Is intermittent fasting different for women?

What trials in women show. A review of human trials found intermittent fasting lowered androgens (testosterone and the free androgen index) and raised sex hormone-binding globulin in premenopausal women with obesity, especially when eating ended before 4 p.m. Estrogen, gonadotropins and prolactin did not change¹. A 2026 review of six time-restricted eating (TRE) studies found randomized trials in women with obesity saw no significant change in testosterone, estrogen, progesterone, luteinizing hormone, cortisol or DHEA². In PCOS, three small TRE studies reported more regular cycles in 33% to 40% of women and lower testosterone³. A meta-analysis of 13 randomized trials in 612 women with overweight or obesity found TRE lowered weight and fasting insulin without lean mass loss, but did no better than ordinary calorie restriction⁵.

When fasting becomes too little. The brain's control of the menstrual cycle reacts to energy availability, meaning calories eaten minus calories burned in exercise. In 29 regularly cycling women, hormone pulses stayed normal at 30 kcal per kg of lean mass per day but were disrupted below that⁴. Pairing long fasts with hard training and low intake is where trouble starts.

Blood sugar. In a three-week study of alternate-day fasting in 8 men and 8 women, women's glucose response to a meal was slightly worse afterward, while men's insulin response improved⁶. It is one small study, but it is a reason for women to watch how they respond.

Cycles, pregnancy and breastfeeding. Ramadan fasting (daytime only) appears to cause mild, temporary menstrual irregularities without major hormone changes, and pregnant and breastfeeding women are advised to consult a physician because the evidence on safety is inconclusive⁷. Birth weight was unaffected across 22 Ramadan studies, but data on other outcomes were insufficient⁸.

Thyroid. In healthy adults, Ramadan fasting keeps thyroid function in the normal range. In people on levothyroxine, TSH rose afterward in several studies, though it is unclear whether fasting, medication timing or sleep explains it⁹. See Thyroid health: the master gland.

Cycle syncing. We found no randomized trial of fasting schedules matched to menstrual-cycle phases. Even for exercise, a meta-analysis found only trivial, low-quality differences between phases and concluded that general phase-based guidelines cannot be formed¹⁰.

Related reading: Who should avoid intermittent fasting, PCOS: natural management, and, for the large trials of daily eating windows (TREAT, the NEJM trial) and the CALERIE calorie-restriction trial, Does limiting your eating window help?.

Reviews of hormones. The review of human trials concluded that intermittent fasting reduced androgen markers and raised sex hormone-binding globulin in premenopausal women with obesity, more so when eating ended by 4 p.m.; estrogen, gonadotropins and prolactin were unchanged. In lean, active young men, testosterone fell without loss of muscle mass or strength. The authors stress that very few studies exist¹. The 2026 systematic review of TRE and female hormones found six studies. Randomized trials in women with a body mass index above 30 consistently showed no significant changes in testosterone, estrogen, progesterone, luteinizing hormone, cortisol, DHEA or sex hormone-binding globulin. Non-randomized PCOS studies reported lower testosterone and luteinizing hormone, possibly confounded by weight loss. The reviewers called TRE hormonally safe in women with general obesity but the evidence inconclusive overall². The PCOS review of three studies reported menstrual regularity in 33% to 40% of women, a 9% fall in testosterone and a 26% fall in the free androgen index, and called for larger trials³.

Weight and body composition in women. Thirteen randomized trials (612 participants, at least 85% women) found TRE reduced body weight by 1.93 kg (95% CI 0.17 to 3.69) and fasting insulin by 2.12 μU/mL (95% CI 0.07 to 4.17), with no significant effect on BMI, fat mass, fat-free mass, lipids, glucose or blood pressure. Against calorie restriction alone there was no extra weight loss (P = 0.295). Four studies reported minor, self-resolving side effects such as hunger and headache⁵.

Energy availability. In a randomized trial of 29 regularly menstruating, sedentary young women, energy availability was set for 5 days at 10, 20 or 30 kcal per kg of lean body mass per day (with 15 kcal per kg of exercise), against a 45 control. Luteinizing hormone pulse frequency fell, and amplitude rose, below the 30 threshold, and the disruption was greater in women with short luteal phases⁴. This tested energy deficit plus exercise rather than fasting windows, but it shows where the cycle starts to react.

Glucose tolerance. In eight non-obese men and eight non-obese women, 22 days of alternate-day fasting slightly impaired the glucose response to a meal in women (P < 0.01) with unchanged insulin response, while men had no change in glucose response and a lower insulin response (P < 0.03)⁶.

Pregnancy, breastfeeding and cycles. In 22 studies covering 31,374 pregnancies (18,920 exposed to Ramadan fasting), birth weight was not affected (standardized difference 0.03, 95% CI 0.00 to 0.05) and preterm delivery was not changed (odds ratio 0.99, 95% CI 0.72 to 1.37, from 5,600 pregnancies), but placental weight was lower in fasting mothers and no data existed on perinatal mortality⁸. A 2026 narrative review found Ramadan fasting caused mild, temporary menstrual irregularities with stable fertility hormones, inconclusive and contradictory safety evidence in pregnancy, and some small long-term associations after first-trimester fasting⁷.

Thyroid. Pooled evidence shows Ramadan fasting preserves hypothalamic-pituitary-thyroid function in healthy adults. In hypothyroid patients on levothyroxine, post-Ramadan TSH rose in several studies while free T4 stayed stable, and no single dosing schedule consistently performed best. Evidence in hyperthyroidism is limited and expert-opinion based⁹.

Eating-disorder and metabolic cautions. In 491 Chinese adults followed for 8 months, current or past intermittent fasting was associated with higher eating-disorder psychopathology and eating-related impairment at follow-up¹¹. In type 2 diabetes treated with insulin or sulfonylureas, intermittent fasting carries an increased hypoglycemia risk¹².

Conflicts and gaps. Most trials are short and in women with obesity; we found little in lean, older or postmenopausal women and none testing cycle-phase-timed fasting. Ramadan data (daytime fasting, no fluids) differ from other patterns. Observational links to eating-disorder symptoms cannot show cause.

Bottom line (confidence 2 of 5). In women with obesity, intermittent fasting seems hormonally neutral to mildly favorable, and early eating windows may modestly lower androgens. For lean women, athletes, people planning pregnancy, and anyone with a thyroid, diabetes or eating-disorder history, the data are too thin to be casual; use medical guidance and stop if your cycle changes.