Background We aimed to determine (i) if other sexually transmitted infections (STIs) increase the risk of incident human papillomavirus (HPV) infection and (ii) if HPV infection predicts the incidence of other STIs.

Methods Women aged 20–38 years were followed semi-annually for 18 months in Thailand (n=1200). Assessment was made on cervical HPV genotypes, cervical cytology, sexual behaviour, demographic factors and diagnoses of other STIs including chlamydia, gonorrhoea, syphilis, genital herpes and trichomoniasis. Incident detection was defined as any type-specific HPV or other STI which was detected at current visit but not at previous visit. Associations were measured by ORs with 95% CIs estimated in generalised estimating equation models.

Results During follow-up, 241 new cases of HPV, 110 incident cases of high risk (HR)-HPV and 46 new cases of other STIs were observed. Diagnosis of other STIs at previous visit was statistically significantly associated with twofold increased odds of any new HPV detection after controlling for sexual behaviour, age, pap smear status and contraceptive use [adjusted OR (aOR): any HPV: 2.16 (95% CI: 1.08% to 4.34%)] (Abstract O1-S02.02 table 1). No significant association was found between diagnosis of other STIs and subsequent incident detection of HR-HPV [aOR: 2.01 (95% CI: 0.74% to 5.48%)] (Abstract O1-S02.02 table 1). Positive detection of any HPV or HR-HPV predicted nearly twofold increased odds of other STIs with the estimates bordering on statistical significance [aORs: any HPV: 1.81 (95% CI: 0.94% to 3.49%); HR-HPV: 2.00 (95% CI: 0.82% to 4.83%)] (Abstract O1-S02.02 table 2).

Conclusions We show that other STIs increase the risk of HPV incidence after controlling for sexual behaviour. The data qualitatively suggest mutual associations of HPV with other STIs. Further studies are warranted to evaluate if these reflect true biologic interactions between HPV and other sexually transmitted microbial agents, or mere confounding from unmeasured sexual risks.