Lipid levels and the risk of hemorrhagic stroke: A dose–response meta-analysis

Hemorrhagic stroke (HS) could damage human health and impose heavy social and economic burden around the world. An accumulating number of studies revealed the effect of lipid levels on HS, whereas the results were inconsistent. Therefore, we conducted a dose–response meta-analysis to evaluate the relationship between lipid levels and HS. We searched the databases for relative cohort studies, which were published before April 2020. We pooled adjusted effect size and performed the dose–response analysis by random-effect model. 31 eligible studies with 2,291,643 participants and 12,147 hemorrhagic stroke cases were included. An inverse association was observed between the risk of hemorrhagic stroke and total cholesterol (TC) (RR: 0.72; 95% CI: 0.64–0.82) or low-density lipoprotein cholesterol (LDL-C) (RR: 0.69; 95% CI: 0.53–0.89). Additionally, in dose–response analysis, the non-linear trend was also found between TC, high-density lipoprotein cholesterol (HDL-C), and risk of HS. When the level of TC and HDL-C was about 6 and 1.3 mmol/L separately, the risk of HS was decreased to the lowest. And we found a linear trend that for every 1 mmol/L triglyceride (TG) increase, the risk of HS decreased by 7%. TC and LDL-C were both inversely related to the risk of HS. In dose–response analysis of TG, we also found the inverse linear trend. Furthermore, the non-linear trend suggested the level of TC and HDL-C was about 6 and 1.3 mmol/L separately could lead to the lowest risk of HS. •Focused on the relationship between multiple types of lipids and the risk of HS.•We observed an inverse association between the risk of HS and TC or LDL-C.•In dose–response analysis, we found for every 1 mmol/L TG increase, the risk of HS decreased by 7%.•The non-linear trend suggested TC and HDL-C cholesterol was about 6 and 1.3 mmol/L separately could lead to the lowest risk of HS.