If the weight arrived after you started a medicine for depression, epilepsy, blood pressure, diabetes or contraception, the most likely explanation is the one nobody offered you. For several widely prescribed drug classes, weight gain is a measured, repeatedly replicated pharmacological effect with a known mechanism, a known time course, and a mean figure attached to it in randomised trials. It is a property of the drug, not a finding about your character.

What follows is what the controlled evidence shows class by class, which drugs are blamed without evidence to support it, and why every route out of this leads back to the prescriber. Nothing here is a reason to change a dose.

• Clozapine, olanzapine, divalproex, and DMPA carry the strongest randomised evidence for weight gain, with mean gains of 3.5 to 5.8 kg in controlled trials.
• Several drugs are blamed without solid evidence to support it. Combined oral contraceptives and lithium show much smaller effects than commonly assumed.
• The mechanisms differ by drug class: some act on appetite signalling, others on fat storage, and others on energy expenditure.
• Stopping a medicine to address weight gain can carry serious risks (relapse, seizures, loss of glycaemic control) that often outweigh the benefit.
• Drug-induced weight gain is a prescriber's problem to solve, not a dietary one. A food diary cannot detect or undo a pharmacological effect.

Clozapine showed a mean weight gain of 4.45 kg over 10 weeks in a meta-analysis of 81 clinical trials using standard doses. It was one of the highest weight-gain drugs in the analysis.

Olanzapine was associated with a mean weight gain of 4.15 kg over 10 weeks. The evidence comes from the same meta-analysis of 81 trials, making it another high-risk antipsychotic for weight gain.

Risperidone produced a mean weight gain of 2.10 kg over 10 weeks in the same meta-analysis. The average gain was lower than clozapine and olanzapine but remained clinically meaningful.

Ziprasidone showed an almost neutral effect on weight, with a mean gain of just 0.04 kg over 10 weeks. In the same analysis, the placebo groups actually lost 0.74 kg.

Divalproex Sodium was linked to a mean weight gain of 5.8 kg over 32 weeks in a randomised monotherapy trial. In comparison, the lamotrigine group gained only 0.6 kg over the same period.

Oral Glucocorticoids were associated with a mean weight gain of 3.6 kg in younger women over months to years. The evidence comes from a primary-care cohort involving 31,516 adults.

DMPA Injection resulted in a mean weight gain of 3.5 kg over 12 to 18 months in a randomised trial of 7,829 women. Women using a copper IUD in the same trial gained 1.5 kg on average.

Insulin for Type 2 Diabetes - Starting insulin therapy was associated with a mean weight gain of 3 to 5 kg over 6 to 12 months across multiple randomised initiation trials.

Pregabalin was linked to a mean weight gain of 2.65 kg over 20 weeks in a randomised neuropathy trial. The placebo group, by comparison, lost 0.55 kg.

Mirtazapine produced a mean weight gain of 1.7 to 2.0 kg within 4 to 12 weeks across short-term randomised trials, with most of the gain occurring early in treatment.

Older Beta-Blockers were associated with an average 1.2 kg excess weight gain over six months or longer in eight randomised hypertension trials.

Pioglitazone showed a dose-dependent weight gain ranging from 0.3 to 3.6 kg over 16 to 24 weeks in type 2 diabetes trials, with higher doses linked to greater weight gain.

A mean hides the thing that matters most: the response is not evenly spread. Across 201 antipsychotic studies, the proportion of patients crossing a 7% weight increase ran from around 2% on short-term blonanserin or brexpiprazole to nearly 50 to 76% on long-term clozapine and first-generation agents.

For valproate, retrospective cohorts found 57 to 71% of treated adults gaining more than 4 kg, with mean gains of 4.9 kg in moderate gainers and 11.4 kg in marked gainers. Some individuals gained over 30 kg.

In insulin initiation trials, roughly 20 to 25% of patients gained 5 kg or more, and the gain tracked the fall in HbA1c at about 0.9 to 1 kg per 1% reduction. A meta-analysis of 31 randomised corticosteroid trials covering 1,740 patients put the pooled prevalence of unintentional weight gain at 13%, with odds of gain about five times placebo and a mean BMI increase of 1.6 kg/m² over roughly 22 weeks.

Within classes, agents differ sharply. A comprehensive meta-analysis of 116 antidepressant studies found amitriptyline, mirtazapine and paroxetine carrying the highest weight-gain risk, while fluoxetine and bupropion were associated with short-term loss and more neutral weight over longer use. A 12-month multicentre cohort comparing two sulfonylureas found a mean change of -2.0 kg on glimepiride against -0.6 kg on glibenclamide at equivalent glycaemic control. None of this is visible in a food diary, which is where most people are told to look.

Some drugs carry a reputation the controlled data do not support, and that is worth as much to a reader as the confirmed cases.

Combined oral contraceptives and progestogen-only pills are the clearest example. Large reviews and guideline statements conclude they do not show consistent, clinically important weight gain against placebo or non-hormonal methods, with mean differences generally under 1 kg and wide individual variation. The injectable progestogen DMPA is the genuine exception in that family, and the contrast between them is why the two should not be discussed as one thing. Why so many women still report gain on the pill has not been resolved. Disentangling a drug effect from age-related and lifestyle change remains an open question.

Lithium is similar. A systematic review and meta-analysis of 20 studies covering 10,812 participants found a mean lithium-associated increase of only 0.46 kg, not statistically significant, and significantly less than several alternative mood stabilisers. Individual maintenance trials still report around 2.2 kg over 52 weeks with 11.8% of patients gaining 7% or more, so a subset does gain. How that distribution splits is not well characterised.

Carbamazepine and gabapentin sit in a third category, where the direction is suspected but reliable randomised magnitude data are absent. First-generation antihistamines sit there too: an observational association between chronic sedating antihistamine use and higher body weight, waist circumference and insulin exists, but the research behind this article could trace it only through secondary summaries rather than to the primary study, so treat both its size and its causal direction as unverified. Cyproheptadine is the exception, having been used deliberately as an appetite stimulant in randomised trials.

Two anti-epileptics run the other way. A six-month randomised trial found mean losses of about 5 to 7 kg on topiramate against 2.8 kg on placebo, and a 32-week zonisamide trial showed about 9.2 kg loss against 1.5 kg. That is a fact about those molecules, not a suggestion to request either.

The mechanisms are specific, and they cluster into three groups:

• Histamine H1 and serotonin 2C blockade (antipsychotics, mirtazapine, tricyclics, sedating antihistamines): Increases hunger, alters leptin and ghrelin signalling, and changes the reward response to food. Trials record increased intake in people who did not consciously change their diet.
• Insulin-mediated fat storage (corticosteroids, valproate, pioglitazone, insulin intensification): Acts through a different route, promoting visceral fat and reducing thermogenesis, so the same intake is more likely to be stored.
• Reduced thermogenesis and lipolysis (beta-blockers): Lowers daily energy expenditure by an estimated 50 to 100 kcal. Genetic variants in HTR2C, DRD2 and MC4R modify how strongly any of this lands on a given person.

Sedation and fatigue add a second pathway: less spontaneous movement and more sitting time. One prospective insulin study attributed part of a mean 2.9 kg gain over 12 months to increased sedentary behaviour after starting therapy. Fluid retention adds a third, putting kilograms on the scale that are not fat at all, as seen with pioglitazone and with some antipsychotics and mood stabilisers.

This is the point where the willpower story collapses. Even where eating more or moving less is part of the pathway, that behaviour is downstream of a drug acting on appetite signalling, energy expenditure or alertness. The clustering by drug and by dose in controlled trials is what tells you which end of the chain the cause sits at.

Time course is another signal: antipsychotic gain is fastest in the first 6 to 12 weeks before settling onto a higher but steadier trajectory, corticosteroid gain accumulates over the first few months, valproate gain is apparent by week 10 and continues through eight months, and most of the beta-blocker difference accrues early. A weight trend that flattens does not rule the medicine out, which is a separate question from why weight loss stalls.

This is the section that matters most, because the obvious inference is the dangerous one.

Meta-analyses of antipsychotic dose reduction or discontinuation in schizophrenia show relapse risk roughly two to three times higher than maintenance, with 12-month relapse rates around 50% in discontinuation arms against 20% in maintenance, most of it in the first year.

In epilepsy, abrupt valproate discontinuation markedly increases seizure recurrence and, in severe cases, the risk of status epilepticus and sudden death. Guidance emphasises gradual, supervised tapering with alternative cover.

Stopping insulin or a sulfonylurea can reverse some of the weight, but the gain is tightly coupled to the improvement in HbA1c, so removing it usually removes the glycaemic benefit and the vascular protection with it.

Do not stop, halve, skip, delay or re-time any of these medicines on the strength of an article. Nothing in this piece is written to be acted on without the person who prescribed the drug.

Everything below is a decision that sits with a Registered Medical Practitioner who knows your diagnosis. It is set out here so you know what exists, not so you can select from it.

Switching within a class has been studied most in psychiatry. Meta-analyses of moving from high-gain agents such as olanzapine to more metabolically favourable ones report mean reductions of only a few kilograms, often without fully reversing what was gained, while increasing the risk of symptom worsening or psychotic relapse. Dose is the second lever: for many antipsychotics weight gain rises with dose and then levels off, but for some, including olanzapine and paliperidone, no clear plateau emerges, and reduction below certain thresholds raises relapse risk.

Adjunctive treatment is best established for antipsychotic-induced gain. A 12-week randomised trial in schizophrenia found metformin at 750 mg/day plus lifestyle intervention reduced BMI by about 1.8 units while placebo patients gained. A meta-analysis of 21 randomised metformin trials in antipsychotic-treated patients (1,547 participants in total) showed a mean weight reduction of roughly 3 kg against placebo. A prevention meta-analysis suggests starting metformin alongside high-risk agents attenuates expected gain by about 3 to 4 kg over 6 to 12 months, though certainty of evidence is rated low to very low.

Timing changes what behavioural support achieves. Intensive dietitian-led programmes begun at the moment insulin was introduced prevented the 4 to 5 kg gain seen in standard care, whereas later or less intensive versions did not. For corticosteroids, a systematic review found only three small randomised lifestyle trials with divergent, low-quality results, so no firm prevention strategy can be drawn from them.

Indian guidance already anticipates this problem in psychiatry. The Indian Psychiatric Society's clinical practice guidelines name clozapine and olanzapine as highest risk for weight, lipids and glucose, and recommend structured monitoring of weight, waist, blood pressure, and fasting glucose and lipids at baseline, at 4 to 6 weeks, at 12 weeks and then at least annually. One Indian hospital-based study found metabolic syndrome in 43% of patients overall and 39% of those on second-generation antipsychotics.

Corticosteroid exposure in India often happens outside any monitoring at all. Cross-sectional Indian dermatology studies report that 42 to 43% of users obtained potent topical steroids over the counter without a prescription, most commonly betamethasone valerate and most commonly for cosmetic use.

Two limits should be stated plainly. Robust national Indian data on drug-induced weight trajectories for insulin, antidepressants, valproate, DMPA and beta-blockers do not yet exist, so almost every figure in this article comes from international trials and single-centre studies. Long-term magnitude data for individual SSRIs in Indian adults are absent altogether.

The most useful version of this conversation is narrow and specific. Tell your prescriber:

• The name of the drug and the date it started
• Any dose changes since, and when they happened
• What the weight did in the weeks that followed

That timeline, set against the known time course for the class, is something a prescriber can work with. It is more informative than any population mean. Weighing a prescription against relapse risk, seizure control and glycaemic control is a prescriber's task, not a dietary one. A nutritionist cannot answer this question.

Bring the weight record, ask whether the timing fits, and ask what monitoring the relevant guideline already recommends for the medicine you are on. If you are also planning to lose weight while staying on it, that combination is itself one of the situations where supervision rather than self-management is the usual advice, because several of these drugs need their dose reviewed as body weight changes.

Drug-related weight gain is something a clinician needs to see. A Voy clinician can review your full medication list alongside your weight history and tell you what is attributable to the drug and what can be addressed within the constraints of your treatment. Book a consultation.

This article is for general information and education only and is not medical advice, diagnosis, or treatment. GLP-1 and other medications referenced are prescription-only and are appropriate only for certain people under the supervision of a qualified clinician. Do not start, stop, or change any medication based on this article. Please consult a registered medical practitioner about your individual circumstances. Information reflects what was available at the time of review and may change.