✓ Practiceby Lab 10 Get started ›
Crossfire

Is it worth tracking my own blood sugar responses to decide what to eat?

What the experts say

Kevin D. HallPhD · metabolism researcher; formerly Section Chief, Integrative Physiology Section, NIDDK, National Institutes of Health

Not from single readings, and not as your only measure

What it rests onInpatient study of 30 adults without diabetes capturing 1,189 CGM responses to duplicate presented meals approximately one week apart across four dietary patterns, using two device types. Weak-to-moderate correlations between duplicate-meal responses, low within-participant reliability, wide limits of agreement, and variability to a repeated meal comparable to variability across different meals.

Source: pubmed.ncbi.nlm.nih.gov ↗
Tim SpectorProfessor of Genetic Epidemiology, King's College London · founder of TwinsUK

Yes, as part of a personalised program

What it rests onPREDICT 1 found severalfold differences between people in their responses to identical meals; the METHOD randomised trial of 347 adults compared an 18-week app-based personalised program against USDA-guideline standard care and found greater triglyceride, weight and waist reduction, and no significant difference in LDL-cholesterol.

Source: pmc.ncbi.nlm.nih.gov ↗

Overview

Neither side is talking about people with diabetes, where continuous glucose monitoring has an established clinical job that nothing here touches. Hall's participants ate duplicate presented meals in a metabolic ward, which is the strictest possible test of reliability and the least like real life.

Spector's participants ate their own food at home with coaching, which is the most realistic setting and the one where personalisation and support cannot be separated.

Where they agree

Both agree that people differ severalfold in their glucose responses to the same meal. That is not the dispute. The dispute is whether one person's response to one food is stable enough over time to act on, which is a reliability question rather than a variability question, and the two are constantly conflated in the marketing on both sides.

What would settle it

A repeat-measures reliability study of the same foods in the same free-living people over weeks. Hall's own paper names the remedy rather than rejecting the idea: aggregated repeated measurements. If a program demonstrated that its food scores are stable on re-test, his objection would be met.

Practice does not pick a winner here

You do, and your call is dated, kept, and yours to change. Follow the people you already trust and what they publish turns into a short daily practice.

Get started

Free to sign up and use. We do not ask for a card.