Your Body Has Been Logging the Evidence

Why "known triggers" lists don't work for you

Search any autoimmune condition and you'll get the same generic trigger list: stress, infection, poor sleep, certain foods. It's not wrong. It's just averaged across thousands of people, and you are not the average. Two people with the same diagnosis can have completely different triggers, different lag times, and different thresholds. The population-level list tells you what's plausible. Only your own data tells you what's true for you.

That's the case for treating yourself as a study with one participant. Not to replace your specialist, but to walk in with evidence instead of impressions.

The data you already have

If you wear an Oura, a Whoop, or an Apple Watch, you're already collecting the hard part: continuous, passive, objective signals.

• Heart rate variability (HRV): a readout of your autonomic state. It tends to drop when your body is under load, sometimes before you feel anything.
• Resting heart rate and skin temperature: slow upward drifts can shadow rising inflammation. In Stanford's wearable research, higher-than-normal heart rate and skin temperature tracked with elevated C-reactive protein, an inflammation marker, and flagged illness before symptoms appeared.^[1]
• Sleep: duration and quality, which feed directly into inflammatory signaling.^[2]

What the wearables can't see is what you did and how you felt. That's the half you log: symptoms, medications, food, alcohol, stress, menstrual cycle. The trick is to make logging cheap enough that you actually do it: a voice note after dinner, a photo of a lab result, a one-tap symptom score. Friction is the enemy; the most rigorous tracking system is the one you'll still be using in week six.

The method

You don't need statistics training. You need four habits.

1. Pick two to four variables, not twenty. Start with one suspected input (say, NSAIDs, or alcohol, or sub-6-hour sleep) and one or two outputs (a symptom score, a flare yes/no). Tracking everything guarantees you'll track nothing consistently.

2. Mind the lag. This is the single biggest reason people miss their own triggers. Autoimmune responses are rarely same-day. The ibuprofen example showed up 24 to 48 hours later. When you look for patterns, line up today's symptoms against the last few days of inputs, not just today's.

3. Wait for repetition. One bad day after a glass of wine is a coincidence. The same sequence happening every time, across several independent instances, is a signal. Resist acting on n=1; look for n=several.

4. Watch the confounders. A stressful week often arrives as a bundle: bad sleep, skipped meds, more alcohol, less movement, all at once. When inputs cluster, you can't tell which one did it. The cleanest test is the old-fashioned one: change a single thing and watch. Suspect NSAIDs? Swap to acetaminophen for a few weeks (with your doctor's okay) and see if the pattern breaks.

Three patterns worth looking for

• NSAIDs and gut symptoms. Ibuprofen and other NSAIDs have long been suspected as an IBD flare trigger, but the population evidence is genuinely mixed. Some reviews find an association;^[3][4] a large 2024 analysis of roughly 35,000 patients suggests much of that link may be confounding rather than cause.^[5] That contradiction is exactly why your own data matters. If you have IBD and reach for ibuprofen, a personal log is the only way to see whether it's true for you. If it is, the effect usually shows up 24 to 48 hours later, not the same day.
• Sleep debt and next-day symptom load. Short or fragmented sleep is associated with higher inflammatory signaling,^[2][6] and many people see their symptoms track their sleep with a one-to-two-day lag. Your wearable already has this data. Pair it with a daily symptom score and look back a day or two, not just at last night.
• Low HRV and rising strain. A sustained drop in HRV reflects rising physiological load, and lower HRV tracks with disease activity and inflammation in autoimmune conditions like rheumatoid arthritis.^[7][8] Whether a dip reliably forecasts a specific flare is still an open question, but watching the trend can prompt you to rest, hydrate, and ease your load on the days your body is signaling strain.

What to do when you find one

Bring it to your specialist. A pattern you've documented across a month ("every time I log X, this happens 36 hours later") is a far better use of a 15-minute appointment than "I think something's been off." It turns a vague visit into a specific conversation, and it gives your doctor something concrete to confirm, rule out, or act on.

The honest limits

A correlation in your own data is a hypothesis, not proof. It tells you where to look, not what's certain. Correlation isn't causation; a third factor could be driving both. And some flares have no findable trigger at all. An empty result isn't a failure, it's information. Most importantly: never stop or change a prescribed medication on a hunch. Use your data to start the conversation with your clinician, not to overrule them.

Doing this without the spreadsheet

Running this by hand (exporting wearable data, logging symptoms, lining up lag windows, filtering out the noise) is tedious, and easy to get wrong. It's exactly the kind of pattern-matching software should do for you.

Immunally is a tracking and insight tool, not a medical device, and doesn't diagnose or treat any condition. Always consult your healthcare provider before making changes to your care.