The Lancet

Logistic regression model was run with a binary outcome h (e.g., 1 if DF, 0 if Ctrl for the outcome comparison DF vs. Control) as the dependent variable and standardized GRS (i.e., GRS divided by its standard deviation), age and sex as the independent variables. The estimate of the beta coefficient and corresponding p-value of GRS from the model fitting provided the estimate of log-odds ratio ... ................
................