correlation with a detailed scene inspection, an extensive exploration into the decedent’s medical and social background, and both gross and microscopic findings, to uncover suicidal ideation or intent. The toxicologic procedures aim to detect all substances present in various body tissues and fluids. Items such as route of administration, acute versus chronic dose, and consistency between drug concentrations and behavioral effects may be critical factors in assessing the manner of death. Toxicologic results can be useful for reconstructing some events before a suicide, and may suggest impaired mental functioning due to intoxication with alcohol or other drugs either acutely, chronically, or both. The presence of prescription medication may indicate recent contact with a physician. Examination for prescribed psychoactive medications may also be useful to estimate the frequency and type of psychiatric treatment before suicide.

The role of toxicologic investigations has only recently received attention as a promising research tool.2 In this article, toxicologic studies in suicide are discussed in the context of known risk factors and the recent black box warnings against antidepressants. More effective suicide prevention will also be discussed. Future studies needed to establish the utility of systematic and comprehensive toxicologic monitoring of suicides for further understanding of suicide.


Psychiatric Impairment

Systematic studies of consecutive suicides show that risk for suicide concentrates heavily in people with psychiatric illness,3 particularly depressive disorders and substance abuse. Postmortem psychiatric investigations4-14 show that a psychiatric disorder is diagnosable in most suicides. Any depressive disorder is present 47% to 89% of the time, any substance use disorder is present 19% to 51% of the time, and any schizophrenic disorder is present 2% to 16% of the time. No disorder was found in ≤19% of cases. In a study of adolescent suicide, the group with no apparent disorder still had more risk factors for suicide (eg, family history of suicide, past suicidal behavior, legal problems) than community controls.15

The lifetime risk of suicide is often quoted as 15% for affective disorders and alcoholism and 10% for schizophrenia. These rates mot likely apply to selected high-risk populations. Lower estimates on suicide rates in the group of affective disorders16,17 and substance use disorders17,18 (ie, approximately 2% to 7% in both groups), have been advanced. The general risk for suicide in schizophrenia is most likely closer to 4%.17


Facilitating Factors, Clinical Syndromes, and Psychopathology

Some risk factors for suicide, including clinical syndromes, are shown in Table 1. The list is not exhaustive. Severe emotional and cognitive turmoil is likely to precede suicide. Symptoms may be described as lowered mood, anxiety, rage, desire/impulse for self destruction, agitation, hopelessness, despair, guilt, or other cognitive distortions. Separating out these psychological states is difficult. Descriptors more often reflect the author’s frame of reference than a distinct abnormality. For example, in readings of Freud,19 it is speculated that anger and hatred can become self-directed, lead to depression, and be a motivating force in suicide. Recent psychodynamic formulations20,21 have stressed that in addition to affective states such as rage, hopelessness, despair, and guilt, it is also important to consider cognitive factors that lead people to view suicide as a permanent solution to their likely temporary predicament. Individuals can come to see suicide as a reunion, rebirth, retaliatory abandonment, revenge, or self-punishment.



Tendencies of suicidal people to engage in extreme aggression are illustrated in cases of murder followed by suicide. The incidence of murder-suicide has been relatively constant (and infrequent) over time across industrialized countries at approximately 0.2–0.3 per 100,000 people each year.22 In another report,23 violent behavior in the past year was approximately five times more frequent in suicides than accidental deaths. This finding was not explained by greater aggression in subjects with alcohol abuse. Reports conflict regarding the importance of aggressive personality traits as predictors of suicide.24,25 A link between suicide and violence is theoretically attractive as both have been associated with disturbances of central serotonin neurotransmission.26 However, further characterization of the type of aggression preceding suicide and further studies on psychological correlates of neurotransmitter disturbances are warranted before conclusions can be drawn.


Evaluation of Suicide Prevention

Suicide is thought to be preventable, although there is currently no foolproof evidence for efficacy of any preventive action. Theoretically, quick and effective treatment of emerging depression or other mental disorders in every person at any time would most likely be most effective (but practically difficult to implement) in reducing suicide rates given the fact that suicide does not or rarely occurs without psychiatric impairment. General approaches such as efforts to educate physicians about diagnosing and treating depressive disorders and to restrict access to lethal means have been supported as most promising ways to reduce suicides while other methods including public education, screening programs, and media education are equivocal and require more testing.27

Lack of sensitive measures poses major methodologic problems to evaluate the success or failure of suicide prevention. This important shortcoming has been discussed in the evaluation of the world’s first comprehensive suicide prevention program in Finland.28 After program implementation, a 20% reduction of the suicide rate was observed until 1996. Since then, the suicide rate has remained stable at 9% below pre-project levels. However, it is unclear which component, if any, of the prevention program was effective, as not enough data to control for confounding variables and to evaluate outcome were included in the program’s design. Reviews of other suicide prevention programs have concluded similarly.29

The two most important obstacles for progress in suicide research are the low base rate of suicide and ethical concerns of studying suicidal people in controlled trials of medications or psychotherapy. The low base-rate of suicide requires the study of large populations to yield meaningful and significant results. These costly and lengthy studies have not been done yet. It is difficult to know if anything actually prevents a suicide because of statistical constraints of the prediction of infrequent events. The issue has been summarized by Murphy,30 who stated that “if suicide is difficult to predict, its prevention is even more difficult to detect.”

Tackling the other problem of including suicidal people in clinical trials of drug or psychotherapy treatment would require acceptance by ethics committees that suicide is at times an unfortunate but difficult-to-avoid outcome of psychiatric illness in some people. Any studies that enroll people with suicidal behaviors should have sufficient safeguards to protect participants from unnecessary risk.


Substance Abuse

The literature supports a strong association between chronic substance abuse, suicide attempts,31 and suicide.32-34 A recent analysis found the highest standardized mortality rates (SMR) for suicide for mixed drug use (SMR 1,685; 95% CI 1,473–1,920), followed by intravenous drug use (SMR 1,373; 95% CI 1,029–1,796), opioid-use disorders (SMR 1,351; 95% CI 1,047–1,715), and alcohol-use disorders (SMR 979; 95% CI 898–1,065).35 In these calculations, an SMR of 100 indicates that the observed number of suicides is the same as the expected number of suicides, whereas an SMR of 1,000 indicates a 10-fold greater number of observed (in substance users) compared to expected (in the general population) suicides.


Psychotropic Medications

Antidepressants, lithium, and clozapine have been ascribed anti-suicide properties for selected patient groups, but there are conflicting results mostly due to methodologic problems. The problem of assessing the impact of psychotropic medications on suicide has long been anticipated by Murphy30:

It is important to realize that the absence of a suicide generates no data. Thus, we can never prove what has been accomplished. Yet, we can hardly doubt that it occurs. This argument would be more satisfying if there had been a gradual reduction in the national suicide rate since studies of the late 1950s clearly linked suicide to psychiatric illness or if it had paralleled the growing use of antidepressants. It might then be concluded that physicians had learned from these studies and were more alert to the indicators of risk, at least that they were recognizing and treating more depressions. In fact, the suicide rate has risen during that period. But it has not done so uniformly. It has actually fallen among the older age groups, those most likely to see a physician. It has risen most—and considerably—among adolescents and young adults. Evidence that is fragmentary at present suggests that this group is far less likely to have been under a physician’s care. Thus, the opportunity to treat may not have occurred in many of these cases.30

Recent studies confirm that approximately 25% to 33% of suicides received psychiatric care in the year before committing suicide.36-38

The evidence that the use of psychotropic medications can prevent suicides is considerable.39-44 The effect is logically stronger if the medication is prescribed appropriately, in sufficient dose, together with other psychotherapeutic and psychosocial interventions, and continually in people at high risk.45,46 Prospective studies providing definitive evidence for anti-suicide effects of psychotropic medications face great methodologic problems and may never be done. In contrast, the evidence that the use of antidepressants increases suicide is equivocal,47,48 although it is possible that in some patients, antidepressant treatment increases suicidal ideas and attempts.49,50


Toxicology of Suicide

Systematic, comprehensive toxicologic studies in suicides and other violent deaths have been infrequently conducted. There is only one study with comprehensive toxicologic data on suicides in the literature that has examined all suicides in a geographically defined area. This study was done in 1,348 suicides from Finland during 1987–1988, representing 97% of all suicides during that period.51 There were 1,032 men (77%) and 316 women (23%) in the sample. The most frequent methods in men were hanging (34%), firearms (26%), and overdose (13%). In women, overdose was the most frequently used method (41%), followed by hanging (29%) and drowning (14%). More than 150 commonly prescribed and/or abused drugs were toxicologically detected. Drugs were found in 42% of cases, in 67% of women, and 34% of men. Alcohol was detected more frequently in men (41%) than in women (20%). Benzodiazepines were the most frequently detected specific prescription drug in both sexes (23% of all cases), followed by neuroleptics (13%) and antidepressants (8%). In cases with overdose, a single drug was found in 29%, two drugs in 30%, three drugs in 23%, four drugs in 15%, and five or more drugs in 4%. The relative risk (RR) of suicide was calculated for different drugs relating the number of suicides committed by use of each drug to its national sales. The highest risk was found for barbiturates (RR=105), followed by imipramine (RR=18), maprotiline (RR=12), doxepin (RR=12), and dextropropoxyphene (RR=10). The lowest risk was found for benzodiazepines (RR=0.33).

Most other studies have typically focused on a particular substance, eg, alcohol and/or cocaine52,53 or one class of substances, eg, prescription psychotropics.54 In most studies, the most frequent substance detected is alcohol. The proportions of alcohol-positive suicides have varied in the 30% to 40% range.51,55-57 Several studies have reported the rates of alcohol detection for various methods of suicide in cases of all ages (Table 2).51,56,58-61 Alcohol detection among suicides by gunshot in these reports tends to be toward the high end of the 30% to 40% range.51,56,58-61 The same is true, however, for overdoses and carbon monoxide poisoning. These findings suggest that, in general, the presence of alcohol is not preferentially associated with any particular suicide method.



Prescription psychotropic medications were the focus of a New York City study.54 Among 1,970 certified suicides from 1990–1992, 1,635 (83%) of cases had valid toxicology findings. Overdose was the method of suicide in 293 cases (18%). Antidepressants and neuroleptic medications were detected in 268 (16%) of 1,635 suicides studied. There were more detections in women and whites. Age was not associated with detection of these medications. In approximately 50% of cases that died by overdose, a prescription psychotropic medication was found. Conversely, approximately 50% of those with positive detection of an antidepressant or neuroleptic drug overdosed. The other 50% used a method other than poisoning, such as guns or fall from height. From this study, it seems that only a small proportion of suicide victims took antidepressants or neuroleptics in the days before their death. The authors of the study54 comment that it is unknown if wider use of psychotropic medication in people with psychiatric conditions can reduce the number of suicide, but that the null hypothesis (ie, that use of psychotropic medications does not affect suicide rates) requires that individuals who commit suicide have been prescribed and have taken those medications. Findings suggest that this is not the case in most suicides.

The Centers for Disease Control and Prevention62 reported that 13 states collected data for the National Violent Death Reporting System (NVDRS) in 2004. None of the states conducted comprehensive alcohol and drug screenings on all suicide victims, despite evidence of substance use among substantial numbers of suicide victims. Descriptions of cases selected for toxicology screening suggest subjective determinations for testing on the basis of local policy and individual coroner or medical examiner preference.63 It was found that the percentage of suicide victims tested varied among states, ranging from 25.9% to 97.7%. Among all suicide victims with positive test results, the greatest percentage tested positive for alcohol (33.3%), followed by opiates (16.4%), cocaine (9.4%), marijuana (7.7%), and amphetamines (3.9%). A similar percentage of poisoning suicide (ie, suspected intentional overdose) and non-poisoning suicide victims tested positive for alcohol or other drugs, with the exception of opiates. Overall, these findings emphasize the need to continue monitoring toxicology test results of suicide victims, which might identify geographic and temporal patterns of substance use that can help guide development of effective suicide interventions. Uniform, comprehensive, toxicology testing practices on a state and national basis should be adopted for better understanding of suicide and development of effective interventions.   


Ex Vivo Marker of Treatment Rate

Several studies support that toxicologic monitoring is a promising marker of treatment rate in suicides. For example, in a study of toxicologic data in 333 consecutive suicides occurring in Mobile County, Alabama, between 1990 and 1998,64 it was found that detection rates of antidepressants were low (ie, 20%) in suicides. This suggests undertreatment of depressive disorders in people who commit suicide. Alcohol was detected in 33% of cases. Extensive overlap in detections of prescribed psychotropics and abusable substances suggests that patients at risk for suicide should be advised not to use abusable substances.

In another study,65 toxicologic findings were compared between 179 suicides in San Diego County, California, from 1981–1982,8 and 333 suicides in Mobile County, Alabama, from 1990–1998. Toxicologic detection rates were similar for most categories of psychoactive substances in suicides from Mobile, Alabama (1990–1998) and San Diego, California (1981–1983). Higher detection rates of antidepressants in Mobile suicides (20% versus 8% in San Diego) suggest higher treatment rates of depression in recent samples. Toxicologic detection rates of neuroleptics in suicides vary greatly between studies, occuring from 2% to 7% in United States suicide samples (Mobile County suicide sample,64 San Diego suicide sample,65 1995 New York City suicide sample54), 7% in a Swedish study,66 and 23% in a Finnish study.51 In the Mobile sample, detection rates of neuroleptics increased from ≤1% to 6% over an 8-year period.64

In order to make maximum use of systematic toxicologic monitoring, future study designs need to collect and correlate three pieces of information. First, all suicides in a defined geographic area need be recorded over a period of time. Second, comprehensive toxicologic findings need to be done in all cases. Last, prescription medication data in the years preceding suicide need to gathered. This design will allow for testing the hypothesis that, first, positive detection rate of prescription psychotropic medications will be higher in suicide victims who have been prescribed psychotropic medications the last year before death than in suicide victims who have not been prescribed psychotropic medications; and second, positive detection rate of alcohol, cocaine, and cannabis will be higher in suicide victims who have not been prescribed psychotropic medications during the year before death than in suicide victims who have been prescribed such medications. Such an ideal study has not been conducted as of yet, but would be necessary to provide further support for routine toxicologic monitoring in all suicides as an important ex vivo measure of treatment rate in suicides.


Ex Vivo Marker of Substance Abuse

Toxicologic monitoring of suicide is also a promising marker for rates of substance use in people at risk for suicide. If interventions to decrease substance abuse in people at risk for suicide are successful, one would like to see a decrease in detection of alcohol, cocaine, and/or cannabis as well as other street drugs. In a review of suicide studies with both comprehensive toxicologic and diagnostic data,67 the sensitivity of alcohol detection to diagnose alcohol and substance use disorders in suicides was low in all studies (range=39% to 42%). The specificity was between 80% and 95%. It was concluded that most suicides with positive alcohol detection seem to suffer from chronic substance abuse problems, although the high rate of false negatives limit the sensitivity of the testing.


Black Box Warnings

In 2003, the US Food and Drug Administration issued a public health advisory about the risk of suicidality in pediatric patients taking selective serotonin reuptake inhibitors (SSRIs) for depression. In 2005, the agency mandated a black box warning and medication guide indicating that pediatric and adult patients may be at risk. Recent studies have reported that the FDA advisory was associated with significant reductions in aggregate rates of diagnosis and treatment of pediatric depression. Evaluation of a large pediatric cohort with newly diagnosed episodes of depression68 showed that from 1999 to 2004 pediatric diagnoses of depression increased from 3 to 5 per 1,000. After the FDA advisory was issued, the national rate decreased to levels observed in 1999, a significant deviation from the historic trend. Pediatricians and nonpediatrician primary care physicians accounted for the largest reductions in new diagnoses. Among patients with depression, the proportion receiving no antidepressant increased to three times the rate predicted by the pre-advisory trend, and SSRI prescription fills were 58% lower than predicted by the trend. Similar findings were presented in analyses of another database.69

Despite the focus of the policy on pediatric patients, the FDA advisory had significant effects on the treatment for adults with depression.70 In a large cohort (1998–2005) with newly diagnosed episodes of depression, the rate of diagnosed depression was significantly lower after the advisory than would have been expected on the basis of the pre-advisory historic trend. The average percentage of adults with new (versus recurrent) depressive episodes was 88.6% in the pre-advisory period (declining at an annual rate of 1.69%), and it decreased significantly to 77.5% (declining more rapidly, at an annual rate of 7.7%). The percentage of adults with depression who did not receive an antidepressant increased from an average of 20% (declining at 0.45% annually) before the policy action to an average of 30% (increasing at an annual rate of 20.6%). The data did not show any compensatory increases in psychotherapy or prescription of atypical antipsychotics or anxiolytics. Specific criticisms of analytic procedures and inferences leading to these black box warnings have been articulated,71 emphasizing that no suicide occurred in clinical trials of approximately 4,400 children and that the composite variable labeled “suicidality” is an unvalidated and inappropriate surrogate.

It is striking that toxicologic studies of suicide have not featured in the discussion of the alleged increased urge for suicide by antidepressants. In fact, the hypothesis that treatment of depressed individuals with SSRIs leads to an increased risk of suicide was rejected in several studies using toxicologic data from adult and youth suicide.72-76 For example, in a Utah study72 of 151 consecutive youth suicides in individuals 13–21 years of age between August 1, 1996, and June 6, 1999, toxicologic studies were completed for 91% (N=137) of subjects. Fifteen percent screened positive for alcohol. Medications were found in 18%, but common psychotropic medications (antidepressants, antipsychotics, and mood stabilizers) were found only in 3% (4 of 137). Studies of youth suicide in New York City reported that in the years 1993 through 1998,73 there were 66 suicides among youths <18 years of age. Toxicology was tested in 58 (87.9%) of the 66 suicides. None of the victims had paroxetine detected in their blood obtained at the time of autopsy. Imipramine was detected in two victims and fluoxetine in another two. The same group found that, from 1999 through 2002, there were 41 individuals in New York City <8 years of age who committed suicide.75 Thirty-six (87.8%) had a serum toxicologic analysis. There was one (2.8%) suicide in which both bupropion and sertraline were detected at the time of autopsy. Antidepressants were not detected in any of the other youth suicides.

Isacsson and colleagues74 compared detections of different antidepressants in the forensic toxicologic screening of 14,857 suicides and 26,422 cases of deaths by accident or natural causes in Sweden between 1992 and 2000. There were 3,411 (20%) detections of antidepressants in the suicide victims and 1,538 in the controls. SSRIs had lower odds ratios than the other antidepressants. In the 52 suicide victims <15 years of age, no SSRIs were detected. Among the victims 15–19 years of age, antidepressants were detected in 13 (4%) of 326 cases. SSRIs had lower relative risk in suicides compared with non-SSRIs. Taken together, these studies support undertreatment of depression and other psychiatric conditions in youths at risk for suicide, rather than supporting antidepressant use as an important risk factor for suicide in this age group. A similar argument applies to adult populations.39-44

In conclusion, evidence obtained from toxicologic studies in adult and youth suicide victims, which goes against the hypothesis that treatment of depressed individuals with SSRIs and other types of antidepressants leads to an increased risk of suicide, was not considered in the FDA advisory and black box warnings. Future toxicologic studies comparing detection rates of antidepressants in suicide victims should be informative as to the extent of decreased rates of depression treatment, especially in the younger age groups.



The routine use of comprehensive toxicologic evaluation of suicide and other violent deaths is a cost-effective tool to ensure completeness of forensic evaluations. Although toxicologic investigation is considered an integral part of medical-forensic investigations in individual cases of suicide, very few states require comprehensive toxicology on all suicide victims. Instead, subjective determination for toxicologic testing based on local policy and individual coroner or medical examiner preference is a common practice. Toxicologic monitoring of suicide should be more widespread. Systematic toxicologic monitoring of suicides have largely confirmed known problems in suicide prevention (ie, the undertreatment of psychiatric disorders, particularly depression, and the deleterious effect of alcohol abuse and other abusable substances on propensity for suicide). Future studies over time and in different areas should assess changes in medical treatment rates of people at risk for suicide as suggested by changes in detection rates of prescription psychotropic substances. In addition, changes in substance use patterns of alcohol and street drugs can be monitored this way. Further research should seek to strengthen the evidence that toxicologic measures are valid ex vivo markers of treatment rate and substance use. PP



1.    Vesalius A. On the Fabric of the Human Body [Latin]. Evanston, IL: Northwestern University; 2003.
2.    Dhossche D. Toxicology of suicide: touchstone for suicide prevention? Med Sci Monit. 2003;9(4):SR9-SR19.
3.    Harris E, Barraclough BM. Suicide as an outcome for mental disorders. Br J Psychiatry. 1997;170:205-228.
4.    Robins E, Murphy GE, Wilkinson RH Jr, Gassner S, Kayes J. Some clinical considerations in the prevention of suicide based on a study of 134 successful suicides. Am J Public Health. 1959;49(7):888-899.
5.    Dorpat TL, Ripley HS. A study of suicide in the Seattle area. Compr Psychiatry. 1960;1:349-359.
6.    Barraclough B, Bunch J, Nelson B, Sainsbury P. A hundred cases of suicide: clinical aspects. Br J Psychiatry. 1974;125(0):355-373.
7.    Chynoweth R, Tonge JI, Armstrong J. Suicide in Brisbane–a retrospective psychosocial study. Aust N Z J Psychiatry. 1980;14(1):37-45.
8.    Rich CL, Young D, Fowler RC. San Diego suicide study. I. Young vs old subjects. Arch Gen Psychiatry. 1986;43(6):577-582.
9.    Arató M, Demeter E, Rihmer Z, Somogyi E. Retrospective psychiatric assessment of 200 suicides in Budapest. Acta Psychiatr Scand. 1988;77(4):454-456.
10.    Runeson B. Mental disorder in youth suicide. DSM-III-R Axes I and II. Acta Psychiatr Scand. 1989;79(5):490-497.
11.    Beskow J. Suicide in mental disorder in Swedish men. Acta Psychiatr Scand Suppl. 1979;(277):1-138
12.    Asgård U. A psychiatric study of suicide among urban Swedish women. Acta Psychiatr Scand. 1990;82(2):115-124.
13.    Cheng AT. Mental illness and suicide. A case-control study in east Taiwan. Arch Gen Psychiatry. 1995;52(7):594-603.
14.    Shaffer D, Gould MS, Fisher P, et al. Psychiatric diagnosis in child and adolescent suicide. Arch Gen Psychiatry. 1996;53(4):339-348.
15.    Brent D, Perper J, Moritz G, Baugher M, Allman C. Suicide in adolescents with no apparent psychopathology. J Am Acad Child Adolesc Psychiatry. 1993;32(3):494-500.
16.    Blair-West G, Mellsop G, Eyeson-Annan M. Down-rating lifetime suicide risk in major depression. Acta Psychiatr Scand. 1997;95(3):259-263.
17.    Inskip H, Harris E, Barraclough BM. Lifetime risk of suicide for affective disorder, alcoholism and schizophrenia. Br J Psychiatry. 1998;172:35-37.
18.    Murphy GE, Wetzel RD. The lifetime risk of suicide in alcoholism. Arch Gen Psychiatry. 1990;47(4):383-392.
19.    Freud S. Mourning and Melancholia. Vol 4. London, UK: Hogarth Press; 1949.
20.    Hendin H. The psychodynamics of suicide. J Nerv Ment Dis. 1963;136:236-244.
21.    Hendin H. Psychodynamics of suicide, with particular reference to the young. Am J Psychiatry. 1991;148:1150-1158.
22.    Marzuk P, Tardiff K, Hirsch C. The epidemiology of murder-suicide. JAMA. 1992;267(23):3179-3183.
23.    Conner KR, Cox C, Duberstein PR, Tian L, Nisbet PA, Conwell Y. Violence, alcohol, and completed suicide: a case-control study. Am J Psychiatry. 2001;158(10):1701-1705.
24.    Angst J, Clayton P. Premorbid personality of depressive, bipolar and schizophrenic patients with special reference to suicidal issues. Compr Psychiatry. 1986;27(6):511-532.
25.    Goldney R, Winefield A, Saebel J, Winefield H, Tiggeman M. Anger, suicidal ideation, and attempted suicide: a prospective study. Compr Psychiatry. 1997;38(5):264-268.
26.    Linnoila V, Virkkunen M. Aggression, suicidality, and serotonin. J Clin Psychiatry. 1992;53(suppl):46-51.
27.    Mann J, Apter A, Bertolote J, et al. Suicide prevention strategies: a systematic review. JAMA. 2005;294(16):2064-2074.
28.    Beskow J, Kerkhof A, Kokkola A, Uutela A. Suicide Prevention in Finland 1986-1996: External Evaluation by an International Peer Group. Helsinki, Finland: Ministery of Social Affairs and Health; 1999.
29.    Breton JJ, Boyer R, Bilodeau H, Raymond S, Joubert N, Nantel MA. Is evaluative research on youth suicide programs theory-driven? The Canadian experience. Suicide Life Threat Behav. 2002;32(2):176-190.
30.    Murphy G. The prediction of suicide: why is it so difficult? Am J Psychotherapy. 1984;38(3):341-349.
31.    Dhossche D, Meloukheia AM, Chakravorty S. The association of suicide attempts and comorbid depression and substance abuse disorders in psychiatric consultation patients. Gen Hosp Psychiatry. 2000;22(4):281-288.
32.    Murphy GE. Suicide and substance abuse. Arch Gen Psychiatry. 1988;45(6):593-594.
33.    Rich CL, Fowler RC, Young D. Substance abuse and suicide. The San Diego study. Ann Clin Psychiatry. 1989;1:79-85.
34.    Mann R, Zalcman R, Smart R, Rush B, Suurvali H. Alcohol consumption, Alcoholics Anonymous membership, and suicide mortality rates, Ontario, 1968-1991. J Stud Alcohol. 2006;67(3):445-453.
35.    Wilcox HC, Conner KR, Caine ED. Association of alcohol and drug use disorders and completed suicide: an empirical review of cohort studies. Drug Alcohol Depend. 2004;76(suppl):S11-S19.
36.    Luoma J, Martin C, Pearson J. Contact with mental health and primary care providers before suicide: a review of the evidence. Am J Psychiatry. 2002;159(6):909-916.
37.    Ho TP. Psychiatric care of suicides in Hong Kong. J Affect Disord. 2003;76(1-3):137-142.
38.    Abe R, Shioiri T, Someya T. Suicide in Japan (letter). Psychiatr Serv. 2007;58(7):1013.
39.    Isacsson G, Bergman U, Rich CL. Epidemiological data suggest antidepressants reduce suicide risk among depressives. J Affect Disord. 1996;41(1):1-8.
40.    Isacsson G, Holmgren P, Druid H, Bergman U. The utilization of antidepressants–a key issue in the prevention of suicide: an analysis of 5,281 suicides in Sweden during the period 1992-1994. Acta Psychiatr Scand. 1997;96(2):94-100.
41.    Gibbons RD, Hur K, Bhaumik DK, Mann JJ. The relationship between antidepressant medication use and rate of suicide. Arch Gen Psychiatry. 2005;62(2):165-172.
42.    Korkeila J, Salminen JK, Hiekkanen H, Salokangas R. Use of antidepressants and suicide rate in Finland: an ecological study. J Clin Psychiatry. 2007;68(4):505-511.
43.    Bramness J, Walby F, Tverdal A. The sales of antidepressants and suicide rates in Norway and its counties 1980-2004. J Affect Disord. 2007;102(1-3):1-9.
44.    Nakagawa A, Grunebaum MF, Ellis SP, et al. Association of suicide and antidepressant prescription rates in Japan, 1999-2003. J Clin Psychiatry. 2007;68(6):908-916.
45.    Angst J, Angst F, Gerber-Werder R, Gamma A. Suicide in 406 mood-disorder patients with and without long-term medication: a 40 to 44 years’ follow-up. Arch Suicide Res. 2005;9(3):279-300.
46.    Søndergård L, Lopez AG, Andersen PK, Kessing LV. Continued antidepressant treatment and suicide in patients with depressive disorder. Arch Suicide Res. 2007;11(2):163-175.
47.    Wessely S, Kerwin R. Suicide risk and the SSRIs. JAMA. 2004;292(3):379-381.
48.    Möller H. Is there evidence for negative effects of antidepressants on suicidality in depressive patients? Eur Arch Psychiatry Clin Neurosci. 2006;256(8):476-496.
49.    Tiihonen J, Lönngvist J, Wahlbeck K, Klaukka T, Tanskanen A, Haukka J. Antidepressants and the risk of suicide, attempted suicide, and overall mortality in a nationwide cohort. Arch Gen Psychiatry. 2006;63(12):1358-1367.
50.    Hall W, Lucke J. How have the selective serotonin reuptake inhibitor antidepressants affected suicide mortality? Aust N Z J Psychiatry. 2006;40(11-12):941-950.
51.    Öhberg A, Vuori E, Ojanperä I, Lönngvist J. Alcohol and drugs in suicides. Br J Psychiatry. 1996;169(1):75-80.
52.    Marzuk PM, Tardiff K, Leon AC, Stajic M, Morgan EB, Mann JJ. Prevalence of cocaine use among residents of New York City who committed suicide during a one-year period. Am J Psychiatry. 1992;149(3):271-375.
53.    Garlow SJ. Age, gender, and ethnicity differences in patterns of cocaine and ethanol use preceding suicide. Am J Psychiatry. 2002;159(4):615-619.
54.    Marzuk PM, Tardiff K, Leon AC, et al. Use of prescription psychotropic drugs among suicide victims in New York City. Am J Psychiatry. 1995;152(10):1520-1522.
55.    McGuire FL, Birch H, Gottschalk LA, Heiser JF, Dinovo EC. A comparison of suicide and non-suicide deaths involving psychotropic drugs in four major U.S. cities. Am J Public Health. 1976;66(11):1058-1061.
56.    Goodman RA, Istre GR, Jordan FB, Herndon JL, Kelaghan J. Alcohol and fatal injuries in Oklahoma. J Stud Alcohol. 1991;52(2):156-161.
57.    Kubo SI, Dankwarth G, Puschel K. Blood alcohol concentrations of sudden unexpected deaths and non natural deaths. Forensic Sci Int. 1991;52(1):77-84.
58.    Rich CL, Dhossche DM, Ghani S, Isacsson G. Suicide methods and presence of intoxicating abusable substances: some clinical and public health implications. Ann Clin Psychiatry. 1998;10(4):169-175.
59.    James IP. Blood alcohol levels following successful suicide. Q J Stud Alcohol. 1966;27(1):23-29.
60.    Riddick L, Luke JL. Alcohol-related deaths in the District of Columbia–a postmortem study. J Forensic Sci. 1978;23(3):493-502.
61.    Hayward L, Zubrick SR, Silburn S. Blood alcohol levels in suicide cases. J Epidemiol Community Health. 1992;46(3):256-260.
62.    Centers for Disease Control and Prevention (CDC). Toxicology testing and results for suicide victims–13 states, 2004. MMWR Morb Mortal Wkly Rep. 2006;55(46):1245-1248.
63.    Shields LB, Hunsaker DM, Hunsaker JC 3rd, Ward MK. Toxicological findings in suicide: a 10-year retrospective review of Kentucky medical examiner cases. Am J Forensic Med Pathol. 2006;27(2):106-112.
64.    Dhossche DM, Rich CL, Ghani SO, Isacsson G. Patterns of psychoactive substance detection from routine toxicology in suicide in Mobile, Alabama, between 1990-1998. J Affect Disord. 2001;64(2-3):167-174.
65.    Dhossche DM, Rich CL, Isacsson G. Psychoactive substances in suicides. Comparison of toxicologic findings in two samples. Am J Forensic Med Pathol. 2001;22(3):239-243.
66.    Isacsson G, Holmgren P, Druid H, Bergman U. Psychotropics and suicide prevention. Implications from toxicological screening of 5281 suicides in Sweden 1992-1994. Br J Psychiatry. 1999;174:259-265.
67.    Dhossche DM. A review of postmortem alcohol detection as a diagnostic test for substance abuse disorders in suicide. Am J Forensic Med Pathol. 2000;21(4):330-334.
68.    Libby AM, Brent DA, Morrato EH, Orton HD, Allen R, Valuck RJ. Decline in treatment of pediatric depression after FDA advisory on risk of suicidality with SSRIs. Am J Psychiatry. 2007;164(6):884-891.
69.    Nemeroff CB, Kalali A, Keller MB, et al. Impact of publicity concerning pediatric suicidality data on physician practice patterns in the United States. Arch Gen Psychiatry. 2007;64(4):466-472.
70.    Valuck RJ, Libby AM, Orton HD, Morrato EH, Allen R, Baldessarini RJ. Spillover effects on treatment of adult depression in primary care after FDA advisory on risk of pediatric suicidality with SSRIs. Am J Psychiatry. 2007;164(8):1198-1205.
71.    Klein DF. The flawed basis for FDA post-marketing safety decisions: the example of anti-depressants and children. Neuropsychopharmacology. 2006;31(4):689-699.
72.    Gray D, Achilles J, Keller T, et al. Utah youth suicide study, phase I: goverment agency contact before death. J Am Acad Child Adolesc Psychiatry. 2002;41(4):427-434.
73.    Leon AC, Marzuk PM, Tardiff K, Teres JJ. Paroxetine, other antidepressants, and youth suicide in New York City: 1993 through 1998. J Clin Psychiatry. 2004;65(7):915-918.
74.    Isacsson G, Holmgren P, Ahlner J. Selective serotonin reuptake inhibitor antidepressants and the risk of suicide: a controlled forensic database study of 14,857 suicides. Acta Psychiatr Scand. 2005;111(4):286-290.
75.    Leon AC, Marzuk PM, Tardiff K, Bucciarelli A, Markham Piper T, Galea S. Antidepressants and youth suicide in New York City, 1999-2002. J Am Acad Child Adolesc Psychiatry. 2006;45(9):1054-1058.
76.    Vieweg WV, Pandurangi AK, Anum EA, Lanier JO, Fierro MF, Fernandez A. Toxicology findings in child and adolescent suicides in Virginia: 1987-2003. Prim Care Companion J Clin Psychiatry. 2006;8(3):142-146.


Dr. Bursztajn is associate clinical professor of psychiatry and Mr. Brodsky is senior research associate, both in the Department of Psychiatry at Harvard Medical School in Boston, MA.

Acknowledgments: The authors thank A. Stone Freedberg, MD, Patricia M.L. Illingworth, PhD, JD, and members of the Program in Psychiatry and the Law at Harvard Medical School, Massachusetts Mental Health Center, for their helpful dialogue. The authors report no financial, academic, or other support of this work.



How does managed care contribute to the psychiatric hazards of medical illness? How can primary care physicians and psychiatrists (as treaters or consultants) recognize and manage the clinical dynamics that result, together with their liability implications? For patients in managed healthcare settings, a latent subjective sense of captivity triggered by care restrictions can exacerbate feelings of helplessness and hopelessness brought on by the threat of serious illness. This sense of captivity can also intensify the patient’s feelings of alienation and betrayal when managed care constrains patient-physician decision making. These emotional dynamics, together with the rigid, defensive reactions to which physicians sometimes fall prey in the face of managed care’s restrictions on professional autonomy, can compromise the patient-physician relationship and the quality of clinical care while heightening liability risk. In the extreme, some patients will experience the escalating distress characteristic of posttraumatic stress disorder. Clinicians can prevent this emerging form of iatrogenesis by using such strategies as crisis anticipation and preparedness, intervention, and damage control, with the goals of enhancing the patient’s decision-making capacity and recovering a sense of choice in the patient-physician relationship. The introductory, ongoing, and termination phases of the treatment alliance are analyzed as focal points for particular clinical interventions.



There has been a growing recognition of the influence of managed health care on the doctor-patient relationship,1 including the forcing of primary care physicians into an alienating, ethically ambiguous, and clinically conflicting “gatekeeper” role.2,3 While financial constraints have always been a factor in clinical cases, one major consequence of managed care practices has been a concern as to the lack of meaningful healthcare choices for both doctors and patients. In this atmosphere, patients and their physicians—who face the threat or actuality of a major illness—may experience heightened feelings of helplessness and hopelessness, especially when healthcare benefits are denied. More dangerous still, some patients and physicians come to distrust each other, as when patients feel that physicians are not advocating for them, and when physicians feel that patients are “shooting the messenger” (ie, blaming the physician for managed care-initiated restrictions). In this atmosphere, physicians often avoid consideration of treatment alternatives they view as likely to be denied by a managed care reviewer, or feel compelled to focus on the catastrophic possibility in the differential diagnosis to obtain otherwise denied benefits.4 The final stage of this downward spiral of the physician-patient relationship can be a kind of mutual resignation, where the patient leaves with silent resentment while the physician acquiesces in the silence of a de facto gag clause.

The sense of virtual clinical captivity that patients and physicians can experience in the face of benefit denial need not be taken as a fait accompli. By becoming aware of the dynamics and dilemmas of the physician-patient relationship under managed care, treating physicians can create greater freedom for their patients and themselves to work together effectively. Useful clinical interventions can be implemented to recover a sense of choice in the doctor-patient relationship, while decreasing liability risk, by effectively anticipating, preparing for, identifying, and responding to the distress that is likely to accompany managed care restriction of clinical care.

There are compelling clinical and ethical considerations favoring a proactive stance by physicians who are aware of potential negative managed care influences on their patients. Given that helplessness and hopelessness have been documented to be predictors of negative patient outcomes (eg, in breast cancer),5 the undermining of the clinical process and the resulting loss of trust also undermine the potential for healing relationships and, in turn, the likelihood of positive health outcomes. Physicians who work to reverse this process are not just protecting themselves from liability, but also providing quality care by protecting patients from the consequences of helplessness, hopelessness, and distrust.

Although specialized consultation (eg, psychiatric) can be helpful in the face of a denial of healthcare benefits, it is important that primary care physicians have the tools and perspectives needed both to advocate for the patient and to support the patient when individually indicated benefits are deemed not “medically necessary.”6 An understanding of the dynamics of the physician-patient relationship in the managed care context can be helpful in adapting one’s clinical skills to the task of supporting the patient’s capacity for choice, hope, and trust.


Patient Dynamics: “Virtual Captivity”

Even as the prevalence of mental health problems such as mood and anxiety symptoms in primary care settings has come to be recognized,7 the cost containment strategies of managed health care have been found to inhibit the recognition and treatment of psychiatric disorders in medical treatment settings.8 Not only has a diagnosis of depression been found to be associated with patient dissatisfaction with medical care,9 but patients with high levels of depressive symptoms are less likely than the general population to act on their dissatisfaction by switching health plans.10 At the same time, the inability to choose one’s personal physician has been found to be a major determinant of patient dissatisfaction with their healthcare plan.11 The existing pool of trapped, frustrated, disillusioned patients also represents a potential tinderbox of litigation in the wake of tragic outcomes associated with managed care-influenced restrictions on effective treatment—in particular, restrictions on the building of therapeutic alliances in the doctor-patient relationship.

Serious illness that threatens an individual’s bodily or psychological integrity already poses a threat to a person’s sense of autonomy and control. Serious illness can heighten dependency while bringing with it feelings of helplessness, hopelessness, and distrust associated with depression. When faced first with a lack of choice of a provider, and then with restriction or even denial of care, the vulnerable patient, already feeling like a prisoner of the threat of serious illness, may now also begin to feel like a lonely captive of the healthcare system. This trapped, “nowhere to turn” feeling may persist after the illness is treated or even cured, as the patient, once burned by the denial of anticipated care or by worry over the prospect of such a denial, may come to think, “What if I get sick again? What care will I receive?” At the extreme, the patient may begin to experience, subtly but significantly, the emotional traumas associated with abandonment12 and captivity.13

Clinically, such experiences can generate cognitive distortions driven by anxiety—and depression—that may impair the patient’s capacity to temper realism with the hope necessary to tolerate uncertainty and to choose wisely from available alternatives.4,14 Moreover, the prospects for a supportive patient-physician alliance are undermined when the physician is seen by the patient as either untrustworthy or largely powerless to implement his or her own clinical recommendations and protect patient choices due to managed care pressure. The patient’s perception of physician trustworthiness may suffer, for example, when managed care drug formularies restrict the physician from prescribing a medication of choice. Even a trusting relationship, especially when it is founded on blind optimism,  can be destroyed by adversity together with perceived abandonment. As the patient’s distrust increases, responses can include such modes as “fight” (litigate), “flight” (drop out of treatment), or “freeze” (become numb, passive, demoralized, and unable to act effectively in the interest of one’s own health care). The resulting anxiety and depression can lead to a greater likelihood of dissatisfaction with medical care9 and to an impaired capacity to act on that dissatisfaction by changing health plans.10


Physician Dynamics: “False Necessity”

The human tendency toward resignation or self-deception and denial of what is too painful to see is not limited to patients. Often, the treating physician may not have freely chosen to be part of a particular managed care organization (MCO) nor to treat a particular patient, except as the best of a set of undesirable choices or the lesser of necessary evils. The physician is also likely to have experienced a substantial reduction of economic and professional autonomy in the shift to managed care. Like the patient, the physician may have few options and insufficient time to recognize, reflect upon, process, and put into perspective the feelings engendered by the managed healthcare context. Under any circumstances physicians are likely to react to a patient’s rejection by withdrawing emotionally from the patient. This reaction is especially likely, however, if the physician faces one frustrated, recalcitrant patient after another in a time-pressured managed care setting. At the same time, from the patient’s perspective, emotional withdrawal by a hurried and frustrated physician can easily be experienced as indifference to the patient’s suffering and a perceived abandonment.

Physicians are far from immune to the contagion of pessimism that can sweep through an institutional atmosphere, as in end-of-life care. Physicians who risk being penalized for caring for patients when, in the judgment of an anonymous third-party reviewer, there is no medical necessity to do so, are more likely to succumb to institutional pressures. Careful “not to raise the patient’s expectations” by effectively disclosing all possible options and advocating for the patient’s right to the best available care, they “hang crepe”15 as a prelude to premature discharge. In the extreme, some clinicians will automatically advise the patient and their family that only low-cost palliative measures be taken, not mentioning the more costly, intensive alternatives that may hold out a slim but real hope for the patient.

Such an atmosphere increases the risk that patients and families will give up prematurely while clinicians who are too distracted or time-squeezed to do the hard work of eliciting the patient’s and family’s deeper intentions go through the motions of obtaining informed consent.16,17 At the same time, some physicians react to their loss of autonomy and choice by making a premature cognitive commitment to diagnostic and treatment decision strategies designed to avoid punitive profiling practices. For example, a physician who is concerned about being identified as readily willing to hospitalize a patient for observation and evaluation will tend to avoid the risk of being deselected by the MCO. This avoidance can manifest itself in the doctor-patient encounter as a fixed, overly rigid stance or a reluctance to present alternatives to the patient other than the treatment least likely to engender MCO scrutiny. Such an attitude interferes with the informed-consent process vital to clinical care, and is often a pivot point in liability.

Of course, some denials of benefits by MCOs do represent a genuine effort to weed out unnecessary treatments and excessive costs. Nonetheless, other treatments that are medically appropriate (in that they are effective relative to individual patient values) come to be discounted as medically unnecessary by reference to an interpretation of “medical necessity” that is insensitive to individual patient values18 as well as to broader health-related social values.19

Medically appropriate care is care that is indicated based on the doctor-patient dialogue as informed by scientific research and accepted practice. Ideally, medically appropriate care considers the whole patient.20 By contrast, the term “medically necessary,” as used by third-party reviewers, is a misapplication of triage principles from military and disaster medicine to individual patients.21 Even treating physicians, when compelled to be time-pressured and hyper cost-conscious (as in many fixed-fee per subscriber capitated systems), tend to narrow their focus to overly concrete, readily measurable “necessary” benefits. Easily overlooked in such calculations are the costs of incomplete treatment22 to a patient’s freedom to live in the least restricted manner, as well as the biopsychosocial benefits of treatment that considers the patient’s overall well-being and level of independent and interdependent functioning.

Excessive reliance on managed care-driven “practice guidelines,” in the guise of being evidence-based, is also used to justify a denial of care. Such guidelines are often very selective as to the evidence they cite, as in the paucity of studies with outcome measures that reflect quality-of-life issues or the widespread neglect of many well-grounded outcome studies showing the efficacy of mental health treatment for patients with many medical and surgical conditions. In any case, decision making for patients in the aggregate is no substitute for individualized clinical decision making.


Prevention and Management of “Managed Care Side Effects”

Whether or not in conjunction with specialized consultation, a variety of clinical tools are available for prevention and management of the increasing iatrogenic harms experienced as side effects of managed care. To begin with, it is helpful for the treating clinician to keep in mind the complex interactions between medical and psychiatric disorders that are often obscured by various managed care influences, such as the lack of time to take a careful history that is objective and empathic. Although psychiatric consultation or referral is helpful in cases that present special difficulties, it is now less accessible than ever given the restrictions of managed health care.7

Thus, diagnostically, it is now more important than ever for treating physicians to be aware of psychiatric comorbidity within both acute and chronic physical illness. Patient suffering accompanying physical illness can present as depression, “sick role” adaptation, chronic pain, exacerbation of substance abuse or dependence, obsessive or dissociative reactions, and conversion reactions. In a person who has had a life-threatening illness, such suffering can sometimes rise to the level of disorders in the posttraumatic stress disorder (PTSD) spectrum.23

The flip side of the tendency to overlook psychiatric disorders exacerbated by managed care is the tendency to use psychiatric disorders as convenient labels to rationalize the denial of medical care and to neglect the existential dimensions of the patient’s suffering. Thus, it is important not to write off patients in panic as simply hypochondriacal because of time pressure associated with managed care. If a person presents with symptoms similar to those that marked a previous life-threatening illness, the physician first rules out a recurrence of that illness. When there has been a recurrence, any posttraumatic sequelae need to be attended to, even in the face of managed care constraints. These sequelae may include symptoms of depression, demoralization, dissociation, flashbacks (eg, “Oh my God, it is happening again!”), and an increased risk of panic and suicide. It is helpful for physicians to keep in mind that a patient who suffers from PTSD (or other emotional vulnerabilities) is especially susceptible to an exacerbation or recurrence of symptoms.

Even if a recurrence of the illness has been ruled out, the illness may have left a vulnerability in that the emotional memory of its painful and frightening initial presentation may be reactivated simply by the recurrence of general symptoms. There is the possibility of life-threatening complications for the patient, such as depression, panic disorder, and suicide or “suicide equivalents.” The latter may include self-medication of panic by excessive drinking or medication overdoses, or counterphobic risk-taking such as driving under the influence of alcohol. With the threat of illness recurrence, the feeling of being alone and the lack of choice in the managed care situation can increase the likelihood of such self-destructive reactions. The patient’s anxiety may also be amplified by acute somatic symptoms associated with vulnerable body image areas.24

Even when practicing in a time- and resource-restricted environment, with access to psychiatric consultation severely limited for patient and physician alike, treating physicians can be alert to make emotional contact with patients. Physicians can create provisional alliances even when they do not have the usual time and ongoing involvement required for sustained alliance building or for providing psychotherapy per se.25 In creating such alliances, it is important to convey certain understandings and attitudes, as delineated here, in whatever wording comes naturally to the individual physician.


Remedies for Each Phase of an Evolving Relationship

In the introductory phase of the relationship, the physician can make contact with the patient while observing the patient for indications of a potential (or hidden) but deep sense of helplessness and hopelessness which accompanies the experience of captivity.13 At the same time, it is helpful to initiate a meaningful dialogue as a first step toward creating a therapeutic alliance and detoxifying feelings of aloneness and abandonment. This step can include engaging in an informed-consent process (not merely a pro forma litany of risks and benefits).4,14 Such a process needs to address clinical and economic risks and potential ethical and role conflicts that might be engendered when a clinician is a dual agent (eg, both a “gatekeeper” and the primary care treating clinician in a capitated system).26 By the same token, the clinician can tactfully but effectively disclose all substantial treatment alternatives, including those not covered by the patient’s health plan.

These disclosures can enable the patient and physician to decide together how to respond to economic restrictions on treatment without the patient being overwhelmed by anxiety and pessimism. Likewise, if it is reasonably foreseeable at the outset that continuity of care will be interrupted by changes in the patient’s insurance coverage, then reminding the patient of how helpful it can be to keep abreast of possible insurance changes will be part of the economic informed-consent process. In our experience, patients can sometimes influence employers’ choice of insurance providers and managed healthcare packages. Subsequently, if a change in coverage is threatened, the physician can support the patient by actively inquiring about how any prospective benefit changes might affect continuity of care.

Given the distrust of physicians engendered by managed care policies,27,28 it is important to sensitively inform the patient of economic considerations such as provider profiling, managed care guidelines, and capitation contracts that may affect the quality of care that the patient receives. Openly acknowledging such dilemmas can enhance the possibilities for a therapeutic alliance. To provide for continuity in sharing uncertainty, it is helpful to articulate questions left open to be addressed in future visits, and to anticipate which questions may arise before the next visit. Such open communication is vital despite the fact that the “gag clauses”29 eliminated as explicit provisions in physician contracts with MCOs may still be implicitly enforced through healthcare provider profiling, economic deselection, and other often hidden rules and procedures.

In the ongoing care phase, the physician can implement treatment with as much continuity and mutual planning as possible while continuing to respond to managed care treatment restrictions in light of the patient’s evolving attitudes and preferences and changing clinical status. Treatment can proceed in a manner that respects the patient’s best interests, including autonomy interests, without being overwhelmed by considerations such as how this will affect the physician’s profile. The physician who needs or wants to apply practice guidelines, as noted above, is also faced with translating what any diagnostic or therapeutic option actually means for this particular patient given the patient’s life history and individual values.

In the event of a denial of benefits, every effort must be made to continue the relationship and avoid abandonment. While the denial of some benefits can reduce the quality of other benefits and the clinical care that the patient receives, it need not result in a catastrophic end to the doctor-patient relationship. For example, even when indicated hospitalization is denied, the physician can work with the patient on an appeal and remain available to help the patient consider the life choices that chronic illness periodically poses.

By working throughout the benefit denial and appeal processes to maintain as much patient confidentiality as possible under the circumstances, the physician can avoid feeling pressured into establishing “secret” manipulative agreements with the patient. An example of such an agreement is selecting, for billing purposes, differential diagnoses that increase the likelihood of receiving managed care benefits. Physicians who feel that their only recourse is to “spin”30 or lie for their patients31 are often expressing the underlying helplessness and hopelessness that they themselves feel. Such secrecy and misalliance based on deceiving the MCO can all too readily undermine the trust necessary for the doctor-patient relationship to be open and healing. The patient may ask, “If my doctor is willing to lie for me, might he/she also be willing to lie to me?” Moreover, a relationship built around secrecy and deception rather than confidentiality and open commitment is vulnerable to a sudden, panicked withdrawal by a guilt-ridden physician, culminating in abandonment.

The termination phase is critically important in any clinical relationship in which an emotional bond has been formed. Prior to the patient’s transferring to another care provider, it is helpful to consider what choices the physician and patient have made together, what other choices they might have made, and how managed care pressures may have influenced those decisions. The physician whose contract is terminated by an MCO can inform the patient of the MCO’s action and coordinate the transfer of care while supporting a patient’s choice to take steps to oppose involuntary termination. If involuntary termination is brought about because the patient’s employer has changed its health plan, the physician can attend actively to both the practical and emotional aspects of termination. The physician need not allow his or her own feelings of anger and frustration toward the MCO to become displaced onto the patient, leading to abandonment via failure to inform patients of the foreseeable consequences of “involuntary” abrupt termination. When the process of saying goodbye is properly attended to, even an involuntary termination can be borne without sliding into an abyss of abandonment and paralysis.



We have offered guidelines for prevention and management of observable clinical harms resulting from managed care’s control of choices made by patients and physicians, including anticipated and actual benefit denial and restrictions on care. Medical outcomes are affected not only by the quality of technical care given, but also by the process of care, including patient participation in decision making.32 Denial of choice reduces quality of care in that the patient loses both the psychological benefits of exercising choice and the medical benefits of individualized treatment. Moreover, irrespective of the structure of healthcare delivery, the physician retains a primary duty to advocate for the patient’s interests, including the right to make informed choices based on effective disclosure of treatment options.3 A considered clinical response to managed care constraints can help physicians fulfill this ethical duty to provide effective and compassionate care.

As risk management, attention to the clinical process can also help prevent malpractice liability in the event of a tragic outcome. Both bad medical outcomes, reasonably attributable to MCO-initiated distractions from a clinical focus on the patient’s best interests, and bad feelings arising from managed care restrictions on patient autonomy tend to feed malpractice risk.1 Moreover, attention to the clinical process allows physicians and patients to initiate, proceed with, and terminate relationships appropriately even when each phase of the relationship is subject to substantial external control.

Although clinical intervention is no substitute for instituting fundamental changes in healthcare financing and regulation of third-party control, even today the ethically sensitive primary care physician or psychiatrist practicing in a marketplace dominated by managed care need not feel too overwhelmed to practice effectively on the individual doctor-patient level. By identifying and then preventing or alleviating the negative biopsychosocial side effects of the restriction of available patient choices, a substantial reduction in the clinical complications and liability risks of MCO control and denial of patient choice and care can be achieved.   PP



1.    Bursztajn HJ, Brodsky A. A new resource for managing malpractice risks in managed care. Arch Intern Med. 1996;156:2057-2063.
2.    Bodenheimer T, Lo B, Casalino L. Primary care physicians should be coordinators, not gatekeepers. JAMA. 1999;281:2045-2049.
3.    Council on Ethical and Judicial Affairs, American Medical Association. Ethical issues in managed care. JAMA. 1995;273:330-335.
4.    Bursztajn HJ, Feinbloom RI, Hamm RM, Brodsky A. Medical Choices, Medical Chances: How Patients, Families, and Physicians Can Cope With Uncertainty. New York, NY: Routledge, Chapman, Hall; 1990.
5.    Watson M, Haviland JS, Greer S, Davidson J, Bliss JM. Influence of psychological response on survival in breast cancer: a population-based cohort study. Lancet. 1999;354:1331-1336.
6.    Levinson W, Gorawara-Bhat R, Dueck R, et al. Resolving disagreements in the patient-physician relationship: tools for improving communication in managed care. JAMA. 1999;282:1477-1483.
7.    Nease DE Jr., Volk RJ, Cass AR. Investigation of a severity-based classification of mood and anxiety symptoms in primary care patients. J Am Board Fam Pract. 1999;12(1):21-31.
8.    Horn SD. Overcoming obstacles to effective treatment: use of clinical practice improvement methodology. J Clin Psychiatry. 1997;58(suppl 1):15-19.
9.    Wyshak G, Barsky A. Satisfaction with and effectiveness of medical care in relation to anxiety and depression: patient and physician ratings compared. Gen Hosp Psychiatry. 1995;17:108-114.
10.    Druss B, Schlesinger M, Thomas T, Allen H. Depressive symptoms and plan switching under managed care. Am J Psychiatry. 1999;156:697-701.
11.    Schmittdiel J, Selby JV, Grumbach K, Quesenberry CP Jr. Choice of a personal physician and patient satisfaction in a health maintenance organization. JAMA. 1997;278:1596-1599.
12.    Pollock GH. Abandonment. In: Rothstein A, ed. The Reconstruction of Trauma: Its Significance in Clinical Work. Madison, Conn: International Universities Press; 1986:105-120.
13.    Bursztajn HJ, Brodsky A. Captive patients, captive doctors: clinical dilemmas and interventions in caring for patients in managed health care. Gen Hosp Psychiatry. 1999;21:239-248.
14.    Gutheil TG, Bursztajn HJ, Brodsky A. Malpractice prevention through the sharing of uncertainty: informed consent and the therapeutic alliance. N Engl J Med. 1984;311:49-51.
15.    Siegler M. Pascal’s wager and the hanging of crepe. N Engl J Med. 1975;293:853-857.
16.    Curtis JR, Rubenfeld GD. Aggressive medical care at the end of life: does capitated reimbursement encourage the right care for the wrong reason? JAMA. 1997;278:1025-1026.
17.    Bursztajn HJ, Brodsky A. Authenticity and autonomy in the managed-care era: forensic psychiatric perspectives. J Clin Ethics. 1994;5:237-242.
18.    Rosenbaum S, Frankford DM, Moore B, Borzi P. Who should determine when health care is medically necessary? N Engl J Med. 1999;340:229-232.
19.    Emanuel EJ, Emanuel LL. Four models of
the physician-patient relationship. JAMA. 1992;267:2221-2226.
20.    Tucker JB. Modification of attitudes to influence survival from breast cancer. Lancet. 1999;354:1320.
21.    Bursztajn HJ, Gutheil TG, Brodsky A. Ethics and the triage model in managed care hospital psychiatry. Psychiatr Times. 1998;15(9):33-40.
22.    Simon G, Ormel J, VonKorff M, Barlow W. Health care costs associated with depressive and anxiety disorders in primary care. Am J Psychiatry. 1995;152:352-357.
23.    Green BL, Epstein SA, Krupnick JL, Rowland JH. Trauma and medical illness: assessing trauma-related disorders in medical settings. In: Wilson JP, Keane TM, eds. Assessing Psychological Trauma and PTSD. New York, NY: Guilford Press; 1997:160-191.
24.    Meissner WW. The self and the body. I. The body self and the body image. Psychoanalysis and Contemporary Thought. 1997;20:419-448.
25.    Meissner WW. The Therapeutic Alliance. New Haven, CT: Yale University Press; 1996.
26.    Miller TE, Sage WM. Disclosing physician financial incentives. JAMA. 1999;281:1424-1430.
27.    Hillman AL. Mediators of patient trust. JAMA. 1998;280:1703-1704.
28.    Grumbach K, Selby JV, Damberg C, et al. Resolving the gatekeeper conundrum: what patients value in primary care and referrals to specialists. JAMA. 1999;282:261-266.
29.    Brody H, Bonham VL Jr. Gag rules and trade secrets in managed care contracts. Arch Intern Med. 1997;157:2037-2043.
30.    Illingworth PML. Bluffing, puffing and spinning in managed care. J Med Philosophy. 2000;25(1):62-76.
31.    Freeman VG, Rathore SS, Weinfurt KP, Schulman KA, Sulmasy DP. Lying for patients: physician deception of third-party payers. Arch Intern Med. 1999;159:2263-2270.
32.    Safran DG, Taira DA, Rogers WH, Kosinski M, Ware JE, Tarlov AR. Linking primary care performance to outcomes of care. J Fam Pract. 1998;47:213-220.



Ms. Schulz is psychiatry research project assistant, Dr. Gotto is director of consult/liaison services, and Dr. Rapaport is chairman in the Department of Psychiatry at Cedars-Sinai Medical Center in Los Angeles, CA. Dr. Rapaport is also vice chairman of the Department of Psychiatry and Behavioral Health at the David Geffen School of Medicine at the University of California—Los Angeles.

Disclosure: Ms. Schulz and Dr. Gotto report no affiliation with or financial interest in any organization that may pose a conflict of interest. Dr. Rapaport is a consultant to Cyberonics, Eli Lilly, Forest, GlaxoSmithKline, Janssen, the National Institute of Mental Health (NIMH), the National Institute on Drug Abuse, Neurocrine Biosciences, Novartis, Pfizer, Roche, Sanofi-Synthelabo, Solvay, Sumitomo, and Wyeth; is on the speaker’s bureaus of Cyberonic, Eli Lilly, Forest, GlaxoSmithKline, Janssen, Novartis, and Pfizer; has received grant support from Abbott, AstraZeneca, Corcept, Cyberonics, Eli Lilly, Forest, GlaxoSmithKline, Janssen, the National Center for Complementary and Alternative Medicine, the NIMH, Novartis, Pfizer, Pharmacia Upjohn, Sanofi-Synthelabo, Solvay, the Stanley Foundation, UCB Pharma, and Wyeth; and owns stock in Forest.

Please direct all correspondence to: Julie Schulz, BA, Department of Psychiatry, Cedars-Sinai Medical Center, 8730 Alden Drive, Los Angeles, CA 90048. Tel: 323-449-1753; E-mail:



Generalized anxiety disorder (GAD) is characterized by excessive anxiety and worry occurring over the course of ≥6 months. GAD has a 5.7% lifetime prevalence in the general adult population, and an 8% lifetime prevalance in the primary care setting. GAD is associated with twice the amount of primary care visits compared with patients without GAD, which leads to significant healthcare costs. This is largely because GAD commonly manifests with musculoskeletal, gastrointestinal, and cardiovascular physical symptoms. The course of GAD is chronic and disabling, with a low rate of spontaneous remission. Two thirds of patients with GAD have a secondary comorbid psychiatric diagnosis, with major depressive disorder being the most frequently occurring. Aggressive treatment is necessary to maximize resolution of symptoms. Fortunately, a variety of effective pharmacotherapies and psychotherapies exist for the acute treatment of GAD. Preliminary work suggests that these benefits are maintained with time and that continued pharmacotherapy may prevent the reoccurrence of GAD.  



Anxiety disorders are the most common form of mental disorders in the United States.1 However, they are often missed because anxious patients typically approach primary care physicians (PCPs) with somatic complaints that obscure the underlying psychiatric diagnosis.2 Patients with generalized anxiety disorder (GAD) are more likely to complain of somatic symptoms than psychological symptoms. As a result, physicians often order costly diagnostic tests, unnecessary treatment trials, and referral for specialty evaluation.

GAD is characterized by excessive anxiety and worry occurring over the course of ≥6 months. Symptoms include restlessness, difficulty concentrating, irritability, muscle tension, sleep disturbance, and fatigue. Patients with GAD worry about physical illness, social and professional performance, interpersonal relations, personal safety, and the safety of people that are close to them. Somatic complaints of GAD patients typically fall into three categories: musculoskeletal, cardiovascular, or gastrointestinal.3 For example, a patient with GAD might complain of tachycardia, chest pain, diarrhea, and fatigue. The disorder is chronic, with periodic exacerbation, and causes significant distress and impairment.4 The spontaneous remission rate, or the rate at which patients recover from the disorder without treatment, is approximately 20% to 25%.5 This compares with a spontaneous remission rate of 32% for patients with major depressive disorder (MDD).6

Untreated GAD causes functional impairment similar to the disability seen in diabetes or congestive heart failure.7 In addition, people experiencing severe anxiety are at risk for suicidal ideation. According to the National Comorbidity Survey Replication (NCS-R), the adjusted odds ratio for lifetime suicidal ideation is .87 (.57-1.32, 95% CI), and the adjusted odds ratio for lifetime suicide attempts is 1.50 (.98-2.30, 95% CI).8 This article reviews the epidemiology, etiology, clinical presentation, diagnosis, clinical course, healthcare costs, and treatment of GAD, with an emphasis on its biological causes and treatment methods.



The Epidemiological Catchment Area study reported that the lifetime prevalence of GAD (according to the Diagnostic and Statistical Manual of Mental Disorders, Third Edition [DSM-III]9) ranges from 4.1% to 6.6%.10 According to the NCS-R, which used DSM-IV11 criteria, GAD has a lifetime prevalence of 5.7% in the US adult population.2  While the age of onset of GAD peaks during the second and third decades,12 first onset of the disorder can occur at any age. Ten percent of GAD patients experience onset of the disorder at <13 years of age, and 10% experience onset of GAD at >51 years of age.12 The 12-month prevalence rates of GAD range from 0.5% to 1.5%.13,14 Women are at twice the risk of developing GAD as men.2

Recognition and treatment of GAD are especially important in the primary care setting since GAD is common and is associated with significant use of medical resources. In one study performed in a primary care setting, GAD was diagnosed in 8.0% of patients.15 In a second study performed in a primary care setting, 3.7% of adult patients met criteria for current GAD and 6.6% presented with subthreshold anxiety symptoms.16 In a third study, 5.3% of the patients seen in primary care met criteria for current GAD.17 Thus, the point prevalence rate for GAD is 3.7% to 8.0% for patients screened in primary care settings.



Only one third of people with a current diagnosis of GAD have no other comorbid psychiatric diagnosis.2,18 In fact, Noyes19 suggests that the rates of comorbid psychiatric disorders might be greatest for patients with GAD. MDD (11% to 47%) is the most common comorbid disorder.19 The National Comorbidity Survey data suggest that preexisting anxiety disorder is a risk factor for depression, and anxiety precedes the onset of depression by about 10 years.2 Social anxiety disorder (17% to 59%), specific phobia (15% to 56%), and panic disorder (11% to 27%) are the most prevalent secondary anxiety disorders diagnosed in patients with GAD.20-25  

Personality disorders are thought to coexist in approximately 60% of patients with GAD.26,27 The most common personality disorders diagnosed in patients with GAD are obsessive-compulsive personality disorder (33.5%), paranoid personality disorder (30.5%), and avoidant personality disorder (21.9%).27

The presence of psychiatric comorbidity increases the severity of GAD and complicates the treatment course for these patients.28 However, the presence of psychiatric comorbidity also increases the likelihood that patients will seek psychiatric care.29


Evolution of the Generalized Anxiety Disorder Diagnosis

The DSM-IV definition of GAD has evolved over the last 125 years. Prior to the DSM-III, the concept of anxiety was amorphous, lumping Freud’s “neurosis” together with phobias and panic attacks.30 The DSM-III-defined generalized anxiety as “uncontrollable anxiety of worry that is not due to a particular life problem,” has a duration of 1 month, cannot be diagnosed in the presence of another mental disorder, and requires the presence of autonomic symptoms. Changes in the DSM-III-R31 definition of GAD required lengthening the duration of symptoms (1 month was changed to 6 months), the presence of six physical symptoms, and the caveat that the anxiety could not be due to another anxiety disorder. In the DSM-III-R, GAD can be diagnosed in the presence of another mental disorder. The DSM-IV criteria require that only three of six physical symptoms be present, and these symptoms must be equally psychologic and somatic (Table 1).


Clinical Presentation

Core features of GAD include excessive anxiety or worry occurring more days than not, over a period of 6 months. These worries may be associated with several events or activities, including job responsibilities, finances, health of friends and family members, safety of children, household chores, car repairs, or tardiness. Patients experience impairment in social, occupational, or other important areas of functioning related to the worries. Individuals with GAD find it difficult to control such worry. The frequency, intensity, and duration of anxiety are out of proportion to the likelihood or impact of a feared event.

A list of common somatic symptoms associated with GAD is presented in Table 2. Patients with anxiety often seek initial treatment from their PCP. However, GAD is frequently not recognized in the primary care setting. In a recent study, PCPs recognized and diagnosed pure GAD in only 34% of patients with the disorder.32

Individual and cultural characteristics may influence the presentation of GAD. Some patients complain primarily of cognitive symptoms such as confusion, perceptual distortion, difficulty concentrating, and indecisiveness. Other patients may express anxiety predominantly through somatic symptoms. While it is accepted that the initial clinical presentation of anxious symptoms may range from cognitive to somatic complaints, the full effect of cultural and personality factors on both the presentation and treatment course of anxiety are not well understood. A review of the subject of cultural effect on anxiety disorders showed that in general, opinions on this topic are mixed.12,33 Cultural attitudes toward mental health, language nuances, and socioeconomic factors all have the potential of affecting a person’s experience of anxiety. This experience may affect the way in which a particular patient presents his or her ailments to the physician. There is a paucity of research on the cultural differences in the presentation of GAD, and this topic merits further study.

Although it is sometimes challenging to differentiate normal concern from pathological anxiety, a key finding about anxiety disorders is that the intensity and distress associated with the worry  far exceed any stimulus. A person with GAD is a “worry wart” with widespread, intense, and persistent anxiety that occurs without an identifiable trigger. Individuals with GAD palpably feel their anxiety, are disabled by it, and have diminished quality of life.


Clinical Course

While the median age of onset of GAD is 31 years, 10% of GAD patients first experience the disorder at <13 years of age, and 10% experience it after 58 years of age.12 Early-onset patients are more likely to develop GAD without precipitating events, and are more likely to have been exposed to domestic disturbances in childhood. These patients report heightened anxiety during childhood, childhood fears and obsessional traits, sensitivity to interpersonal relationships, and social inhibition and maladjustment.33 In contrast, late-onset GAD is associated with a precipitating adverse life experience.33 The development of a full-blown anxiety disorder often occurs when an already anxious person is subjected to great life stress, such as losing a job or taking on a second mortgage.

The severity of symptoms of GAD and impairment may fluctuate considerably over an individual’s lifetime. The symptoms are often more severe when exogenous stressors are present. In women, symptoms may be exacerbated premenstrually. It is clear that for the majority of patients, extended periods of remission are unlikely without vigorous treatment.

GAD has a chronic course. Treatment should be continued on a long-term basis in order to maximize resolution of symptoms, remediate functional disability, and increase the intervals between episodes of illness.34 The longer term spontaneous remission rate is low for GAD, with a .38 probability of remission at 5 years.35 The 12-year follow-up results from the Nottingham Study found that 59% of the patients originally diagnosed with GAD (using DSM-III criteria) were symptomatic at follow-up.36 Of the patients originally diagnosed with GAD, the most common diagnoses after 12 years were dysthymic disorder (18%) and major depressive episode (10%). Only 3% had a diagnosis of GAD. Thus, people with GAD are at significant risk for developing depressive disorders.  The Harvard-Brown Anxiety Research Program found that 60% of subjects identified with an index episode of GAD had active GAD at 2-year follow-up. At 5-year follow-up, 66% of index cohort was classified as having either GAD or GAD in partial remission.37


Disability and Healthcare Utilization

Patients with GAD have twice the rate of primary care visits compared to primary care patients without GAD.32 Patients with comorbid GAD and other Axis I disorders have greater rates of hospitalization, greater utilization of diagnostic and laboratory tests, higher pharmacy costs, and more absenteeism from work than patients with GAD without any psychiatric comorbidity.38 In fact, comorbid GAD is commonly associated with a number of physical disorders, such as irritable bowel syndrome. The prevalence of comorbid GAD ranges from 13% to 25% in patients with irritable bowel syndrome.39,40

Biological Etiology

Relatively few studies have addressed the biological aspects of GAD. The development of a sound biological basis for GAD, distinct from other anxiety or mood disorders, has been complicated by three major factors. First, while there are animal models available for stress and anxiety, there is no true animal model specific for GAD. Second, the pharmacologic probes available for investigating the neurobiology of GAD are relatively insensitive to their targets. Finally, the lack of a consistent definition for GAD makes it difficult to compile studies using the same criteria. Many studies investigating the biology of GAD use DSM-III or DSM-III-R criteria for diagnosis.

A significant portion of the biological research investigating GAD is based on the success of pharmacologic treatments. Since these interventions tend to be effective across most of the anxiety disorders, it is challenging to identify neurotransmitter abnormalities that may be specific to GAD. In fact, many clinical studies now address behavioral or symptomatic phenomena that cut across several diagnoses, rather than focusing on a single disorder.


Molecular Biology

It is generally accepted that dynamic interactions among several different neurotransmitter systems are likely to underlie different anxiety states. Evidence from animal and human studies suggests that alteration of function of the γ-aminobutyric acid (GABA) neuronal system plays a significant role in the pathophysiology of GAD. Benzodiazepine receptors (BDZ-Rs) and GABAA receptors are part of the same macromolecular complex. Animal studies suggest that GABAA receptor dysfunction plays a role in the development of GAD.41 In addition, clinical studies have shown that benzodiazepines are effective in treating patients with GAD.42 Patients with GAD have been shown to have low levels of peripheral benzodiazepine receptors (pBR) and low levels of pBR messenger ribonucleic acid in lymphocytes.43 These levels increased to normal after 2 months of treatment with a diazepam-related compound, and remained normal up to 1 month after discontinuing treatment. In this study, the level of pBR was also correlated with the amount of clinical improvement. These results suggest that the lower expression of pBR is a state-related biological abnormality in patients with GAD that can be remediated with appropriate treatment. However, the mechanism by which the levels of pBRs are normalized is still unknown and one must always be careful when extrapolating from studies of the pBRs to the function of the central BDZ-Rs.

Medications that affect serotonin and norepinephrine, neurotransmitters  thought to be central to the pathophysiology of GAD, prove efficacious. However, research investigating these neurotransmitters is inconsistent. One study found elevated plasma levels of norepinephrine and decreased α2-adrenergic function in GAD patients.44 The authors of that study concluded from these findings that norepinephrine activity is elevated in GAD, causing receptor downregulation. Other studies were unable to replicate these findings.45,46 The one investigation exploring the metabolic pathways of catecholamines failed to find a difference between GAD subjects and normal controls,47 and studies have failed to consistently demonstrate that plasma catecholamine levels of GAD subjects are higher than normal comparison subjects.45,46 Furthermore, research investigating catecholamine receptors in GAD patients are mixed and inconclusive.48-50 In summary, there is little evidence suggesting abnormalities in the catecholamine system for patients with GAD.

Studies investigating the role of serotonin in GAD are also inconclusive. Both overactivity or underactivity of the serotonergic system has been hypothesized to be involved in the pathogenesis of GAD.51,52 Some of the strongest biological evidence linking the serotonin system and GAD have been studies with the serotonin receptor agonists buspirone, ipsapirone, and gepirone. These agents decrease the firing rate of serotonergic neurons and exert anxiolytic effects in patients with GAD.53

The hypothalamic-pituitary-adrenal  axis has been implicated in other anxiety disorders and may play a role in GAD. Cortisol, the primary hormone implicated in stress, has been shown to be elevated in GAD patients.54 In normal subjects, exogenous cortisol induces an increase in the serotonin transporter gene transcription.55 However, the incubation of peripheral lymphocytes from GAD patients with cortisol failed to increase serotonin transporter gene product.55 Corticotropin-releasing factor (CRH) is associated with heightened states of anxiety in animal models. However, there is no conclusive association between CRH dysfunction and GAD.56-58



Genetic studies suggest a moderate degree of heritability for GAD. The prevalence rate for GAD is 20%59 to 22%60 for individuals who have a first-degree relative with GAD, as compared to a 4% prevalence rate in the control sample.59 Twin studies demonstrate that 21% of monozygotic twins are concordant for GAD, versus 13% of dizygotic twins.61 Candidate genes for further study include those influencing norepinephrine and serotonin neurotransmission as well as GABA and CRH. Of particular interest is the serotonin transporter gene located on chromosome 17q. One study reported that polymorphism in the second intron of this gene occurs at significantly higher rates in individuals with GAD.62


Treatment Options in Generalized Anxiety Disorder

There are evidence-based pharmacologic and psychologic treatment options for GAD (Table 3). Ideally, the physician and patient should work together to define the problem and the treatment plan. It helps to get the patient’s dedication to the treatment plan, and for the patient to understand that improvement may take months or even years. The patient should be aware that the course of GAD waxes and wanes, and that complete remission is unlikely. Treatment should be individualized for each patient, with the patient’s symptom profile and ability to tolerate side effects as the primary considerations. Other treatment factors may include history of previous treatment and response, presence of comorbid depression or other anxiety disorders, presence of Axis II disorder, presence of comorbid medical condition, and history of substance abuse. There are no clear evidence-based guidelines for determining the length of time a patient should receive a particular treatment. A review of eight long-term GAD treatment studies was inconclusive, and no drug was identified as a reference for long-term treatment.63 It is generally recommended that treatment continue for 1 year after the patient responds to it before considering cessation.64


Pharmacologic Therapies


Benzodiazepines are a well-studied class of medications with proven efficacy in the treatment of GAD.65 Approximately 75% of patients have either a marked (35%) or moderate (40%) response to treatment.66 Anxious symptoms respond to benzodiazepines rapidly, with the most dramatic improvement occurring in hypervigilance and somatic symptoms, such as muscle tension, restlessness, and insomnia. Psychological symptoms, such as worrying and fretting, are not as responsive to benzodiazepine medication, and may require additional treatment with cognitive-behavioral therapy (CBT) or another specific modality.

Mechanism of Action

Benzodiazepines exert their effect at BDZ-Rs, which are located throughout the brain in the cerebellum, limbic system, and cerebral cortex. The BDZ-R is structurally linked with the GABA receptor. When GABA is present at the GABA receptor, chloride ions pass through a channel into the neuron, which hyperpolarizes the cell membrane. This results in decreased firing, which “quiets” the cell. A benzodiazepine, when present at the BDZ-R, potentates the GABA-induced opening of the chloride channels. GABA must be present at the GABA receptor in order for benzodiazepines to work. This explains why an overdose on a benzodiazepine alone is usually not fatal; the body has a finite supply of GABA. However, if a benzodiazepine overdose is combined with a substance that opens chloride channels without needing to bind to a receptor (such as alcohol or phenobarbital), then a benzodiazepine overdose can result in central nervous system depression and death.

The main differences between benzodiazepines are their half-life and potency. Half-life is important when considering dosing; a medication with a short half-life, such as alprazolam, may require multiple daily dosing, whereas clonazepam, with a half-life of 34 hours, may be given QD or BID. Benzodiazepines with shorter half-lives are more likely to be associated with “clock watching.” This occurs when an anxious patient notices the anxiolytic effect of medication is wearing off, and watches the clock, counting the minutes until the next dose. Shorter-acting benzodiazepines are also associated with interdose rebound anxiety, greater potential for abuse, and greater risk of withdrawal syndromes.    

Absorption and the onset of action are most rapid for diazepam, lorazepam, alprazolam, triazolam, and estazolam.  The rapid-onset medications are useful in patients who use benzodiazepines on an as-needed basis for a burst of anxiety or for falling asleep.

The 2-keto benzodiazepines (Table 4) have half lives of 30–120 hours. This occurs because they are all metabolized to desmethyldiazepam, which is pharmacologically active and takes a long time to break down to oxazepam and then to a form that can undergo glucuronidation.67

The half-lives of the 3-hydroxy benzodiazepines are much shorter, at 10–30 hours. This is because they are directly metabolized by glucuronidation and have no active metabolites.  Alprazolam and triazolam are triazolobenzodiazepines that are hydroxylated prior to undergoing glucuronidation.  

Benzodiazepines are generally well tolerated. Side effects such as somnolence, fatigue, reduced concentration, and slowed psychomotor function may occur early in treatment, and can be handled either by dose reduction or changing the time the medication is taken. Patients usually develop tolerance to the sedating side effect of benzodiazepines after 7–10 days of treatment. Although patients become tolerant to the side effects of benzodiazepines, they do not develop tolerance to the anxiolytic effect of the medication.68 It is unclear how long therapy with a benzodiazepine should continue. Recent recommendations suggest a treatment trial of 6 weeks using the lowest possible effective dose, and then making an attempt to taper the medication dose. The reason for the 6-week time frame is that approximately 50% of patients do not relapse after 6 weeks of treatment with benzodiazepines.69 Given this, the clinician should instigate an initial taper after 6 weeks to see if the patient is one of the 50% of patients who tolerate early benzodiazepine discontinuation. If the patient does not tolerate the taper and experiences a relapse, then treatment resumes for a longer period of time. A second attempt at benzodiazepine discontinuation with a slower taper may be tried. If the patient still experiences relapse, that patient may require longer-term therapy, or the addition of antidepressant medication or CBT.

Even though the majority of patients do not escalate their dose or abuse their medication,70 the long-term use of benzodiazepines for anxiety remains controversial. There are several reasons for this. Physicians are concerned about the long-term side effects and the abuse potential of benzodiazepines. In reality, patients with anxiety typically are wary of becoming dependent on medication. The tolerance that occurs with chronic benzodiazepine use may be confused with the idea of drug dependence, and this may be unsettling for GAD patients. If a patient treated with twice daily alprazolam is admitted to the hospital for a medical illness, this patient is at great risk for rebound anxiety and even benzodiazepine withdrawal unless the admitting physician is aware of the patient’s anxiety disorder and specific treatment. Even then, a well-intentioned physician might schedule the alprazolam as needed when the patient has been using it regularly, thus setting the anxious patient up for potential demise. The withdrawal that occurs with the shorter-acting benzodiazepines can occur remarkably fast, as the patient’s mental state may move from anxious to agitated to psychotic to delirious within 48–72 hours of the last missed dose.

The benzodiazepines that undergo rapid metabolism and have no active metabolites (lorazepam, oxazepam) are safer choices when using benzodiazepines in the elderly or in patients with liver or respiratory disease. The elderly are at risk for confusion, falls, and becoming paradoxically agitated when exposed to benzodiazepines, so these medications should be used judiciously in these patients. Benzodiazepines should be used with caution in any medically ill patient, though they are quite useful in treating anxiety associated with procedures or anxiety due to certain medications.  



Tricyclic Antidepressants and Monoamine Oxidase Inihibitors

Over 2 decades of data demonstrate that antidepressants are effective in the treatment of GAD. Since anxiety and depression commonly coexist, this makes antidepressant pharmacology an efficient option for treating both problems. Most of the studies using tricyclic antidepressants (TCAs) in anxiety are older, and the criteria used for GAD are less strict than in the more recent DSM-IV-based studies, so it is possible that patients included in the earlier studies may have had disorders other than GAD. In any case, the TCAs are at least as effective as benzodiazepines in the short-term management of anxiety,71 and are more effective in reducing the psychic symptoms of anxiety.72

TCAs work by inhibiting the reuptake of norepinephrine (NE) and serotonin (5-HT), thereby increasing the available amount of both neurotransmitters in the synaptic cleft. The TCAs differ in their abilities to inhibit NE and 5-HT.  For example, clomipramine is considered to be a “serotonergic” tricyclic. The tricyclics, like the selective serotonin reuptake inhibitors (SSRIs), are useful in treating the patient with mixed anxiety and mood disorder, where anxiety and mood pathology coexist.

The clinical utility of the TCAs is limited by their side-effect and safety profile. The typical TCA side effects, which include dry mouth, constipation, blurred vision, passing out, or gaining weight, may be more than anxious patients want to tolerate. This class of agents has anticholinergic, antiadrenergic, and antihistaminic side effects.  The monoamine oxidase inhibitors (MAOIs), are used in the treatment of atypical depression, panic disorder, and as an alternative therapy in treatment-resistant depression. MAOIs are known to reduce anxiety when used in the treatment of depression. They work by irreversibly inhibiting monoamine oxidase, the enzyme that breaks down NE, dopamine, and 5-HT. MAOIs are infrequently prescribed. In a 1999 survey of prescribing practices, Balon and colleagues73 found that although 92% of respondents believed MAOIs are useful for atypical depression, only 2% said they would use them as a first-line treatment. MAOIs have fallen out of favor due to side effects and dietary restrictions. Moclobemide, a reversible MAOI, is an effective treatment for depression and does not carry the risk for tyramine reaction. However, it is not available in the US.


Selective Serotonin Reuptake Inhibitors

There is great deal of placebo-controlled evidence supporting the efficacy of SSRIs as a treatment for GAD. Clomipramine was the first TCA “serotonergic” agent studied for the treatment of GAD; the study was an 8-week open trial where 10 subjects were treated with doses ranging from 50–250 mg/day.74 Five subjects dropped out because of the clomipramine side effects. The remaining five subjects had a good response to the medication.  

The first controlled study of an SSRI compared the outcomes of 81 patients with DSM-IV-based GAD treated with paroxetine 20 mg/day, imipramine 75 mg/day, or 2-chlordesmethyldiazepam 4.2 mg/day.75 Two thirds of patients in all three treatment groups improved moderately or markedly (imipramine, 67%; paroxetine, 68%; 2-chlordesmethyldiazepam, 60%). In a multi-center, double-blind, placebo-controlled trial, 566 patients with DSM-IV-diagnosed GAD were treated with paroxetine 20 mg/day, paroxetine 40/day, or placebo. Paroxetine 40 mg alleviated symptoms better than paroxetine 20 mg, and both were superior to placebo.76 In another placebo-controlled study of 324 outpatients with DSM-IV-diagnosed GAD, a flexible dose of paroxetine (20–50 mg/day) was compared to placebo. The patients treated with paroxetine experienced improvement of core symptoms of GAD as early as 4 weeks, and a reduction in disability compared to patients treated with placebo.77

One double-blind, placebo-controlled study of citalopram 20–30 mg/day in 34 older (≥60 years of age) outpatients with mostly DSM-IV-diagnosed GAD found that 65% of patients responded by 8 weeks of treatment with citalopram versus 24% of patients treated with placebo. The most common side effect experienced by the treatment group was sedation.78 A study of escitalopram, the s-enantiomer of citalopram, compared 8 weeks of treatment with escitalopram 10–20 mg/day to placebo in 315 patients with GAD. Sixty-eight percent of escitalopram-treated subjects responded, compared to 41% of patients treated with placebo.79 There was one double-blind, prospective trial comparing 8 weeks of paroxetine to sertraline in the treatment of 55 patients with GAD.80 This comparison found that both SSRIs equally reduced Hamilton Rating Scale for Anxiety (HAM-A) scores (paroxetine 57%±28%, and sertraline 56%±28%) and were similarly well-tolerated.


Mechanism of Action   

SSRIs work by inhibiting the reuptake of serotonin so that more serotonin is available in the synaptic cleft. It is unclear exactly how this increase affects changes in anxiety and mood. The SSRI antidepressants are better tolerated than the TCAs, and are safer in overdose situations. Anorgasmia is a significant side effect of SSRI therapy, and occurs in both genders. There are several possibilities for dealing with this problem. Dose reduction may help in some cases. Adding a small dose of trazodone, bupropion, or buspirone may be helpful in others. Some patients respond to sildenafil therapy, as needed. If none of these interventions work, the patient and physician should weigh the benefit of treating the anxiety with the current SSRI versus adequate sexual functioning. Sometimes switching to a different SSRI medication may improve symptoms. CBT is an option for patients who decide to taper off SSRI medication.  

Other side effects of SSRI therapy include initiation activation, restlessness, nausea, and diarrhea. Starting patients on a very low dose of SSRI and increasing the dose slowly averts initiation activation. Alternatively, starting a benzodiazepine simultaneously with the SSRI may also prevent initiation anxiety. The benzodiazepine can be tapered and discontinued once the SSRI dose is within a therapeutic range.


Serotonergic Noradrenergic Reuptake Inhibitors

Venlafaxine, a serotonergic noradrenergic reuptake inhibitor, is the best studied antidepressant medication used in the treatment of GAD. In a series of placebo-controlled, randomized, short-term treatment trials comparing venlafaxine extended release (XR) to placebo, venlafaxine was consistently superior to placebo in improving both somatic and psychic symptoms of anxiety. In one study that compared venlafaxine XR 75 mg and 150 mg, buspirone 30 mg, and placebo, both venlafaxine doses improved anxious mood and tension indices by week 8, and both venlafaxine doses were better than placebo and more effective than buspirone after week 1.81 Another 8-week, randomized, placebo-controlled study evaluated fixed doses of 75 mg, 150 mg, or 225 mg venlafaxine XR against placebo and found that venlafaxine was consistently superior to placebo on all of the primary outcome measures. The most robust improvements were seen with the higher doses.82 An 8-week, double-blind, placebo-controlled study of Greek outpatients with GAD compared flexible dosing of venlafaxine XR (75–150 mg) with placebo, and found that the 62.5% of venlafaxine-treated patients achieved remission versus 9.1% in the placebo group. Remission was defined as a HAM-A score ≤7.83

The long-term studies of venlafaxine validate the clinical lore that the improvements gained in the acute treatment phase can be maintained and even enhanced over a longer period of time. The first long-term study evaluated flexible dosing of venlafaxine XR over 6 months. In this study, 251 patients were assigned to receive either venlafaxine XR at doses between 75 mg/day and 225 mg/day to control symptoms, or placebo. Venlafaxine XR was superior to placebo with a response rate of 69% from week 6 through week 28, versus placebo response rates of 42% to 46%.84 Another 6-month, fixed-dose trial comparing venlafaxine XR 37.5 mg/day, 75 mg/day, or 150 mg/day with placebo found that in the short term (8 weeks), both the 75-mg and 150-mg dose groups showed significant differences from placebo on all of the primary outcome measures. The outcome measures included the HAM-A total score, the HAM-A psychic anxiety factor, the Hospital Anxiety and Depression (HAD) anxiety subscale and the Clinical Global Impressions (CGI) Improvement rating. At 8 weeks, the 37.5-mg dose was only significantly better than placebo on the HAD anxiety subscale.  After 24 weeks of treatment, the greater efficacy for the higher venlafaxine doses was preserved over both placebo and the 37.5-mg dose. The higher doses also demonstrated improved social functioning by the Social Adjustment Rating Scale compared to placebo.85

Like the TCAs, venlafaxine works by inhibiting the reuptake of serotonin and norepinephrine. It lacks the anticholinergic, antiadrenergic, and antihistaminic effects of the TCAs, and is consequently better tolerated. Somnolence, nausea, and dry mouth are the side effects most commonly reported. Anorgasmia may also occur as a side effect of venlafaxine treatment. Venlafaxine needs to be tapered when discontinued.     


Other Antidepressants


Little evidence supports the role of trazodone for the treatment of GAD. One 8-week, randomized, placebo-controlled trial comparing imipramine, trazodone, and diazepam in 230 patients with a DSM-III diagnosis of GAD, found trazodone to be more effective than placebo. In this study, diazepam was initially the most effective, especially against somatic symptoms, but by the end of the 8 weeks a moderate-to-marked improvement was reported by 73% of imipramine-treated patients, 69% of trazodone-treated patients, 66% of diazepam-treated patients, and 47% of the placebo group. Both the imipramine and trazodone groups had a better response on measures of tension, apprehension, and worry.71 Trazodone’s mechanism of action is not well understood; it is believed to work by inhibiting the reuptake of serotonin, and by acting as an antagonist at the serotonin receptor 5-HT2A/1C.  

 Nefazodone is an antidepressant with activity similar to both SSRIs and trazodone. There is evidence to suggest that nefazodone is a reasonable treatment option for GAD. One open trial of 21 patients with DSM-IV-based GAD were treated with nefazodone for 8 weeks. Of the 15 patients who completed the trial, 80% were rated as both very much or much improved, 7% as minimally improved, and 13% as unchanged.86 Unfortunately, nefazodone is associated with an elevation in liver enzymes that has led to a black box warning. This has diminished its use. 

Mirtazapine, a noradrenergic and specific serotonergic antidepressant, has a dual mechanism of action. It binds to and inhibits the α2 autoreceptor and the α2 heteroreceptor, which prevents negative feedback of synaptic noradrenaline on 5-HT and noradrenaline receptors. This results in enhanced serotonergic and noradrenergic activity. Mirtazapine also blocks 5-HT2 and 5-HT3 receptors on the postsynaptic membrane, which results in enhanced 5-HT neurotransmission. There was one 8-week, open-label trial of mirtazapine in patients with comorbid MDD and GAD. In this study, 10 patients were initially treated with mirtazapine 15 mg, which was titrated up to a maximum of 45 mg/day. Significant improvement in HAM-A and Hamilton Rating Scale for Depression scores occurred after the first week of treatment and persisted throughout the rest of the study.87



The clinical evidence for buspirone in the treatment of GAD is fairly extensive. There are 15 double-blind, placebo-controlled trials of buspirone, most of which compare it to a benzodiazepine, placebo, or both. Buspirone was more effective than placebo in 10 of these studies, but was never more effective than benzodiazepines.30 One study reanalyzed previous study data of 735 patients with DSM-III GAD treated with either placebo, buspirone, or a benzodiazepine to determine whether or not prior treatment with benzodiazepines might predict a poorer response with buspirone.  The results showed that patients with recent benzodiazepine treatment report much less improvement with buspirone than patients who are benzodiazepine-naïve, or patients who have a remote history of benzodiazepine treatment.88

Buspirone acts by modulating the serotonergic system. It has activity as a full agonist at the presynaptic 5-HT1A, where it inhibits serotonin synthesis and firing of the raphe nucleus. It acts as a partial agonist at the postsynaptic 5-HT1A, which results in increased receptor activity.89 Buspirone provides a nonbenzodiazepine option for the treatment of GAD, and may be useful in the benzodiazepine-naïve patient. Unfortunately, the population most likely to be prescribed buspirone, patients with comorbid alcohol, personality, or substance abuse disorders, are the least likely to feel its antianxiety effects. Buspirone has a slow onset of action, and may take several weeks to demonstrate a significant clinical effect. It may be difficult for an anxious patient to wait for the anxiolytic effect. Additionally, buspirone does not improve insomnia, which is a common complaint in GAD.  


Other Medications

A few French studies demonstrate some efficacy for hydroxyzine in the treatment of GAD. Hydroxyzine works by competetively blocking the effects of histamine at H1 receptor sites. One 4-week, double-blind, placebo-controlled study compared hydroxyzine 50 mg/day to placebo in 133 patients with DSM-III GAD. The hydroxyzine group had significantly more improvement on measures of anxiety. However, 52% of the treatment group complained of sleepiness, weight gain, dry mouth, and loss of concentration, all common antihistamine-related side effects.90 A longer, 3-month, double-blind, placebo-controlled, randomized trial comparing hydroxyzine 50 mg/day, bromazepam 6 mg/day, and placebo found that hydroxyzine was as effective as bromazepam and more effective than placebo in reducing CGI severity scale scores and HAD scores.91 Finally, a European multicenter, double-blind comparison of hydroxyzine 50 mg, buspirone 20 mg, and placebo found that only hydroxyzine was significantly better than placebo on the HAM-A, which was the primary outcome measure. Both hydroxyzine and buspirone were more effective than placebo on CGI and HAD ratings. However, this study is confounded by the idiosyncratic dosing of buspirone.92

Kava kava, a compound from a plant that grows in the Pacific Islands region, is reported to have analgesic and anxiolytic properties. It is also reported to inhibit monoamine oxidase B.93 However, the evidence supporting its use in the treatment of GAD is scant. An 8-week, randomized, double-blind, German trial treated 129 GAD patients with either 400 mg Kava LI 150, buspirone 10 mg, or opipramol 100 mg. There were no significant differences between any of the outcome measures in 127 of the treated patients. Seventy-five percent of patients were classified as responders and 60% achieved remission.94 Another study that compared kava kava against placebo in 37 patients with DSM-IV GAD found that improvements occurred in both treatments, and kava did not differ significantly from placebo.95 Studies of kava kava have been discontinued because of concerns about hepatoxicity.



CBT has been demonstrated to be an effective treatment for GAD. A study comparing nondirective therapy, CBT, and applied relaxation in outpatients with GAD found that both applied relaxation and CBT were superior to nondirective therapy by the end of the study. Longer-term follow-up of the subjects determined that the nondirective therapy group had lost its treatment gains, but both CBT and applied relaxation maintained their treatment gains, with the greatest gains maintained by the CBT group.96 Another study comparing CBT, behavior therapy, and a wait-list control group found that CBT was consistently more effective than behavior therapy on measures of cognition, mood, and anxiety.97

Although psychodynamic psychotherapy was employed for years in the psychotherapeutic treatment of anxiety, there is no evidence to support its efficacy in the treatment of GAD.



GAD is a common disorder and is highly comorbid with other psychiatric syndromes. GAD alone and GAD comorbid with other psychiatric syndromes is remarkably disabling and significantly impedes on the quality of life of individuals and their loved ones. At this time, little is understood about the biology of GAD, but it is hoped that advances in neurobiology and genomics will enhance our knowledge about the pathogenesis of this syndrome. Patients with GAD commonly approach PCPs with somatic concerns such as headaches, backaches, and muscle spasms. Fortunately, there are safe and effective pharmacotherapies and psychotherapies for GAD. However, the longer-term prognosis for GAD is still troubling. This emphasizes the need for more longitudinal treatment research as well as exploration of alternative modalities of therapy.  PP



1.    DuPont RL, Rice DP, Miller LS, Shiraki SS, Rowland CR, Harwood HJ. Economic costs of  anxiety disorders. Anxiety. 1996;2(4):167-172.

2.    Kessler RC, McGonagle KA, Zhao S, et al. Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States. Results from the National Comorbidity Survey. Arch Gen Psychiatry. 1994;51(1):8-19.

3.    Weiner H. The psychobiology and pathophysiology of anxiety and fear. In: Tuma AH, Maser J, eds. Anxiety and the Anxiety Disorders. Hillsdale, NJ: Lawrence Erlbaum; 1985:333-345.

4.    First MB, Spitzer RL, Gibbon M, Williams J. Structured Clinical Interview for DSM-IV Axis I Disorders, Clinician Version (SCID-CV). Washington, DC: American Psychiatric Press; 1997.

5.    Ballenger JC, Davidson JR, Lecrubier Y, et al. Consensus statement on generalized anxiety disorder from the International Consensus Group on Depression and Anxiety. J Clin Psychiatry. 2001;62(suppl 11):53-58.

6.    Posternak MA, Zimmerman M. Short-term spontaneous improvement rates in depressed outpatients. J Nerv Ment Dis. 2000;188(12):799-804.

7.    Fifer SK, Mathias SD, Patrick DL, Mazonson PD, Lubeck DP, Buesching DP. Untreated anxiety among adult primary care patients in a health maintenance organization. Arch Gen Psychiatry. 1994;51(9):740-750.

8.    Sareen J, Houlahan T, Cox BJ, Asmundson GJ. Anxiety disorders associated with suicidal ideation and suicide attempts in the National Comorbidity Survey. J Nerv Ment Dis. 2005;193(7):450-454.

9.    Diagnostic and Statistical Manual of Mental Disorders, 3rd ed. Washington, DC: American Psychiatric Association; 1980.

10.    Blazer DG, Hughes D, George LK, et al. Generalized anxiety disorder. In: Robins LN, Reiger DA, eds. Psychiatric Disorders in America: The Epidemiologic Catchment Area Study. New York, NY: The Free Press; 1990:180-203.

11.    Diagnostic and Statistical Manual of Mental Disorders, 4th ed. Washington, DC: American Psychiatric Association; 1994.

12.    Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry. 2005;62(6):593-602.

13.    Wittchen HU, Nelson CB, Lachner G. Prevalence of mental disorders and psychological impairments in adolescents and young adults. Psychol Med. 1998;28(1):109-206.

14.    Carter RM, Wittchen HU, Pfister H, Kessler RC. One-year prevalence of subthreshold and threshold DSM-IV generalized anxiety disorder in a nationally representative sample. Depress Anxiety. 2001;13(2):78-88.

15.    Ustun TB, Sartorius N (eds). Mental Illness in General Health Care: An International Study. New York, NY: John Wiley & Sons; 1996: 18-23.

16.    Olfson M, Broadhead WE, Weissman MM, et al. Subthreshold psychiatric symptoms in a primary care group practice. Arch Gen Psychiatry. 1996;53(10):880-886.

17.    Wittchen HU, Krause P, Hoyer J, et al. Prevalence and correlates of generalized anxiety disorders in primary care. Fortschr Med Orig. 2001;1999(suppl 1):17-25.

18.    Wittchen HU, Zhao S, Kessler RC, et al. DSM-III-R generalized anxiety disorder in the National Comorbidity Survey. Arch Gen Psychiatry. 1994;51(5):355-364.

19.    Noyes R Jr. Comorbidity in generalized anxiety disorder. Psychiatr Clin N Am. 2001;24(1):41-55.

20.    Sanderson WC, Barlow DH. A description of patients diagnosed with DSM-III-R generalized anxiety disorder. J Nerv Ment Dis. 1990;178:588-581.

21.    Brown TA, Barlow DH. Comorbidity among anxiety disorders: implications for treatment and DSM-IV. J Consult Clin Psychol. 1992;60(6):835-844.

22.    Noyes R Jr, Woodman C, Garvey MJ, et al. Generalized anxiety disorder vs. panic disorder. Distinguishing characteristics and patterns of comorbidity. J Nerv Ment Dis. 1992;180(6):369-379.

23.    Brawman-Mintzer O, Lydiard RB, Emmanuel N, et al. Psychiatric comorbidity in patients with generalized anxiety disorder. Am J Psychiatry. 1993;150(8):1216-1218.

24.    Starcevic V, Fallon S, Uhlenhuth EH, Pathak D. Comorbidity rates do not support the distinction between panic disorder and generalized anxiety disorder. Psychopathology. 1994;27(6):269-272.

25.    Goisman RM, Goldenberg I, Vasile RG, et al. Comorbidity of anxiety disorders in a multicenter anxiety study. Compr Psychiatry. 1995;36(4):303-311.

26.    Zimmerman M, Rothschild L, Chelminski I. The prevalence of DSM-IV personality disorders in psychiatric outpatients. Am J Psychiatry. 2005;162(10):1911-1918.

27.    Grant BF, Hasin DS, Stinson FS, et al. Co-occurrence of 12-month mood and anxiety disorders and personality disorders in the US: Results from the national epidemiologic survey on alcohol and related conditions. J Psychiatr Res. 2005;39(1):1-9.

28.    Kessler RC. The epidemiology of pure and comorbid generalized anxiety disorder: a review and evaluation of recent research. Acta Psychiatr Scand Suppl. 2000;406:7-13.

29.    Bland RC, Newman SC, Orn H. Help-seeking for psychiatric disorders. Can J Psychiatr. 1997;42(9):935-942.

30.    Hoge E, Oppenheimer J, Simon N.  Generalized anxiety disorder. Focus. 2004;2(3):346-358.

31.    Diagnostic and Statistical Manual of Mental Disorders, 3rd ed rev. Washington, DC: American Psychiatric Association; 1987.

32.    Wittchen HU, Kessler RC, Beesdo K, Krause P, Hofler M, Hoyer J. Generalized anxiety disorder and depression in primary care: prevalence, recognition, and management. J Clin Psychiatry. 2002;63(suppl 8):24-34.

33.    Hoehn-Saric R, Hazlett RL, McLeod DR. Generalized anxiety disorder with early and late onset of anxiety symptoms. Compr Anxiety. 1993;34(5):291-298.

34.    Keller MB. The long-term clinical course of generalized anxiety disorder. J Clin Psychiatry. 2002;63(suppl 8):11-16.

35.    Yonkers KA, Dyck IR, Warshaw M, Keller MB. Factors predicting the clinical course of generalised anxiety disorder. Br J Psychiatry. 2000;176:544-549.

36.    Tyrer P, Seivewright H, Johnson T. The Nottingham Study of Neurotic Disorder: predictors of 12-year outcome of dysthymic, panic and generalized anxiety disorder. Psychol Med. 2004;34(8):1385-1394.

37.    Yonkers KA, Warshaw MG, Massion AO, Keller MB. Phenomology and course of generalised anxiety disorder. Br J Psychiatry. 1996;168(3):308-313.

38.    Souetre E, Lozet H, Cimarosti I, et al. Cost of anxiety disorders: impact of comorbidity. J Psychosom Res. 1994;38(suppl 1):151-160.

39.    Walker EA, Roy-Byrne PP, Katon WJ, et al. Psychiatric illness and irritable bowel syndrome: a comparison with inflammatory bowel disease. Am J Psychiatry. 1990;147(12):1656-1661.

40.    Lydiard RB, Fossey MD, Marsh W, et al. Prevalence of psychiatric disorders I patients with irritable bowel syndrome. Psychosomatics. 1993;34(3):229-234.

41.    Crestani F, Lorez M, Baer K. Decreased GABAA receptor clustering results in enhanced anxiety and a bias for threat cues. Nat Neurosci. 1999;2(9):833-839.

42.    Rickels K, Case WG, Schweizer E. The drug treatment of anxiety and panic disorder. Stress Med. 1988;4(4):231-239.

43.    Rocca P, Beoni AM, Eva C, Ferrero P, Zanalda E, Ravizza L. Peripheral benzodiazepine receptor messenger RNA is decreased in lymphocytes of generalized anxiety disorder patients. Biol Psychiatry. 1998;43(10):767-773.

44.    Sevy S, Papadimitriou GN, Surmont DW, Goldman S, Mendlewicz J. Noradrenergic function in generalized anxiety disorder, major depressive disorder, and healthy subjects. Biol Psychiatry. 1989;25(2):141-152.

45.    Munjack DJ, Baltazar PL, DeQuattro V, et al. Generalized anxiety disorder: some biochemical aspects. Psychiatry Res. 1990;32(1):35-43.

46.    Mathew RJ, Ho BT, Francis DJ, Taylor DL, Weinman ML. Catecholamines and anxiety. Acta Psychiatr Scand. 1982;65(2):142-147.

47.    Khan A, Lee E, Dager S, et al. Platelet MAO-B activity in anxiety and depression. Biol Psychiatry. 1986;21(8-9):847-849.

48.    Tiihonen J, Kuikka J, Rasanen P, et al. Cerebral benzodiazepine receptor binding and distribution in generalized anxiety disorder: a fractal analysis. Mol Psychiatry. 1997;2(6):463-471.

49.    Abelson JL, Glitz D, Cameron OG, et al. Blunted growth hormone response to clonidine in patients with generalized anxiety disorder. Arch Gen Psychiatry. 1991;48(2):157-162.

50.    Charney DS, Woods SW, Heninger GR. Noradrenergic function in generalized anxiety disorder: effects of yohimbine in healthy subjects and patients with generalized anxiety disorder. Psychiatry Res. 1989;27(2):173-182.

51.    Jetty PV, Charney DS, Goddard AW. Neurobiology of generalized anxiety disorder. Psychiatr Clin North Am. 2001;24(1):75-97.

52.    Nutt DJ. Neurobiological mechanisms in generalized anxiety disorder. J Clin Psychiatry. 2001;62(suppl 11):22-27.

53.    Gray JA. The neuropsychological basis of anxiety. In: Last CG, Hersen M, eds. Handbook of Anxiety Disorders. New York, NY: Pergamon Press; 1988:10-37.

54.    Tafet GE, Idoyaga-Vargas VP, Abulafia DP, et al. Correlation between cortisol level and serotonin uptake in patients with chronic stress and depression. Cogn Affect Behav Neurosci. 2001;1(4):388-393.

55.    Tafet GE, Toister-Achituv M, Shinitzky M. Enhancement of serotonin uptake by cortisol: a possible link between stress and depression. Cogn Affect Behav Neurosci. 2001;1(1):96-104.

56.    Butler PD, Weiss JM, Stout JC, Nemeroff CB. Corticotropin-releasing factor produces fear-enhancing and behavioral activating effects following infusion into the locus coeruleus. J Neurosci. 1990;10(1):176-183.

57.    Griebel G. Is there a future for  neuropeptide receptor ligands in the treatment of anxiety disorders? Pharmacol Ther. 1999;82(1):1-61.

58.    Koob GF, Gold LH. Molecular biological approaches in the behavioural pharmacology of anxiety and depression. Behav Pharmacol. 1997;8(7):652.

59.    Noyes R Jr, Clarkson C, Crowe RR, Yates WR, McChesney CM. A family study of generalized anxiety disorder. Am J Psychiatry. 1987;144(8):1019-1024.

60.    Skre I, Onstad S, Edvardsen J, Torgersen S, Kringlen E. A family study of anxiety disorders: familial transmission and relationship to mood disorder and psychoactive substance use disorder. Acta Psychiatr Scand. 1994;90(5):366-374.

61.    Andrews G, Stewart G, Allen R, Henderson AS. The genetics of six neurotic disorders: a twin study. J Affect Disord. 1990;19(1):23-29.

62.    Ohara K, Suzuki Y, Ochiai M, Tsukamoto T, Tani K, Ohara K. A variable-number-tandem-repeat of the serotonin transporter gene and anxiety disorders. Prog Neuropsychopharmacol Biol Psychiatry. 1999;23(1):55-65.

63.    Mahe V, Balogh A. Long-term pharmacological treatment of generalized anxiety disorder. Int Clin Psychopharmacol. 2000;15(2):99-105.

64.    Pollack MH. Optimizing pharmacotherapy of generalized anxiety disorder to achieve remission. J Clin Psychiatry. 2001;62(suppl 19):20-25.

65.    Shader RI, Greenblatt DJ. Use of benzodiazepines in anxiety disorders. N Engl J Med. 1993;328(19):1398-1405.

66.    Dubovsky SL.  Generalized anxiety disorder: new concepts and psychopharmacologic therapies. J Clin  Psychiatry. 1990;51(suppl):3-10.

67.    Sadock BJ, Sadock VA. Biological therapies. In: Sadock BJ, Sadock VA, eds. Synopsis of Psychiatry: Behavioral Sciences/Clinical Psychiatry. Baltimore, MD: Lippincott, Williams & Wilkins; 1991:907-911.

68.    Ballenger J. Current treatment of the anxiety disorders in adults. Biol Psychiatry. 1999;46(11):1579-1594.

69.    Rickels K, Downing R, Winokur A. Long-term diazepam therapy and clinical outcome.  JAMA. 1983;250(6):767-771.

70.    Hollister LE, Conley FK, Britt RH, Shuer L. Long-term use of diazepam. JAMA.  1981;246(14):1568-1570.

71.    Rickels K, Downing R, Schweizer E, Hassman H. Antidepressants for the treatment of generalized anxiety disorder. A placebo-controlled comparison of imipramine, trazodone, and diazepam. Arch Gen Psychiatry. 1993;50(11):884-895.

72.    Hoehn-Saric R, McLeod DR, Zimmerli WD.  Differential effects of alprazolam and imipramine in generalized anxiety disorder: somatic versus psychic symptoms. J Clin Psychiatry.  1988;49(8):293-301.

73.    Balon R, Mufti R, Arfken C. A survey of prescribing practices for monoamine oxidase inhibitors. Psychiatr Serv. 1999;50(7):945-947.

74.    Wingerson D, Nguyen C, Roy-Byrne PP. Clomipramine treatment for generalized anxiety disorder. J Clin Psychopharmacol. 1992;12(3):214-215.

75.    Rocca P, Fonzo V, Scotta M, Zanalda E, Ravizza L. Paroxetine efficacy in the treatment of generalized anxiety disorder. Acta Psychiatr Scand. 1997;95(5):444-450.

76.    Bellew KM, McCafferty JP, Iyengar M, et al.  Paroxetine treatment of GAD: A double blind, placebo-controlled trial. Presented at: the 153rd Annual Meeting of the American Psychiatric Association; May 13-18, 2000; Chicago.

77.    Pollack MH, Zaninelli R, Goddard A, et al. Paroxetine in the treatment of generalized anxiety disorder: results of a placebo controlled, flexible-dosage trial. J Clin Psychiatry. 2001;62(5):350-357.

78.    Lenze EJ, Mulsant BH, Shear MK, et al. Efficacy and tolerability of citalopram in the treatment of late-life anxiety disorders: results from an 8-week randomized, placebo-controlled trial. Am J  Psychiatry. 2005;162(1):146-150.

79.    Davidson JR, Bose A, Korotzer A, Zheng H.  Escitalopram in the treatment of generalized anxiety disorder: double-blind, placebo controlled flexible-dose study. Depress Anxiety. 2004;19(4):234-240.

80.    Ball SG, Kuhn A, Wall D, Shekhar A, Goddard AW. Selective serotonin reuptake inhibitor treatment for generalized anxiety disorder: a double-blind, prospective comparison between paroxetine and sertraline. J Clin Psychiatry. 2005;66(1):94-99.

81.    Davidson JR, DuPont RL, Hedges D, Haskins JT. Efficacy, safety, and tolerability of venlafaxine extended release and buspirone in outpatients with generalized anxiety disorder.  J Clin Psychiatry. 1999;60(8):528-535.

82.    Rickels K, Pollack MH, Sheehan DV, Haskins JT. Efficacy of extended-release venlafaxine in nondepressed outpatients with generalized anxiety disorder. Am J Psychiatry. 2000;157(6):968-974.

83.    Nimatoudis I, Zissis NP, Kogeorgos J, Theodoropoulou S, Vidalis A, Kaprinis G. Remission rates with venlafaxine extended release in Greek outpatients with generalized anxiety disorder. A double-blind, randomized, placebo controlled study. Int Clin Psychopharmacol.  2004;19(6):331-336.

84.    Gelenberg A, Lydiard RB, Rudolph R, Aguiar L, Haskins JT, Salinas E. Efficacy of venlafaxine extended-release capsules in nondepressed outpatients with generalized anxiety disorder: A 6-month randomized controlled trial. JAMA. 2000;283(23):3082-3088.

85. Allgulander C, Hackett D, Salinas E.  Venlafaxine extended release (ER) in the treatment of generalised anxiety disorder: twenty-four-week placebo-controlled dose-ranging study. Br J Psychiatry. 2001;179:15-22.

86. Hedges DW, Reimherr FW, Strong RE, Halis CH, Rust C. An open trial of nefazodone in adult patients with generalized anxiety disorder. Psychopharmacol Bull. 1996;32(4):671-676.

87. Goodnick PJ, Puig A, DeVane CL, Freund BV. Mirtazapine in major depression with comorbid generalized anxiety disorder. J Clin Psychiatry. 1999;60(7):446-448.

88. DeMartinis N, Rynn M, Rickels K, Mandos L. Prior benzodiazepine use and buspirone response in the treatment of generalized anxiety disorder. J Clin Psychiatry. 2000;61(2):91-94.

89. Apter J, Allen L.  Buspirone: future directions.  J Clin Psychopharmacol. 1999;19(1):86-93.

90. Ferreri M, Hantouche EG, Billardon M.  Value of hydroxyzine in generalized anxiety disorder: controlled double blind study versus placebo. Encephale. 1994; 29(6): 785-791.

91. Llorca PM, Spadone C, Sol O, et al. Efficacy and safety of hydroxyzine in the treatment of generalized anxiety disorder: a 3-month double-blind study. J Clin Psychiatry. 2002;63(11):1020-1027.

92. Lader M, Scotto JC. A multicentre double-blind comparison of hydroxyzine, buspirone and placebo in patients with generalized anxiety disorder. Psychopharmacology (Berl). 1998;139(4):402-406.

93. Uebelhack R, Franke L, Schewe HJ. Inhibition of platelet MAO-B by kava pyrone-enriched extract from Piper methysticum Forster (kava-kava). Pharmacopsychiatry. 1998;31(5):187-92.

94. Boerner R, Sommer J, Berger W, Kuhn U, Schmidt U, Mannel M. Kava-kava extract li 150 is as effective as opipramol and buspirone in generalized anxiety disorder—an 8 week randomized, double blind multicentre clinical trial in 129 outpatients. Phytomedicine. 2003;10(suppl 4):38-49.

95. Connor KM, Davidson JR. A placebo-controlled study of kava kava in generalized anxiety disorder. Int Clin Psychopharmacol. 2002;17(4):185-188.

96. Borkovec TD, Costello E. Efficacy of applied relaxation and cognitive-behavioral therapy in the treatment of generalized anxiety disorder.  J Consult Clin Psychol. 1993;61(4):611-619.

97. Butler G, Fennell M, Robson P, Gelder M. Comparison of behavior therapy and cognitive behavior therapy in the treatment of generalized anxiety disorder. J Consult Clin Psychol.



Dr. Yusim is a third-year psychiatry resident at the New York University (NYU) Medical Center in New York City.

Disclosure: Dr. Yusim reports no affiliation with or financial interest in any organization that may pose a conflict of interest.

Acknowledgments: Dr. Yusim thanks Dr. Stephen J. Ferrando, MD, in the Department of Psychiatry at New York-Presbyterian/Weill Cornell Medical Center and Benjamin Sadock, MD, in the Department of Psychiatry at NYU Medical Center for their insight and advice in the preparation of this manuscript.

Please direct all correspondence to: Anna Yusim, MD, 207 E. 88th St, #3D, New York, NY 10128; Tel: 203-704-0274; E-mail:


Focus Points

• Although limbs are most commonly affected, phantom sensation or pain may appear in virtually any part of the body, somatic or visceral, that is accessible to sensory perception. 
• Phantom pain has been described in many body parts including the tongue, bladder, breast, rectum, and penis.  Phantom pain in the genital region is particularly rare. 
• The pathophysiologic mechanism of phantom pain is not well understood, though the dominant view holds that it is created by activity in the parietal sensory cortex that formerly subserved somatosensation in the amputated part.
• Although physicians believe that there are effective treatments for phantom pain, <10% of phantom pain patients receive relief with prescribed medication.
• Further study into the etiology, mechanism, and treatment of phantom pain is necessary in order to create an evidence-based treatment approach and subsequently reduce the unnecessary suffering caused by this elusive clinical syndrome.



Phantom sensation or pain is the persistent perception that a body part exists or is painful after it has been removed. Although limbs are most commonly affected, phantom sensation or pain may appear in virtually any part of the body, somatic or visceral, that is accessible to sensory perception. The following is a case of phantom testicular pain that presented in a young man 18 years after bilateral orchiectomy.


“One might say that a phantom limb is a part of the self, poorly or incompletely forgot; an obliteration not yet registered in the corporeal world. Once the physical form has changed, the memory of a prior self becomes indistinct over time as the body reconciles itself to its new body shape. The body is suspended between a long-acquired memory and a newly acquired need to forget. The dysfunctional conversation that the body is having with its absent limb gives rise to fanciful forms, distortions, enigmatic sensations and eventually, the patient hopes, silence.”1

Phantom sensation or pain is the persistent perception that a body part exists or is painful after it has been removed. It can occur from as early as 1 week after removal of the body part to as late as 40 years later.2,3 Although limbs are most commonly affected, phantom sensation or pain may appear in virtually any part of the body, somatic or visceral, that is accessible to sensory perception. It has been described in many body parts including the tongue,4 bladder,5 breast,6,7 rectum,8 and penis.9 Phantom pain in the genital region is particularly rare. The following is a case of phantom testicular pain that presented in a young man 18 years after bilateral orchiectomy.

Case Report

Mr. C, a 33-year-old domiciled employed African-American male with a history of depression who sustained a bilateral orchiectomy after severe trauma to his groin at 15 years of age, presented to our emergency department for complaints of severe pain in the area where his testicles used to be. After his bilateral orchiectomy, the patient was free of sensation or pain in this area for 18 years. Seven months prior to his emergency department presentation, the patient began to feel a faint sensation in his “testicles,” which evolved into pain shortly thereafter. The patient was aware that he did not have testicles and was confused at having sensory experiences as if he did. The patient described the pain as a bilateral burning, tingling, and warm sensation. The pain was episodic in nature, would occur several times per day, and would last on average 1–2 hours per episode. The severity of the pain was 6–8 out of 10 on the pain scale, but rose to 10 out of 10 on the day he presented to the emergency department. Motion and touch exacerbated the pain, whereas strong pressure to the area attenuated it. The pain could be reproduced by pressing on several trigger points on the lateral scrotum. It was not associated with urination, defecation, coughing, or temperature changes. Physical exam of the empty scrotal sac was unremarkable. Urologic and musculoskeletal examinations were likewise unremarkable. The patient’s admission labs were within normal limits and scrotal infection was ruled out, including herpes zoster. The patient was not immunocompromised. Spinal magnetic resonance imaging ruled out disc herniation, as radiculitis is one of the most common causes of scrotal pain. Computerized axial tomography and ultrasound showed absence of testes, but were otherwise unremarkable. Referred pain to the “testes” caused by a ureteral stone was also ruled out, as were other common etiologies of scrotal and testicular pain.10 By diagnosis of exclusion, the patient was believed to have developed phantom testicular pain 18 years after removal of his testes.


Unlike pain which serves as a warning of bodily injury, phantom pain has no “ulterior motives,” prophylactic purpose or obvious evolutionary significance. Psychodynamic theory suggests that phantom pain may represent an important unconscious defense, through denial, against the loss of bodily integrity and castration anxiety.11 When the lost body part is genital in origin, castration anxiety becomes castration reality, thereby heightening the need for such a defense.

The pathophysiologic mechanism of phantom pain is not well understood, though the dominant view holds that it is created by activity in the parietal sensory cortex that formerly subserved somatosensation in the amputated part.12 In the sensory homunculus of Penfield and Rasmussen,13 the genitalia are depicted on the medial surface of the hemisphere, in the most inferior position, just posterior to the region of the toes. In their electrical exploration of the human cortex, sensory responses referred to the genital region were rare.

In patients whose phantom pain increases in severity or begins a long time after removal of the affected body part, a differential diagnosis must be entertained. In a patient with a history of depression, a psychotic etiology of the pain must be considered. Although our patient did not have evidence of other psychotic features, it is possible that his “testicular” pain was actually a somatic delusion resulting from a mood disorder with psychotic features. The distinction between phantom pain and somatic delusions is itself questionable, as phantom pain is arguably a “hallucination” of a body part that is no longer there.

The treatment of phantom pain is particularly challenging. Although physicians believe that there are effective treatments,14 <10% of phantom pain patients receive relief with prescribed medication.15 Lack of clinical trials and evidence-based treatment guidelines further complicate this problem.16 Further study into the etiology, mechanism, and treatment of phantom pain is necessary in order to create an evidence-based treatment approach and subsequently reduce the unnecessary suffering caused by this elusive clinical syndrome. PP


1.    Patrizio A. The Phantom Limb in Contemporary Art and Exhibition Practice. Available at: Accessed October 2, 2008.
2.    Ribera H, Cano P, Dora A and Garrido GP. Phantom limb pain secondary to post-trauma stump hematoma 40 years after amputation: description of one case. Rev Soc Esp Dolor. 2001;8(3):217-220.
3.    Rajbhandari SM, Jarratt JA, Griffiths PD, Ward JD. Diabetic neuropathic pain in a leg amputated 44 years previously. Pain.1999;83(3):627-629.
4.    Hanowell ST, Kennedy SF. Phantom tongue pain and causalgia: case presentation and treatment. Anesth Analg. 1979;58(5):436-438.
5.    Biley FC. Phantom bladder sensations: a new concern for stoma care workers. Br J Nurs. 2001;10(19):1290-1296.
6.    Jamison K, Wellisch DK, Katz RL, Pasnau RO. Phantom breast syndrome. Arch Surg. 1979;114(1):93-95.
7.    Krøner K, Krebs B, Skov J, Jørgensen HS. Immediate and long-term phantom breast syndrome after mastectomy: incidence, clinical characteristics and relationship to pre-mastectomy breast pain. Pain. 1989;36(3):327-334.
8.    Ovesen P, Krøner K, Ornsholt J, Bach K. Phantom-related phenomena after rectal amputation: prevalence and clinical characteristics. Pain. 1991;44(3):289-291.
9.    Fisher CM. Phantom erection after amputation of penis. Case description and review of the relevant literature on phantoms. Can J Neurol Sci. 1999;26(1):53-56.
10.    Holland JM, Feldman JL, Gilbert HC. Phantom orchalgia. J Urol. 1994;152(6 Pt 2):2291-2293.
11.    Noble D, Price DB, Gilder R. Psychiatric disturbances following amputation. Am J Psychiatry. 1954;110(8):609-613.
12.    Melzack R. Phantom limbs and the concept of a neuromatrix. Trends Neurosci. 1990;13(3):88-92.
13.    Penfield W, Rasmussen T. The Cerebral Cortex of Man. New York, NY: The MacMillan Company; 1950.
14.    Sherman RA, Sherman CJ, Gall NG. A survey of current phantom limb pain treatment in the United States. Pain. 1980;8(1):85-99.
15.    Sherman RA, Sherman CJ, Parker L. Chronic phantom and stump pain among American veterans: results of a survey. Pain. 1984;18(1):83-95.
16. Manchikanti L, Singh V. Managing phantom pain. Pain Physician. 2004;7(3):365-375.




Needs Assessment: Despite available treatment strategies, numerous patients with depression remain symptomatic. There is an urgent need to expand available augmentation strategies for depression. The use of atypical antipsychotics represents one such strategy, but clinicians must be updated on the status of clinical evidence supporting this treatment strategy.

Learning Objectives:
• Describe the urgent need to expand our treatment armamentarium for treatment-resistant major depressive disorder (MDD).
• Understand the extent of the evidence supporting the use of the atypical antipsychotics for treatment-resistant MDD.
•  Understand limitations in knowledge regarding the use of the atypical antipsychotics for treatment-resistant MDD.

Target Audience: Primary care physicians and psychiatrists.

CME Accreditation Statement: This activity has been planned and implemented in accordance with the Essentials and Standards of the Accreditation Council for Continuing Medical Education (ACCME) through the joint sponsorship of the Mount Sinai School of Medicine and MBL Communications, Inc. The Mount Sinai School of Medicine is accredited by the ACCME to provide continuing medical education for physicians.

Credit Designation: The Mount Sinai School of Medicine designates this educational activity for a maximum of 3 AMA PRA Category 1 Credit(s)TM. Physicians should only claim credit commensurate with the extent of their participation in the activity.

Faculty Disclosure Policy Statement:
It is the policy of the Mount Sinai School of Medicine to ensure objectivity, balance, independence, transparency, and scientific rigor in all CME-sponsored educational activities. All faculty participating in the planning or implementation of a sponsored activity are expected to disclose to the audience any relevant financial relationships and to assist in resolving any conflict of interest that may arise from the relationship. Presenters must also make a meaningful disclosure to the audience of their discussions of unlabeled or unapproved drugs or devices. This information will be available as part of the course material.

This activity has been peer-reviewed and approved by James C.-Y. Chou, MD, associate professor of psychiatry at the Mount Sinai School of Medicine, and Norman Sussman, MD, editor of Primary Psychiatry and professor of psychiatry at New York University School of Medicine. Review Date: October 6, 2008.

Dr. Sussman reports no affiliation with or financial interest in any organization that may pose a conflict of interest. Dr. Chou receives honoraria from AstraZeneca, Bristol-Myers Squibb, Eli Lilly, Janssen, and Pfizer.

To receive credit for this activity: Read this article and the two CME-designated accompanying articles, reflect on the information presented, and then complete the CME posttest and evaluation. To obtain credits, you should score 70% or better. Early submission of this posttest is encouraged: please submit this posttest by November 1, 2010 to be eligible for credit. Release date: November 1, 2008. Termination date: November 30, 2010. The estimated time to complete all three articles and the posttest is 3 hours.


Primary Psychiatry. 2008;15(11):44-47


Dr. Papakostas is an assistant professor of psychiatry at Harvard Medical School and director of Treatment-Resistant Depression Studies at the Massachusetts General Hospital Depression Clinical and Research Program in Boston.

Disclosure relevant towards this article (as of October 1, 2008): Dr. Papakostas has served (over the past 12 months) or has agreed to serve (over the next 12 months) as a consultant for Eli Lilly Co. (makers of olanzapine). Dr. Papakostas has received (over the past 12 months) or is due to receive (over the next 12 months) honoraria from Bristol-Myers Squibb Company (makers of aripiprazole), Eli Lilly Co. (makers of olanzapine), and Pfizer Inc (makers of ziprasidone). Dr. Papakostas has received (over the past 12 months) research support or is due to receive research support (over the next 12 months) from Bristol-Myers Squibb Company (makers of aripiprazole), the National Institute of Mental Health, and Pfizer Inc. (makers of ziprasidone). Dr. Papakostas has served (over the past 12 months) on the speaker’s bureau for Bristol-Myers Squibb Company (makers of aripiprazole), and has agreed to do so throughout 2008.

This article includes discussion of the following unapproved medications for major depressive disorder: olanzapine, quetiapine, risperidone, and ziprasidone.

Please direct all correspondence to: George I. Papakostas, MD, Massachusetts General Hospital, Department of Psychiatry, Depression Clinical and Research Program, 15 Parkman St, WACC 812, Boston, MA 02114; Tel: 617-726-6697; Fax: 617-726-7541; E-mail:


Given the relatively modest response and remission rates seen during monotherapy of major depressive disorder (MDD) with all contemporary, first-line antidepressants, there is an urgent need to develop treatments which are both safer and more effective than those currently available. Augmentation of standard antidepressants with atypical antipsychotics represents one such treatment development approach. The preclinical rationale for using atypical antipsychotics as adjunctive therapy in MDD lies in their affinity for a number of serotonergic structures, including the serotonin-1 and -2 receptor subtypes. As a result, numerous randomized, double-blind, placebo-controlled trials have recently been conducted focusing on evaluating whether or not this novel treatment strategy is truly efficacious. This article reviews such studies to inform the reader of the evidence supporting this treatment approach.


Major depressive disorder (MDD) is a prevalent illness that can often be quite challenging to treat. Traditionally, antidepressants used as monotherapy have represented the most common (especially as first-line agents) approach to target the illness for >50 years. Over time, clinicians and researchers have come to realize that this approach, while relatively convenient and user friendly, is only effective for a subset of patients.1 As a result, during the course of the last 25 years we have seen a steady growth of studies focusing on testing the efficacy, safety, and tolerability of adjunctive therapies for MDD, either as first-line therapy or for treatment-resistant populations.1 While some studies have concluded that adjunctive therapy is, for the most part, not more effective than monotherapy, others have identified adjunctive treatment therapies that, when employed, have been shown to incrementally improve the outcome of patients with MDD. As a result, these treatments have been incorporated in the standard treatment armamentarium of clinicians nation wide.1 The results of the recently published sequenced treatment alternatives to relieve depression (STAR*D) trial suggest that many MDD patients remain symptomatic despite such adjunctive treatment therapies.2 Results of studies like STAR*D serve to remind clinicians and researchers alike of the urgent need to continue to develop new treatments for MDD. Atypical antipsychotics have recently received attention as potential novel adjunctive therapies for MDD, with numerous randomized, double-blind, placebo-controlled trials either published or presented at major scientific meetings. This article reviews such studies to inform the reader regarding the evidence in support of this treatment approach.

Atypical Antipsychotics: Neuropharmacology

The preclinical rationale for the use of the atypical antipsychotics for the treatment of MDD derives from their complex neuropharmacologic effects at various monoaminergic receptors and transporters.3 Specifically, all atypical antipsychotics appear to be serotonin (5-HT)2 receptor antagonists.4,5 Atypical antipsychotics also vary in terms of neuropharmacologic effects. Ziprasidone and aripiprazole possess affinity for the 5-HT1A receptor,4,5 while ziprasidone and risperidone possess affinity for the 5-HT1D receptor.4,5 Ziprasidone has also been shown to inhibit the reuptake of serotonin and norepinephrine,6 while aripiprazole has been shown to possess mixed agonist and antagonist effects at various dopamine receptors.7 These effects were thought to be suggestive of potential antidepressant activity.3

Atypical Antipsychotics: Clinical Studies

Not unlike many alternative augmentation strategies for treatment-resistant depression (TRD), until recently, there was a paucity of double-blind, placebo-controlled studies examining whether augmentation of standard antidepressants with atypical antipsychotics for TRD is, truly, an efficacious treatment strategy. Shelton and colleagues8 randomized 28 outpatients with fluoxetine-resistant MDD to continue treatment with fluoxetine monotherapy, undergo a switch to olanzapine monotherapy, or receive olanzapine augmentation for 8 weeks. Greater remission rates were reported for patients who received augmentation of fluoxetine with olanzapine than patients who received treatment with either fluoxetine or olanzapine monotherapy. However, two subsequent trials also focusing on the use of olanzapine failed to demonstrate the superiority of olanzapine plus fluoxetine versus fluoxetine monotherapy for patients with nortriptyline- or venlafaxine-resistant MDD.9,10 In 2006 alone, however, seven new double-blind, placebo-controlled trials focusing on the use of adjunctive olanzapine (included two trials),11 risperidone,12,13 or quetiapine14-16 were either presented at major scientific meetings or published. Six of these demonstrated greater efficacy for adjunctive treatment with an atypical antipsychotic than placebo in TRD (Thase and colleagues11 reported two separate but identical trials of olanzapine augmentation of fluoxetine, and found olanzapine augmentation to be effective in one but not the second study).

These results appear to be very promising for the use of this treatment strategy. In fact, a meta-analysis pooling data from these first 10 studies demonstrated greater remission rates for patients who received adjunctive atypical antipsychotics than placebo (Figure 1).17 However, tolerability appeared to be a considerable limitation of this treatment approach, with a greater than three-fold higher rate of premature discontinuation of treatment for patients who received atypical augmentation than those who received placebo (37% vs. 12%). While these discontinuation rates appear somewhat high, STAR*D2 found that discontinuation due to intolerance increased with each treatment failure. For example, at levels 3 and 4, dropout rates were 25.6% and 30.1%, respectively.18-20 Thus, the higher rates in the atypical trials may reflect inclusion of more resistant patients. Other more serious side effects, such as weight gain, dyslipidemia, hypercholesterolemia, hypertriglyceridemia, glucose dysregulation, metabolic syndrome, hyperprolactinemia, neuroleptic malignant syndrome, and tardive dyskinesia have not been as well studied in the augmentation trials and would benefit from longer duration trials in any case. Further, there is a paucity of data from randomized, double-blind, placebo-controlled trials focusing on the risk of such adverse events during adjunctive therapy with atypical antipsychotics specifically for patients with MDD.



Placebo-controlled trials focusing on the use of olanzapine report significantly higher rates of somnolence,8,10 hypersomnia,11 peripheral edema,10,11 dry mouth,11 and tremor9 among patients treated with the combination of olanzapine and fluoxetine versus those treated with fluoxetine monotherapy. In addition, significantly greater weight-gain (mean difference in change in weight between treatment groups of 5.79 Kgr,8 4.7 Kgr,9 4.3 Kgr,10 and 4.5 Kgr,11) increases in cholesterol (mean difference in the degree of change in levels between treatment groups of 0.3 mmol/L9 or 0.4 mmol/L,10) triglyceride (mean difference in the degree of change in levels between the two treatment groups of 23.9 mg/dL,12) and prolactin levels (mean difference in the degree of change in levels between the two treatment groups of 0.31 mmol/L10) were reported among patients treated with the olanzapine plus fluoxetine versus those treated with fluoxetine monotherapy. There was no significant difference in the change over time in QTc interval duration,9-11 akathisia or dyskinesia scores,9,10 or mean serum glucose levels9-11 between the two treatment groups.

In contrast to studies focusing on the use of olanzapine, placebo-controlled trials focusing on the use of risperidone do not report on the majority of relevant safety and tolerability outcomes (ie change in cholesterol, triglyceride, prolactin, glucose levels, or QTc interval). Significantly greater weight gain (mean difference in change in weight between treatment groups of 2.5 lbs12 or 4.0 lbs13) was reported among patients treated with the risperidone plus antidepressants versus those treated with antidepressants alone. Finally, it is worth noting that, unlike studies employing the use of either olanzapine or risperidone, the three placebo-controlled trials focusing on the use of quetiapine14-16 were too small and underpowered to allow for an accurate assessment of the safety and tolerability of this treatment approach.

More recently, two separate but identical, double-blind, placebo-controlled trials investigating the use of adjunctive aripiprazole in MDD were published.21,22 These studies enrolled MDD patients resistant to 1–3 historic antidepressant trials, followed by an 8-week, open-label trial with either a selective serotonin reuptake inhibitor (SSRI; including fluoxetine, sertraline, paroxetine, or escitalopram, but not fluvoxamine) or the serotonin norepinephrine reuptake inhibitor venlafaxine. Non-responders to antidepressant monotherapy then randomized to have had either aripiprazole or placebo added to their antidepressant treatment regimen, for a total of 6 weeks. In both trials, a statistically significant difference in remission rates was observed (Figure 2),21,22 results that led to the approval by the Food and Drug Administration of aripiprazole as adjunctive therapy for TRD. This was the first ever approval for the use of any medication for TRD by the FDA. Significantly more patients treated with adjunctive aripiprazole than those treated with antidepressant monotherapy complained of akathisia,21,22 restlessness,22 insomnia,22 fatigue,22 tremor,22 and constipation.22 There was also significantly greater weight gain (mean difference in change in weight between treatment groups of 1.05 Kgr22 or 1.67 Kgr21), as well as greater increases in the scores of scales measuring the severity of extrapyramidal side-effects (Barnes Akathisia Scale; Simpson-Angus scale21,22) among patients treated with adjunctive aripiprazole than those treated with antidepressant monotherapy. There was no significant difference in the change in prolactin levels22 or QTc interval duration21 between these two treatment groups.


While studies conducted so far are in support of this treatment strategy, several questions remain. First, it is unclear whether all atypical antipsychotics are efficacious as adjunctive treatments in MDD and, if so, whether they are equally efficacious. For example, at the present time, evidence supporting the use of adjunctive ziprasidone in MDD derives exclusively from open-label,23 but not placebo-controlled, trials. However, a randomized, double-blind, placebo-controlled trial focusing on the use of ziprasidone as adjunctive treatment to the SSRIS escitalopram for escitalopram-resistant MDD is currently underway (NCT00633399).24

In addition, very little is known regarding the long-term efficacy of this therapeutic approach. Rapaport and colleagues25 examined patients with MDD who failed to experience sufficient symptom improvement following treatment with citalopram monotherapy, who then went on to receive adjunctive treatment with risperidone for a total of 4–6 weeks. Patients who experienced a significant improvement in depressive symptomatology following citalopram-risperidone treatment were then randomized, under double-blind conditions, to either continue to receive the combination of risperidone and citalopram or to continue with citalopram but to undergo a substitution of risperidone for placebo for a total of 24 weeks. Relapse rates between the two groups were not statistically significant (53.3% following continued treatment with risperidone and citalopram versus 54.6% following citalopram monotherapy). Of note, however, the unusually high rate of initial non-response to citalopram, 89%, followed by a high rate of remission with open label augmentation, 59%, suggest that the randomized population may have been enriched with placebo responders, resulting in the lack of “separation” in efficacy (differential relapse rates) between the two treatment groups (patients who continued on risperidone plus citalopram versus those who continued on risperidone plus placebo). Finally, much remains to be determined regarding the short- and long-term tolerability and safety of this treatment strategy in MDD populations, as well as how this treatment compares (both in terms of efficacy and tolerability) with other therapeutic strategies for TRD.


In light of the challenge which TRD poses to clinicians and patients, there is an urgent need to develop novel treatment strategies. From the evidence available to date, it appears that augmentation of antidepressants with atypical antipsychotics is effective in some cases of TRD, at least during the acute phase of treatment. However, the long-term efficacy, tolerability, and safety of this treatment has not yet been established. In particular, the risk of metabolic (weight gain and hyper- or dyslipidemia), endocrine (hyperprolactinemia), cardiac (QTc prolongation and arrhythmogenesis), and central nervous system (akathisia, parkinsonism, tardive dyskinesia, neuroleptic malignant syndrome) adverse events during treatment with these agents specifically in MDD populations needs to be better quantified. Further research in necessary to examine the efficacy, safety, and tolerability of augmentation of antidepressants with atypical antipsychotics for TRD. In addition, further research is also required examining how this treatment strategy compares, both in terms of efficacy as well as tolerability/safety, with other treatment strategies for TRD. PP



1.    Fava M, Rush AJ. Current status of augmentation and combination treatments for major depressive disorder: a literature review and a proposal for a novel approach to improve practice. Psychother Psychosom. 2006;75(3):139-153.
2.    Rush AJ, Trivedi MH, Wisniewski SR, et al. Acute and longer-term outcomes in depressed outpatients requiring one or several treatment steps: a STAR*D report. Am J Psychiatry. 2006;163(11):1905-1917.
3.    Papakostas GI. Augmentation of standard antidepressants with atypical antipsychotic agents for treatment-resistant major depressive disorder. Essent Psychopharmacol. 2005;6(4):209-220.
4.    Richelson E, Souder T. Binding of antipsychotic drugs to human brain receptors, focus on newer generation compounds. Life Science. 2000;24;68(1):29-39.
5.    Shapiro DA, Renock S, Arrington E, et al. Aripiprazole, a novel atypical antipsychotic drug with a unique and robust pharmacology. Neuropsychopharmacology. 2003;28(8):1400-1411.
6.    Schmidt AW, Lebel A, Howard HR Jr, Zom SH. Ziprasidone: a novel antipsychotic agent with a unique human receptor binding profile. Eur J Pharmacol. 2001;425(3):197-201.
7.    Burris KD, Molski TF, Xu C, et al. Aripiprazole, a novel antipsychotic, is a high-affinity partial agonist at human dopamine D2 receptors. J Pharmacol Exp Ther. 2002;302(1):381-389.
8.    Shelton RC, Tollefson GD, Tohen M, et al. A novel augmentation strategy for treating resistant major depression. Am J Psychiatry. 2001;158(1):131-134.
9.    Shelton RC, Williamson DJ, Corya SA, et al. Olanzapine/fluoxetine combination for treatment-resistant depression: a controlled study of SSRI and nortriptyline resistance. J Clin Psychiatry. 2005;66(10):1289-1297.
10.    Corya SA, Williamson D, Sanger TM, et al. A randomized, double-blind comparison of olanzapine/fluoxetine combination, olanzapine, fluoxetine, and venlafaxine in treatment-resistant depression. Depress Anxiety. 2006;23(6):364-372.
11.    Thase ME, Corya SA, Osuntokun O, et al. A randomized, double-blind comparison of olanzapine/fluoxetine combination, olanzapine, and fluoxetine in treatment-resistant major depressive disorder. J Clin Psychiatry. 2007;68(2):224-236.
12.    Mahmoud RA, Pandina GJ, Turkoz I, Kosik-Gonzalez C, Canuso CM, Kujawa MJ, Gharabawi-Garibaldi GM. Risperidone for treatment-refractory major depressive disorder: a randomized trial. Ann Intern Med. 2007;147(9):593-602.
13.    Keitner GI, Garlow SJ, Ryan CE, et al. Risperidone augmentation for patients with difficult-to-treat major depression. Paper presented at: the 159th Annual Meeting of the American Psychiatric Association; May 20-25, 2006; Toronto, Canada.
14.    McIntyre A, Gendron A, McIntyre A. Quetiapine adjunct to selective serotonin reuptake inhibitors or venlafaxine in patients with major depression, comorbid anxiety, and residual depressive symptoms: a randomized, placebo-controlled pilot study. Depress Anxiety. 2007;24(7):487-494.
15.    Khullar A, Chokka P, Fullerton D, McKenna S, Blackman A. Quetiapine as treatment of non-psychotic unipolar depression with residual symptoms: double blind, randomized, placebo controlled study. Paper presented at: the 159th Annual Meeting of the American Psychiatric Association; May 20-25, 2006; Toronto, Canada.
16.    Mattingly GW, Ilivicky HJ, Canale JP, Anderson RH. Quetiapine augmentation for treatment-resistant depression. Paper presented at: 159th Annual Meeting of the American Psychiatric Association; May 20-25, 2006; Toronto, Canada.
17.    Papakostas GI, Shelton RC, Smith J, Fava M. Augmentation of antidepressants with atypical antipsychotic medications for treatment-resistant major depressive disorder: a meta-analysis. J Clin Psychiatry. 2007;68(6):826-831.
18.    Fava M, Rush AJ, Wisniewski SR, et al. A comparison of mirtazapine and nortriptyline following two consecutive failed medication treatments for depressed outpatients: a STAR*D report. Am J Psychiatry. 2006;163(7):1161-1172.
19.    Nierenberg AA, Fava M, Trivedi MH, et al. A comparison of lithium and T(3) augmentation following two failed medication treatments for depression: a STAR*D report. Am J Psychiatry. 2006;163(9):1519-1530.
20.    McGrath PJ, Stewart JW, Fava M, et al. Tranylcypromine versus venlafaxine plus mirtazapine following three failed antidepressant medication trials for depression: a STAR*D report. Am J Psychiatry. 2006;163(9):1531-1541.
21.    Berman RM, Marcus RN, Swanink R, et al. The efficacy and safety of aripiprazole as adjunctive therapy in major depressive disorder: a multicenter, randomized, double-blind, placebo-controlled study. J Clin Psychiatry. 2007;68(6):843-853.
22.    Marcus RN, McQuade RD, Carson WH, Hennicken D, Fava M, Simon JS, Trivedi MH, Thase ME, Berman RM. The efficacy and safety of aripiprazole as adjunctive therapy in major depressive disorder: a second multicenter, randomized, double-blind, placebo-controlled study. J Clin Psychopharmacol. 2008;28(2):156-165.
23.    Papakostas GI, Petersen TJ, Nierenberg AA, et al. Ziprasidone augmentation of selective serotonin reuptake inhibitors (SSRIs) for SSRI-resistant major depressive disorder. J Clin Psychiatry. 2004b;65(2):217-221.
24. Available at: Accessed September 15, 2008.
25.    Rapaport MH, Gharabawi GM, Canuso CM, et al. Effects of risperidone augmentation in patients with treatment-resistant depression: Results of open-label treatment followed by double-blind continuation. Neuropsychopharmacology. 2006;31(11):2505-2513.



Needs Assessment: Despite numerous published efficacy studies, use of lithium augmentation for the treatment of refractory depression has declined markedly among clinicians. This paper critically reviews the literature in order to clarify the current role of this strategy in light of emerging trends and findings in managing treatment-resistant depressed patients.

Learning Objectives:
• Be familiar with the controlled and large-scale uncontrolled acute efficacy studies on lithium augmentation.
• Be aware of data concerning the effects of lithium augmentation on long-term outcome.
• Understand the strengths and limitations of the clinical database on lithium augmentation.
• Appreciate the role of lithium augmentation in contemporary practice.

Target Audience: Primary care physicians and psychiatrists.

CME Accreditation Statement: This activity has been planned and implemented in accordance with the Essentials and Standards of the Accreditation Council for Continuing Medical Education (ACCME) through the joint sponsorship of the Mount Sinai School of Medicine and MBL Communications, Inc. The Mount Sinai School of Medicine is accredited by the ACCME to provide continuing medical education for physicians.

Credit Designation: The Mount Sinai School of Medicine designates this educational activity for a maximum of 3 AMA PRA Category 1 Credit(s)TM. Physicians should only claim credit commensurate with the extent of their participation in the activity.

Faculty Disclosure Policy Statement: It is the policy of the Mount Sinai School of Medicine to ensure objectivity, balance, independence, transparency, and scientific rigor in all CME-sponsored educational activities. All faculty participating in the planning or implementation of a sponsored activity are expected to disclose to the audience any relevant financial relationships and to assist in resolving any conflict of interest that may arise from the relationship. Presenters must also make a meaningful disclosure to the audience of their discussions of unlabeled or unapproved drugs or devices. This information will be available as part of the course material.

This activity has been peer-reviewed and approved by James C.-Y. Chou, MD, associate professor of psychiatry at the Mount Sinai School of Medicine, and Norman Sussman, MD, editor of Primary Psychiatry and professor of psychiatry at New York University School of Medicine. Review Date: October 6, 2008.

Dr. Sussman reports no affiliation with or financial interest in any organization that may pose a conflict of interest. Dr. Chou receives honoraria from AstraZeneca, Bristol-Myers Squibb, Eli Lilly, Janssen, and Pfizer.

To receive credit for this activity: Read this article and the two CME-designated accompanying articles, reflect on the information presented, and then complete the CME posttest and evaluation. To obtain credits, you should score 70% or better. Early submission of this posttest is encouraged: please submit this posttest by November 1, 2010 to be eligible for credit. Release date: November 1, 2008. Termination date: November 30, 2010. The estimated time to complete all three articles and the posttest is 3 hours.


Primary Psychiatry. 2008;15(11):35-42


Dr. Price is professor of psychiatry and human behavior, Dr. Carpenter is associate professor of psychiatry and human behavior, and Dr. Tyrka is assistant professor of psychiatry and human behavior, all at the Warren Alpert Medical School of Brown University in Providence, Rhode Island. Dr. Price is director of research and clinical director, Dr. Carpenter is chief, and Dr. Tyrka is associate chief of the Mood Disorders Research program, all at Butler Hospital in Providence.

Disclosures: Dr. Price is consultant to Gerson Lehrman, Oxford University Press, Springer, and Wiley; serves on the speaker’s bureau for Jazz; and receives research support from Cyberonics, the National Institutes of Health (NIH), Sepracor, UCB Pharma, and the United States Department of Defense. Dr. Carpenter is consultant to AstraZeneca, Cyberonics, Novartis, and Wyeth; is on the speaker’s bureau for Cyberonics; and receives research support from Cyberonics, National Alliance for Research on Schizophrenia and Depression, the NIH, Sepracor, and UCB Pharma. Dr. Tyrka receives research support from Cyberonics, Sepracor, UCB Pharma, and the US Department of Defense.

Please direct all correspondence to: Lawrence H. Price, MD, Butler Hospital, 345 Blackstone Blvd, Providence, RI 02906; Tel:  401-455-6533; Fax: 401-455-6534;  E-mail:


Lithium augmentation, first described in 1981, is considered one of the best-supported strategies for treating refractory depression. However, clinical use of this approach has fallen dramatically in recent years, and some recent studies have cast doubt over its efficacy. This article reviews published findings in order to clarify the current role of lithium augmentation in treating refractory depression. Ten placebo-controlled studies, eight comparator-controlled studies, and 13 uncontrolled large-scale prospective studies of acute efficacy were reviewed in addition to six studies of effects on long-term outcome. Detailed examination found that controlled studies of lithium augmentation suffer from inadequate criteria for refractoriness, marked variability in duration of augmentation, variability in lithium levels, inadequate criteria for evaluating response, and idiosyncratic designs. Even more recent studies, while methodologically superior to earlier trials, have significant limitations, especially with respect to variability in lithium levels. “True” response rates to lithium augmentation are likely between 30% to 40%, rather than the 50% assumed by clinicians. A more balanced appraisal of the benefits and risks of this underutilized approach might encourage its wider use by clinicians.


Numerous groups in the 1970s studied the effects of combining lithium with tricyclic antidepressants (TCAs) or monoamine oxidase inhibitors (MAOIs),1,2 but broad interest in this approach developed after an open-label report by De Montigny and colleagues in 1981.3 The researchers described dramatic improvement in eight inpatients with unipolar depression within 48 hours of adding lithium 900 mg/day to an ongoing ineffective TCA regimen. Early investigators hypothesized that a specific pharmacodynamic interaction between the TCAs and lithium accounted for such effects, so that lithium acted to “augment” the effects of the TCA rather than exerting independent antidepressant effects of its own.4 This interpretation is complicated; lithium alone does appear to have some antidepressant efficacy.5 Lithium augmentation has come to mean the addition of lithium to a primary antidepressant in order to increase efficacy over what might be obtained with the primary antidepressant alone.6

Since De Montigny and colleagues’ initial report,3 1,263 patients have been described in 31 published prospective studies of lithium augmentation. Attention to the problem of treatment-resistant or refractory depression has expanded greatly, and articles on the topic have generally identified lithium augmentation as one of the best-substantiated pharmacologic approaches to treatment.7-10 Despite the extent of the supporting evidence base, several observers have noted that use of lithium augmentation in general psychiatric practice has declined markedly in recent years.7,9

This article reviews published findings on the efficacy of lithium augmentation in refractory depression, focusing on acute-controlled and large-scale prospective studies and long-term outcome studies. Based on this article, and in light of evolving trends and emerging findings, the current role of lithium augmentation in treating refractory depression should be reconsidered.

Efficacy in Acute Treatment

Placebo-controlled Studies

Ten studies, comprising a total of 269 patients, have evaluated lithium augmentation in the context of a placebo-controlled, double-blind design (Table 1).11-20 Six of these11,14,16-19 containing 194 patients supported the efficacy of the intervention; 4 studies12,13,15,20 with 75 patients did not. A recent meta-analysis21 of these studies demonstrated superiority of lithium augmentation over placebo. The cumulative response rates in these studies, omitting three reports in which categorical response under double-blind conditions was not clearly reported,11,14,17 are 42% for lithium augmentation and 20% for placebo. Including two reports in which single-blind categorical response was reported,14,17 cumulative response rates for lithium augmentation yielded 41% and 21% for placebo.


Nine studies reported complete diagnostic data on 194 of 208 patients who had unipolar illness; one study involving 61 patients included both unipolar and bipolar patients without providing a breakdown.18 All studies utilized validated criteria for depression as defined by either the Diagnostic and Statistical Manual of Mental Disorders, Third Edition (DSM-III)22 or its successors, or the Research Diagnostic Criteria. All studies provided data on the severity of depression at entry to lithium augmentation based on the Hamilton Rating Scale for Depression (HAM-D), with minimum scores ranging from >12 to >18; two studies19,20 required a <50% decrease from pre-antidepressant baseline but did not specify a minimum score. Primary antidepressant treatment in most of the studies involved TCAs; one study19 of 24 patients, used a selective serotonin reuptake inhibitor (SSRI) exclusively, while another with 61 patients, treated 50% of patients with an SSRI and 50% with a TCA.18 Only a few patients were treated with an MAOI or trazodone. Antidepressant doses (or plasma levels) were generally adequate when reported.

Duration of primary antidepressant treatment was ≥21 days in all studies, and ≥28 days in six studies collectively comprising 186 patients.13,15,16,18-20 Although some data support the use of 28 days as a benchmark for adequacy of antidepressant trial duration,23 several authorities recommend 4–8 weeks.24,25 Experts on refractory depression advise a duration of at least 6 weeks,26,27 and there is empirical support for trials of 8–10 weeks.28 Studies in Table 1 meeting a criterion of 6 weeks comprised only 112 of 269 patients,13,18,20 meaning that a large proportion of patients participating in these studies might not meet current criteria for refractoriness. However, use of the parallel-group design for the augmentation component of these trials would mitigate this limitation, since patients treated with placebo would be as likely to have a delayed response to the primary antidepressant as patients treated with lithium. These factors might lead to an underestimate of lithium augmentation’s efficacy. The same could not be said of open-label trials, in which enrollment of patients insufficiently treated with a primary antidepressant would be expected to inflate the apparent response to augmentation.

Assignment to lithium or placebo augmentation was randomized in all but two studies, one of which was alternating11 and in the other unreported.12 Duration of lithium augmentation was extremely variable, ranging from 2–42 days, with 123 patients in six studies11,12,14-16,19 treated for ≤14 days; 75 patients in four studies were treated for only ≤7 days.12,14,15,19 The reason many investigators used such short treatment durations derives from the initial understanding of lithium augmentation as a synergistic pharmacodynamic interaction between acutely administered lithium and chronic primary antidepressant treatment.4 In that conceptualization, the behavioral (ie, therapeutic) effects of adding lithium were expected to occur almost immediately, and early uncontrolled reports supported that expectation.4,29 Other investigators, conceptualizing lithium augmentation from a more clinical perspective, utilized longer trial durations more in keeping with standard pharmacologic approaches to depression. In their cumulative meta-analysis of the first nine studies in Table 1, Bauer and Döpfmer30 found that significant effects of lithium augmentation were first evident at the point of including studies lasting ≥7 days. While this is reassuring regarding the sensitivity of these short-term trials in detecting the positive effects of lithium augmentation, it does not mitigate the fact that such studies provide no information on the duration of any sustained benefit.

Dosing of lithium ranged from 250–1,200 mg/day; plasma levels reported in all but one study18 ranged from 0.3–1.1 mmol/L. In general, no relationship between dose or plasma level and clinical outcome was observed within studies. The exception to this was the study by Stein and Bernadt,17 which was specifically designed to evaluate this issue. These investigators found that low-dose lithium (250 mg/day; mean ± SD plasma level=0.3±0.1 mmole/L) was no better than placebo, whereas intermediate-dose lithium (750 mg/day; mean ± SD plasma level=0.8±0.5 mmole/L) was superior to the low dose. In their meta-analysis, Bauer and Döpfmer30 observed a significant effect of lithium augmentation beginning with studies dosing ≥600–800 mg/day.

The HAM-D was used to evaluate outcome in all but one study, which used the nurse-rated Short Clinical Rating Scale (SCRS).11 Most studies categorized response as a >50% decrease in the HAM-D. Exceptions to this were two studies in which a maximum final HAM-D score was required (ie, either <713 or <1,018). One was a study11 in which response criteria were not clearly specified, and the other was a study in which response criteria were not specified but a single responder was identified based on a >50% decrease in the HAM-D and final score <7.12 One study specified the criterion of a >50% decrease plus a final score of <10.16 Categorical response rates to double-blind treatment were available for all but three studies. In one study, the HAM-D was not used,11 and in the other two studies, the data were not reported, although single-blind categorical outcome was presented.14,17 Aggregate response was evaluated in all studies. Most of these studies would not meet current standards for evaluating response or remission.31

Comparator-controlled Studies

Eight studies, collectively comprising a total of 469 patients, have evaluated lithium augmentation in the context of a comparator-controlled design (Table 2).16,32-38 One of these studies included a placebo comparison group in addition to an active comparator16 and was included in the discussion of placebo-controlled studies mentioned above. The cumulative response rates in these studies are 30% for lithium augmentation and 38% for the active comparators. In general, these studies show no significant differences in response rates between lithium and the comparators used. However, in all but one study,33 lithium augmentation was numerically less effective than the comparator. In another study, lithium was significantly less effective than quetiapine augmentation based on continuous but not categorical measures of outcome.38


All but one patient in these studies were unipolar,32 and all studies used validated diagnostic criteria for depression. Most studies provided data on the severity of depression at entry to augmentation based on the HAM-D, with minimum scores ranging from >10 to >20. One study used an entry criterion of >5 on the Quick Inventory of Depressive Symptomatology-Clinician Rating.36 Three studies required a <50% decrease from pre-antidepressant baseline in addition to a minimum score.34,35,37 Primary antidepressant treatment involved a mix of TCAs, SSRIs, and other second-generation agents (eg, bupropion,36,38 mirtazapine,37,38 venlafaxine36-38). In contrast to the placebo-controlled studies, only 114 of 469 patients in these studies had primary treatment with a TCA.16,32,33 This reflects the fact that the comparator studies tend to be more recent than placebo-controlled studies, during which time usage of TCAs has declined.16,32,33 Primary antidepressant doses were adequate.

Duration (≥28 days) of primary antidepressant treatment in all studies was longer than that of placebo-controlled trials; 369 of  452 patients had ≥6 weeks of treatment,33,37 and 284 had ≥8 weeks.34-36 These durations meet stringent current criteria for trial adequacy.

Assignment to lithium augmentation or comparator was randomized in all of these studies, but assessment of outcome was conducted on a double-blind basis in only half.16,33-35 This constitutes a significant limitation of these studies in comparison with the placebo-controlled trials, all of which were double-blind. Comparators used included electroconvulsive therapy,32 switch to an alternative full-dose antidepressant (reversible MAO-A inhibitor brofaromine),33 dose escalation of the primary antidepressant (fluoxetine increased from 20 mg/day to 40–60 mg/day),34,35 and augmentation with an agent other than lithium (triiodothyronine,16,36 low-dose desipramine,34,35 full-dose lamotrigine,37 and quetiapine38).

Duration of lithium augmentation or comparator was ≤14 days in all studies, and in most studies, it was ≥4 weeks; two studies specified minimum durations of <4 weeks,16,32 while one study, which did not establish a minimum, reported a mean lithium augmentation duration of 9.3±5.4 weeks.36 This constitutes a major strength of these studies in comparison with the placebo-controlled trials, in 50% of which lithium augmentation lasted ≤14 days. Again, this reflects the more recent provenance of these studies and evolution in the management of refractory depression in the 25 years since the first reports on lithium augmentation.

Dosing of lithium, reported in all but one study,37 ranged from 300–1,200 mg/day while plasma levels ranged from 0.1–1.1 mmol/L. Three studies employed low doses of lithium (300–600 mg/day). The first had a range of levels between 0.1–0.8 mmol/L (mean=0.4±0.2),35 the second reported only mean level (0.2±0.1 mmol/L),34 and the third reported titration to an unspecified dose range with a mean level of 0.7 mmol/L (no SD reported).38 A fourth study reported mean dosage of 860±373 mg/day and a median level of 0.6 mmol/L.36 Based on the meta-analysis of Bauer and Döpfmer,30 efficacy of lithium augmentation might be attenuated at these low doses and levels, which appears to be the case in all four reports.

The HAM-D was used to evaluate outcome in all studies, and criteria for categorical response were notably more stringent than in the placebo-controlled studies. In those trials, the most common criterion for response was a >50% decrease in the HAM-D scores, but this was used as a criterion in 4 of the 8 comparator trials. It was the sole criterion in only one study.33 In one of the remaining three, it was combined with a maximum final HAM-D score of <10,16 while in the other two a maximum final HAM-D score of <7 was set to define a more stringent criterion of remission.37 All other comparator trials used the more stringent requirement of a maximum final HAM-D score to define response, either <734-36 or <10.32

The STAR*D study39 is the largest and most widely publicized study of refractory depression. The component of this study examining lithium augmentation36 included the largest sample by far (N=142) of any controlled or uncontrolled lithium augmentation trial reported in the literature. Major strengths of this study included a form of randomized assignment, stringent criteria for refractoriness and response, prolonged primary antidepressant and augmentation treatment duration, and an overall meticulous attention to standardization of methods, as well as the broad generalizability deriving from its design as an effectiveness trial. Major limitations, however, are the modest mean doses (860±373 mg/day) and median levels (0.6 mmol/L) of lithium achieved, coupled with significant findings of poor tolerability. These features raise questions as to the impact the open-label nature of this study, using non-expert clinicians, may have had on its outcome. In this sense, the generalizability of the study as an effectiveness trial may have come at the expense of a more accurate assessment of the efficacy of lithium augmentation as implemented by experts.

Uncontrolled Studies

There are scores of uncontrolled, open-label reports on lithium augmentation. In an effort to minimize the effects of ascertainment bias, reporting bias, and inter-study variability in this literature, the authors of this article identified studies meeting the following three criteria. First, the study had to be of prospective design. Second, the sample size had to be larger (N≥15). Last, a categorical assessment of outcome had to be conducted. Thirteen studies collectively comprising a total of 525 patients, met these criteria (Table 3).29,40-51 In keeping with current standards on outcome assessment,31 and in contrast to previous reviews,6 the authors of this article classified patients as “responders” only if they met the study’s criteria for response or remission; “partial responders” were classified as nonresponders.


The cumulative response rate in these reports is 49%; in trials restricted to unique clinical subgroups (ie, adolescent46 and geriatric43,47 patients) in which response rates to antidepressants might be lower or lithium is less tolerable, the cumulative response rate is 52%. This is substantially higher than the 42% response rate observed in the placebo-controlled trials and the 30% response rate in the comparator-controlled studies. Interestingly, response rates of approximately 50% for lithium augmentation are most commonly cited in reviews on the treatment of refractory depression.7,9,52

Effects on Long-term Outcome

Data concerning the impact of lithium augmentation on long-term outcome in depression are still limited. One study has examined the efficacy of continued lithium augmentation in preventing the return of depressive symptoms following an acute positive response. In a randomized, placebo-controlled, double-blind trial of 29 unipolar depressed patients, Bauer and colleagues53 found relapse rates of 47% with placebo, including two patients who became manic, and 0% with lithium during a 4-month continuation period following 8–10 weeks of acute treatment and stabilization. During a subsequent open-label, 6-month follow up after taper and discontinuation of all medications, five of 14 lithium- and two of eight placebo-augmented patients relapsed.54

Three studies, all in unipolar depressed geriatric patients, examined the effects of lithium discontinuation following successful augmentation.55 In the only randomized, double-blind trial involving 12 patients maintained on lithium augmentation for a mean 5.8 years, Hardy and colleagues56 observed relapse rates of 33% in the group randomized to placebo and 33% in those continuing on lithium over a 2-year period. Reynolds and colleagues57 reported a relapse rate of 46% over an unspecified follow-up period in 11 patients discontinued from a median 10 weeks of lithium augmentation, while Fahy and colleagues58 observed a relapse rate of 52% over a mean 19.5-month follow up in 21 patients tapered from maintenance lithium augmentation. In the study by Fahy and colleagues,58 likelihood of relapse correlated with duration of lithium maintenance.

Finally, Nierenberg and colleagues59 studied 66 depressed patients (60 unipolar, 6 bipolar) who had undergone an acute trial of lithium augmentation. In this retrospective study, 63% of acute responders had a good outcome during a mean 29-month follow-up period; there was only 40% of partial or non-responders.


This critical reevaluation of evidence supporting lithium augmentation reveals serious deficiencies in the current database. Taken together, these deficiencies are sufficiently fundamental and pervasive to limit the usefulness of formal meta-analytic approaches to literature review. The placebo-controlled trials suffer from inadequate criteria for refractoriness, marked variability in duration of augmentation, variability in lithium levels, inadequate criteria for evaluating response, and idiosyncratic designs. The comparator trials, as a group newer than the placebo-controlled studies, improve upon numerous methodologic flaws of the older work, albeit with even more variability in lithium levels. However, the lack of placebo controls in the comparator studies is a fundamental limitation that is not readily mitigated. The large-scale prospective uncontrolled trials suffer from the usual constraints of such studies, including an overly generous evaluation of responses to the intervention. Again, early enthusiasm for lithium augmentation must be considered in light of the looser methodologic practices, more modest expectations of benefit, and dearth of alternatives that characterized the field 20 years ago.

While the “true” response rate to lithium augmentation is still unclear, it is certainly less than the 50% assumed by many clinicians and closer to 30% to 40%. It is increasingly problematic to consider this approach a “gold standard” in treating refractory depression given emerging large-scale controlled trials of other interventions employing superior designs and methodologies.60,61 At the same time, its current underutilization by clinicians in the United States,62,63 is highly unfortunate, given its supporting database and the limited number of studies supporting newer approaches. In addition to the efficacy studies discussed here, that database includes extensive preclinical64 and clinical10 research on its mechanism of action, such research on basic mechanisms being fairly unique among augmentation strategies. Particularly promising are recent findings on potential genetic markers for response.65,66 Consideration of these factors might help promote a more balanced appraisal of the benefits and risks of this neglected approach. PP


1.    Lingjaerde O, Edlund AH, Gormsen CA, et al. The effects of lithium carbonate in combination with tricyclic antidepressants in endogenous depression. A double-blind, multicenter trial. Acta Psychiatr Scand. 1974;50(2):233-242.
2.    Himmelhoch JM, Detre T, Kupfer DJ, Swartzburg M, Byck R. Treatment of previously intractable depressions with tranylcypromine and lithium. J Nerv Ment Dis. 1972;155(3):216-220.
3.    De Montigny C, Grunberg F, Mayer A, Deschenes JP. Lithium induces rapid relief of depression in tricyclic antidepressant drug non-responders. Br J Psychiatry. 1981;138:252-256.
4.    De Montigny C. Enhancement of the 5-HT neurotransmission by antidepressant treatments. J Physiol (Paris). 1981;77(2-3):455-461.
5.    Souza FG, Goodwin GM. Lithium treatment and prophylaxis in unipolar depression: a meta-analysis. Br J Psychiatry. 1991;158:666-675.
6.    Price LH, Carpenter LL, Rasmussen SA. Drug combination strategies. In: Amsterdam JD, Hornig M, Nierenberg AA, eds. Treatment-Resistant Mood Disorders. Cambridge, United Kingdom: Cambridge University Press; 2001:194-222.
7.    Thase ME. Therapeutic alternatives for difficult-to-treat depression: a narrative review of the state of the evidence. CNS Spectr. 2004;9(11):808-816,818-821.
8.    Fleck MP, Horwath E. Pharmacologic management of difficult-to-treat depression in clinical practice. Psychiatr Serv. 2005;56(8):1005-1011.
9.    DeBattista C. Augmentation and combination strategies for depression. J Psychopharmacol. 2006;20(3 suppl):11-18.
10.    Bschor T, Bauer M. Efficacy and mechanisms of action of lithium augmentation in refractory major depression. Curr Pharm Des. 2006;12(23):2985-2992.
11.    Heninger GR, Charney DS, Sternberg DE. Lithium carbonate augmentation of antidepressant treatment. An effective prescription for treatment-refractory depression. Arch Gen Psychiatry. 1983;40(12):1335-1342.
12.    Kantor D, McNevin S, Leichner P, Harper D, Krenn M. The benefit of lithium carbonate adjunct in refractory depression–fact or fiction? Can J Psychiatry. 1986;31(5):416-418.
13.    Zusky PM, Biederman J, Rosenbaum JF, et al. Adjunct low dose lithium carbonate in treatment-resistant depression: a placebo-controlled study. J Clin Psychopharmacol. 1988;8(2):120-124.
14.    Schopf J, Baumann P, Lemarchand T, Rey M. Treatment of endogenous depressions resistant to tricyclic antidepressants or related drugs by lithium addition. Results of a placebo-controlled double-blind study. Pharmacopsychiatry. 1989;22(5):183-187.
15.    Browne M, Lapierre YD, Hrdina PD, Horn E. Lithium as an adjunct in the treatment of major depression. Int Clin Psychopharmacol. 1990;5(2):103-110.
16.    Joffe RT, Singer W, Levitt AJ, MacDonald C. A placebo-controlled comparison of lithium and triiodothyronine augmentation of tricyclic antidepressants in unipolar refractory depression. Arch Gen Psychiatry. 1993;50(5):387-393.
17.    Stein G, Bernadt M. Lithium augmentation therapy in tricyclic-resistant depression. A controlled trial using lithium in low and normal doses. Br J Psychiatry. 1993;162:634-640.
18.    Katona CL, Robertson MM, Abou-Saleh MT, et al. Placebo-controlled trial of lithium augmentation of fluoxetine and lofepramine. Int Clin Psychopharmacol. 1993;8(4):323.
19.    Baumann P, Nil R, Souche A, et al. A double-blind, placebo-controlled study of citalopram with and without lithium in the treatment of therapy-resistant depressive patients: a clinical, pharmacokinetic, and pharmacogenetic investigation. J Clin Psychopharmacol. 1996;16(4):307-314.
20.    Nierenberg AA, Papakostas GI, Petersen T, et al. Lithium augmentation of nortriptyline for subjects resistant to multiple antidepressants. J Clin Psychopharmacol. 2003;23(1):92-95.
21.    Crossley NA, Bauer M. Acceleration and augmentation of antidepressants with lithium for depressive disorders: two meta-analyses of randomized, placebo-controlled trials. J Clin Psychiatry. 2007;68(6):935-940.
22. Diagnostic and Statistical Manual of Mental Disorders. 3rd ed. Washington, DC: American Psychiatric Association; 1968.
23.    Nierenberg AA, McLean NE, Alpert JE, Worthington JJ, Rosenbaum JF, Fava M. Early nonresponse to fluoxetine as a predictor of poor 8-week outcome. Am J Psychiatry. 1995;152(10):1500-1503.
24.    Practice guideline for the treatment of patients with major depressive disorder (revision). American Psychiatric Association. Am J Psychiatry. 2000;157(4 suppl):1-45.
25.    Crismon ML, Trivedi M, Pigott TA, et al. The Texas Medication Algorithm Project: report of the Texas Consensus Conference Panel on Medication Treatment of Major Depressive Disorder. J Clin Psychiatry. 1999;60(3):142-156.
26.    Souery D, Amsterdam J, de Montigny C, et al. Treatment resistant depression: methodological overview and operational criteria. Eur Neuropsychopharmacol. 1999;9(1-2):83-91.
27.    Petersen T, Papakostas GI, Posternak MA, et al. Empirical testing of two models for staging antidepressant treatment resistance. J Clin Psychopharmacol. 2005;25(4):336-341.
28.    Quitkin FM, Petkova E, McGrath PJ, et al. When should a trial of fluoxetine for major depression be declared failed? Am J Psychiatry. 2003;160(4):734-740.
29.    de Montigny C, Cournoyer G, Morissette R, Langlois R, Caille G. Lithium carbonate addition in tricyclic antidepressant-resistant unipolar depression. Correlations with the neurobiologic actions of tricyclic antidepressant drugs and lithium ion on the serotonin system. Arch Gen Psychiatry. 1983;40(12):1327-1334.
30.    Bauer M, Dopfmer S. Lithium augmentation in treatment-resistant depression: meta-analysis of placebo-controlled studies. J Clin Psychopharmacol. 1999;19(5):427-434.
31.    Rush AJ, Kraemer HC, Sackeim HA, et al. Report by the ACNP Task Force on response and remission in major depressive disorder. Neuropsychopharmacology. 2006;31(9):1841-1853.
32.    Dinan TG, Barry S. A comparison of electroconvulsive therapy with a combined lithium and tricyclic combination among depressed tricyclic nonresponders. Acta Psychiatr Scand. 1989;80(1):97-100.
33.    Hoencamp E, Haffmans PM, Dijken WA, Hoogduin CA, Nolen WA, van Dyck R. Brofaromine versus lithium addition to maprotiline. A double-blind study in maprotiline refractory depressed outpatients. J Affect Disord. 1994;30(3):219-227.
34.    Fava M, Rosenbaum JF, McGrath PJ, Stewart JW, Amsterdam JD, Quitkin FM. Lithium and tricyclic augmentation of fluoxetine treatment for resistant major depression: a double-blind, controlled study. Am J Psychiatry. 1994;151(9):1372-1374.
35.    Fava M, Alpert J, Nierenberg A, et al. Double-blind study of high-dose fluoxetine versus lithium or desipramine augmentation of fluoxetine in partial responders and nonresponders to fluoxetine. J Clin Psychopharmacol. 2002;22(4):379-387.
36.    Nierenberg AA, Fava M, Trivedi MH, et al. A comparison of lithium and T(3) augmentation following two failed medication treatments for depression: a STAR*D report. Am J Psychiatry. 2006;163(9):1519-1530.
37.    Schindler F, Anghelescu IG. Lithium versus lamotrigine augmentation in treatment resistant unipolar depression: a randomized, open-label study. Int Clin Psychopharmacol. 2007;22(3):179-182.
38.    Doree JP, Des Rosiers J, Lew V, et al. Quetiapine augmentation of treatment-resistant depression: a comparison with lithium. Curr Med Res Opin. 2007;23(2):333-341.
39.    Fava M, Rush AJ, Trivedi MH, et al. Background and rationale for the sequenced treatment alternatives to relieve depression (STAR*D) study. Psychiatr Clin North Am. 2003;26(2):457-494.
40.    Price LH, Charney DS, Heninger GR. Variability of response to lithium augmentation in refractory depression. Am J Psychiatry. 1986;143(11):1387-1392.
41.    Thase ME, Kupfer DJ, Frank E, Jarrett DB. Treatment of imipramine-resistant recurrent depression: II. An open clinical trial of lithium augmentation. J Clin Psychiatry. 1989;50(11):413-417.
42.    Cowen PJ, McCance SL, Ware CJ, Cohen PR, Chalmers JS, Julier DL. Lithium in tricyclic-resistant depression. Correlation of increased brain 5-HT function with clinical outcome. Br J Psychiatry. 1991;159(X):341-346.
43.    Zimmer B, Rosen J, Thornton JE, Perel JM, Reynolds CF 3rd. Adjunctive lithium carbonate in nortriptyline-resistant elderly depressed patients. J Clin Psychopharmacol. 1991;11(4):254-256.
44.    Ontiveros A, Fontaine R, Elie R. Refractory depression: the addition of lithium to fluoxetine or desipramine. Acta Psychiatr Scand. 1991;83(3):188-192.
45.    Rybakowski J, Matkowski K. Adding lithium to antidepressant therapy: factors related to therapeutic potentiation. Eur Neuropsychopharmacol. 1992;2(2):161-165.
46.    Strober M, Freeman R, Rigali J, Schmidt S, Diamond R. The pharmacotherapy of depressive illness in adolescence: II. Effects of lithium augmentation in nonresponders to imipramine. J Am Acad Child Adolesc Psychiatry. 1992;31(1):16-20.
47.    Flint AJ, Rifat SL. A prospective study of lithium augmentation in antidepressant-resistant geriatric depression. J Clin Psychopharmacol. 1994;14(5):353-356.
48.    Sluzewska A, Sobieska M, Rybakowski JK. Changes in acute-phase proteins during lithium potentiation of antidepressants in refractory depression. Neuropsychobiology. 1997;35(3):123-127.
49.    Bruijn JA, Moleman P, Mulder PG, van den Broek WW. Comparison of 2 treatment strategies for depressed inpatients: imipramine and lithium addition or mirtazapine and lithium addition. J Clin Psychiatry. 1998;59(12):657-663.
50.    Hoencamp E, Haffmans J, Dijken WA, Huijbrechts IP. Lithium augmentation of venlafaxine: an open-label trial. J Clin Psychopharmacol. 2000;20(5):538-543.
51.    Birkenhager TK, van den Broek WW, Mulder PG, Bruijn JA, Moleman P. Comparison of two-phase treatment with imipramine or fluvoxamine, both followed by lithium addition, in inpatients with major depressive disorder. Am J Psychiatry. 2004;161(11):2060-2065.
52.    Carvalho AF, Cavalcante JL, Castelo MS, Lima MC. Augmentation strategies for treatment-resistant depression: a literature review. J Clin Pharm Ther. 2007;32(5):415-428.
53.    Bauer M, Bschor T, Kunz D, Berghofer A, Strohle A, Muller-Oerlinghausen B. Double-blind, placebo-controlled trial of the use of lithium to augment antidepressant medication in continuation treatment of unipolar major depression. Am J Psychiatry. 2000;157(9):1429-1435.
54.    Bschor T, Berghofer A, Strohle A, et al. How long should the lithium augmentation strategy be maintained? A 1-year follow-up of a placebo-controlled study in unipolar refractory major depression. J Clin Psychopharmacol. 2002;22(4):427-430.
55.    Ross J. Discontinuation of lithium augmentation in geriatric patients with unipolar depression: a systematic review. Can J Psychiatry. 2008;53(2):117-120.
56.    Hardy BG, Shulman KI, Zucchero C. Gradual discontinuation of lithium augmentation in elderly patients with unipolar depression. J Clin Psychopharmacol. 1997;17(1):22-26.
57.    Reynolds CF 3rd, Frank E, Perel JM, et al. High relapse rate after discontinuation of adjunctive medication for elderly patients with recurrent major depression. Am J Psychiatry. 1996;153(11):1418-1422.
58.    Fahy S, Lawlor BA. Discontinuation of lithium augmentation in an elderly cohort. Int J Geriatr Psychiatry. 2001;16(10):1004-1009.
59.    Nierenberg AA, Price LH, Charney DS, Heninger GR. After lithium augmentation: a retrospective follow-up of patients with antidepressant-refractory depression. J Affect Disord. 1990;18(3):167-175.
60.    Berman RM, Marcus RN, Swanink R, et al. The efficacy and safety of aripiprazole as adjunctive therapy in major depressive disorder: a multicenter, randomized, double-blind, placebo-controlled study. J Clin Psychiatry. 2007;68(6):843-853.
61.    Mahmoud RA, Pandina GJ, Turkoz I, et al. Risperidone for treatment-refractory major depressive disorder: a randomized trial. Ann Intern Med. 2007;147(9):593-602.
62.    Valenstein M, McCarthy JF, Austin KL, Greden JF, Young EA, Blow FC. What happened to lithium? Antidepressant augmentation in clinical settings. Am J Psychiatry. 2006;163(7):1219-1225.
63.    Fredman SJ, Fava M, Kienke AS, White CN, Nierenberg AA, Rosenbaum JF. Partial response, nonresponse, and relapse with selective serotonin reuptake inhibitors in major depression: a survey of current “next-step” practices. J Clin Psychiatry. 2000;61(6):403-408.
64.    Chenu F, Bourin M. Potentiation of antidepressant-like activity with lithium: mechanism involved. Curr Drug Targets. 2006;7(2):159-163.
65.    Adli M, Hollinde DL, Stamm T, et al. Response to lithium augmentation in depression is associated with the glycogen synthase kinase 3-beta -50T/C single nucleotide polymorphism. Biol Psychiatry. 2007;62(11):1295-1302.
66.    Stamm TJ, Adli M, Kirchheiner J, et al. Serotonin transporter gene and response to lithium augmentation in depression. Psychiatr Genet. 2008;18(2):92-97.



This interview took place on August 26, 2008, and was conducted by Norman Sussman, MD.


This interview is also available as an audio PsychCastTM at

Disclosure: Dr. Turk is a consultant to EMpi Recovery Sciences, Janssen, Ortho-McNeil, and Pricara; is on the advisory boards of Abbott, Eli Lilly, Ferring Pharmaceuticals, and Pfizer; receives grant support from the National Institute of Arthritis, Musculoskeletal and Skin Diseases; and is a special government employee of the United States Food and Drug Administration.


Dr. Turk is John and Emma Bonica Professor of Anesthesiology and Pain Research and director of the Fibromyalgia Research Program at the University of Washington in Seattle, Washington. He is also editor-in-chief of The Clinical Journal of Pain and co-director of the Initiative on the Methods, Measurement, and Pain Assessment in Clinical Trials. While his research has been funded by the National Center for Health Statistics, National Center for Medical Rehabilitation Research, and numerous private foundations, he has consistently received funding from the National Institutes of Health since 1977. Dr. Turk has published over 450 journal articles and chapters in erudite texts and written and edited 13 volumes, including The Pain Survival Guide: How to Reclaim Your Life.


What is fibromyalgia?

Fibromyalgia, most cases of which occur in women, is a chronic condition characterized by hyperalgesia (ie, hypersensitivity to pain), multiple comorbid symptoms, psychosocial distress, functional disability, and physical deconditioning. People with fibromyalgia tend to be inactive, resulting in loss of muscle strength, endurance, and flexibility.

Does physical activity help fibromyalgia patients avoid symptoms of the disorder?

Exercise is important for people with fibromyalgia. Many patients claim that engaging in physical activity makes them feel worse and, as a result, they withdraw from activity. It is difficult to have people engage in activity beneficial to them when they believe it is harmful. It is important for fibromyalgia patients to learn the difference between “hurtful” and “harmful.” Patients may think physical activity hurts them and ultimately worsens their condition. In reality, inactivity makes tasks difficult to accomplish. If patients slowly engage in the seemingly arduous activities, gradually building their physical strength, they eventually increase their endurance. Increasing strength and flexibility diminishes numerous problems related to physical activity. Patients should start activities at a low level; intensity should increase gradually. Successful increase and completion of activities should provide corrective feedback and reduce fears regarding the potential of activity to exacerbate symptoms. In addition, optimism should be conveyed and reassurance should be given at all times when working with fibromyalgia patients. They frequently question themselves and their condition in general, as doctors, physicians, and specialists often fail to provide a diagnosis; evidence of the disease does not appear in x-rays; and blood work is negative. Until they receive a diagnosis, patients feel anxious about the meaning and cause of their symptoms. However, while there are numerous treatments that can help patients manage symptoms and reduce the deconditioning of physical inactivity, there is no cure for fibromyalgia.

Has fibromyalgia been described in other terms before prior to its current terminology?

Yes. There has been a wide range of different diagnoses for fibromyalgia dating back to hundreds of years (ie, neurosthenia, psychogenic rheumatism). However, the official diagnosis of fibromyalgia was developed in 1990. Prior to that, it was called fibromyositis, but since no inflammatory process was discovered in the condition, fibromyositis became fibromyalgia. There was a multi-centered study1 sponsored by the American College of Rheumatology in which an official diagnosis of fibromyalgia and its specific criteria were outlined.

What are the signs and symptoms of fibromyalgia?

The diagnosis of fibromyalgia is determined by the presence of widespread pain, defined  as pain in three of the four body quadrants (above and below the waist, right and left side of the body) and axial, pain of at least 3 months duration, and the presence of pain reports following palpation of 11 or more 18 specific locations (muscle and muscle insertion)—tender points. The American College of Rheumatology study1 investigated many other locations but a specific set of 18 best discriminated between patients with fibromyalgia and rheumatoid arthritis and back pain.

How is fibromyalgia distinguished from other pain syndromes?
Examination of blood chemistry can help physicians to rule out different rheumatologic conditions that mimic signs of fibromyalgia. For example, there is some overlap between the presenting symptoms of fibromyalgia and rheumatoid arthritis. Examination of erythrocyte sedimentation levels can assist in the differential diagnosis. There is also a good deal of overlap between fibromyalgia and chronic fatigue syndrome. Approximately 70% of patients who meet diagnostic criteria for chronic fatigue syndrome meet the diagnostic criteria for fibromyalgia. The central features of fibromyalgia are widespread pain and fatigue. However, fibromyalgia patients also report a high prevalence of back pain, stiffness, irritable bowel syndrome, migraine headache, and temporomandibular disorders.

Colleagues and I use a 40-item symptom checklist with our fibromyalgia patients. The average number of symptoms the fibromyalgia patient population reports is 20, meaning these patients endure 20 other symptoms in addition to the pain already caused by the primary diagnosis. Approximately 74% of patients with fibromyalgia report that they experienced depressed mood and >70% indicate that they feel nervous and irritable.

What causes fibromyalgia in terms of etiology?

Patients with fibromyalgia are viewed as a homogeneous group of people, but it is likely that there are different factors contributing to their condition. For example, approximately 30% to 35% of people with fibromyalgia say their symptoms began after physical trauma (eg, car accident, surgery). Approximately 65% to 70% say they experienced flu-like symptoms one day, and after, their illness progressively worsened. Whether these indications belong to the same category is unknown, and there may be genetic differences instigating the condition that need to be investigated. Further, a disordered central processing of noxious sets of stimulation characterizes people with fibromyalgia, but the actual mechanisms causing the symptoms are currently under much investigation. There is some evidence2,3 suggesting that fibromyalgia could be an autonomic nervous system dysfunction or an endocrinologic problem, but to date the cause of fibromyalgia is essentially unidentified. It is only known that it likely occurs in a heterogeneous group of people and likely involves different mechanisms.

Has the bimodal impact of stress, fear, and depression on the physical symptoms of fibromyalgia been examined?

Yes. As with any chronic disease, fibromyalgia impacts different domains of people. However, there is debate concerning whether psychiatric problems are causal or a reaction to living with a chronic set of symptoms. Although the lifetime prevalence of psychiatric disorders is higher in people with fibromyalgia than it is within the general population, a significant minority have no prior history of any psychiatric diagnoses. Although there is no question that some people may have had a premorbid psychiatric history contributing to fibromyalgia, living with fibromyalgia or any chronic disease, for that matter, has significant impact on multiple areas of functioning. It is hardly surprising that there is a high prevalence of depressed mood accompanying the presence of distressing symptoms. When doctors and specialists cannot find the cause, they may think that the symptoms are fabriacted or magnified. The patient then feels even greater distress, become more isolated focusing on the symptoms. This withdrawal from activities and preoccupation with symptoms may exacerbate the presence of fibromyalgia with the myriad of accompanying symptoms.

Do people with fibromyalgia complain of tinnitus?

Tinnitus is not highly prevalent in fibromyalgia. I do not think it is any more prevalent than what exists in the general population. However, like fibromyalgia, tinnitus is a symptom that is hard to objectify.

What treatments are beneficial to people with fibromyalgia?

There are 57 pharmacologic agents and 56 non-pharmacologic treatments reported to have beneficial effects in patients with fibromyalgia. Pharmacologic treatments, including antidepressants, opioids, anticonvulsants, N-methyl-D-aspartate excitotoxic amino acid receptor antagonists, 5-HT3 receptor antagonists, muscle relaxants, sedatives, and non-steroidals, have been reported in the literature to have beneficial effects on some symptoms. In terms of non-pharmacologic treatments, everything from magnetized mattresses to electroconvulsive therapy has been reported as effective for some symptoms in some patients. People have different symptoms, indicating it is possible that the range of treatments has different effects on different symptoms. With pharmacologic treatment, approximately 30% to 40% of patients had an approximately 30% to 50% improvement in their symptoms, particularly pain.4 However, no patients reported the elimination of pain by any newly developed treatments.

Is that degree of improvement significant to improve quality of life?

There is a distinction between statistically significant and clinically meaningful. The studies discussed earlier are of statistical significance. In terms of pain, most of the literature is not specifically about fibromyalgia but about chronic pain conditions. Pain reduction of approximately 30% is viewed as being clinically meaningful to the patient. Improvement depends on the severity of one’s pain with which he or she starts. For example, if a patient’s rating of pain is a “9” on an 11-point scale (0–10), a greater percentage of pain reduction is required for patients to judge that the improvement is clinically meaningful than if the patient’s initial rating is a “4” on this scale. The results on physical and emotional functioning are not as good as those reported for pain. Pain may be relieved, but there is no guarantee that pain reduction leads to the same levels of improvement in physical function and emotional function.

Which drugs are officially approved for the treatment of fibromyalgia?

There are two medications—one antidepressant and one anticonvulsant—approved by the Food and Drug Administration, and there are more in various phases of seeking approval. Duloxetine is an antidepressant and pregabalin is an a2d ligand. Approximately 30% to 40% of patients will find approximately 30% to 50% improvement in pain with either drug. However, an improvement in emotional functioning or physical functioning may not be as substantial and there are side effects with either drug. This may be due to the short-term nature of the trials. Typically, these trials tend to be ≤6 months long, and it may take longer for functioning to improve as much as observed pain.

Does fibromyalgia ever spontaneously go away?

No. There have been community longitudinal studies suggesting it may diminish, especially as a person ages.5-7 The peak for symptoms of fibromyalgia is between 40–60 years of age, but even people with fibromyalgia >60 years of age tend to report that symptoms improve but  do not disappear.

Like those with neuropathic pain syndromes, do people learn to live with the pain associated with fibromyalgia?

It is probably more difficult to learn to live with neuropathic pain, as it can be more intense than the pain fibromyalgia patients experience. The characteristics of neuropathic pain are completely different than those of fibromyalgia. People with fibromyalgia have a whole range of other symptoms to manage. Neuropathic pain is more focal. However, when one has co-ocurring disorders with fibromyalgia, such as irritable bowel syndrome or headaches, it is difficult to function. Even with the available treatments, one can continue to have some set of these symptoms for the foreseeable future.

In addition to medications, is there rehabilitation for people with fibromyalgia?

Yes. I was on the panel of the American Pain Society’s development of an evidence-based, clinical practice guideline for fibromyalgia. Approximately 3 years ago, we reviewed all the available studies on the condition and concluded that the optimal treatment for people with fibromyalgia includes information, reassurance, an exercise program, cognitive-behavioral therapy, and most likely antidepressants or an anticonvulsant. The combination of these is probably the best treatment for fibromyalgia patients. My own  research has delved into rehabilitation approaches, and it appears that this combination is the best available option. Any of the treatments used as monotherapy would not be as nearly as effective as a combination.

Upon consultation, is there any way to know if someone will respond to treatment?

That area is getting a lot of attention right now in terms of pain in general as well as fibromyalgia specifically. Through the use of responder analysis we are trying to identify those patients who are most likely to respond to a certain combination of these treatments. People may respond differently depending on the kinds of symptoms they present, psychosocial issues, and social support.

Unlike other chronic pain conditions, numerous fibromyalgia patients report not receiving adequate support from friends and family, as they might be skeptical of the diagnosis. Another fairly substantial subgroup of these people tend to be individuals living on their own or in isolation; they do not receive much support in general. Approximately 8 females to 1 male seek treatment. Prior to the FDA’S approval of two drugs approximately 1.5 years ago, there was great skepticism about this condition and whether it was a psychiatric problem, symptom magnification, or an attempt to obtain attention. Hopefully, the approval of the two drugs and others in the pipeline will change this viewpoint. PP


1.    Wolfe F, Smythe HA, Yunus MB, et al. The American College of Rheumatology 1990 Criteria for the Classification of Fibromyalgia. Report of the Multicenter Criteria Committee. Arthritis Rheum. 1990;33(2):160-172.
2.    Adler GK, Geenen R. Hypothalamic-pituitary-adrenal and autonomic nervous system functioning in fibromyalgia. Rheum Dis Clin North Am. 2005;31(1):187-202.
3.    Pillemer S, Bradley LA, Crofford LJ, Moldofsky H, Chrousos GP. The neuroscience and endocrinology of fibromyalgia. Arthritis Rheum. 1997;40(11):1928-1939.
4.    Turk DC. Fibromyalgia: a patient-oriented perspective. In: Dworkin RH, Breitbart WS, eds. Psychosocial and Psychiatric Aspects of Pain: a Handbook for Health Care Providers. Seattle, WA: IASP Press, 2004;308-339.
5.    Felson DT, Goldenberg DL. The natural hisory of fibromyalgia. Arthritis Rheum. 1986;29(12):1522-1526.
6.    White KP, Nielson WR, Harth M, Ostbye T, Speechley M. Does the label “fibromyalgia” alter health status, functioning, and health service utilization? A prospective, within-group comparison in a community cohort of adults with chronic widespread pain. Arthritis Rheum. 2002;47(3):240-265.
7.    Wolfe F, Anderson J, Harkness D, et al. Health status and disease severity in fibromyalgia. Results of a six-center longitudinal study. Arthritis Rheum. 1997;40(9):1571-1579.



Print Friendly 

Psychotherapies and Other Psychosocial Interventions for Depression in Late Life: Innovation Through Hybridization

Gary J. Kennedy, MD



Controversy continues as to which form of psychotherapy should be used with each type of depression1, 2 and under what circumstances psychotherapy may be recommended without antidepressant medication.3 The evidence base supporting empirically validated therapies, while sufficient, is limited compared to that which supports antidepressants. Nonetheless, the effect size of 0.78 for psychosocial interventions in geriatric depression compared to placebo or no intervention contrasts favorably with antidepressant medication.4 More limited but nonetheless promising are reports of psychosocial interventions modeled after disease management programs or psychoeducation which are adjuncts to medication management but are not considered psychotherapy.5-7 This process of innovation through hybridization is also apparent with the incorporation of cognitive-behavioral techniques within interpersonal therapy (IPT); redesigned to treat complicated grief or prevent recurrent episodes of affective illness.8-11

The availability of trained therapists as well as patient preference may narrow the choice of interventions to a simple few.12,13 Yet, the informed practitioner should be able to provide direction when the evidence suggests that under certain circumstances the interventions are not therapeutically equivalent. The psychotherapies and psychosocial interventions which have promise for geriatric depression5-8,10,11,14-25 are summarized in the Table. What follows is a more detailed discussion of some of the emerging, empirically proven interventions. 

Interpersonal and Social Rhythm Therapy  

Social rhythm therapy arises from the social zeitgeber hypothesis, which states that instability or disruptions in personal relationships and life’s daily routine destabilize circadian rhythms, triggering episodes of affective illness in vulnerable individuals.26 When added to IPT, the social rhythms component seeks to prevent future problems in the areas of grief, interpersonal disputes, role transitions, and role deficits, by stabilizing both personal and interpersonal routines. Treatment focuses on the connection between mood and the quality and regularity of social roles and relationships. The patient is asked to complete the Social Rhythm Metric, which is designed to assess and then enhance the regularity of daily routines. The patient records the time that target activities occur daily for 1 week. These activities include meal times, bedtime, getting up, going to work, engaging in hobbies, or leisure time pursuits. Each activity is assigned a score with higher values indicating greater regularity. The patient and therapist then work on identifying and managing potential precipitants of rhythm disruptions. Efforts to resolve current interpersonal issues and prevent future problems in these areas are also emphasized. Sessions last 45 minutes and occur weekly until remission of the present episode of illness, then every other week for 3 months, and then monthly. Patients with bipolar I disorder who are able to improve the regularity of social rhythms experience a reduced likelihood of recurrence.10 A unique component designed for younger patients with bipolar disorder but applicable to seniors as well is the problem area of “grief for the loss of the healthy self.” Here, the patient is helped to mourn the loss of what life might have been were it not for the illness. Although designed to reduce recurrence rates in bipolar disorder, social rhythm therapy has particular appeal for older patients with recurrent depression whose social roles, routines, and networks have diminished due to retirement, bereavement, or physical illness.

Problem-Solving Therapy

Problem-solving therapy is directive and brief, usually spanning six sessions. Written lists are generated in the therapy session and homework is prescribed. The patient is asked to specify and prioritize problems whether they are thoughts, feelings of fatigue, anxiety, depression, or interpersonal difficulties. The patient and therapist assess the circumstances surrounding the problem. The therapist urges the patient to formulate a feasible approach as well as to triage problems with less probability of change to a lower priority. Once a problem is identified and its determinants are specified, a brief list of potential solutions is generated by the patient and therapist. For each solution, a set of pros and cons are drawn up, which serve to prioritize the solutions from most to least difficult to effect. Pessimism (“it has not worked before”) and lack of motivation (“I forgot to practice”) are addressed by taking the problem-solving approach. However, the focus is kept purposefully narrow on the initial problem the patient identified to minimize a diffusion of effort. The patient is also reminded that the number of sessions set aside may be specific to the identified problem and that additional problems may be beyond the scope of the present therapy. This is not to say that genuine problems do not arise during therapy or that they never displace the identified difficulty. However, the aim is for the patient and therapist to focus the effort on realistic solutions.

In randomized controlled trials of interventions for minor depression or dysthymia in primary care settings, neither problem-solving therapy nor paroxetine were superior to placebo in helping patients achieve remission.27 Furthermore, the majority of patients who had received medication or psychotherapy and experienced remission discontinued the intervention at the end of the trial.19 However, the response rate among the placebo group was greater than expected. When coupled with the spontaneous remission rates seen in patients with minor depression in primary care, it may well be that the added but nonspecific attention given the placebo group functioned as low-dose therapy. Moreover, because of its brevity and narrow focus, problem-solving therapy shows promise of usefulness in home care settings.28

Social Support Intervention

Oxman and colleagues19 studied treatment response and naturally occurring social support among seniors randomized to one of three treatments for depression. In the placebo group, patients expressing higher levels of perceived social support experienced decreases in subthreshold depression. Neither of the active treatment groups (paroxetine or problem-solving therapy) exhibited this phenomenon, suggesting the potential value of efforts to change perceived social support in milder depressions. In a prospective study of clinically depressed older patients, Bosworth and colleagues29 found that baseline perceived social support was as important as clinical and diagnostic variables in predicting depression remission. They concluded that interventions directed at social support were likely to improve rates of depression remission beyond that achieved with conventional interventions.

Social support is most often assessed in behavioral, cognitive, or structural components reflecting such variables as frequency or kind of contact, perception of adequacy of support, and size or symmetry of the supportive network. If support influences depression through perceived adequacy, then interventions should focus on the cognitive processes. If contact affects depression directly, then interventions would focus on manipulations to increase or improve contacts. Several studies suggest that perceived emotional support has a significant impact on physical and mental health.30-32 In contrast, instrumental support is either unrelated to or negatively associated with well being.33,34

Given the same level of actual support, some people will see particular relationships as more or less adequate than others.31 For patients with low perceived support, techniques from cognitive therapy of depression offer strategies to address their interpretive bias.31 Such techniques focus on changing distorted, unrealistic perceptions of supportive relationships. In addition, perceptions of support from specific relationships tend to be distinct from perceptions of support in general.35,36 The opportunity to change perceptions is greatest when there is an expressed need during a time of stress.37

 The Enhancing Recovery in Coronary Heart Disease (ENRICHD) study employed a psychosocial intervention to successfully increase social support and alleviate depression in older adults with a recent acute myocardial infarction.8 The ENRICHD social support intervention (SSI) utilizes behavioral, cognitive, and network intervention methods to improve perceived emotional support. The intervention has its origins in both cognitive-behavioral therapy14 and social cognitive theory38 in which psychosocial functioning is seen to emerge from the interplay of cognitive, behavioral, and environmental elements. In SSI terminology, the therapeutic focus highlights the patient’s behavioral repertoire, cognitive schema, and network interactions (especially among family). Among patients with chronic illness, psychosocial interventions are more likely to have positive effects on depression when they include family members.39

Thus, SSI seeks to socially activate the patient through active problem solving, alteration of counterproductive automatic thoughts, and enhanced coping skills. A major goal is the alteration of perceived emotional support through modification of the environmental, behavioral, and cognitive factors that lead to the perception of inadequate emotional support. Marital, family, and network interactions are included to identify modifiable attributes deemed most responsible for the participant’s perception of inadequate emotional support. The SSI intervention is tailored to the patient’s individual deficits in psychosocial functioning with the Social Networks in Adult Life Questionnaire. Counseling can then be mapped onto specific problems underlying the perception of inadequacy. The problem is conceptualized in terms of social behavior and beliefs so that the therapist can offer one or more specific work modules. For example, social isolation would be conceptualized as an environmental deficit requiring the social outreach and network development module. Those who have automatic thoughts that negate opportunities for supportive relationships (“They think they are better than me!”) would be offered the cognitive therapy module. Ineffective communication skills and passivity (“How do you introduce yourself to a stranger?”) would be targeted with the social communication and assertiveness module. The involvement of network members is a key therapeutic component requiring the identification of potential, but currently disengaged, sources of support and connecting them to the therapeutic process. However, when a network member attends the therapy session, the focus remains the participant’s adjustment.  

Psychotherapy for Complicated Grief

Bereavement is considered complicated grief when it evolves into a set of persistent cognitions and behaviors characterized by a sense of disbelief regarding the death, anger, bitterness over the loss, recurrent waves of painful yearning for the deceased, preoccupation with the lost loved one, and intrusive thoughts related to the death.9 Avoidance of an expanding array of situations and activities which remind the person of the loss reinforces the preoccupation and truncates opportunities to recover spontaneously. Moreover, treatments for bereavement-related depression are minimally beneficial for symptoms of complicated grief.22 Although it resembles posttraumatic stress disorder (PTSD), factor analyses demonstrate that the symptoms of complicated grief load separately from both depression and anxiety. Shear and colleagues9 describe a psychotherapy which achieved superior results compared to IPT for symptoms of complicated grief. While therapy for complicated grief is similar to IPT for the bereaved, it incorporates imaginal and in vivo exposure techniques that commonly work in PTSD patients. In the introductory phase, the therapist provides information to distinguish normal bereavement from complicated grief, including a model of adaptive coping which addresses not only adjustment to the loss but also restoration of a life’s satisfactions. The model portrays optimal bereavement as a process alternating between attention to the loss and restorative behavior and entails a focus on personal life goals.

In the middle phase, the PTSD-like character of complicated grief is addressed with symptom-specific techniques. These include recounting the circumstances of the death as well as exercises to confront avoided situations. The therapist asks the patient to do a “revisiting exercise” in which the “story” of the death is related with closed eyes. At critical junctures, the therapist will ask the patient to gauge the level of distress associated with various elements of the story. The exercise is tape recorded for patient playback between sessions. The experience of listening to themselves recount the event helps patients detach from the intensity of the experience and places them in the role of therapist. The process also evokes memories not previously expressed that can be examined in the sessions. When the patient is able to complete the home playback without being overwhelmed, the homework is complete.

To reduce the distress of yearning and preoccupation with the loss and to promote a sense of ongoing connection with the deceased, the patient is asked to carry on a mock conversation. The therapist asks the patient to close his or her eyes and begin speaking to the deceased as if the person could hear and respond. The therapist then directs the patient to play the role of the lost loved one and answer over a period of 10–20 minutes. This call-and-response exercise is facilitated by having the patient complete a set of memory questions focused on positive remembrances but also inviting reminiscence of the negative ones as well.  

For the restorative component, life goals are evoked from the perspective of “what would you want for yourself if the grief were not so intense?” The therapist then works with the patient to identify ways of recognizing that he or she is working toward these goals. Concrete, behaviorally operationalized plans are discussed and the therapist encourages the patient to put them into action. As in conventional IPT, role transitions and disputes are also addressed to reengage the patient in meaningful satisfying relationships. In the termination phase, the patient is reminded that intrusive thoughts and pangs of longing may briefly reappear in the future, triggered by anniversaries or intimate situations. However, they will not be so intense or disabling. The transient recurrence of symptoms does not mean the patient has failed. Rather, the return of symptoms overcome in the past only reinforces the value of the steps that were taken to cope with them. If the recurrence persists, a short course of therapy may be advised.

Family-focused Treatment for Recurrent Disorders

Miklowitz and colleagues11 describe a bipolar patient family-focused intervention which achieves superior medication adherence and reduced hospital readmission rates compared to crisis management or intensive individual therapy alone. Because the approach involves the patient, family, and practitioner in a collaborative effort at relapse prevention, it is also appealing for recurrent depression, particularly when delusions or suicidality have complicated the illness. At the outset, the therapist emphasizes that collaboration between the patient and family reinforces rather than sacrifices autonomy by preserving independence. An educational approach is used to combat stigma by informing the patient and family that depression (or mania) is part of an illness and is not a moral weakness. The therapist reinforces the importance of social stimulation, rewarding activities, physical activity, and regular sleep schedule, as well as medication, psychotherapy, and social support. For both the patient and family, the therapist provides training to enhance communication. This includes lessons in “active listening” that seeks clarification rather than closure, focused rather than global positive and negative feedback, framing requests for change in positive rather than negative language, and role playing assignments.

However, a significant amount of time is devoted to relapse prevention planning. In-session practice on brainstorming solves the problem of what to do when symptoms emerge. The therapist offers feedback but insures that solutions are selected by the patient and family. This includes devising and rehearsing a relapse prevention drill with a family-wide response plan. The plan will specify risk reduction for suicidal ideation if present, actions family and social supports should institute,  an established threshold for emergency communication and intervention, and emergency contact information and backup for times when a particular professional is not available (physician, therapist, emergency room, dialing 911). In sum, the goal of preserving independence is conceived as multilateral with responses to specified threats foreordained by the patient and family through the assistance of the therapist. Although risk may not be eliminated, procedures to reduce vulnerability offer both hope and support.  


There is increasing evidence that psychotherapeutic interventions from different traditions offer benefits for older persons with depressive disorders and complicated grief. The differing psychotherapies for late-life depression share common elements that are easily adapted across techniques. These include a problem-focused, here-and-now approach with distinct educational and social components. Homework may be assigned, faulty cognition confronted, and interpersonal changes suggested. For the motivated primary care practitioner, use of short-term psychotherapeutic techniques can enliven practice, lessen the need for referrals, and provide greater acceptance of mental health services for those patients who need specialty care. For mental health specialists accustomed to psychotherapy with younger patients, necessary adjustments in approach include allowances for sensory impairments and cognitive slowing, greater collaboration with the patient’s family and other care providers, and identification of improved function as well as symptom reduction as worthwhile goals. Although pharmacotherapy for depression has advanced greatly in the last decade, innovation through hybridization is leading to a similar change of pace in psychosocial interventions. PP


1.    Gallagher-Thompson DE, Thompson LW. Psychotherapy with older adults in theory and practice. In: Bonger B, Beutler L, eds. Comprehensive Textbook of Psychotherapy. New York, NY: Oxford University Press; 1995:357-379.

2.    Karasu TB. The specificity versus nonspecificity dilemma: toward identifying therapeutic change agents. Am J Psychiatry. 1986;143:687-695.

3.    Areán PA, Cook BL. Psychotherapy and combined psychotherapy/pharmacotherapy for late life depression. Biol Psychiatry. 2002;52:293-303.

4. Scogin F, McElreath L. Efficacy of psychosocial treatments for geriatric depression: a quantitative review. J Consult Clin Psychol. 1994;62:69-74.

5.    Kennedy GJ. Telephone-facilitated treatment of depression in primary care using the PHQ-9. Primary Psychiatry. 2004;11:18-21.

6.    Sirey AA, Bruce ML, Alexopoulos GS. The treatment initiation program: an intervention to improve depression outcomes in older adults. Am J Psychiatry. 2005;162:184-186.

7.    Rosen J, Rogers JC, Marin RS, Mulsant BH, Shahar A, Reynolds CF. Control-relevant intervention in the treatment of minor and major depression in a long-term, care facility. Am J Geriatr Psychiatry. 1997;5:247-257.

8.    The ENRICHD Investigators. Enhancing recovery in coronary heart disease (ENRICHD) study intervention: rationale and design. Psychosom Med. 2001;63,747-755.

9. Shear K, Frank E, Houck PR, Reynolds CF III. Treatment of complicated grief: a randomized controlled trial. JAMA. 2005;293:2601-2608.

10.    Frank E, Kupfer DJ, Thase ME, et al. Two year outcomes for interpersonal and social rhythm therapy in individuals with bipolar I disorder. Arch Gen Psychiatry. 2005;62:996-1004.

11. Miklowitz DJ, George EL, Richards JA, Simoneau TL, Suddath RL. A randomized study of family-focused psychoeducation and pharmacotherapy in the outpatient management of bipolar disorder. Arch Gen Psychiatry. 2003;60:904-912.

12.    Wei W, Sambamoorthi U, Olfson M, Walkup JT, Crystal S. Use of psychotherapy for depression in older adults. Am J Psychiatry. 2005;162:711-717.

13.    Glick ID. Adding psychotherapy to pharmacotherapy: Data, benefits, and guidelines for integration. Am J Psychother. 2004;58:186-208.

14.    Thompson LW, Gantz F, Florsheim M, et al. Cognitive-behavioral therapy for affective disorders in the elderly. In: Myers WA, ed. New Techniques in the Psychotherapy of Older Patients. Washington, DC: American Psychiatric Press; 1991:3-19.

15.    Miller MD, Cornes C, Frank E, et al.  Interpersonal psychotherapy for late-life depression: Past, present and future. J Psychother Pract Res. 2001;10:231-238.

16.    Lazarus LW, Sadavoy J. Individual psychotherapy. In: Sadavoy J, Lazarus LW, Jarvik, LF, Grossberg GT, eds. Comprehensive Review of Geriatric Psychiatry-II. 2nd ed. Washington, DC: American Psychiatric Press Inc; 1996:819-850.

17.    Butler RN, Lewis MI. Aging and Mental Health: Positive Psychosocial Approaches. St. Lewis, MO: CV Mosby; 1977.

18.    Fry PS. Structured and unstructured reminiscence training and depression among the elderly. Clin Gerontol. 1983;1:15-36.

19.    Oxman T, Barrett JE, Sengupta A, et al. Status of minor depression or dysthymia in primary care following a randomized controlled treatment. Gen Hosp Psychiatry. 2001;23:301-310.

20.    Areán, PA, Perri MG, Nezu AM, Schein RL, Christopher F, Joseph TX. Comparative effectiveness of social problem-solving therapy and reminiscence therapy as treatment for depression in older adults. J Consult Clin Psychol. 1993;61:1003-1010.

21.    Zarit SH, Orr NK, Zarit JM. The Hidden Victims of Alzheimer’s Disease. New York, NY: New York University Press; 1985.

22.    Reynolds CF III, Miller MD, Pasternak RE, et al. Treatment of bereavement-related major depressive episodes in later life: a randomized, double-blind, placebo controlled study of acute and continuation treatment with nortriptyline and interpersonal psychotherapy. Am J Psychiatry. 1999;156:202-208.

23.    Shear K. Traumatic Grief; Guidebook for Enhanced Services Providers. Pittsburgh, PA:University of Pittsburgh; 2003.

24.    Lewinshon PM. The behavioral study and treatment of depression. In: Hersen M, Eisler RM, Miller PM, eds. Progress in behavior modification. New York, NY: Academic Press; 1975:19-64.

25.    Lynch TR, Morse JQ, Menddelson T, Robins CJ. Dialectical behavior therapy for depressed older adults: A randomized pilot study. Am J Geriatr Psychiatry. 2003;11:33-45.

26.    Ehlers CL, Frank E, Kupfer DJ. Social zeitgebers and biological rhythms: a unified approach to understanding the etiology of depression. Arch Gen Psychiatry. 1988;45:948-952.

27. Frank E, Rucci P, Katon W, et al. Correlates of remission in primary care patients treated for minor depression. Gen Hosp Psychiatry. 2002;24:12-19.

28. Gum A, Areán PA. Current status of psychotherapy for mental disorders in the elderly. Curr Psychiatry Rep. 2004;6:32-28.   

29.    Bosworth HB, McQuoid MS, George LK, Steffens DC. Time-to-remission from geriatric depression.  Psychosocial and clinical factors. Am J Geriatr Psychiatry. 2002;10:551-559.

30.    Bolger N, Zuckerman A, Kessler RC. Invisible support and adjustment to stress.  J Pers Soc Psychol. 2000;79:953-961.

31.    Lakey B, Lutz CJ. Social support and preventive and therapeutic interventions. In: Pierce GR, Sarason BR, Sarason IG, eds. Handbook of Social Support and the Family. New York, NY: Plenum Press; 1996:435-465.

32.    Oxman TE, Hull JG. Social support, depression, and activities of daily living in older heart surgery patients.  J Gerontol B Psychol Sci Soc Sci. 1997;52:P1-P14.

33. Reinhardt JP, Boerner K, Horowitz A. Good to have but bad to use:  Differential impact of perceived and received and support on well-being.  J Soc Pers Relat. In press.

34.    Wethington E, Kessler RC. Perceived support, received support, and adjustment to stressful life events. J Health Soc Behav. 1986;27:78-89.

35.    Pierce GR, Sarason IG, Sarason BR. General and relationship-based perceptions of support: Are two constructs better than one?  J Pers Soc Psychol. 1991;61:1028-1039.

36. Sarason BR, Pierce GR, Sarason IG. Social support: the sense of acceptance and the role of relationships. In: Sarason BR, Sarason IG, Pierce GR, eds. Social support: An interactional view. New York, NY: Wiley; 1990:97-128.

37. Beach SRH, Fincham FD, Katz J, Bradbury TN. Social support in marriage: A cognitive perspective. In: Pierce GR, Sarason BR,  Sarason IG, eds. Handbook of Social Support and the Family. New York, NY: Plenum Press; 1996:43-64.

38.    Bandura A. Social Foundations of Thought and Action: A Social Cognitive Theory.  Englewood Cliffs, NJ: Prentice Hall; 1986.

39.    Martire LM, Lustig AP, Schulz R, Miller, GE, Helgeson VS. Is it beneficial to involve a family member? A meta-analysis of psychosocial intervention for chronic illness. Health Psychol. 2004;23:599-611.

Dr. Kennedy is professor of psychiatry and behavioral sciences at Albert Einstein College of Medicine, and director of the Division of Geriatric Psychiatry at Montefiore Medical Center in Bronx, NY.

Disclosure: Dr. Kennedy has received research support or honoraria from AstraZeneca, Eli Lilly inc., Forest, Janssen, and Pfizer Inc.

Please direct all correspondence to: Gary J. Kennedy, MD, Director, Department of Geriatric Psychiatry MMC, Dept. of Psychiatry, 111 East 210th Street, Klau One, Bronx, NY 10467; Tel: 718-920-4236; Fax: 718-920-6538; E-mail:



Print Friendly 

Remission of Depression: How Effective are Antidepressants?

Donald S. Robinson, MD



There exist longstanding concerns that antidepressant drugs produce modest clinical improvement, with only a minority of patients achieving full remission of their depressive episodes. Uncertainties about the therapeutic benefits of antidepressant treatment have been heightened by awareness of how difficult it is to demonstrate differences between active drugs and placebo in controlled clinical trials. This was evidenced by an analysis of data for several approved antidepressants submitted to the Food and Drug Administration over the past 2 decades.1-3

Doubts continue to be raised about the inability of antidepressants to fully relieve depressive symptoms and to induce remission of a major depressive episode.4-5 Data from controlled clinical trials infer that response rates with antidepressant treatment differ from placebo by only a small percentage. Recent media reports have emerged claiming that a majority of patients are insufficiently helped by antidepressants.6 Yet, the perception that antidepressants yield poor therapeutic response conflicts with clinical experience. Such experience suggests that most patients benefit from drug treatment when attention is paid to response to the drug and when antidepressant therapy is switched if improvement is insufficient.

Remission Rates in Placebo-Controlled Efficacy Trials

Consensus about what constitutes remission varies among experts. In placebo-controlled efficacy studies, a generally accepted definition of remission is achieving a score of ≤7 on the 17-item Hamilton Rating Scale for Depression (HAM-D), corresponding to minimal symptoms. Using this criterion for pharmaceutical industry efficacy trials, the data suggest remission rates of 35% to 45% for the newer antidepressants, such as venlafaxine and selective serotonin reuptake inhibitors.7 However, one should recognize that efficacy trials are typically short-term studies (6–8 weeks duration). In most short-term trials, group mean severity scores are still trending downward at the end of treatment, suggesting that additional patients would have achieved remission if therapy were continued longer. Consequently, most efficacy trials tend to underestimate the degree of improvement attributable to the study drug as compared with placebo. In addition, patients with an unsatisfactory response to a study drug might very well have achieved remission if switched to another antidepressant.

Remission rates in clinical practice settings are hard to quantify because well-controlled trials that allow longer treatment and switching of drugs for poor response necessitate complex study designs that are difficult to implement. However, a recent study attempted to address the question by examining eventual remission rates for patients who participated in a series of efficacy trials at Columbia University.8 Effectiveness was assessed in patients unimproved after initial treatment who were subsequently switched to a second or third drug using a therapeutic approach analogous to clinical practice.

Remission Rates in Effectiveness Studies

The investigators examined clinical remission in participants involved in three clinical trials of second-generation antidepressants conducted at the depression clinic at New York State Psychiatric Institute and Columbia University, and compared results with remission rates following a previously published study.9 Follow-up therapy after the initial study drug treatment differed among the three studies. In one study, nonresponders received up to two additional courses of the drug of the clinician’s choice; in two other studies, nonresponders were switched to a protocol-specified second antidepressant drug under blinded conditions, and then, if necessary, received a third course of antidepressant of the clinician’s choice. The investigators compared the results of these second-generation drug trials with aftercare of subjects involved in a previous trial of first-generation antidepressants (imipramine, phenelzine, or placebo).9 In this latter study, drug nonresponders received up to two additional courses of antidepressant treatment in order to seek remission.

In these clinical trials, patients were evaluated using the clinician-rated HAM-D and the self-rated Symptom Checklist-90 (SCL-90) scales. In the few instances where no final HAM-D or SCL-90 score was available to make a determination of remission, the clinical chart was reviewed. In such cases, remission only applied if the patient was described as being either euthymic or depression free.

Remission Rates with Second-Generation Antidepressants

A total of 171 patients entered the three second-generation antidepressant trials, and 50% (n=86) of the patients achieved remission during the first course of treatment (8–12 weeks); 73 patients were unimproved, and 44 of these received up to 12 weeks of treatment with a second antidepressant. Twenty-two of the latter group achieved remission. Of 20 unimproved patients following the second course of treatment, 14 were switched to a third drug, with 5 remitting. In total, 66% (n=113) of the 171 original patients ultimately achieved remission of their depressive episode. These findings show that in spite of being treated in a research setting where some constraints on clinical management exist, two-thirds of patients experienced remission of their depressive episodes. An additional 10% (n=17) improved significantly, although without achieving remission.

Remission Rate with First-Generation Antidepressants

These results were compared with clinical management of 420 patients following participation in a previously published double-blind trial of first-generation antidepressants (imipramine or phenelzine versus placebo).9 In this 6-week trial, drug nonresponders were switched to double-blind treatment with the other active drug (either imipramine or phenelzine) for 6 additional weeks, and placebo nonresponders were randomly assigned to either imipramine or phenelzine. If a third course of treatment was necessary, patients were treated open label with the antidepressant of the clinician’s choice for 6 additional weeks.

During initial double-blind treatment, 48% (n=205) of patients met remission criteria by 6 weeks. Of 167 nonresponders, 121 patients were switched to a second course of double-blind treatment, with 58 patients achieving remission. There were 24 nonresponders who underwent a third 6-week course of open-label treatment, with 12 patients remitting. Thus, out of 450 patients initially entering treatment for depression in this study, a total of 65% (n=275) ultimately remitted, while 15% (n=65) additional patients improved significantly but failed to remit.

The Importance of Continuing Treatment

The fact that 66% of patients ultimately achieved remission in a combined sample totaling more than 600 subjects involved in first- and second-generation antidepressant trials suggests that a majority of depressed patients will benefit if they remain in treatment. Since not all patients agreed to switch antidepressants and chose to drop out instead, 66% probably represents an underestimate of the potentially achievable rate of remission with drugs. In fact, in this study >90% of patients remaining in treatment for up to three courses of drugs remitted. It is likely that data in new drug applications inherently underestimate the effectiveness of antidepressants because industry-funded efficacy studies are predominately short-term trials that impose restrictions on dose and duration of drug treatment and do not permit switching therapy.

The overall favorable therapeutic outcome in these Columbia studies infers that the majority of depressed patients will remit if they receive appropriate antidepressant therapy. Good therapeutic management requires closely monitoring patients, with dose titration based on clinical response, and switching drugs if remission is not achieved with an adequate course of treatment. Motivating patients to continue treatment until they experience therapeutic benefit is a key aspect of optimal clinical management.


The perception that antidepressants produce low remission rates derives largely from results of short-term efficacy trials conducted by the pharmaceutical industry during the approval process. These studies tend to underestimate drug effectiveness and the remission rates potentially achievable in the practice setting. Effectiveness studies where patients are switched for reason of unsatisfactory response to another antidepressant drug indicate that the majority of patients will ultimately remit if they remain in treatment. Motivating depressed patients to remain in treatment is therefore a critical part of good clinical management. PP


1. Khan A, Warner HA, Brown WA. Symptom reduction and suicide risk in patients treated with placebo in antidepressant clinical trials: an analysis of the food and drug administration database. Arch Gen Psychiatry. 2000;57(4):311-317.

2. Robinson DS, Rickels K. Concerns about clinical drug trials. J Clin Psychopharmacol. 2000;20(6):593-596.

3. Khan A, Khan SR, Walens G, Kolts R, Giller EL. Frequency of positive studies among fixed and flexible dose antidepressant clinical trials: an analysis of the food and drug administration summary basis of approval reports. Neuropsychopharmacol. 2003;28(3):552-557.

4. Olfson M, Marcus SC, Druss B, Elinson L, Tanielian T, Pincus HA. National trends in the outpatient treatment of depression. JAMA. 2002;287(2):203-209.

5. Schneider LS, Nelson JC, Clary CM, et al. An 8-week multicenter, parallel-group, double-blind, placebo-controlled study of sertraline in elderly outpatients with major depression. Am J Psychiatry. 2003;160(7):1277-1285.

6. Goode E. Antidepressants lift clouds, but lose ‘miracle drug’ label. New York Times. June 30, 2002.

7. Thase ME, Entsuah AR, Rudolph RL. Remission rates during treatment with venlafaxine or selective serotonin reuptake inhibitors. Br J Psychiatry. 2001;178:234-241.

8. Quitkin FM, McGrath PJ, Stewart JW, et al. Remission rates with 3 consecutive antidepressant trials: effectiveness for depressed outpatients. J Clin Psychiatry. 2005;66(6):670-676.

9. Quitkin FM, Stewart JW, McGrath PJ, et al. Columbia atypical depression. A subgroup of depressives with better response to MAOI than to tricyclic antidepressants or placebo. Br J Psychiatry Suppl. 1993;(21):30-34.

Dr. Robinson is a consultant with Worldwide Drug Development in Melbourne, Florida.

Disclosure: Dr. Robinson is a consultant to Bristol-Myers Squibb, Genaissance, Organon, Somerset, and Takeda.



Dr. Kennedy is professor in the Department of Psychiatry and Behavioral Sciences at Albert Einstein College of Medicine, and director of the Division of Geriatric Psychiatry at Montefiore Medical Center in the Bronx, New York. Dr. Leon is a fellow in the Division of Geriatric Psychiatry at Montefiore Medical Center.

Disclosure: Dr. Kennedy is consultant to Myriad; on the speaker’s bureau of Pfizer; and has received grant support from Forest, Janssen, Myriad, Novartis, Pfizer, and Takeda. Dr. Leon reports no affiliation with or financial interest in any organization that may pose a conflict of interest.

Please direct all correspondence to: Gary J. Kennedy, MD, Director, Department of Geriatric Psychiatry, MMC, 111 East 210th St, Klau One, Bronx, NY 10467; Tel: 718-920-4236; Fax: 718-920-6538; E-mail:

Recent reports from the 2008 International Conference on Alzheimer’s Disease and numerous articles indicate substantial progress in the diagnosis and treatment of Alzheimer’s disease. However, the progress was partially a result of two well-designed anti-amyloid studies with disappointing results. Paradoxically, the effect is a heightened awareness of alternative therapeutic avenues that have emerged with substantial promise.


What follows is a selection of abstracted reports1 available online from the 2008 International Conference on Alzheimer’s Disease and two articles2 published within a month of the conference. This column is not meant to be a comprehensive or representative survey of data. Rather, it is an effort to highlight trends in current research. Though some outcomes were disappointing, the sheer volume of work on diagnosis and treatment promises substantial near-term progress.

Diagnostic Measures and Procedures

With the growing incidence of Alzheimer’s disease, the development of early detection methods identifying at-risk individuals is crucial. Recognizing affected individuals before their cognitive symptoms become evident allows for early intervention as well as further research and potential preservation of function with future disease-modifying therapies. Several studies reported advances in this area.

Protein in WBC

Brain cells in patients with Alzheimer’s disease have an abnormal tendency to enter the process of division and replication, making them more vulnerable to programmed cell death or apoptosis. This defect is also found in lymphocytes of patients with Alzheimer’s disease. In a study by Arendt and colleagues,3 the expression of a protein involved in white blood cell (WBC) production (CD-69) was measured on multiple cell lines of subjects with probable Alzheimer’s disease (n=32), healthy controls (n=30), and other dementias, namely Parkinson’s disease (n=26). CD-69 values showed variations in levels that allowed the researchers to differentiate between patients with Alzheimer’s disease (91% accuracy) and those with Parkinson’s disease (92% accuracy). The levels also distinguished people with Alzheimer’s disease from normal subjects 88% of the time when they had Alzheimer’s disease and 82% of the time when they had no cognitive deficits.

Brain Amyloid in Cerebrospinal Fluid

Accumulation of amyloid plaques in the brain is considered one of the primary causes of Alzheimer’s disease. Fagan and colleagues4 demonstrated an inverse relation between amounts of brain amyloid (according to positron emission tomography [PET] scans) and cerebrospinal fluid (CSF) levels of A42, which is the major constituent of amyloid plaques independently of patient’s cognitive status in a cohort of 132 patients 45–88 years of age. The group included individuals without cognitive deficits and those with very mild and mild dementia. Individuals with high amounts of brain amyloid according to PET scans had low levels of A42 in their CSF 97% of the time (n=37). People with low levels of brain amyloid had high CSF A42 in 84% of cases (n=95). This association of altered A42 dynamics between brain and CSF was found even in pre-clinical stages, which indicates that this is a very promising early detection marker.

Brain Enzyme in CSF

Beta-secretase (BACE1) is one of two enzymes involved in processing the amyloid precursor protein and producing toxic beta-amyloid. Hampel and Shen5 conducted a study that found higher levels of BACE1 activity in CSF of people with mild cognitive impairment (MCI) when compared to healthy controls and subjects with Alzheimer’s disease. Initially, the researchers measured CSF BACE1 in 80 people with Alzheimer’s disease, 59 people with MCI, and 69 healthy controls in two international centers. MCI subjects had a significantly higher level of BACE1 activity when compared to the other two groups. BACE1 activity was correlated with levels of b-amyloid. In a second part of the study, 47 MCI subjects were evaluated over the course of 2 years to assess BACE1 levels in combination with CSF tau and phosphorylated tau to determine the predictive value of these markers for determining conversion to Alzheimer’s disease. Fifteen MCI subjects converted to Alzheimer’s disease, and it was shown that BACE1 protein levels and the ApoE genotype were the strongest predictors of conversion to Alzheimer’s disease after controlling for age and gender (accuracy 78%, sensitivity 80%, and specificity 77% for the combination). A blood-based test for BACE1 is in process.

Amyloid Imaging Agents

PET scans create images of brain amyloid using a radioactive tracer that is injected into the patient. One of the challenges in widespread use of PET scanning is that the first amyloid tracer is short lived and, therefore, must be produced on-site. Of the compounds being studied, 18F-AV-45 promising based on a study of 42 cognitively healthy individuals and 39 patients with Alzheimer’s disease. It was favored because of its rapid uptake and ability tot maintain brain levels for 50–90 minutes post-injection. It is now being used for research, but it could  potentially be useful in a community setting.6

Anti-amyloid Agents

Because there are numerous recognized milestones along the path to amyloid deposition, an array of potential targets is available for intervention. Among several reports of amyloid-related studies, two stood out for their disappointing results despite rigorous methodologies. Bapineuzumab is a humanized monoclonal antibody engineered to reduce beta-amyloid in the brain. As such it would harness the natural immune processes to reduce the cause or entity near the cause of Alzheimer’s disease. However, amyloid in humans is produced in the absence of disease; provoking an immune reaction to amyloid is not without risk. Administration of an externally derived antibody or passive immunization to amyloid might prevent the adverse events resulting from active immunization.7 To study the safety and tolerability of bapineuzumab, 234 patients with mild-to-moderate Alzheimer’s disease were randomized to either placebo intravenous infusions or to one of four doses of bapineuzumab ranging from .15–1 mg/kg. The placebo group included 110 participants; the bapinezumab groups included roughly 30 participants each. Infusions of either placebo of bapienuzumab were scheduled every 13 weeks for a total of six treatments over 18 months.

Primary outcome measures specified prior to participant enrollment included the Alzheimer’s Disease Assessment Scale-Cognitive Subscale (ADAS-cog) and the Disability Assessment Scale for Dementia, neither of which exhibited statistically significant differences between bapineuzumab and placebo at 78 weeks. Post hoc analyses of participants who received all six infusions favored bapineuzumab group but were not significant. Post hoc analyses of participants free of the ApoE4 allele, which is associated with elevated risk of Alzheimer’s disease, showed significant benefits compared to placebo on the ADAS-cog as well as the Neuropsychological Test Battery and Clinical Dementia Rating scale and brain volume, as measured by magnetic resonance imaging (MRI). However, vasogenic edema of the brain was observed by routine MRI in 12 patients, all of whom received bapineuzumab. Eight  cases occurred in the highest dose of bapineuzumab. Six cases presented with clinical symptoms, one requiring steroids. In six of the 12 cases, re-dosing with bapineuzumab did not result in recurrence of vasogenic edema.

The study is noteworthy for its use of brain imaging and genetics as well as its multiple measures of cognitive performance and disability. Although benefits were limited to people not genetically predisposed to the disease, the study was powered with sufficient participants to detect small effect sizes. In addition, infusion therapy was burdensome and treatment- related brain edema was dose related. Without a measure of the dynamics of amyloid metabolism, it remains unclear whether altering the deposition of amyloid plaques or removing plaques once established would be the mechanism of therapeutic action. The study6 sponsors authorized a phase III trial of 4,100 participants in December 2007.

More disappointing than the bapineuzumab phase II study were results from the phase III 18-month trial8 of tarenflurbil. Tarenflurbil modulates g-secretase activity to selectively reduce beta-amyloid without interfering with other critical activities of the enzyme. A phase II study suggested tarenfluribil’s disease-modifying potential for people with mild Alzheimer’s disease. Green7 reported that 1,649 people with mean Mini-Mental Status Examination (MMSE) scores of 23.3 were randomized to either tarenflurbil 800 mg or placebo for 18 months. Participants were not excluded if they used cholinesterase inhibitors or memantine. The primary endpoints were the ADAS-cog and Alzheimer’s Disease Cooperative Study Activities of Daily Living Inventory (ADCS-ADL) scales, with the clinical dementia rating being the secondary endpoint, none of which showed significant differences between drug and placebo. The placebo group declined at the expected rate, making the possibility that placebo responders or sampling irregularities masked benefits unlikely. Given the length of observation and sample size, failure to detect a small but genuine effect due to the confounding influence of cholinesterase inhibitors or memantine was ruled out as well. The design contrasts with that of the beta-amyloid antagonist tramiprosate trial in which lack of demonstrated efficacy may have been the result of the confounding presence of cholinesterase inhibitors and memantine taken by the study participants.7 In addition, the study is the longest, largest placebo-controlled treatment trial of Alzheimer’s disease, regardless of the degree of severity.

A Cholinesterase-Inhibiting Antihistamine

Dimebon is a non-selective antihistamine with a weak capacity to inhibit butyryl- and acetylcholinesterase. It also weakly blocks the N-methyl-D-aspartate receptor-signaling pathway. In theory, these properties mimic both the Food and Drug Administration-approved cholinesterase inhibitors and memantine, which makes dimebon an appealing candidate for the treatment of Alzheimer’s disease. However, dimebon may exert its effects at the level of mitochondria to enhance neuronal function. It also inhibits neuronal death in models for Alzheimer’s disease. Doody and colleagues2 enrolled 183 patients, 155 of whom completed 26 weeks of either placebo or dimebon 20 mg three times daily. Participants had a mean age of 68 years, with an MMSE score of 18 and a mean of 5 years duration of dementia. Entry criteria included mild-to-moderate Alzheimer’s disease based on the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders9 and computed tomographies/MRIs of the brain consistent with Alzheimer’s disease from the National Institute of Neurological and Communicative Disorders and Stroke/Alzheimer’s Disease and Related Disorders Association. None of the participants had taken memantine or a cholinesterase inhibitor 6 months prior to entering the study.

The group-administered dimebon improved significantly from baseline to 26 weeks with a mean decrease in errors on the ADAS-cog of 1.9 (95% Confidence Interval [CI] -2.92 to -0.85; P=.0005). Compared to placebo, the dimebon group exhibited a mean decrease in ADAS-Cog errors of 4 (95% CI -5·73 to -2·28; P<.0001). Secondary outcomes also improved in association with dimebon. The MMSE score rose 1.8 points from 18.7 (.35) at baseline to 20.5 (.46; 95% CI 1.14 to 2.39; P<.0001). The ADCS-ADL score rose 1.3 points from 52.7 (1.32) to 54 (1.44; 95% CI -0.09 to 2.70; P=.024). The Neuropsychiatric Inventory (NPI) showed a decline in behavioral symptoms of 1 point from 11.8 (1.22) to 10.7 (1.33; 95% CI -2.62 to 0.56; P=.050). The Clinician’s Interview-based Impression of Change plus caregiver input (CIBIC-plus) also favored dimebon over placebo (0.6; 95% CI .92 to -0.31; P<.0001).

Dimebon showed statistically reliable improvements in the scientific measures of cognition required by the FDA (ADAS-cog), as well as dementia-related disability (ADCS-ADL), and behavioral disturbances (NPI). It also demonstrated benefits apparent to clinicians blind to the patient’s treatment or placebo status (CIBIC-plus) and on a cognitive screening exam in common clinical practice (MMSE). Dry mouth (14% vs. 1%) and depressed mood (14% vs. 5%) were more frequent among the dimebon than placebo group but were not associated with discontinuation of treatment. Although initially designed as a 26-week study the investigators extended the double-blind period of observation for a total of 52 weeks, ending with 61 individuals in the dimebon group and 59 in the placebo group. At 52 weeks the difference between drug and placebo on the ADAS-Cog scale expanded to -6.9 points (95% CI -9.43 to -4.28; P<.0001). The dimebon group’s cognitive performance remained above their starting point while the placebo group had continued to decline linearly. However the investigators do not suggest this is evidence of disease modification. Thus, dimebon shows modest efficacy with substantial safety. Additional studies will be necessary to confirm the results as well as test the safety and efficacy of dimebon in combination with other medications approved for the treatment of Alzheimer’s disease and other dementias.

Anti-Tau Agents

There have been two recent favorable reports on anti-tau agents. Schmechel and colleagues10 used AL-108, an experimental peptide derived from the activity-dependent neuroprotective protein. Incorporated into a nasal spray, AL-108 is designed to target neurofibrillary tangles, one of the early changes seen in MCI and Alzheimer’s disease. The double-blind, randomized, placebo-controlled study assessed safety, tolerability, and effects of 5 mg and 15 mg of AL-108 after 12 weeks of treatment. Subjects (n=144) included men and women 55–85 years of age with MMSE scores of ≥24 self-reported memory problems corroborated by a companion, and a Wecshler Memory Scale III age-adjusted Logical Memory score of ≤5. Cognitive testing was conducted at baseline as well as at 4, 8, 12, and 16 weeks. The results showed that the medication was well tolerated, as similar rates of adverse events (eg, headaches) were reported in placebo and AL-108 treated subjects. Twelve weeks of treatment showed significant dose-dependent and consistent improvement on various measures of memory. More specifically, at the higher dose of AL-108, there was statistically significant improvement in the delayed-match-to-sample test with a 34.2% change from baseline at 4 weeks (P=.067 compared to placebo) and a 62.4% change from baseline at week 16 (P=.038 compared to placebo). Digit span forward test was also significantly improved at week 8 with 11.2% change (P=.032) and maintained improvement at week 16 with 11.7% change from baseline (P=.052).

A second report11 is based on a study of methylthioninium chloride (MTC), a tau aggregation inhibitor that has been shown to dissolve tau tangle filaments and prevent aggregation in vitro. Wischik and colleagues11 conducted a 24-week, double-blind, randomized trial of MTC treatment in 321 people with Alzheimer’s disease at 17 centers in the United Kingdom and Singapore, followed by a 60-week, double-blind, treatment extension. Results showed that subjects with moderate impairment taking MTC had significant improvement on cognitive function after 24 weeks of treatment compared to placebo (-5.5 ADAS-cog units at the 60 mg dose with a P=.0208). MTC also showed to reduce the rate of cognitive decline (P=.0014) with no significant decline from baseline compared to non-MTC groups at the final 84-week analysis. Clinical findings were also correlated by imaging brain metabolism using single photon emission computed tomography and position emissoin tomograph, which showed that treatment with MTC at 60 mg eliminated the decline in cerebral blood flow seen in controls. These results are encouraging in terms of progress toward obtaining disease-modifying treatments.


Although the cerebral cascade of amyloidosis remains a prominent target for the modification of Alzheimer’s disease, its salience has been diminished first by the disappointing results with tramiprosate and more recently by the tarenflurbil and bapineuzumab trials. In contrast, the two reports with favorable outcomes of anti-tau medications are among the first to provide empirical evidence for those who suggest that amyloid may not be proximal to the cause of the disease. Subsequent studies support the preliminary findings, other tauopathies, such as fronto-temporal dementia, would come into therapeutic play. Enthusiasm for dimebon should be tempered by the lack of evidence of its safety when combined with the cholinesterase inhibitors and memantine. Long-term benefits and the possibility of disease modification await subsequent data.

The number of reports on promising diagnostic measures and procedures continues to grow. Yet, none of them approach the reliability and convenience of tests such as hemoglobin A1c for the diagnosis of diabetes.12 Nonetheless, the promise of early detection combined with disease modification promises a substantial reduction in the associated cost of both private and public dementia care. Recent reports further indicate a possible treatment that will prevent the crippling disability of the disease. They promise proof of Fries’s13 1980 hypothesis that morbidity can be compressed to the end of the life span, countering fears of an epidemic of disabled, dependent seniors. PP


1. ICAD. Alzheimer’s Association International Conference on Alzheimer’s Disease 2008. Available at: Accessed October 10, 2008.
2. Doody RS, Gavrilova SI, Sano M, et al. Effect of dimebon on cognition, activities of daily living, behaviour, and global function in patients with mild-to-moderate Alzheimer’s disease: a randomised, double-blind, placebo-controlled study. Lancet. 2008;372(9634):207-215.
3. Arendt T. Diagnosis of Alzheimer’s Disease using peripheral blood lymphocyte expression of CD-69 following a mitogenic stimulus. Available at: Accessed October 3, 2008.
4. Fagan AM. Update on the relationship between in vivo amyloid imaging with 11c-PIB and CSF A42. Available at: Accessed October 3, 2008.
5. Hampel H. Alteration of beta secretase (BACE1) functional candidate biomarkers in subjects with mild cognitive impairment and Alzheimer’s disease. Available at: Accessed October 3, 2008.
6. Pontecorvo MJ. Development of 18F-AV-45, a novel 18F-labeled A amyloid imaging agent. Available at: Accessed October 3, 2008.
7. Available at: http://alzforum/org/drc/detail.asp?id=1984. Accessed October 3, 2008.
8. Green RC. Safety and efficacy of tarenflurbil in subjects with mild Alzheimer’s disease: results from an 18-month multi-center phase 3 trial. Available at: Accessed October 3, 2008.
9. Diagnostic and Statistical Manul of Mental Disorders. 4th ed. Washington, DC: American Psychiatric Association; 1994.
10. Schmechel DE. A phase 2, double-blind, placebo-controlled study to evaluate the safety, tolerability, and effect on cognitive function of AL-108 after 12 weeks of intranasal administration in subjects with mild cognitive impairment. Available at:{5FB5340C-F07B-4DD5-BEC7-AEAE4DECBA86}&MKey={CFC5F7C6-CB6A-40C4-BC87-B30C9E64B1CC}&AKey={50E1744A-0C52-45B2-BF85-2A798BF24E02}&SKey={5075100B-928F-42CD-B0CD-19E3F395979F}. Accessed October 10, 2008.
11. Wischik CM. Tau aggregation inhibitor (TAI) therapy with remberTM arrests disease progression in mild and moderate Alzheimer’s disease over 50 weeks. Available at:{E7C717CF-8D73-41E0-8DB0-FA92205978CD}&MKey={CFC5F7C6-CB6A-40C4-BC87-B30C9E64B1CC}&AKey={50E1744A-0C52-45B2-BF85-2A798BF24E02}&SKey={68E04DB5-AB1C-4F7B-9511-DA3173F4F755}. Accessed October 3, 2008.
12. Consensus Committee. Consensus statement on the worldwide standardization of the hemoglobin A1C measurement: the American Diabetes Association, European Association for the Study of Diabetes, International Federation of Clinical Chemistry and Laboratory Medicine, and the International Diabetes Federation. Diabetes Care. 2007;30(9):2399-2400.
13. Fries JF. Aging, natural death, and the compression of morbidity. N Engl J Med. 1980;303(3):130-135.