Investigating the historic long-term population health impact of the US National School Lunch Program

Objective: The present research aimed to compare historic participation in the US National School Lunch Program (NSLP) during childhood and subsequent prevalence of overweight and obesity among adults at the population level. Design: Regression models examined cross-sectional, state- and age-based panel data constructed from multiple sources, including the Behavioural Risk Factor Surveillance System, US Congressional Record, US Census and the US Department of Agriculture. Models controlled for cohorts' racial/ethnic composition and state poverty rates. Subjects: Adult-age cohorts (18-34, 35-49, 50-64 and 18-64 years) by US state over a 25-year period (1984-2008). Setting: The cohorts' prevalence of overweight and obesity was compared with the cohorts' estimated NSLP participation during schooling (1925-2007; the NSLP began in 1946). Results: Among adults aged 18-64 years, a one percentage-point increase in estimated NSLP participation during schooling between 1925 and 2007 was significantly associated with a 0.29 percentage-point increase in the cohort's later prevalence of overweight and obesity. Analysis of narrower age cohorts and different schooling periods produced mixed results. Conclusions: The NSLP might have influenced population health historically. Longitudinal analysis of individuals from studies now underway will likely facilitate more robust conclusions about the NSLP's long-term health impact based on more recent experiences.