Check wearables data before changing meals

A commenter used Claude to review tracked health details and meal logs, then updated one repeat eating day after spotting a link to medication response.

Done with Claude· Intermediate· Jun 2026

What they did
They organized their setup into two Claude projects, one focused on food planning and the other on health. Rather than upload everything, they entered a few readings by hand from a Hume band, body-composition scales, and their medicine routine.
How it went
In conversation, Claude pointed out that they ate far more on the same day each week and that this seemed to line up with Mounjaro working less well a few days after the injection. They then rebuilt the meals for that day so they could handle a larger intake without side effects. The discussion did not include exact metrics, and other commenters noted that LLMs can turn odd blips into patterns if the context is thin.
Tips
Start with a small set of manual entries if you want a quick test. Use the agent to surface hypotheses, then verify them against your own experience and medical advice.

Try it yourself with Claude

I’m giving you my [date range] Garmin export plus my handwritten notes on [sleep, activity, heart rate, meals, medication, and symptoms]. Look for repeated timing or behavior patterns, especially anything that may be connected to how well [medication or symptom] is working. After that, draft a seven-day meal plan for [goal] using those patterns, and pause for my approval before you change, send, submit, or delete anything. Also list questions I can take to my doctor instead of making medical claims.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Read the original on reddit.com ↗

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.