Only the group that lifted weights kept its diabetes-risk gain a year after the diet ended

The uncomfortable fact about weight-loss research is that almost everybody regains. The interesting question is what is left behind when they do.

Writing in Obesity, researchers followed 221 sedentary women who had been randomized to Diet+Aerobic, Diet+Resistance, or Diet-only and put on an 800-calorie-a-day diet until their body mass index dropped below 25.

Mean BMI decreased from 28.3 to 23.9, with regain to 25.8 at 1 year. About a third of the loss came back, which is unremarkable. What happened to their risk estimates was not.

Two risk scores, two different fates

The team calculated predicted 10-year risk for two things at each stage: major adverse cardiovascular events and type 2 diabetes.

Both fell during the diet. Cardiovascular risk decreased from baseline to post-weight-loss and rebounded from post-weight-loss to 1 year in all groups, tracking the weight up and down as you would expect.

Diabetes risk did something different. It decreased in all groups but rebounded in Diet+Aerobic and Diet-only. The Diet+Resistance group maintained a significant net reduction at 1 year.

One arm out of three still had something to show for the diet a year later, and it was the arm that had been lifting.

Why muscle is the plausible explanation

Skeletal muscle is where most of the glucose you eat ends up. It is the largest disposal site in the body, and how much of it you have, and how well it responds to insulin, is a large part of what a diabetes-risk equation is measuring.

An 800-calorie diet takes lean tissue as well as fat. Regain, when it comes, arrives disproportionately as fat. Run that cycle without resistance training and a person can return to their old weight with less muscle than they started with, which is a worse metabolic position than the one they left.

Adding resistance work interrupts that. It does not stop the regain, and in this trial it did not: the resistance group put weight back on like everyone else. It changes the composition of what comes back, which is exactly the kind of effect that would show up in a diabetes estimate and not in a cardiovascular one.

Adherence supports the reading. Follow-up exercise adherence was associated with lower diabetes-risk rebound regardless of arm.

What a predicted risk is and is not

These are modeled numbers, and the article should be blunt about that.

Nobody in this trial had a heart attack that was counted, or was diagnosed with diabetes on the study record. Predicted 10-year risks were calculated using cardiometabolic disease staging equations, which convert measurements like waist, blood pressure and glucose into an estimate of what would happen over a decade.

That is a reasonable way to detect change in a one-year study of 221 people, because waiting for real events would take a trial twenty times the size and ten times as long. It is not the same as observing that fewer people got ill.

The oddity in the cardiovascular result

One finding runs against the grain and deserves flagging rather than burying: Diet-only achieved greater cardiovascular risk reduction than Diet+Resistance.

That is the opposite of the diabetes result, in the same trial, in the same women. It may be noise in a small secondary analysis. It may reflect that resistance training preserves lean mass, which keeps weight higher, and weight is an input to the cardiovascular equation in a way that penalizes muscle.

Either way, a study reporting one comparison in its favor and one against it is describing something more complicated than “lifting is better”, and this one does not have the size to resolve which.

What it means for anyone about to start a diet

More than 70 percent of adults in the United States are overweight or have obesity, and a great many of them will attempt this exact sequence: lose it fast, keep some of it off, watch the rest return.

What this analysis argues is that the exercise attached to the diet is not there to accelerate the loss. All three groups lost the same amount. It is there for what happens in year two, when the weight is back and the only thing that differs between two people at the same BMI is what that weight is made of.

Regular exercise is one of the best things you can do for your health, which is the sort of sentence that usually appears at the end of an article like this and means nothing. Here it has a specific version attached: the resistance half is the part that was still doing something a year after the diet stopped.