[Association between high-density lipoprotein cholesterol level and cardiovascular disease and all-cause mortality in the elderly population].

To investigate the relationship between high density lipoprotein cholesterol (HDL-C) and cardiovascular disease (CVD) and all-cause mortality in the elderly population. A total of 14 355 elderly persons aged ≥65 years, who participated in the annual physical examination in Kailuan Group in 2006 were included in this prospective cohort study. According to HDL-C level, the participants were divided into 4 groups: low-level group (HDL-C<1.30 mmol/L), intermediate-level group (1.30 mmol/L ≤HDL-C≤1.54 mmol/L), medium-high-level group (1.55 mmol/L ≤HDL-C≤1.80 mmol/L), high-level group (HDL-C≥1.81 mmol/L). Baseline data such as age, sex and blood lipid levels were collected and compared. Inpatient medical records and death information were obtained through the social security system, and CVD and all-cause mortality were analyzed. After adjusting for confounding factors, the medium-high-level group was used as the reference group. Cox proportional risk regression model was used to evaluate the impact of HDL-C on CVD and all-cause mortality events. The linear or nonlinear relationship between HDL-C level and CVD and all-cause mortality events was evaluated by restricted cubic spline regression model. Death competitive risk analysis was conducted, and sensitivity analysis was performed after excluding subjects with CVD or all-cause mortality within 1 year of follow-up and female participants. The average age of this cohort was (71.5±5.5) years and follow-up time was (10.9±3.3) years. Compared with medium-high-level group, Cox proportional risk regression analysis showed that the (95%) of CVD and all-cause mortality in low-level group were 1.21 (1.06-1.38) (<0.05) and 1.02 (0.95-1.11) (>0.05), respectively; the (95%) of CVD events in high-level group was 1.17 (1.03-1.33) (<0.05), and there was a marginal significant association with all-cause mortality, the (95%) was 1.07 (1.00-1.16) (0.05<<0.1). The restricted cubic spline regression analysis showed that HDL-C was nonlinearly correlated with CVD (nonlinear correlation <0.1), and presented a U-shaped curve trend, while HDL-C was linearly correlated with all-cause mortality (nonlinear correlation >0.1). In the elderly population, the risk of CVD is lowest when the HDL-C level is 1.55-1.80 mmol/L, either high or low HDL-C is a risk factor for CVD. High HDL-C tends to be related to increased risk of all-cause mortality and low HDL-C is not related to increased risk of all-cause mortality.