Antidepressants and Weight Gain in PCOS: What Differs by Drug

Does PCOS Make This Question More Likely to Come Up?

Women with PCOS have 3.78 times the odds of any depressive symptoms and 5.62 times the odds of any anxiety symptoms compared with women without it, independent of weight, in a meta-analysis of close to 7,000 women across 30 studies (Cooney et al., 2017). That prevalence gap — covered in full, with the screening recommendation behind it, in PCOS, depression and anxiety — means this readership encounters antidepressant prescriptions more often than the general population does. What almost no PCOS-focused page covers is that the weight effect of that prescription depends heavily on which drug it is, not on whether “antidepressants cause weight gain” as a single category. The 2023 international guideline names psychological features, including depression and anxiety, as a core disease feature of PCOS requiring recognition and screening at diagnosis, not an incidental side issue (Teede et al., 2023) — which is part of why an antidepressant prescription is common enough in this population to deserve its own page on what it actually does to weight.

Note: in May 2026, PCOS was renamed polyendocrine metabolic ovarian syndrome, or PMOS, by a global consensus of more than 50 organisations. Same condition, same pharmacology — only the label changed. This article uses PCOS, since that is still the term most readers search.

Which Antidepressants Carry the Highest Weight-Gain Risk?

Mirtazapine, paroxetine, and amitriptyline showed a significantly greater risk of weight gain than other antidepressants, in a comprehensive meta-analysis pooling 116 studies on antidepressants and body weight (Serretti & Mandelli, 2010). The review’s own framing matters: these three were flagged specifically because their weight-gain signal was strong and consistent enough to separate from the rest of the field, not because every other antidepressant is weight-neutral by comparison. The same review found that fluoxetine and bupropion were associated with some weight loss, though fluoxetine’s effect appeared to be mostly limited to the acute treatment phase rather than continuing indefinitely. Most other compounds showed a transient or negligible short-term effect that, the authors noted, “became more evident in the long term to a variable degree across compounds” — meaning a drug that looks weight-neutral at eight weeks is not guaranteed to still look that way at a year.

What Does Real-World, Longer-Term Data Actually Show?

Two large studies measured what actually happened to real patients’ weight over a year or more, and one of them complicates the short-term picture above rather than confirming it. A 12-month electronic health records study of 22,610 adults starting one of 11 antidepressants found that, relative to citalopram, people prescribed bupropion, amitriptyline, or nortriptyline gained significantly less weight (all p ≤ .02) — no other drug in the study, including mirtazapine and paroxetine, showed a statistically significant difference from citalopram over that year (Blumenthal et al., 2014). Amitriptyline landing in the “less weight gain” group here, opposite to its short-term “highest risk” flag in the review above, is the real state of the evidence, not an error to smooth over: short-term trial data and 12-month real-world prescribing data do not always agree, and a reasonable read is that individual dosing, the condition being treated, and how long someone actually stays on a drug all shift the outcome in ways a single average can’t capture.

Is Avoiding the Medication the Safer Choice for Weight?

No — untreated depression carries its own, separately documented metabolic cost. Depression predicted the later onset of obesity with an odds ratio of 1.58 across 15 longitudinal studies and almost 59,000 people (Luppino et al., 2010), through mechanisms that have nothing to do with medication: appetite change in either direction, disrupted sleep, and reduced activity are core features of depression itself, not side effects of treating it. Weighing “the drug might cause weight gain” against “doing nothing has no weight consequence” is comparing against a baseline that does not exist. Undertreating depression or anxiety specifically to avoid a possible weight effect is not a neutral or safer trade — it exchanges a manageable, monitorable risk for a less visible one.

Why Do the Numbers Move Over Time, Not Just Between Drugs?

The Gafoor cohort found the elevated weight-gain risk from antidepressant use persisted through at least six years of follow-up, not just the first months of treatment — this is a chronic, cumulative effect for the drugs that carry one, not a short adjustment period that resolves on its own. That is a different claim from Serretti’s finding that some drugs’ short-term profile changes by the one-year mark; both are true at once. A drug that looks weight-neutral at eight weeks can still be showing a real, sustained effect — favourable or unfavourable — years later, which is exactly why a single early checkpoint isn’t enough to settle the question for any individual person.

What Is Actually Worth Raising at Your Next Review?

Three things belong in that conversation, and none of them are a request to change medication on the spot. First, ask which weight-gain category your specific current drug falls into, using the tables above as a starting point rather than a verdict — your prescriber can weigh it against why that drug was chosen for you in the first place. Second, ask whether a baseline weight was recorded when you started, and request one now if not, since both large studies above measured change from a starting point, not an absolute number — without a baseline, “weight gain” isn’t measurable at all. Third, if weight change has been substantial and troubling, ask directly whether a weight-neutral or weight-favourable option — bupropion is the most consistently supported example across both major sources here — is appropriate for your specific diagnosis and history, understanding that not every option fits every person or every condition being treated.

Who This Does Not Help, and Where the Evidence Runs Out

None of this applies cleanly to someone already stable and doing well on a drug flagged here as higher-risk: switching a working treatment purely to chase a weight profile carries its own risk of relapse, and the evidence above describes group averages, not a guarantee about what will happen to any one person. The Blumenthal and Serretti findings also disagree with each other on amitriptyline specifically, which means neither study alone is a complete answer — a genuinely honest limitation, not a reason to distrust both. Pregnancy, breastfeeding, and a personal history of disordered eating all change this calculus in ways this page does not cover; each needs its own direct conversation with a prescriber rather than an extrapolation from the numbers above.