Health pattern detection

Your watch might know you're getting sick before you do

It sounds like science fiction, but large-scale research confirms it: wearable biomarkers often start shifting one to three days before illness symptoms appear. In a 3 318-person cohort, a smartwatch alerting system flagged 67 of 84 infections (80 %), with pre-symptomatic signals a median of three days before symptoms (Alavi 2022). In controlled influenza and rhinovirus challenge studies, wrist sensors told infected from healthy participants 12 to 36 hours before symptom onset, most models above 80 % accuracy (Grzesiak 2021). The same alerting system also fired on days with no infection, after stress, alcohol or travel, at a mean of 1.15 alert days per person (Alavi 2022). Your body begins fighting an infection well before you feel the first sniffle, and that fight leaves traces in your data.

Why the signals move before you feel anything

The immune response starts hours after exposure, long before the symptoms that are its by-products. Cytokines released in that first phase raise body temperature slightly, push the autonomic nervous system toward its sympathetic branch, and make the heart work harder to move more blood through a warmer body. Each of those has a wearable signature: a higher overnight heart rate, a lower HRV, faster breathing, a warmer wrist, more restless sleep. None of them is specific to infection, which is why the pattern matters more than any single signal.

The early warning signs

• Elevated resting heart rate: even 5 to 10 bpm above your personal baseline can signal the immune system ramping up. The overnight minimum moves first
• Declining HRV: your autonomic nervous system shifts toward sympathetic (fight) mode, typically a drop of 20 % or more against your baseline
• Overnight SpO2 dips: subtle drops in blood oxygen during sleep, most relevant for respiratory infections
• Disrupted sleep architecture: less deep sleep, more awakenings, a longer time in bed with less sleep in it
• Faster overnight breathing: respiratory rate rises with fever and with airway involvement; a jump of two breaths a minute above baseline is unusual for a healthy night
• Increased skin temperature: where the sensor supports it, the wrist warms a night or two before a measured fever

What the research does not say

Two limits are worth being clear about. First, the same pattern follows a hard training session, a night of drinking, a long flight or a stressful week, which is why the Stanford system raised more than one false alert per person over the study. A warning is a reason to take it easier, not a diagnosis. Second, the studies were done with continuous overnight data from a snug wrist device; a watch worn loosely or taken off at night has nothing to compare. The signal exists, and it is only as good as the nights that feed it.

How MyBodyAI flags it

A multi-signal index does not read measurements one at a time. It looks at patterns over time, compares them with the shapes described in the literature, and widens or narrows the window it looks back over according to how dense the data is. Each signal is weighed against your own baseline and any of them can raise the index on its own; when several move together, resting heart rate up, HRV down, sleep disturbed, breathing faster, it rises faster and the warning is firmer. The app names the signals that moved, so you can tell a warning that follows a race from one that follows nothing. It is not a diagnosis; it is a reason to take it easier and support your immune system. How many days of warning it gives you personally has not been measured, and the app says so rather than borrow the study's number.

What to do with a warning

• Swap the hard session for an easy day or rest; the immune response needs the energy
• Go to bed early and keep the room cool; the next night's data will tell you whether the pattern is resolving
• Drink, eat normally, skip alcohol
• If symptoms follow, you have a day's head start on rest. If nothing follows and the signals settle, it was a hard day or a late night, and that is worth knowing too