The peer-reviewed literature does not support the confident phase-based training prescriptions many apps market. Group-mean cycle effects on endurance performance are trivial (McNulty 2020, 78 studies, 1,193 participants), and most primary studies used inadequate phase verification. What IS settled: luteal HRV is genuinely suppressed by progesterone (relevant to any readiness score), iron deficiency hits 15 to 35 percent of female endurance athletes (up to 45 percent in trained subgroups), and REDs prevalence in trail-running populations exceeds 40 percent. If you are training for a mountain, the honest priorities are: fuel enough, test ferritin annually, and use your cycle as a health signal rather than an intensity dial.
Why this article exists
A companion piece on this site (Mountaineering Training for Women) covers sex-specific physiology across five areas: fatigue resistance, iron, altitude response, RED-S, and injury. It is deliberately calm about menstrual-cycle prescription because the underlying evidence base was not (and still is not) strong enough to justify the confident claims elsewhere in the industry.
This piece goes underneath that. What does the peer-reviewed literature actually show about cycle phase and endurance training? What is genuinely settled? What is contested or under-powered? And where does TTM's approach fit inside those honest limits?
The settled science, one row per finding
Read the table below as a research map. Everything cited to a URL.
| Finding | Status | Reference |
|---|---|---|
| Group-mean effect of cycle phase on endurance performance is trivial | Settled | McNulty 2020 meta-analysis, 78 studies, 1,193 participants, SMD -0.06 to -0.08 (Sports Medicine) |
| Most primary studies use inadequate phase verification | Settled | Elliott-Sale et al 2021 methodological standards (Sports Medicine) |
| Luteal HRV suppression is real and progesterone-driven | Settled | Schmalenberger 2020 (J Clinical Medicine); replicated at wearable scale in Clinical Autonomic Research 2023 |
| Iron deficiency prevalence 15 to 35 percent, up to 45 percent in trained women | Settled | Endurance iron literature summary (ScienceDirect) |
| REDs prevalent in endurance sport; ~43 percent at risk in trail running | Settled | Kuikman 2024 (PMC); IOC 2023 REDs consensus (BJSM) |
| Perimenopause accelerates VO2max decline beyond age effect | Settled | 2025 scoping review, women masters athletes (Sports Medicine Open) |
| Individual phase-based prescription improves endurance training outcomes | Contested | Evidence thin; no adequately powered trial demonstrates group-mean gain |
| Phase-adjusting HRV readiness gate improves training decisions | Contested | Mechanism is settled; the applied algorithmic fix has not been tested |
| ACL and connective-tissue injury risk in mountain sports across cycle | Contested | Team-sport data (Herzberg 2017) does not clearly transfer to trail or alpine |
| REDs prevalence in alpinism specifically | Contested | No peer-reviewed alpinism data; trail-running data is best proxy |
The McNulty result and why it matters
McNulty and colleagues' 2020 systematic review and meta-analysis in Sports Medicine pooled 78 studies covering 1,193 eumenorrhoeic participants. Performance was trivially reduced during the early follicular phase versus all other phases, standardised mean difference −0.06 to −0.08, with wide confidence intervals crossing zero and substantial between-study heterogeneity. The authors' own conclusion: general phase-based guidelines cannot be formed, and an individualised approach is recommended (McNulty 2020 full text).
This is a big deal because most confident phase-based prescriptions in the app and coaching industry cite selected small studies that showed larger effects, without accounting for how those effects wash out when you pool the literature honestly. Elliott-Sale et al 2021 in Sports Medicine went further and highlighted that most primary studies use inadequate phase verification (no hormone confirmation, no ovulation test), so a share of the reported phase signal is likely misclassification noise (Elliott-Sale 2021). A 2025 methodological audit of elite-athlete performance studies replicated this finding: only a small minority meet the recommended standard for phase determination (Tandfonline 2025 audit).
Individual variance is real. If you have a personal pattern (worse sessions in the late luteal phase, for example) that is consistent over several cycles, honour it. But the peer-reviewed evidence does not support a fixed rule that everyone should reduce intensity in one phase or push it in another. Individual pattern beats one-size-fits-all phase prescription.
Luteal HRV suppression is the one clean signal
This one is worth understanding in detail because most amateur mountaineers now use a wearable readiness score, and most of those scores rest on a rolling HRV baseline. Schmalenberger and colleagues' 2020 within-person study in the Journal of Clinical Medicine showed that vagally-mediated HRV (RMSSD, HF-HRV) declines significantly from follicular to luteal phase, and the decline tracks progesterone rather than estradiol (Schmalenberger 2020). A 2022 Physiological Reports study by Ramesh confirmed the pattern and extended it across menopausal status (Ramesh 2022). A 2023 study using continuous wearable data replicated luteal HRV suppression at population scale (Clinical Autonomic Research 2023).
What does that mean in practice for a readiness score? If your HRV gate compares today against a 14-day rolling baseline (a common setup), for many cycling women it will systematically read "suppressed" in the mid- to late-luteal phase, without any real training-stress signal driving the change. The physiology is doing exactly what progesterone tells it to do; the baseline window is too short to see it as normal.
The correct fix is a longer baseline window (28 to 35 days, capturing a full cycle) or explicit phase-aware normalisation. Neither has a published trial showing improved training outcomes yet, but the mechanism is well-enough understood that a longer baseline is worth using even in the absence of trial data. If you use a wearable with a short readiness baseline and notice a luteal-phase pattern of low reads, take it as noise you can now name, not a training red flag.
Iron: the single highest-return blood test
Iron deficiency in female endurance athletes is one of the most under-recognised limiters in the whole space. Prevalence in the literature is commonly reported at 15 to 35 percent, and up to about 45 percent in highly trained subgroups (Sim 2019 discussion, ScienceDirect). Two things drive it: menstrual blood loss (a physiological monthly draw on iron stores) and exercise-induced hepcidin elevation, which acutely suppresses iron absorption after hard sessions.
Cycle phase also modulates iron absorption. Alfaro-Magallanes and colleagues found that menses is associated with lower hepcidin (a wider window for iron absorption from food or supplements), while the luteal phase raises resting hepcidin (Alfaro-Magallanes 2022). The common threshold in sports medicine practice is ferritin below 30 micrograms per litre as a clinical treatment threshold, with 40 to 50 as a performance-relevant floor (ISSN 2023 Position Stand on the Female Athlete).
If you are a woman training seriously for a mountaineering objective, an annual ferritin check is the single highest-return blood test on the whole list. Iron deficiency does not present as "I feel tired". It presents as "I am fit but my sessions feel harder than the plan says they should". If that is you, test it before you conclude the plan is wrong.
RED-S in mountain sports: the alarming adjacent data
The 2023 IOC REDs consensus statement (Mountjoy et al, British Journal of Sports Medicine) is the authoritative document. It expanded the model beyond the female athlete triad to include effects on carbohydrate availability, overtraining overlap, males, and Para athletes (IOC 2023 REDs consensus).
Mountain-sport-specific data are thin but growing. Kuikman and colleagues 2024 found approximately 43 percent of trail runners at risk for low energy availability using the LEAF-Q screening tool, alongside high rates of disordered eating and exercise dependence (Kuikman 2024). A 2026 screening study in female trail runners reported over 50 percent at risk (2026 female trail runners). A 2024 study of cross-country skiers found LEA prevalence varies substantially by phase of the annual training cycle (XC skiing 2024).
Direct alpinism data does not exist yet. Extrapolation from trail running is the current best proxy. If nearly half of female trail runners are at risk of a physiological cascade that suppresses endurance, iron, bone density, thyroid function and eventually performance itself, the honest expectation is that the mountaineering population is not immune. The defence is not restrictive dieting. It is fueling training rather than under-eating it: daily energy availability at or above 40 kcal per kg of fat-free mass is the commonly-cited floor.
Perimenopause: what the record actually shows
A large share of amateur first-time mountaineers are in the 40 to 55 year cohort, which sits directly on the perimenopause transition. A 2025 scoping review in Sports Medicine Open on determinants of endurance performance in women masters athletes summarised age-related decline in VO2max, lactate threshold, and running economy, and specifically highlighted that the perimenopause transition accelerates VO2max decline beyond the linear age effect, mediated by estrogen decline (Sports Medicine Open 2025). Cross-sectional masters running data show VO2max drops of roughly 5 to 7 percent per decade in trained women, with the largest single-decade decrement often falling in the perimenopause window.
Structured endurance training in women aged 40 to 60 still elicits improvements in systemic and muscle oxygen utilisation over multi-year timeframes (Bell 2016). Practical implications commonly discussed in the literature: greater need for higher-intensity work to preserve VO2max, higher protein intake (roughly 1.6 to 2.0 g per kg per day), and attention to bone density and iron. These are informed recommendations rather than randomised-trial-proven prescriptions. If you are in this cohort and training seriously, the plan does not become impossible; it becomes less forgiving.
The cycle-tracking app landscape
Descriptive only, no endorsement. FitrWoman (Orreco) provides phase-based nutrition and training tips and is used by professional teams. Its scientific basis rests on peer-reviewed work by Bruinvels and colleagues; no independent validation study of the app itself as a training-decision tool exists in the peer-reviewed literature we could locate. Wild.AI offers a readiness score and integrates with Garmin, Oura, Strava, and Apple Health; marketing cites internal scientific advisory but no peer-reviewed independent validation. Garmin cycle tracking is a Connect app feature without peer-reviewed validation of its algorithm as a training tool.
None of this means these apps are useless. It means the confident, phase-specific training prescriptions they market are running well ahead of the underlying evidence. Treat them as tracking and awareness tools, not as a training-plan authority.
What TTM does today (honest)
The reason this article exists in this form is that we would rather be accurate than aspirational. So here is what the plan does and does not do today.
The training-load model uses sex-specific TRIMP coefficients. TRIMP (training impulse) is the standard endurance-training-load metric that combines duration and heart-rate response into a single load number. Version 1.1 of the TTM algorithm applies sex-specific coefficients so that a given duration and heart-rate response translates into an appropriate load for a female athlete rather than defaulting to a male reference. That is real and shipped.
The plan does not vary session prescription by menstrual cycle phase. Sessions are not moved, intensity is not shifted, and volume is not tuned based on where you are in your cycle. The reason: the evidence base described above does not support confident phase-based prescription for endurance training. Shipping something we cannot back would be worse than shipping nothing.
The fuelling calculator does account for luteal-phase fluid and sodium shifts. That is a different physiological question with cleaner evidence: plasma volume falls modestly in the luteal phase and core temperature runs slightly higher, so hydration and sodium needs rise. The calculator adjusts targets when you note the luteal phase. This is separate from training-load prescription.
The readiness score reads HRV from your wearable and monitors trends. It does not yet apply phase-aware normalisation to the HRV baseline. If you cycle regularly and use the readiness score, the caveat above applies: expect systematically lower reads in the mid- to late-luteal phase, especially against a 14-day baseline.
Cycle-aware readiness and prescription
Phase-aware HRV normalisation (longer baseline window, plus explicit cycle-tracking input where the athlete provides it) is on the roadmap. So is optional per-individual load adjustment where the athlete has logged a consistent pattern across multiple cycles.
We are not shipping either yet because the evidence for group-mean phase-based prescription is thin, and we would rather build cycle-aware features on a base of individual pattern-matching than on a fixed rule the peer-reviewed record does not support. When we ship, we will say so plainly and cite the evidence.
If you take one thing from this article
Fuel enough. Test ferritin annually. Use your cycle as a health signal, not an intensity dial. Do not let a confident-sounding app convince you to skip hard sessions on a schedule the evidence does not support, and do not let a slightly low HRV read in the luteal phase talk you out of a session your body is fine to do.
Individual pattern beats population-level phase prescription every time. If you notice a consistent personal pattern across three or four cycles, respect it. If you do not, train the plan.
Same plan structure, honest inputs
The training plan for a female mountaineer is not a smaller version of the male one, and it is not a fundamentally different one. It is the same plan, with sex-specific training-load calibration, fuelling adjusted for the physiology, and honest inputs on iron, energy availability, and cycle health. Read the pillar guide on training for women for the five-area breakdown, or test your readiness with the Summit Readiness Calculator.
The confident industry story about phase-based training is not what the peer-reviewed record says. The honest story is duller and more useful. Fuel enough. Watch iron. Respect your own pattern. Train the plan.
Common questions
Should women train differently in different phases of the menstrual cycle?
The largest meta-analysis to date (McNulty et al 2020, Sports Medicine, 78 studies, 1,193 participants) found the group-mean effect of cycle phase on endurance performance is trivial, with wide individual variance. Follow-up methodological reviews have flagged that many primary studies used inadequate phase verification, so signals attributed to phase may partly be misclassification noise. Practical read: a stable regular cycle is a health signal to monitor, not an intensity dial the plan should be tuned to. Individual pattern matters more than a fixed phase-based prescription. If your cycle is disrupted or missing, that is a warning sign about energy availability worth addressing directly.
Does HRV change across the menstrual cycle?
Yes, reliably. Vagally-mediated HRV (RMSSD, HF-HRV) declines significantly from follicular to luteal phase, tracked to progesterone rather than estradiol. Schmalenberger et al 2020 established the pattern, and larger continuous-wearable studies have replicated it at population scale. Practical implication: an HRV readiness score that compares today against a short 14-day rolling baseline will systematically read low in the mid- to late-luteal phase, without any real training stress driving it. A longer baseline window (28 to 35 days) or phase-aware normalisation is the correct fix. This is a real physiological signal, not a training red flag.
Is iron deficiency more common in female endurance athletes?
Yes. Prevalence of iron deficiency in female endurance athletes is commonly reported at 15 to 35 percent, and up to about 45 percent in highly trained subgroups. Menstrual blood loss is one driver; exercise-induced hepcidin elevation (which suppresses iron absorption) is another. Ferritin below 30 micrograms per litre is a clinical treatment threshold in sports medicine, with 40 to 50 micrograms per litre commonly used as a performance-relevant floor. For a woman training seriously for a mountaineering objective, an annual ferritin check is one of the highest-return blood tests available.
What is RED-S and how common is it in mountain sports?
Relative Energy Deficiency in Sport (REDs, per the 2023 IOC consensus) is the physiological cascade that follows chronically eating fewer calories than training and basic function require. Prevalence data specific to alpinism does not exist, but adjacent mountain-sport data are alarming: a 2024 screening study of trail runners found roughly 43 percent at risk of low energy availability, and a 2026 study of female trail runners specifically reported over 50 percent at risk. Symptoms include cycle disruption or amenorrhea, iron and thyroid drops, bone stress injuries, immune suppression, and eventually endurance performance decline. Prevention is fueling training rather than under-eating it.
Does TTM's training plan adjust for menstrual cycle phase?
Not today. TTM's training-load model uses sex-specific TRIMP coefficients (v1.1) that account for the average physiological differences in training-load response between men and women, but the plan does not vary session prescription by menstrual cycle phase. Cycle-phase-aware readiness and prescription is a roadmap item, not a shipped feature. The practical reason we have not shipped it yet: the peer-reviewed evidence base for phase-based training prescription is thinner than the tracking-app industry suggests, and we do not want to ship something we cannot back. TTM's fueling calculator does account for luteal-phase fluid and sodium shifts, which is a separate physiological question with cleaner evidence.