Abstract
BACKGROUND: While prior meta-analyses have demonstrated the benefits of lipid-lowering agents for atherosclerotic cardiovascular disease prevention in high-risk populations with definite risk, the applicability of these findings to global community-based populations, where hemorrhagic risk profiles and optimal serum lipid levels might differ, remains less defined. This gap underscores the need to clarify the relationship between serum lipids and hemorrhagic stroke (HS) risk in a broader community-based population. METHODS: PubMed, Ovid, and the Cochrane Library were searched from inception to May 21, 2025, for observational prospective cohorts in populations with no major disease, baseline serum lipid assessment, and first-attack HS. Risk ratios (RRs) with 95% confidence intervals (95% CIs) of continuous variables and dose-response modeling were both analyzed. Potential nonlinear trends were validated through model comparison. RESULTS: A total of 50,244,342 participants from 54 cohorts were included. Continuous variable analysis revealed that serum total cholesterol (sTC) was inversely related to total HS (RR = 0.94, 95% CI 0.90 ~ 0.97, P < 0.01) and intracerebral hemorrhage (ICH) (RR = 0.90, 95% CI 0.87 ~ 0.93, P < 0.01) risk; serum low-density lipoprotein cholesterol (sLDL-C) was inversely related to total HS (RR = 0.92, 95% CI 0.86 ~ 0.98, P = 0.012) and ICH (RR = 0.83, 95% CI 0.74 ~ 0.94, P < 0.01) risk; and serum high-density lipoprotein cholesterol (sHDL-C) was inversely related to subarachnoid hemorrhage (SAH) (RR = 0.77, 95% CI 0.63 ~ 0.92, P < 0.01) risk. In the dose-response analysis, linear trends were detected in total HS, with a decreasing risk of 7.8% (95% CI 3.6%~11.8%, P < 0.01) per mmol/L sTC increase and a decreasing risk of 9.6% (95% CI 2.4%~16.2%, P < 0.01) per mmol/L sLDL-C increase; in ICH, with a decreasing risk of 17.0% (95% CI 7.5%~25.6%, P < 0.01) per mmol/L sLDL-C increase; and in SAH, with a decreasing risk of 20.7% (95% CI 1.5%~36.1%, P = 0.036) per mmol/L sHDL-C increase and a decreasing risk of 54.0% (95% CI 5.5%~77.6%, P = 0.035) per mmol/L triglyceride increase. U-shaped trends were detected between sTC and ICH risk, with the lowest RR at 6.07 mmol/L; between sLDL-C and SAH risk, with a minimum RR of 3.69 mmol/L; between sHDL-C and total HS risk, with a minimum RR of 1.53 mmol/L; and between sHDL-C and ICH risk, with a minimum RR of 1.50 mmol/L. CONCLUSION: When the minimal important difference was set at an RR of 0.87 ~ 1.15, sTC between 3.48 ~ 5.20 mmol/L, sLDL-C between 1.09 ~ 3.40 mmol/L, and sHDL-C between 0.99 ~ 2.34 mmol/L might provide balanced lipid management, contributing to stroke prevention in global community-based populations. TRIAL REGISTRATION: The protocol of this study was registered on PROSPERO (CRD420250650371). SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12944-025-02698-0.