Underfueling vs Overtraining: A Wearable-Driven RED-S Framework to Catch Low Energy Availability Before Performance Crashes

If your HRV drops, resting heart rate rises, and workouts feel harder than expected, you are usually told one of two stories: “you are overtraining” or “you need to eat more.” The problem is both can be true, and both can look identical in wearable dashboards for days.

That overlap is where athletes lose months. You deload when you actually need aggressive refueling. Or you keep fueling harder when the true issue is a training-load spike and autonomic strain. SensAI built this guide for that exact decision point: use a 7-day, multi-signal protocol to separate likely underfueling from likely overreaching, then pick the right action early.

Why this is hard: underfueling and overtraining produce overlapping symptoms and wearable noise

Underfueling and overtraining both present as “under-recovery” at first: lower readiness, mood changes, reduced performance, disturbed sleep, and autonomic shifts. The ECSS/ACSM consensus has warned for years that overtraining syndrome is a diagnosis of pattern and duration, not a single metric.¹

At the same time, RED-S can produce many of the same outward signs. The IOC consensus describes RED-S as a syndrome of health and performance consequences caused by low energy availability (LEA), and it explicitly affects both male and female athletes.²

So a single-day HRV drop is not a diagnosis. As Dr. Christopher Bellenger and colleagues wrote, “Resting HRV is largely unaffected by overreaching… additional measures of training tolerance may be required.”³ SensAI treats this as a core rule: one number never decides your day.

RED-S in male endurance athletes: definition, prevalence, and why athletes are missed

RED-S in men is underdetected because athletes and coaches often look for dramatic endocrine signs while ignoring quieter performance and recovery drift. But prevalence data shows the risk is not niche.

A 2024 systematic review/meta-analysis reported that 2737 of 6118 athletes (44.7%) had LEA, including 49.4% of male athletes.⁴ In the same evidence base, 460 of 730 athletes (63.0%) were at risk of RED-S.⁴ That is not edge-case territory. That is mainstream endurance and field-sport reality.

In elite cohorts, risk stratification tools are now exposing clinically meaningful gradients. In one prospective cohort (n=213), IOC CAT2 classification was 55% green, 36% yellow, 5% orange, 4% red.⁵ Critically, bone-stress-injury odds were significantly higher in orange vs green athletes (OR 7.71, 95% CI 1.26-39.83).⁵

As Prof. Margo Mountjoy and the IOC group put it: “REDs was first introduced in 2014… identifying a syndrome of deleterious health and performance outcomes experienced by female and male athletes exposed to low energy availability.”²

LEA vs RED-S vs functional overreaching/non-functional overreaching

Here is the practical separation:

• LEA = energy intake is insufficient relative to exercise expenditure and total physiological needs.²
• RED-S = the broader syndrome (performance + health consequences) that can emerge from persistent LEA.²⁶
• Functional overreaching (FOR) = short-term planned overload with rebound adaptation.
• Non-functional overreaching (NFOR) = prolonged maladaptation and performance decrement from stress-recovery mismatch.¹

These categories can overlap. You can be in NFOR while also in LEA. That mixed state is exactly where most wearable-only interpretations break.

As Prof. Romain Meeusen and colleagues wrote, “A keyword in the recognition of OTS might be ‘prolonged maladaptation’.”¹ Duration and trajectory matter more than one bad week.

The SensAI Signal Disambiguation Protocol (7-day rolling view)

The protocol is simple: do not diagnose from isolated daily scores. Score a 7-day cluster across autonomic trend + sleep trend + load context + fueling context + performance trend.

SensAI uses this sequence:

1. Establish personal 28-day baseline for HRV, resting/overnight HR, sleep regularity, and training load.
2. Review 7-day drift (not just yesterday): direction, consistency, and volatility.
3. Cross-check energy context: intake logging quality, missed fueling windows, appetite suppression, or weight/mood drift.
4. Cross-check load context: acute load spike, intensity density, travel/work stress.
5. Assign likely cluster (A, B, or C) and choose action for 72 hours, then reassess.

Signal cluster A (likely underfueling): HRV suppression or instability + rising resting/overnight HR + sleep disruption + low intake/high expenditure context

Cluster A is most likely when autonomic and sleep stress occur without a clear external load spike, but with strong fuel mismatch context (reduced intake, long gaps between meals, high expenditure blocks).

Evidence supports this caution. In a 2025 randomized 24h energy-availability manipulation trial (n=20), exercise-induced LEA changed sleep and overnight heart-rate behavior, yet overnight HRV itself did not always differ significantly by condition.⁷ In plain language: LEA can distort recovery signals without giving you a clean “HRV warning” every day.

Seasonal fueling swings can also be large in male endurance athletes. Male cross-country skiers reported mean intake of 4050 ± 797 kcal/day on training days versus 5986 ± 924 kcal/day on competition days.⁸ If intake fails to scale with load phase, Cluster A risk rises fast.

Within-day energy balance matters too. Male collegiate soccer data shows that energy deficit patterns across the day can align with metabolic suppression signals even when total daily intake appears acceptable on paper.⁹

Signal cluster B (likely overreaching): load spike + performance drop + autonomic changes without clear intake deficit

Cluster B is more likely when there is a clear training-dose error: abrupt load/intensity increase, performance decrement, and autonomic change despite adequate fueling behavior.

This is where many athletes overreact to single-day HRV values. Meta-analytic work suggests resting RMSSD changes with overreaching are often small (SMD 0.26), with trivial changes in some related indices.³ Weekly trend quality is more informative than snapshots.

Schaal et al. showed stronger discrimination when HRV is averaged across weeks in functionally overreached athletes (SMD 0.81; 95% CI 0.35-1.26).¹⁰ So if your weekly trend is deteriorating while load has clearly spiked, Cluster B becomes more likely even if one morning score looks “okay.”

Signal cluster C (mixed stress): concurrent load escalation and low energy availability

Cluster C is common in ambitious athletes: you increase training load and simultaneously underfuel key sessions. This mixed pattern often produces the worst confusion because both causal pathways amplify each other.

In practice, Cluster C looks like this: rising overnight HR, unstable HRV, poor sleep continuity, a load spike, and obvious fueling misses. If this persists for 7-14 days, treat it as high risk for RED-S progression and prolonged performance decline unless corrected quickly.²¹

Threshold-led action rules: Fuel More vs Reduce Load vs Full Recovery + Clinical Screen

Action should follow pattern, not emotion. Use these 7-day rules:

Pattern over last 7 days	Most likely issue	First action (72h)	Re-check trigger
Autonomic stress + sleep disruption + no major load spike + clear intake deficit	Cluster A (underfueling dominant)	Fuel More (carbohydrate timing + total energy restoration)	If no improvement in 3-5 days, add load reduction + clinical screen
Load spike + performance drop + autonomic drift + intake appears adequate	Cluster B (overreaching dominant)	Reduce Load (volume/intensity deload microcycle)	If persistent >2 weeks, escalate medical review
Both load spike and intake deficit with worsening trends	Cluster C (mixed)	Fuel More + Reduce Load simultaneously	If red flags present, move directly to clinician escalation

This is where SensAI is intentionally practical: it translates ambiguous wearable trends into a concrete daily prescription rather than another generic readiness score.

Fuel-first protocol (carbohydrate timing, total energy restoration, within-day energy balance)

When Cluster A is likely, prioritize restoring energy availability before trying to “train through” autonomic stress.

Fuel-first sequence:

1. Fix within-day deficits first: avoid long daytime low-energy windows around training sessions.⁹
2. Anchor carbohydrate around key sessions: pre-session and early post-session fueling improve recovery trajectory.
3. Raise total intake to match current block: competition-like or double-session weeks need explicit intake scaling.⁸
4. Track 3 outcomes daily: sleep continuity, overnight HR drift, and session RPE at expected pace/power.

SensAI applies this as a dynamic nutrition adjustment layer tied to your training calendar and wearable response, not static macro targets.

Training-dose protocol (volume/intensity reductions, recovery microcycle sequencing)

When Cluster B is likely, reduce stress dose rapidly and restore adaptability.

Training-dose sequence:

1. Cut volume first for 3-4 days while preserving technical movement quality.
2. Reduce high-intensity density (especially stacked hard days).
3. Insert recovery microcycle sequencing: easy day -> moderate day -> re-evaluate -> only then progress.
4. Do not judge recovery from one morning score; use 3-7 day trend plus performance feel.

SensAI operationalizes this as adaptive programming: if weekly trend improves, load progresses; if not, deload persists.

How to use WHOOP/Oura/Garmin trends without overinterpreting single-day HRV

Treat wearable ecosystems as signal collectors, not verdict engines. WHOOP, Oura, and Garmin each capture useful autonomic and sleep data, but none should be interpreted in isolation.

A practical cross-device rule:

• Use weekly medians/rolling averages for HRV and overnight HR.
• Look for agreement across domains (autonomic + sleep + load + fueling context).
• Ignore one-off outliers unless supported by symptoms/performance.
• Prioritize direction over absolute score (your baseline matters more than someone else’s “normal”).

Bellenger’s and Schaal’s findings support this trend-first approach: weekly or repeated measures outperform single snapshots for training-status interpretation.³¹⁰

SensAI’s value is exactly this normalization layer: combine wearable telemetry, nutrition logs, and training load into one athlete-specific recommendation each morning.

RED-S red flags needing clinician escalation (bone stress injury risk, endocrine signs, persistent dysfunction)

Do not delay escalation when risk signs cluster. Wearables guide training decisions, but they do not replace clinical assessment tools like IOC REDs CAT2.⁶

Escalate to a qualified sports medicine clinician if you see:

• Persistent performance decline despite 2+ weeks of correction
• Recurrent bone pain or suspected bone stress injury history/risk
• Endocrine or reproductive changes (including low libido in men)
• Mood, sleep, and fatigue dysfunction that does not normalize with fuel/load correction

Why urgency matters: in CAT2-classified athletes, orange-risk status was linked to substantially higher prospective bone-stress-injury odds (OR 7.71), and in male athletes, low sex drive in CAT2 was associated with markedly higher BSI odds (OR 16.0).⁵

Brand-differentiated angle: How SensAI unifies wearable telemetry + nutrition logs + training load into a daily prescription

Most apps tell you what happened. SensAI tells you what to do next.

SensAI combines:

• Wearable trend data (WHOOP/Oura/Garmin/Apple Watch)
• Nutrition timing and total intake context
• Training-load progression and intensity density
• Symptom/performance check-ins

Then it outputs a clear daily action: Fuel More, Reduce Load, or Recovery + Clinical Screen. That is the differentiator for athletes trying to avoid false calls between underfueling and overtraining.

Internal links to continue your protocol with SensAI

• Data-Driven Deload Weeks: HRV, Sleep Debt, and Training Load
• Overtraining vs Overreaching: How Wearables Detect the Difference
• Fitness Apps Integrated with Oura and Whoop for HRV Control
• SensAI FAQ
• Contact SensAI

FAQ mapped to high-intent queries (carbs when HRV drops, high resting HR despite deload, male RED-S symptoms)

Am I overtraining or underfueling?

If there is a clear load spike and no obvious intake deficit, overreaching is more likely. If load is stable but energy mismatch is clear, underfueling is more likely. If both are present, treat as mixed stress and intervene on both fronts immediately.

Can HRV detect low energy availability?

Not reliably as a single-day diagnostic. LEA can alter sleep and overnight HR patterns even when overnight HRV does not change clearly after short exposure.⁷ Use clustered signals, not one HRV reading.

Why is my resting heart rate high even after a deload?

Persistent high resting/overnight HR after load reduction suggests unresolved stress outside training dose, often fueling mismatch, illness, or mixed stress. Re-check intake timing, sleep continuity, and non-training stress before reloading volume.

What are male RED-S symptoms I should take seriously?

Watch for persistent underperformance, fatigue, recurrent injury risk (especially bone stress), mood/sleep disruption, and endocrine signals such as low libido. Male athletes are frequently under-screened despite high LEA prevalence.⁵⁴

Should I increase carbs when HRV drops?

If HRV drop occurs with signs of low energy availability (sleep disruption, rising overnight HR, intake deficit context), yes—prioritize fuel-first correction. If HRV drop is tied to an acute load spike with adequate fueling, reduce load first and reassess in 72 hours.

How does SensAI help me make this call day by day?

SensAI applies the 7-day disambiguation protocol automatically, merges wearable and nutrition context, and gives a daily recommendation with confidence level. That removes guesswork and lowers the chance of missing early RED-S drift.

The bottom line: do not ask one metric to solve a multi-system problem. Use trend clusters, act early, and let SensAI translate complexity into an actionable plan before performance crashes.

---

Footnotes

1. Meeusen R, Duclos M, Foster C, et al. “Prevention, diagnosis and treatment of the overtraining syndrome: ECSS/ACSM consensus statement.” Medicine & Science in Sports & Exercise, 2013. https://pubmed.ncbi.nlm.nih.gov/23247672/ ↩ ↩² ↩³ ↩⁴

2. Mountjoy M, Sundgot-Borgen J, Burke L, et al. “IOC consensus statement on Relative Energy Deficiency in Sport (REDs): 2023 update.” British Journal of Sports Medicine, 2023. https://pubmed.ncbi.nlm.nih.gov/37752011/ ↩ ↩² ↩³ ↩⁴ ↩⁵

3. Bellenger CR, Fuller JT, Thomson RL, et al. “Monitoring athletic training status through autonomic heart rate regulation: a systematic review and meta-analysis.” Sports Medicine, 2016. https://pubmed.ncbi.nlm.nih.gov/26888648/ ↩ ↩² ↩³

4. Gallant TL, Eirale C, O’Reilly M, et al. “Low energy availability and relative energy deficiency in sport: a systematic review and meta-analysis.” 2024. https://pubmed.ncbi.nlm.nih.gov/39485653/ ↩ ↩² ↩³

5. Koltun KJ, Strock NCA, Southmayd EA, et al. “Application of the IOC REDs CAT2 in elite athletes and associations with prospective bone stress injury.” British Journal of Sports Medicine, 2024. https://pubmed.ncbi.nlm.nih.gov/39164063/ ↩ ↩² ↩³ ↩⁴

6. Mountjoy M, Sundgot-Borgen JK, Burke LM, et al. “IOC REDs CAT2: Clinical Assessment Tool Version 2 (2023).” British Journal of Sports Medicine, 2023. https://pubmed.ncbi.nlm.nih.gov/37752002/ ↩ ↩²

7. Scott TJ, Driller MW, Lastella M, et al. “Twenty-four-hour low energy availability effects on sleep and overnight cardiac autonomic activity in endurance athletes.” 2025. https://pubmed.ncbi.nlm.nih.gov/40523229/ ↩ ↩²

8. Łuszczyk M, Kacperak K, Książek A, et al. “Low energy availability in male cross-country skiers across an annual training cycle.” Nutrients, 2024. https://pubmed.ncbi.nlm.nih.gov/39064722/ ↩ ↩²

9. Taghavi N, Keenan S, Verdoucci T, et al. “Within-day energy balance and markers of metabolic suppression in male collegiate soccer players.” 2021. https://pubmed.ncbi.nlm.nih.gov/34444803/ ↩ ↩²

10. Schaal K, Le Meur Y, Louis J, et al. “Heart rate-derived indices for monitoring training status in functionally overreached athletes: a meta-analysis.” 2021. https://pubmed.ncbi.nlm.nih.gov/33533045/ ↩ ↩²