In the development of the prognostic models, age at diagnosis (number of years), radiological stage (stage IA-IV) and treatment (curative vs palliative) were included because these are strong predictors of survival [26].
The following set of variables were included in all models: age at enrollment, race, gender, age at first rubella vaccination, age at second rubella vaccination, and cohort status (first versus second).
In the final models, age at diagnosis was treated as a continuous variable and adjustments were included for stage, decade of diagnosis, and geographic region.
In logistic regression models and Cox models, age at diagnosis was considered in its original scale, and after a cubic spline transformation, only results in the original scale were reported since both gave similar results.
In the analyses, the following factors were considered in the multiplicative models: age at cancer diagnosis, calendar time at diagnosis, the cancer site, time since diagnosis, and stage at diagnosis.
The following covariates were adjusted in the multivariate models: age at diagnosis, education, income, body mass index (BMI), radiotherapy, chemotherapy, immunotherapy and tamoxifen use, TNM stage, histologic grade, and tumor size.
In these models, age, age at menarche, age at menopause, weight, waist-to-hip ratio and birth weight were entered as continuous variables.
We model age at migration as both a continuous and dichotomous explanatory variable.
We made novel use of parametric time-to-event analysis to model age at initial use of assistive device for ambulation.
Therefore, results were presented separately by attained follow-up duration (≤5 vs >5 years), and adjusted for sex (as covariate in the regression model), age at entry, and area of residence (as stratification variables).
Under this model, age at first exposure was inversely related with mesothelioma risk.
