HTR2C Genotype

If you or someone close to you is about to start clozapine, olanzapine, or risperidone, one question matters more than almost any other: how much weight is this going to add? Some people gain a few pounds. Others gain enough to develop the blood sugar and blood pressure problems that follow. This genotype is one of the few things available in advance that may shift the odds.

HTR2C, the gene for the serotonin 2C receptor, is a research-grade marker, not a guideline-endorsed test. No major professional guideline recommends testing it before prescribing, and every recent systematic review reaches the same conclusion: no single HTR2C variant is ready to change prescribing. Metabolic monitoring is already recommended for everyone starting a second-generation antipsychotic, whatever their genotype. What the pooled data can do is tell you how much attention that monitoring deserves in the first months.

What the Receptor Does

The serotonin 2C receptor is a cell-surface protein that responds to serotonin. In the brain it helps regulate appetite, glucose control, and energy balance. Many second-generation antipsychotics block this receptor, which is one reason clozapine and olanzapine can make people hungry.

The gene is on the X chromosome. That matters more than it sounds. Men carry a single copy, so whatever variant they have is the only one they express. Women carry two, and which one is active varies cell to cell through random X inactivation. This may help explain why the findings keep splitting by sex, with some variants predicting metabolic problems in women but not men, or with stronger signals in men.

Genetic testing reads the DNA. It does not measure how much receptor you have or how active it is. A small postmortem brain study looked directly for that link and found none: individual variants and nearby inherited patterns did not explain baseline receptor gene expression or the small RNA changes that can tune receptor function. So for the common variants on these panels, the genotype is a probability shift, not a readout of your serotonin system.

Antipsychotic Weight Gain

This is where the evidence is strongest. A meta-analysis of 27 studies of people on clozapine found that carriers of the T allele at rs3813929, the promoter variant often written as -759C/T, gained less: about 0.6 BMI points less than non-carriers, with the effect most apparent in men. For someone 5 feet 9 inches tall, that's roughly four pounds of difference. A separate meta-analysis focused on antipsychotic-induced weight gain reached the same direction, with T-allele carriers less likely to gain substantially.

Across the broader pharmacogenetic literature, HTR2C is one of the better-supported weight-gain genes alongside ADRA2A, DRD2, and MC4R. A 2026 meta-analysis of antipsychotic-induced weight gain named rs3813929 and ADRA2A rs1800544 as the two largest effects it found.

The direction matters. The T allele is the protective one here. Carrying the common C allele is the unremarkable result, not the good news, and that is the opposite of how most people read a genetic report.

Metabolic Syndrome and Uneven Replication

A second variant, rs1414334, points at a more concrete endpoint. In the same clozapine meta-analysis, C-allele carriers had roughly twice the odds of developing metabolic syndrome, the cluster of high blood sugar, high triglycerides, low HDL, high blood pressure, and a larger waist that drives heart disease and diabetes.

Doubling the odds of metabolic syndrome is not a small thing for someone facing years on an antipsychotic. Two caveats sit next to it. The C allele is carried by roughly 10% of people in the United States and about 15% in Europe, but only around 1% of people of Asian ancestry, so the finding applies to a minority of patients and to almost none in some populations. And the replication record is uneven. A study of 166 people on atypical antipsychotics found no significant link between rs1414334 and metabolic syndrome at all.

Individual cohorts often come up empty. Among 108 Thai psychiatric patients taking risperidone alone, the -759C/T variant predicted neither weight gain nor metabolic syndrome. Drug, dose, ancestry, and sample size all move these results around, which is why pooled analyses carry more weight here than any single cohort, and why no result from this gene should be treated as a forecast.

Why Sex Changes the Answer

In 456 people with schizophrenia on olanzapine, risperidone, or clozapine, the C allele at rs498177 raised metabolic syndrome risk in women, and the effect survived correction for multiple testing. In men, it did not. A two-variant pattern combining rs521018 and rs498177 looked protective instead.

An older replication study in 186 people on antipsychotics also found the rs1414334 C allele associated with metabolic syndrome. Put the X chromosome biology next to these results and the picture is coherent: which HTR2C variant matters depends on whether you have one copy or two, and possibly on which copy is active in the tissue that counts.

Reconciling the Contradiction

The largest study of HTR2C in the general population, 20,981 adults in the EPIC-Norfolk cohort, found that common HTR2C variants do not meaningfully affect obesity or mental health traits. An initial association between -759T and major depression vanished on full-cohort replication. Postmortem brain sequencing in psychiatric cases turned up no disease-specific mutations either.

Both findings are true because they answer different questions. The common variants tested here do not make you fat or depressed on their own. They shape how your body responds to a drug that blocks the receptor those variants encode. For these variants it is a drug-response marker, not a disease-risk marker, and without the pharmacological trigger the genotype usually does little by itself.

Rare variants in the same gene are a different story. Researchers have identified rare loss-of-function changes in HTR2C, ones that knock out the receptor's activity rather than nudging it, that cause severe obesity starting in childhood along with behavioral difficulties, inherited in an X-linked pattern. Those are rare, they are found far more often in men, and they are not the common promoter and coding variants these pharmacogenetic panels report. But they show the receptor does control body weight in humans directly, which is part of why the drug-response signal is biologically plausible in the first place.

What the Genotype Also Touches

Two findings outside antipsychotic weight gain are worth knowing about, both from single lines of evidence and neither ready to act on.

• Stress hormone response: in two samples totaling 313 people, men with a single Cys23 copy and women with two showed higher cortisol release during a standardized social-stress test. Cortisol is the hormone your adrenal glands release under pressure.
• Dopamine release: in 54 healthy adults given a pain challenge during brain imaging, Ser23 carriers released more dopamine in reward-related parts of the striatum. The variant accounted for about 12% of the variation in one of those regions, which is a large share for a single common variant.

Neither finding has been connected to a clinical outcome you can do anything about. Response data are thinner too. One clozapine study in 185 people found no link to rs6318, while a later analysis reported sex-specific response signals. None of that has become usable prescribing guidance. A study of 195 people found no link to alcohol dependence. Note also that rs6318 (Cys23Ser) is a coding change in the receptor itself, mechanistically separate from the -759C/T promoter variant behind the weight-gain meta-analyses; the two are easy to conflate on a report and should not be.

One Test, For Life

Your genotype will not change, so there is nothing to retest. The value lies in what you do with it over years, and what needs repeating is the phenotype, not the gene.

If you're starting or already taking an atypical antipsychotic, the variables that actually determine your outcome are weight, waist, fasting glucose, HbA1c, lipids, and blood pressure. Those directly measure the organ dysfunction the genotype only hints at. In 13,554 adults, plasma metabolic patterns separated pre-metabolic syndrome and metabolic syndrome risk groups. That is the point: direct metabolic data answer today's question better than this genotype can.

Get baseline measurements before the first dose. The ADA and APA consensus schedule then calls for weight at 4, 8, and 12 weeks and every three months after that, blood pressure and fasting glucose again at 12 weeks and then yearly, and fasting lipids at 12 weeks and then at least every five years, though many clinicians now repeat lipids annually. That early cadence applies to everyone on these drugs, genotype aside. A higher-risk result is a reason to weigh yourself more often and to repeat labs closer to every three months through the first year, not the thing that triggers monitoring in the first place.

What to Do With an Unexpected Result

A higher-risk genotype does not mean avoid the drug. Clozapine is often the only thing that works for treatment-resistant schizophrenia, and no HTR2C result should override that. What it should do is change how closely you watch and lower your threshold for acting early.

Order the companion labs yourself: fasting glucose, HbA1c, insulin, a full lipid panel, and blood pressure at home rather than once a year in an office. Track your weight weekly for the first three months, since that early trajectory predicts where you land. If your weight climbs fast and your fasting glucose or triglycerides start drifting in the same direction, bring that combination of findings to your prescriber rather than waiting for a diagnosis. Weight-gain management, metformin, a dose change, or a switch to a different antipsychotic are all real options. Metformin is the best-studied of them, averaging around 3 kg more weight loss than placebo in pooled trials and working best when it is started early rather than after the weight has accumulated. The argument for any of these is far stronger with three months of your own data behind it.

Most CYP2D6 and CYP2C19 pharmacogenetic panels people order for psychiatric drugs look at how fast you clear a drug, a different question from how your receptors respond to it. Those metabolizer genes are also the only antipsychotic pharmacogenetics with actionable prescribing guidance today. If you want both pieces, you need both tests.

Where This Result Can Mislead You

Sequencing and genotyping have limits, and a clean report means less than it looks like.

• Panel coverage: the assay finds what it was designed to find. Most of the clinical value here comes from named variants such as rs3813929 and rs1414334. If your assay does not include a position, a negative report says nothing about it. Rare changes elsewhere in the gene are not interpretable for antipsychotic weight gain, and the rare loss-of-function variants tied to severe childhood obesity are usually found through broader sequencing, not a pharmacogenetic panel.
• Ancestry: allele frequencies differ by population. The rs1414334 C allele, for instance, is about fifteen times more common in Europe than in Asian populations. Exome sequencing of 64 Amazonian Indigenous people found HTR2C variants at frequencies far from continental reference groups. Most HTR2C pharmacogenetic work comes from Chinese, Thai, and European cohorts, so how well it transfers to your background depends on which cohorts resemble you.
• Uncertain variants: you may get back a change nobody can interpret yet. That is not a finding; it is a placeholder.
• Sample source: buccal, saliva, and blood usually give the same inherited call. A donor stem-cell transplant or a contaminated sample can break that rule, so an unexpected result should be confirmed from a clean germline sample.
• Confounding: sex, diet, activity level, and other medications all shift metabolic outcomes. Individual cohorts can lose their HTR2C associations once those factors are included, and age and starting body mass index remain the most consistent predictors of who develops metabolic syndrome on these drugs.

Be skeptical of consumer genotype reports covering these positions. A clinical report should tell you exactly which variants were assayed, whether a low-confidence call was confirmed, and which regions were not covered. For this gene, the usual risk is not that food or stress changed the result. The risk is that the report overinterprets a variant.