Health Services/Policy
Comparison Group and Non-Random Selection Bias as Potential Modifiers on the Relationship between Supplemental Nutrition Assistance Program (SNAP) Participation and Obesity Risk: A Systematic Review and Meta-Analysis Arin A Balalian* Arin A Balalian A Balalian A Balalian A Balalian A Balalian A Balalian A Balalian A Balalian A Balalian A Balalian A Balalian A Balalian A Balalian A Balalian A Balalian A Balalian A Balalian Arin A Balalian Agriculture, Food, and Nutrition Evidence Center
Background: Conclusions about SNAP participation and obesity in the U.S. are inconsistent, potentially due to several methodological factors, such as selecting an appropriate comparison group and controlling for non-random selection into SNAP (NRS). This systematic review assessed if these factors modify the relationship between SNAP participation and obesity risk.
Methods: We searched 9 databases for peer‑reviewed studies and government reports from 1980 to Nov 2025. Randomized controlled trials, non-randomized interventions, and prospective and retrospective cohorts comparing obesity risk (adults: BMI ≥30.0 kg/m2; children: BMI-for-age and sex ≥95th percentile) between SNAP participants and non-participants (NP) with household income ≤200% of the federal poverty level were eligible. We assessed risk of bias (RoB) with ROBINS-I and certainty of evidence (CoE) with GRADE. Data were synthesized using multilevel-multivariate meta‑analyses with robust variance estimation. Comparison group and NRS adjustment were assessed as potential modifiers.
Results: 10 articles from 3 prospective cohorts were eligible. SNAP participation was not associated with obesity risk (RD=0.01, 95%CI: -0.00, 0.02; CoE: very low). One article that could not be pooled reported higher but insignificant odds of obesity with SNAP (OR=1.58; 95%CI: 0.79, 3.19). Most articles had serious RoB due to confounding or misclassification. 9 articles used income-eligible NP and 2 used NP with unreported eligibility as the comparison group. 5 adjusted for NRS. Comparison group or adjusting for NRS did not modify the relationship (p>0.05).
Conclusions: Prospective cohorts suggest no association between SNAP participation and obesity risk. More studies are needed to increase precision of the modification analyses. More research is needed to investigate other sources of heterogeneity like population characteristics and statistical approach.