Mauritius: OGTT surveys show type 2 diabetes prevalence falling after 2009

Design

Seven repeated national cross-sectional surveys, 1987 to 2021, OGTT-based

Population

36,744 survey attendances by adults aged 25 to 74 in Mauritius

Primary outcome

Age- and sex-standardised prevalence of type 2 diabetes and intermediate hyperglycaemia

Effect

Diabetes 23.3% (2009) to 19.9% (2021); intermediate hyperglycaemia 24.9% to 15.6%

Mauritius has run seven standardised population surveys since 1987, each with more than 85% participation and an oral glucose tolerance test for anyone not already on glucose-lowering drugs. Across 36,744 attendances by adults aged 25 to 74, age- and sex-standardised type 2 diabetes prevalence rose from 14.7% in 1987 to a peak of 23.3% in 2009, then fell to 19.9% in 2021.

Intermediate hyperglycaemia followed the same arc, peaking at 24.9% and falling to 15.6%. The share of diabetes that was newly found at survey dropped from 67.0% to 32.5%, so far more people now know their diagnosis. The fall held across sexes, age groups and ethnic groups.

The island's population is largely of South Asian ancestry, which makes this more relevant to Indian practice than most Western trend data. It suggests that sustained national prevention and screening can turn the curve, although a repeated cross-sectional design cannot say which measures did it.

It also argues for the OGTT where it is feasible: fasting glucose or HbA1c alone miss a share of people with impaired glucose tolerance.

• Population diabetes prevalence can fall with sustained national effort, even in a high-risk South Asian population
• Undiagnosed diabetes halved as a share of cases over 34 years, a measurable target for screening programmes
• Consider an OGTT when fasting glucose or HbA1c is borderline and the result would change management
• Intermediate hyperglycaemia fell faster than diabetes, so the next generation of cases may be smaller

Why it matters

It challenges the assumption that diabetes prevalence in South Asian populations can only keep rising.

Don't overread it

The design cannot show which policy, if any, caused the decline.

The statistics, in plain English

The 2009 and 2021 confidence intervals (22.4 to 24.3% and 18.6 to 21.1%) do not overlap, so the fall is unlikely to be chance. Repeated cross-sectional surveys sample different people each time, so they show population trends, not what happened to individuals.

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.