Combined BMI and NHHR assessment for stroke risk stratification in middle-aged and older Chinese: a nationwide cohort study

Body mass index (BMI) is an established risk factor for stroke but does not fully capture metabolic abnormalities associated with excess adiposity. The non-high-density lipoprotein cholesterol to high-density lipoprotein cholesterol ratio (NHHR) reflects atherogenic lipid imbalance. This study investigated the joint associations of BMI and NHHR with incident stroke and assessed the incremental value of NHHR beyond BMI for stroke risk stratification. We analyzed 9,101 stroke-free participants aged ≥ 45 years from the China Health and Retirement Longitudinal Study (CHARLS). BMI was categorized as <24.0 kg/m 2 , 24.0–27.9 kg/m 2 , and ≥28.0 kg/m 2 . NHHR was calculated as (total cholesterol−HDL-C)/HDL-C and categorized into quartiles. Incident stroke was identified by self-reported physician diagnosis during follow-up. Cox models estimated hazard ratios (HRs) and 95% confidence intervals (CIs). Joint associations were evaluated using 12 BMI–NHHR categories. Predictive performance for 4-year stroke risk was assessed using the time-dependent area under the receiver operating characteristic curve, continuous net reclassification improvement, integrated discrimination improvement, and decision curve analysis. During a mean follow-up of 6.3 years, 635 incident strokes occurred. BMI 24.0–27.9 kg/m 2 , BMI ≥ 28.0 kg/m 2 , and NHHR Q4 were each associated with higher stroke risk after multivariable adjustment (adjusted HRs, 1.24, 1.34, and 1.49, respectively). No statistically significant additive or multiplicative interaction was observed between BMI and NHHR. Joint-category estimates were heterogeneous rather than strictly monotonic. The BMI ≥ 28.0 kg/m 2 /NHHR Q4 group had the highest Kaplan–Meier-estimated 7-year stroke risk at 14.45%, compared with 5.85% in the BMI < 24.0 kg/m 2 /NHHR Q1 group. Adding NHHR to the covariate-adjusted BMI model produced little change in 4-year discrimination (AUC, 0.733 to 0.734) and did not significantly improve risk reclassification (continuous NRI = 0.107, 95% CI, −0.022 to 0.368; p = 0.096; IDI = 0.0005, 95% CI, −0.0003 to 0.0019; p = 0.156). Combined BMI–NHHR assessment identified heterogeneous stroke-risk profiles among middle-aged and older Chinese adults, although adding NHHR to BMI did not materially improve 4-year predictive performance. NHHR may provide complementary metabolic information for characterizing stroke risk, but its incremental predictive utility beyond BMI requires further validation.