Twenty-five years ago, long before the introduction of selective serotonin-reuptake inhibitors (SSRIs), the adolescent suicide rate was increasing rapidly, having tripled over the previous two decades, but the risk factors involved were unknown. Adolescents who committed suicide were regarded as misunderstood teenagers who had been under too much stress. There was debate about whether depression could occur in children, and the prevailing view was that moodiness was normal in adolescents. Furthermore, even if we could have diagnosed depression and recognized young people who were at risk for suicide, there were no empirically validated treatments to offer.

Eventually, we learned that depression did indeed affect children and adolescents. Through retrospective interviews with family members and friends, this disorder emerged as the single most important risk factor for adolescent suicide,1 although it often acted in concert with substance abuse and impulsive aggression. Adolescents who committed suicide frequently had a history of suicidal thoughts or behavior, disclosed only to a friend who was sworn to secrecy. Most commonly, adolescents killed themselves with a gun, and guns were much more frequently available to those who had died by suicide than to those who had attempted suicide but lived.

These findings suggested some straightforward approaches to prevention. Although it had been thought that people who talk about suicide don't kill themselves, these results showed that previous suicidal behavior and current suicidal thoughts are potent risk factors for suicide and must be taken seriously. The association between the availability of guns and their use in suicides suggested that guns should be removed, or at least secured. Finally, the development and testing of treatments for pediatric depression should be given high priority.

Today, we are able to identify young people who are at high risk for suicide and to offer empirically validated treatments for depression.2 It is ironic that concern about the risks posed by antidepressants has arisen now, when the adolescent suicide rate has been decreasing for a decade (see Figure), for the first time in more than half a century. This trend is accounted for primarily by a drop in the rate of suicide by means of firearms, suggesting that more restrictive gun-control laws may be partially responsible.3 A portion of the decrease may be related to better detection of depression and suicidality (suicidal ideation, behavior, or both) and the dissemination of validated treatments. There is some ecologic evidence that increases in the number of prescriptions for SSRIs for adolescents are associated with a decrease in adolescent suicide.4

Nevertheless, given findings showing a relationship between suicidality and completed suicide, one must take seriously the possibility that antidepressants might increase the risk of suicidality. And yet the concern about SSRIs in pediatric depression that has been aroused by the British Medicines and Healthcare Products Regulatory Agency (MHRA) (http://medicines.mhra.gov.uk/ourwork/monitorsafequalmed/safetymessages/ssrioverview_101203.htm) is based, in my opinion, on an overestimation of risk and an underestimation of benefit.

Current clinical practice with regard to SSRI use for pediatric depression is based on six published studies, although five unpublished studies were much less favorable. There is the most incontrovertible evidence of efficacy for fluoxetine, which had positive results in three clinical trials. For sertraline, the results were positive but modest, with a 10 percent difference in response between drug and placebo. One study of citalopram was positive and one negative, but the latter involved both inpatients and outpatients and had a very high dropout rate. There were two negative trials of venlafaxine, but the doses used were often well below the minimal therapeutic dose for adults, and when the results were stratified according to age, venlafaxine was superior to placebo among adolescents. There have been three clinical trials of paroxetine, of which only one was positive.

In addition to questioning the benefit of these drugs, the MHRA and the Food and Drug Administration (FDA) focused attention on a possible increase in the likelihood of suicidality. In response, the MHRA declared that all antidepressants except fluoxetine were contraindicated in pediatric depression. The FDA initially advised against the use of paroxetine only, since there was some evidence of efficacy for sertraline and citalopram. More recently, the FDA labeled all antidepressants with a warning about their possible potential for inducing suicidal thoughts or behavior. The FDA also recently commissioned a blind independent review of these adverse events by a consensus panel of international experts.

With regard to the main outcome, suicidality (the combination of new suicide attempts, new-onset suicidal ideation, and worsening of existing suicidal ideation), the FDA analysis, presented to an advisory committee in a public hearing September 13 and 14, 2004, found an increase by a factor of 1.8 associated with drug treatment, which translates to a difference of 1.7 percentage points between drug and placebo (3.8 percent vs. 2.1 percent). Although the difference is small, it seems likely that the effect is real, because the findings were statistically significant in aggregate and are consistent across multiple studies of various agents.

Although the initial MHRA report and the FDA analysis found that fluoxetine treatment was not associated with suicidality, an FDA analysis of a new clinical trial found otherwise — but the results help to put the benefits and risks into perspective.3 In this study, cognitive–behavioral therapy and fluoxetine treatment, alone and in combination, were compared with each other and with placebo. Fluoxetine was much more likely than placebo to result in a significant clinical improvement (in 61 percent of cases vs. 35 percent) but, according to the FDA analysis, was associated with a significant increase by a factor of 4.6 in the rate of suicidal events (8.3 percent vs. 1.8 percent). Once a suicidal event was detected, the patient was withdrawn from the trial. Although depression and suicidality are both significant risk factors for suicide, depression improved in these patients four times as frequently as suicidality developed, which seems to represent an acceptable risk–benefit ratio. Fluoxetine plus cognitive therapy was not superior to fluoxetine alone according to most measures of depression, but the combination was superior to all other treatments in reducing the intensity of suicidal thoughts. This finding suggests that the optimal treatment for suicidality in a depressed patient may be multimodal — a logical approach, given the multiple risk factors for suicide.

In addition to increasing the risk of suicidality, SSRIs are twice as likely as placebo to result in agitation and hostility. The FDA could not test whether hostility and suicidality were linked, but the drugs most closely associated with one were also most closely associated with the other. In addition, treatment with antidepressants is much more likely to unmask an underlying bipolar disorder in children or adolescents than it is in adults,5 which can result in the induction of a mixed manic and depressive state — a condition that carries a very high risk of suicidal behavior.

As SSRIs have gained in popularity, their ease of use and relatively favorable side-effect profile may have led to an overly casual approach to the treatment of depression. All depressed patients who are treated with antidepressants must be closely monitored for emergent suicidality, hostility, agitation, and mania. Families and patients must be informed about the benefits and risks of these drugs and should be educated about monitoring for emergent side effects, as recommended in recent public statements and labeling changes made by the FDA. Because children and adolescents generally metabolize antidepressants more rapidly than adults, they must receive doses adequate to achieve a clinical response. Suicidality in depressed patients may be best treated by a combination of psychotherapy and medication.

The FDA's recent analysis suggests that the risk of emergent suicidality in children and adolescents receiving SSRIs is real — but small. The FDA's advisors recommended stronger warnings in labeling and better information for patients and caregivers, but they stopped short of recommending contraindications for these drugs. However, many participants in the public hearing seemed convinced that the pharmacologic treatment of pediatric depression should be banned or severely curtailed. That would turn the clock back 25 years, to a time when the only thing we could offer the families of suicide victims was the hope that someday we would have effective treatments. Ideally, the FDA, families, and clinicians will find the right balance between the risk of suicidality and another, greater risk: the risk that lies in doing nothing.

A. Five (or more) of the following symptoms have been present during the same 2-week period and represent a change from previous functioning; at least 1 of the symptoms is either (1)depressed mood or (2)loss of interest or pleasure.

Depressed mood most of the day, nearly every day, as indicated by either subjective report or observation made by others

Markedly diminished interest or pleasure in all, or almost all, activities most of the day, nearly every day

Significant weight loss when not dieting or weight gain, or decrease or increase in appetite nearly every day

Insomnia or hypersomnia nearly every day

Psychomotor agitation or retardation nearly every day

Fatigue or loss of energy nearly every day

Feelings of worthlessness or excessive or inappropriate guilt (which may be delusional) nearly every day

Diminished ability to think or concentrate, or indecisiveness, nearly every day

Recurrent thoughts of death (not just fear of dying), recurrent suicidal ideation without a specific plan, or a suicide attempt or specific plan for committing suicide

B. The symptoms do not meet criteria for a mixed episode.

C. The symptoms cause clinically significant distress or impairment in social, occupational, or other important areas of functioning.

D. The symptoms are not due to the direct physiologic effects of a substance or general medical condition.

E. The symptoms are not better accounted for by bereavement, ie, after the loss of a loved one, the symptoms persist for longer than 2 months or are characterized by marked functional impairment, morbid preoccupation with worthlessness, suicidal ideation, psychotic symptoms, or psychomotor retardation.

Demographic Factors

Male gender

White race

Aged 24 to 35 or >50 years

Factors Associated with the Depressive Episode

Suicidal thoughts or acts

Lack of treatment or inadequate treatment of depressive episode

Presence of hopelessness

Low self-esteem, feelings of failure

More severe depressive symptoms (diminished concentration, insomnia, anhedonia, but not diminished energy level)

Recent bereavement

Presence of psychosis

Concomitant anxiety or panic attacks

Being in the first 3 months after the onset of the depressive episode

Factors Related to Personality Characteristics

Personality variables including aggression, hostility, or impulsiveness

Comorbidity

Comorbid substance use disorders

Comorbid borderline personality traits or disorder

Comorbid conduct disorder or antisocial personality traits or disorder

Comorbid general medical conditions

Bipolar disorder or mood cycling

Acute substance intoxication

Factors Revealed by History

History of suicide attempt

History of taking precautions against being discovered after an attempt

Presence of a family history of suicide

Early onset of depression, being younger at first hospitalization, more previous hospitalizations

Parental loss through death before the age of 11

Childhood history of physical or sexual abuse

Corporal punishment in adolescence

Psychosocial Factors

Being recently widowed/separated/divorced

Chronic physical illness

Social, financial, or family crisis or loss (negative life events)

Unemployment or financial problems

Lack of religious or moral constraints against suicide

Not living with a child younger than 18

Social isolation

Miscellaneous

Access to means with greater lethality

Contagion or recent exposure to suicide

Appearance: diminished self-care and grooming, quiet and withdrawn stance

Attitude: irritability, restlessness, sadness

Behavior: diminished eye contact, stooped posture, hand-wringing, psychomotor slowing or agitation

Speech: increased response latency, decreased rate and rhythm of speech

Mood: down, depressed, sad, unhappy, empty, helpless, hopeless

Affect: sad, emotionally blunted, tearful, anxious

Thought processes: slow thoughts, fewer expressed thoughts

Thought content: excessive and inappropriate guilt; thoughts of worthlessness, hopelessness, or helplessness; mood-congruent psychotic features (hallucinations or delusions); suicidal thoughts, intent, plans, or acts

Insight/judgment: impaired to good

Cognition: impairment in memory, attention, and concentration

Episode: A period lasting longer than 2 weeks (as defined by the DSM-IV-R) during which the patient is consistently within the fully symptomatic range of a sufficient number of symptoms to meet syndromal criteria for the disorder. For the clinician, this typically triggers a decision to treat.

Partial remission: A period during which an improvement of sufficient magnitude is observed that the individual is no longer fully symptomatic (ie, no longer meets syndromal criteria for the disorder but continues to evidence more than minimal symptoms). If a partial remission fails to become a full remission after a reasonable period, the clinician will typically alter treatment (by increasing the intensity of the current treatment or by augmentation).

Response: The point at which a partial remission begins. A response, unlike a partial remission, does require treatment and thus implies that the cause of the change in the patient's condition is known, which may not be a valid assumption.

Full remission: A relatively brief period during which an improvement of sufficient magnitude is observed that the individual is asymptomatic (ie, no longer meets syndromal criteria for the disorder and has no more than minimal symptoms). No increase in the intensity of the treatment regimen is required, although continuation treatment should be used to prevent relapse.

Recovery: A remission that lasts for a specified period of time. The term is used to describe recovery from the episode, not the illness per se. In the clinical setting, a declaration of recovery raises the possibility that treatment can be discontinued or, if treatment is continued, the aim is prevention of a subsequent recurrence of an episode (maintenance treatment).

Relapse: A return of symptoms satisfying the full syndrome criteria for an episode that occurs during the period of partial or full remission, but before recovery as defined above. A relapse signals a need for treatment intervention or modification of ongoing treatment.

Recurrence: The appearance of a new episode of major depressive disorder occurring during a recovery. A recurrence implies the need for treatment and a revision in the history of the course of the illness. The latter may have prognostic and treatment implications. In studies of maintenance therapy, recurrence is typically the outcome of primary interest. The term relapse (vidra supra) represents the return of the symptoms of a still ongoing but symptomatically suppressed episode, whereas the term recurrence represents an entirely new episode.

Generic Name (Trade Name)	Typical Daily Starting Dosage	Typical Daily Dosage Range
Fluoxetine (Prozac®)	20 mg	20 mg-80 mg
Paroxetine (Paxil®)	20 mg	20 mg-50 mg
Sertraline (Zoloft®)	50 mg	50 mg-200 mg
Fluvoxamine (Luvox®)	50 mg	100 mg-300 mg
Citalopram (CelexaTM)	20 mg	20 mg-60 mg
Venlafaxine (Effexor® XR)	37.5 mg	75 mg-375 mg
Bupropion (Wellbutrin SR®)	100 mg	100 mg-200 mg BID
Nefazodone (Serzone®)	100 mg	200 mg-600 mg
Mirtazapine (Remeron®)	15 mg	15 mg-45 mg