WP 56 Dose-Response Relationships Between Cigarette Smoking and Trichomonas Vaginalis Acquisition over 18 Months Among African American Adolescent Females

Tuesday, June 10, 2014
International Ballroom
Andrea Swartzendruber, MPH, PhD1, Jennifer L. Brown, PhD2, Jessica M. Sales, PhD1, Eve S. Rose, MSPH1 and Ralph J. DiClemente, PhD1, 1Rollins School of Public Health, Department of Behavioral Sciences and Health Education, Emory University, Atlanta, GA, 2Department of Psychology, Texas Tech University, Lubbock, TX

Background:  Studies have shown a dose-response relationship between cigarette smoking and bacterial vaginosis.  Bacterial vaginosis is associated with increased Trichomonas vaginalis (TV) acquisition risk; however, few studies have examined dose-response relationships between cigarette smoking and TV.  The objective was to examine dose-response relationships between cigarette smoking and TV acquisition among African American adolescent females.

Methods:  At baseline, 6, 12 and 18 months, African American females 14-20 years (n=701) enrolled in an HIV prevention trial completed audio computer-assisted self-interviews and provided self-collected vaginal swab specimens assayed for TV.  Participants were categorized as light (1-3), moderate (4-5) or heavy (≥6) smokers based on self-reported number of cigarettes smoked per day and as short- (<1 year), medium- (1-2 years) and long-term (≥3 years) smokers.  Generalized estimating equations examined associations between TV acquisition, defined as a positive test result subsequent to a negative result or documented treatment, and amount and duration of cigarette smoking relative to non-smokers.

Results:  Of 605 (86.3%) participants completing ≥1 follow-up assessment, 20.0% (n=121) acquired TV.  Smoking prevalence during follow-up was 12.9-14.1%.  The likelihood of TV acquisition was marginally increased among light (OR: 2.1, 95% CI: 1.0, 4.5) and significantly increased among moderate (OR: 3.0, 95% CI: 1.6, 5.5) and heavy (OR: 3.2, 95% CI: 1.8, 5.8) smokers (test for trend p<0.001).  The likelihood of TV acquisition was also marginally increased among short-term (OR: 1.8, 95% CI: 0.9, 3.6) and significantly increased among medium- (OR: 3.2, 95% CI: 1.5, 6.7) and long- (OR: 3.7, 95% CI: 2.1, 6.5) term smokers (test for trend p<0.001).  Tests for trend remained significant (p<0.05) after adjusting for known correlates of TV acquisition in this sample.

Conclusions:  Dose-response relationships were observed between self-reported measures of cigarette smoking and TV acquisition.  Future research should investigate mechanisms through which cigarette smoking may be associated with TV acquisition.