Sarcopenia in Indian adults with type 2 diabetes: about one in four, with wide uncertainty

Design

Systematic review and meta-analysis of 7 cross-sectional studies

Population

1,359 Indian adults with type 2 diabetes

Primary outcome

Prevalence of sarcopenia by AWGS 2019 criteria

Effect

27.4% (95% CI 18.9% to 37.9%); I2 91.7%; prediction interval 6.9% to 65.6%

A systematic review and meta-analysis pooled seven cross-sectional Indian studies (1,359 adults with type 2 diabetes) that used Asian Working Group for Sarcopenia criteria. Confirmed sarcopenia was 27.4% (95% CI 18.9% to 37.9%), severe sarcopenia 11.0% and sarcopenic obesity 25.2%.

The headline needs its caveats. Heterogeneity was very high and the prediction interval for a new Indian clinic ran from about 7% to 66%. Studies that measured muscle mass by bioimpedance found more sarcopenia (31.7%) than those using DXA (20.2%). GRADE certainty was very low, and only two studies reported adjusted risk factors, which pointed loosely at hypertension and diabetes duration of 10 years or more.

The authors themselves say this is not enough to justify a screening programme. What it does support is noticing: muscle loss is common enough in Indian type 2 diabetes to deserve a question and a grip test in older patients, and it matters when choosing glucose-lowering drugs that cause weight loss.

• Ask older patients with long-standing type 2 diabetes about falls, difficulty rising from a chair or loss of strength.
• A handgrip measurement is a quick first check where muscle loss is suspected.
• Consider muscle mass, protein intake and resistance exercise when starting drugs that cause substantial weight loss.
• Bioimpedance gives higher estimates than DXA, so compare results only within the same method.

Why it matters

Weight loss is now a treatment goal in diabetes, and in a population this prone to muscle loss it matters what kind of weight is lost.

Don't overread it

Certainty is very low; the true prevalence in a given clinic could be anywhere from about 7% to 66%.

The statistics, in plain English

An I-squared of 91.7% means almost all the variation between studies reflects real differences in populations and methods, not chance. That is why the prediction interval, the range expected in a new setting, is so much wider than the confidence interval around the average.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.