Objective: To evaluate the impact of dyslipidaemia on cardiovascular disease (CVD) mortality in relation to fasting (FPG) and 2-h (2hPG) plasma glucose levels in individuals without a prior history of diabetes.
Methods: Data from 14 European population-based prospective studies of 9132 men and 8631 women aged 25-89 years were jointly analysed. A total of 871 CVD deaths occurred during the average 10 years of follow-up. Subjects were classified into normoglycaemia, isolated fasting hyperglycaemia (IFH, FPG> or =6.10 mmol/l and 2hPG<7.80 mmol/l), isolated post-load hyperglycaemia (IPH, FPG<6.10 mmol/l and 2hPG> or =7.80 mmol/l) and combined fasting and post-load hyperglycaemia (CH, FPG> or =6.10 mmol/l and 2hPG> or =7.80 mmol/l). Multivariate-adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) for CVD mortality were estimated using Cox proportional hazard analysis.
Results: Multivariate-adjusted HRs (95% CIs) for high-density lipoprotein cholesterol (HDL-C) were 0.84 (0.75-0.94), 0.66 (0.48-0.92), 1.03 (0.84-1.27) and 0.67 (0.51-0.89) in individuals with normoglycaemia, IFH, IPH and CH, respectively. For total cholesterol (TC) to HDL-C ratio they were 1.14 (1.03-1.27), 1.44 (1.13-1.84), 0.94 (0.77-1.15) and 1.26 (1.05-1.50), respectively. HRs for TC and triglycerides (TG) were not significant in most of the glucose categories except for TG in those with CH [HR 1.12 (1.00-1.27)].
Conclusions: Low HDL-C and high TC/HDL-C increase CVD mortality in either diabetic or non-diabetic individuals defined based on the fasting glucose criteria, but not the 2-h criteria. TG is a significant CVD risk predictor only in the presence of combined hyperglycaemia or diabetes. The difference between fasting and post-load hyperglycaemia with regard to the lipid-CVD relation may suggest a different pathophysiology underlying these two prediabetic states.