Genital human papillomavirus prevalence over the lifespan among females and males in a national cross-sectional survey, United States, 2013-2016.

BACKGROUND Patterns of human papillomavirus (HPV) prevalence by age differ by sex. To further the descriptive epidemiology of genital HPV, we analyzed prevalence by age for non-vaccine-(non-4vHPV)-type and vaccine-(4vHPV)-type HPV by sex using 2013-2016 National Health and Nutrition Examination Survey data, the first 4 years of national data from both sexes. METHODS Penile and cervicovaginal swabs were self-collected from 15-59-year-olds and tested for 37 HPV types. 4vHPV-type (6/11/16/18) and non-4vHPV-type (any of 33 other types) prevalences were estimated by 3-year age group and participant characteristics. Average percent changes (APC) in prevalence were estimated using segmented log-binomial regression. RESULTS Among females, a positive relationship between non-4vHPV-type prevalence and age was seen from 15-17 to 21-23 years (APC: 56.5), followed by a negative relationship through 30-32 years (APC: -13.2); thereafter, prevalence was not related to age. 4vHPV-type prevalence was positively related to age through 24-26 years (APC: 56.9), then negatively related through 57-59 years (APC: -6.0). Among males, non-4vHPV-type prevalence had a positive relationship with age through 21-23 years (APC: 102.4) with a smaller positive relationship through 57-59 years (APC: 1.4). For both sexes, modeled joinpoints for 4vHPV-type prevalence occurred at older ages compared to joinpoints for non-4vHPV-type prevalence. CONCLUSIONS Sex differences in age-specific non-vaccine-type HPV prevalence may reflect natural history and sexual behavior. Differences in vaccine-type and non-vaccine-type modeling results suggest vaccine impact as joinpoints occur in mid-late-20s for vaccine-type HPV but early-20s for non-vaccine-types. These data can assist in refining HPV vaccination models and inform HPV vaccination practices and policy.