Bicycle helmet laws and persistent racial and ethnic helmet use disparities among urban high school students: a repeated cross-sectional analysis

BACKGROUND Bicycle helmet laws generally increase helmet usage, but few studies assess whether helmet laws reduce disparities. The objective of this study is to assess changes in racial/ethnic disparities in helmet use among high school students in urban jurisdictions where laws were previously determined to increase overall helmet use. METHODS Log-binomial models were fit to four districts' 1991-2013 Youth Risk Behavior Survey (YRBS) data. Post-regression predictive margins were used to calculate adjusted bicycle helmet use proportions, assess before-to-after changes in race/ethnicity specific helmet use, and estimate changes in disparities from jurisdictions' white subpopulations. RESULTS Helmet use among white students increased by 10.2 percentage points in two Florida counties (p < 0.001), 20.1 points in Dallas (p < 0.001), and 24.4 points in San Diego (p < 0.001). Increases among African Americans were 6.1 percentage points in the Florida counties (p < 0.001), 8.2 points in Dallas (p < 0.001), and 6.3 points in San Diego (p = 0.070). Use increased among Latino students in the Florida counties (4.3 percentage points, p = 0.016) and Dallas (6.2, p = 0.002), but not significantly in San Diego. San Diego helmet use among Asian students increased by 12.8 percentage points (p < 0.001). Because helmet use increased more for white students, helmet laws were associated with increased disparities. In the Florida counties, disparities increased significantly by 5.9 percentage points for Latino students (p = 0.045). San Diego disparities worsened by 18.1 (p < 0.001), 21.3 (p < 0.001), and 11.6 (p = 0.013) percentage points among African American, Latino, and Asian students respectively. Dallas disparities increased by 11.9 (p = 0.015) and 14.0 (p = 0.003) percentage points among African American and Latino students. Increased disparities generally persisted for follow-up time of at least a decade. Main study limitations include the possibility of helmet use reporting error and limited socioeconomic variables in YRBS datasets. CONCLUSIONS Helmet use increased across racial/ethnic subpopulations, but greater increases among white students increased disparities. Policymakers should couple laws with other approaches to reduce helmet disparities and cycling injuries.