Regional variations in human papillomavirus prevalence across time in NHANES (2003-2014).

Introduction: The consequences of low human papillomavirus (HPV) vaccination in Census regions with higher incidence of cervical cancer may contribute to continued disparities. Our purpose was to evaluate regional variations in HPV prevalence across time.

Methods: Repeated cross-sectional data from the National Health and Nutrition Examination Survey (NHANES), 2003-2014 were examined. Participants included females 14 to 34 years old who provided adequate vaginal samples for HPV DNA typing (N = 6387). Region of residence and HPV vaccination status associations with HPV prevalence were examined using chi-square and multivariable logistic regression. HPV types were grouped according to vaccine-type HPV (types 6, 11, 16, 18) and risk (high or low-risk). Time and vaccination status were included in subsequent models for post-licensure survey cycles (2007-2014) to assess their effects on observed associations.

Results: No decreases in vaccine-type HPV prevalence were found between the prevaccine cycles (2003-2006) and early post-licensure cycles (2007-2010, p > 0.05). Vaccine-type HPV prevalence decreased in late post-licensure years (2011-2014) compared to prevaccine years (2003-2006, p = 0.001). The highest prevalence of vaccine-type HPV occurred in the South (8.6%) and Midwest (8.6%), followed by the West (4.8%), and the Northeast (3.5%) in late post-licensure years. Lower odds of vaccine-type HPV across time in post-licensure survey cycles were found to be attributable to time, and more strongly to HPV vaccination.

Conclusions: There were regional variations in vaccine-type HPV prevalence between prevaccine and post-licensure years. These decreases appeared to be at least partially attributable to HPV vaccination. Programs are needed to address geographical disparities in HPV vaccination.