To assess the impact of missing covariates, we applied multiple imputation for missing BMI, education, cigarette smoking status, exercise, walking, histories of hypertension and diabetes data using PROC MI and PROC MIANALYSE. All analyses were performed using SAS 9.4, and we determined when a two-tailed p -value was less than 0.05 as the statistical significance and 0.05-<0.10 as borderline significance.
← all excerpts
Alcohol Consumption and Long-Term Mortality in Men with or without a History of Myocardial Infarction.
1
—
—