This study illustrates the use of correspondence analysis to identify distinct profiles characterizing hormone replacement therapy (HRT) users and non-users, allowing the visualization of concurrent characteristics associated with this self-selection process. We analyzed cross-sectional information provided by 195 women reporting natural menopause, among 2,240 participants in the Pro-Saude Study (Rio de Janeiro, Brazil). Non-users had an unfavorable profile, including greater weight gain during adult life, older age, later menopause, and less schooling. This analytic tool should be used more extensively in public health research, in exploratory analyses of the relations among a large number of variables in key populations.
Hormone Replacement Therapy; Selection Bias; Longitudinal Studies
