Introduction Relationship-based sexual behaviours are important elements in sexual health approaches to STI prevention.

Method STI-clinic recruited men and women (n=423, 18–30 years.) contributed quarterly sexual health information and weekly urine-based STI specimens over 12 weeks. Sexual health was a standardised, 10-item index of partner-associated measures (relationship quality, sexual satisfaction, control over sexual activity, shared social and sexual decision making, condom use efficacy, emotional commitment, partner meeting needs, knowing partner well, frequency of seeing partner, perceived STI chances). STI- behaviours were: condom use during vaginal sex, condom use during anal sex, given money for sex (no/yes) and partner forced sex (no/yes). STI was NAAT diagnosis of chlamydia (CT), gonorrhoea (GC) and trichomonas (TV). Analyses were mixed-effect logistic or poisson regression (Stata, v.13); models were stratified by gender and monogamy, controlling for age, race and partner age difference, and number of partners and condom use (STI models).

Results Sexual health was associated with more frequent condom use during vaginal sex for men and women (IRR=1.92–2.18) and for monogamous and non-monogamous (IRR=1.72–3.01) relationships, with lower odds of men’s or women’s (OR=0.16–0.19) being paid for sex, with lower odds of women’s (OR=0.09) and monogamous and non-monogamous relationships’ (OR=0.24–0.15) reporting forced sex. Controlling for total partners and condom use, sexual health lowered men’s and women’ odds of CT (OR=0.38–0.43), men’s (OR=0.40) and non-monogamous relationships’ odds of GC (OR=0.41), and men’s odds of TV (OR=0.33).

Conclusion Among men and women, and even in non-monogamous partnerships, a relationship-based measure of sexual health is linked to fewer STI risk behaviours and lower STI incidence. Our data are empirical evidence of the importance of sexual health in relationships as a means of STI control.