Antidepressants Are Not Metabolically Neutral

Leon Wirz
Dec 22, 2025
5 min read

The Lancet, November 2025, King’s College London, University of Oxford, University of Bern & international partners

Introduction

Antidepressants are among the most widely prescribed medications in Europe and Switzerland. They are used not only for major depressive disorder, but also for anxiety disorders, chronic pain, sleep disorders, and other psychiatric or psychosomatic conditions. For many patients, they are life-saving therapies.

Yet antidepressants are often discussed as if they mainly act on mood and emotions. Side effects such as nausea, fatigue, or sexual dysfunction are commonly addressed, while broader physiological effects receive far less attention. This is striking, given that many patients receiving antidepressants already have elevated cardiovascular or metabolic risk.

A new, large-scale analysis published in The Lancet now challenges the idea that antidepressants are physiologically interchangeable. The study systematically compares how different antidepressants affect body weight, blood pressure, heart rate, glucose, cholesterol, and liver parameters—providing the most comprehensive overview to date of their non-psychiatric effects.

The Core Discovery

The central message of the study is clear: antidepressants differ markedly in their cardiometabolic and physiological effects.

Across commonly prescribed drugs, the authors observed clinically meaningful differences within only eight weeks of treatment. These differences were not subtle. Depending on the specific antidepressant, patients experienced substantially different changes in weight, heart rate, and blood pressure. In some cases, the contrast between drugs reached levels that are known to influence long-term cardiovascular risk.

Crucially, the study found no association between improvement in depressive symptoms and the degree of metabolic or cardiovascular change. In other words, drugs that worked equally well for mood differed substantially in how they affected the body.

How the Study Was Conducted

The researchers performed a systematic review and network meta-analysis, a method that allows comparison of many treatments even when they have not all been tested head-to-head.

They searched major medical databases and regulatory sources from inception to April 2025 and included randomised, blinded trials comparing antidepressant monotherapy with placebo or with other antidepressants.

In total, the analysis included:

151 randomised controlled trials and 17 FDA study reports
More than 58,000 adult participants
30 different antidepressants
A median treatment duration of eight weeks

Physiological outcomes included body weight, blood pressure, heart rate, glucose, total cholesterol, liver enzymes, electrolytes, kidney markers, and cardiac QT interval. The authors also examined whether age, sex, or baseline body weight influenced these effects.

Because the analysis is based on randomised trials, it provides strong evidence for causal short-term effects, rather than associations driven by confounding or disease severity.

Key Findings

First, weight change differed substantially between antidepressants. Some drugs were associated with clinically relevant weight gain in a large proportion of patients (e.g. maprotiline, amitriptyline, mirtazapine, mianserin), while others led to modest weight loss (e.g. agomelatine, fluoxetine, bupropion, sertraline, venlafaxine, duloxetine). The difference between the most extreme drugs (particularly between agomelatine and maprotiline) approached four kilograms within eight weeks.

Second, cardiovascular parameters showed striking variation. Certain antidepressants markedly increased resting heart rate (notably nortriptyline, clomipramine, imipramine, amitriptyline, doxepin, levomilnacipran), while others reduced it (fluvoxamine, moclobemide). Similarly, systolic and diastolic blood pressure responses differed substantially between drugs, with relevant increases observed for several serotonin–noradrenaline reuptake inhibitors (duloxetine, desvenlafaxine, venlafaxine, levomilnacipran) and tricyclic antidepressants (amitriptyline, imipramine, maprotiline)—changes large enough to matter for long-term stroke and cardiovascular risk.

Third, metabolic markers did not always track with weight change. Some antidepressants reduced body weight but still increased cholesterol or glucose levels (paroxetine, duloxetine, desvenlafaxine, venlafaxine), highlighting that metabolic risk cannot be inferred from weight alone.

Fourth, liver enzymes increased with several antidepressants. Elevations in transaminases and alkaline phosphatase were observed particularly with (duloxetine, desvenlafaxine, levomilnacipran, venlafaxine, paroxetine, bupropion). Although these changes were generally small and not immediately clinically alarming, they point to drug-specific hepatic effects that may become relevant in long-term treatment or in high-risk patients.

Finally, the study did not find strong short-term effects on QT interval, electrolytes, or kidney function for most antidepressants (including SSRIs such as citalopram, escitalopram, sertraline, fluoxetine, as well as SNRIs and atypical agents) in the relatively healthy, middle-aged trial populations studied.

Limitations of the Study

Despite its scale and methodological strength, the study has important limitations.

Most trials were short, typically lasting around eight weeks. This means the analysis captures early physiological changes but cannot determine whether these effects persist, worsen, or normalise over time.

Participants were generally younger and healthier than many real-world patients. Older adults, people with multiple chronic conditions, and those taking several medications simultaneously were under-represented. As a result, real-world risks (particularly for arrhythmias or electrolyte disturbances) may be higher than observed here.

Finally, outcomes were analysed as average changes rather than rare but severe adverse events. This is appropriate for population-level risk assessment but cannot replace post-marketing surveillance or registry studies.

Relevance for Switzerland

The findings are highly relevant for the Swiss healthcare system.

Switzerland combines high antidepressant prescription rates with a population that is ageing and increasingly affected by cardiometabolic disease. Even small drug-induced increases in weight, blood pressure, or heart rate can translate into higher long-term costs through cardiovascular events, diabetes, and multimorbidity.

For health insurers, these results underline that antidepressant choice is not cost-neutral beyond drug price. Downstream effects on chronic disease risk, monitoring needs, and long-term care costs matter. For primary care physicians and psychiatrists, the study supports more differentiated prescribing, especially in patients with existing metabolic or cardiovascular risk.

At a system level, the data argue for closer integration of mental and physical health considerations, an area where Switzerland has traditionally been strong, but where prescribing guidelines often lag behind emerging evidence.

Potential Impacts of a Successful Therapy

The study does not question the effectiveness of antidepressants. On the contrary, effective treatment of depression is associated with reduced suicide risk and lower overall mortality.

The potential impact lies in choosing the right antidepressant for the right patient. If clinicians systematically consider cardiometabolic profiles alongside psychiatric symptoms, it may be possible to maintain antidepressant efficacy while reducing avoidable physical harm. Over time, this could lower cardiovascular disease burden and improve overall health outcomes in people with depression.

Risks

The main risk highlighted by the study is therapeutic inertia. The tendency to treat antidepressants as interchangeable and to continue prescriptions without reassessing physiological effects.

When antidepressants are prescribed without considering metabolic or cardiovascular consequences, patients may accumulate risk silently. Weight gain, rising blood pressure, or subtle metabolic changes often go unnoticed until they contribute to overt disease.

Another risk is over-reaction. The findings should not lead to abrupt discontinuation of effective treatments, but rather to informed, individualised decision-making.

Overall Assessment

This Lancet meta-analysis provides strong evidence that antidepressants differ substantially in their cardiometabolic and physiological effects. These differences are clinically meaningful, emerge early, and are independent of antidepressant efficacy.

The study strengthens the case for personalised prescribing and challenges guidelines that focus almost exclusively on psychiatric outcomes. For clinicians, insurers, and health systems, the message is not to avoid antidepressants, but to use them more intelligently.

What Comes Next

Future research needs to address long-term effects, older and more complex patient populations, and real-world prescribing patterns. Individual-patient data meta-analyses and registry studies will be essential to bridge the gap between trial populations and everyday clinical practice.

In parallel, prescribing guidelines and digital decision-support tools could integrate physiological risk profiles, enabling shared decision-making that reflects both mental and physical health.

Reference

The effects of antidepressants on cardiometabolic and other physiological parameters: a systematic review and network meta-analysis Link

Pillinger, Toby et al.

The Lancet, Volume 406, Issue 10515, 2063 - 2077