Purpose: This study evaluated the reliability of assessing transgender status in African populations using questions about current gender identification and sex assigned at birth. Methods: Data were obtained from the HIV Prevention Trials Network 075, a study designed to assess the feasibility of recruiting and retaining 400 men who have sex with men (MSM) in a 1-year prospective cohort study in Kenya, Malawi, and South Africa and conducted from 2015 to 2017. We compared responses of 401 participants to questions about gender identity obtained at the screening and enrollment visits and, for a subset of participants, in open interviews 6-9 months after enrollment. Results: One hundred and eleven of the 401 persons reported to identify as female or transgender at either the screening visit and/or the enrollment visit. Of those 111 persons, 42 (37.8%) switched between the two assessment moments from male to female or transgender, or vice versa. Furthermore, most transgender women were sexually attracted to men and identified as gay. In the follow-up interview with a subset of participants, almost all persons categorized as transgender identified as MSM. Conclusion: Categorizing persons as transgender based on self-identified gender and sex assigned at birth has limited reliability in African populations. Possible explanations for the observed inconsistencies in the responses to gender-related questions are discussed. For future studies, an understanding is needed of what specific words and labels such as gender, female, or transgender mean to study participants.
Keywords: Africa; assessment; men who have sex with men; transgender women.