The research team, led by Mariana Rodrigues, Dorian S. Odems and Anna E. Austin at the University of North Carolina at Chapel Hill, focused on a policy mechanism known as broad-based categorical eligibility, or BBCE. Under standard federal rules, households generally qualify for SNAP only if their gross income does not exceed 130 percent of the federal poverty level and they fall below a strict asset limit. States, however, can use BBCE to raise that income threshold to as much as 200 percent of the poverty level and to eliminate the asset test altogether, allowing families to keep modest savings without losing food benefits. Because states adopt these options independently, the country is effectively a natural laboratory of food-assistance generosity, and the researchers exploited that variation to ask a question rarely examined at the individual level: does wider access to nutrition benefits shape violence and mental health in the sensitive window before, during and after pregnancy?

To answer it, the team turned to the Pregnancy Risk Assessment Monitoring System, a CDC-run surveillance survey that samples women who recently gave birth in participating states. They pooled data from 2016 through 2020 and restricted the sample to 66,677 women ages 18 and older who had household incomes at or below 200 percent of the federal poverty level or who were enrolled in Medicaid before pregnancy — criteria designed to capture women plausibly eligible for SNAP. The investigators also included only women who had at least one prior live birth, since childless adults face punishing time limits and work requirements under SNAP rules that make enrollment difficult. About half of the women lived in states with both BBCE policies in place, 27 percent lived in states with one policy, and 23 percent lived in states with neither.

The outcomes were drawn directly from mothers’ survey responses. Physical intimate partner violence was measured with questions asking whether an ex-husband or ex-partner had pushed, hit, slapped, kicked, choked or otherwise physically hurt the respondent in the 12 months before pregnancy or during the pregnancy itself. Postpartum depressive symptoms were assessed with two validated screening questions about feeling down, depressed or hopeless, and experiencing little interest or pleasure in usual activities, since the baby was born; women answering always or often to at least one question were classified as symptomatic. These two questions have been shown to have high sensitivity and specificity compared with structured clinical interviews for major depression.

Crude numbers already told a striking story. Physical IPV in the preconception or prenatal period affected 7.1 percent of women in states with no BBCE policies, compared with 5.8 percent in states with one policy and 5.5 percent in states with both. Postpartum depressive symptoms followed the same gradient, affecting 18.9 percent of women in the most restrictive states, 17.6 percent in states with one policy, and 15.0 percent in states with both. The researchers then used log-binomial regression with generalized estimating equations to account for the clustering of women within states, adjusting for a wide range of maternal characteristics — age, race and ethnicity, education, insurance type, pregnancy intention, and receipt of WIC benefits — as well as state economic conditions such as unemployment, median income, minimum wage, Earned Income Tax Credit rates, Medicaid expansion and paid family leave, and a linear time trend.

After all of these adjustments, the association held. Women in states with both BBCE policies had a 21 percent lower prevalence of preconception and prenatal physical IPV than women in states with no expansions, with a prevalence ratio of 0.79 and a 95 percent confidence interval of 0.66 to 0.94. Their prevalence of postpartum depressive symptoms was 17 percent lower, at a ratio of 0.83 with a confidence interval of 0.75 to 0.93. Women in states with only one BBCE policy showed smaller, directionally consistent reductions — roughly 11 percent lower for both outcomes — but the confidence intervals included the null, meaning those estimates were less certain. The graded pattern, strongest where eligibility was most expansive, appeared across both outcomes simultaneously.

The investigators ran a battery of robustness checks. Results were similar when they included all women regardless of prior births, when they examined each BBCE policy type separately, when they dropped 2020 data to remove pandemic-era distortions, and when they analyzed preconception and prenatal violence separately. They also conducted a placebo test using first-trimester prenatal care initiation as an outcome — a behavior expected to share the same confounders but not to be plausibly affected by SNAP eligibility — and found no association, bolstering confidence that the observed links were not artifacts of unmeasured confounding. Formal E-value calculations indicated that only moderate to strong unmeasured confounding could fully explain the findings.

Why would food policy show up in patterns of violence and depression? The authors ground their interpretation in the Family Stress Model, a well-established framework holding that economic hardship erodes emotional regulation, heightens psychological distress and intensifies relational conflict. Food insecurity, in particular, is both a form of material deprivation and a chronic stressor, and prior reviews have consistently linked it to interpersonal violence and maternal distress. Pregnancy and the postpartum period amplify these dynamics: financial demands rise, employment often shifts, and reliance on outside support deepens, making households more sensitive to the presence or absence of material resources. In this reading, SNAP eligibility expansions operate upstream, easing the economic strain that feeds both relational conflict and psychological symptoms.

The authors also point to a more specific mechanism relevant to violence. Economic control and financial abuse are well-documented tactics of coercion in abusive relationships, and limited access to independent resources constrains survivors’ ability to leave. Food assistance available before and after separation may reduce one structural barrier to safety planning and to establishing a separate household, particularly in states where higher income thresholds and the elimination of the asset test make benefits accessible to survivors rebuilding on their own. The finding that only states adopting both BBCE policies showed robust associations also underscores the importance of policy design: raising income limits and removing asset tests together appears to lower barriers to program access more effectively than either measure alone, a pattern consistent with prior research linking dual BBCE adoption to larger reductions in child protective services reports and foster care entries.

The study has limitations that its authors acknowledge candidly. Most states did not change their BBCE status during the study window, so the analysis could not track how policy changes produce outcome changes over time, and a quasi-experimental design was not feasible with so little policy variation. Self-reported measures of violence and depression may understate true prevalence, though such underreporting is unlikely to differ systematically by state policy context, and the surveillance data capture only physical forms of IPV, leaving emotional, sexual and psychological abuse unexamined. Even so, the implications are timely. As federal and state policymakers weigh stricter SNAP work requirements and proposed limits on BBCE, the findings suggest that tightening food assistance access could raise exposure to material hardship among pregnant and postpartum women, with downstream consequences for violence and mental health. Screening and clinical treatment for IPV and perinatal depression remain essential, the authors conclude, but the results point to a complementary and often overlooked lever: the economic policies that shape the conditions under which risk emerges in the first place.

Subject of Research: The association between state SNAP eligibility expansion under broad-based categorical eligibility and intimate partner violence and postpartum depressive symptoms among low-income women during the preconception, prenatal, and postpartum periods.

Article Title: More Expansive State Supplemental Nutrition Assistance Program eligibility, intimate partner violence, and mental health during the preconception, prenatal, and postpartum periods

Article References: Rodrigues, M., Odems, D. S., & Austin, A. E. (2026). More expansive state supplemental nutrition assistance program eligibility, intimate partner violence, and mental health during the preconception, prenatal, and postpartum periods. SSM – Population Health, 36, Article 101969. https://doi.org/10.1016/j.ssmph.2026.101969

Image Credits: AI Generated

DOI: 10.1016/j.ssmph.2026.101969

Keywords: SNAP, intimate partner violence, postpartum depression, maternal health, food insecurity, BBCE, social safety net, preconception, prenatal, PRAMS, Family Stress Model, health policy