Eating with your menstrual cycle
The question
Do your energy needs and appetite really shift across the menstrual cycle, and should the way you eat shift with them?
What the evidence says
Cycle syncing is one of the biggest trends in nutrition content: eat one way in your follicular phase, another in your luteal phase, time your carbohydrates to your hormones. The protocols are detailed and confident, and the research behind them is thin. The cyclical shifts in metabolism and appetite are real, but they are small, and they keep shrinking as studies measure the cycle more carefully. The newest, best-controlled work is starting to erase some of them entirely.
Start with energy burn, where the claim is that the luteal phase, the two weeks after ovulation, raises your needs. A meta-analysis of 26 studies covering 318 women did find a small rise in resting metabolic rate after ovulation, but the effect shrank and stopped being statistically clear in studies published after 2000. A newer review of seven recent studies puts the rise at roughly 30 to 120 calories a day, about 3 to 5%, and warns that it overlaps normal day-to-day variability. The best-controlled recent trial, with 38 naturally cycling women, measured only about 40 extra calories a day, and even that reached significance only after four outliers were excluded. A realistic estimate of the luteal energy rise is 30 to 60 calories a day. A single extra bite of food covers it.
Appetite follows the same pattern. Older reviews put the premenstrual rise in intake anywhere from 90 to 500 calories a day, a spread that already hints at measurement trouble. A cohort of about 259 women, with phases timed by hormone tests rather than calendars, found shifts in macronutrients and food groups that were detectable but small. Then a 2025 study measured intake precisely, in the lab and in free-living conditions, and found appetite and energy intake essentially flat across the cycle. That null suggests part of the earlier effect was an artefact of how intake was measured.
Cravings deserve their own verdict, because as a reported experience they are among the most consistent findings in this literature. What the data question is their interpretation, not their existence. In a cohort of 259 women, premenstrual cravings tracked inflammation markers and mood rather than energy need, and a separate study found that women with higher mid-luteal progesterone reported fewer cravings, the opposite of a simple hormones-demand-chocolate story. Cravings track mood and inflammation more closely than they track any energy shortfall. They are worth managing as a symptom, and they do not rewrite your macros.
The exercise story is the least settled. The classic account says estrogen tilts fuel use toward fat and spares carbohydrate, with any endurance advantage sitting in the mid-luteal phase. Newer controlled trials point in conflicting directions: one crossover found peak fat oxidation identical across cycle and pill phases, another found no phase effect on fuel use after exercise, and a third, with 15 women, found more carbohydrate burned in the luteal phase after a glucose load. On performance itself, the largest analysis, spanning 78 studies, found only a trivial dip in the early follicular phase, on evidence it rated low quality. That is not a foundation for phase-specific fuelling advice for anyone who is not an elite athlete with individual data.
Two things shape how much weight to put on any of this. Hormonal contraception flattens the effects: trials found pill phases identical in both resting energy expenditure and fat oxidation, so phase-based eating advice does not apply on hormonal birth control. Calendar tracking also misplaces phases often, because cycle length varies mostly on the follicular side while the luteal phase holds near 11 to 17 days, and the research field is now moving to hormone-verified phasing, with the measured effects shrinking as it does. What holds up best is the lived-experience part: bloating, fatigue, low mood, and cravings change what you feel able to eat far more than they change what you need. A 2026 review of athlete nutrition asks in its title whether symptoms and nutrition matter more than cycle phase, and on the evidence so far they do.
Key studies
| Study | Year | Design | n | Finding |
|---|---|---|---|---|
| Benton et al. | 2020 | Systematic review + meta-analysis | 26 studies / 318 women | Small luteal rise in resting metabolic rate (ES 0.33); attenuates to ES 0.23 and loses significance in studies published after 2000. DOI: 10.1371/journal.pone.0236025 |
| Hurtová et al. | 2026 | Systematic review (2020-2025 studies) | 7 studies | Luteal resting metabolic rate roughly 30-120 kcal/day higher (~3-5%); overlaps normal day-to-day variability and measurement error. DOI: 10.3389/fphys.2026.1778735 |
| Löfberg et al. | 2024 | Crossover, indirect calorimetry | 38 cycling + 19 on the pill | Luteal resting energy expenditure only ~40-44 kcal/day higher (d ~0.20), significant only after excluding four outliers; oral-contraceptive phases flat. DOI: 10.1249/MSS.0000000000003518 |
| Smith et al. | 2025 | Lab + free-living, precise intake measurement | small | Appetite and energy intake essentially consistent across cycle phases; a recent null. DOI: 10.1016/j.appet.2025.108314 |
| Gorczyca et al. | 2016 | Prospective cohort (BioCycle) | ~259 | Small but detectable macronutrient and food-group shifts across hormone-timed phases. DOI: 10.1007/s00394-015-0931-0 |
| Agarwal et al. | 2023 | Prospective cohort (BioCycle) | 259 | Premenstrual cravings associated with inflammation markers and mood, not pure energy need. DOI: 10.1016/j.clnesp.2023.06.004 |
| Hamidovic et al. | 2023 | Observational (hormones and cravings) | — | Higher mid-luteal progesterone inversely associated with cravings. DOI: 10.3390/nu15051097 |
| Oosthuyse & Bosch | 2010 | Narrative review | — | Estrogen promotes fat oxidation and carbohydrate sparing; any endurance edge concentrated mid-luteal. DOI: 10.2165/11317090-000000000-00000 |
| Löfberg et al. | 2025 | Crossover (fat oxidation) | cycling + pill users | Peak fat oxidation during submaximal exercise consistent across cycle phases and pill phases. DOI: 10.1249/MSS.0000000000003676 |
| Govette et al. | 2026 | Within-subject crossover, carbon-13 tracer | 15 | Higher carbohydrate oxidation after a glucose load in the luteal vs follicular phase. DOI: 10.14814/phy2.70876 |
| McNulty et al. | 2020 | Systematic review + Bayesian meta-analysis | 78 studies | Only a trivial exercise-performance reduction in the early follicular phase; effects small, evidence quality low. DOI: 10.1007/s40279-020-01319-3 |
| Natalucci et al. | 2026 | Narrative review (athletes) | — | Argues symptoms and nutrition status may matter more than cycle phase itself. DOI: 10.3390/nu18071144 |
| Bisdee et al. | 1989 | Whole-body calorimetry | 8 | Sleeping metabolic rate 6.1% higher in the late luteal phase; 24-hour expenditure change not significant. Predates DOIs (PMID 2706224). |
Confidence rating and why
Emerging. The direction is consistent: a small luteal rise in resting energy expenditure and a modest premenstrual appetite and craving signal show up across decades of work, so a conservative “there is a small effect” statement holds. Everything needed to turn that into a protocol is weak. Effect sizes sit around 3 to 5% of resting metabolism and shrink as phase verification improves, the newest well-measured intake study is a null, substrate-oxidation findings conflict, samples typically run 8 to 40 women, person-to-person variability is large, and hormonal contraception erases the effects entirely. The field is mid-correction toward hormone-verified phasing, so older, larger-looking effects from calendar-only studies deserve a discount. The gap between that consistent direction and protocol-grade evidence is the space cycle-syncing content has filled.
What to do with this
Do not restructure your eating around cycle phases. The measured shifts are smaller than the error in estimating anyone’s daily calorie needs, so a plan that works for you in one phase works in all of them.
If the premenstrual week leaves you hungrier, give yourself honest headroom and keep it in proportion: the physiology accounts for tens of extra calories a day, not hundreds. Treat cravings as a symptom to handle kindly. Planning for the chocolate beats pretending it will not happen.
Keep protein steady across the whole cycle. No controlled study supports raising or lowering it by phase.
The two nutrients genuinely tied to menstruation are daily-intake questions, not timing ones. Iron matters if your periods are heavy, and a blood test settles whether you need more. Calcium has some trial support for easing PMS symptoms at adequate daily intakes.
If you use hormonal contraception, skip phase-based eating advice entirely. And if you do track your phase, treat the label as an estimate: calendar apps misplace ovulation often, and the steadiest anchor is counting back about 14 days from your predicted next period to find the start of the luteal phase.
What we do NOT claim
- A per-day calorie prescription by cycle phase. The measured luteal rise of 30 to 60 calories a day is smaller than the error in estimating anyone’s needs.
- That protein or macro targets should change across the cycle. No controlled evidence supports it.
- That changing what or when you eat treats PMS. Symptom-aware eating helps you cope with the harder days; it does not treat them.
- That calendar-based tracking can tell you your phase precisely. It misplaces ovulation often.
- Any of this for people on hormonal contraception. The pill flattens the cyclical effects these claims rest on.
- That cravings signal a nutritional requirement. The evidence ties them to mood and inflammation more than to energy need.
- Anything clinical. Menstrual disorders, amenorrhea, PCOS, and fertility need medical care, not meal timing.