The multifactorial etiology of cervical dysplasias (intraepithelial neoplasia (CIN), the precursor and preinvasive lesions of cervix cancer, will be examined in a case-control study in the Bronx, New York. Histopathologically diagnosed CIN cases (n=686) will be identified from colposcopically directed biopsies. Equal numbers of women who have two consecutive negative Pap smears will be controls. Controls will be frequency matched to the cases by race. Cervico-vaginal lavage (CVL) and peripheral venous blood samples will be collected. Data on dietary and vitamin intake in the past year and exposure to other CIN risk factors, such as smoking, use of oral contraceptives, level of education, sexual history factors, and menstrual as well as reproductive histories, will be obtained by questionnaires. Presence of human papillomavirus (HPV) DNA in CVL cells will be determined by both Southern blot hybridization and polymerase chain reaction techniques. The confidential counselling and screening for the presence of HIV-1 antibodies will be carried out. Plasma levels of ascorbic acid, beta-carotene, and alpha-tocopherol will be measured by high pressure liquid chromatography (HPLC) methods. It is estimated that 323 Hispanic, 288 Black, and 75 White cases and equal numbers of controls will be recruited in a 2-year period. The smaller number of white CIN cases still provide a revealing comparison with the black and hispanic minority populations. Univariate and logistic regression analyses will be done within each ethnic group to examine independent effects of implicated risk factors. Homogeneity of odds ratios across different ethnic-group factors, between race, HPV and HIV or HPV and smoking, will be tested in multiplicative regression models with all ethnic groups combined (a total sample size of 1,372). Finally polychotomous logistic regression analyses will compare patients with different CIN grades to identify the risk factors that may be related to histopathologic severity of dysplasia at the time of diagnosis.
