A longitudinal analysis of human papillomavirus 16 infection, nutritional status, and cervical dysplasia progression

To evaluate the effect of potential risk factors, especially human papillomavirus type 16 (HPV-16) infection and nutritional status on the course of cervical dysplasia, we analyzed data from an intervention trial conducted from 1985 to 1990, in Birmingham, Alabama. With the use of data of four repeated evaluations of dysplasia at an interval of 2 months, specific relationships between HPV-16 infection, plasma retinol and zinc levels, and dysplasia progression were evaluated through longitudinal data analysis of generalized estimating equations. Repeated assessments of nutritional status from blood samples, HPV-16 infection, and dysplasia diagnosis were available from 206 women. Dysplasia diagnosis was confirmed by both Papanicolaou smear and colposcopy examinations and was classified as normal, low, or high grade squamous intraepithelial lesions according to the Bethesda system and assigned a score of 0, 1, or 2, respectively. Generalized estimating equation analyses were performed with assumptions of variance of Poisson and link of logarithm. Separate analyses were also conducted for HPV-16-positive and HPV-16-negative women. By multivariate modeling with adjustment for age, race, smoking, oral contraceptive use, and plasma levels of nutrients, HPV-16 infection was found to be related to the progression of cervical dysplasia, with a relative risk of 1.19 and a 95% confidence interval of 1.03-1.38. High plasma levels of retinol were related to the regression of cervical dysplasia, especially in HPV-16-positive women. A protective effect was also observed for high levels of zinc.(ABSTRACT TRUNCATED AT 250 WORDS)