Abstract
AIM: Extremely high high-density lipoprotein cholesterol (HDL-C) may increase cardiovascular death risk, but whether such non-linear patterns extend to other major causes of death is unclear. This study examined dose-response associations between HDL-C and mortality from the ten leading global causes of death.
METHODS: This study included 429,759 UK Biobank participants with baseline HDL-C data. Cause-specific deaths were obtained from national registries. HDL-C was modeled using Cox proportional hazards and Fine-Gray subdistribution hazard models. Restricted cubic splines assessed non-linear associations, stratified by sex.
RESULTS: Over a median follow-up of 13.8 years, 37,785 deaths occurred. U-shaped associations were observed between HDL-C and death risk from ischemic heart disease, lower respiratory infections, trachea, bronchus, or lung cancers, diabetes mellitus, and kidney disease. The optimal HDL-C range for the lowest death risk from above causes was 58-74 mg/dL in females and 50-60 mg/dL in males. J-shaped curves were observed for chronic obstructive pulmonary disease and liver disease, with the lowest death risk at 30-50 mg/dL. Stroke and Alzheimer’s disease/dementias death risk displayed sex-specific patterns: an L-shaped curve in females and U-shaped in males for stroke, and the reverse for Alzheimer’s disease/dementias. Extremely high HDL-C levels were associated with increased risk of death across several causes.
CONCLUSIONS: HDL-C is nonlinearly and sex-specifically associated with the top ten global causes of death. Both low and high HDL-C confer increased risk through different mechanisms. These findings highlight the importance of evaluating HDL functionality rather than just quantity in future research and clinical care.