Social programme receipt and child health, healthcare access and utilisation among low-income US households: a cross-sectional analysis of the National Survey for Child Health Authors: Sonia Taneja, Jie He, Lucy Marcil, Lawrence Chang, Edie Weller, Lois Lee Published: 2026 ABSTRACT AbstractBackgroundPoverty is associated with poor health and increased healthcare utilisation. Social programmes may mitigate these effects by enhancing the ability to purchase direct inputs into health, improving access. We aimed to examine associations between receipt of cash assistance, Supplemental Nutrition Assistance Program, Special Supplemental Nutrition Program for Women, Infants and Children, and free/reduced cost school meals with child health measures among children from low-income US households.MethodsWe conducted a retrospective, cross-sectional analysis of data from the 2019 National Survey for Child Health. Households with incomes <185% of the federal poverty level were included. Multivariable logistic regressions assessed associations, reporting OR with 95% CIs for the following outcomes: parent-reported child health; health access as defined by unmet healthcare need, inability to pay medical bills and experiencing frustrating access to care; and healthcare utilisation defined by the number of preventative healthcare visits and number of emergency visits in the past year.ResultsThe sample included 7279 children aged 0–17, representing 27 million children nationally. Among low-income households, 24.6% received none of the four programmes. Programme receipt was not associated with child health. Programme receipt was associated with frustrating healthcare access (OR 1.72, 95% CI 1.23 to 2.41) and increased emergency department utilisation (OR 1.63, 95% CI 1.16 to 2.29), with no association with preventative care.ConclusionSocial programme receipt was associated with frustrating healthcare access and greater emergency care utilisation. Policy efforts to facilitate access for low-income families may improve child health outcomes. What is known on this topic Poverty is associated with poor health outcomes and increased healthcare utilisation and there is some evidence of individual social programmes improving health outcomes. There is still little studied, however, on what the relationship of multiple programmes in the social safety net is on child health, in particular. Poverty is associated with poor health outcomes and increased healthcare utilisation and there is some evidence of individual social programmes improving health outcomes. There is still little studied, however, on what the relationship of multiple programmes in the social safety net is on child health, in particular. What this study adds Parents of children in low-income households who receive social programmes report more frustrating access to care. They also have increased emergency department utilisation but not preventative care. Parents of children in low-income households who receive social programmes report more frustrating access to care. They also have increased emergency department utilisation but not preventative care. How this study might affect research, practice or policy There continue to be critical gaps in healthcare access for low-income households in the USA, which can guide targets to improve primary and preventative care access for children and families who receive social programmes. There continue to be critical gaps in healthcare access for low-income households in the USA, which can guide targets to improve primary and preventative care access for children and families who receive social programmes. Introduction Poverty, affecting one in seven US children,1 has an established association with poor health outcomes.2 3 Material hardship including food insecurity,4 5 housing instability6 7 and neighbourhood deprivation8 9 is all associated with greater disease burden in children. In adults and children, material hardship is also associated with increased healthcare utilisation for emergency department (ED) visits10 11 and hospitalizations12 13 and decreased preventive healthcare visits.14 While families in poverty may have worse health outcomes, this can be buffered by antipoverty programmes.15 16 These programmes include cash assistance, which is primarily Temporary Assistance for Needy Families (TANF) in the USA,17 the Supplemental Nutrition Assistance Program (SNAP),18 Special Nutritional Assistance for Women, Infants and Children (WIC)19 and free/reduced-cost school meals.20 These programmes improve overall child health and optimise healthcare utilisation by three mechanisms: (1) enhancing a family’s ability to purchase direct inputs into child health, (2) reducing stress and conflict which encourages a healthier environment21 and (3) increasing access to care, specifically preventative healthcare. However, less is understood about the impact of participation in multiple programmes on child health outcomes and utilisation. The objective of this study is to examine the association between receipt of social programmes and child health, access to healthcare and healthcare utilisation for children in low-income households. Our hypothesis is that receipt of at least one social programme is associated with better overall child health and healthcare access due to increased family investment into child healthcare expenditures. A better understanding about the current social safety net may inform future social policy improvements to improve child health. Methods This is a cross-sectional study using the 2019 National Survey of Children’s Health (NSCH) from the Health Resources and Services Administration Maternal and Child Health Bureau. The 2019 survey was used to avoid confounding from the COVID-19 pandemic and related economic policies. The NSCH is conducted annually using address-based sampling and is administered via web or paper surveys in Spanish and English. The overall response rate was 42.4% (68 500 screeners, 35 760 full questionnaires and 29 433 interviews). Data were weighted to account for non-response and generate nationally representative rates.22 This study follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for cross-sectional studies. This study was deemed exempt from institutional review board review. Children 0–17 years old from households with incomes less than 185% of the federal poverty level (FPL) were included, as this threshold broadly captures eligibility for benefit programmes.17–20 While eligibility for each programme varies by state, <185% FPL is a common national cut-off. Households below 185% FPL are eligible for WIC and those below 130% FPL for SNAP. Cash assistance eligibility varies widely by state, but overall income limitations include households <185% FPL. Children with missing data on social programme receipt were excluded. Outcome measures and exposures The primary outcome was caregiver-reported overall child health in response to “In general, how would you describe this child’s health?” Secondary outcomes included healthcare access and utilisation. Healthcare access was measured by the NSCH items: (1) unmet medical need (“During the past 12 months, was there any time when this child needed health care, but it was not received?”); (2) inability to pay for healthcare (“During the past 12 months, did your family have problems paying for any of this child’s medical or health care bills?”) and (3) frustration accessing healthcare (“During the past 12 months, how often were you frustrated in your efforts to get services for this child?”). Utilisation was measured by the number of preventative and emergency healthcare visits in the past 12 months (none, one or more than one). The exposure was household receipt of any of the following programmes in the past 12 months: cash assistance, SNAP, WIC or free/reduced-cost school meals. Individual level covariates included child characteristics including age, race/ethnicity and health insurance status (adequate or not adequate); and adult caregiver characteristics including primary language, education level, and maternal and paternal mental health status. Covariates were selected based on relevance and availability in the NSCH.23 Missingness was minimal across all covariates (<3% for the exposure variable; 0% for all other covariates) and the outcome variables (<2% per variable); therefore, complete-case analysis was used. Statistical analysis Summary statistics (observed counts, survey-based counts and percentages) were calculated for baseline characteristics. We used univariate logistic regression to assess for associations between programme receipt and both demographics and prevalence of the three outcomes. Multivariable logistic regressions were performed for all three outcomes after adjusting for covariates and reporting adjusted OR (aOR) with 95% CIs. Interactions between benefit receipt and covariates were evaluated. As a sensitivity analysis, we fit survey-weighted proportional odds (cumulative logit) regression models for outcomes with ordinal response scales (overall child health, frustration accessing healthcare and ED utilisation) to assess whether dichotomisation affected our findings. The proportional odds assumption was evaluated by comparing common ORs from the ordinal models against cut point-specific binary logistic regression models. As additional sensitivity analyses, multivariable models were fit with age in years included as a continuous (linear) covariate rather than as a categorical term. To assess whether a linear age term was adequate, we then compared each linear-age model with an otherwise identical model in which age was entered as a natural cubic spline (df=3), using a design-based Wald test. We defined a priori a p value <0.05 as evidence that a linear age term did not adequately describe the association between age and the outcome. For any outcomes where this test indicated non-linearity, we conducted another multivariable model with age included as a natural cubic spline (df=3). Multicollinearity among predictors in the multivariable model was assessed using the condition index, excluding factors with values higher than 10. All p values were two-tailed with significance set at <0.05. Weighted estimates using the sampling weights were calculated using R Survey package to adjust for the complex survey design.24 Patient involvement Patients and the public were not involved in the design or conduct of this cross-sectional study, as the study used a publicly available nationally representative dataset and thus direct involvement of individual families was not applicable. However, the findings are intended to inform the public and policymakers about the administrative and access-related challenges faced by families receiving social safety net support. These results will guide future research and policy efforts focused on improving healthcare access and system navigation for vulnerable populations and should inform patient-centred studies using qualitative analysis in which direct patient involvement in study design and community dissemination would be fruitful. Results There were 7279 children from low-income households aged 0–17 years old in the NSCH sample, representing an estimated 27 million children nationally. Age groups were equally represented and 20.4% identified as non-Hispanic Black, 33.0% as non-Hispanic White and 37.1% of Hispanic ethnicity (table 1). In our sample, 24.6% of households received none of four programmes, and 29.9% received only one (figure 1), with highest participation in free/reduced cost school meals (58.1%) (figure 2). Compared with households without benefits, households receiving at least one benefit had higher odds of identifying as Black, non-Hispanic, of Hispanic ethnicity or Other/Multiracial compared with White households. Households receiving benefits had higher odds of being led by an adult with less education, a single parent or a mother with poorer mental health (table 2). Univariate analysis demonstrated that households with benefits had lower odds of adequate health insurance over the last 12 months (OR 0.47, 95% CI 0.37 to 0.59) (table 2). Families receiving at least one programme had lower odds of excellent/very good/good self-reported child health (OR 0.46, 95% CI 0.22 to 0.96) (table 3). Table 1Demographic characteristics of children with household Incomes <185% federal poverty levelParticipation in 1 or more benefitsn=19 500 225 (71.9%)Participation in 0 benefitsn=6 669 655 (24.6%)Total samplen=27 115 277N (total sample)N (%)N (%)N (%)Age (years) 0–532.1%34.6%32.9% 6–1134.9%31.5%33.7% 12–1733.0%33.9%33.4%Sex Female48.5%47.5%48.1% Male51.5%52.5%51.9%Race and ethnicity American Indian or Alaska Native non-Hispanic0.7%0.1%0.6% Asian, non-Hispanic3.2%7.4%4.2% Black, non-Hispanic22.9%13.2%20.4% Hispanic (all races)39.4%28.7%37.1% Native Hawaiian and Other Pacific Islander0.3%0.2%0.3% Multirace non-Hispanic4.8%4.0%4.5% White, non-Hispanic28.7%46.5%33.0%Metropolitan statistical area Yes76.4%77.1%76.7% No13.0%11.2%12.5% Other10.6%11.7%10.9%Adequate health insurance (12 m) Yes77.1%63.0%72.5% No13.7%23.9%16.8% Unknown10.6%13.1%10.7%Childbirth location USA95.3%92.9%94.1% Outside of the USA4.4%5.3%4.8% Unknown0.3%1.8%1.1%Parental nativity Parent(s) born in the USA57.9%60.0%56.7% Any parent born outside USA30.9%29.5%29.7% Other/unknown11.3%10.5%13.6%Primary language spoken in household English75.2%80.5%75.8% Spanish19.7%11.8%17.8% Other/unknown5.1%7.7%6.4%Highest education of adult in household Less than high school19.8%15.9%19.0% High school degree or GED37.8%21.8%33.7% Some college/technical school28.7%25.3%27.7% College degree or higher13.8%37.0%19.6%Family structure Two parents*54.0%66.6%55.7% Single parent35.7%21.8%31.3% Grandparent household/other8.9%8.1%8.4% Unknown1.4%3.4%4.6%Mental health of mother Excellent or very good54.3%62.4%54.8% Good21.6%16.7%19.6% Fair or poor8.5%4.0%7.1% Other15.7%16.8%18.5%Mental health of father Excellent or very good40.1%57.3%43.1% Good13.2%10.4%12.0% Fair or poor4.1%3.1%3.7% Other42.6%29.2%41.1%*Currently married or currently not married.GED, General Educational Development. Demographic characteristics of children with household Incomes <185% federal poverty level Currently married or currently not married. GED, General Educational Development. Per cent of children among households with income <185% FPL enrolled in multiple social benefit programmes. FPL, federal poverty level. Per cent receipt of social programmes in the last 12 months among children in households with income <185% federal poverty level. SNAP, Supplemental Nutrition Assistance Program; WIC, Women, Infants and Children. Table 2Demographic factors associated with programme participation, univariate logistic regression analysisParticipation in 1+programmes, OR (95% CI)Cash assistance participation, OR (95% CI)SNAP participation, OR(95% CI)WIC participation, OR (95% CI)Free/reduced price school meal,OR (95% CI)Age in years, ref: age 12–17 0–50.95(0.74 to 1.23)1.05(0.68 to 1.63)1.49*(1.17 to 1.9)10.6*(7.46 to 14.9)0.35*(0.27 to 0.44) 6–111.14(0.89 to 1.46)1.02(0.65 to 1.62)1.32*(1.04 to 1.67)2.20*(1.49 to 3.25)0.95(0.75 to 1.21)Sex, ref: male Female1.04(0.84 to 1.29)1.10(0.75 to 1.61)1.04(0.85 to 1.26)1.09(0.86 to 1.37)0.95(0.78 to 1.16)Race and ethnicity, ref: White, non-Hispanic Hispanic2.22*(1.66 to 2.97)1.40(0.88 to 2.24)1.03(0.8 to 1.32)1.59*(1.22 to 2.09)1.94*(1.52 to 2.49) Black, non-Hispanic2.79*(2.09 to 3.73)1.55*(1.02 to 2.36)2.52*(1.99 to 3.19)1.40 *(1.06 to 1.85)2.14*(1.67 to 2.76) Asian, non-Hispanic0.70(0.47 to 1.06)0.60(0.24 to 1.53)0.50(0.32 to 0.78)0.65(0.40 to 1.06)0.84(0.55 to 1.27) Other/multiracial, non-Hispanic†2.21*(1.57 to 3.1)1.18(0.67 to 2.07)1.85*(1.35 to 2.52)1.30(0.90 to 1.90)1.75*(1.28 to 2.39)Metropolitan statistical area status (MSA), ref: not living in an MSA Yes0.86(0.67 to 1.09)0.92(0.61 to 1.41)0.81(0.64 to 1.01)0.81(0.62 to 1.05)0.92(0.73 to 1.15)Health insurance status, ref: adequate and continuous health insurance over last 12 m No0.47*(0.37 to 0.59)0.81(0.5 to 1.3)0.46*(0.36 to 0.59)0.45*(0.32 to 0.63)0.64*(0.51 to 0.80)Child’s birthplace, ref: USA Outside of the USA0.80(0.47 to 1.35)1.30(0.65 to 2.61)0.81(0.49 to 1.34)1.24(0.70 to 2.21)0.61*(0.38 to 0.98) Unknown0.16(0.04 to 0.70)0.29(0.05 to 1.55)0.17(0.05 to 0.63)0.99(0.18 to 5.39)0.24(0.05 to 1.08)Parental nativity, ref: any parent born outside the USA Parent(s) born in the USA0.92(0.71 to 1.19)1.09(0.7 to 1.67)1.80*(1.39 to 2.32)0.76(0.58 to 1.01)0.94(0.74 to 1.20) Other/unknown1.03(0.64 to 1.64)2.62(1.49 to 4.59)1.70(1.16 to 2.49)0.66(0.44 to 1.00)1.08(0.73 to 1.60)Primary language spoken in household, ref: English Spanish1.79*(1.15 to 2.79)1.02(0.57 to 1.82)0.67(0.48 to 0.94)1.70*(1.20 to 2.41)1.48*(1.05 to 2.09) Other0.71(0.5 to 1.01)1.06(0.55 to 2.04)0.76(0.52 to 1.09)1.08(0.71 to 1.64)0.83(0.59 to 1.18)Highest education of adult in household, ref: some college/technical school or higher education High school degree/GED or less2.24*(1.78 to 2.82)1.63*(1.1 to 2.43)1.80*(1.48 to 2.19)1.31*(1.04 to 1.64)1.73*(1.42 to 2.11)Family structure, ref: two-parent household Single parent household2.02*(1.62 to 2.51)1.82*(1.19 to 2.8)2.61*(2.11 to 3.23)0.84*(0.66 to 1.07)1.70*(1.39 to 2.08) Grandparent-led household/other family1.36(0.85 to 2.17)3.08*(1.83 to 5.21)1.65*(1.15 to 2.38)0.74(0.49 to 1.10)1.21(0.83 to 1.78) Unknown0.50(0.19 to 1.31)0.32(0.06 to 1.74)1.37(0.57 to 3.28)0.60(0.24 to 1.50)0.63(0.26 to 1.53)Mental health status of mother, ref: excellent or very good mental health Good1.48*(1.13 to 1.95)1.62(0.91 to 2.91)1.32*(1.03 to 1.71)0.94(0.71 to 1.25)1.43*(1.11 to 1.83) Fair or poor2.43*(1.38 to 4.28)1.85*(1.11 to 3.08)2.09*(1.42 to 3.07)1.13(0.74 to 1.72)1.87*(1.22 to 2.87) Other1.07(0.78 to 1.48)2.93*(1.89 to 4.52)1.29(0.97 to 1.70)0.77(0.55 to 1.07)1.14(0.86 to 1.52)Mental health status of father, ref: excellent or very good mental health Good1.82*(1.27 to 2.59)2.22*(1.08 to 4.57)1.13(0.8 to 1.59)1.03(0.70 to 1.51)1.68*(1.21 to 2.32) Fair or poor1.88(0.9 to 3.9)1.00(0.35 to 2.85)2.12*(1.2 to 3.77)1.09(0.51 to 2.32)1.55(0.83 to 2.90) Other2.09*(1.63 to 2.67)1.92*(1.26 to 2.95)2.55*(2.03 to 3.19)0.83(0.65 to 1.06)1.74*(1.40 to 2.16)*Statistically significant test results (p<0.05).†Incl American Indian or Alaska Native Non-Hispanic, Native Hawaiian and Other Pacific Islands.GED, General Educational Development; SNAP, Supplemental Nutrition Assistance Program; WIC, Women, Infants and Children. Demographic factors associated with programme participation, univariate logistic regression analysis Statistically significant test results (p<0.05). Incl American Indian or Alaska Native Non-Hispanic, Native Hawaiian and Other Pacific Islands. GED, General Educational Development; SNAP, Supplemental Nutrition Assistance Program; WIC, Women, Infants and Children. Table 3Association between participation in 1+social benefits and child health outcomes (unadjusted)OutcomesOR (95% CI)Overall child health0.46 (0.22 to 0.96)*Health access Frustration accessing healthcare1.54 (1.15 to 2.06)* Problems paying medical bills0.82 (0.60 to 1.13) Forgone medical care1.33 (0.68 to 2.61)Healthcare utilisation Preventative health utilisation1.00 (0.78 to 1.29) Emergency department utilisation1.82 (1.35 to 2.49)**Statistically significant test results (p<0.05). Association between participation in 1+social benefits and child health outcomes (unadjusted) Statistically significant test results (p<0.05). In the final multivariable model, programme receipt was not significantly associated with overall child health (aOR 0.56, 95% CI 0.23 to 1.35). Decreased odds of excellent/very good/good overall child health, however, was reported for non-Hispanic, Black children (aOR 0.42, 95% CI 0.22 to 0.82), children with inadequate health insurance (aOR 0.46, 95% CI 0.24 to 0.85) and in households with lower adult education (aOR 0.50, 95% CI 0.28 to 0.88) (table 3). Receipt of one or more social programmes was associated with increased odds of frustration accessing healthcare (aOR 1.72, 95% CI 1.23 to 2.41), but not other health access outcomes (table 4). Sensitivity analyses with ordinal outcomes for overall child health and frustration accessing healthcare yielded similar results (online supplemental table 1). Table 4Multivariable logistic regression analysis for healthcare access for children in households with incomes <185% federal poverty levelForgone medical care*OR (95% CI)Frustrationaccessing healthcare*(always/usually/sometimes vs never)OR (95% CI)Problems payingmedical bills*OR (95% CI)Participation in 1+social programmes2.02 (0.97 to 4.22)1.72 (1.23 to 2.41)†1.28 (0.9 to 1.83)Age (years), ref: 12–17 0–50.34 (0.17 to 0.66)†0.72 (0.53 to 0.98)†0.74 (0.5 to 1.09) 6–110.66 (0.40 to 1.10)0.98 (0.73 to 1.32)0.95 (0.63 to 1.45)Race and ethnicity, ref: White, non-Hispanic Black, non-Hispanic0.65 (0.34 to 1.26)0.94 (0.66 to 1.34)0.95 (0.61 to 1.49) Asian, non-Hispanic0.73 (0.13 to 4.03)0.74 (0.41 to 1.34)0.42 (0.19 to 0.94)† Hispanic1.07 (0.49 to 2.35)1.21 (0.85 to 1.72)1.48 (0.99 to 2.21) Other0.76 (0.41 to 1.41)0.74 (0.51 to 1.07)0.80 (0.48 to 1.34)Health insurance, ref: adequate and continuous health insurance over last 12 m Inadequate2.47 (1.46 to 4.17)†2.90 (2.20 to 3.82)20.7 (14.6 to 29.4)† Unknown4.17 (2.43 to 7.16)†3.05 (2 to 4.64)†10.0 (5.84 to 17.2)†Language, ref: English Spanish0.82 (0.3 to 2.24)0.79 (0.48 to 1.3)0.64 (0.31 to 1.31) Others0.68 (0.14 to 3.27)1.84 (1.11 to 3.05)†0.49 (0.19 to 1.22)Education, ref: some college or higher High school degree/GED or less0.71 (0.44 to 1.15)0.95 (0.74 to 1.22)0.64 (0.45 to 0.89)†Maternal mental health status, ref: excellent/very good Good0.78 (0.28 to 2.17)1.64 (1.18 to 2.27)†1.42 (0.93 to 2.16) Fair/poor12.5 (3.42 to 45.6)†2.61 (1.71 to 3.99)†2.06 (1.26 to 3.37)† Other1.69 (0.49 to 5.86)1.28 (0.91 to 1.81)1.06 (0.63 to 1.78)Paternal mental health status, ref: excellent/very good Good2.95 (0.92 to 9.43)1.92 (1.23 to 2.99)†0.94 (0.57 to 1.57) Fair/poor3.87 (1.56 to 9.59)†1.48 (0.8 to 2.76)1.57 (0.84 to 2.93) Other1.88 (1.11 to 3.18)†1.27 (0.95 to 1.7)0.96 (0.64 to 1.46)*All interaction terms were not statistically significant.†Statistically significant test results (p<0.05).GED, General Educational Development. Multivariable logistic regression analysis for healthcare access for children in households with incomes <185% federal poverty level All interaction terms were not statistically significant. Statistically significant test results (p<0.05). GED, General Educational Development. No association was detected between programme receipt and preventative care utilisation (aOR 0.81, 95% CI 0.58 to 1.13). There were increased odds of preventative healthcare utilisation for children 0–5 (aOR 1.75, 95% CI 1.28 to 2.40) compared with age 12–17 and for families with fair/poor compared with excellent/very good paternal mental health (aOR 2.30, 95% CI 1.22 to 4.33) (table 5). In contrast, increased ED utilisation was associated with social programme receipt (aOR 1.63, CI 1.16 to 2.29), children 0–5 years (aOR 1.70, 95% CI 1.28 to 2.25) and good compared with excellent/very good maternal mental health (aOR 1.53, 95% CI 1.14 to 2.04) (table 5). Sensitivity analyses with ordinal outcomes for ED utilisation yielded similar results (online supplemental table 1). The sensitivity analyses including age as a linear variable, also yielded similar results (online supplemental table 2). The spline results indicated age non-linearity at p<0.05 for the outcomes of: frustration accessing healthcare and emergency department utilisation: one or more visits. For these outcomes, the multivariable regression analyses with age modelled as a natural cubic spline also yielded similar results (online supplemental table 2). Table 5Multivariable logistic regression analysis for overall health and healthcare utilisation by children in households with incomes <185% federal poverty levelOverall child health*(excellent/very good/good vs fair/poor)OR (95% CI)Preventative healthcare utilisation in the last 12 months*OR (95% CI)Emergency department utilisation (2+visits) in the last 12 months*OR (95% CI)Participation in 1+programmes0.56 (0.23 to 1.35)0.81 (0.58 to 1.13)1.63 (1.16 to 2.29)†Age (years), ref: age 12–17 0–51.75 (0.81 to 3.75)1.75 (1.28 to 2.40)†1.70 (1.28 to 2.25)† 6–110.93 (0.49 to 1.80)0.98 (0.75 to 1.28)1.05 (0.78 to 1.42), 0.741Race and ethnicity, ref: White, non-Hispanic Black, non-Hispanic0.42 (0.22 to 0.82)†1.05 (0.77 to 1.45)1.33 (1.00 to 1.76) Asian, non-Hispanic1.74 (0.42 to 7.28)0.71 (0.38 to 1.36)0.44 (0.22 to 0.89)† Hispanic0.98 (0.43 to 2.20)1.49 (0.99 to 2.24)1.05 (0.76 to 1.44) Other0.94 (0.35 to 2.53)0.91 (0.60 to 1.38)1.04 (0.71 to 1.51)Health insurance, ref: adequate health insurance over the last 12 m Inadequate0.46 (0.24 to 0.85)†0.83 (0.62 to 1.10)1.13 (0.86 to 1.47) Unknown0.60 (0.16 to 2.16)0.32 (0.23 to 0.46)†1.14 (0.75 to 1.72)Language, ref: English Spanish0.90 (0.29 to 2.81)0.61 (0.37 to 0.98)†0.94 (0.58 to 1.52) Others0.75 (0.28 to 2.03)0.8 (0.46 to 1.38)0.98 (0.61 to 1.58)Education, ref: some college or higher High school degree/GED or less0.50 (0.28 to 0.88)†0.8 (0.62 to 1.04)1.14 (0.91 to 1.44)Maternal mental health, ref: excellent/very good Good0.55 (0.30 to 1.04)0.69 (0.40 to 1.19)1.53 (1.14 to 2.04)† Fair/poor0.54 (0.22 to 1.34)1.14 (0.40 to 3.26)1.25 (0.84 to 1.85) Other0.70 (0.34 to 1.44)0.7 (0.37 to 1.33)0.85 (0.61 to 1.17)Paternal mental health, ref: excellent/very good Good0.78 (0.31 to 1.99)1.25 (0.79 to 1.98)0.69 (0.45 to 1.06) Fair/poor0.36 (0.10 to 1.352.30 (1.22 to 4.33)†1.50 (0.84 to 2.70) Other0.68 (0.34 to 1.36)1.17 (0.89 to 1.54)1.44 (1.11 to 1.87)†*All interaction terms were not statistically significant.†Statistically significant test results (p<0.05).GED, General Educational Development. Multivariable logistic regression analysis for overall health and healthcare utilisation by children in households with incomes <185% federal poverty level All interaction terms were not statistically significant. Statistically significant test results (p<0.05). GED, General Educational Development. Discussion In this cross-sectional study of children in low-income families, receipt of one or more social programmes was not associated with overall child health. However, it was associated with frustrating healthcare access and increased odds of ED utilisation, but not preventive care. This study is novel in its examination of multiple child health outcomes with receipt of all four of the common social safety net programmes for low-income children. Results did not demonstrate an association between social programme receipt and overall child health. This suggests that the financial buffer provided by social programmes takes longer than the 1 year studied to impact child health.25 Furthermore, as these programmes are supplemental by definition, they do not fully counteract the effects of poverty on children.19 Still, social programmes serve crucial functions that support child well-being, including reducing material hardship and food insecurity.26 Receipt of at least one social programme was associated with frustration accessing healthcare. As social programme receipt may actually represent general material hardship and families in poverty are more likely to be uninsured or covered by public health insurance, this may contribute to more limited care access.27 Frustrating access to public services, or ‘administrative burden,’ is an established barrier for families in poverty.28 For healthcare utilisation, social programme receipt was not associated with preventative healthcare visits. Prior studies have demonstrated mixed results. Some demonstrate an association between increased purchasing power from SNAP and increased preventative healthcare utilisation,17 29 and others demonstrate an indirect relationship between benefit enrolment and preventative care.30 This is likely due to administrative churn—short-term disenrolment due to administrative barriers—which is associated with decreased preventative healthcare utilisation.31 Consistent with prior studies, social programme receipt was associated with increased odds of ED utilisation, likely reflecting challenging access to primary or urgent care, rather than true acuity.11 16 32 Our study has limitations. Its cross-sectional design prevents causal inferences. The analysis also did not account for specific programme eligibility criteria, which vary by state, work requirements and individual circumstances. The NSCH also does not capture the full spectrum of social programmes or the duration of benefit receipt. We also did not include medical complexity or chronic health needs in our final model, which plays a role in the frequency of care utilisation. As a large database, which suggests generalisability of our results, there is the possibility of misclassification of income level, social programme participation and demographic characteristics in the sample. Our findings support policy implications to address the healthcare and social needs of low-income families. As one-quarter of our sample received no benefits and 30% received one, there are opportunities to improve existing enrolment of low-income families. These may include streamlining applications through co-enrolment with insurance, reducing bureaucratic eligibility barriers and expanding outreach.28 33 Financial coaching in primary care settings and expanding access to preventative care may also help eligible families enrol.34 35 Future social policy research can further explore how the social safety net can improve outcomes including healthcare utilisation, not just during childhood but over the lifespan. In our study, a minority of low-income households participated in one or more common social programmes. Receipt of at least one social programme was associated with increased frustration accessing healthcare and ED visits but not overall improved child health. Policy implications include recommendations to strengthen antipoverty policies, reform enrolment protocols and improve healthcare access to promote child well-being. Supplementary material