In exploratory analyses, we used stratified models and also ran interaction p-values to examine potential effect modification by age in categories (4 to < 10, 10 to < 14, 14 to < 19 years), sex (male, female) and race/ethnicity (black, Hispanic, Asian, white, and other in addition to a missing category).
These extensions would more accurately model cofactor dynamics of parasitic conditions like schistosomiasis when interacting with the primary disease HIV, as has been done for the parasites alone in Lloyd-Smith et. al. (2008) [ 35], and for malaria and tuberculosis using stratified models [ 16, 17].
We considered potential effect modification by age, sex, or interval between symptom onset and diagnosis using stratified models.
Next we estimated effects of lead and hemoglobin on IQ according to genotype (three categories) using stratified models adjusted for the covariates above (model 2).
We examined effect modification by menopausal status using stratified models with each dietary variable dichotomized at the median.
Further research on the topic could investigate the results of including interactions terms or investigate the effect on using stratified models.
The time scale can be divided into several intervals using stratified models, with specific baseline hazards with a separate origin for each, defining a piecewise (e.g., piecewise Weibull) model.
In addition, we used stratified models to estimate associations between urinary cadmium and DNMT1 expression according to genotype in the 75 participants that were not first-degree relatives.
We also used stratified models to investigate the possibility of vulnerable subgroups.
In addition, we used stratified models to estimate heat wave effects by cause of death (cardiovascular causes, ICD-10 codes I00 I99; respiratory causes, ICD-10 codes J00 J99) and effect modification by individual characteristics such as sex, age (0 14, 15 64, 65 74, and ≥ 75 years), education level (none, 1 12, and > 12 years), and place of death (out of hospital or in hospital) for Seoul.
