Every athlete knows the morning: you feel slightly off, and you are not sure whether it is yesterday's session or the start of a cold. Wearables promise to settle it, and the research behind that promise is real. So we asked our own data the same question: across the sick days our athletes log, what do the overnight numbers do in the week before, and in the weeks after?
The answer changed where IntervalCoach looks. The week before is quieter than the headlines suggest. The weeks after are where the signal is, and it is clearer and more useful than we expected.
How we measured it
IntervalCoach reads each athlete's overnight resting heart rate, HRV, sleep and, where the wearable provides them, breathing rate, blood oxygen and skin temperature. Athletes mark sick days on their calendar. For every logged illness we compared each night in the week before the first sick day, and the two weeks after the last one, with that athlete's own normal over the four weeks before. Nights that fell inside a different illness were left out, so one cold cannot distort the next.
The result is an average across hundreds of illnesses, expressed in standard deviations from each athlete's own normal. Around 0.2 is a small but real shift; 0.5 and above is a large one.
The week before looks almost normal
In the week before a logged sick day, resting heart rate and HRV stay close to normal. Resting heart rate edges up in the last night or two, reaching about 0.15 the night before. That rise is consistent, and it is too small to act on by itself: resting heart rate also moves with training, so a shift this size cannot tell an oncoming illness from a hard day.
This differs from what wearable research has shown, and the difference is instructive. Mishra and colleagues found resting heart rate rising before symptoms in most confirmed COVID-19 cases in their smartwatch cohort (Nature Biomedical Engineering, 2020). Radin and colleagues tracked influenza-like illness across a US state from the share of Fitbit users with raised resting heart rate (The Lancet Digital Health, 2020). Those studies looked at a specific, confirmed infection, or at a whole population at once. An athlete's logged sick day is a mixture: a cold, a stomach bug, the flu, sometimes marked when the symptoms start and sometimes a day later. Averaged across that mixture, and across athletes whose heart rate also moves with training, the early signal mostly washes out.
Skin temperature is the one exception worth naming. It rose in the two nights before illness, by about 0.3, which is what the research predicts: Smarr and colleagues showed that a wearable's temperature sensor picks up illness-related fever (Scientific Reports, 2020). Only a small share of our athletes' wearables report skin temperature, so at this depth we cannot separate that rise from noise. It matches the literature, and we treat it as unconfirmed until more data arrives.
Recovery outlasts the illness
The picture after the illness is the opposite of quiet. The disruption is largest around four days after the last sick day, when resting heart rate sits about 0.35 above normal. HRV drops furthest in the first days back and stays below normal for more than a week. Averaged over the first week, resting heart rate is about 0.25 above normal and HRV about 0.2 below. In the second week HRV is essentially back, and resting heart rate is still slightly raised.
That timing matters because it is out of step with how athletes feel. Most people call an illness over the day the symptoms go, and that is when the body's overnight numbers are only starting to show the cost.
A longer illness leaves a deeper mark
Split by how long the illness lasted, the pattern is ordered. In the first week back, resting heart rate averages about 0.12 above normal after one or two days ill, about 0.29 after three to six days, and about 0.41 after a week or more. HRV follows the same order.
The second week is where the longest illnesses stand out. After a week or more ill, resting heart rate is still about 0.3 above normal and HRV still below it. After three to six days, HRV is back and resting heart rate has settled to a small elevation of about 0.12. The shortest illnesses leave that same small elevation through both weeks, which is why even a one-day bug gets a few easier days.
What IntervalCoach does with this
Spotting an illness on the way leans on the signals that do not move just because you trained hard: skin temperature, breathing rate and blood oxygen. When two of those line up, IntervalCoach eases your training toward recovery. A sustained rise in resting heart rate can count as one of the two, so a single illness-specific signal can be confirmed by it, but it never triggers the response on its own. The data above is why: on average it moves too little, too late, to carry that decision alone.
Coming back is where the data is strongest, so that is where the rules are firmest:
- At least five days. Even a short illness gets a graded return.
- Longer after a longer illness. The ramp scales with how many days you were ill, up to two weeks.
- Your numbers give the all-clear. Easing holds while resting heart rate is 5% or more above your normal, or HRV 10% or more below it, even if your readiness looks green.
What it means for your training
Treat the end of an illness as the start of recovery. The nights after you feel better are when your overnight numbers are furthest from normal, and the longer you were ill, the longer that lasts. Ease back over several days, and let your resting heart rate tell you when you are ready.
On the way in, watch for more than one sign at once. A single high resting heart rate is noise. A raised temperature, faster breathing or a drop in blood oxygen, together with resting heart rate creeping up, is worth an easy day.
As more athletes connect wearables that measure skin temperature, we will repeat this analysis and see whether that early rise holds. The earlier read on how fitness changes recovery uses the same overnight data.
— Martijn
Frequently asked questions
Can my watch tell me I am getting sick before I feel it?
Sometimes, but less reliably than headlines suggest. Studies of confirmed COVID-19 cases found resting heart rate rising before symptoms in a majority of people. Across the everyday sick days our athletes log, the week before looks almost normal on average: resting heart rate edges up only in the last night or two, by too little to act on alone. An early warning needs more than heart rate.
Which signal is most useful for spotting an oncoming illness?
The ones that do not move just because you trained hard: skin temperature, breathing rate and blood oxygen. In our data skin temperature rose in the two nights before illness, as the research predicts, but too few wearables report it for us to confirm that yet. IntervalCoach treats a sustained resting heart rate rise as backing for those signals, never as a verdict on its own.
How long does it take to recover after being sick?
Longer than the illness. In our data overnight resting heart rate and HRV are most disrupted around four days after the last sick day, and resting heart rate is still slightly raised in the second week. The disruption comes after you feel better, which is when it is tempting to go back to hard training.
Does a longer illness need a longer comeback?
Yes. The longer the illness lasted, the deeper the disruption afterwards: in the first week back, resting heart rate sits more than three times further from normal after a week or more of illness than after a day or two. After a week or more ill, it is still clearly raised in the second week, and HRV is still below normal.
When can I train hard again after being sick?
When your resting heart rate and HRV are back at your own normal, which is a better all-clear than a countdown from the calendar. IntervalCoach eases you back over at least five days, longer after a longer illness, and holds the easing while either number is still clearly off.
Is this a medical study?
No. It is an observational read of the sick days our athletes log against their own overnight data. Logged sick days cover everything from a cold to the flu, and athletes mark them when they notice them. It describes our population and it shapes how the coach behaves; it is not medical advice.