The prevalence of diabetes mellitus in the Netherlands. A quantitative review

Abstract

The prevalence of diabetes above the age of 30 years in The Netherlands was estimated and the influence of methodological variables on the reported prevalence rates was quantified. Fifteen Dutch studies performed since 1970 were entered in logistic regression models with the presence or absence of known diabetes above the age of 30 years as a dependent variable. In order to quantify the variability among studies, the study methodology and population characteristics were chosen as independent variables. The age-standardized prevalence of known diabetes varied between the studies. The total prevalence of known diabetes above the age of 30 years in 1993 can be estimated at 2.7% on the basis of the cases reported by general practitioners and at 3.2% based on self-reported cases in surveys. The prevalence according to age increased by 7.1% per life year for men and by 7.7% for women. These associations were essentially similar in all studies. Systematic screening, performed with an oral glucose tolerance test, revealed a prevalence which was 1.5–2 times higher.Depending on the method used, the number of subjects with known diabetes in the Netherlands in 1993 varied between 235,000 and 285,000. After systematic screening, this prevalence will be 1.5–2 times higher. These estimates may serve various goals. The prevalence rates based on self-reports or reports by general practitioners are important for estimating health care use. Prevalence rates based on screening are relevant for preventive strategies and latent health care needs.