Development and internal validation of a multivariable prediction model for 6-year risk of stroke: a cohort study in middle-aged and elderly Chinese population

Objective To develop and internally validate a prediction model for 6-year risk of stroke and its primary subtypes in middle-aged and elderly Chinese population. Design This is a retrospective cohort study from a prospectively collected database. Participants We included a total 3124 adults aged 45-80 years, free of stroke or myocardial infarction at baseline in the 2009-2015 cohort of China Health and Nutrition Survey. Primary and secondary outcome measures The outcome of the prediction model was stroke. Investigated predictors were: age, gender, body mass index (BMI), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), total cholesterol (TC), hypertension (HBP), drinking status, smoking status, diabetes and site. Stepwise multiple Cox regression was applied to identify independent predictors. A nomogram was constructed to predict 6-year risk of stroke based on the multiple analysis results. Bootstraps with 1000 resamples were applied to both C-index and calibration curve. Result The overall incidence of overall stroke was 2.98%. Age, gender, HBP and TC were found as significant risk predictors for overall stroke; age, gender, HBP and LDL-C were found as significant risk predictors for ischaemic stroke; age, gender, HBP, BMI and HDL-C were found as significant risk predictors for haemorrhagic stroke. The nomogram was constructed using significant variables included in the model, with a C-index of 0.74 (95% CI: 0.72 to 0.76), 0.74 (95% CI: 0.71 to 0.77), and 0.81 (95% CI: 0.78 to 0.84) for overall stroke, ischaemic stroke, and haemorrhagic stroke model, respectively. The calibration curves demonstrated the good agreements between predicted and observed 6-year risk probability. Conclusion Our nomogram could be convenient, easy to use and effective prognoses for predicting 6-year risk of stroke in middle-aged and elderly Chinese population.