Serotonin and DepressionA Disconnect between the Advertisements and the Scientific Literature

Depression is a serious medical condition that may be caused by a chemical imbalance in the brain. Paxil can help. That sentence appeared in millions of American homes — on television, in magazines, and on websites. It sounds like medicine. It sounds like settled science. According to Jeffrey Lacasse and Jonathan Leo, writing in two thousand five, it is not supported by the scientific literature. It is not contested at the margins. It is not awaiting confirmation. It is simply not there. That gap — between what the ads say and what the research actually shows — is what their paper maps, in careful, uncomfortable detail. Start with the history, because the chemical imbalance story didn't arrive from nowhere. In nineteen sixty-five, a psychiatrist named Joseph Schildkraut proposed that depression might be associated with low levels of norepinephrine, a neurotransmitter. Researchers later shifted their attention to serotonin. Over the following decades, that idea migrated from specialist journals into mass marketing, and a hypothesis became, in the public mind, a fact. The drugs built around that narrative are the selective serotonin reuptake inhibitors. The consumer explanation goes like this: serotonin is a chemical messenger sent from one nerve cell to the next. Normally, the receiving cell returns some of that serotonin to the sending cell. In depression, the ads say, too much serotonin gets reabsorbed, leaving a deficit. An SSRI blocks that reabsorption. More serotonin stays available. Balance is restored. It's a clean story. Clean stories sell. And sell they did. Lacasse and Leo report that Zoloft alone was the sixth best-selling medication in the United States in two thousand four, with over three billion dollars in sales. One study found that more than seventy percent of surveyed patients reported exposure to antidepressant direct-to-consumer advertising. That is a lot of people receiving a lot of confident information. Here is what that information actually sounded like. Celexa's advertising stated that the drug "helps to restore the brain's chemical balance by increasing the supply of a chemical messenger in the brain called serotonin." Lexapro's copy explained that "in people with depression and anxiety, there is an imbalance of serotonin — too much serotonin is reabsorbed by the first nerve cell." Prozac's ads told consumers that "when you're clinically depressed, one thing that can happen is the level of serotonin may drop" and that Prozac would "help bring serotonin levels closer to normal." Paxil stated flatly that it "works to correct the chemical imbalance believed to cause the disorder." Zoloft told viewers that "Zoloft works to correct this imbalance. You just shouldn't have to feel this way anymore." That last line is worth sitting with. You just shouldn't have to feel this way anymore. It's not just a mechanism claim. It's a promise tied to a biological explanation — one that turns suffering into a solvable chemistry problem. Now here is the problem. Lacasse and Leo examined the actual scientific literature, and the serotonin deficit model does not hold up. Studies measuring serotonin metabolites in the cerebrospinal fluid of depressed patients produced inconsistent findings, and a German Medical Board review concluded that the reported associations were likely premature given the flawed methodology — small sample sizes and uncontrolled confounding variables. When researchers tried to induce depression experimentally by depleting serotonin, the results were inconsistent. When they tried the opposite — administering high-dose L-tryptophan to flood the brain with serotonin — that failed to relieve depression too. The experiment worked in neither direction. The prescribing information for paroxetine — Paxil — actually reflects this uncertainty, if you read it carefully. It states that the drug's efficacy "is presumed to be linked to potentiation of serotonergic activity." Presumed. That word does not appear in the advertisements. And then there is the clinical trial evidence, which Lacasse and Leo present in full. Irving Kirsch and colleagues obtained all company-sponsored trials submitted to the Food and Drug Administration, including the unpublished ones. When those trials were pooled together, placebo duplicated about eighty percent of the antidepressant response. Fifty-seven percent of the submitted trials failed to show a statistically significant difference between antidepressant and inert placebo. A Cochrane review suggested even those results might be inflated, because most trials used an inactive placebo — meaning patients could often tell whether they had received the drug, which contaminates the comparison. Other results cut against the serotonin-specific story even more directly. A Cochrane systematic review found no meaningful difference in efficacy between selective serotonin reuptake inhibitors and tricyclics — a much older class of antidepressants that work differently. Bupropion and reboxetine, which do not significantly affect serotonin, performed just as well as the selective serotonin reuptake inhibitors in randomized trials. In recent trials, St. John's Wort and placebo outperformed selective serotonin reuptake inhibitors. One randomized trial found exercise to be as effective as the selective serotonin reuptake inhibitor sertraline. Eli Lilly, the maker of Prozac, shifted to duloxetine — a drug that targets both norepinephrine and serotonin — which Lacasse and Leo read as a sign that even the industry was backing away from a purely serotonergic explanation. Lacasse and Leo quote researchers across the field saying as much. Elliot Valenstein, David Healy, Kenneth Kendler, Peter Kramer — scientists who differ on many things — all made variations of the same point: there is no established evidence that a serotonin deficiency causes depression. The Diagnostic and Statistical Manual, psychiatry's official diagnostic reference, does not list serotonin as the cause of any mental disorder. Lacasse and Leo state, in their own words, that to their knowledge, there is not a single peer-reviewed article that can be accurately cited to directly support claims of serotonin deficiency in any mental disorder. That is not a small claim. It means the central selling point of a multi-billion-dollar advertising enterprise lacks a direct citation in the scientific literature. So why were these ads allowed to run? Lacasse and Leo walk through the regulatory structure, and the answer is partly institutional. The Food and Drug Administration does not require preapproval of pharmaceutical advertisements. It monitors ads after they appear and enforces rules reactively — typically in response to complaints. Federal regulations do prohibit advertising claims about mechanisms of action that are not established unless the ad discloses the limitations of the supporting evidence. In their review, Lacasse and Leo found not a single selective serotonin reuptake inhibitor advertisement that disclosed such limitations. The Food and Drug Administration did issue ten warning letters to antidepressant manufacturers between nineteen ninety-seven and the paper's publication — covering Effexor, Paxil, Remeron, Sarafem, Celexa, and Zoloft. But the agency never cited a pharmaceutical company specifically for the kind of chemical imbalance claims under examination here. One Food and Drug Administration regulatory reviewer told the authors, in an email, that the reductionistic language was used to explain putative mechanisms "to the fraction of the public that functions at no higher than a sixth-grade reading level." That explanation does not resolve the tension. It deepens it. Ireland, notably, did something different. The Irish Medical Board banned GlaxoSmithKline from claiming that paroxetine "corrects a chemical imbalance" even in patient information leaflets. A whole country decided that was too far. The United States, with its much larger market, did not reach the same conclusion. The stakes of all this extend beyond regulatory policy. When patients believe their suffering has a known, specific, neurochemical cause, it shapes what they ask for and what they accept. Lacasse and Leo cite a randomized trial by Kravitz and colleagues in which people presenting with adjustment disorder symptoms were frequently prescribed paroxetine when they asked specifically about it — a direct demonstration that advertising-driven patient requests affect prescribing. The UK's National Institute for Clinical Excellence had recommended non-pharmacological treatment for mild depression. But patients who believe they have a chemical deficiency may be skeptical of cognitive behavioral therapy or other alternatives, which complicates shared decision-making at its root. Informed consent requires accurate information. If the information patients receive is a commercial narrative that the scientific literature cannot support, then consent is something less than fully informed — regardless of whether any individual drug helps any individual patient. Lacasse and Leo are careful to separate those questions. Their critique targets the advertised mechanism, not a blanket claim that selective serotonin reuptake inhibitors cannot help anyone. That distinction matters. The question the paper raises is precise. It's not: do these drugs do anything? It's: when a patient is handed a brain diagram and told that their illness is a serotonin imbalance that the pill will fix, is that an accurate account of what science knows? According to the literature Lacasse and Leo surveyed, the answer is no. The advertisements said otherwise. And for years, hundreds of millions of dollars ensured that version of the story reached you first. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.