Jacob Z. Hess, Ph.D., with Robert Whitaker
The list of 20 studies below (now 22 studies – with a recent update) comprises all the long-term studies of antidepressant outcomes of which we’re aware (with the large majority of existing studies focused short-term, 3-6 weeks, etc.). I give credit and thanks to award-winning journalist, Robert Whitaker – for aggregating this data so families and professionals are able to make more sense of it. I have interviewed Bob twice and find him a man of high integrity and great modesty – which is why I believe so many scholars and medical doctors, including at institutions like Harvard and Yale, have welcomed him for talks and grand rounds.
One doctor famously threw Whitaker’s book across the room, but then admitted publicly, “but I couldn’t dispute the data he was presenting.” He subsequently made some meaningful adjustments to his practice that made his patients facing depression safer. That’s what Whitaker and others like him (myself included) are after: a more nuanced conversation that would allow meaningful adjustments to treatment protocols that reduce risk and increase benefit.
In what follows, many of Whitaker’s own analyses are provided verbatim, with adjustments for brevity – along with summaries of his graphics, statistics and explanations for simplicity sake. Each study is provided chronologically, with a hyperlink to the original study for additional details. Whitaker’s own written review of this same material can be accessed on page 164-169 in this chapter of his book as well – along with other helpful presentations available online (see, for instance, this excerpt). Separately, I’ve analyzed how these patterns are showing up in the Latter-day Saint faith community that I call home (this piece also includes a number of other excerpts from my own interview with Bob).
Although the long-term patterns across these 20 studies (and others like it for antianxiety and antipsychotic medications) are strikingly similar, these findings have unfortunately been largely overlooked or explained away in both public and professional conversations about antidepressants today. This must changed, many of us are now convinced, if we are to see any substantial reduction in our suicide and depression numbers.
(1) 1973: A long-term review of 94 depressed patients. [Van Scheyen, J. (1973), Recurrent vital depressions. Psychiatry, Neurologia, Neurochirugia, 76, 93-112].
J.D. Van Scheyen, a Dutch psychiatrist, concluded from his own original research and a review of the existing literature that “it was evident, particularly in the female patients, that more systematic long-term antidepressant medication, with or without ECT [electronconvulsive therapy], exerts a paradoxical effect on the recurrent nature of the vital depression. In other words, this therapeutic approach was associated with an increase in recurrent rate and a decrease in cycle duration.”
As he notes in the paper, other psychiatrists had observed that antidepressants were causing a “chronification” of the disease.
Dr. Scheyen went on to ask, “Should [this increase] be regarded as an untoward long-term side effect of treatment with tricyclic antidepressants?”
(2) 1973: Six-month follow-up of 92 depressed patients. [Mindham, R. (1973). An evaluation of continuation therapy with tricyclic antidepressants in depressive illness. Psychological Medicine, 3, 5-17].
Between the 1970’s and early 1980’s, the NIMH and other groups reported on four different occasions that patients withdrawn from antidepressants were showing uniquely high relapse rates (comprising the next four studies – #2, 3, 4 & 5). In this 1973 paper, British researchers found that 50% of drug-withdrawn patients relapsed within six months.
(3) 1980: Six-month follow-up of 55 depressed patients. [Stein, M. (1980). Maintenance therapy with amitriptyline. American Journal of Psychiatry, 137, 370-1].
Investigators at the University of Pennsylvania reported that 69% of patients withdrawn from an antidepressant relapsed within six months. There was “rapid clinical deterioration in most of the patients.”
(4) 1984: Two-year follow-up of 267 depressed patients [Prien, R. (1984) Drug therapy in the prevention of recurrences in unipolar and bipolar affective disorders. Archives of General Psychiatry, 41, 1096-1104].
Robert Prien at the NIMH reported that 71% of depressed patients relapsed within 18 months of drug withdrawal.
(5) 1986: Two-year follow-up of depressed patients [Blackburn, I . M., Eunson, K., & St Bishop, S. (1986). A two-year naturalistic follow-up of depressed patients treated with cognitive therapy, pharmacotherapy and a combination of both. Journal of Affective Disorders, 10, 67-75].
This research team likewise found that significantly more patients in the group given anti-depressants alone relapsed at both 6 month and 2 year time-points, when compared to psychotherapy in various combinations.
(6) 1992: 16-week results compared to 18-month results [Shea, M. et al. (1992) Course of depressive symptoms over followup. Findings from the National Institute of Mental Health Treatment of Depression Research Program. Archives of General Psychiatry, 49, 782-787].
In an 18-month NIMH study that compared four types of treatment (two forms of psychotherapy, an antidepressant, and placebo), the group that was initially treated with the antidepressant had the lowest stay-well rate by the end of the study (19%) – and the highest relapse rate (50%). If study dropouts were included in the analysis, then the results for the imipramine patients “look even worse . . . patients receiving the antidepressant were most likely to seek treatment following termination, produced the highest probability of relapse, and exhibited the fewest weeks of reduced or minimal symptoms during the follow-up period.”
(7) 1995: Six-year NIMH follow-up on 547 people diagnosed with depression [Coryell, W. (1995). Characteristics and significance of untreated major depressive disorder. American Journal of Psychiatry, 152, 1124-9].
NIMH-funded investigators led by University of Iowa psychiatrist William Coryell tracked the outcomes of medicated and unmedicated depressed people over a period of six years; those who were “treated” for the illness were three times more likely than the untreated group to suffer a “cessation” of their “principal social role” and nearly seven times more likely to become “incapacitated.” While many of the treated patients saw their economic status markedly decline during the six years, only 17 percent of the unmedicated group saw their incomes drop, and 59 percent saw their incomes rise. The NIMH researchers noted that overall: “The untreated individuals described here had milder and shorter-lived illness [than those who were treated], and, despite the absence of treatment, did not show significant changes in socioeconomic status in the long term.”
(8) 1997: Six-month follow-up up of 148 depressed patients. [Ronalds, C. Outcome of anxiety and depressive disorders in primary care.” British Journal of Psychiary, 171 (1997): 427-33].
In a 1997 study of the outcomes of 148 depressed patients at a large inner-city facility, British scientists reported that ninety-five never-medicated patients saw their symptoms decrease by 62 percent in six months, whereas the fifty-three drug-treated patients experienced only a 33 percent reduction in symptoms. The medicated patients, they concluded, “continued to have depressive symptoms throughout the six months.”
(9) 1998: Harvard meta-analysis of 27 studies across 3037 depressive patients treated for up to 48 months (6 month average), then followed for up to 66 months (17 month average) [Viguera, A. (1998). “Discontinuing antidepressant treatment in major depression.” Harvard Review of Psychiatry, 5, 293-305].
In a meta-analysis of the relapse literature, Havard researchers concluded that at least fifty percent of all drug-withdrawn patients relapsed within 14 months. They also noted that the longer the patient had been on an antidepressant prior to drug withdrawal, the higher the relapse rate. (In the pre-antidepressant era, this was the relapse rate seen in studies that lasted 15 years or more.)
(10) 1998: Twelve-month World Health Organization follow-up on 740 people screened for depression in 15 cities around the world [Goldberg, D., et al. (1998). The effects of detection and treatment of major depression in primary care: A naturalistic study in 15 cities. British Journal of General Practice, 48, 1840-44].
A World Health Organization study of depressed patients in 15 cities around the world was designed to assess the merits of screening for the disorder. The hypothesis behind the study was that those treated with antidepressants would have the best long-term outcomes, and those whose depression was not detected (and thus didn’t receive treatment) would fare the worst. The results were the opposite of expected. At the end of one year, the subset of 484 people who didn’t receive psychotropic medications enjoyed much better “general health” and had depressive symptoms that were “much milder.” The non-medicated group was also less likely to still be “mentally ill.” By comparison, the group that suffered most from “continued depression” were the patients receiving ongoing antidepressant treatment.
As the lead researcher summarized: “Patients not given drugs had milder illnesses but did significantly better than those receiving drugs, both in terms of symptoms lost and their diagnostic status.” This was so “even after adjustment for initial scores on each instrument.” The researchers concluded, the “study does not support the view that failure to recognize depression has serious adverse consequences.”
(11) 1998: One-year follow-up of 100 people [Rost, K., Zhang, M., Fortney, J., et al. (1998). Persistent poor outcomes of undetected major depression in primary care: implications for intervention. General Hospital Psychiatry, 20, 12-20].
Patients in general practice receiving standard antidepressant treatment (in accordance with practice guidelines for maintenance over time) had higher rates of relapse than those receiving no therapy.
(12) 2000: Ten-year follow-up of 222 people who had suffered a first episode of depression [Weel-Baumgarten, E. (2000). “Treatment of depression related to recurrence,” Journal of Clinical Psychiatry & Therapeutics, 25, 61-66].
In a retrospective study tracking 10-year case history outcomes, Dutch investigators found that 76% of those not treated with an antidepressant recovered and never relapsed, compared to 50% of those prescribed an antidepressant. The lead researcher wrote, “Even when a diagnosis of Major Depressive Disorder has been made, spontaneous recovery should be considered for a number of cases in general practice and watchful waiting could prove worthwhile.”
(13) 2003: 1281 people assessed for anti-depressants and rates of disability [Dewa, C. (2003). “Pattern of antidepressant use and duration of depression-related absence from work.” (here’s a hard-copy version from the journal) British Journal of Psychiatry, 183, 507-13].
In Canada, Carolyn Dewa and her colleagues at the Center for Addiction and Mental Health in Ontario set out to explore the relationship between antidepressants and disability rates. They first identified 1,281 people who went on short-term disability between 1996 and 1998 because they missed ten consecutive work days due to depression. The 564 people who opted to not go on antidepressants returned to work, on average, in 77 days – while the medicated group took 105 days to get back on the job, on average. More important, those who took an antidepressant were more than twice as likely to go on to long-term disability, with only 9 percent of the unmedicated group submitting a request for long-term disability, compared to 19 percent of those who took an antidepressant. “Does the lack of antidepressant use reflect a resistance to adopting a sick role and consequently a more rapid return to work?” the researcher wondered.
Of course, alternative hypotheses could be made about those choosing to go on medication being more severe cases, as an explanation for the numbers. Neither hypothesis can be dismissed from the current evidence – with both deserving consideration.
(14) 2004: Five-year follow-up of 9,508 depressed patients [Patten, S. (2004). “The Impact of antidepressant treatment on population health.” Population Health Metrics, 2, 9].
In a five-year University of Calgary study of 9,508 depressed patients in Canada, medicated patients were depressed on average 19 weeks a year, versus 11 weeks for those not taking the drugs. Episodes occurring in antidepressant users lasted longer than those in non-users. The apparent incidence of major depressive episodes among those taking antidepressants was higher than that among respondents not taking antidepressants. The lead author, Scott Patten, wrote that “these findings are consistent with Giovanni Fava’s hypothesis that ‘antidepressant treatment may lead to a deterioration in the long-term course of mood disorders.’”
(15) 2004: One-year NIMH follow-up of 118 people [Rush, J., (2004). “One-year clinical outcomes of depressed public sector outpatients,” Biological Psychiatry 56: 46-53].
126 patients originally treated with antidepressants were given emotional and clinical support “specifically designed to maximize clinical outcomes.” 26% responded to antidepressants (defined as a 50% reduction in symptoms), with half of those who responded (13%) staying better for a significant period of time. At the one year mark, only 5-6% of patients remained in remission and doing well, which is a much lower remission rate than is typically found in studies of unmedicated depressed patients. Lead author John Rush acknowledged, “These findings reveal remarkably low response and remission rates.”
Similarly, in further analysis of the large STAR*D trial, only 108 of the 4041 (3%) patients who entered the study remitted and then stayed well during the one-year followup – with all of the others either never remitting, relapsing, or dropping out of the study.
(16) 2006: Up to 15-year NIMH follow-up of 84 unmedicated people facing depression [Posternak, M. et al., (2006). “The naturalistic course of major depression in the absence of somatic therapy Journal of Nervous and Mental Disease, 194, 324-349].
Noting that much of the existing knowledge for poor long-term outcomes in textbooks told the story of medicated depression, Dr. Michael Posternak, a psychiatrist at Brown University wrote that “unfortunately, we have little direct knowledge regarding the untreated course of major depression.” In an NIMH study of “untreated depression,” Posternak and his team subsequently identified a subset of 84 people in a larger NIMH study who had relapsed but chosen not to go back on anti-depressants.
How did these 84 people fare over the years they were tracked? Twenty-three percent of the non-medicated patients recovered in one month; 67% in six months; and 85% within a year. Given these numbers, Posternak and colleagues pointed out that German psychiatrist Emil Kraepelin had observed untreated depressive episodes usually cleared up within six to eight months – adding that their own results provided “perhaps the most methodologically rigorous confirmation of this estimate.” Overall, most people in this study struck by a bout of major depression naturally recovered: “These results suggest that there is a high rate of recovery in individuals not receiving somatic treatment of their depressive illness, particularly in the first 3 months of an episode.” In particular, for all 84 subjects who went unmedicated, the “median time to recovery was 13 weeks.”
They went on to note, “treatment-seeking behavior is known to be associated with a worse prognosis.” The investigators concluded, “If as many as 85% of depressed individuals who go without somatic [medical] treatments spontaneously recover within one year, it would be extremely difficult for any intervention to demonstrate a superior result to this.”
(17) 2008: Two-year follow-up of 172 patients facing depression [Bockting, C. (2008).“Continuation and maintenance use of antidepressants in recurrent depression.” Psychotherapy and Psychosomatics, 77, 17-26].
Given the positive short-term effect that many have reported on starting an antidepressant, lead researcher Claudi Bockting, wrote, “Continued antidepressant treatment may oppose the initial acute effects of [the] antidepressant . . . neurobiological mechanism(s) may be involved in increasing vulnerability.”
(18) 2010: Patterns across over 260,000 patients facing depression [MN Community Measures, 2010 Health Care Quality Report]
This is a study looking one year out at all patients in Minnesota treated for depression – that documented over the course of several years well over 80% of patients still depressed while being administered long-term medical treatment. Bob points out in his talk that this is the opposite of outcome numbers in previous eras, where large majorities were finding longer-term healing.
(19) 2011: Meta-analysis of 46 studies exploring antidepressant discontinuation. [Andrews, P. (2011). Blue Again: Perturbational Effects of Antidepressants Suggest Monoaminergic Homeostasis in Major Depression. Frontiers in Psychiatry, 2, 159].
In a meta-analysis of 46 studies looking at discontinuation of anti-depressant medications, researchers found that the relapse rate for placebo responders during a followup period was 24.7%, compared to 44.6% of drug responders who were then withdrawn from the drug.
(20) 2012: Comprehensive review of the long-term research on anti-depressants. [Andrews, P. (2012). Primum non nocere: An evolutionary analysis of whether antidepressants do more harm than good. Frontiers in Psychology, 3, 1-18].
This comprehensive review of the literature included an original meta-analysis of studies in which anti-depressants had been discontinued. The authors found more than twice as much depression relapse on those treated by SSRI anti-depressants long-term, when compared to those receiving placebo (see graph below). Looking across all the long-term evidence, the authors conclude, “it is widely believed that antidepressant medications are both safe and effective; however, this belief was formed in the absence of adequate scientific verification. The weight of current evidence suggests that, in general, antidepressants are neither safe nor effective; they appear to do more harm than good.” The lead researcher, Paul Andrews, wrote, “The more antidepressants perturb monamine levels in the brain, the more the brain appears to push back, which increases the risk of relapse when the drug is discontinued . . . antidepressant use appears to increase [biological] susceptibility to depression.”
(21) 2017: Comparative analysis of outcomes 9 years out. [Vittengl, J. R. (2017). Poorer long-term outcomes among persons with major depressive disorder treated with medication. Psychotherapy and Psychosomatics, 86, 302-304].
Jeffrey Vittengl at Truman University conducted a study finding that taking antidepressant medications resulted in more severe depression symptoms after nine years. As Peter Simons summarizes, “Even after controlling for depression severity, participants who took medication had significantly more severe symptoms at the nine-year follow-up than participants who did not. In fact, even people who received no treatment at all did better than those who received medication.”
In addition to collecting information on depression, generalized anxiety disorder, panic disorder, the survey used documented data on other medical conditions, family history of mental health conditions, childhood trauma, personality factors, social support, daily functioning, and alcohol use. While these factors did impact depressive symptoms, Vittengl found they did so equally between the groups. As Simons notes, “That is, initial depression severity does predict lack of improvement—but it does so whether the person is taking medication or not. Therefore, it does not explain how outcomes could be worse with medication.”
(22) 2018: 30-year follow-up of 591 adults in a community sample. [Hengartner, M. P., Angst, J., & Rössler, W. (2018). Antidepressant use prospectively relates to a poorer long-term outcome of depression: Results from a prospective community cohort study over 30 years. Psychotherapy and Psychosomatics, 87, 181–183].
Researchers from Zurich University of Applied Sciences and the University of Zurich followed 591 Swiss adults from the age of 20/21 until they were 49/50 years old, finding that those who took antidepressants at some point in the study were more likely to have worse depression symptoms after 30 years—even when controlling for initial symptoms and other factors. This finding was independent of illness severity as well as a large number of other potential confounding factors.
As Peter Simons summarizes, “Assessments began in 1979 (baseline assessment) when participants were all 20-21 years old, and assessments were conducted again in 1981, 1986, 1988, 1993, 1999, and finally in 2008 (when they were 49-50 years old). At each assessment, the primary outcome was the severity of depressive symptoms within the previous year. Also at each assessment, participants reported whether they had been prescribed antidepressants within the previous year. In order to create their predictive model, the authors tested whether being prescribed antidepressants at one assessment (e.g. 1988) increased the likelihood of more severe depressive symptoms at the next time point (e.g. 1993).”
The authors controlled for numerous factors—including gender, education level, marital status, any affective disorder at baseline, suicidality at baseline, family history of depression, subjective distress, childhood adversity, and low parental income. Although each could be a third variable explaining the results, the researchers found that even taking them into account, antidepressant use was still associated with an 81% increased likelihood of depression severity increase.
The authors (which include Jules Angst, a world leader in mood disorders), write, “These findings are in line with a growing body of evidence from several naturalistic observational studies suggesting that (long-term) antidepressant use may produce a poor long-term outcome in people with depression.”
As reflected above, this pattern showed up again and again, across multiple converging evidence bases. Generally speaking, people taking antidepressants long-term don’t fare so well. That’s the simple truth. Far back in a 1994 article, the psychiatrist Giovanni Fava summarized that “Psychotropic drugs actually worsen, at least in some cases, the progression of the illness which they are supposed to treat.” More recently in a 2003 article, he noted that “A statistical trend suggested that the longer the drug treatment, the higher the likelihood of relapse.”
[For those interested in more details, I highly encourage you to access Bob’s books and many presentations. And once again, to watch Bob’s own presentation of some of these findings, check out the video below (Long-term Anti-depressant Outcome Research)]:
Same patterns with anti-psychotics and anti-anxiety drugs. To those skeptical of these conclusions, be aware that this pattern of increasing chronicity associated with long-term treatment is not exclusive to anti-depressants. Other classes of psychiatric medications show the same patterns. For instance, there is extensive evidence that anti-psychotics lead to the same patterns long-term, as evident in this 2016 talk: Long-term outcomes from anti-psychotics (watch 20:55-29:49). For a more detailed summary of the same literature, watch his 2015 talk given to the Council for Evidence-based Psychiatry, starting with 20:56 and ending with 49:25. Or check out this clip from our own interview with Bob (Takeaways from Long-term Anti-Psychotic Outcome Research):
One representative 2007 study is cited for illustration: [Harrow M. & Jobe, T H. “Factors involved in outcome and recovery in schizophrenia patients not on antipsychotic medications.” Journal of Nervous and Mental Disease 195 (2007):406-14].
Harrow shared other data regarding work history:
Harrow later said about his data, “How unique among medical treatments is it that the apparent efficacy of antipsychotics could diminish over time or become ineffective or harmful? There are many examples for other medications of similar long-term effects, with this often occurring as the body readjusts, biologically, to the medications.”
So why does this appear to be happening? What’s the best explanation for it?
Biological Explanations for the Long-Term Outcomes of Psychiatric Meds
Stephen Hyman, former director of the NIMH, first raised an explanation in 1996, noting that psychiatric medications “create perturbations in neurotransmitter functions.” In response, the brain goes through a series of compensatory adaptations in order “to maintain their equilibrium in the face of alterations in the environment or changes in the internal milieu.” The “chronic administration” of the drugs then cause “substantial and longlasting alterations in neural function.” After a few weeks, the person’s brain is now functioning in a manner that is “qualitatively as well as quantitatively different from the normal state.”
In 2003, Fava added the following: “In order to cope with the antidepressant’ perturbation of neurotransmitter activity, the brain undergoes compensatory adaptations, and “when drug treatment ends, these (compensatory) process may operate unopposed, resulting in appearance of withdrawal symptoms and increased vulnerability to relapse” [Can long-term treatment with antidepressant drugs worsen the course of depression? Fava, G. Journal of Clinical Psychiatry 64 (2003):123-33].
There are many studies documenting various brain changes elicited by antidepressants and other psychiatric medication. For instance, in 2004, chronic antidepressant treatment was documented as resulting in a down-regulation of the synaptic function of forebrain 5-HT2 receptors. [Chronic Antidepressant Treatment Alters Serotonergic Regulation of Gaba Transmission In Prefrontal Cortical Pyramidal Neurons, by P. Zhong And Z. Yan, Neuroscience 129 (2004) 65–73].
In studies with rats, long-term treatment with an SSRI led to markedly reduced serotonin in “nine areas of the brain.” As a team led by Dr. El-Mallakh at the University of Louisville found in 2011, treatment with an SSRI leads to a reduced density of receptors for serotonin in the brain. In experiments with animals, such impairments in serotonergic functions were “associated with increased depressive and anxious behaviors.”
Dr. El-Mallakh concluded that over time, antidepressants “may induce processes that are the opposite of what the medication originally produced.” “Rather than raise serotonin levels, the drugs over the long-term impaired serotonergic pathways in the brain. As he put it, “Continued drug treatment may induce processes that are the opposite of what the medication originally produced.” This may “cause a worsening of the illness [and may] continue for a period of time after discontinuation of the medication.” The researcher writes: “A chronic and treatment-resistant depressive state is proposed to occur in individuals who are exposed to potent antagonists of serotonin reuptake pumps for prolonged time periods. Due to the delay in the onset of this chronic depressive state, it is labeled tardive dysphoria.” [Tardive Dysphoria: The Role of Long-term Antidepressant Use in Inducing Chronic Depression. El-Mallakh, R. Medical Hypotheses 76 (2011):769-773].
As Giovanni Fava wrote, “A chronic and treatment-resistant depressive state is proposed to occur in individuals who are exposed to potent antagonists of serotonin reuptake pumps (i.e. SSRIs) for prolonged time periods. Due to the delay in the onset of this chronic depressive state, it is labeled tardive dysphoria. Tardive dysphoria manifests as a chronic dysphoric state that is initially transiently relieved by – but ultimately becomes unresponsive to – antidepressant medication. Serotonergic antidepressants may be of particular importance in the development of tardive dysphoria.”
Giovanni Fava elaborated further in 2011, “When we prolong treatment over 6-9 months, we may recruit processes that oppose the initial acute effects of antidepressant drugs (loss of clinical effects)…We may also propel the illness to a malignant and treatment-unresponsive course that may take the form of resistance or episode acceleration. When drug treatment ends, these processes may be unopposed and yield withdrawal symptoms and increased vulnerability to relapse.”
The same year, Dr. Andrews and colleagues published their examination of “oppositional tolerance” – referring to “the forces that develop when a homeostatic mechanism has been subject to prolonged pharmacological perturbation that attempt to bring the system back to equilibrium. When pharmacological intervention is discontinued, the oppositional forces cause monoamine levels to overshoot their equilibrium levels. Since depressive symptoms are under monoaminergic control, this overshoot should cause a resurgence of depressive symptoms that is proportional to the perturbational effect of the ADM” [Blue again: perturbational effects of antidepressants suggest monoaminergic homeostasis in major depression, by Paul W. Andrews1, Susan G. Kornstein, Lisa J. Halberstadt, Charles O. Gardner, and Michael C. Neale, Frontiers of Evolutionary Psychology, July 2011, Vol., Art 159].
As Peter Simons summarizes, “The prevailing theory on why antidepressants might make depression worse is receptor sensitization—the idea that long-term use modifies the ways that neuroreceptors work, causing the medication to become ineffective, and potentially making people vulnerable to worsening depression.”
What about short-term outcomes? While the above review focuses almost exclusively on long-term outcomes, it would be a mistake to conclude that short-term outcomes are a settled matter when it comes to antidepressants. While there is growing awareness of the unique risks of even short-term antidepressant use on teenage brains, the research literature on short-term antidepressants for adults has been widely assumed to be a settled matter. However, when more critical analyses are conducted of direct clinical studies, particularly from scholars independent of industry monies, the conclusions about both safety and effectiveness vary considerably from the predominant consensus (see, for instance, this recent set of reviews published in 2018).
It’s become more well-known that previous research has also found that antidepressants are no more effective than placebo for mild-to-moderate depression, with other studies questioning whether these medications are as effective as we thought for severe depression. More substantial examination has also gone to health risks of taking antidepressants—such as a recent study which found that taking antidepressants increases one’s risk of death by 33% (see MIA report).
 Hyman, S. “Initiation and adaptation: A paradigm for understanding psychotropic drug action.” Am J Psychiatry 153 (1996):151-61.
 El-Mallakh, R. “Tardive dysphoria: The role of long-term antidepressant use in inducing chronic depression. Medical Hypotheses 76 (2011): 769-773.
 Giovanni Fava, “The mechanisms of tolerance in antidepressant action.” Progress in NeuroPsychopharmacology & Biological Psychiatry 35 (2011): 1593-1602.