In addition, for any significant base-model adjusted nutrient survival associations that were observed, we did a check for other potential confounders (e.g. income, comorbid conditions, protein intake), although none were found to appreciably change hazard ratio (HR) estimates.
Adjustment for risk factors did not significantly change the hazard ratio (hazard ratio, 1.15; 95% CI 0.77 1.72).
Covariables that changed the hazard ratio (HR) of cancer risk by more than 10% or were considered clinically relevant were included in the model.
Factors with a P value of <0.10 or those that changed the hazard ratio for TST or ELISPOT response by greater than 10% were considered for the multivariable analysis.
Cofactors that changed the hazard ratio of a fall incident according to the different echocardiographic abnormalities by more than 5% or that were biologically plausible, and drug withdrawal as defined above, were maintained in the final model.
The confounding factors that changed the hazard rate ratios most were included in the final model.
None of the factors considered, however, changed the hazard ratios by ≥ 15%.
Neither timing nor duration or dosage of statin use changed the hazard of death significantly.
Adding the neighborhood SES or combined neighborhood SES/enclave variable to these models changed the hazard ratio by 3.6% or less.
