Antidepressants and Adolescent Brain Development

Emily Karanges; Iain S McGregor


Future Neurology. 2011;6(6):783-808. 

In This Article

Antidepressant Treatment of Children & Adolescents

Rates of antidepressant prescription have increased over the past two decades, with antidepressants now the most commonly prescribed class of medications in the USA.[44] Prescription of antidepressant drugs, particularly the SSRIs, also increased over this period in pediatric and adolescent populations,[16,44] although declines followed the 2004 US FDA warnings on the hazards of antidepressant use in children and teenagers.[45] Despite these declines, approximately 2.5% of US children under the age of 18 years are thought to currently receive antidepressant drugs.[44,45]

One factor driving these high rates of antidepressant prescription is the difficulty in finding effective treatments for depression in children and adolescents. While psychotherapeutic options such as cognitive–behavioral therapy and interpersonal therapy show some benefit,[46] they tend to be slow to take effect and are less effective in cases of greater severity or with externalizing comorbid conditions.[47,48] TCAs and monoamine oxidase inhibitors (MAOIs) are not recommended in young people due to their unfavorable side-effect profile, cardiotoxicity, lack of efficacy and association with fatal overdose.[48,49] Consequently, the SSRIs and to a certain extent, SNRIs are widely used as first-line treatments for young people with depression.

The SSRI Controversy

In recent years, the use of SSRIs in the treatment of childhood and adolescent depression and anxiety has become a topic of considerable debate. This debate is concerned with two issues, namely efficacy (i.e., SSRIs, with the possible exception of fluoxetine,[50] appear to have minimal efficacy in young people with depressive disorders[51,52]) and safety (i.e., SSRIs may cause serious psychiatric side effects in children and adolescents, being associated with worsening of depression and increased risk of suicidal ideation and behavior,[53–56] particularly during the early stages of treatment[57]). There are some indications that the SNRIs and TCAs may carry similar risks in adolescent populations.[58]

Concerns about the safety of the SSRIs began in the early 1990s with several case reports detailing the emergence of suicidal preoccupation and deliberate self-harm in adults and young people treated with fluoxetine.[59,60] Subsequent investigations largely dismissed safety concerns in adults: many well-controlled studies revealed that, overall, adults treated with SSRIs had a similar or lower risk of suicidal ideation and attempts compared with those treated with placebo.[53,54,61] However, there is a recognition that as many as 50% of adult patients do not show a clinically significant response to SSRI treatment,[62,63] and a subset are susceptible to treatment-emergent suicidality.[54]

By contrast, concerns about pediatric suicidality due to SSRI treatment has intensified. In the late 1990s, the FDA requested well-controlled pediatric studies from the manufacturers of various antidepressants in an attempt to resolve the issue.[301] Although no completed suicides occurred in any of these studies, the FDA reviewers noted that many of the reports suggested an increased risk of suicidality among treated children and adolescents compared with placebo-treated controls. Furthermore, many of the studies reported little evidence of efficacy.[301]

The most concerning findings, perhaps, were associated with the SSRI paroxetine. In 2001, a controversial and apparently ghostwritten paper published in the Journal of the American Academy of Child and Adolescent Psychiatry under the authorship of Keller and colleagues presented the results of SmithKline Beecham's Study 329 on paroxetine in adolescent major depression (Box 1).[64] Contrary to the misleading claims of the published paper that "paroxetine is generally well-tolerated and effective for major depression in adolescents",[65] Study 329 actually showed evidence of elevated rates of adverse effects, including suicidal ideation/gestures, worsening depression and aggression in paroxetine-treated adolescents. Furthermore, paroxetine failed to show efficacy on the two predefined primary measures. Later studies supported these initial findings.[50,66,67] In 2003, the FDA recommended against the use of paroxetine in depressed children and adolescents.[68] The UK Medicines and Healthcare Regulatory Agency (MHRA) took a more drastic approach, prohibiting its use entirely in those under 18 years of age.[302]

Paroxetine was not the only antidepressant to show such effects. Although some studies demonstrated efficacy of sertraline[69] and citalopram[70] for the treatment of depression in young people, the majority showed no benefit over placebo.[50,51,67] Furthermore, elevated risk of suicidal ideation and other potentially related adverse effects (e.g., mania, hypomania, agitation and aggression) have been reported in pediatric trials involving sertraline, citalopram and fluvoxamine.[55,61,71] The only exception appears to be fluoxetine, which demonstrates efficacy in pediatric randomized controlled trials[72,73] and appears to be associated with a lower risk of suicide-related adverse effects than other SSRIs.[50,55] Consequently, fluoxetine is currently the only antidepressant approved for the treatment of depression in pediatric patients in the UK and USA.[301] Even so, questions remain about the safety and efficacy of fluoxetine in adolescent populations.[74]

Although the pediatric use of all other SSRIs is contraindicated in the UK,[303] the FDA stopped short of such a move in light of the limited treatments available for depression in young people. Instead, in 2004 the FDA introduced a black box warning on all SSRIs, notifying prescribers and consumers of the potential increased risk of suicidal ideation and behavior in people under 18 years of age.[75] This warning was modified in 2007 to include people aged 24 years and under, following evidence that the increased risk extends into young adulthood.[54]

Nevertheless, the association between SSRIs and suicidality in pediatric patients remains uncertain. Although evidence from randomized controlled trials appears to point to a causal relationship, methodological shortcomings relating to sample selection[76,77] suggest that the results should be treated with caution. The rarity of suicide also necessitates the use of less definite measures of suicidality in these trials, such as attempts and ideation, which may not necessarily predict suicide completion.[78] In addition, some epidemiological studies suggest a negative relationship between SSRI prescription and suicide rates,[76,79] although these claims are contested.[78,80–82]

Overall, however, the evidence suggests that adults and young people respond differently to antidepressant drugs, with SSRIs less likely to show efficacy and more likely to lead to adverse psychiatric effects in those under 25 years of age. The mechanism underlying these effects is unknown, although pharmacokinetic differences[83] and increased susceptibility to activating side effects[84] have been suggested to play a role in the emergence of suicidal behavior. Studies investigating developmental differences in neural responses to antidepressants have provided additional clues. These studies will be reviewed in the following sections. Summaries of the key preclinical and pharmacogenomic findings are presented in Table 1 & Table 2, respectively.

Neural Effects of Antidepressants in Adolescents: Considerations in Reviewing the Literature

Why Animal Studies? Given the complications involved in the investigation of brain structure and function, it is not surprising that the majority of studies exploring developmental differences in neural effects of antidepressants have employed laboratory animals. In addition to this prime advantage, animal studies overcome many limitations associated with the use of clinical samples, such as attrition, prior, current and future treatment exposure, difficulties in cause–effect determination and the prohibitive time span of studies looking for long-term effects.[14] In addition, clinical trials often employ patients with a wide age range, limiting their ability to detect developmental differences.[85] By contrast, animal studies allow for examination of effects over a clearly restricted age range with key variables under strict experimental control.

Of course, the limitations of animal studies cannot be ignored. Extrapolation to the human condition can be complicated by species differences, dose selection and route of drug administration. These issues are of particular relevance when conducting developmental comparisons.[12] For example, age-related pharmacokinetic differences may produce dramatically different drug concentrations in the brain and plasma of adolescent rodents compared with those seen in adults given an equivalent dose.[86,87] Strain differences can also complicate conclusions drawn from animal studies as different strains often differ in baseline behavior and neurochemistry, as well as responsiveness to antidepressants and other drugs.[88,89]

Regardless, adolescent rodents provide some surprising analogs of the behavioral response of young persons to antidepressants, generally displaying behaviors reflective of minimal efficacy and increased risk of adverse effects. For example, SSRI administration to adolescent rodents has been associated with increases in anxiety-like[90] and depression-like behaviors,[87] minimal antidepressant-like effects (see Figure 1[86]) and decreased sociability.[86]

Figure 1.

Effects of paroxetine (10 mg/kg in drinking water for 22 days) on depression-like behavior in the forced swim test in adolescent and adult rats.
The number of 5-s intervals throughout the 5-min test period in which the specified behavior (swimming, climbing or immobility) is the dominant behavior.
*Significant treatment effect compared with age-matched controls (p < 0.05).
Ado: Adolescent; CON: Control; PRX: Paroxetine.
Adapted with permission from [86].

Normal Animals versus Animal Models of Depression

An additional advantage of animal studies is the ability to separate drug effects from the underlying disease state, allowing easier detection of adverse neural and behavioral effects.[14] For example, animal studies have an important role to play in resolving the recent controversy on the role of antipsychotic drugs in the regional brain atrophy observed in schizophrenic patients.[91] Similarly, studies employing 'normal' animals can help to disentangle the effects of antidepressant drug exposure from the effects of the underlying depressive or anxious states. Furthermore, although most people who are prescribed antidepressants suffer from a mood disorder, a significant proportion of prescriptions are for alternative diagnoses such as anxiety disorders, eating disorders, substance abuse, dementia, headache, fibromyalgia and chronic pain.[92,93] Concerns about off-label prescribing, overdiagnosis and overprescription[94] suggest the use of 'normal' animals may be increasingly relevant to the clinical situation.

Nonetheless, the primary aim of therapeutic drugs is to normalize aberrant behavior and/or brain function, and this cannot be examined in normal animals.[14] Animal models of depression have been developed in an attempt to reflect the etiology of depression and its neural correlates. Given the association between stressful life events and the development of depression in humans (see [95,96] for reviews), many paradigms expose standard or genetically susceptible strains to early-life or chronic stress.[97] Use of such models facilitates detection of interaction effects occurring between the disease state and the drug treatment, whereby the treatment has one effect in a model of depression and an opposite (or null) effect in normal animals (e.g., [98]). The use of a combination of animal models, normal animals and clinical studies is needed to obtain a full picture of drug actions.

Timing of Drug Administration & Outcome Assessment

Critical or sensitive periods are time-limited windows when development requires, or is strongly influenced by, certain environmental factors.[6,14] The timing of a critical or sensitive period is influenced by the developmental trajectory of the affected system, and even small shifts in timing may dramatically alter the behavioral or neural effects of a drug. Indeed, a shift of timing by 1 week during a key period of noradrenergic development alters the antidepressant-like response of rats to TCAs from nonresponsive at P21 to responsive at P28.[99]

Equally important is the timing of outcome assessment. Conclusions regarding beneficial and/or adverse effects of adolescent drug exposure will differ depending on whether the outcomes are assessed during treatment, shortly after treatment cessation or following an extended phase of drug washout. Andersen and Navalta propose an elegant model describing the 'equal, but opposite' enduring effects of developmental drug exposure:[13,14,100] although a drug may produce similar short-term effects in the developing and adult brain (e.g., the inhibition of 5-HTT by SSRIs), the enduring effects on the developing system may well be opposite to those seen during treatment in adults. Such enduring 'opposite' behavioral and neural effects are clearly seen following in utero or early life 5-HTT blockade.[3,101,102] Similar consequences may conceivably occur following alteration of maturational processes by adolescent drug exposure. Indeed, unlike the anxiolytic effects of adult antidepressant treatment,[103] SSRI exposure during adolescence appears to have anxiogenic consequences in adulthood.[104] Similarly, as described in the 'Serotonin system' section, the enduring increases in 5-HTT expression following developmental SSRI exposure contrast with the decreases in expression normally observed during treatment in adults.[105,106]

Choice of Antidepressant Drug: Differences Between SSRIs

Despite their similarities, the SSRIs differ markedly from one another in pharmacokinetic and pharmacological profiles, and therefore in their efficacy, safety and suitability for clinical and animal studies. For example, the relative safety of fluoxetine in adolescents has been attributed to its long half-life and active metabolite, while paroxetine's short half-life has been implicated in its association with treatment-emergent suicidality.[107] Half-life is also a consideration in animal studies using once- or twice-daily drug administration, where a short half-life may prevent attainment of the steady-state levels needed for the detection of neural effects.[108] By contrast, many pharmacogenomic studies favor citalopram or escitalopram for its selectivity for the 5-HTT and limited interaction with the liver cytochrome P450 system.[109]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.