Table 4 presents the results of a series of multiple logistic regressions that modeled the use of herbal medicine in this population. The first analysis included only the sociodemographic characteristics as independent variables. As can be seen in the table, female gender, being age 75 and older, living alone, being an immigrant, and reporting financial strain remained significant predictors of herbal medicine usage even when controlling for other sociodemographic factors. Table 4 also shows the results of individual logistic regressions
modeling herbal medicine use on each of the health-related measures, controlling for the influence of the sociodemographic variables. All variables shown in Table 3 were analyzed, but only those that remained significant predictors are shown in Table 4. Arthritis, hip fracture, urinary