Model development and apparent performance of model 2, model 3 and model 4 using multivariable logistic regression.

<p>Regression coefficients with 95% confidence intervals, c-statistic, and apparent performance measures of model 2, model 3 and model 4. Age was divided into three subgroups (shown as age, age’ and age”) using restricted cubic spline function to account for non-linearity. *The predictor history of cardio- and cerebrovascular disease is a combination of history of stroke, history of TIA and history of cardiovascular disease. **The predictor cardiovascular risk factors is a combination of diabetes, hypercholesterolaemia and hypertension. CI = confidence interval.</p> <p>(DOCX)</p>