PCOS and Fatty Liver: Prevalence, Screening and What Reverses It

How common is fatty liver in PCOS?

Roughly 43% of women with PCOS have fatty liver, according to the largest pooled estimate to date. A 2023 systematic review and meta-regression of 36 studies screened 817 articles and calculated a pooled prevalence of 43% (95% CI, 35–52%), though the studies varied widely in how they defined and detected it, which is reflected in a wide confidence interval. An earlier, more tightly controlled analysis found the same pattern from a different angle: a 2017 meta-analysis of 17 studies, comparing 2,734 women with PCOS against 2,561 controls matched for age and BMI, found PCOS roughly doubled the odds of fatty liver (odds ratio 2.54, 95% CI 2.19–2.95) even after accounting for weight. That BMI-matched design matters: the excess risk isn’t just “PCOS involves more obesity,” since the comparison already controlled for that.

What is MASLD, and is it the same thing as NAFLD?

MASLD is the current name for the condition most PCOS research still calls NAFLD, and the change is a diagnostic tightening, not a rebrand. A 2023 multisociety Delphi consensus involving 236 panelists from 56 countries voted to replace “non-alcoholic fatty liver disease” with metabolic dysfunction-associated steatotic liver disease, or MASLD — 74% of respondents judged the old name flawed enough to warrant a change, and majorities found “nonalcoholic” and “fatty” stigmatizing terms. The practical difference is the entry criteria: a MASLD diagnosis requires liver fat on imaging or biopsy plus at least one of five cardiometabolic risk factors — a category that includes obesity, elevated blood pressure, elevated triglycerides, low HDL cholesterol, or prediabetes/type 2 diabetes. NAFLD, by contrast, only required ruling out other causes of liver fat, mainly alcohol. A related new term, MetALD, describes people who meet the MASLD criteria but also drink alcohol above a set threshold. Almost everyone diagnosed with NAFLD under the old definition also meets the MASLD one, so the change mostly affects terminology and research classification, not who gets caught by screening.

The same kind of consensus renaming happened to PCOS itself. In May 2026, a global consensus of more than 50 medical organisations — published in The Lancet — renamed PCOS polyendocrine metabolic ovarian syndrome, or PMOS. Same condition, same mechanism; this article uses PCOS because that’s still what most readers search, the same way most liver research still says NAFLD out of habit even after the 2023 change.

Is it insulin resistance or high androgens driving the fat into the liver?

Both drive it, and the research is specific about how much each contributes rather than picking one winner. The mechanism runs largely through insulin: chronic hyperinsulinemia increases free fatty acid delivery to the liver and drives unrestrained hepatic lipogenesis even as insulin’s normal brake on glucose production fails — the hallmark of what’s called selective hepatic insulin resistance. The 2023 meta-regression above found that rising NAFLD prevalence tracked most closely with rising HOMA-IR (the standard insulin-resistance index), free androgen index and total testosterone across studies — insulin resistance and androgen excess moving together, not one substituting for the other.

Androgens add an independent effect on top of that. The 2017 meta-analysis found that among women with PCOS, those with fatty liver had meaningfully higher total testosterone (mean difference 0.40 nmol/L) and free androgen index (mean difference 4.46) than those without it — and multivariate analysis that adjusted for age, BMI, triglycerides and insulin resistance still found androgens an independent predictor. The same analysis found the hyperandrogenic phenotype of PCOS carries a higher fatty-liver prevalence than the non-hyperandrogenic phenotype. If your PCOS presents with high androgens — significant acne or hirsutism alongside labs showing elevated testosterone — that phenotype carries more of this specific risk than a phenotype without androgen excess, independent of weight.

Does PCOS cause elevated liver enzymes?

Liver enzymes run higher in PCOS, but only in the overweight or obese subgroup — not across the board. A 2009 study of 83 women with PCOS and 64 controls, split by BMI into lean and overweight/obese subgroups, found no enzyme difference at all between lean PCOS and lean controls. Among the overweight/obese subgroups, though, ALT ran higher in PCOS (17.7 vs 14.1 U/L) and so did gamma-glutamyl transpeptidase (17.9 vs 13.4 U/L). In that same overweight/obese group, ALT correlated with free androgen index (r=0.25) and total testosterone (r=0.33) — a smaller but consistent version of the androgen link described above.

An ALT or GGT reading a lab flags as “high normal” is worth a second look in PCOS specifically, not dismissed as a lab-to-lab quirk — it’s exactly the pattern this study associated with the metabolic and androgen picture above.

What actually gets screened, and what does the PCOS guideline say about it?

There is no PCOS-specific liver screening protocol — the guideline this article is checked against doesn’t carry one. The 2023 international PCOS guideline sets out detailed cardiometabolic screening recommendations for PCOS — glucose tolerance testing, lipid panels, blood pressure — but does not include a specific recommendation for liver-disease screening. What exists instead is general hepatology guidance that PCOS happens to qualify for on metabolic grounds. The AASLD’s 2023 practice guidance recommends risk-factor-based screening — using the FIB-4 index, a simple calculation from age, AST, ALT and platelet count — for anyone with prediabetes, type 2 diabetes, two or more cardiometabolic risk factors, or an unexplained elevation in liver enzymes. PCOS with obesity, insulin resistance or dyslipidaemia meets that bar through the underlying risk factors, even though PCOS itself isn’t named as a trigger.

Everything above describes fatty liver at the stage screening is meant to catch it — before it has produced any symptom you’d notice. Advanced liver disease does eventually announce itself, and those symptoms are worth knowing regardless of where your FIB-4 score sits.

What actually reverses it

Weight loss reverses measurable liver fat in dose-dependent steps, not all-or-nothing. The AASLD’s 2023 guidance states that losing 3–5% of body weight improves steatosis, while more than 10% is generally required to improve steatohepatitis and fibrosis — and notes plainly that fewer than 10% of patients sustain that larger amount of weight loss at one year even with structured support. That’s a general hepatology finding, not a PCOS-specific trial, but the mechanism — reduced free fatty acid flux to the liver as insulin sensitivity improves — is the same pathway described above for PCOS. Since insulin resistance is the shared engine behind fatty liver, dyslipidaemia and central fat storage in PCOS, treating it tends to move more than one marker at once; PCOS and cholesterol covers the lipid side of the same pattern, and waist-to-hip ratio in PCOS covers the central-fat side, which tracks with liver fat through the same visceral-fat pathway. GLP-1 receptor agonists are also being studied for liver fat specifically, though that evidence and the separate question of who can actually access one are worth reading on their own rather than folded in here.

Where this will not apply to you

If you have lean PCOS with a BMI under 25, the enzyme data above found no measurable liver-enzyme difference from lean women without PCOS at all — this isn’t a risk that applies evenly across every PCOS phenotype, and a lean presentation with no other metabolic risk factors doesn’t automatically warrant liver screening on PCOS grounds alone. The screening approach above also doesn’t apply if you don’t have prediabetes, type 2 diabetes, or at least two cardiometabolic risk factors — FIB-4 is a risk-stratification tool for people who already qualify for it, not a universal test everyone with PCOS needs annually. And none of the prevalence figures above tell you your own liver’s current state; they describe group-level risk, not an individual diagnosis, which only imaging or a clinician’s assessment of your actual labs can do.

Your next step

At your next PCOS-related bloodwork, ask specifically for ALT, AST and GGT to be included if they aren’t already, and ask your clinician to calculate a FIB-4 score if any of the three come back above the lab’s reference range. This is one entry in the site’s weight-loss section, which treats liver fat the same way it treats waist size and cholesterol — as a metabolic marker worth tracking accurately, not a verdict on anything you did.