SSRIs, Suicide Warnings & the FDA: What the Record Shows

By The Recover Editorial Team | Updated October 19, 2026

If you or someone you love is taking an antidepressant right now

Do not stop an SSRI suddenly because of anything you read here. Stopping abruptly can cause discontinuation symptoms and a return of depression. Any change should be planned with the prescriber.

New agitation, severe restlessness, or new or worsening thoughts of suicide after starting, raising, lowering, or stopping an antidepressant should be reported to the prescriber the same day.

In a crisis, call or text 988 (Suicide & Crisis Lifeline) or call 911. More options are on our crisis resources page.

Key Takeaways

Since October 2004, every antidepressant sold in the United States has carried a boxed warning about increased suicidal thinking and behavior in children and teens. In 2007 the warning was extended to young adults through age 24.
The FDA’s pooled analysis of 24 pediatric trials found suicidal thinking or behavior in about 4% of young patients on antidepressants versus about 2% on placebo. No completed suicides occurred in those trials.
An FDA medical officer, Dr. Andrew Mosholder, reached a similar conclusion in 2003 but was not allowed to present it at the February 2004 advisory meeting, a decision that triggered congressional investigations.
Akathisia, a drug-induced state of intense inner restlessness, is a recognized SSRI side effect and one proposed pathway between these drugs and sudden suicidal behavior, though the size of that link is still debated.
Whether SSRIs work better than placebo is a real scientific dispute, not a settled “they don’t work.” The largest analysis to date found all 21 antidepressants studied outperformed placebo, while other researchers argue the margin is clinically small for many patients.
The warnings themselves had unintended effects: antidepressant use among young people fell sharply, and at least one large study found a simultaneous rise in suicide attempts.

Why This History Still Matters

Selective serotonin reuptake inhibitors (SSRIs) such as fluoxetine (Prozac), sertraline (Zoloft), paroxetine (Paxil), citalopram (Celexa), and escitalopram (Lexapro) are among the most widely prescribed medications in the country. They are used for depression, anxiety disorders, panic disorder, PTSD, and obsessive-compulsive disorder, and they are a routine part of care for people in addiction recovery who also live with a mood or anxiety disorder (often called dual diagnosis).

For most people who take them, SSRIs are tolerated without serious problems. But for a minority, especially young people and anyone in the first weeks of treatment or after a dose change, these drugs have been linked to agitation and to new or worsening suicidal thinking. How the U.S. Food and Drug Administration came to acknowledge that risk took roughly fifteen years, several contentious public hearings, a suppressed internal report, and multiple congressional inquiries.

This article lays out that record as accurately as we can, drawing on FDA actions, congressional correspondence, peer-reviewed research, and court documents. It also separates what is documented from what is still argued, because families making decisions about these medications deserve both the warning signs and the context.

Akathisia: The Side Effect at the Center of the Debate

Akathisia, from the Greek for “inability to sit,” is a movement disorder most often associated with antipsychotic drugs. It is also a recognized side effect of SSRIs. People describe it as an intense, hard-to-name inner restlessness, sometimes with pacing, fidgeting, or an inability to stay still, and sometimes with very little visible movement at all.

Clinicians have long been concerned about akathisia because it can be profoundly distressing. A pharmacy review in the Journal of the American Pharmacists Association noted that akathisia can be very bothersome and may give rise to suicidal thoughts, and identified risk factors that include recent SSRI dose increases, use of more than one akathisia-causing drug, a prior episode of akathisia, and brain injury. Clinicians can screen for it with a standard tool, the Barnes Akathisia Rating Scale.

Two cautions belong here. First, akathisia is frequently missed or mistaken for worsening anxiety, which can lead to a dose increase that makes it worse. Second, the strength of the causal link between akathisia and suicidal behavior is not settled. Some reviews treat it as a key mechanism; a systematic review of antipsychotic-induced akathisia concluded the evidence could not reliably establish that link. What is not in dispute is that new, severe restlessness after starting or changing an SSRI is a reason to call the prescriber promptly.

You may encounter other terms in advocacy material, such as “medication spellbinding” or “intoxication anosognosia,” used to describe people who do not recognize that a drug is changing their behavior. These are concepts proposed by individual critics, notably psychiatrist Peter Breggin, rather than recognized clinical diagnoses. They describe a real concern (people often cannot see their own drug-induced changes, which is why families’ observations matter), but they are not established medical conditions.

A Timeline of the FDA Hearings and Warnings

1990 to 1991: The first alarm

Public concern about Prozac and suicidality grew after a 1990 case report by Harvard psychiatrists Martin Teicher, Carol Glod, and Jonathan Cole described intense, new suicidal preoccupation in several patients taking fluoxetine. On September 20, 1991, the FDA’s Psychopharmacologic Drugs Advisory Committee met to consider whether antidepressants could cause or intensify suicidality or violent behavior.

The committee concluded there was not credible evidence of a causal link. The FDA did not close the question entirely: it stated it could not dismiss the possibility that antidepressants, or fluoxetine specifically, might cause or intensify injurious behavior, and asked the manufacturer to develop further studies. Later litigation produced documents that critics argue the committee never saw, including earlier German regulatory concerns about Prozac and agitation.

2003 to February 2004: The pediatric data and the Mosholder report

Interest revived when pediatric trial data, submitted under a law that rewarded companies for testing drugs in children, became available. FDA medical officer Dr. Andrew Mosholder of the Office of Drug Safety analyzed those data and became the first FDA reviewer to state openly that antidepressants increased the risk of suicidal thoughts and behaviors in children and adolescents.

Mosholder was scheduled to present his findings at the February 2, 2004 advisory meeting. He was told he would not be allowed to. The FDA later acknowledged to the BMJ that he was prevented from presenting and that, if questioned at the meeting, he could respond only from a script approved by his supervisors. FDA officials said the concern was that the committee might mistake one reviewer’s conclusions for the agency’s official position, and that the agency wanted an independent reanalysis of how suicidal events had been classified, which it commissioned from suicidality experts at Columbia University. FDA’s then-psychiatric drug team leader, Dr. Thomas Laughren, told the BMJ that the underlying risk data were presented and only Mosholder’s conclusions were withheld.

After the report was leaked to the press, the agency opened an investigation into the leak. Senator Charles Grassley, who chaired the Senate Finance Committee, wrote to the FDA questioning both the decision to sideline Mosholder and the resources spent on the leak inquiry. On March 22, 2004, the FDA issued a public health advisory on antidepressants and suicidality in young people.

September 2004: Congress and the boxed warning

In September 2004, a House Energy and Commerce subcommittee examined the FDA’s handling of the issue. Committee chairman Joe Barton asked what harm there would have been in letting Mosholder present his data, and testimony revealed that signals of suicidality in pediatric trials had appeared as early as 1996 and 1997 without aggressive follow-up. Dr. Robert Temple, then director of FDA’s Office of Drug Evaluation I, defended the agency’s handling.

That same month, after the Columbia reanalysis confirmed the signal, the advisory committee voted 15 to 8 to recommend a boxed warning, the FDA’s strongest label warning. The FDA ordered the warning on October 15, 2004. Although the hearings focused on SSRIs and SNRIs, the warning was applied to all antidepressants so that clinicians would not simply switch young patients to older drugs.

December 2006 to May 2007: Extending the warning to young adults

On December 13, 2006, the committee reviewed a much larger analysis of adult trials covering about 77,000 patients. It found a slight increase in suicidality among adults 18 to 24, roughly five additional cases per 1,000 patients, and a decrease among older adults, with people 65 and older showing reduced risk. The committee voted 6 to 2 to extend the warning, and each yes vote came with the condition that the label also caution about the dangers of untreated depression.

On May 2, 2007, the FDA asked manufacturers to extend the boxed warning to age 24 and, in a then-unusual step, to add language noting that depression itself raises suicide risk. That warning remains on antidepressant labels today.

What the Numbers Actually Showed

Precision matters here, because both critics and defenders of SSRIs tend to round the numbers in their favor.

Children and adolescents: The FDA’s pooled analysis of 24 short-term trials in about 4,400 young patients found suicidal thinking or behavior in roughly 4% of those on antidepressants versus 2% on placebo. That is a doubling of relative risk, but an absolute difference of about two percentage points. No completed suicides occurred in these trials.
Young adults 18 to 24: A small increase, around five additional cases per 1,000 patients.
Adults 25 to 64: No increase was found.
Adults 65 and older: Risk of suicidality was lower on antidepressants than on placebo.
Timing: The label emphasizes the first one to two months of treatment and any period of dose change as the highest-risk windows.

The trials measured suicidal thoughts and behaviors, not deaths, and most were short and excluded people at high suicide risk. That means the data cannot tell us the full real-world rate in either direction, which is part of why the issue remains contested.

Industry Conduct: What Court Records Show

Some of the strongest evidence that manufacturers shaped the public picture of these drugs comes not from hearings but from litigation and federal prosecutions.

Forest Laboratories (Celexa and Lexapro). In September 2010, Forest pleaded guilty to criminal charges related to the promotion of Celexa, paying more than $39 million in criminal fines and about $149 million to settle civil False Claims Act claims, and entered a five-year corporate integrity agreement. Federal prosecutors alleged Forest promoted pediatric use while failing to disclose a negative pediatric study run by its partner Lundbeck. Documents later unsealed in consumer litigation include a March 2000 email exchange about how Forest described a packaging error that had unblinded patients in its pediatric citalopram trial. When a colleague called the draft wording to the FDA a euphemism, Forest regulatory manager Amy Rubin replied that part of her job was to create “‘masterful’ euphemisms to protect Medical and Marketing.” That wording was ultimately sent to the agency.

GlaxoSmithKline (Paxil). In 2012, GSK agreed to pay $3 billion to resolve federal criminal and civil charges that included the unlawful promotion of Paxil for patients under 18. Documents unsealed in earlier Paxil litigation led Senator Grassley to ask GSK in 2008 when it first knew of an association between Paxil and suicidality.

State prescribing guidelines. Allen Jones, an investigator in the Pennsylvania Office of Inspector General, uncovered payments from drug companies to a state pharmacist involved in developing medication guidelines that favored newer, more expensive psychiatric drugs. Jones said he was told to back off and that drug companies wrote checks to politicians on both sides of the aisle. He was removed from his job after speaking to the press, later pursued a successful whistleblower case in Texas against Johnson & Johnson over its antipsychotic Risperdal, and the Pennsylvania pharmacist was eventually convicted of accepting kickbacks. His case concerned antipsychotic marketing rather than SSRIs specifically, but it became a widely cited example of how industry money reached the people setting prescribing policy.

User Fees and the Question of Regulatory Capture

A structural issue runs underneath this history. The Prescription Drug User Fee Act of 1992 (PDUFA) allowed the FDA to collect fees from drug companies to fund the review of new drug applications, in exchange for agreed performance goals such as faster review times. Congress has reauthorized the law every five years since.

The share of drug-review funding that comes from industry has grown substantially. According to the Congressional Research Service, user fees provided about 65% of the FDA’s Human Drugs Program budget in fiscal year 2016, and a later federal analysis put the figure at 66% in fiscal year 2022.

Critics argue that this dependence creates an incentive to prioritize speed and industry relationships over safety, and they point to the movement of officials between the agency and the companies it regulates. Supporters counter that user fees dramatically shortened review times for important drugs, that fee-funded staff cannot be directed by any individual company, and that Congress, not industry, sets the program’s terms. Both arguments are part of the legitimate policy debate. What the SSRI record does show is that the agency’s safety apparatus was slow to act on an emerging signal and that its leadership chose to limit how one of its own reviewers’ conclusions reached the public.

The Families Who Testified

The 2004 and 2006 hearings were unusual for the number of parents, spouses, and survivors who spoke during public comment, many describing a loved one who became suddenly agitated and then died by suicide or acted violently shortly after starting or changing an antidepressant. Former pharmaceutical sales representative and author Gwen Olsen, patient-rights advocate Vera Sharav of the Alliance for Human Research Protection, and Harvard psychiatrist Dr. Joseph Glenmullen, author of Prozac Backlash, were among the critics who urged stronger warnings. Some speakers directed their anger at named FDA officials, telling them they bore responsibility for children’s deaths.

Those testimonies did not, on their own, prove causation in any individual case, and that was never their purpose. What they did was force the agency to look harder at its own data, and the boxed warning that followed is a direct result.

Do SSRIs Work? The Efficacy Debate in Plain Terms

A common claim in critical material is that SSRIs simply do not beat placebo. The evidence is more nuanced than that.

A 2008 meta-analysis by Irving Kirsch and colleagues, published in PLOS Medicine and based on all trial data submitted to the FDA for four newer antidepressants, including unpublished studies, found that the difference between drug and placebo was very small for people with moderate depression and grew larger only at the most severe end. The authors argued the benefit for most patients fell below common thresholds for clinical significance, and that the larger effect in severe depression came mainly from a weaker placebo response rather than a stronger drug response.

A 2018 network meta-analysis in The Lancet led by Andrea Cipriani, covering 522 trials and about 116,000 adults, found that all 21 antidepressants studied were more effective than placebo for the acute treatment of major depression, with effects ranging from small to moderate. Even its authors noted that this does not mean medication should always be the first-line treatment.

Read together, these studies support a middle position: SSRIs do outperform placebo on average, the average advantage is modest, and the benefit is more reliable in more severe depression. That makes individual decisions about whether to start, continue, or taper a medication genuinely personal, and best made with a clinician who knows the patient’s history.

The Unintended Consequences of the Warnings

The boxed warning also has to be judged by what happened after it. Following the 2003 to 2004 advisories and heavy media coverage, antidepressant use among young people dropped sharply. A 2014 study in the BMJ led by Christine Lu of Harvard Medical School, covering about 2.5 million young people in 11 health plans, found that psychotropic drug poisonings, used as a proxy for suicide attempts, rose by about 22% among adolescents and 34% among young adults over the same period, an absolute increase of a few cases per 100,000 people.

That study was itself challenged. Columbia psychiatrist Mark Olfson and others noted that psychotropic poisonings are an imperfect stand-in for suicide attempts, and the authors found no change in completed suicides. Still, a later systematic review in Health Affairs concluded that reductions in depression diagnosis, therapy, and medication after the warnings support reevaluating them. The American Psychiatric Association has opposed the warning since its introduction, arguing that untreated depression is the larger danger.

The takeaway for families is not that the risk was imaginary or that the warning was wrong to exist. It is that the right response to the warning is closer monitoring, not avoidance of care.

What Patients and Families Should Watch For

The FDA’s antidepressant labeling asks patients, families, and caregivers to watch closely for the following, especially in the first months of treatment and whenever the dose goes up or down:

New or worse depression, anxiety, or panic attacks
Agitation, restlessness, pacing, or an inability to sit still
Trouble sleeping
New or worse irritability, anger, aggression, or acting on dangerous impulses
An extreme increase in activity or talking (possible mania)
Thoughts about suicide or dying, or any attempt
Other unusual changes in behavior or mood

If any of these appear, contact the prescriber promptly; for thoughts of suicide or any immediate danger, call or text 988 or call 911. Family members are often the first to notice changes the patient cannot see in themselves, so it helps for a trusted person to know when a new medication or dose change begins.

Equally important: do not stop an SSRI abruptly. Discontinuation can cause dizziness, flu-like symptoms, irritability, sleep problems, and sensory disturbances, and stopping suddenly raises the risk of relapse. A clinician can plan a gradual taper if stopping is the right choice.

When Mental Health and Substance Use Overlap

Many people searching for addiction treatment are also taking, or have been prescribed, an antidepressant. Alcohol and other drugs can worsen depression, interact with medications, and make side effects such as agitation harder to recognize. Programs that treat both conditions together, through dual diagnosis care, an intensive outpatient program, or telehealth therapy, can coordinate medication decisions with the rest of a person’s recovery plan.

The Recover is a behavioral health referral network and editorial publication. We do not prescribe medication or provide treatment, but we can help you find licensed providers who offer psychiatric evaluation and dual diagnosis care. Call 888-510-3898, or explore our guides on mental health resources, helping a loved one, and family support.

Frequently Asked Questions

Do SSRIs carry a black box warning?

Yes. Since 2004, all antidepressants in the U.S. carry a boxed warning about increased risk of suicidal thoughts and behaviors in children and adolescents. In 2007 it was extended to young adults through age 24, and it also notes that depression itself increases suicide risk.

What is SSRI-induced akathisia?

Akathisia is a drug-induced state of intense inner restlessness, sometimes with pacing or an inability to sit still. It is a recognized SSRI side effect, most likely early in treatment or after a dose change, and should be reported to the prescriber promptly.

Who was Andrew Mosholder?

Dr. Andrew Mosholder was an FDA medical officer who analyzed pediatric antidepressant trials in 2003 and concluded the drugs raised the risk of suicidal thoughts and behaviors in young people. He was prevented from presenting those conclusions at a February 2004 advisory meeting, which led to congressional investigations.

Do SSRIs work better than placebo?

On average, yes, but modestly. A 2018 analysis of 522 trials found all 21 antidepressants studied outperformed placebo, while a 2008 analysis of FDA data found the advantage was small except in very severe depression.

Should I stop my antidepressant because of the warning?

Not on your own. Stopping suddenly can cause discontinuation symptoms and relapse. Talk with your prescriber about any concerns; if you have thoughts of suicide, call or text 988 right away.

SSRIs, Suicide Warnings, and the FDA: What the Public Record Actually Shows

Key Takeaways

Why This History Still Matters

Akathisia: The Side Effect at the Center of the Debate