We evaluated the models at both patient and hospital level by comparing the full model (model 4) including variables derived from the two data sources with more parsimonious models, following the procedure of Dimick et al. [ 20].
To address this concern with residual confounding, we added to the mixed-effect model (model 2) 23 indicator variables representing the effects of unmeasured activities that presumably affect the SDNN hour by hour.
In this model (Model 1), the variable a represents the concentration of the species undergoing polarization and whose spatial dynamics are of interest (Fig. 1B).
In the allocation model (Model 5) the variable is significant and has a positive coefficient.
In the final adjusted model (Model 4), neighbourhood-level variables were adjusted for all individual variables, including self-reported urinary infection.
The final model, Model 4, incorporated sociodemographic variables, aggravation by movement, and anxiety and depression.
The final model, Model 4, incorporated sociodemographic variables, short sleep time, insomnia, poor sleep quality, anxiety, and depression.
Notably, in the final model (Model 8) all independent variables indicate roughly the same power of association.
The final model, Model 4, included demographic variables, vascular risk factors and a measure of psychological distress.
To adjust data, we first used a full model (model 1) including all variables listed in Table 1 (except for total cholesterol, due to collinearity).
