They swapped 150 g of staple food for walnuts. After 12 weeks, the waist-and-fat numbers looked better — but the two groups still did not differ at the finish

Take 150 grams of staple food out of a calorie-cutting eating plan and replace it with walnuts: that was the change tested in a small Chinese trial published in October. After 12 weeks, people in the walnut-based, lower-carbohydrate group had lower average BMI, less estimated visceral fat and a lower trunk-to-leg fat ratio than at the start. It sounds like a straightforward walnut result. The direct comparison at the end tells a more cautious story.

Eighty adults with central obesity were randomly assigned to one of two active diets. Both were designed to reduce daily energy intake, included vegetables, fruit, protein foods and a salt target, and came with five follow-ups. The comparison diet kept 100 grams of staple food at each meal. The walnut plan replaced 50 grams at each meal with walnuts — 150 grams per day. The researchers provided the nuts free only for the first two weeks; after that, participants paid for them. By the end, 35 people in the walnut group and 29 in the comparison group remained for the main analysis.

Both groups were eating far less by week 12. Reported energy intake fell from 1,746 to 1,162 calories per day in the walnut group, and from 1,892 to 1,211 in the comparison group. The bigger dietary separation was fat: it rose slightly in the walnut group, from 52 to 55 grams a day, while falling from 55 to 31 grams in the comparison group. Carbohydrate ended at 112 grams versus 175 grams a day. That means this was not walnuts added to an otherwise unchanged everyday diet. It was one part of two fairly intensive, low-energy programmes.

The walnut group’s estimated visceral-fat area moved from 113 to 97 square centimetres, and its average BMI from 27.6 to 26.1. The comparison group also changed, from 116 to 109 square centimetres and from 27.3 to 27.0. The authors’ repeated-measures model found that the pattern of change over time differed between groups. But when they compared the groups at the 12-week endpoint, the p values were 0.132 for visceral-fat area and 0.180 for BMI — not statistically significant by the study’s own conventional threshold. A faster-looking trend is not the same as a settled endpoint difference.

The dropouts make that distinction more important. Five of 40 people left the walnut group, compared with 11 of 40 in the comparison group. Four comparison participants were lost to follow-up, two left on their own, one became pregnant and one had poor adherence. The researchers also ran an intention-to-treat analysis that kept all 80 people in their original groups, using each person’s last measurement when an endpoint was missing. They say the broad findings were similar, but the complete-case analysis is still the clearer source for the eye-catching before-and-after numbers.

There are other limits. The study lasted 12 weeks, involved adults aged 18 to 60 in one Chinese setting, and used a bioelectrical-impedance device rather than a scan to estimate visceral fat. Food intake came from three-day records. The diets were not made equal in calories, and the authors list that as a limitation. The nuts were also an expense for most of the trial. So this is useful evidence that a walnut-based, lower-carbohydrate plan can be tested and may shift body-composition measures alongside calorie restriction. It does not show that walnuts alone melt abdominal fat, or that everyone should swap 150 grams of staple food for them. For someone considering a major diet change, the more durable question is whether the plan is safe, affordable and workable with their clinician or dietitian.