at baseline, 12, 24, and 48 weeks. We used generalized estimating equations to describe sexual behavior over 48 weeks for each study arm, and Cox proportional hazards to compare STI risk between arms. Results After randomization, 209 HIV-positive participants completed at least 12 weeks of study. The number of anal sex partners (in the prior 3 months) declined in both arms (Immediate: 12-week IRR=0.66; 95%CI=0.45-0.97, 48-week IRR=0.75; 95%CI=0.51-1.12; Deferred: 12-week IRR=0.68;

more »

... 8, 48-week IRR=0.56; 95%CI=0.40-0.79). The proportion reporting >50% condom use with main partners (in the prior 30 days) doubled among Deferred participants (12-week RR=2.04; 95%CI=1.40-2.95, 48-week RR=1.97; 95%CI=1.35-2.87) and increased by 75% among Immediate participants (12-week RR=1.75; 95%CI=1.31-2.35, 48-week RR=1.74; 95%CI=1.26-2.41). Condom use with casual partners increased in both arms (Immediate: 12week RR=1.37; 95%CI=1.16-1.61, 48-week RR=1.23; 95% CI=1.02-1.49; Deferred: 12-week RR=1.47; 95%CI=1.20-1.79, 48-week RR=1.25; 95%CI=0.99-1.58). Relative to Immediate participants, Deferred participants had higher risk of chlamydia (Hazard Ratio=1.85; 95%CI=1.09-3.15), with a trend toward higher risk of gonorrhea (HR=1.62; 95% CI=0.88-2.97), and syphilis (HR=2.05; 95%CI=0.82-5.16). Conclusion Despite reporting protective behavior at levels similar to or slightly higher than those of participants who started ART immediately, participants who deferred ART initiation had increased risk of bacterial STIs. This warrants further investigation. Disclosure No significant relationships.