Both groups ate whole grains. In 160 adults with diabetes, the less-milled ones finished 0.2 percentage points lower on HbA1c

Wholemeal bread and intact grain kernels can both be called whole grain, even though the body does not meet them in quite the same form. A randomised trial from New Zealand, published in The American Journal of Clinical Nutrition on 26 October, asked whether that physical difference matters to people already living with type 2 diabetes.

The researchers randomly assigned 160 adults to 12 weeks of either less-processed or more-processed whole-grain foods. Both packages contained four foods made from whole-grain wheat, rice and oats. Enough was delivered for the household — 4.5 to 9 kilograms every two weeks, depending on family size — and participants were asked to swap these for the grain foods they ordinarily ate.

The average participant was 58.5 years old, had a BMI of 33.7 and began with an HbA1c of 56.4 mmol/mol, about 7.3% by standard conversion. At the end, 143 people, or 89%, provided blood samples. HbA1c was 2.2 mmol/mol lower in the less-processed group than in the milled-whole-grain group. That is about 0.2 percentage points by our conversion; the 95% confidence interval ran from 4.2 lower to 0.1 lower, so its least optimistic edge was close to no difference.

This was not simply a whole-grain-versus-refined-grain result. Reported whole-grain, refined-grain and nutrient intakes did not differ between groups, and the supplied foods provided roughly 20% of dietary energy. Body weight did not differ either. That makes the structure of the grains a plausible explanation for the glucose result, although the abstract does not report how faithfully each food was eaten or give meal-by-meal glucose measurements.

Of the secondary and exploratory outcomes, only self-reported bowel habits showed an indicated improvement with the less-processed grains. That restraint matters: the trial did not find a broad sweep of metabolic changes, and 17 randomised participants did not provide the final blood sample. It lasted three months, so it cannot tell us whether the small HbA1c separation persists or changes complications over years.

The practical distinction is between a grain that largely keeps its physical structure and one ground into flour, even when the flour still contains bran and germ. Steel-cut oats, wheat berries and brown rice are examples closer to the first end; wholemeal bread sits closer to the second. This trial supports paying attention to that distinction, not treating intact grains as a replacement for diabetes medicines or suggesting that every product labelled whole grain has the same effect.