I set out to test whether walking after a meal lowered the spike. It is one of the better-supported ideas in this field and I expected to confirm it.
I ended up with three walked-after meals. They showed an average rise of 34. The meals I didn't walk after showed 20.
On the face of it, walking made things worse.
It didn't. I designed the test badly. I walked after the meals I was already worried about — the heavy ones, the ones I could feel — so I compared my worst meals against my ordinary ones and called the difference an effect. Three meals, chosen after the fact, by the person being measured.
Done properly it needs matched pairs: the same meal, half the time with a walk and half without, decided in advance rather than in the moment. That is what I would do differently, and it is the part of the plan I got wrong.
The rest of what these sixteen days can't tell you:
It was one person. No control group, no second participant, no blinding. Nothing here is a study and I won't dress it up as one.
I knew what I was testing while I tested it, which changes how a person eats whether they intend it or not.
A sensor measures interstitial fluid, not blood, and lags real blood glucose by several minutes. It also drifts over a wear period, and I didn't cross-check with finger-sticks.
The meal-order result held, but modestly. Protein and vegetables before carbohydrate gave an average rise of 20 against 24 without, across 40 meals. Same direction as the published research, about a 4-point difference. I'd rather report it that size than round it up into something more impressive.
I'm writing this because the failed test is the one that tells you how much to trust the other two. If I only sent you the results that worked, you'd have no way to judge any of it.
If you want to read your own sensor graph without over-reading it, that is what the CGM reading guide is for.
One question. What would you want measured on yourself? Reply and tell me — the next test should answer a real question rather than one I guessed at.
— Alistair
PureNovus
Educational, not medical advice.