Looking Again at SSRIs in Adolescent Depression and Anxiety
“Society gets the type of adolescent that it expects and deserves.”
On October 02, 2025, Martin Plöderl (a PhD clinical psychologist and psychotherapist in Austria) published a blog post, “A call for evidence based, fear-free practice,” in which he takes a close look at two statements by prominent American psychiatrists (Olfson, 2025; Strawn & Walkup, 2025) responding to the White House’s “Make America Healthy Again” (MAHA) report and its criticisms of the use of psychiatric medications in children and adolescents. While acknowledging the scientific as well as political problems with MAHA, Plöderl insists that psychiatrists should still respond to it with scientific rigor and he finds the responses lacking in this regard. Plöderl focuses on SSRI use in children and adolescents for pediatric depression, an area where he sees both overuse and uncertainty about benefits versus harms (he doesn’t really tackle pediatric anxiety, which, as I’ll discuss below, is relevant and weakens Plöderl’s position).
The first target of critique is Marc Olfson, whose commentary in JAMA Pediatrics stressed (among other things) antidepressants’ protective role against suicidality. Plöderl argues that Olfson misrepresents the evidence, especially findings from the Treatment for Adolescents with Depression Study (TADS). Contrary to Olfson’s claim, fluoxetine alone showed little to no benefit over placebo for suicidal ideation, and suicide attempts were actually more frequent among those receiving fluoxetine. Later reanalysis of TADS by the RIAT project in fact revealed additional unreported suicide-related adverse events, further undermining Olfson’s narrative.
Plöderl then challenges Olfson’s handling of the FDA black box warning on antidepressants. While Olfson implied the warning discouraged treatment and caused more suicides, Plöderl emphasizes that much of the supporting ecological research has been challenged and cannot be accepted at face value. He cites methodological flaws and contradictory evidence, noting that longer-term trends do not support the claim that the warning increased suicides. Randomized controlled trials and high-quality observational studies, in fact, point to a small but real increase in suicide attempts and suicides among youth taking SSRIs.
Turning to Jeffrey Strawn and John Walkup’s article “Fact Versus Fear,” Plöderl critiques their dismissal of the FDA black box warning and their selective use of outdated evidence. He argues they misrepresent trial results, exaggerate the benefits of fluoxetine and other SSRIs, downplay or ignore evidence of harm, and present their opinions as evidence. He criticizes Strawn and Walkup for overstating the efficacy of antidepressants, particularly fluoxetine, which they describe as having a “robust” evidence base. Plöderl notes that while early trials suggested benefit, more recent meta-analyses—including a 2021 Cochrane review—show the efficacy of fluoxetine and other antidepressants to be clinically negligible, essentially indistinguishable from placebo. Plöderl points out that reanalysis of clinical trials have shown consistent underreporting and misclassification of harms. He challenges Strawn and Walkup’s claim that benefits outweigh risks, pointing to evidence from Strawn’s own trial of escitalopram in pediatric anxiety, which showed more participants becoming suicidal than achieving remission.
Plöderl is a formidable critic with a deep grasp of evidence and a strong commitment to evidence-based medicine. I don’t always agree with him but I respect him and take him seriously. The issue of SSRI efficacy is a rather tired one at this point, but I think there are useful things to be said in the context of adolescent depression and anxiety. Aside from reviewing relevant evidence, I also want to focus on how the psychiatric community can better acknowledge the current state of affairs and how we can communicate to the clinical and scientific communities as well as the public in a manner that demonstrates integrity, humility, transparency, and, hopefully, repairs some degree of lost trust.
A fundamental point is that we have to acknowledge when evidence from RCTs reveals that treatments have limited efficacy and are accompanied by a considerable risk of harm. This seems like a basic point, but as Plöderl reveals, it is apparently difficult for many in the psychiatric community. The attitude demonstrated mostly appears to be one of forced reassurance, with nods towards “appropriate use” and condemnation of critics as ignorant or ideologically motivated.
It is painful, but it is true that data from RCTs in support of SSRI use in pediatric depression is rather pitiful. If we look at meta-analytic evidence alone, we’d be forced to conclude that SSRIs have trivial effects above placebo in childhood/teenage depression.
Consider the 2021 Cochrane network meta-analysis of antidepressants for depression in children and adolescents:
“Most antidepressants may be associated with a “small and unimportant” reduction in depression symptoms on the CDRS‐R scale (range 17 to 113) compared with placebo…”
Or the 2016 Lancet network meta-analysis:
“The quality of evidence was rated as very low in most comparisons. For efficacy, only fluoxetine was statistically significantly more effective than placebo… When considering the risk–benefit profile of antidepressants in the acute treatment of major depressive disorder, these drugs do not seem to offer a clear advantage for children and adolescents. Fluoxetine is probably the best option to consider when a pharmacological treatment is indicated.”
Or consider the 2020 AHRQ systematic review:
“For adolescents (study participants’ ages range from 12 to 18 years) with MDD, cognitive behavioral therapy (CBT), fluoxetine, escitalopram, and combined fluoxetine and CBT may improve depressive symptoms (1 randomized controlled trial [RCT] each, n ranges from 212 to 311); whether the magnitude of improvement is clinically significant is unclear.”
Now I don’t personally think that RCTs are the end-all be-all of clinical practice, and I am in favor of triangulating results of controlled trials with observational data, clinician observations (and accumulated impressions of the clinical communities), patient experiences, and animal studies. Although, of course, these sources of information do not possess equal weight, and well-designed, high-quality RCTs with representative samples and clinically meaningful measures are usually definitive (sadly, we don’t have many of them). But if we are going to bang the drum of “evidence-based medicine” and if we are going to invoke the authority of empirical evidence, then we must admit when the evidence is unfavorable and resist the urge to rationalize it away through weak arguments.
Secondly, the risk of suicidality with SSRIs (suicidal thoughts, behaviors, and possibly deaths) is real and genuine, and no amount of squinting one’s eyes to look at data this or that way will make it go away. Yes, the absolute risk is generally small and, in competent clinical hands, quite manageable, but acknowledging it is the right and prudent thing to do. I have seen many teenagers and young adults (and at times other adults) who experience new suicidal thoughts or worsening of existing suicidal thoughts on starting a new antidepressant that rapidly goes away once the particular antidepressant is discontinued (and not infrequently, for mysterious reasons, it happens with some antidepressants but not with others). I have seen it happen so many times that the reality of this phenomenon is as apparent to me as the reality of clinical benefit from medications, and I am always puzzled when respected clinicians make fools of themselves by trying to deny that this happens. The idea that reduced antidepressant prescriptions following the FDA black box warning led to increased suicides in the US is an appealing one, but it is not well-supported, is likely false, and at any rate is believed too uncritically by psychiatrists because it fits in with their biases and with the story they’d like to tell the public.
Thirdly, while psychiatrists often appeal to the risk of suicide in depression to advocate for pharmacological treatment, saying, “But think of all the suicidal children,” the reality is that antidepressants, whatever their efficacy in depression and anxiety, show no clear improvements in suicidal behaviors compared to placebo. I do not believe that SSRIs, on average, improve suicidality beyond what we see with good clinical contact and a therapeutic relationship. This puts medication advocates in an awkward position: while teen suicide is the most drummed-up talking point, it is also a situation where we have no strong evidence of any specific benefit from antidepressants. I therefore advise against appeals to preventing suicide as an argument in favor of antidepressants; it’s always going to backfire.
Now I have to clarify here that efficacy evidence for SSRIs in adolescents is not uniformly bleak. SSRIs do work decently well for generalized anxiety and obsessive-compulsive disorders.
2015 systematic review and meta-analysis by Strawn et al.:
“Nine trials involving 1,673 patients and six medications were included. All SSRI/SSNRIs evaluated demonstrated efficacy, and the meta-analytic estimate of effect was of moderate magnitude (Cohen’s d = 0.62, confidence interval [CI]: 0.34-0.89, P = .009) and there was evidence of modest heterogeneity (I(2) = 0.29, P = .103).”
2017 systematic review and meta-analysis by Wang et al.:
“A total of 7719 patients were included from 115 studies. Of these, 4290 (55.6%) were female, and the mean (range) age was 9.2 (5.4-16.1) years. Compared with pill placebo, selective serotonin reuptake inhibitors (SSRIs) significantly reduced primary anxiety symptoms and increased remission (relative risk, 2.04; 95% CI, 1.37-3.04) and response (relative risk, 1.96; 95% CI, 1.60-2.40). Serotonin-norepinephrine reuptake inhibitors (SNRIs) significantly reduced clinician-reported primary anxiety symptoms. Benzodiazepines and tricyclics were not found to significantly reduce anxiety symptoms. When CBT was compared with wait-listing/no treatment, CBT significantly improved primary anxiety symptoms, remission, and response. Cognitive behavioral therapy reduced primary anxiety symptoms more than fluoxetine. The combination of sertraline and CBT significantly reduced clinician-reported primary anxiety symptoms and response more than either treatment alone. Head-to-head comparisons were sparse, and network meta-analysis estimates were imprecise. Adverse events were common with medications but not with CBT and were not severe. Studies were too small or too short to assess suicidality with SSRIs or SNRIs.”
2024 AHRQ systematic review of OCD in children:
“SSRIs are more effective than placebo for OCD symptoms and global severity (high SoE). Clomipramine is probably more effective than placebo (moderate SoE). When used together, ERP and an SSRI are probably more effective than treatment with an SSRI alone for OCD symptoms (moderate SoE). ERP combined with an SSRI are as effective as ERP alone for OCD symptoms (high SoE). The side effects of SSRIs and clomipramine were inconsistently reported…”
Jennifer Dwyer and Michael Bloch (2019) have argued that poor meta-analytic efficacy of SSRIs in pediatric depression should take into account the difference seen in NIMH-sponsored TADS trial vs industry-sponsored trials:
“Selective serotonin reuptake inhibitors are also effective for treating pediatric MDD; however, the literature is more complex for this disorder compared to GAD and OCD as there are considerable differences in effect sizes between National Institute of Mental Health (NIMH)–funded studies and industry-sponsored trials. The major NIMH-sponsored adolescent depression trial, TADS (Treatment for Adolescents and Depression Study), showed that SSRIs (fluoxetine in this case) were quite effective, with an NNT of 4 over the acute phase (12 weeks). Ultimately, approximately 80% of adolescents improved over 9 months. Many industry-sponsored trials for MDD in pediatric patients had large placebo response rates (approximately 60%), which resulted in smaller between-group differences, and estimates of an NNT closer to 12 which has muddied the waters in meta-analyses that include all trials. Improvement in depressive symptoms also appears to be bolstered by concomitant CBT in MDD, but not as robustly as in GAD and OCD.”
A recent reanalysis of TADS using more rigorous statistical methods casts some doubts on the rosy results presented earlier. In the primary ITT analysis based on CDRS-R scores, there was no statistically significant difference between fluoxetine and placebo (p = 0.480) as well as CBT and placebo (p = 1.000). Combination treatment was statistically significant versus placebo (p < 0.001), and, interestingly, combination treatment separated statistically from CBT alone but not fluoxetine alone. However, even in the TADS re-analysis, fluoxetine was statistically superior to placebo in terms of response rate (p = 0.004). Combination treatment showed the highest response rate (73.7%), followed by fluoxetine (63.3%), CBT (48.9%) and placebo (39.6%).
So, the picture is rather mixed and doesn’t clearly show fluoxetine to be ineffective. On their own, response rates of 74% and 63% are not trivial, whatever the mechanisms by which such improvement is obtained.
Plöderl is silent on these issues of efficacy in anxiety and OCD and efficacy signals in TADS, but I believe they are important. They are enough to convince me of the inadequacy of any straightforward position that SSRIs are clinically useless in children and adolescents.
The other thing is that evidence-based medicine critics such as Plöderl who take a position of skepticism with regard to the efficacy of medications and psychotherapy have very little to offer as an alternative.
Plöderl:
“Nobody would argue that depressed people should be left without support or “untreated”. However, in my experience, all too often “treated” is equaled with “prescribing antidepressants”, and there seems to be a widespread automatism in prescribing antidepressants, driven by the assumption that antidepressants are safe and effective and that “something must be done”, and because of a seeming lack of alternatives. With the known poor efficacy of antidepressants and their harms, alternatives to antidepressants should be offered more often, such as watchful waiting, or basic interventions such as sleep management, light exposure etc (Selalmazidou & Bschor, 2023).”
I agree with Plöderl about “widespread automatism in prescribing antidepressants,” and I’d agree that we shouldn’t use ineffective treatments just because something must be done, but as I’ve discussed above, I am also not convinced that SSRIs and CBT are ineffective. Watchful waiting and reassurance… I think that’s fine for mild depression and anxiety or where there is a strong situational component and distress is expected to be short-lived, or where patients and families prefer it. But frequently the presentations we see in the clinic demand utilizing whatever treatments, including imperfect ones, are available and have a reasonable chance of helping the person.
Plöderl:
“With the repeated claim that untreated depression is most dangerous, critics of MAHA are doing the same fear-mongering as proponents of MAHA do for the harms of the drugs. Usually, authors who claim that untreated depression is super risky either did not reference related studies or they misrepresent the evidence… Taking the evidence seriously would mean that we don’t know very much about untreated depression. However, the few available studies show that people who do not seek treatment fare well in the longer term (Whiteford et al., 2013). And as repeatedly pointed out above, treatment with antidepressants hardly outperforms treatment with placebo, at least for the majority of patients. Thus, instead of spreading fear we should provide hope. And this is evidence based hope.”
It takes a serious and restrictive commitment to evidence to maintain that we cannot assert that untreated depression and anxiety are dangerous or threats! I respect that in a way, although this sort of commitment is too confining for my taste. Perhaps we are not in a position to say much with confidence about the dangers of treated vs untreated depression, but adolescent depression itself is associated with a variety of poor health and psychosocial outcomes (Naicker et al, 2013; Clayborne et al, 2019; Jonsson et al, 2011) and is considered by WHO to be a leading cause of illness and disability in adolescents. To what degree are bad outcomes prevented by treatments, I don’t know, but that by itself does little to reassure me, or parents, or the public, about the benign nature of adolescent depression and anxiety.
On his substack,
discussed the results of a recent paper in Lancet Psychiatry that analyzed outcomes for 1.5 million people treated for anxiety or depression through the UK’s Talking Therapies program. “Young adults experienced a smaller change in symptoms pre–post treatment and had between 17% and 26% lower odds of reliable recovery and reliable improvement.”Insel wonders:
“Maybe the answer isn’t more therapy, but a focus on more normative development. Erik Erikson taught us that every stage of life poses a developmental challenge: industry versus inferiority (tweens), identity versus role confusion (adolescents), intimacy versus isolation (young adults). Now imagine growing up with seven hours of daily screen time (inferiority), psychiatric labels where once there were growing pains (role confusion), and friendship filtered through devices (isolation). We talk endlessly about treatments, but barely mention the basics needed for healthy development: sleep, community, and purpose are three that need our attention.”
“Decades ago in his book on Adolescence, the British psychoanalyst James Anthony wrote, “Society gets the type of adolescent that it expects and deserves.” Perhaps we expect too little. Maybe our rush to diagnose, treat, and medicalize ordinary distress has crowded out the harder but healthier path: expecting responsibility, engagement, and contribution. If the new data tell us anything, it’s that the solutions for youth mental health won’t be found only in clinics. They may be found in bedrooms, classrooms, and communities—where sleep, friendship, and purpose quietly do their developmental work.” (emphasis in original)
I am inclined to agree with Insel, while acknowledging the complexity of adolescent mental health problems and the existence of serious psychopathology. The adolescent patients that I see—admittedly, a rather biased sample of kids that require a higher level of care than their pediatrician or regular-frequency outpatient psychiatric visits—often show maladaptive traits (e.g., high neuroticism or disinhibition) or characterological difficulties in early stages (e.g., an evolving borderline personality organization) or significant developmental trauma with its psychological sequelae.
There are powerful structural forces at play, and adolescent patients experience unique developmental challenges. In my interactions with teenagers, beyond dysfunctional home environments and characterological issues, I am often struck by how many struggle with an absence of meaning in life. They struggle to envision a future that resonates with them. Often, there isn’t even a person they look up to that serves as a model of healthy ambition (on more than one occasion, kids have offered the name of some Twitch streamer as someone they admire, something that leaves me feeling rather dispirited, but maybe I’m just old).
How do you give someone ambition, a sense of purpose, a sense that life can be beautiful, that there are people and tasks worth devoting your life to? I sit in my office with struggling, lost, untethered kids, and I find myself thinking, “What you need most of all is connection and purpose.” And I am fortunate that I work in a clinical setting (intensive outpatient) that does seek to facilitate that. More fundamentally, I encourage them to look at their emotional and behavioral challenges in the context of their lives, with medications playing a supportive role at best.
I don’t believe that we should judge how many are “untreated” by looking at what percentage of kids are on SSRIs, as Strawn & Walkup attempt to do. But I do believe—if I’m allowed to believe things without evidence from controlled trials—that more kids do need to receive good clinical care with an emphasis on relational aspects and that they should be provided non-reductive formulations of their problems that set them on a path of psychological growth. And right now, I suspect, we are utterly failing at that.
See also:
Thank you for this, as a practicing Child and Adolescent psychiatrist it is a hugely underdiscussed and misunderstood territory and your article is informative and balanced. Three things come to mind, one is the paucity of data on SSRI related sexual side effect in adolescent populations (if we look at the range of sexual side effects in adults including loss of libido, anorgasmia etc it ranges from 58-73%), there is little to no meaningful data on the incidence of these sorts of side effects (which would have enormous implications for normal adolescent psychosexual development) in younger populations. Is this because we are afraid or uncomfortable to ask? I often wonder about this with my patients. Secondly, the discussion of SSRIs and suicidality- it is my understanding that only two drugs have real data to support any influence on incidence of suicide, lithium and clozapine. SSRIs do not and have not shown efficacy around reducing suicidality for either adults or adolescents, though by reducing anxiety load they may downstream reduce discomfort and therefore secondarily suicidality. Thirdly, with all medications there is a tension between the individual and the data. Meaning that while data may indicate a medication has limited efficacy on a study level, that same intervention may be profoundly influential on an individual level. This is what makes evidence based medicine ideology so complex in psychiatry. Personally, I tend to think of SSRIs in adolescents as possibly being able to soften symptom burden, which would allow the teen to engage in more meaningful lifestyle changes, tolerate therapeutic interventions more deeply, and overall be able to engage in non-medication interventions that would actually be making the bulk of the difference.
I’m grateful for your persistence with this issue, because it makes me very tired. These debates keep going around of circles because of the underlying assumption that there is one universal best treatment for depression (SSRIs, therapy, social reform, exercise, resilience) despite the fact that depression is an incredible heterogeneous condition and frequently co-occurs with a highly diverse range a other condition (treating some over with concurrent MDD and GAD is very different to treating MMD+PTSD, and very different again if there is substance use involved, or they are neurodiverse) Even though there is more acceptance and research into personalized medicine, we still get caught up in wanting generalized, universal solutions. Further complicating this (and this is the issue I don’t see discussed as much) is too often treatment decisions aren’t based on ‘best evidence based practice’, but on ‘best available treatment’. These debates keep wanting to frame that question as ‘what is the best treatment for depression’, when the reality for clinicians is ‘what is the best treatment for this person with these difficulties in this context, that is also acceptable AND available to the patient.’ Trying to find treatments that tick all of those boxes feels like an impossible task some days (especially if you want to write a snappy headline’, so I get the appeal of wanting blaming everything on something like screentime!