Does Keto Help or Harm Fertility?

Nobody knows, because nobody has measured it. Every study of ketogenic eating in women with PCOS has stopped at weight, blood sugar and hormone levels. Not one has reported how many women ovulated more often, conceived, stayed pregnant or had a baby.

That is not a gap you can reason across. A diet that lowers testosterone on a blood test has not thereby been shown to produce a pregnancy, and the two are routinely presented as if they were the same finding. If someone is selling you keto as a fertility treatment, this is the question to ask them.

Two meta-analyses exist, and both are in the same narrow population: women with PCOS who also had raised body weight. Not women trying to conceive in general.

• A systematic review of 10 studies — 3 randomised trials, 1 non-randomised intervention, 4 cohorts and 2 case series. Compared with low-calorie diets, ketogenic protocols lowered serum glucose and the HOMA insulin-resistance index, lowered total cholesterol and triglycerides, and lowered serum LH.
• A separate meta-analysis pooling 11 trials of a very low-carbohydrate ketogenic diet. Weight fell by 9.57 kg on average, waist by 7.75 cm, BMI by 3.45 kg/m2; total testosterone, LH, glucose, insulin, HOMA-IR and triglycerides all fell; SHBG rose; FSH did not change significantly.

Those are real changes and they matter for metabolic health. They are not fertility outcomes, and the papers do not claim they are.

This is the distinction the whole subject turns on, and it has actually been tested. Within that 10-study review, two of the randomised trials compared ketogenic protocols directly against ordinary low-calorie diets — and found no significant difference in body weight between them.

So the pattern that survives is: eat less, lose weight, and glucose, insulin and androgens improve. There is no evidence that the ketosis is doing something a calorie deficit would not do. The 2023 International Evidence-based Guideline for PCOS puts it directly: there is no evidence to support any one type of diet composition over another for anthropometric, metabolic, hormonal, reproductive or psychological outcomes. The same guideline advises sustainable eating tailored to what a person will actually keep to, and notes there are benefits to a healthy lifestyle even where weight does not change.

What weight does and does not do to fertility is covered separately in weight and its role in female fertility.

Less reliably than the advice implies. A Cochrane review of 10 trials of weight-loss strategies in women with subfertility found the evidence insufficient to determine whether diet or lifestyle intervention affects live birth: pooled odds ratio 0.85, 95% confidence interval 0.65 to 1.11, across 918 women in 3 studies, rated low-quality. Clinical pregnancy was similarly uncertain.

The largest single trial is starker, and it needs reading carefully. It randomised 577 infertile women with a BMI of 29 or above either to six months of lifestyle intervention before fertility treatment, or to prompt fertility treatment. The women in the lifestyle arm did lose more weight — 4.4 kg against 1.1 kg. They had fewer healthy term births: the primary outcome occurred in 27.1% of them against 35.2% of the prompt-treatment group, a rate ratio of 0.77 (95% CI 0.60 to 0.99).

What that trial tested was a strategy, not a food. It does not show that losing weight is bad for you. It shows that spending six months on diet and exercise before starting fertility treatment left fewer women with a baby than getting on with treatment did — and for anyone over about 35, six months is not a neutral cost.

This is the part that decides the question in practice, and it is usually left out. Rice, roti, idli, dosa, poha, upma, dal with rice — these are not accompaniments to an Indian meal, they are the meal. In a dietary survey of 2,042 urban adults in Chennai, carbohydrate supplied 64% of daily energy, and refined cereals alone supplied 45.8% of energy in that cohort.

Ketosis needs carbohydrate down to roughly a tenth of energy intake. From a 64% baseline, in a household cooking one meal for everyone, that is not a substitution — it is removing the centre of the plate at every meal, indefinitely. For a predominantly vegetarian household it is harder still, because the protein and fat that would replace the cereal are narrower in range.

The Indian evidence points somewhere more achievable. In the ICMR-INDIAB national survey of 18,090 adults, the highest carbohydrate intakes carried a higher risk of newly diagnosed type 2 diabetes versus the lowest (odds ratio 1.30, 95% CI 1.14 to 1.47). But swapping refined cereals for whole wheat or millet flour without reducing the quantity was not associated with lower risk at all — odds ratio 0.94, with a confidence interval spanning no effect — while modelled substitution of some carbohydrate with plant, dairy, egg or fish protein was. Less rice, more dal, curd, egg or fish, same plate — that is a change people keep.

Nothing, in the sense that matters. No trial has put men on a ketogenic diet and measured their semen parameters, and none has measured pregnancies. A search of PubMed for ketogenic or low-carbohydrate eating together with semen, sperm or male fertility in humans returns reviews, a mouse study, and one laboratory experiment in which a ketone body helped human sperm keep moving in a dish. A result in a dish is not a result about eating.

If a man has been handed a diet sheet, the more useful reading is what weight actually does to male fertility, where at least the measurements exist.

Six things this page cannot tell you.

• That a ketogenic diet increases ovulation, conception, pregnancy or live birth in anyone. Not measured in either meta-analysis.
• That ketosis adds anything beyond eating less. The two randomised comparisons against low-calorie diets found no weight difference.
• That the hormonal changes translate into fertility. Falling testosterone and LH on a blood test is a biochemical result, and no study followed it through to a pregnancy.
• That keto is safe or unsafe in pregnancy. It has not been tested in pregnancy for a fertility indication, and this page is not the place to decide it.
• That keto does anything for men. No dietary trial in men exists.
• That any of it applies to women with PCOS at a normal body weight, or to women without PCOS. Both meta-analyses studied women with PCOS and raised body weight only.

One further caution, from outside the PCOS literature. In a US case-control study of 1,740 pregnancies affected by anencephaly or spina bifida, against a comparison group of 9,545 unaffected, women whose carbohydrate intake was in the lowest 5% had less than half the dietary folic acid intake of the rest, and higher adjusted odds of a neural-tube-defect pregnancy versus the rest (1.30, 95% CI 1.02 to 1.67). That finding drew published criticism from two authors and rests on food-frequency questionnaires, so it is not settled. It is also US data from a country that fortifies cereal grains by law, so the arithmetic does not transfer directly to India. The practical point survives the uncertainty: if you cut out cereals while trying to conceive, folate has to come from somewhere, and that means a supplement.

Diets get reached for because they are the one part of this you can start today without an appointment. That is worth naming rather than moralising about — but it is also why the sequence gets inverted, and months go by.

1. Get the diagnosis confirmed and ovulation established. The 2023 guideline notes that ovulatory dysfunction can occur even when cycles look regular, and that a blood progesterone level can confirm whether ovulation happened.
2. For anovulatory infertility in PCOS with no other infertility factor, letrozole is the guideline's first-line drug for ovulation induction — and it carries the highest evidence grade in the document, which the dietary recommendations do not. Ask about it by name; it is used off-label in many countries, so ask about that too.
3. Make the eating change you can sustain, on the guideline's own terms: any composition consistent with healthy eating, tailored to what you will keep to. Benefits exist even without weight loss.
4. Do not defer the fertility workup to do the diet first. That specific sequence is the one the LIFEstyle trial tested, and it produced fewer babies.

The fuller picture of what does and does not work in PCOS is in PCOS and fertility: how it affects conception and what works.