About the SA Blog Network



Critical views of science in the news
Cross-Check Home

Did Antidepressant Play a Role in Navy Yard Massacre?

The views expressed are those of the author and are not necessarily those of Scientific American.

Email   PrintPrint

Once again, antidepressants have been linked to an episode of horrific violence. The New York Times reports that Aaron Alexis, who allegedly shot 12 people to death at a Navy facility in Washington, D.C., earlier this week, received a prescription for the antidepressant trazodone in August.

Aaron Alexis, accused of shooting to death 12 people this week, was reportedly prescribed the antidepressant trazodone in August.

When I first researched antidepressants almost 20 years ago, I encountered claims that they sometimes triggered violent episodes—for example, a 1989 incident in which a Kentucky man taking fluoxetine (brand name Prozac) shot to death eight co-workers and then himself. I dismissed the claims, reasoning that, because people prescribed psychiatric drugs are disturbed to begin with, it is not surprising that a tiny fraction hurt themselves and/or others.

By 2004, however, in part because of lawsuits that forced pharmaceutical companies to disclose data on adverse effects, the FDA ordered antidepressant manufacturers to include a warning that antidepressants “increased the risk compared to placebo of suicidal thinking and behavior (suicidality) in children, adolescents, and young adults in short-term studies of major depressive disorder (MDD) and other psychiatric disorders.”

Alexis, who was 34, was reportedly seeking treatment for insomnia when he received his prescription for trazodone. Originally marketed as an antidepressant after its approval by the FDA in 1981, trazodone is also prescribed for anxiety and insomnia. Trazodone was a precursor of the extremely popular selective serotonin reuptake inhibitors (SSRIs); like the SSRIs, trazodone boosts levels of the neurotransmitter serotonin.

website maintained by the National Institutes of Health states that trazodone and other antidepressants have been associated with “new or worsening depression; thinking about harming or killing yourself, or planning or trying to do so; extreme worry; agitation; panic attacks; difficulty falling asleep or staying asleep; aggressive behavior; irritability; acting without thinking; severe restlessness; and frenzied abnormal excitement.”

Could trazodone have played some role in the Navy Yard rampage? I put this question to David Healy, a professor of psychiatry at Cardiff University in England and an authority on side effects of psychiatric medications.* He has testified in numerous court cases involving suicides and homicides whose perpetrators were ingesting antidepressants. Healy is not opposed to psychiatric medications–he prescribes them to his own patients—but he has long been active in pointing out medications’ risks. *[Healy and other health-care experts have founded an organization called RxISK to gather data on side effects of drugs. See Postscript.]

Healy responded by email that he “would need to know much more details” to judge whether trazodone might have contributed to this week’s massacre. Indeed, as the Times reported, Alexis had displayed signs of mental illness and acted violently well before being prescribed trazodone. In 2004 he fired bullets into a car in Seattle, during what he described later as a “black out” episode. So far, moreover, there are no reports that he had trazodone in his system during his Navy Yard rampage.

But Healy said that, although data on antidepressants and violence are much more scarce than data on suicide, there is evidence that “you can put healthy volunteers on these drugs and some will become violent.” A study by the Drug Safety Research Unit in Southampton of paroxetine (Paxil) and fluoxetine (Prozac) involving more than 25,000 subjects showed that one out of every 250 subjects were involved in “a violent episode,” including 31 assaults and one homicide, Healy said.

Another study involving more than 9,000 subjects taking the antidepressant paroxetine (Paxil) for depression and other disorders showed that subjects experienced more than twice as many “hostility events” as subjects taking a placebo. Healy added: “I have had clinical experience of at least one older man, with no prior history of violence, who became homicidal after a week on citalopram where the problem cleared up once treatment had stopped.”

Healy suspects that the main causal factor behind suicide and violence toward others is increased mental and/or physical agitation, which leads about 5 percent of subjects taking antidepressants to drop out of clinical trials, compared to only 0.5 percent of people on placebos.

In their excellent overview “Antidepressants and Violence: Problems at the Interface of Medicine and Law,” published in PLoS Medicine in 2006, Healy and two co-authors note that “serious violence on antidepressants is likely to be very rare.” But they call for “more clinical trial and epidemiological data to be made available and for good clinical descriptions of the adverse outcomes of treatment.”

They add: “Legal systems are likely to continue to be faced with cases of violence associated with the use of psychotropic drugs, and it may fall to the courts to demand access to currently unavailable data. The problem is international and calls for an international response.”

Antidepressants and other psychiatric drugs clearly help alleviate mental illness in some patients. But as I have written previously, such medications may on balance harm patients more than they help. We need better data on the risks that these medications pose not only to patients but also to others.

*Postscript: David Healy and other health-care experts from around the world have formed an organization called RxISK to gather data on adverse effects of pharmaceutical drugs, including violence. The RxISK website calls it “the first free, independent website where patients, doctors, and pharmacists can research prescription drugs and easily report a drug side effect–identifying problems and possible solutions earlier than is currently happening.” I hope RxISK succeeds, because we badly need it.

Post-Postscript: For more on links between psychiatric medications and violence, see this 2010 paper in PLOS One, “Prescription Drugs Associated with Reports of Violence Towards Others,” and a followup blog post by journalist Robert Whitaker, author of Anatomy of an Epidemic.

Photo: Federal Bureau of Investigation.

John Horgan About the Author: Every week, hockey-playing science writer John Horgan takes a puckish, provocative look at breaking science. A teacher at Stevens Institute of Technology, Horgan is the author of four books, including The End of Science (Addison Wesley, 1996) and The End of War (McSweeney's, 2012). Follow on Twitter @Horganism.

The views expressed are those of the author and are not necessarily those of Scientific American.

Rights & Permissions

Comments 19 Comments

Add Comment
  1. 1. 4:09 pm 09/20/2013

    The prescription-medication-equals-violence link is well-established.

    Most perpetrators of school shootings and similar mass murders in our modern era were either on, or just recently coming off of, psychiatric medications:

    * Recently confirmed: Aurora shooter James Holmes, who murdered 12 people and injured 58 in a movie theatre, was on prescription meds —the names of these medications have been redacted from court documents.

    * Columbine (April 20, 1999) mass-killer Eric Harris was taking Luvox® (Fluvoxamine maleate)

    * Patrick Purdy, who murdered five children and wounded 30 at Cleveland Elementary School in San Diego, on January 17, 1989, was using Amitriptyline (an antidepressant), as well as the antipsychotic drug Thorazine.

    * Kip Kinkel, 15, murdered his parents and then opened fire on his classmates at Thurston High School in Springfield, Oregon, May, 1998, killing two and wounding 22 others. He had been prescribed both Prozac and Ritalin.

    * In 1988, 31-year-old Laurie Dann went on a shooting rampage in a second-grade classroom in Winnetka, Illinois, killing one child and wounding six. She had been taking the antidepressant Anafranil as well as Lithium.

    * In Paducah, Ky., in late 1997, 14-year-old Michael Carneal, traveled to Heath High School and started shooting students in a prayer meeting taking place in the school’s lobby, killing three and leaving another paralyzed. Carneal reportedly was on Ritalin.

    * Jeff Weise, living on Minnesota’s Red Lake Indian Reservation, shot and killed nine people and wounded five others before killing himself. Weise had been taking Prozac.

    * Joseph T. Wesbecker, just a month after he began taking Prozac in 1989, shot 20 workers at Standard Gravure Corp. in Louisville, Ky., killing nine. Prozac-maker Eli Lilly later settled a lawsuit brought by survivors.

    Link to this
  2. 2. drshua 5:43 pm 09/20/2013

    did you also know that as people’s interest in eating ice cream increases there is also a greater likelihood of death by drowning. thus, it could be said that eating ice cream causes people to drown…it must be the case! we must warn everyone about this awful tragic coverup by the ice cream companies in the world.

    have you ever heard of a thing in science called confounds or third variable explanations? i wish i had time to explain it but i have to go do science.

    Link to this
  3. 3. drshua 5:46 pm 09/20/2013

    i will add that the third variable explanation in the classic ice cream drowning scenario is that both happen with much greater frequency due to summer weather. the real culprit is summer. likewise, the real culprit in these other cases is simply sever mental illness not medication.

    Link to this
  4. 4. sandragon 6:25 pm 09/20/2013

    I would expect problems to arise with changes in medication. For example if one were too depressed to do much of anything then the improvement that commences when he starts taking an antidepressant could be just enough to allow him to follow through on a suicide attempt. We all learn to control our thoughts and urges but changes in brain chemistry for whatever reason will interfere with that. Giving drugs without careful observation and support (which seems to be the norm) is probably dangerous for some individuals.

    Link to this
  5. 5. SAReadersince67 7:49 pm 09/20/2013

    In the larger context, it is perhaps important to recognize that primates are essentially violent. Humans are particular adept at violence, employing tools (weapons) to enhance the consequences of their violent thoughts and actions. Whether a human being is caused to be a little more or less violent by a medication, by sugar, by social competition or whatever, is not very distinguishable from the “noise” in the behavioral spectrum.

    It may also be argued that if primates were less violent, humans in particular, we would not exist to be making these observations and debate.

    What is truly amazing is that our society is not more violent given our propensities and abilities. Also, in that respect, we might consider why we have not invented more substances which are designed to make us more violent. If we can create a molecule that is untented to alleviate clinical depression, but know it can also appear at least to cause violent behavior, why have we not used this knowledge to exploit violence and aggression so that some groups may obtain more resources or offspring. Well perhaps we have done so already…

    The role of morality in human communities is off course “the” factor in moderating violence… A big subject of enduring debate…

    Link to this
  6. 6. SAReadersince67 7:51 pm 09/20/2013

    Should have read:

    The role of morality in human communities is, of course, “the” factor in moderating violence…

    Link to this
  7. 7. SAReadersince67 8:00 pm 09/20/2013

    Also should have read:
    If we can create a molecule that is “intented” to alleviate clinical depression, but know it can also appear at least to cause violent behavior, why have we not used this knowledge to exploit violence and aggression so that some groups may obtain more resources or offspring?

    I wonder if there is any correlation between propensity to violence and the accuracy of ten-finger typing skills? Could there be a role for anti-depressants in improving keyboarding skills? Or proof reading.. and I am not making light of the subject. How are skills and knowledge correlate with a persons ability to control violent thoughts and actions?

    Link to this
  8. 8. lynnoc 11:24 pm 09/20/2013

    One fact rarely mentioned –when the “news” came out that some depressed kids/adolescents might be more suicidal on antidepressants, a “black box” warning appeared on antidepressants (SSRIs, selective serotonin reuptake inhibitors) for kids and teens. MDs became afraid to prescribe. In the ten years since, as the rate of antidepressant use for depressed kids/adolescents went down, the suicide rate went significantly up. Lets get the whole story here.

    Link to this
  9. 9. rossm 4:22 am 09/21/2013

    Are people committing these crimes because they are have a condition for which a medication is prescribed, or because they have a condition that the medication cannot moderate (or has not yet moderated). A major statistical analysis is required to address the issue, not anecdotes.

    Link to this
  10. 10. thompa2 3:56 pm 09/21/2013


    The black box warning regarding increased risk of suicidality in teens and young adults has a well established explanation. People who are experiencing severe depression may exhibit something called “psychomotor retardation” where they are literally too depressed to move. Many of these people have a plan to commit suicide but simply cannot muster the motivation to do it. Initiating SSRI use in these patients can give these patients more energy before the depression begins to resolve, so paradoxically, they become more likely to carry out a suicide plan because they have more motivation but are still suicidal. This is seen more often in kids and young adults because they are more prone to compulsive behavior due to the fact that their frontal lobes have not finished developing. Thus, there is an increased risk of suicide in a very specific population of depressed patients, which are pediatric and young adult patients with severe depression showing features of psychomotor retardation. For ALL OTHER groups of depressed persons, risk of suicide goes down with appropriate pharmaceutical treatment.

    That’s why suicide rates went up when SSRI use when down, even though there’s a black box warning about increased risk of suicide.

    Link to this
  11. 11. softwarematters 7:02 pm 09/21/2013

    Another outstanding entry John. Keep up the good work!


    You have it backwards. An increase in the rate of suicide is correlated with an increase in SSRI usage,

    “Annual suicide rates for this age group increased 28 percent over this period (from 13.7 suicides per 100,000 people in 1999 to 17.6 per 100,000 in 2010)”

    “According to a report released yesterday by the National Center for Health Statistics (NCHS), the rate of antidepressant use in this country among teens and adults (people ages 12 and older) increased by almost 400% between 1988–1994 and 2005–2008.”

    You are entitled to your own opinion, but not to your own facts!

    Link to this
  12. 12. Nickihealing 9:53 am 09/22/2013

    Ultimately, once everyone agrees that the medication approach is a problem, the question is what can you do to help people in extreme distress? I can tell you that after decades of being paralyzed by the medication-only approach I finally searched outside of psychiatry, into mind-body modalities and energy healing and guess what, I am for the most post cured and healed. I am not “doing the best I can given that I am emotionally unwell” but totally good, in fact healthier emotionally and physically than most people I meet. I don’t know when psychiatry will shed it’s biological model complete with (meaningless) diagnoses and labels and instead opt to help people find true health, but I hope it’s soon. Until then there will continue to be too many lost lives, wasted years, hopelessness seething on the edges of it all.

    Link to this
  13. 13. BigYin 12:31 pm 09/22/2013

    A significant metabolite of the antidepressant Trazodone (now widely prescribed off-label for insomnia) is m-chlorophenylpiperazine (mCCP).

    There is always the possibility of a Trazodone-related adverse effect, especially in higher doses and when co-prescribed with other psychiatric drugs.

    “Adverse effects reported after the abuse of mCPP include anx­iety, dizziness, hallucinations, nausea, warm and cold flushes, migraine and panic attacks. mCPP often induces severe hallucina­tions and nausea…..The agonist properties of mCPP at 5-HT2 receptors can explain its hallucinogenic features since other hallucinogenic substances such as lysergic acid diethylamide exert their effects through activation of these receptors (Bossong et al., 2005,2010).”

    Arbo MD, Bastos ML & CarmoHF. Piperazine compounds as drugs of abuse. Drug & Alcohol Dependence. 122, 174-185

    Link to this
  14. 14. thompa2 8:25 pm 09/22/2013


    Ignoring your unsubtle snark at the end of your post, let me first agree with you: yes, the last decade increasing SSRI use is correlated with increased suicide.

    Correlation is NOT causation. You can look in any basic medical textbook, or advanced psychiatric textbook, and find that SSRIs and other, newer generation antidepressants statistically decrease suicidal ideation and behavior. The data is pretty robust. There is no controversy in the medical community about it. That alone should tell you something.

    The problem with suicide is that it’s a phenomenally complicated thing. For example, economic recessions are associated with increased rates of suicide. We had a pretty major recession in the last decade, which for many people is still ongoing. So it would be easy to make the argument that the increased use of SSRIs has help prevent MORE people from committing suicide, and that without them we would have seen a much sharper rise in the numbers.

    Superficially, your argument “There are more SSRIs in use and more suicides, therefore antidepressants increase likelihood of suicide” makes sense. But that’s about it. Drowning is also positively correlated with consumption of ice cream. Why? Because during the summer, when it’s hot out, people go out for ice cream and they go swimming…where some of them drown. Thus ice cream and drowning are positively correlated, but have nothing to do with each other.

    Antidepressants treat depression in many patients, which lowers their risk of suicide. Depression is hardly the only risk factor, but it is an important one. Let’s try to stop vilifying life saving medications. They don’t work for everyone, but they do work for a lot of people, and there is no good data to suggest they increase suicidality with the sole exception of the group I described in my earlier post.

    Actually, there is one other group, which is bipolar patients. Nearly 10% of individuals receiving a first-time diagnosis of depression are actually bipolar. SSRIs and other antidpressants can, in some bipolar patients, “flip” them from depressed to manic. If there’s anything that has a stronger connection to suicide than depression, it’s mania. So there is that.

    Link to this
  15. 15. Chryses 6:02 am 09/23/2013

    thompa2, (14)

    As documented here ( ), beginning at post 28 and going on (and on, and on …), I fear you’ll find some posters on this topic are indifferent to facts that fail to fit their Beliefs.

    Link to this
  16. 16. portland17 4:48 pm 09/23/2013

    Sorry, Thompa, the concept that drug activation causes inert suicidal people to get the energy to act does not hold up to scientific scrutiny. First off, any ethical trial of antidepressant drugs screens for suicidal people and eliminates them from the trial, for obvious reasons having to do with patient safety. We don’t want to do experiments that end up with dead people! So the increased suicide rates over placebo can be assumed to be from people who were not overtly suicidal at the beginning of the trials. Sure, there may have been a few who didn’t talk about their urges, but overall, the rate should be LOWER for these pre-selected subjects than for the general population. And yet there were more suicides/suicidal thinking that predicted and in comparison to placebo.

    Additionally, there have been studies with healthy volunteers that have shown increasing suicide rates. It’s not a huge effect, but it does exist and is documented scientifically, and is clearly not due to people who were thinking about it getting the energy to act.

    As for aggression, I’ve seen examples myself in my work as an advocate for foster youth – kids who are not aggressive but become so after starting SSRI antidepressants or stimulants, and who become less aggressive if I can ever convince the prescribers to stop the drugs. I have seen multiple instances of this in my work. Again, not extremely common, but neither are they terribly rare.

    I would imagine a very special set of circumstances would have to come together for a person to become fully homicidal on these drugs. For instance, a number of the students who went on rampages (Columbine and Red Lake, for instance) were bullied and isolated. But most bullied and isolated kids don’t go on shooting rampages. I remember lots of kids being bullied in my school, but you never heard anything about school shootings when I was a kid. What big changes have happened in our society since then? Well, lots, but one of them is the heavy use of psychiatric drugs, especially in kids.

    Don’t be so quick to dismiss this possibility out of hand. There are way too many correlations to consider it a “conspiracy theory” or mere mindless correlation (like the ice cream example). I think it’s something we really need to take a hard look at, but most of the US establishment is turning a blind eye to this. Not saying we know that these things are connected, but we don’t know that they aren’t, either. Any decent scientist who cared about the people they’re trying to help would want to look into this further. Kudos to the author for finally putting the question out there in the public eye.

    —- Steve

    Link to this
  17. 17. softwarematters 10:38 pm 09/23/2013


    Our dear Steve did a terrific job debunking your nonsense. But I have an even better way you debunk this,

    “You can look in any basic medical textbook, or advanced psychiatric textbook, and find that SSRIs and other, newer generation antidepressants statistically decrease suicidal ideation and behavior. The data is pretty robust. There is no controversy in the medical community about it. ”

    It comes directly from the FDA .

    You should have said that “There is no controversy in the BIG PHARMA BRIBED medical community about it”, to which I would say: sure!

    Now, while it is true that correlation does not imply causation, the data that correlates SSRI usage with increase in suicide rate is just a piece of the puzzle. The other is the studies Steve mentioned that pushed the FDA to ask the manufacturers of antidepressants to include the black box warning. If what the black box warning is true, the expected effect of massive usage of antidepressants is an increase in suicide rates. That is precisely what the data shows.

    Of course, for those whose prime motivation is Big Pharma bribes (and this year again, psychiatrists topped the list of Big Pharma payments), the data and the scientific evidence are irrelevant if they get in the way of the dollar.

    Link to this
  18. 18. PaulJScott 10:33 am 09/24/2013

    The small risk of suicidal behavior or violence (a form of suicidal behavior) associated with some Rx drugs is unfortunately not as easily dismissed as a confound or third variable as some commenters here would suggest. As David Healy’s work has shown the risk departs and reappears in challenge-dechallenge-rechallenge experiments, is dose-specific, and has been demonstrated (indeed was first discovered with reserpine for hypertension) in populations using the medications for no mental health problems whatsoever. The long list of non-psychiatric treatments that have warranted a BBW should speak for this question as well.

    The notion that the BBW caused a drop in prescriptions and a rise in suicides is folk wisdom that has been thoroughly debunked save for all but one highly creative bio-epidemiologist partial to the cause. The notion that SSRI’s prevent suicide is simply not proven; WHO, for one, disagrees. Indeed, in the clinical trials the placebo arm was the far less self-destructive place to be, in terms of numbers if not statistical sugnificance. Suicide is a rare event and it can be argued that the trials are not powered sufficiently to determine risk, perhaps by design. This question about trazedone is legitimate and worthy of further examination.

    Link to this
  19. 19. Chryses 10:25 pm 09/24/2013


    Res ipsa loquitur, eh?

    Link to this

Add a Comment
You must sign in or register as a member to submit a comment.

More from Scientific American

Email this Article