DDeep Dive with Ali Abdaal
← All frameworks
Self-MasteryRupy Aujla, The Doctors Kitchen

The Personal Health Feedback Loop

Track symptoms, adjust one variable, and use feedback to improve health

Difficulty
Moderate
Time to result
~months to results
Steps
6
Confidence
97%

This framework creates a feedback loop between daily behaviour and a measurable health outcome. First, record when symptoms occur, how long they last, and what happened beforehand. Next, inspect possible inputs such as sleep, meals, caffeine, exercise, stress, and work patterns. Change small variables gradually rather than attempting a total lifestyle overhaul, then compare the frequency or severity of subsequent episodes. Rupy used this process while continuing cardiology follow-up, progressively changing breakfast, packed lunches, sleep habits, meditation, and movement. The output is not a universal medical conclusion but a clearer personal pattern that supports better choices and more informed clinical conversations. Its power comes from making long-term health behaviours feel immediate through visible feedback.

Origin

Rupy Aujla developed this approach after experiencing recurrent atrial fibrillation as a young doctor and tracking how gradual diet, sleep, stress, and movement changes related to his episodes.

Core principles

  • 01Measure what happens instead of relying on memory
  • 02Look for patterns between symptoms and preceding behaviours
  • 03Change manageable variables gradually
  • 04Treat personal data as evidence for discussion, not universal proof
  • 05Continue appropriate medical supervision

How to run it

  1. 1

    Define the observable outcome

    Choose a symptom or performance marker that can be recorded consistently, such as episode frequency, duration, sleep quality, or next-day energy.

    Pro tip Use the simplest measure you can record reliably.

    Watch out Do not invent a proxy for a condition that requires clinical measurement.

  2. 2

    Log every occurrence

    Record when the outcome occurs, how long it lasts, and its severity. Consistent records make trends easier to distinguish from isolated events.

    Pro tip Keep the log on your phone so recording takes less than a minute.

    Watch out Memory alone is vulnerable to recency and confirmation bias.

  3. 3

    Capture preceding inputs

    Note relevant behaviours before each occurrence, including meals, caffeine, exercise, sleep, stress, and working patterns.

    Pro tip Use the same short set of input fields every time.

    Watch out A repeated association does not automatically establish causation.

  4. 4

    Optimise gradually

    Change small, sustainable elements one at a time or in limited groups. Examples include replacing refined breakfasts, packing lunch, meditating, or moving bedtime earlier.

    Pro tip Start with the change requiring the least friction.

    Watch out Changing everything simultaneously makes it difficult to identify what mattered.

  5. 5

    Review the trend

    Compare the frequency and severity of outcomes over weeks and months. Keep helpful changes and reassess changes that produce no observable benefit.

    Pro tip Review on a fixed weekly or monthly schedule rather than reacting daily.

    Watch out Short symptom-free periods may not indicate lasting resolution.

  6. 6

    Maintain clinical oversight

    Share meaningful patterns with an appropriate clinician and continue recommended monitoring. Use the data to improve the conversation rather than replace it.

    Pro tip Bring a concise timeline instead of an unstructured diary.

    Watch out Never delay urgent care or independently discontinue prescribed treatment.

In the wild

Tracking atrial fibrillation episodes

Rupy recorded when each atrial fibrillation episode happened, how long it lasted, and possible preceding triggers. He gradually changed breakfast, packed lunches, stress management, sleep, and movement while continuing cardiology follow-up. Over time, episodes moved from roughly weekly to fortnightly, monthly, and then stopped occurring for extended periods.

The records gave him immediate feedback that lifestyle changes were associated with a substantial improvement in his condition.

Testing caffeine sensitivity

Rupy removed caffeine for 30 days while examining sleep data from his Oura Ring. His deep and REM sleep improved even though he had usually consumed caffeine before noon. After reintroducing coffee and noticing poorer sleep, he planned to drink it earlier.

A bounded experiment produced a personalised rule for caffeine timing.

Common mistakes

Changing everything overnight

A sweeping overhaul creates friction and obscures which change influenced the outcome. Gradual optimisation is easier to sustain and interpret.

Treating correlation as proof

Personal tracking can reveal useful associations but does not establish a universal treatment or eliminate alternative explanations.

Replacing medical care with tracking

The framework complements clinical assessment; it does not replace diagnosis, prescribed treatment, or emergency care.

Is it for you?

Best for

People with trackable symptoms or performance markers who want to test sustainable lifestyle changes alongside medical care.

Not ideal for

Medical emergencies, rapidly worsening symptoms, or anyone using self-experimentation as a substitute for professional treatment.

From the transcript

what I would be fastidious at would be tracking when I was having these AF episodes.

Rupy Aujla · 27:25

I'd also look at triggers before what was I doing that morning? Had I had worked out? Had I gone for a run had I…

Rupy Aujla · 27:25

after about a year of optimising things slowly, but slowly

Rupy Aujla · 27:25

From the episode

How To Embrace Vulnerability, Imposter Syndrome and Self-Doubt - Rupy Aujla, The Doctors Kitchen

Rupy Aujla, The Doctors Kitchen