Practiceby Lab 10 Get started ›
Crossfire

Is there a test that tells me whether low-carb is right for me?

What the experts say

Benjamin T. BikmanPhD · Professor of Cell Biology and Physiology, Brigham Young University; director of the Laboratory of Obesity and Metabolism

Yes, measure your insulin

What it rests onHis model makes insulin the variable that matters, so measuring it is the diagnostic that follows. He also recommends the triglyceride-to-HDL ratio off an ordinary lipid panel, with bands at 2, 3 and roughly 3.5, as a marker of insulin resistance almost everyone already has the numbers for.

Source: benbikman.com ↗
Christopher D. GardnerPhD · nutrition scientist, Stanford Prevention Research Center

No, and the specific test you are thinking of was tested and failed

What it rests onDIETFITS randomised 609 adults for 12 months with both arms built to be good versions of themselves. Weight change was 5.3 kg on healthy low-fat and 6.0 kg on healthy low-carbohydrate, not significantly different. The trial's stated purpose was to test these two predictors: genotype pattern P = .20, baseline insulin secretion measured as INS-30 P = .47. NCT01826591.

Source: pubmed.ncbi.nlm.nih.gov ↗

Overview

Gardner's participants were overweight adults without diabetes, randomised, so nobody chose their arm. Bikman is addressing people who suspect they are already insulin resistant and want to know before a diagnosis arrives, which is a population DIETFITS excluded and a question DIETFITS did not ask.

The trial tested whether insulin secretion predicts which diet loses more weight; it did not test whether a fasting insulin result is useful information about your metabolic health. Those are two different questions and it is easy to run them together.

Where they agree

Both accept that a substantial minority of people do very well on low-carbohydrate eating, and Gardner says so explicitly. Both accept that insulin resistance is real, matters, and is worth knowing about. Both would tell the same user to cut refined carbohydrate. The dispute is narrow and expensive: whether a number should decide it.

What would settle it

A trial randomising by measured fasting insulin rather than by insulin secretion after a glucose load, which is the version of the test Bikman actually recommends and not the one DIETFITS measured. That distinction is the strongest thing his side has, and it has not been tested.

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.