Worldwide trends in diabetes prevalence and treatment from 1990 to 2022: a pooled analysis of 1108 population-representative studies with 141 million participants.
Level 4 - case-series / case-control
Pooled epidemiological analysis of cross-sectional population surveys using meta-regression modeling
PubMed 39549716 · doi:10.1016/S0140-6736(24)02317-1
What was done
The authors pooled data from 1108 population-representative studies covering 141 million participants aged 18 years and older across 200 countries and territories from 1990 to 2022. Diabetes was defined as fasting plasma glucose of 7.0 mmol/L or higher, glycated hemoglobin (HbA1c) of 6.5% or higher, or current use of diabetes medication. Treatment coverage was defined as the proportion of individuals with diabetes who reported taking medication. Trends were estimated using a Bayesian hierarchical meta-regression model.
What was found
In 2022, an estimated 828 million adults (95% credible interval [CrI] 757 to 908) had diabetes, representing an increase of 630 million (554 to 713) since 1990. Age-standardized prevalence increased in 131 countries for women and 155 countries for men (posterior probability > 0.80), with the largest rises occurring in low- and middle-income nations in south Asia, southeast Asia, the Middle East and north Africa, and Latin America. In 2022, 445 million adults aged 30 years or older (59% of that demographic with diabetes; 95% CrI 401 to 496) were untreated, a 3.5-fold increase compared to 1990. Treatment coverage remained below 10% in some African countries, while exceeding 55% in South Korea and several high-income western nations.
Why it matters
This study provides the most comprehensive worldwide assessment of diabetes burden to date, highlighting a widening global equity gap where surging prevalence in low- and middle-income countries is met with persistently low treatment coverage.
Limits
The findings rely on hierarchical modeling to estimate values for country-years with missing or sparse survey data, particularly in low-income regions. Treatment was defined strictly as taking medication, without assessing adherence, specific drug classes, lifestyle interventions, or achievement of glycemic control targets. The dataset does not distinguish between type 1 and type 2 diabetes.
Cited by
- contradicts The United States ranks around 70th in the world for diabetes prevalence, while Singapore ranks around 9th and Middle Eastern countries occupy ranks 1 through 8.