So to start off I have a spinal cord injury (T1 ASIA D) and I’m using Whoop to help backoff when I’m not recovered enough to be active. I think there is a problem with your algorithm for people who don’t have normal spinal cord function.
I’ve been hooking up my Whoop data to a bunch of other data and using claude to analyze it against SCI literature. Here’s the question I asked:
Any signals that can help detect when my nervous system is overworked? My PT thinks this makes sense that it happens, but feel like I am always guessing. There were definitely episodes on March 15th and March 22nd. It seems to even happen even when whoop says I have had good recovery. i want to know when to back off. Also can you look through the literature about this so I can understand it better and maybe find ways to improve it or make myself more resilient
And here is the full response:
The research agent is still working. I have enough data to give you a solid answer. Here’s a comprehensive breakdown:
Your Nervous System Episodes: What the Data Shows
March 15 — The “WHOOP Said 93” Problem
Date Recovery HRV RHR Mar 11 33 52.6 57 Mar 13 80 48.8 60 Mar 14 93 52.5 55 Mar 15 93 66.7 50
The problem is right there: you went from 33 → 80 → 93 in two days. And your HRV (66.7) was 63% above your long-term baseline of 41 ms. WHOOP read this as “great” — but a spike that far above personal baseline, especially after a rapid rebound from very low scores, can actually signal autonomic instability, not readiness. The system is oscillating, not stable.
March 22 — The Accumulation Crash
Date Recovery HRV Notes Mar 15 93 66.7 episode Mar 17 70 43.0 Mar 18 89 52.5 Mar 19 74 50.6 2x circus arts Mar 20 88 55.3 Mar 22 47 44.5 episode Mar 24 31 35.9 Mar 25 26 36.3 Mar 26 12 28.5 Mar 28 17 24.7
This is a textbook accumulation crash. Five consecutive high-recovery days post-episode, with two circus arts sessions on March 19 — and then the floor fell out. WHOOP didn’t catch the cumulative load building. Note also: recovery scores of 31, 26, 12 on March 24–26, but you still trained (weightlifting + multiple circus arts) on March 27–28. That’s likely extending the crash.
Why WHOOP Misses This for You
WHOOP’s recovery algorithm is trained primarily on able-bodied people. For SCI, there are fundamental differences:
1. Sympathetic nervous system is partially disconnected below your lesion. WHOOP’s HRV (RMSSD) measures parasympathetic tone via vagal activity — primarily above the injury level. It doesn’t capture what’s happening below. You can have good vagal tone while your sub-lesional autonomic system is dysregulated.
2. HRV spikes above baseline can mean instability. In intact nervous systems, higher HRV = more parasympathetic = more recovery. In SCI, a very high RMSSD can reflect autonomic hypersensitivity or parasympathetic rebound after sympathetic suppression — the system oscillating rather than being in steady state. See: Krassioukov et al., 2009 — autonomic cardiovascular dysfunction in SCI.
3. Circus arts is neurologically expensive in ways calories don’t capture. Complex, novel movement recruits spinal interneuron circuits and proprioceptive processing that is metabolically cheap but neurologically demanding — especially with your injury pattern. The 55-minute circus session on March 19 likely cost far more “neural budget” than it cost in calories.
Signals That Actually Predict Your Episodes
Based on the data, here’s what correlates:
Signal 1: HRV spike >50% above your personal baseline
Your baseline HRV is ~41 ms. The day before March 15 it was 52.5, and on the day itself it was 66.7. A reading >55 ms (~35% above baseline) in context of preceding instability appears to be a warning, not a green light. This is the opposite of the typical interpretation.
Practical rule: If WHOOP HRV is >55 ms and the preceding 3 days had high variance in recovery scores, treat it as a yellow flag, not green.
Signal 2: Recovery swing >50 points in 3 days
You went 33 → 93 in 2 days before March 15. That kind of rebound suggests the nervous system is oscillating, not recovered. A smooth recovery trends up 10–20 points per day.
Practical rule: If recovery goes up >30 points in a single day after a low, treat it as unstable, not recovered.
Signal 3: 5+ consecutive days with recovery ≥70, following an episode
The week after March 15 had five days ≥70. This feels good but may represent an “allostatic window” where the system is primed for a crash if you push into it. The March 19 circus arts was likely the trigger.
Practical rule: After a nervous system episode, even if scores look good, treat the following 7–10 days as “modified load” regardless of what WHOOP says.
Signal 4: HRV trend direction, not just value
Look at the 3-day trend. If HRV is falling (55 → 52 → 45) that’s more concerning than a steady 45.
What the Research Says
The closest analog in the literature is Long COVID autonomic dysfunction and post-exertional malaise (PEM), which has similar features: intact HRV above a lesion/disruption, misleading wearable signals, and crash patterns after exertion. A 2025 study (medRxiv preprint) found:
HRV remains depressed for 24+ hours after exercise at or above the first ventilatory threshold (VT1) in autonomic-dysregulated patients, vs. 3–6 hours in healthy controls
Morning HRV drops are early warning signs of impending crashes, often appearing before subjective symptoms
Wearable HRV monitoring is valid for identifying overexertion threshold
For SCI specifically:
West et al., 2012 — established that HRV is significantly reduced in SCI vs. able-bodied, and that the degree of reduction correlates with lesion completeness and level. RMSSD in particular reflects the degree of intact vagal efferents.
Krassioukov & Claydon, 2006 — reviewed how cardiovascular and autonomic control is compromised at and above T6, with particular emphasis on the loss of sympathetic modulation meaning that the remaining parasympathetic system can become hypersensitive/unbalanced.
Phillips et al., 2012 — found that autonomic dysreflexia episodes are preceded by measurable cardiovascular instability patterns, and that beat-to-beat BP variability increases before full AD episodes.
The concept directly matching your experience is Autonomic Cardiovascular Instability (ACI) — a subthreshold state of autonomic dysfunction that doesn’t always trigger full AD but causes symptoms (fatigue, brain fog, tone changes, pain amplification). It’s poorly studied but clinically recognized at SCI rehab centers like Craig Hospital.
How to Build Resilience / Practical Changes
1. Add a morning resting HR trend check. RHR rising 3+ bpm above your personal average over 3 consecutive days is a validated overreaching signal (even in SCI patients). Your RHR jumped from 50–51 → 59–65 in the days after March 22. Track this directly.
2. Use HRV stability (day-to-day variance) as your primary metric, not the absolute value. High day-to-day variance (e.g., 66.7 one day, 43 the next) is more concerning than a consistent 40. A stable, slightly-below-average HRV often means more than an erratic high one.
3. Pacing windows for circus arts / novel movement. After any session involving complex proprioceptive or novel motor patterns, apply a 48-hour low-load window regardless of WHOOP score. The neural recovery cost isn’t reflected in the caloric cost.
4. Establish a “traffic light” protocol:
Green: HRV within ±20% of baseline, stable 3-day trend, recovery 50–75
Yellow: HRV >35% above baseline OR >20% below, recovery <40 or >85 after episode, any rapid swing in recovery
Red: Recovery <30, post-episode within 7 days, HRV falling 3 days in a row
5. Aerobic threshold training for autonomic conditioning. The most evidence-supported intervention for improving autonomic resilience in SCI is consistent sub-threshold aerobic exercise — staying below VT1 (the point where you can’t hold a conversation). Bhambhani et al., 2010 showed this improves cardiovascular autonomic function in incomplete SCI. FES cycling, arm ergometry, or low-intensity wheelchair work at conversational pace 3x/week.
6. Heart rate during workouts as a proxy. If your HR is going above ~110–120 bpm during activity, you’re likely pushing the autonomic system harder than the activity “looks.” For incomplete SCI, maintaining HR below that threshold during recovery periods may help prevent accumulation.
“circus arts” btw was me trying to learn to juggle on a beach. I’m not sure if this is something Whoop could support better. I imagine you’d need to add some indication of existing conditions, but it would be really helpful to adjust the algorithm.
Thanks