The role of childhood Epstein Barr Virus and Cytomegalovirus infections and pubertal timing

Authors: McDonald JA, Goldberg M, Wei Y, Andrulis I, Buys S, Bradbury A, Daly M, John EM, Santella RM, Terry MB

Category: Molecular Epidemiology & Environment, Inflammation & Cancer
Conference Year: 2018

Abstract Body:
Purpose: Examine the independent associations between Epstein Barr Virus (EBV) and Cytomegalovirus (CMV) and pubertal timing using the LEGACY Girls Study. Studies suggest that a higher burden of childhood infections are associated with later onset of puberty; however, an alternative hypothesis suggests that microbes are obesogenic and promote earlier onset of puberty. Methods: LEGACY is a multiethnic cohort of girls recruited at ages 6-13 years, half of whom had a breast cancer family history. Using the subcohort of LEGACY girls who were pre-menarcheal at their first available blood biospecimen (N=498), we measured seropositivity to EBV, CMV, Herpes Simplex Virus 1 (HSV1), and HSV2. We used prospective guardian-reported data on Tanner Stage (TS) for the onset of breast development (TS2-TS5 compared with TS1) and the onset of menarche. We used prospective parametric Weibull survival models to model the association between a given viral exposure and pubertal outcomes. For girls that reached the pubertal event during the study period, we used interval censoring. We adjusted for age at blood draw, body mass index (BMI)-for-age percentile at time of blood draw, race/ethnicity, and maternal education. We also examined whether any of the key exposures and pubertal outcomes varied by time through the addition of interaction terms. Results:The racial/ethnic composition included 69% non-Hispanic white, 15% Hispanic, 8% Asian American, 6% African American, and 2% of mixed race/ethnicity with a mean (SD) age of 11.3 years (2.4). Seropositivity was 38% EBV (17% EBV+ only), 32% CMV (13% CMV+ only), 14% HSV1, and 0.4% HSV2. CMV-only infection was significantly associated with an earlier age at breast TS2+, but not menarche, in age-adjusted models and associations remained after further adjusting for BMI percentile, race/ethnicity, and maternal education (β -0.05, 95% CI -0.09, -0.01). EBV-only infection was associated with earlier age at breast TS2+ and menarche in age-adjusted models; however, associations were attenuated when further adjusted for BMI percentile. Conclusion: Viral exposures were common in this contemporary cohort of girls. CMV infection was associated with earlier breast development suggesting that common childhood viruses may impact pubertal timing.

Keywords: Viral Infection Breast Development Menarche