314 - Neighborhood vulnerability and pubertal development in US Children

Publication Number: 314.5984

Sabrina Karim, Boston Children's Hospital, Boston, MA, United States; Sheryl Rifas-Shiman, Harvard Medical School and HPHCI, Boston, MA, United States; Allison J. Wu, Boston Children's Hospital, Boston, MA, United States; Li Yi, Harvard Medical School and Harvard Pilgrim Health Care Institute, Boston, MA, United States; Marie-France Hivert, Harvard Medical School, Boston, MA, United States; Emily Oken, Harvard Pilgrim Health Care Institute, Boston, MA, United States; Izzuddin Aris, Harvard Medical School, Boston, MA, United States

Background: Earlier pubertal onset has been associated with increased risk of obesity, metabolic disorders, and cancers in adulthood. While studies have shown how biological, environmental, and socioeconomic factors may affect pubertal timing, the role of early-life neighborhood environment is less well understood.

Objective: To examine the associations of neighborhood vulnerability during pregnancy and early childhood, and timing of pubertal outcomes.

Design/Methods: We used data from Project Viva, a prospective pre-birth cohort in eastern Massachusetts. We geocoded residential addresses at enrollment during pregnancy (1999-2002) and early childhood (mean age 3.2 years; SD = 0.4) and linked them to Social Vulnerability Index (SVI), a validated measure to quantify communities’ susceptibility to external stressors on human health. SVI was categorized as quintiles, with higher scores representing higher vulnerability. Primary outcomes included age at peak height velocity (APHV) in all children and age at menarche for females only. We used multivariable linear regression stratified by sex to evaluate associations of SVI with pubertal outcomes, adjusting for socioeconomic status (SES), maternal biological factors, and self-reported maternal race and ethnicity.

Results: Of 1295 mother-child dyads included, 885 (69%) mothers identified as non-Hispanic White (Table 1); 671 (52%) children were females. In males, very high SVI in pregnancy was associated with earlier APHV adjusting for SES only (β=-0.28y; 95% CI=-0.54, -0.02; Table 2, Model 1) compared to the very low SVI group. This association persisted after adjustment for maternal factors with medium SVI also showing significance with earlier APHV (β=-0.25y, 95% CI=-0.50, -0.01; Table 2, Model 3). However, early childhood SVI was not associated with APHV in males. In females, SVI in pregnancy or early childhood was not associated with APHV. Very high SVI in pregnancy was associated with earlier menarche after adjusting for SES only (β=-0.42y, 95% CI=-0.83, -0.01; Table 3, Model 1), but was attenuated to non-significance after additional adjustment for maternal factors, and race and ethnicity. Medium SVI in early childhood was consistently associated with earlier menarche across all models (Table 3).

Conclusion(s): This study shows that residence in vulnerable neighborhoods is associated with earlier pubertal timing, particularly for males’ APHV and females’ age at menarche. These findings highlight the importance of addressing neighborhood-level vulnerabilities to promote healthier developmental outcomes and reduce the risks associated with earlier puberty.

Table 1. Demographics by Social Vulnerability Index (SVI) US Total Percentile Ranking at Enrollment


Table 2. Associations of Social Vulnerability Index (SVI) US Total Percentile Ranking During Pregnancy with Age at Peak Height Velocity in Males and Females


Table 3. Associations of Social Vulnerability Index (SVI) US Total Percentile Ranking During Pregnancy and in Early Childhood with Age at Menarche in Females