Diet and exercise helped frail older veterans; adding metformin did not

Metformin is the standard first pill for type 2 diabetes, and animal studies have linked it to longer, healthier lives. A US trial in older veterans now tested it in people. Six months of weight-loss diet and exercise made frail older adults with obesity noticeably fitter. Adding metformin on top made no measurable difference to that main result. One caution up front: these are results the team posted on ClinicalTrials.gov, the public trial registry, not a peer-reviewed paper.

Extra weight makes the normal decline of muscle with age worse, and many older people slide into frailty: they become slow, weak and easily exhausted, and lose independence. Losing weight helps, but it has a catch. Part of what is lost is muscle and bone, which older people can least afford to lose. The researchers wanted to know whether metformin could add to the benefits of diet and exercise. Their registry entry says they hoped it would also help protect muscle and bone, partly because metformin is reported to reduce senescent cells, cells that have stopped dividing and release inflammatory substances.

What they did

The trial ran at the Michael E. DeBakey VA Medical Center in Houston, Texas, and was sponsored by the VA Office of Research and Development. It enrolled 114 people aged 65 to 85 (average age about 73) with a body mass index of 30 or more, mild to moderate frailty and no diabetes. Most were men: 101 of 114. People were randomly split into three groups of 38:

• Lifestyle plus metformin: a diet with 500 to 750 fewer calories a day, aiming at about 10% weight loss, plus supervised aerobic and strength exercise three times a week, plus daily metformin pills.
• Lifestyle plus placebo: the same diet and exercise, with dummy pills.
• Metformin alone: metformin plus a monthly group class on healthy living, but no diet or exercise program.

Participants, investigators and assessors did not know who got metformin and who got placebo; the diet and exercise part could not be hidden.

The main measure was the Physical Performance Test: nine everyday tasks such as walking 50 feet, putting on a coat and climbing stairs, scored from 0 to 36. To join, people had to score between 18 and 31.

What they found

After six months (average changes, as reported in the registry):

• Physical performance: up 3.8 points with lifestyle plus metformin, 3.6 with lifestyle plus placebo and 1.2 with metformin alone. Lifestyle plus metformin beat metformin alone (difference 2.5 points, p = 0.0002) but not lifestyle plus placebo (difference 0.1 points, p = 0.92).
• Strength and fitness followed the same pattern. Knee strength rose by about 11 and 10 units in the two lifestyle groups and did not change with metformin alone. Peak aerobic fitness rose by 3.6 and 3.5 units against 0.2. Gait speed improved only in the two lifestyle groups.
• Body composition: body fat fell by 8.3 kg and 6.8 kg in the two lifestyle groups and 2.7 kg with metformin alone. But lean mass (mostly muscle) also fell by 2.3 and 2.1 kg against 0.3 kg, and hip bone density dropped slightly. Metformin did not prevent this loss: the lifestyle groups with and without the drug were not clearly different.
• Other measures: lifestyle plus metformin did better than metformin alone on blood pressure, fasting insulin, triglycerides, two quality-of-life questionnaires and a cognitive test score. Against lifestyle plus placebo, none of these differed clearly.
• Where metformin did show up: compared with placebo, it lowered two blood markers of bone turnover. It also raised GDF-15, a blood protein the team counted among senescence-related signals: by about 40% with lifestyle plus metformin and about 30% with metformin alone, while it stayed flat with placebo. The registry gives no interpretation of this.

What to keep in mind

The team lists its own limits in the record: the trial lasted only six months, it may have been too small to detect differences in some secondary outcomes, and most participants were older men. All 114 people were included in the analyses, but 13 did not complete the study (2, 7 and 4 in the three groups). Planned measurements in muscle samples, including markers of senescent cells and telomere length, have not been posted.

Serious adverse events were reported in 4, 6 and 5 people in the three groups. Two people died, both in the lifestyle plus placebo group; the record does not give the cause. The most common milder events included diarrhea, COVID-19, dizziness and knee or shoulder pain. Metformin is a prescription drug. It often causes stomach and bowel complaints, and it is not given to people with poor kidney function, which is why such people were excluded from this trial.

Why it matters

Metformin is one of the drugs most often discussed as a possible "anti-aging" pill. In this trial, in frail older adults with obesity, it added nothing measurable to what diet and exercise achieved, and on its own it did much less. The clearer message is about the lifestyle program: supervised exercise and modest weight loss made people stronger and fitter within six months, though some muscle and bone were lost along the way. Keeping that muscle and bone remains an open problem, and this trial gives no sign that metformin solves it.