The side effects prevalence profile of the sample was calculated as a percentage of participants who scored one or
more on the relevant LUNSERS items. Accumulated side effects scores were subsequently analysed by summing the
values of all the items (excluding those from the RH subscale). Sub-scores were calculated by totalling the values
of individual items on the subscales. Hormonal side effects were analysed separately for men and women because of
gender differences in hormonal items (e.g. item no. 50 ‘period problems’ only applies to women). For categorical variables, the chi-squared test for independence was performed. Prior to further analyses, data were screened for normality and homogeneity using the Shapiro–Wilk and Levene tests respectively.