Dependent variable. Our dependent variable was PrEP uptake (occasionally and daily PrEP uptake) during anal intercourse with a casual partner in the previous 6 months.
Mediator. The mediation variable was HL, which was evaluated by using the HLQ—the “ability to actively engage with health care providers” scale (Debussche et al., 2018 (link); Osborne et al., 2013 (link)). The score of the scale is the average score across all five items and ranges from 1 (low HL level) to 5 (high HL level). We then categorized participants into two categories: poor (HL score ≤3 1st quartile) and adequate (HL score >3 1st quartile) HL.
Independent variables. Independent variables were measured by SES, which, in turn, was separately evaluated through education level (higher than upper secondary school certificate (high hereafter) versus upper secondary school certificate or lower (low hereafter) and perceived financial situation (Ousseine et al., 2020 (link)). The latter was evaluated with the question “Would you say that financially…?” (answer options: you are comfortable; you get by; you have to be careful; you find it hard to make ends meet; and you can't makes ends meet without going into debt). These five response categories were merged to form a binary variable (comfortable and you get by versus all other responses).
Covariates. Covariates were sociodemographic characteristics (age, place of birth, place of residence, in a stable relationship with a man) and social support. The latter was defined as having LGBTQ+ friends, and frequenting LGBTQ+ community social venues (sauna, bars) or websites.