Background: The precise relationship between the lipid profile and mortality in elderly patients with type 2 diabetes mellitus (T2DM) remains unclear. The aim of this study was to investigate the relationship between the lipid profile over time, and mortality in elderly patients with T2DM.
Methods and findings: In 1998, 881 primary care patients with T2DM aged 60 years and older participated in the ZODIAC study, a prospective observational study. The cohort was divided into two age categories: 60-75 years and older than 75 years. Updated means of all lipid profile indices were calculated after a median follow-up time of 9.8 years. These values were used as time dependent covariates in a Cox proportional hazard model. The cholesterol-HDL ratio and LDL-cholesterol were positively related to both all-cause and cardiovascular mortality in the low age group. In contrast, except for the triglyceride level, none of the other lipid profile indices were related to all-cause mortality in patients aged over 75 years. The mortality risk decreased by 17% (95%CI: 5% to 27%) for each 1 mmol/L higher serum level of triglycerides. The relationships between the various lipid profile indices and cardiovascular mortality were not significant. However, the results were different after stratification for diabetes duration. In the subgroup of elderly patients with a diabetes duration of 8 years and longer, higher lipids were predictive of increased cardiovascular mortality. The main limitation of this study is its observational design, which prevents us drawing conclusions about causality.
Conclusion: Although the lipid profile was not predictive in the overall group of elderly patients, higher lipids were related to increased cardiovascular mortality in patients with diabetes of long duration. In order to make valid recommendations concerning lipid-lowering treatment, a randomized controlled trial or a meta-analysis concerning this specific population is mandatory.