Tip · Fasting

Fasting for women – cycle-adjusted

Q: Should women fast differently from men?

There is no controlled study that has tested this directly. What there is, is a laboratory threshold: if energy availability drops below 30 calories per kilogram of fat-free mass per day, the release of luteinizing hormone is disrupted in women. This depends on the amount of energy, not on the length of the eating break – a time window is only a problem if it leads to eating too little.

Q: Do I have to fast for shorter periods in the luteal phase?

There is no study for this rule. The usual rationale, that carbohydrates are utilized less well in the second half of the cycle, is not supported by measured data: in a tracer study with 15 women, carbohydrate oxidation was higher in the luteal phase than in the follicular phase, and the area under the glucose curve did not differ. Only the peak value was higher.

Q: Does intermittent fasting disrupt my hormones?

In the longest study to date, over twelve months with an 8-hour window in 90 adults with obesity, testosterone, DHEA and SHBG remained unchanged, and in postmenopausal women so did estradiol, estrone and progesterone. An analysis with an even narrower window over eight weeks found a drop in DHEA of 13 to 14 percent, with very large scatter. The systematic review of all six available studies calls the evidence inconclusive overall.

Q: I have PCOS – does fasting make sense then?

The largest randomized study to date, with 76 women over six months, found a six-hour eating window just as effective as a 25 percent calorie reduction: a good four percent weight loss compared with control in both groups, with no difference between them and without serious adverse events. For the hormone values themselves, the data in PCOS are inconsistent and come mostly from non-randomized studies.

Q: When do I have to stop?

If menstruation stops or becomes irregular. This is the clinical counterpart to what was measured in the laboratory as disrupted LH pulsatility below the energy threshold. Young women in the first years after their first menstrual period should be particularly attentive – in this group, the axis reacted considerably more sensitively than in adults.

Q: Who should not fast at all?

Pregnant and breastfeeding women, children and adolescents, underweight people and people with a history of eating disorders. In addition, women with existing functional hypothalamic amenorrhea and women who already eat sparingly while training at high volume – precisely this combination of little energy and a lot of exercise is what disrupted cycle regulation in the experiments.

A precautionary rule, not proof of efficacy: there are no controlled studies on cycle-adjusted fasting.

Cycle-adjusted fasting is one of the best-known rules in the scene: a long window in the first half of the cycle, a short one in the second. This tip is here because there is not a single controlled study on it – and because there is instead a measured number that is far more useful.

In short

The established rule depends on the amount of energy, not on the length of the window: if energy availability drops below 30 kcal per kg of fat-free mass per day, the pulsatility of luteinizing hormone is disrupted in regularly menstruating women – at 30 kcal/kg nothing happens, below it, it does. On cycle-adjusted fasting itself there is no study. The reassuring side is now well supported: in the longest study to date, over twelve months with an 8-hour window, testosterone, DHEA and SHBG remained unchanged, and in the largest study in women with PCOS, a six-hour window over six months was just as effective as a 25 percent calorie reduction, without serious adverse events. The clinical warning sign remains: absent or irregular menstruation is a reason to stop.

What lies behind it

In women, the hypothalamic-pituitary-gonadal axis does not respond to body weight but to energy availability – that is, to the dietary energy that remains after subtracting the energy expended in exercise. In the central study, 29 regularly menstruating, sedentary young women with normal body composition were set to defined values for five days in the early follicular phase: 45 kcal per kg of fat-free mass per day as the reference, and in a second run 10, 20 or 30. All exercised daily, expending 15 kcal/kg of fat-free mass at 70 % of maximal oxygen uptake.

The result is a threshold, not a straight line: at 30 kcal/kg of fat-free mass per day, LH pulsatility was unchanged (p > 0.3); below that, pulse frequency fell and amplitude rose (each p < 0.04); in women with a short luteal phase, the disruption was more pronounced (p < 0.01). This leads to the point for fasting: it is not the length of the window that decides, but whether enough energy is left at the end of the day. A 16-hour window with full daily energy does not fall below the threshold; a 13-hour window with a marked deficit and a lot of exercise can.

What happens to the hormones

Over twelve months, 90 adults with obesity were randomized to an 8-hour window from 12 noon to 8 p.m. without calorie counting, to a 25 percent calorie reduction, or to control. Weight fell in both intervention groups compared with control (p < 0.01), in men, premenopausal and postmenopausal women. Nothing happened to the hormones: total testosterone, DHEA and SHBG changed neither over time nor between groups; estradiol, estrone and progesterone remained unchanged in postmenopausal women.

An even narrower window over eight weeks shows the same picture with one exception. In 12 premenopausal and 11 postmenopausal women with a 4- to 6-hour eating window, testosterone, androstenedione and SHBG remained unchanged; only DHEA fell – premenopausally by 14 % ± 32 %, postmenopausally by 13 % ± 34 % (main effect of time p = 0.03).

The luteal phase rationale

The rule “fast for shorter periods in the second half of the cycle” is usually justified by the claim that carbohydrates are utilized less well then. In 15 eumenorrheic women, a within-person crossover measured the opposite: one metabolic test each in the follicular phase (day 7 ± 2) and in the luteal phase (day 22 ± 3) after an overnight fast with 75 g of glucose plus tracer. In the luteal phase, whole-body carbohydrate oxidation was higher (p = 0.03), as was oxidation of the ingested carbohydrates (p = 0.002). Only the peak glucose concentration over 3 hours was higher (p = 0.04), with no differences in the 3-hour mean or in the area under the curve. So there is a cycle difference – but it does not support the rule derived from it.

What is well supported

The most robust figure in this field is the energy availability threshold of 30 kcal per kg of fat-free mass per day, below which LH pulse frequency falls and amplitude rises. Added to this is the age dependence: in a controlled experiment with 9 adolescents of 5 to 8 years gynecological age and 10 adults of 14 to 18 years, an energy availability of 10 kcal/kg of fat-free mass over five days lowered LH pulse frequency in the adolescents (p < 0.01), not in the adults (p = 0.39). On the reassuring side is the PCOS study: 76 participants, six months, eating window 1 p.m. to 7 p.m. without calorie counting versus 25 percent calorie reduction versus control. Weight change after 6 months: time window −4.32 % (95% CI −6.20 to −2.44; p < 0.01), calorie reduction −4.66 % (95% CI −7.13 to −2.19; p < 0.01), difference between the two 0.34 % (p = 0.79), no serious adverse events.

What the studies show

The threshold study on energy availability

29 regularly menstruating, habitually sedentary young women with normal body composition. Over 5 days in the early follicular phase, energy availability was controlled: 15 kcal per kg of fat-free mass of supervised exercise daily, plus a clinical nutrition product that set energy availability to 45 and, in two randomized runs at least two months apart, to 10, 20 or 30 kcal per kg of fat-free mass per day. What the study did not examine: fasting, eating windows, cycle phases beyond the early follicular phase, and periods longer than five days.

Twelve months of time-restricted eating and sex hormones

90 adults with obesity, randomized to an 8-hour window from 12 noon to 8 p.m. without calorie counting, a daily calorie reduction of 25 %, or control. Weight fell in both intervention groups compared with control. Total testosterone, DHEA and SHBG remained unchanged, as did estradiol, estrone and progesterone in postmenopausal women. This is the strongest finding against concerns about hormonal disruption – albeit in women with obesity and with a window that only began at 12 noon.

A six-hour eating window in PCOS over half a year

76 women with polycystic ovary syndrome, randomized to a 6-hour time window from 1 p.m. to 7 p.m. without calorie counting, a 25 percent calorie reduction, or control without dietary change. The primary endpoint, the percentage change in weight after six months, was met, with no difference between the two intervention groups. A six-hour window is considerably narrower than any of the circulating cycle rules, and it was tolerated by these women over half a year.

The systematic review: six studies in total

A systematic search of the literature on time-restricted eating and female reproductive hormones found six studies. Randomized studies in women with general obesity consistently found no significant changes in androgens or other key hormones – testosterone, estrogen, progesterone, luteinizing hormone, cortisol, DHEA and SHBG. Non-randomized studies in women with PCOS reported favorable hormonal changes that could be confounded by weight loss. Conclusion: the evidence is inconclusive.

Where the data stop

There is not a single controlled study on cycle-adjusted fasting. No study has compared a longer fasting window in the follicular phase with a shorter one in the luteal phase – neither for weight nor for hormones, cycle regularity or well-being. The threshold study did not examine fasting either: it manipulated energy availability over five days in exercising women using a liquid study product. Whether an eating window with the same daily energy has the same effect on LH pulsatility has never been tested. The data base as a whole is small: six studies, almost all of the hormone analyses secondary analyses of weight-loss studies in women with obesity. The drop in DHEA is the only consistently measured hormonal effect and its significance is unclear – measured after eight weeks with a very narrow window, but not in the twelve-month study, and with very large scatter. And for normal-weight women who practice 16:8 long term, there are no hormone data at all.

How to do it

What was actually studied: an energy availability of 30 kcal per kg of fat-free mass per day as a threshold, measured over 5 days in the early follicular phase with a daily exercise expenditure of 15 kcal/kg of fat-free mass. An 8-hour window from 12 noon to 8 p.m. without calorie counting over 12 months – weight fell, all measured sex hormones unchanged. A 6-hour window from 1 p.m. to 7 p.m. over 6 months in PCOS – equivalent to a 25 percent calorie reduction, no serious adverse events. And a 4- to 6-hour window over 8 weeks – testosterone, androstenedione and SHBG unchanged, DHEA down by 13 to 14 %. On the cycle rule itself – a long window on days 1 to 14, at most 13:11 with more protein and more carbohydrates in the evening from day 15 – there is no study, and the usual rationale via carbohydrate utilization is refuted by measured data. The rule that can be derived is a different one: more important than the length of the window is that the daily energy fits into it. Anyone who exercises should plan for the exercise energy in addition, because the threshold refers to the energy remaining after subtracting exercise expenditure.

Safety

The established core of the caution is energy availability, not the time window: below 30 kcal per kg of fat-free mass per day, LH pulsatility is disrupted, and this risk arises from the combination of eating too little and exercising a lot. A fasting window is only a problem if the daily energy can no longer be accommodated as a result. Young women in the first years after menarche are particularly sensitive: at a gynecological age of 5 to 8 years, an energy availability of 10 kcal/kg of fat-free mass over five days significantly lowered LH pulse frequency, at a gynecological age of 14 to 18 years it did not. The clinical warning sign is clear: absent or irregular menstruation is a reason to stop, not a side issue. Fasting is not suitable for pregnant or breastfeeding women, children and adolescents, underweight people and people with a history of eating disorders; for women, known functional hypothalamic amenorrhea, a wish to conceive with an irregular cycle, and high training volumes with tight nutrition are added to this list. What the studies provide in terms of safety data, on the other hand, is rather reassuring: no serious adverse events in six months with a six-hour window, unchanged sex hormones over twelve months. Consistent with tolerability is a finding from a crossover study with 23 young women: total mood disturbance was more pronounced after alternate-day fasting than with time-restricted eating (p = 0.03), but disappeared by the second fasting day.

BK-Score Long used, barely studied

Human evidence		2
Mechanism		4
Safety data		5
Hype gap		3
Track record of use		6

There are no controlled studies on cycle-adjusted fasting. What there are, are indications that severe energy deficits can disrupt the hypothalamic-pituitary axis – the precautionary rule is derived from this, not from proof of efficacy.

The score rates the state of knowledge, not the effect. “Safety data 9” means well studied – not harmless.
Subjective assessment by Biohacking Kompakt based on published scoring rules – not a scientific rating and not a medical recommendation. Rules and all ratings (German)

Frequently asked questions about fasting for women – cycle-adjusted

Should women fast differently from men?