Abstract P103: Childhood- and Adult Neighborhood-Level Variation in Adult Metabolic Syndrome Measures

Abstract

Objectives: Metabolic syndrome (MS) risk varies by residential neighborhood across the life-course. We compared the proportional variance of adult MS measures explained at the level of adult and childhood neighborhoods. Methods: We defined neighborhoods as groups of Framingham Offspring Cohort (OS) members whose examination 1 (1971-4) residences were geocoded to within 100 m of ≥1 other group member. We included neighborhoods whose residents had ≥5 children in the Generation 3 Cohort (Gen3) aged 0-19 years at the time of their parent’s first examination. Outcomes were measured from 1971-4 for OS and 2002-5 for Gen3; we analyzed residuals of log-transformed body mass index (BMI), systolic blood pressure (SBP), diastolic blood pressure (DBP), and serum high-density lipoprotein (HDL) and -triglycerides, regressed on age and sex. We estimated neighborhood-level intraclass correlations (ICCs) and 95% confidence intervals (95% CIs) of all outcomes using multilevel linear regression. Results: Among 598 Gen 3 participants (mean age 39.8 years, 49.8% men), parental-residence-defined childhood neighborhoods (n=58) explained 6.6% (95% CI: 2.3, 15.5%) and 6.0% (95% CI: 2.1, 15.9%) of the variance in HDL and triglyceride levels, respectively. The neighborhood-level ICC for BMI was 2.4% (95% CI: 0.5, 11.9%); ICCs for blood pressure measures were similar to BMI. Neighborhoods explained none of the variance in any outcome among 772 OS participants (mean age 38.0 years, 49.2% men). Conclusion: Childhood, but not adult, neighborhoods explained a modest percentage of the variance in adult MS measures.