Table 3

Direct and indirect (mediated) effects of PrEP use on any recent bacterial STI diagnosis via CAS, oral sex (OS) and number of anal sex partners (NSPs); baseline data, Engage cohort study (n=2009)

PredictorOutcomeβ95% CIP value
Direct effect
PrEP useSTI0.04−0.12 to 0.210.62
OSSTI0.500.21 to 0.790.001
CASSTI0.220.07 to 0.370.003
PrEP useOS0.06−0.17 to 0.300.59
NSPsOS0.001−0.15 to 0.150.98
PrEP useCAS0.120.01 to 0.230.03
NSPsCAS0.530.47 to 0.59<0.001
PrEP useNSPs0.300.24 to 0.35<0.001
Indirect effect
PrEP use–OSSTI0.033−0.08 to 0.140.57
PrEP use–CASSTI0.027−0.01 to 0.060.11
PrEP use–NSPs–OSSTI0.000−0.02 to 0.020.98
PrEP use–NSPs–CASSTI0.0350.01 to 0.060.003
Total effect0.1370.03 to −0.240.01
  • The estimated coefficients are standardised coefficients.

  • CAS, condomless anal sex; PrEP, pre-exposure prophylaxis.