Prevalence and risk factors associated with prediabetes and undiagnosed diabetes in France: The national CONSTANCES cohort

Aims

To assess the prevalence of prediabetes and diabetes in France between 2013 and 2014 using data from the CONSTANCES cohort, and to identify factors associated with prediabetes and undiagnosed diabetes.

Methods

The study population comprised participants recruited in 2013–2014 in CONSTANCES, an ongoing French national prospective cohort following participants aged 18–69 years who are covered by France's general health insurance scheme. Participants completed a questionnaire at baseline and underwent a medical examination which included providing blood samples. Undiagnosed diabetes was defined as a fasting plasma glucose (FPG) ≥ 7 mmol/l and diagnosed diabetes as self-report or identification of reimbursements for anti-diabetics. Prediabetes was defined as a FPG ≥ 6 mmol/l but < 7 mmol/l.

Results

25,137 participants were included in the analyses. The overall prevalence of prediabetes was 7.2% [95% confidence interval: 6.7–7.7], 1.6% [1.4–1.9] for undiagnosed diabetes, and 4.0% [3.6–4.4] for diagnosed diabetes. These rates were significantly higher in men, in older persons, in persons with obesity, and in those with lower education levels. In multivariate regression models, excessive corpulence was the variable most strongly associated with undiagnosed diabetes (adjusted Odds Ratio=9.31) and prediabetes (aOR=3.85). Additionally, male sex, older age, family history of diabetes, at-risk alcohol use, and lower education level were all positively associated with undiagnosed diabetes and prediabetes.

Conclusion

Diabetes and prediabetes prevention together with screening for undiagnosed diabetes must be strengthened for persons with low socioeconomic status and for those with obesity or overweight.