Why Your HRV Drops Before Your Period: A Cycle-Aware Train/Modify/Rest Framework for Oura, WHOOP, and Garmin

If you want the practical answer first: HRV often drops in the late luteal phase for physiological reasons, not because your training suddenly stopped working. The right move is not to ignore the signal or panic over it. The right move is to combine cycle context with other recovery markers and decide: Train, Modify, or Rest.

That is exactly how SensAI handles this pattern across Oura, WHOOP, and Garmin: trend-first, cycle-aware, and explicit about uncertainty.

Why HRV Often Drops in the Late Luteal Phase (Progesterone, Autonomic Shift, and Temperature)

The late luteal phase often pushes autonomic balance toward lower vagal activity, which can show up as lower RMSSD/HRV in wearables.¹² In short: a lower HRV before your period can be expected physiology.

Schmalenberger and colleagues summarized prior evidence showing vagally mediated HRV tends to decrease from follicular to luteal phase with a moderate effect size (d = -0.39).¹ In their own within-person datasets (n=40 and n=50), they found that higher-than-usual progesterone predicted lower-than-usual HRV in the same person.¹

As Kathleen M. Schmalenberger et al. wrote: “Higher-than-usual P4 significantly predicted lower-than-usual HRV within a given participant.”¹

Thermoregulation shifts matter too. Core temperature is typically 0.3°C to 0.7°C higher post-ovulation in the luteal phase, which can increase cardiovascular strain for a given training session.³ During exercise in heat, luteal-phase thermoregulatory burden is also higher; one meta-analysis (9 papers, n=83) found higher initial and post-exercise core temperatures in luteal vs follicular phases.⁴

The coaching implication from SensAI: a late-luteal HRV dip is often real physiology, but it still needs multi-signal context before you change training.

What Your Wearable Is Actually Seeing (HRV, Resting HR, Temperature, Respiratory Rate, Sleep, and Load)

Wearables are not reading “readiness” directly. They infer it from a signal bundle.

For premenstrual decision-making, these are the most useful features:

HRV (often RMSSD-based): commonly trends downward near cycle end in naturally cycling users.²
Resting heart rate: often trends upward as HRV trends downward late luteal.²
Temperature trend: often elevated post-ovulation and can amplify perceived strain.³
Respiratory rate: usually steadier than HRV, but useful as a strain/illness cross-check.
Sleep continuity/efficiency: poor sleep can compound luteal autonomic stress.
Recent training load: high strain plus low recovery signals should influence session choice.

A large free-living wearable dataset (11,590 participants, 1,241,929 days, 45,811 cycles) showed clear cyclical offsets in both RHR and RMSSD.² In that analysis, RHR reached about -1.83 bpm near day 5 and +1.64 bpm near day 26, while RMSSD showed the inverse (+3.57 ms near day 5 and -3.22 ms near day 27).²

So if your app flags low HRV and slightly higher RHR right before bleeding starts, it may be reflecting expected cycle dynamics rather than poor training adaptation.

Device-Specific Interpretation Differences (Oura Readiness vs WHOOP Recovery vs Garmin HRV Status)

The same physiology can look different across devices because each platform uses different baselines, weighting, and output labels.

Oura Readiness + cycle-aware adjustments

Oura has explicitly updated Readiness logic to better account for menstrual cycle physiology.⁵ Oura reports a major reduction in disproportionately low luteal Readiness outputs after these changes: 81% decrease in disproportionate low-luteal penalties, +4 to +5 average Readiness points for affected users, and 35% of cycling users no longer seeing disproportionate luteal penalty.⁵

As Neta Gotlieb, PhD (Oura), explained: “Our goal in making these changes is not to ignore the natural physiological differences women experience across their cycles… Rather, we aim to ensure your Oura Scores more accurately reflect your health.”⁵

WHOOP Recovery + menstrual insights

WHOOP surfaces menstrual-cycle context inside its coaching ecosystem, but Recovery still reflects broader recent strain and sleep dynamics.⁶ In practice, a “yellow/red” during late luteal needs cycle-context interpretation, not automatic cancellation of training.

WHOOP HR/HRV measurement validity is generally acceptable for overnight trend use when interpreted longitudinally rather than as a one-off verdict.⁷

Garmin HRV Status + baseline dependence

Garmin’s HRV Status depends heavily on personal baseline construction, and Garmin documentation emphasizes a baseline period requirement (roughly three weeks of sleep data for stable status behavior).⁸⁹ If baseline is immature, recent routine changed, or data quality is inconsistent, premenstrual “strained” labels can be overinterpreted.

For SensAI users, this means: trust trend maturity before trusting color labels.

The SensAI Cycle-Aware Decision Engine (Natural Cycles Branch vs Hormonal Contraception Branch)

A device-agnostic decision engine works better than score chasing because it combines cycle context, physiology, and recent load into one daily action.

SensAI uses two branches:

Natural cycles branch
- Expect larger cyclical amplitude in RHR/HRV.
- Late luteal low-HRV days are not automatically “stop.”
- Weigh symptom direction and multi-signal clustering before downshifting.
Hormonal contraception branch
- Expect flatter or altered cyclic amplitude in many users.
- A sharp HRV/RHR disturbance may be less likely to be normal cycle noise and may deserve faster load adjustment.
- Use trend disruptions, symptoms, and sleep strain as stronger decision inputs.

In the large npj Digital Medicine cohort, naturally cycling users showed higher cyclic amplitudes (RHR 2.73 bpm, RMSSD 4.65 ms) while birth-control-pill users showed blunted amplitudes (RHR 0.28 bpm, RMSSD -0.51 ms), with significant between-cohort differences.² Separate work also shows HRV pattern differences between natural menstrual and oral-contraceptive cycles.¹⁰

That is why SensAI does not apply one universal threshold to everyone.

TRAIN Day Criteria (when low HRV is expected noise, not red alert)

Proceed with planned training when most of these are true:

You are in expected late luteal window (or a known personal low-HRV window).
HRV is down, but within your usual cyclic range.
Resting HR increase is mild and consistent with prior cycles.
Temperature and respiratory rate are near personal expectation.
Symptoms are minimal and not worsening.
Last 48-hour training load is manageable.

Training note: keep session quality high, but monitor in-session RPE drift. SensAI often recommends normal main sets with tighter cooldown and hydration control on these days.

MODIFY Day Criteria (keep session quality, reduce autonomic strain)

Modify instead of cancel when moderate strain signals cluster:

HRV clearly below expected cyclic band
Resting HR meaningfully elevated vs your normal luteal pattern
Sleep quality reduced or fatigue symptoms rising
Heat exposure, poor recovery, or high recent load likely amplifying strain

Typical SensAI modifications:

Reduce high-intensity density by ~20-40%
Keep key technical work, trim volume
Extend warm-up/cooldown
Prioritize fueling, sodium/hydration, and sleep extension

This preserves training continuity without forcing a high-autonomic-cost session on a borderline day.

REST Day Criteria (when multi-signal strain outweighs planned load)

Shift to rest/recovery when multiple red flags align:

HRV suppression is large and persistent (not one-day noise)
Resting HR and temperature are both elevated beyond your usual premenstrual pattern
Respiratory rate rises or illness-like symptoms appear
Felt readiness is strongly negative and worsening
Prior days already had high strain load

On REST days, SensAI prioritizes active recovery, symptom monitoring, and next-day reassessment instead of forcing adaptations that likely will not stick.

Can You Do HIIT in the Luteal Phase? A Risk-Managed Rule Set

Yes, you can do HIIT in the luteal phase. But eligibility should be conditional, not automatic.

A practical SensAI rule set:

Green-light HIIT only if at least 4 of 5 are favorable:
- HRV near expected personal luteal range
- Resting HR not materially elevated
- Sleep acceptable
- Symptoms low/stable
- No excessive heat stress or dehydration context
Downgrade to MODIFY if 2+ caution signals appear.
Skip HIIT (REST or easy aerobic) if multi-signal strain clusters or symptoms trend upward.

This individualized approach aligns with the performance evidence base. McNulty et al. (78 studies) reported that broad universal menstrual-cycle training rules are weak, with pooled effects generally trivial/small and evidence quality frequently low.¹¹ Their direct recommendation supports personalization over rigid phase dogma.

As Kelly Lee McNulty et al. concluded: “General guidelines on exercise performance across the MC cannot be formed; rather, it is recommended that a personalised approach should be taken…”¹¹

Symptom-Signal Conflict Resolution (When Wearable Data and Felt Readiness Disagree)

Conflicts happen. Here is the practical tie-breaker logic SensAI uses:

Data low, you feel good: treat as caution, not stop. Keep session but reduce autonomic cost and reassess post-warm-up.
Data normal, you feel poor: trust symptoms and reduce load. Wearables can miss local pain, migraine, GI issues, or cumulative psychosocial stress.
Data low + symptoms rising: bias toward REST and recovery.
Data mixed for >3 days: run a calibration check (sleep consistency, wear time quality, training monotony, hydration, heat exposure).

The key is consistency: one framework, every day, regardless of app color.

14-Day Practical Playbook for Athletes and General Population (including premenstrual week)

Use this playbook as an implementation layer for the final luteal week through early follicular transition.

Days -10 to -7 (relative to expected bleed)

Establish or confirm current cycle phase context.
Keep baseline collection quality high (sleep wear compliance, routine timing).
Complete key quality sessions if signals are stable.

Days -6 to -3 (common inflection window)

Increase sensitivity to combined HRV + RHR + temperature patterns.
Pre-plan fallback versions of hard workouts (A plan: Train, B plan: Modify).
Tighten sleep and hydration targets.

Days -2 to 0 (premenstrual high-variance window)

Use strict Train/Modify/Rest criteria from above.
Prefer decision confidence over ego pacing.
If doing intensity, use shorter high-quality intervals with longer recoveries.

Days 1 to 3 (early follicular transition)

Reassess trends rather than assuming instant rebound.
Return load progressively if markers normalize.
Document personal response patterns for next cycle.

Days 4 to 7

Resume normal progression when signals and symptoms agree.
Preserve what worked (fueling, scheduling, workout sequencing).

For general population users (not performance-focused), the same framework works with simpler outputs:

TRAIN: full planned session
MODIFY: lighter intensity or shorter duration
REST: movement snack/walk/mobility + recovery focus

Over 2-3 cycles, SensAI typically sees clearer personal patterns and fewer false red days.

Internal links and next steps with SensAI

If you want one implementation takeaway: build your rule set around your own cycle-aware baseline, not generic internet thresholds. SensAI’s device-agnostic coaching layer is designed to do exactly that—turn noisy premenstrual data into clear daily actions.

Schmalenberger KM, et al. “Menstrual Cycle Changes in Vagally-Mediated HRV Are Associated with Progesterone.” Journal of Clinical Medicine, 2020. https://pmc.ncbi.nlm.nih.gov/articles/PMC7141121/ ↩ ↩² ↩³ ↩⁴
A Novel method for quantifying fluctuations in wearable derived daily cardiovascular parameters across the menstrual cycle. npj Digital Medicine, 2024. https://www.nature.com/articles/s41746-024-01394-0 ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶
Charkoudian N, Stachenfeld NS. “Temperature regulation in women: Effects of the menstrual cycle.” 2020 review. https://pmc.ncbi.nlm.nih.gov/articles/PMC7575238/ ↩ ↩²
Giersch GE, et al. “Menstrual cycle and thermoregulation during exercise in the heat.” Journal of Science and Medicine in Sport, 2020. https://pubmed.ncbi.nlm.nih.gov/32499153/ ↩
Oura. “Readiness Score Now Takes Cycles Into Account.” 2025. https://ouraring.com/blog/readiness-score-cycle-consideration/ ↩ ↩² ↩³
WHOOP. “Menstrual Cycle Insights on WHOOP.” https://www.whoop.com/us/en/thelocker/whoop-feature-menstrual-cycle-coaching/ ↩
Miller DJ, et al. “Wrist-Based PPG Assessment of Heart Rate and HRV: Validation of WHOOP.” Sensors, 2021. https://pmc.ncbi.nlm.nih.gov/articles/PMC8160717/ ↩
Garmin. “Understanding the HRV Status on Your Garmin Smartwatch.” 2024. https://www.garmin.com/en-US/blog/fitness/understanding-the-hrv-status-on-your-garmin-smartwatch/ ↩
Garmin Owner Manual. “Heart Rate Variability Status” (3-week baseline requirement). https://www8.garmin.com/manuals/webhelp/GUID-2DA54DF8-8084-40ED-954F-EDA09C13B47F/EN-US/GUID-9282196F-D969-404D-B678-F48A13D8D0CB.html ↩
Blake K, et al. “Heart rate variability between hormone phases of the menstrual and oral contraceptive pill cycles.” Clinical Autonomic Research, 2023. https://pubmed.ncbi.nlm.nih.gov/37294472/ ↩
McNulty KL, et al. “The Effects of Menstrual Cycle Phase on Exercise Performance in Eumenorrheic Women.” Sports Medicine, 2020. https://pubmed.ncbi.nlm.nih.gov/32661839/ ↩ ↩²