You wake up feeling slightly off — not sick, exactly, just not quite yourself. You glance at your Apple Watch and see your resting heart rate ran higher than usual overnight. Coincidence, or did your body notice something before you did?
It turns out that in the day or two before many common infections take hold, several of the metrics your watch already records tend to drift in a recognisable direction. Your Apple Watch can't tell you that you have the flu, and it certainly can't diagnose anything. But it can show you when your body is quietly working harder than normal — and sometimes that nudge arrives before the first sniffle.
The Signals That Shift Before You Feel It
When your immune system ramps up to fight something off, it changes how your body idles at rest. That effort shows up in the overnight numbers your watch is already collecting. Four metrics tend to move together.
Resting heart rate creeps up
Mounting an immune response is metabolically expensive. Your body raises its baseline effort, and one of the clearest places that shows is your heart rate while you sleep. A rise of a few beats per minute above your usual overnight figure — sustained across a night or two — is one of the most commonly reported early signals. On its own it means little, which is why context matters, but a jump you can't explain by a late workout or a stressful day is worth noticing. Our guide to resting heart rate explains why this number is such a sensitive gauge of what's going on beneath the surface.
Heart rate variability drops
Heart rate variability (HRV) measures the tiny differences in timing between consecutive heartbeats. Higher variability generally reflects a relaxed, well-recovered nervous system; a lower value points to strain. When your body diverts resources to an immune response, HRV often dips below your normal range a day or two before you feel unwell. Like resting heart rate, it's noisy from night to night — but a clear step down from your personal norm, alongside the other signals here, adds to the picture.
Breathing rate rises
Your overnight breathing rate is one of the most stable metrics your Apple Watch records. Most people vary by less than a breath per minute from one night to the next, which is exactly what makes a change stand out. An increase of even one or two breaths per minute can accompany the early stages of a respiratory or systemic infection. Because it's normally so steady, it's a quietly powerful early indicator — our piece on respiratory rate digs into why this overlooked number deserves attention.
Wrist temperature ticks up
Newer Apple Watch models estimate how your skin temperature at the wrist deviates from your own overnight baseline. This isn't a thermometer reading and it can't tell you your core body temperature, but a sustained rise above your usual level can appear before a noticeable fever. Temperature is influenced by plenty of unrelated things — a warm bedroom, alcohol, the menstrual cycle — so it's most useful read alongside the other signals rather than alone. Our guide to wrist temperature covers what this sensor can and can't do.
What the Research Actually Shows — and What It Doesn't
The idea that wearables can flag an oncoming illness isn't just marketing. Several large studies — including research during the COVID-19 pandemic — found that metrics like resting heart rate, HRV and skin temperature often shifted one to three days before people reported symptoms. In many participants, the combination of signals moving together was detectable before the person felt sick.
That's genuinely useful, but it comes with heavy caveats worth stating plainly:
- It doesn't happen for everyone, or every illness. Plenty of infections produce no clear early signature, and some people simply don't show one.
- False alarms are common. A hard workout, a few drinks, a bad night's sleep, travel across time zones, or a stressful week all push these same metrics in the same direction. Your watch can't tell an immune response apart from a tough training block.
- It's a probability, not a test. At best these signals raise the odds that something is going on. They never confirm what, and they never rule anything out.
In other words, the research supports a modest, honest claim: patterns in your data can hint that your body is under unusual load. It does not support the idea that a watch can detect or diagnose a specific illness.
Why This Is a Rest-Day Signal, Not a Diagnosis
Given all that, what's the point? The value is behavioural, not medical. If your numbers suggest your body is working harder than usual, that's a reasonable prompt to ease off — take a lighter training day, prioritise sleep, drink more water, and give your immune system room to do its job. None of that requires knowing what, if anything, is actually brewing.
What these signals are emphatically not is a diagnosis. A wearable cannot tell you whether you have a cold, the flu, or nothing at all, and it should never be the reason you delay or avoid seeing a doctor. If you have real symptoms — a persistent fever, breathlessness, chest pain, or anything that worries you — the right move is to talk to a healthcare professional, regardless of what your watch shows. Treat the data as a gentle heads-up to look after yourself, never as a verdict.
Reading the Signals Without Overreacting
The difference between a useful early warning and a source of needless anxiety comes down to how you read the data.
Look for several signals, not one. A single high-heart-rate night is noise. A night where resting heart rate is up, HRV is down, and breathing rate has risen — all at once — is a more meaningful cluster. And one night rarely means much on its own; a shift that holds for a couple of nights says more. This is the whole reason trends beat single readings when interpreting health data.
Know your own normal. A resting heart rate of 60 is meaningless without knowing whether yours usually sits at 52 or 68. What matters is deviation from your personal baseline, not any textbook range. Everyone's numbers are different, and the signal lives in the change, not the absolute value.
Account for context. Before reading anything into a shift, ask what else was going on. Late alcohol, an intense workout, poor sleep, a stressful deadline, or a long flight can all move these metrics without any illness involved. If there's an obvious explanation, that's usually the story.
Turning Scattered Signals Into One Clear Picture
Watching four metrics across several nights, remembering your own baseline for each, and mentally filtering out the workout you did on Tuesday is a lot to track by hand. This is exactly where an interpretation layer helps.
Health AI Insight reads these metrics from Apple Health and builds a personal 90-day baseline for each one — your resting heart rate, HRV, breathing rate, blood oxygen and wrist temperature. On paid plans it lets you know when a metric drifts and stays outside your own norm, so you don't have to scroll through charts to spot it. It's framed exactly as it should be: a deviation from your personal baseline worth a look, never a diagnosis.
The context piece matters just as much. When you mark "illness" in the app's life context diary, your weekly report reads that week's dip in the right light — as recovery from being unwell, not as lost fitness or a training setback. That way a rough week doesn't quietly drag down the story your data tells about your progress.
Used this way, your Apple Watch becomes something realistic and genuinely useful: not a device that diagnoses illness, but one that occasionally taps you on the shoulder and suggests today might be a good day to take it easy.