Dr. Hollander is professor of psychiatry; director of the Compulsive, Impulsive, and Anxiety Disorders Program; director of clinical psychopharmacology; and director of the Seaver and New York Autism Center of Excellence at Mount Sinai School of Medicine in New York City. He has received a Research Scientist Development Award from the National Institute of Mental Health to investigate the psychobiology of obsessive-compulsive disorder and related disorders, such as body dysmorphic disorder and pathological gambling. His other research interests include autism, neuropsychiatry, and functional imaging.

This interview took place on February 7, 2005 in New York City.


As a foremost researcher and clinician on obsessive-compulsive disorder (OCD), how would you define or describe OCD? 

OCD is currently classified as an anxiety disorder and is characterized by obsessional thoughts and/or compulsive rituals that cause distress or interfere with functioning. Obsessions can be thoughts, impulses, or images that cause anxiety and are intrusive. These anxiety-provoking thoughts frequently cause compulsive rituals, which are designed to either neutralize the thoughts or reduce the anxiety.

While there are a broad range of symptoms, these can be arranged into four subcategories or subfactors. For example, people with OCD can be classified as washers and checkers; as individuals who have issues concerning symmetry (ie, they want things to be perfect, even, or “just so”); individuals with intrusive disturbing obsessions with sexual, aggressive, or religious content; or hoarders.

Doubt and uncertainty are characteristic across the different subtypes. The patients feel that if they do not give in to their habits or rituals, something terrible will happen, and as a result of that they request a lot of reassurance. Clearly people who are checkers have a lot of issues about doubt and uncertainty and ask for reassurance over and over again. They can, for example, stare at the lock on the door or stare at the oven and never get the sense that the lock is really locked or the oven is really off.


Could it be argued that a bit of obsessive-compulsiveness is actually a good thing?

These kinds of obsessional thoughts, rituals, and routines are relatively common in the general population. They can be a good thing, because they help us to stay organized, think ahead, and seek information. From a survival standpoint, it makes sense that these symptoms persist within the general population, as they may be helpful for us as a species.  They only become a disorder when they start to become time-consuming, cause distress, or interfere with functioning.

OCD symptoms can, for example, be beneficial in some ways to pregnant women. There is often a peak of OCD symptoms in women during childbirth, pregnancy, and breast feeding. Perhaps one of the things that is going on here is that there is a release of the peptide oxytocin, which is involved in uterine contractions, milk letdown, and bonding of the mother to the infant. This peptide may play a role in causing the mother to scan the environment to look for aggressors or for contamination that could harm the infant, which results in OCD symptoms.


Sometimes the only evidence of OCD that patients present with is something as vague as procrastination, trouble making simple decisions, or trouble throwing things away. But the Diagnostic and Statistical Manual of Mental Disorders (DSM) does not really reflect procrastination and doubting. How would you change the DSM? Would you write it differently than it exists right now?

First of all, there may be big changes coming down the road with regard to the DSM and OCD. I co-chair the American Psychiatric Association (APA)/National Institute of Mental Health Research Planning Agenda for the fifth edition of the DSM (DSM-V), along with Joseph Zohar, MD, of the Chaim Sheba Medical Center in Tel Hashomer, Israel, which is currently examining whether OCD should be taken out of the anxiety disorders and clustered with a group of related or spectrum conditions. OCD in some respects differs from the other anxiety disorders in terms of phenomenology, brain circuitry, family history, and treatment response. Instead, it shares features of basic etiology, brain circuitry, and genetics with a group of other related or OCD spectrum disorders. These may include Tourette’s syndrome; body dysmorphic disorder; autism and the developmental disorders; eating disorders, including binge-eating disorder; Huntington’s disorder and Parkinson’s disorder; pediatric autoimmune neuropsychiatric disorders associated with streptococcal infections (PANDAS) or Sydenham’s chorea; some of the impulse control disorders; some of the newly emerging compulsive and impulsive disorders; and obsessive-compulsive personality disorder. At issue for the DSM-V is also whether the hoarders that are currently considered a subtype of OCD should be thought of as distinct from OCD and placed into one of the obsessive-compulsive spectrum disorders.

Currently, the OCD spectrum disorders are grouped into three basic clusters; the first cluster being those disorders where doubt and uncertainty are coupled with a heightened sense that something bad may happen. The focus is on body image or body size, body sensations, body weight, and compulsively-driven disorders around those body-related issues.

The second cluster are the more impulsive disorders in which people initiate certain behaviors to obtain some element of pleasure, arousal, or gratification; however, over time a compulsively-driven component keeps these behaviors going. Within the impulse-control disorders, there are two groups: the first consists of the impulse-control disorders that currently exist in the DSM, including kleptomania, pyromania, intermittent explosive disorder, pathological gambling, and trichotillomania. In addition to that, there will probably be four newer disorders in the DSM-V, which we have started to call the impulsive-compulsive disorders. These are impulsive-compulsive shopping, impulsive-compulsive sexual behavior, impulsive-compulsive Internet use, and impulsive-compulsive psychogenic excoriation or skin picking. These behaviors have some element of pleasure, arousal, or gratification in the beginning, but a compulsively driven mechanism later causes the behavior to persist.

For example, we just finished a study on Internet addiction using escitalopram as a treatment, which we have not fully analyzed. These people get “stuck” on the Internet searching for information (ie, they constantly surf the Web with a “need-to-know” feeling of incompleteness) or they get stuck doing all kinds of impulsive things on the Internet, such as gambling, looking at sexual material, shopping, and trading stocks.

The third and final cluster of OCD spectrum disorders are the neurologically-based disorders that frequently involve some impairment of basal ganglia function, where one of the symptom domains includes obsessive-compulsive symptoms. Again, this obsessive-compulsive spectrum disorder is made up of a different symptom domain from the others, but it shares repetitive intrusive thoughts and compulsive or repetitive behaviors with them and may be characterized by autism or Tourette’s syndrome.


Have any of your theories been modified since you introduced the OCD spectrum idea?

In 1993, I published a book called Obsessive-Compulsive Related Disorders, and it was essentially the first systematic examination of the group of OCD spectrum or related disorders that shared features based on clinical phenomenology, comorbidity, family history, and underlying neurobiology. It also examined how these different conditions share a similar treatment response. If we understand what drives repetitive thoughts and behaviors in OCD, we can apply that knowledge to related or spectrum disorders and then develop specific treatments for the repetitive thoughts and behaviors across the OCD spectrum.

This idea has undergone refinement over the last 12 years; to some extent, this area is more complex than we originally thought. At first we thought that this whole area could be viewed from a dimensional standpoint: there are some disorders that are risk-aversive and harm-avoidant, such as OCD, and other disorders that are much more impulsive or novelty-seeking, such as the impulse-control disorders or the Cluster B personality disorders, including borderline personality disorder. Now we see that a lot of these disorders have both compulsive and impulsive features.

The four impulsive-compulsive disorders, for example (shopping, sexual behavior, Internet use, and skin picking) have compulsive and impulsive features, and are relatively complex. They can have features of both compulsively-driven behaviors to reduce anxiety, but also impulsive behaviors driven to obtain pleasure, arousal, or gratification. Patients with autism who have many rigid routines and rituals coupled with impulsive-aggressive behavior can also be included in this group, as can patients with Tourette’s syndrome who have compulsive rituals and impulsive-aggressive behavior. So these are not necessarily unitary disorders but are disorders made up by different symptom domains, one of them being repetitive thoughts and behaviors and others being, for example, inattention coupled with motor hyperactivity or affective instability. These associated symptom domains or comorbid conditions can influence phenomenology or how these behaviors are expressed and can certainly influence treatment response. The treatment response then becomes much more complex as well. We have to understand how repetitive thoughts and behaviors interact with these other symptom domains or other comorbid conditions.


On the subject of treatment: the selective serotonin reuptake inhibitors (SSRIs) and clomipramine (which is a serotonergic-like tricyclic antidepressant) are the only drugs that are indicated for OCD. Are they in fact the only ones that work?

Not necessarily, but for the drugs that have specific Food and Drug Administration indications (eg, clomipramine, fluvoxamine, sertraline, paroxetine, fluoxetine), there is a large database available of both efficacy and safety. All of the SSRIs seem to be effective and the serotonin norepinephrine reuptake inhibitors (SNRIs) seem to be effective as well. That would include clomipramine, although there is some evidence that venlafaxine, particularly in higher doses, is also effective for OCD. I would think that duloxetine may be effective also, but there have not been any kind of systematic studies on duloxetine done to date.

Interestingly, citalopram and escitalopram do not have FDA indications for OCD. That does not mean that they are not effective (data with citalopram shows efficacy); it only means that the companies behind them have not pursued an indication for OCD.


All of the drugs you mention have serotonergic side effects (eg, sexual dysfunction, weight gain). What do you do when you have a patient who is absolutely intolerant of serotonergic drugs? Are there other treatments that work that are not either SNRIs or SSRIs?

One of the things I am involved in now is the APA practice guideline for OCD, along with Lorrin Koran, MD, of Stanford University Medical Center in Stanford, California, who chairs the committee. This involves a systematic review of the literature to determine best practices. In areas where there is not a large database, expert consensus can be utilized to rank different treatments in terms of what would be recommended.

Clearly, clomipramine and the SSRIs have the largest database and the strongest support. The good news with these drugs is that 40% to 60% of patients have a good response to these agents. The bad news is that 40% to 60% of the patients are only partial or nonresponders.

Even when people respond, remission is not always the case, so there is a large universe of patients with OCD and obsessive-compulsive spectrum conditions that are partial or nonresponders. The field as a whole needs to address these individuals, because many of these OCD patients who are nonresponders function at very low levels. In fact, their level of functioning can be approximately equal to the level of patients with schizophrenia. And if there are 2% to 3% of the United States population with OCD and a much larger percentage of patients who have obsessive-compulsive spectrum conditions, and overall 40% to 60% of these are partial or nonresponders, that is a large number of patients who need alternate treatment.

If a patient is unable to tolerate clomipramine and the SSRIs due to side effects, there are a couple of different approaches. One approach is the intravenous (IV) administration of clomipramine or citalopram: when these drugs are given in an IV form there are actually fewer side effects. This IV administration is not available in a clinical setting in the US. (It is available in a research setting and is more widely available in Europe.) It does seem that IV administration of some of these medications may be associated with fewer side effects.

As far as what is available in the US, now we are in the realm of more novel treatments where there has been less systematic study. For example, Dr. Koran has been interested in the use of opiate agonists and has done some studies giving morphine sulfate 30 mg once or twice weekly. He found that there was some benefit in a group of patients who were SSRI nonresponders or who could not tolerate SSRIs. The database there is relatively small though, so it cannot be recommended as a standard treatment at this point.

Some people actually may respond to other types of agents, such as mood-stabilizing agents and/or stimulants. There has been some study, for example, of anti-anxiety medicines. It is clear that benzodiazepines as monotherapy do not seem to be effective for OCD, although sometimes they may be helpful as an augmentation or add-on. Likewise, buspirone as a monotherapy does not seem to be helpful. It is more helpful in higher doses as an add-on. There is a little bit of data on inositol as a monotherapy, which can be of some benefit in some patients, but again that needs further study. Inositol is a natural supplement that works on a phosphatidyl-inositol pathway.

Stimulants may be helpful, especially in light of the association between attention-deficit/hyperactivity disorder (ADHD) and OCD. This is frequently found in boys who have an onset of OCD before puberty and almost invariably have a triad of symptoms: they either have the symmetry/perfectionism/just-so/evening-up subtype or the disturbing, intrusive, sexual, aggressive, or religious obsession subtype. They frequently have tic-related symptoms when they have ADHD. These individuals are the OCD patients who have high rates of oppositional-defiant or conduct disorder upon follow-up; as they get older, they have high rates of impulse-control disorders and cluster B personality disorders.

Jeffrey M. Halperin, PhD, at the Mount Sinai School of Medicine in New York City has done an interesting follow-up study of kids with early ADHD in New York. He found that 60% of these individuals would meet criteria for Cluster B personality disorder as young adults. Earlier, we had found that individuals with OCD were four times more likely to have comorbid antisocial personality disorder. So the subtype of boys with early-onset OCD with comorbid ADHD may tend to have more impulsive and compulsive features. Those individuals may benefit from the use of stimulant-type medicines to address the ADHD component.

One of the advantages of stimulants (or SNRIs) is that they increase the ability for sustained attention, which allows people with OCD to redirect their attention into other meaningful activities or meaningful relationships. One problem with OCD patients is that the more they avoid and do nothing, the more the OCD symptoms fill up their free time. The more interest, initiative, and motivation they have to redirect their attention and get involved in meaningful activities, the better they do.

There have also been studies with topiramate in Tourette’s syndrome. People with Tourette’s syndrome have both compulsive rituals but also a lot of impulsive aggression. There is a multicenter study looking at topiramate as an augmentation strategy in treatment-resistant OCD. Topiramate is a mood-stabilizing anticonvulsant that has some effects not only on g-aminobutyric acid, but also on glutamate; it has been shown to be helpful in some of these impulses and with problems such as alcohol abuse and binge-eating disorder. Another multicenter study I am conducting is looking at topiramate as a treatment for compulsive gambling; the idea is that this is an agent that may decrease thoughts and urges, and which could also be of some benefit for the thoughts and urges in compulsive gambling.


Do you think OCD still remains underdiagnosed or misdiagnosed in general medical practice?

I certainly do think so, and I think one of the reasons OCD is underdiagnosed is that many patients think that these thoughts are bizarre or crazy and try to resist having them; as a result, they experience a lot of shame and humiliation, and they may not tell other people about the thoughts. They may not tell their mental health practitioners either. So I do think that clinicians really need to screen for these things systematically. Once clinicians start to ask routine screening questions, they will pick up on OCD more frequently. OCD is often comorbid with other conditions, other anxiety or mood-related disorders, certainly other OCD spectrum disorders, and other conditions, such as ADHD or bipolar disorder; when patients come in, sometimes they present with one of these comorbid conditions and, again, clinicians need to screen for OCD to pick it up.

Also, we think that the same basic phenomena may be playing an important role across these other obsessive-compulsive spectrum disorders—if we understand what drives the symptoms in OCD and how to treat them in OCD, that may be applicable to a much larger universe of conditions.


It seems that many treating clinicians, even psychiatrists, get caught up with the complaint of depression or anxiety, and never look for OCD. What advice would you give to primary care physicians in order to help them pick up the diagnosis? Do you describe patients who come in with all their medical records, who write on every available white space on their intake forms, and who telephone multiple times before and after the first appointment?

I think that those are good clinical pearls that you have brought up. It helps to specifically ask a few screening questions. One is “Do you have some thoughts that you cannot get out of your head that are upsetting?” And then, “Do you have to do certain kinds of things over and over again, such as washing your hands or checking the oven or the locks?” “Do you need things to be perfect or just so or symmetric?” and “Do you have trouble throwing out things?” If people answer yes to any of these things, then more systematic questioning is needed to look at the different kinds of symptoms they may have.


People often think that if you have OCD, you must be fastidious and well-organized. Is the opposite also true?

Yes. In fact, many patients who have OCD have houses that are total disasters. They are a wreck partly because they are hoarders and pile up newspapers or magazines. For example, in the movie “The Aviator,” there was an interesting scene about Howard Hughes (who had OCD), who was holed up in his movie projection room. He was hoarding his urine in bottles and at the same time was unable to put clothes on because he thought that they were contaminated. There is a dynamic struggle between perfectionism and complete disorganization.


Is the paranoia that Howard Hughes developed in “The Aviator” an accurate portrayal of someone with OCD? Do OCD patients sometimes deteriorate like that, into paranoia?

It is not typical for OCD, but there are many patients with OCD who get stuck ruminating and can develop a major depression or a bipolar illness with psychotic-like symptoms. There are also others who develop delusional certainty about their obsessions.


Is there an emerging understanding of the anatomy and neurochemistry of OCD, and has this led to the development of any lab tests or imaging findings that are clinically useful?

There does seem to be unique brain circuitry that mediates the repetitive thoughts and behaviors in OCD and many obsessive-compulsive spectrum conditions, including PANDAS and even autism, for example. In these disorders, there is some evidence for increased caudate metabolic activity, frontal lobe activity, and limbic activity (in the thalamus and anterior cingulate).

There are a couple of unique subtypes that clinicians should look for; if they find that people have a waxing and waning course of OCD with exacerbation of symptoms with recurrent strep throats, then they can culture for strep and look for anti-strep antibodies.

There have been some research studies looking at certain markers, such as D8/17, which seems to be expressed at high levels in patients with Tourette’s syndrome, pediatric OCD, PANDAS, and also autism with repetitive behavior. Therefore, D8/17 may be a marker of these repetitive behaviors across different neuropsychiatric disorders. In those individuals, a clinician should think about intervening with a prophylactic, aggressive antibiotics, or immunomodulatory-type treatments to clear out circulating auto-antibodies that may bind to and cause swelling of the caudate.

In autism, for example, the severity of repetitive behaviors seems to map onto the expression of this D8/17 marker, and is tightly correlated with the size of the caudate on the right side. First-degree family members of patients with autism also tend to express repetitive behaviors and have very high rates of OCD. Within autism, there are a number of different symptom domains that come together to form the full syndrome. One of these symptom domains, repetitive thoughts and behaviors, may have common genetic, neurobiological, and environmental factors that also contribute to the development of OCD.  PP