Background Predictors of rectal health are needed for public health programs and planning rectal microbicide clinical trials.

Methods From October 2006 to December 2008, 896 men and women in Los Angeles and Baltimore in a rectal health and behaviours study (UCLA IPCP U19 0606414) completed interviews and rectal sexually transmitted infections (STIs) testing including Gonorrhoea (GC), Chlamydia (CT) and high-risk human papilloma-virus (HR-HPV). Rectal signs detected by High Resolution Anoscopy (HRA) and self-reports of symptoms in the past month were analysed for associations with STIs (GC (n=7), CT (n=15) and HR-HPV (n=178)), demographics and receptive anal intercourse (RAI) in the past month. Those with active haemorrhoids were excluded (n=77) from analyses of STI predictors.

Results SIGNS—Men reporting RAI (n=234) had a higher prevalence of swelling seen by HRA than men not reporting RAI (6.4% vs 1.9%; p value 0.02). Prevalence of erythema or haemorrhoids was lower among men reporting RAI (2.2% vs 6.8%; p value=0.02 and 18.3% vs 30.8%; p value<0.01 respectively). There were no differences in signs between women by RAI. More erythema and discharge were HRA detected among HIV-positive men; more HIV positive men and women than HIV negative had ulceration. SYMPTOMS—Significantly more RAI+ men reported symptoms (swelling, itching, burning and irritation) but there were no differences in reported symptoms among women by RAI. HIV positive men (n=218) and women (n=186) reported more symptoms (swelling, itching, discharge and burning) than HIV negative men and women; HIV positive men reported more pain and irritation than HIV negative men (n=227). STIs—No HRA detected signs were associated with rectal GC or CT. Bleeding (OR 5.06, 95% CI 1.09 to 23.45) and leukoplakia (OR 5.06 95% CI 1.22 to 25.66) were significantly associated with detection of HR-HPV. No symptoms were associated with GC or CT; discharge in the past month was associated with HR-HPV (OR 2.16, 95% CI 1.07 to 4.35).

Conclusions Lack of association between signs or symptoms and prevalent rectal STI suggests syndromic approaches may be problematic requiring laboratory testing for detection of rectal STIs. Prevalence of signs varied little by RAI practice but there was a higher prevalence of self-reported symptoms among men reporting RAI. These findings can serve as a reference for expected rates to be found in clinical trials testing rectal microbicides and interpretation of adverse events.