Mobile Menu - OpenMobile Menu - Closed

Exploring the Relationship Between Medication and Veteran Suicide.

Printer-Friendly Version











FEBRUARY 24, 2010

SERIAL No. 111-62

Printed for the use of the Committee on Veterans' Affairs





For sale by the Superintendent of Documents,  U.S. Government Printing Office
Internet:  Phone: toll free (866) 512-1800; DC area (202) 512-1800
Fax: (202) 512-2104  Mail: Stop IDCC, Washington, DC 20402-0001



BOB FILNER, California, Chairman


VIC SNYDER, Arkansas
JOHN J. HALL, New York
TIMOTHY J. WALZ, Minnesota
JOHN H. ADLER, New Jersey
GLENN C. NYE, Virginia

STEVE BUYER,  Indiana, Ranking
HENRY E. BROWN, JR., South Carolina
BRIAN P. BILBRAY, California
DAVID P. ROE, Tennessee




Malcom A. Shorter, Staff Director

Pursuant to clause 2(e)(4) of Rule XI of the Rules of the House, public hearing records of the Committee on Veterans' Affairs are also published in electronic form. The printed hearing record remains the official version. Because electronic submissions are used to prepare both printed and electronic versions of the hearing record, the process of converting between various electronic formats may introduce unintentional errors or omissions. Such occurrences are inherent in the current publication process and should diminish as the process is further refined.



February 24, 2010

Exploring the Relationship Between Medication and Veteran Suicide


Chairman Bob Filner
    Prepared statement of Chairman Filner
Hon. David P. Roe
    Prepared statement of Congressman Roe
Hon. Harry E. Mitchell, prepared statement of



U.S. Department of Veterans Affairs, Ira Katz, M.D., Ph.D., Deputy Chief Officer, Mental Health Services, Office of Patient Care Services, Veterans Health Administration
    Prepared statement of Dr. Katz
U.S. Department of Defense, Brigadier General Loree K. Sutton, M.D., Director, Defense Centers of Excellence for Psychological Health and Traumatic Brain Injury, Special Assistant to the Assistant Secretary of Defense for Health Affairs
    Prepared statement of General Sutton

American Psychiatric Association, Annelle Primm, M.D., MPH, Deputy Medical Director for Minority Affairs, and Associate Professor of Psychiatry, John Hopkins School of Medicine, Baltimore, MD
    Prepared statement of Dr. Primm
American Psychological Association, M. David Rudd, Ph.D., ABPP, Dean, College of Social and Behavioral Science, The University of Utah, Salt Lake City, UT
    Prepared statement of Dr. Rudd
Breggin, Peter R., M.D., Ithaca, NY
    Prepared statement of Dr. Breggin
Farber, Commander Donald J., Esq., USN (Ret.), San Rafael, CA
    Prepared statement of Commander Farber
Leon, Andrew C., Ph.D., Professor of Biostatistics in Psychiatry and Public Health, Weill Cornell Medical College, New York, NY
    Prepared statement of Dr. Leon


Billings, Bart P., Ph.D., Carlsbad, CA


Wednesday, February 24, 2010
U. S. House of Representatives,
Committee on Veterans' Affairs,
Washington, DC.

The Committee met, pursuant to notice, at 10:00 a.m., in Room 334, Cannon House Office Building, Hon. Bob Filner [Chairman of the Committee] presiding.

Present:  Representatives Filner, Michaud, Herseth Sandlin, Mitchell, Halvorson, Perriello, Teague, Rodriguez, Donnelly, Space, Walz, Adler, Bilirakis, and Roe.


The CHAIRMAN.  Good morning.  The Committee on Veterans’ Affairs will come to order.

I ask unanimous consent that all Members may have 5 legislative days in which to revise and extend their remarks.  Hearing no objection, so ordered.

Thank you all for attending today’s hearing.  I think it is an important hearing to look at the potential relationship between psychiatric medications and suicides.  Not an attractive topic, but one that I think we have to address.

Certainly, we know with post-traumatic stress disorder (PTSD) and traumatic brain injury (TBI) being so prevalent in the current wars in Iraq and Afghanistan, mental health issues have taken and should take center stage.

Research has shown that mental disorders and substance abuse disorders are linked to more than 90 percent of people who die by suicide.  Today, as you know, suicides among servicemembers and veterans continue to increase at an alarming rate, far exceeding the comparable suicide rates among the general population, and I think higher than we had during the Vietnam War.

It is a tragedy that our servicemembers and veterans survive the battle abroad only to return home from the theater of war to fall by suicide.

We know there is a widespread availability and use of psychiatric medications to address mental health disorders, but there is apparently some dispute about whether these drugs prevent or lend a hand in suicide.  Some doctors are convinced by their clinical experience that psychiatric drugs often adversely impact the individual’s better judgment and lead people to lose control over their emotions and actions.

Suicides may be driven by so-called drug-induced adverse reactions and intoxications.  There are, on the other hand, some studies that show suicide attempts were lower among patients who are treated with antidepressants than those who are not.

Through this hearing, we will explore the two opposing schools of thought on the relationship with psychiatric medicine and suicide.  In this process, we will also seek to better understand the reasons why more and more servicemembers and veterans are taking their own lives and what the U.S. Department of Veterans Affairs (VA) and U.S. Department of Defense (DoD) are doing to prevent such deaths.

Before we hear from our first witnesses, I will recognize Dr. Roe for an opening statement.

[The prepared statement of Chairman Filner appears in the Appendix.]


Mr. ROE.  Thank you, Mr. Chairman, and thank you for holding this hearing.

I think those of us who are gathered here today would be hard pressed to find a topic more heart breaking than when a servicemember makes the decision to end his or her own life.  This hearing is one of the many hearings and meetings this Committee has had in an effort to combat veteran suicide and I can tell you that the stories we hear in these proceedings much like those in Dr. Breggin’s book always raise difficult questions.

As painful as such anecdotal accounts are, we must take heed not to be so quick to point out to a single cause or mistaken theory for a solution.  It is sound research that is critical to our efforts to put an end to these tragedies and understand the entire story.

On that front, there are many encouraging signs.  In 2008, the Army and the National Institutes of Mental Health (NIMH) began a 5-year study into the factors that contribute to suicide in the Armed Forces and how to prevent them.  Called the ARRM Study to Assess Risk and Resilience in Servicemembers, this is the largest study of suicide and mental health among military members ever conducted.

In addition, there is a great deal of ongoing public and private research into the causes of suicide and treatment options, including medication, to prevent it. 

I am hopeful that with this research, practitioners will be able to better identify risk factors for veteran suicide and design prevention, outreach, and treatment options that are effective and practical within the VA setting.

The psychology behind why a person may see death as the only way out is more complex than any of us have the ability to fully comprehend and it is the interaction of a number of factors that may lead to this catastrophe.

In addressing these issues, one cannot simply place blame on the veteran, their military service, their illness, or their chosen treatment option.  As the research goes on, we must allow our veterans and servicemembers to have the full range of approved treatment options that they decide upon with their doctors.

I want to thank our witnesses for being here this morning and I look forward to hearing and learning from each of you.  It is only by working together that we can convince every courageous yet struggling American veteran and their country that supports them that hope and help are out there.

I yield back the balance of my time.

[The prepared statement of Congressman Roe appears in the Appendix.]

The CHAIRMAN.  Thank you, Dr. Roe.

I want to introduce the first panel.  We have Dr. Peter Breggin, Psychiatrist and Author from Ithaca, New York, and Dr. Andrew C. Leon, Professor of Biostatistics in Psychiatry and Public Health at Weill Cornell Medical College.

Thank you for being with us. 

Dr. Breggin, I just have one question to start off with, to test your mental state—do you willingly live in Ithaca, New York?

Dr. BREGGIN.  I lived here most of my life, DC.

The CHAIRMAN.  I spent 10 years at Cornell, so I know something about the background.

Dr. BREGGIN.  That is right.

The CHAIRMAN.  Okay, test passed.  Please enlighten us.  You have the floor.



Dr. BREGGIN.  Well, I am Peter R. Breggin, M.D.  I am a psychiatrist.  And I was in this area of DC for most of my career and then we moved to Ithaca, New York, to be in the country.

In the early 1990s, I became the first psychiatrist to speak and write extensively about violence and suicide caused by the newer antidepressants beginning with Prozac, later going on to Paxil, Zoloft, Celexa, and other drugs.

I also, as a result of that early research, became a scientific expert for more than 100, I think it was like 170 product liability cases against Eli Lilly, the manufacturer of Prozac, that were combined by a court to provide the opportunity for one person to research the data and look into the company files for all of the suits.  And I was chosen to be that one medical expert.

This ended up giving me experience that literally no one else in the world has had in terms of looking at the basic data from Eli Lilly concerning the development and marketing and then from some other drug companies.

I was shocked at what I found inside the company.  For example, the German equivalent of our Food Drug Administration (FDA) had become concerned that they were finding an increased suicide rate on studies of Prozac.  So they asked Eli Lilly, and this is back now in the late 1980s, to go back and look at all of their clinical trials and report to them on the rate of suicide attempts in the controlled clinical trials of Prozac compared to another drug or placebo.

Lilly found, depending on how you count it, a six to twelve to one ratio of suicide attempts, not just thinking, attempts in the control or comparison group compared to placebo.  Lilly never made the results public.  They never gave this report that I found to the Germans.  They never made it available to the FDA.

I also found memos inside Lilly explaining guilt and shame on the part of some German investigators working for Lilly that the company was classifying suicides and suicide attempts reported by doctors to them as no drug effect or other harmless kinds of entities, thereby disguising the suicide attempts and the completed suicides.

And one of these memos, the gentleman declared, how am I going to explain this to my family.  He expressed a genuine feeling of  shame.

At the same time, the FDA conducted a study comparing Prozac to an older antidepressant, Trazodone.  After factoring in the increased number of prescriptions for Prozac and also factoring in the controversy because the controversy had not broken out yet, there were far more reports of suicidality and violence and other mental adverse effects on Prozac.

I worked on these issues for many, many years, as you know, and then testified before the FDA in 2004 on a couple of occasions and the agency distributed one of my papers written in 2003 to the panel, the FDA panel.  And a lot of the language in the current label virtually reads actually very, very similar to what I had to say in my papers and books.

Now, my conclusions in this testimony are based not only on these very early studies that I discovered inside of Eli Lilly, which, by the way, you can find on my Web site, the Lilly documents I am describing, and I also described them in a couple of my books.

But my conclusions are based in part on the many citations in the paper I wrote specifically for this Committee.  I actually sat down and wrote you a paper, Antidepressant Induced Suicide and Violence: Risks for Military Personnel, and in the hundreds of citations in my book.

My recent book, which Mr. Roe was kind to mention and which I know that you have read, Mr. Chairman, Medication Madness, gives an overview of my clinical experience, which now included in the book more than 50 cases of violence, suicide and crime, most of them on antidepressants. 

And I actually interviewed survivors.  I actually went to crime scenes, read all the medical records, police records, and clearly documented in Medication Madness from a clinical viewpoint that there are many, many cases like this.  I actually have over 100 that are mentioned in the book and 50 documented in detail.

In 2004, after the various hearings, the FDA actually before them, required the antidepressant manufacturers to review their clinical trials.  The FDA itself concluded that the newer antidepressants doubled the rate of suicidal thoughts and behaviors in children, youth, and young adults up to age 24, which, of course, is very menacing for the soldier population, the military population.

Now, you get a doubling of rates.  Well, what does this mean?  Well, the clinical trials are very short.  Most of them average about 6 weeks.  Some of the Prozac trials were 4 weeks.  Suicidal patients are excluded from clinical trials. 

The patient is monitored every week by experts and informed of all the dangers presumably and the patient is given huge hope.  You are in this wonderful research setting where you are getting something new and wonderful.  And, furthermore, there is no attempt to look for suicide attempts and to categorize them.

Now, when you get a doubling of suicide attempts and ideation under those conditions, you can assume that in the military or clinical practice it is going to be multiples, unknown multiples because there it is given for months, there it is not monitored, there psychotic patients are included, their suicidal patients are included, and all of that is excluded from the clinical trials.

Now, one of the questions that may come up today is that there were in this particular batch of trials no completed suicides.  The shock is how many attempts there were because the best way to treat suicide, if there were one, would be to simply put somebody in a clinical trial and give hope because suicide is loss of hope.  That is why when you get all the doctors looking at you and testing you and working with you, you almost never get suicides in any kind of clinical trial of that kind.

Now, the FDA warnings that came out of these hearings are identical for all antidepressants.  The Zoloft label is the model I am going to use.  And it begins with a huge black box, huge black box, very rare thing, with the title Suicidality and Antidepressant Drugs.  And I will read you just the first line of it.

"Antidepressants increase the risk compared to placebo of suicidal thinking and behavior, parentheses, suicidality in children, adolescents, and young adults in short-term studies of major depressive disorder and other psychiatric disorders." And later in the label, they will say that a lot of the adverse effects occur in nonpsychiatric patients.

This black box is very lengthy and many of the items are repeated over and over again in the warnings and further on.  It is the only label like that that I know of.  The black box is followed by a very ominous section, still in the warnings, entitled "Clinical Worsening and Suicide."  This idea of clinical worsening that is repeated in the label has not been given enough attention.

It states in this section that the following symptoms I am going to list, quote "have been reported in children and adults taking antidepressants both for psychiatric and nonpsychiatric purposes."  And the list includes "anxiety, agitation, panic attacks, insomnia, irritability, hostility, aggressiveness, impulsivity, akathisia," which is psychomotor restlessness, and the DSM-IV, our major document, points out that akathisia leads to violence and suicide.

So I interrupted.  "Akathisia, hypomania, and mania."  And mania is an out of control state that increases vastly the risk of violence and suicide.  This is the list that is virtually taken from several of my earlier publications and note the mention of irritability, hostility, aggressiveness, and impulsivity. 

Imagine causing that in young men and women who are heavily armed and under a great deal of stress.  And irritability, hostility, aggressiveness, impulsivity not only lead to violence but to suicide.  Many suicides are out of anger and irritability and resentments.

The CHAIRMAN.  Dr. Breggin, I do not mean to interrupt.  I just want to ask a specific question.  If an active-duty soldier is given these medications, they may not even see that warning, right?  I mean—

Dr. BREGGIN.  Well, my experience, last year, I spoke at the oldest military stress conference given.  Bart Billings, whom you know, retired Army officer and psychologist, runs that.  And I talked to Generals and I talked to mental health professionals and they all agreed that these warnings were hardly ever presented to the soldiers and that the Army was in a sense acting as if it was unaware. 

And some of these people gave me estimates not of the 15 percent of active-duty soldiers on psychiatric drugs that we often hear but up to 30 percent of soldiers in some sections.  Marines in particular was one that was mentioned to me.

The CHAIRMAN.  So they are not even informed of the risks?

Dr. BREGGIN.  No, no.  And as we go on further, we will see that the FDA tells doctors you should, and the word "should" is in the label, you should share this information with the patient and the family and make sure they understand it.  It is not just you repeat it to them.  You sort of, you know, "hey, I want you to understand this is what may happen to you."

That is what I do in my clinical practice.  I do not say by the way, the drug may cause this or that, you know.  I just make sure over a period of many sessions that the person understands the risks.

Did I answer your question, sir?

In addition to this list that is associated with the drug itself, the antidepressants, and this is mentioned in the label and mentioned in the medication guide I will tell you about, the antidepressants often cause severe withdrawal reactions in which all of those symptoms, all of those adverse reactions can develop. 

In fact, I spend probably half my practice even in remote Ithaca treating people who come to me to try to get off of these drugs and are suffering violent feelings, suicidal feelings when they try to stop.  And, in fact, in the last 3 weeks, I have had at least three or four patients who as we went down lower on their doses developed really, really frightening reactions and then I had to treat those, usually by raising the dose back up for a while.

Let me mention some of the science more specifically at this point.  This list of mania and hypomania and agitation and aggression and so on, that list in one FDA document is stated to be a "known" effect, this whole series that I read to you, are known effects of the drugs.

And what I am going to read to you now involved mostly controlled clinical trials or epidemiological studies.  No one should be able to say that causality has not been demonstrated.  The gold standard for causality is the controlled clinical trial and it shows a doubling of the rate of suicidality.  That is the gold standard. 

And I have a discussion that is documented between the top members of the FDA agreeing at the hearings that unless somebody is cheating or there is some other malfeasance going on when you have a causal association in controlled clinical trials.  I would add epidemiological studies that demonstrates causality.

In addition, Federal regulations say that the warning labels must have a reasonable degree of certainty about causality before it gets put into the label.

An overview of some of the studies that are involved, because I want you to know this is not merely my personal opinion.  This is what the science has overwhelmingly taught me starting with my looking inside the drug companies.

In addition to the studies done under the auspices of FDA, we have, and this is for children and adults, we have a study by Aursnes, A-U-R-S-N-E-S, in 2005.  He looked at 16 placebo controlled clinical trials in which Paxil was randomized against placebo and found increased suicidal behavior. 

The references are in the article that I gave to you, attached to my testimony, as well as my books and paper.

Ferguson in 2005 searched the adult literature, found 702 randomized clinical trials of 87,000 patients and found a significant increase in suicidality on antidepressants.

Donovan in 1999, in a large British study involving 229 completed suicides, it is a big study in England, found a higher suicide rate in patients treated with the newer antidepressants.

Donovan in 2000 examined 2,776 consecutive cases of deliberate harm in individuals age 17 and older, not children, 17 and older, like soldiers, seen at the emergency department of a British infirmary.  Again, the suicide rates were increased in people taking the newer antidepressants.

A fellow named Jick, J-I-C-K, in 1995 conducted an epidemiological study in the United Kingdom involving 172,000 adult patients and Prozac was associated with more suicides than the older antidepressants.

In my home State for many years of Maryland, Frankenfeld and some very respected researchers at the University of Maryland studied coroners’ cases in Maryland and found that suicides were more violent, which is my clinical experience, in patients taking Prozac compared to older antidepressants.

Now, GlaxoSmithKline in 2006, the manufacturer of Paxil, conducted a new meta-analysis of all its adult trials, the FDA said you have got to do a meta-analysis of all the adult trials, and found a statistical increase in the rate of suicidality in depressed patients of all ages, an increased rate from the clinical studies, their own, which were not oriented toward this, an increased statistically significant rate, all ages in patients with major depressive disorder.  It is in a Dear Doctor letter that they sent out to all the health care, it is really now called Healthcare Letter, to all the health care professionals in the country.

A study of 1,255 suicides in 2006 in Sweden found that, which was 95 percent of all the suicides in Sweden, by Ljung, et al, L-J-U-N-G, published in 2009, found there was a greatly increased number of completed suicides exposed to the antidepressant drugs.  In fact, 52 percent of the Scandinavian women who killed themselves had filled a prescription for these drugs.

Now, this is not as causal as the clinical trials.  It could be other factors.  But it is just part of this mountain, mountain of evidence.

A retrospective study examined the suicide rate in the VA involving a cohort, a group of people that were 887,000, that is six digits, 887,000 VA patients treated for depression and found, quote, "completed suicide rates were approximately twice the base rate following antidepressant starts in VA clinical settings."  That is Valenstein, et al, in 2009.

Now, again, you can look at something like this and say, well, maybe the worst patients were put on the antidepressants and that is the correlation.  That is not what they concluded.  And, of course, this now comes against the backdrop of the clinical trials which show causality.

Juurlink, J-U-U-R-L-I-N-K, et al, in 2006 reviewed more than a thousand cases of actual suicide in the elderly and found that during the 1st month of treatment with selective serotonin reuptake inhibitors (SSRIs) there was a five-fold increase in risk in the elderly.  This is no surprise because the elderly are more susceptible to adverse effects.

Fisher, et al, in a really interesting study did a phone survey of people who went to a pharmacy and got drugs, medications, and found that there was a higher rate of suicidality in people who got the SSRIs compared to other antidepressants.

I do not think there is a question about causality, although some people will raise questions of causality today, I am sure.

Finally, just to look at the literature on mania, because mania, if you read the DSM-IV, is caused by antidepressants.  This is in our diagnostic manual, that all of the phenomena of mania are caused by antidepressants as well as by just simply bipolar disorder.  And mania results in suicide, violence, crime.  I have had a whole bunch of just dreadful cases like that.

Well, in 2001, Preda, P-R-E-D-A, found that 8.1 percent of adult psychiatric admissions could be attributed to antidepressant induced mania and psychosis, eight percent of hospital admissions.

Another group found that eight percent of patients treated with Paxil developed mania.  In other words, they are not even looking for it.  They go back and they look at records and they find that eight percent of the patients got mania when the drug was started.  Causality has been definitely established in studies like this.

Howland in 1996 again found a six percent rate.  Look at these rates, six, eight.  Very, very consistent of SSRI induced mania.  To induce mania in a soldier, in an armed young man or woman is an incredibly risky affair.

Ebert, et al in 1997 found a 17 percent rate of hypomania in patients on SSRIs.  Some were suicidal or dangerous.

Martin used a national database of 7 million privately insured individuals and he found that if you look at people given antidepressants, all of a sudden, they are getting bipolar disorder diagnoses afterward, after the antidepressants.

I could go on and on, but I will not.  I want to tell you one more study because this one comes out of the heart of the advocacy group for psychiatric drugs.  It comes from Harvard Medical School where most of what they do is financed or much of it by the drug companies and where some of the prominent doctors were recently under investigation by Senator Grassley for taking money from the drug companies and not informing people.

Well, they did a study, this is Wilens, et al, 2003, of adverse psychiatric events on children taking these drugs, children and adolescents, and they are just more susceptible than adults, but it is the same phenomena.  And they found that 22 percent had adverse psychiatric events, quote, "most commonly related to disturbance of mood."

Then they did something that is called a re-challenge.  We have got a few studies like this.  A re-challenge is where somebody develops a symptom like suicidality.  You stop the drug, the symptom goes away.  You restart the drug, the suicidality comes back.  You stop the drug, it goes away.

Rothschild did a study like this, not even in this paper, but you can find it my books and scientific papers.  The FDA says this is a very, very good thing to do.  Well, in this case, when they re-exposed the children to an SSRI, 44 percent of that group again became disturbed.  And what drug-induced symptoms did they develop?  The things we have been hearing about.  They became irritable, anxious, manic.  They developed insomnia.  Four percent of the children became aggressive. 

The CHAIRMAN.  Dr. Breggin, I need you to—

Dr. BREGGIN.  I will finish up now.

The CHAIRMAN [continuing].  Finish up right now.  Okay.  Thank you.

Dr. BREGGIN.  I really appreciate this time to share with you my work and the literature.

Now, under FDA regulations, I want to talk about efficacy very briefly, the pharmaceutical companies can cherry pick their studies.  They can do six or eight studies and just provide two that are marginally effective, statistically significant to the FDA for the purpose of pricing efficacy.  It is not hard when you are purchasing the investigators and writing the protocols for them and then analyzing the data inside the drug company for the companies to develop positive studies, but it is still hard.

And when all of the antidepressant studies that are done, not just the cherry picked ones, are combined in a meta-analysis, the antidepressants are no better than placebo. 

Now, as you may discover today, psychiatric associations and other groups that rely heavily on financial support are going to try to reject and deny all this. 

In conclusion, there is overwhelming evidence that the newer antidepressants commonly prescribed in the military can cause or worsen suicidality, aggression, and other dangerous mental states.  The documented increase of suicides in the military as well as any discovered, and I hope you will look into this, Mr. Chairman, any discovered increase in random violence among soldiers is in part caused or exacerbated by the widespread use of prescriptions for antidepressants.

Finally, little will be lost and much will be gained by curtailing the prescription of antidepressants in the military.  The military instead should rely upon newly developed psychological and educational programs, many of which are being implemented and which Dr. Bart Billings, who is familiar to this Committee, has written about in his report to the Committee, including his Human Assistance Rapid Response Team (HARRT) program.

Thank you very, very much for the time.

[The prepared statement of Dr. Breggin appears in the Appendix.]

The CHAIRMAN.  Thank you so much for that, rather chilling testimony.

Dr. Leon, you are recognized.


Dr. LEON.  Thank you for the opportunity, Mr. Chairman and distinguished Committee Members, to discuss this topic.

My name is Dr. Andrew Leon.  I know this is clearly an emotional issue.  My family has been profoundly impacted by mental illness, so much so that I devoted my career to the field of psychiatry.

I am Professor of Biostatistics in Psychiatry and Public Health at Weill Cornell Medical College where I have been on the faculty for over 20 years.  I have published over 200 peer-reviewed scientific manuscripts.  Nearly all of my research has been funded by the National Institutes of Health (NIH).

I have served as a consultant to FDA, to the NIMH, and to industry primarily to monitor the safety of patients who are enrolled in clinical trials on data and safety monitoring boards.

All of us here in this room share a common goal and that is to do the very best for our veterans.  My perspective is that doing the best requires the discipline to use empirical methods to understand optimal mental health care and suicide prevention.

I was a biostatistician on the FDA’s Psychopharmacologic Drug Advisory Committee from 2003 to 2008 and participated in the FDA hearings on antidepressants and suicide, antidepressants and suicidality, I should say, not suicide deaths.

The class of medications that I will discuss is antidepressants.  First, depression is a life-threatening illness.  Suicidality is a symptom of depression, whether treated or untreated.  My main points that I will make today are depression increases the risk of suicide.  Antidepressants reduce suffering from depression that has been demonstrated in several hundred randomized controlled clinical trials where the investigators and the assessors were blinded to the treatment received by the subjects in the trial.  There is not a way that it can be manipulated by a pharmaceutical company when they are blinded to the treatment received when the ratings are done.

The CHAIRMAN.  I do not mean to interrupt, but I do not think the issue was whether somebody is cheating on a study.  It is the selection of studies when they are finished.

Dr. LEON.  Oh, no.  Absolutely.  That is a very important point.

The CHAIRMAN.  You cannot say that cherry picking cannot result from this.

Dr. LEON.  I will address that in just 1 minute.

So my three points that I want to make, then I will address, Mr. Chairman, your point, is depression increases the risk of suicide.  Antidepressants, antidepressant medication can reduce the suffering from depression.  And to reduce risk of suicide, clinicians must carefully monitor veterans with depression, whether treated or untreated.

Now, with regard to the clinical trials, Mr. Chairman, that you are referring to, all clinical trials conducted by a pharmaceutical company for a particular drug must be submitted to the FDA.  They cannot cherry pick.  They submit all.  The point Dr. Breggin was making that the FDA regulations require at least two positive trials where the medication meets another cell, typically placebo.

So the cherry picking that Dr. Breggin was referring to, whether it was submitting just part of the results or manipulating the data that come in, is simply not true.  I have reviewed those data.  I reviewed the data from many hearings at clinical in the FDA, not just on suicidality, but for many other indications that were being sought for other psychiatric medications.

Today I am going to discuss three different types of studies, randomized controlled clinical trials where antidepressants are typically compared to placebo, observational studies, and postmortem studies.

Three types of suicidality are reported in these studies, suicidal thinking, suicide attempts, and suicide deaths.

In 2004, the FDA reviewed 25 pediatric clinical trials for antidepressants involving over 4,400 subjects and found that patients randomized to antidepressants were about twice as likely to report suicidality.  Nearly all of that was suicidal thinking.  These were for pediatric clinical trials, children and adolescents under the age of 18.

There were about three percent of those on medication or placebo.  Three percent reported feelings of suicidality.  And that was mostly suicidal thinking.  About 80 to 90 percent of that was suicidal thinking.  There were no suicide deaths in those clinical trials, no suicide deaths.

In 2006, we reviewed 295 clinical trials of antidepressants for adults involving over 77,000 participants.  Less than one percent of those participants reported suicidality, most suicidal thinking.

Unlike the pediatric trials, adults randomized to antidepressants were not more likely to report suicidality.  In fact, antidepressants conveyed significant protection for adults over the age of 65.  For the age group that was referred to earlier, ages 18 to 25, there was not a significant increase in the risk of suicidality.  I mean, I reviewed the data.  I wrote five papers on this topic.  I have one of them here.  The rate was not elevated more than we would expect by chance. 

So we have clinical trials here, say maybe nearly 300 clinical trials.  From those, the 11 new antidepressants, new meaning starting in 1985, a newer class, a newer generation of antidepressants, was being developed and approved.  All of those antidepressants had demonstrated efficacy, that is clinical benefit for treating the symptoms of depression in more than one trial.  So the efficacy, the clinical value of these was very clearly presented empirically. 

Now, I am involved in the NIMH collaborative depression study, which just ended last year.  It was a 31-year follow-up study of patients who presented with mood disorders, bipolar disorder, and depression starting in the late 1970s.  And we continued to assess as many as we could follow until 2009.

One relevant paper from that is, I was able to look at those subjects during the course of the study, and, actually, this was before the 2009 data came in, so we had at the time up to 28 years of follow-up data, in order to examine the risk of suicidality and antidepressants.

And what is not included in the clinical trial is those who received no treatment and those subjects who were not quite severely ill enough to qualify to be enrolled in a randomized trial and those subjects who were too severely ill to be enrolled in a randomized trial.  We included subjects who were suicidal.  We included subjects who had co-morbidity, both psychiatric co-morbidity, substance abuse, alcohol abuse, and other medical co-morbidity.  We included subjects who were taking other medications. 

And during the course of this from 757 subjects, we had over 6,700 intervals during which they either received or did not receive antidepressants.  We found from that that antidepressants significantly reduced the risk of suicide attempts and suicide deaths.  We did not assess suicidal thinking, but there was a significant reduction in the risk of suicidal behavior, that is attempts and deaths.

The data from our observational study are much more generalizable than a randomized controlled trial because we include much more the subjects and the situations they are in more typically reflects those who are receiving antidepressants in the United States.

Our research group at Cornell has conducted several postmortem studies of suicide deaths in New York City.  In fact, we examined all suicide deaths in New York City.  We looked at the autopsy records over a 10-year period for youth suicides, children and adolescents under the age of 18.  We looked at the toxicology to determine whether or not antidepressants were taken before their suicide death. 

There were 105 adolescents and children who killed themselves in New York City over a 10-year period.  Ninety-five percent of those deaths had not, of the suicides, had not taken antidepressants.  Only five percent had been treated with antidepressants.  It is a tragic story. 

We repeated it again in adults.  We looked at a 4-year period and there we had over 1,400 suicides.  We had autopsy records of 1,400 suicides.  Seventy-seven percent of those suicide deaths had not received antidepressants prior to their suicide.

This suggests that prevention of suicide requires intervention primarily among patients who are not receiving antidepressants. 

A cause and effect relationship has not been established between antidepressants and suicide.  This is one of the most controversial issues in the field of psychiatry.  It is an issue about which many people write and speak without access to the proper data.  The randomized controlled clinical trial data that the FDA reviewed is the most comprehensive database about antidepressant exposure ever assembled in the field of psychiatry.

Antidepressants have clearly been shown to reduce the suffering from depression and suffering from depression is accompanied by significant functional impairment, job loss, family disruption, and inability to get out of bed in the morning.  And a great deal of that can be reduced by taking antidepressants.

However, there is a risk-benefit ratio we have to look at as with any medication.  As with any medication, they are not going to be perfect.  But a great deal of those patients who take the medications will get better and a very tiny percentage will have thoughts of suicidality.  They may have had those thoughts before they started. 

Okay.  Let me wind up by saying, so the cause and effect relationship has not been established.  However, antidepressants do successfully reduce symptoms of depression. 

In light of the suicide risk in depression, a prudent recommendation is that veterans, whether treated or untreated, must be appropriately monitored by clinicians.

In conclusion, I would like the Committee to recognize that depression is itself a risk factor for suicide.  To leave these men and women untreated is to accept suffering from the disorder itself.

I thank the Committee for the opportunity to speak here today on this critically important issue.

[The prepared statement of Mr. Leon appears in the Appendix.]

The CHAIRMAN.  I thank both of you for your testimony.

In your last sentence, Dr. Leon, I do not think anybody was ever suggesting not to treat people.

Dr. LEON.  Good.  I am glad to hear that.

The CHAIRMAN.  You are setting up a false straw man there, but we will get to that.

Mr. Teague, do you have any questions?

Mr. TEAGUE.  No.  For the second time, I will pass for right now.

The CHAIRMAN.  Mr. Rodriguez?

Mr. RODRIGUEZ.  I was listening to the comments and I thought I heard that you indicated that the FDA looks at the top two studies to determine if they are positive. 

Is there a review of the literature before deciding on anything or is it just the two top positive studies?  I guess both of you all can comment.

Dr. LEON.  Well, there will be very little literature on medication that has not been approved for sale in the United States.  I mean, the literature is presented.

And just as a clarification, the FDA does not just review the top two studies.  They review all of the studies.  But what is required for approval, it is one of the many criteria requirement, is that they have at least two positive studies. 

But when I was on the Advisory Committee, we reviewed data from all of the studies that were conducted with a particular medication.

Dr. BREGGIN.  Yes.  The drug company can do as many studies as it wants.  And, again, they are very heavily programmed by the drug company.  And then it only has to submit for efficacy to get the drug approved only two studies that are marginal.

When you put all those studies together, including the ones that were not cherry picked, say the five or six or seven others that each of the drug companies is doing, the most recent meta-analysis shows that, in fact, the drugs do not work.

It is easy to pick a study that says the drugs are efficacious.  We are looking at a mountain of evidence that they cause suicidality.

I am very surprised at Dr. Leon’s comments that suicidality has not been proven.  He is willing to use two of the cherry picked studies to say there is efficacy and it has been approved by the FDA, but he is not willing to use the FDA’s own studies, which the drug companies were forced to reevaluate, to use the FDA studies to show causality for suicidality.  What is good for the goose is good for the gander.  And this, I believe, is an extraordinary omission.

And also he is contradicting the FDA’s basic conclusion that the risk goes to age 24.  It is in the Black Box Warning.  Federal regulations require that there be evidence for causality before you have a Black Box Warning.

Dr. LEON.  Well, that is incorrect.  Okay.  The Federal regulations do not require evidence of causality.  I have been involved—

Dr. BREGGIN.  The warnings.

Dr. LEON.  That is not correct.

Dr. BREGGIN.  I can get them for you.

Dr. LEON.  Dr. Breggin, you do not know what you are talking about.  There are many things you have said that are incorrect and that was the most blatant error you have made today.

Dr. BREGGIN.  I think—

Dr. LEON.  I am speaking right now, Dr. Breggin.  Thank you.

Dr. BREGGIN.  Mr. Chairman, I think you actually have the regulation.

Mr. RODRIGUEZ.  I think if I have time, I would like to ask another question.

Now that you are talking about suicides, one of the things you mentioned was that medication was reducing the deaths and attempts of suicides, however, that the suicidal thinking was still there.

And so, how do you determine a situation such as that?  Because I know that if you are working with an individual and he is suicidal during the period of time when you are engaged with them, a lot of times if they still have the suicidal thinking, as soon as you let them go, the possibility of committing suicide or attempts will probably come back, but when you continue to be engaged with them in the studies, closer the monitoring the less they are likely to do it? 

I think that is common sense.  If I want to commit suicide and I am in a study, and you are watching me a lot closer, it is less likely I am going to kill myself or less likely to make an attempt; however, I still would have the suicidal thinking which you have indicated stays with them.

So what did the medication actually do?

Dr. LEON.  Okay.  Congressman, apparently I was not clear.  I was talking about two different studies.  The one study, that large observational study funded by the National Institute of Mental Health, we only evaluated, we only asked questions about—recorded information about suicide attempts and suicide deaths.  We did not ask about suicidal thinking.

So in that, we did not—

Dr. BREGGIN.  Why not?

Dr. LEON.  Why not?

Dr. BREGGIN.  Yeah.  Why not?

Dr. LEON.  Well, we had to—

Dr. BREGGIN.  It is a very important clinical practice.  It is a much more—

Dr. LEON.  Thank you. 

Okay.  We had 28 years of follow-up at the time.  We developed our assessment criteria back in the 1970s and the assessment criteria served as the standard for psychiatric research that is conducted today.

We actually did ask about suicidal thinking once every 5 years, but our weekly assessment records recorded suicide attempts, suicide deaths, all medications taken and—

Mr. RODRIGUEZ.  But you said you only did it every 5 years?

Dr. LEON.  No, no.  The suicidal thinking question was once only every 5 years.

Mr. RODRIGUEZ.  Because I know that as soon as a person indicates any kind of tendency for suicide, a flag goes up.  And so in the studies, you would think that would be accounted for because then the family is engaged also and thus it would be less likely for the attempt to occur.  If other factors are also there, then it is less likely to occur without even any medication.  And that just makes sense.

But, you know, I guess we would have to see which prescription and which item because if the suicidal thinking is there, then it is still there.  And as soon as you maybe take them off the support in terms of the family or anything else, whether they are taking medication or not, they might commit suicide.

Dr. LEON.  It is not like a clinical trial where a comprehensive assessment battery is administered every week or 2 weeks and the medication received is controlled by the investigator.  An observational study is very different. 

A randomized controlled clinical trial has very strict inclusion and exclusion criteria.  Typically for antidepressants, it excludes about 80 or 90 percent of the patients who want to become subjects in the trial. 

An observational study on the other hand observes the treatment and the responses to treatment that subjects and the patients have and it also observes responses and the clinical course among people who are not treated.  But in an observational study, the investigator does not have any control over the treatment.  They receive the treatments that their clinicians choose to give them.

Mr. RODRIGUEZ.  I ran out of my time.  I do not know if you will allow Dr. Breggin to answer.

The CHAIRMAN.  Go ahead, Doctor.

Dr. BREGGIN.  Notice that Dr. Leon is using a noncontrolled study, a naturalistic study, which allows for multiple interpretations, multiple variables, all kinds of things going on in the practice, but is rejecting the gold standard that gentlemen like Dr. Leon always said was the gold standard which are the controlled clinical trials.

Every single study, save maybe two that I read to you today, the epidemiological studies and the clinical trials, were all controlled.  He is basing his rebuttal or his argument on uncontrolled data.  And that is notoriously not good for determining causation.

Mr. RODRIGUEZ.  Now, how do you determine uncontrolled data?

Dr. BREGGIN.  There is no control group to the work he is doing.  He is just looking at collaborative unity. 

Dr. LEON.  That is incorrect, Dr. Breggin.

Dr. BREGGIN.  But it is not a controlled clinical trial.

Dr. LEON.  No.  The difference as a researcher, as a—

Dr. BREGGIN.  For causation, it is not as good, period.

The CHAIRMAN.  Let Dr. Leon answer, please.

Dr. BREGGIN.  Sir, you are an epidemiologist.

Dr. LEON.  I am a biostatistician.  I have designed dozens, hundreds of trials, Dr. Breggin.  And the difference is we did have a control in our observational—

Mr. RODRIGUEZ.  In layman’s term, control means you have a group on one side and a group on the other?  Is that correct?

Dr. BREGGIN.  That are identical but receiving different treatments and you do not know which one is which.

Mr. RODRIGUEZ.  Okay.  I understand.

Dr. BREGGIN.  You did not have that.

Dr. LEON.  The investigator and the subject do not know, the patients do not know which one is which.  But we did have a control in our study and that was the control that we see out in the general population, subjects either—I mean, those with depression in the general population, those veterans with depression are either treated or they are not treated.  Our control in our observational study was some subjects received antidepressants, the other subjects did not receive antidepressants.

The CHAIRMAN.  Is an observational study just based on data or are you looking at people in real time?

Dr. LEON.  Oh, no.  We are watching them over time and this was 28 years of follow-up.  It is called observational because we observe but manipulate the treatment that is received.

The CHAIRMAN.  Are you familiar with Heisenberg’s Uncertainty Principle?

Dr. LEON.  Well, it applies if assessing the subjects has a therapeutic benefit.

Dr. BREGGIN.  The Heisenberg—

The CHAIRMAN.  Right.  That is the point that Mr. Rodriguez raised, who has a lot of mental health background, that is if you are studying people and showing an interest in them, that affects—

Dr. LEON.  Wait, wait.

The CHAIRMAN [continuing].  In and of itself—

Dr. LEON.  Absolutely, and that is—

The CHAIRMAN [continuing].  Affects the outcome.  That is, as Mr. Rodriguez said, if they are being watched and followed and advised—

Dr. LEON.  Yeah.  That is—

The CHAIRMAN [continuing].  That may be the effect of your observation, not of whether they are receiving or not receiving treatment.

Dr. LEON.  Yeah, absolutely.  And that is the reason we included a control group in this, because they were also receiving those same assessments.  So the change in psychopathology, the change in suicide risk that might be brought about by conducting the assessments would be held constant across the treated and the untreated groups.

The CHAIRMAN.  Okay.  We will get back to that.  Thank you.

Mr. RODRIGUEZ.  Mr. Chairman.

The CHAIRMAN.  Mr. Walz?

Mr. RODRIGUEZ.  I am sorry.  I really need to add one additional question.

The CHAIRMAN.  Sure.  Go ahead.

Mr. RODRIGUEZ.  I know we are talking about veterans right now, but as it deals with kids, because I am really concerned about the medication given to kids that has never been tested on kids.  And now I just wanted to see if you care to comment on medication for kids.

Dr. LEON.  Oh, absolutely.  And I will tell you I was at several hearings on this topic at the FDA.  And when we reviewed the antidepressants and suicidality in children and adolescents, those under age 18, there was a very big difference between the data that were available because with children and adolescents, we reviewed those data and there were only a few clinical trials that ever showed efficacy in children with antidepressants.

Now, the one medication that showed efficacy in children in more than one trial was fluoxetine, Prozac.  That is the one.  And they did get an indication or approval from the FDA.  So the risk-benefit ratio when we look at most medications, in most antidepressants in children, there is not a well-established benefit.

So a very small risk really raises alarm.  However, in looking at adult data, there is a great deal of data showing a benefit of antidepressants.  There is several decades of data available showing us that there is a benefit.

So then when we look at a very small risk and a very large benefit, the risk-benefit ratio is much less alarming.  It cannot be ignored completely.  Instead what we need to do and what we insisted on when we were helping the FDA, advising the FDA in putting together the Black Box Warning, we, and I have looked at the transcripts of our meetings to confirm this, those of us who voted for a Black Box Warning, and I voted for the Black Box Warning, for antidepressants both in kids and in adults, and it is because I saw from these data, no, I cannot say there is no risk at all.  I want the clinician to be aware of the risk.  I want the patient to be aware of the risk and I want the family members to be aware.  There is a very small risk.  If we see any hint of agitation, hostility, akathisia, or suicidality, contact the physician immediately so we can deal with that problem.

The CHAIRMAN.  Thank you.

Mr. Walz?

Mr. WALZ.  Well, thank you, Mr. Chairman, for holding this hearing.  As usual, you are a strong advocate for making sure we get things right as we care for our veterans. 

And I want to thank both of you for coming.

Dr. Leon, you did mention that our ultimate goal here is to make sure we provide the ultimate care to our veterans and I am appreciative of that statement.

Just a couple things.  And it might either be the school teacher in me or the parent of a small child, I want to bring us back to something we can agree on on this.  And I appreciate listening to both sides of this.  Just a couple things.

Am I right, in the Black Box Warning on this, I am trying to get this side of it, then bring it back to the veterans’ care, the Black Box Warning just it may increase the risk of suicidal thinking?  That is where it came from.  If I am right, there was about a four percent rate amongst those children who were looked at?

Dr. LEON.  Well, yeah.  In children, I think it was at three percent.

Mr. WALZ.  And it was amongst children is where it was put out on that.  Okay.

Dr. LEON.  Okay.

Mr. WALZ.  And NIH does go on to say SSRIs can be beneficial and FDA says the benefits far outweigh the risk.

Now, my question to you, Dr. Breggin, and I am very appreciative of this on interactions and how things come together, do you believe there are any benefits or places where antidepressants should be prescribed?

Dr. BREGGIN.  I think they do more harm than good.  In this I am certainly in a position that is different than most psychiatrists who practice psychiatry and I do not want that to influence all the science I am presenting, the controlled clinical trials, on the viewpoint on the suicidality. 

In my experience the way antidepressants work, if they do work, is they cause either apathy or a mild euphoria.  We are now seeing patients who have been on these drugs for 5 or 10 years and they have lost their interest in life.  It is an incredible tragedy.  They do not love anymore.  They do not care as much.  They have lost their musical abilities.  The drugs do not have a magic way of fixing depression, which is basically a loss of hope.  Depression is where a person feels so bound down in choices, or—

Mr. WALZ.  You do not think there is any physical evidence on serotonin, and—

Dr. BREGGIN.  Oh no, sir.  There is not.  There is not.  The way that—

Mr. WALZ.  I say this coming from Minnesota, where the sun does not shine much.