The association between blood glucose value and long-term mortality.

This study aims at estimating the association between different fasting blood glucose levels (FBG) and total mortality during a long-term follow-up. In all 2,300 subjects were health examined, out of a stratified sample of 32,185 individuals aged 18-64 years drawn from the population in Stockholm County from the years 1969-70. FBG values were divided into following groups:<3.0, 3.0-4.4, 4.5-5.5, 5.66.0, 6.1-6.6, and > 6.6 mmol/l (corresponding to fasting plasma glucose, FPG,<3.5, 3.5-4.9, 5.0-6.0, 6.1-6.9, 7.0-7.7 and > 7.7 mmol/l), and known diabetes mellitus. All participants were followed up in the National Cause of Death Register up to the end of 1996. Multivariate analysis was performed by Cox regression, with three models, the first age- and sex-adjusted, the second also adjusted for care need category and hypertension, and the third with added BMI-category, with hazard ratios (HR) and 95% confidence interval (CI). Smoking habits were available for around half of the sample. Compared to the FBG showing the lowest mortality, i.e. FBG 5.6-6.0 mmol/l, we found an age- and sex-adjusted excess risk for subjects with known diabetes (HR 7.39, 95% CI 3.78-14.45), with FBG > 6.6 mmol/l (HR 2.30, 95% CI 1.20-4.39), and with FBG<3.0 mmol/l (HR 3.44, 95% CI 1.47-8.06). The excess risk persisted when adjusting for care need, hypertension, BMI, and also for smoking. The cause of the increased mortality risk with low FBG values is unclear, but low FBG value seems to be a risk marker of poor health.