Abstract
Background
Area-level investment in education may influence individual health. We evaluated whether county-level public school expenditure per student around time of birth was associated with cardiovascular health (CVH) in young adulthood.
Methods
Among Future of Families and Child Wellbeing Study participants enrolled at birth and followed through year 22, we used multivariable linear regression to evaluate the associations of annual school expenditure per student in the county of residence at birth (low: z-score ≤ -0.5; intermediate: z-score -0.5 to 0.5; high: z-score ≥0.5) with cardiovascular risk factors and Life’s Essential 8 (LE8) scores at year 22, adjusting for sociodemographic characteristics.
Results
Among 1089 participants (54% female), annual county-level expenditure per student was: low, $4020–6245; intermediate, $6317–7392; high, $7561–9902. By year 22, 9% of participants graduated college, 41% had some college education, and 39% graduated high school; mean LE8 score was 69 (SD 14). Compared with living in a high expenditure county at birth, living in a low expenditure county was associated with 7.5 mg/dL higher non-HDL-C (95% CI 2.9, 12.0) and 3.4-point lower LE8 cholesterol score (95% CI -6.6, -0.23) in young adulthood; and living in an intermediate expenditure county was associated with a 2.3-point lower LE8 glucose score in young adulthood (95% CI -4.6, -0.04). There was no significant difference in overall LE8 score.
Conclusions
Living in a county with lower school expenditure per student around time of birth was associated some worse CVH factors in young adulthood. County-level school expenditure is likely a reflection of a county’s socioeconomic standing. Policies that increase funding for children’s education may improve CVH in young adulthood.
Keywords: Health policy, Socioeconomic status, Education, Cardiovascular health, Youth
1. Introduction
Childhood is a formative period during which adverse social exposures can influence cardiovascular health (CVH) in adulthood [1]. Educational attainment is a key social determinant, as greater attainment is associated with CVH and years lived without cardiovascular disease [2]. Area-level educational expenditure is a policy-level factor that is correlated with educational attainment. However, the associations between community-level school expenditure and the patterns of CVH of children who are born in these communities is not well understood. Therefore, we used data from the Future of Families-Cardiovascular Health Among Young Adults Study (FF-CHAYA) to examine the associations between county-level educational expenditure around time of birth and CVH in young adulthood. We hypothesized that those born in counties with higher education expenditure around time of birth would have better CVH in young adulthood compared to those born in counties with lower expenditure.
2. Methods
Study population. The Future of Families and Child Well-Being Study (FFCWS) is a longitudinal cohort study that enrolled 4897 children at birth from primarily low-income families with unmarried mothers, in 1998–2000 from 20 U.S. metropolitan areas [3,4]. At study year 22, a subset of 1421 participants enrolled in the FF-CHAYA ancillary study and had CVH data measured, with full data collection protocols previously described [4,5]. Our final sample of 1089 individuals composed of FF-CHAYA participants that had data available regarding annual county-level school expenditure per student in their county of residence at birth. Written informed consent was provided by the participants’ parents at baseline and by the participant at year 22. Protocols were approved by the institutional review board at Princeton University. Reporting follows the STROBE guideline.
Variables. The exposure was annual school expenditure per student in US dollars in a child’s county of residence at birth [6,7]. County-level school expenditure per student at year of birth was categorized by z-score as low (z-score ≤ −0.5), intermediate (z-score −0.5 to 0.5), and high (z-score ≥0.5). Primary outcomes measured at year 22 included systolic blood pressure (SBP, mmHg), non-high-density lipoprotein cholesterol (non-HDL-C, mg/dL), hemoglobin A1c (%), and body mass index (BMI, kg/m2). Secondary outcomes included CVH modeled as the American Heart Association’s overall Life’s Essential 8 (LE8) score (range 0–100, higher score indicating better CVH), and individual component metric scores for diet quality, physical activity, smoking, BMI, blood pressure, cholesterol, and blood glucose [8]. LE8 score definitions have been described previously [9], with overall score calculated as the unweighted average of each available component. Covariates included sex, family income in childhood, childhood primary caregiver’s educational attainment, and mother’s age at birth, which were obtained during survey collection at year 0, and child’s educational attainment at year 22. Child’s educational attainment was included to account for individual variations in education in order to better analyze the influence of community-level spending. Given that this may be on the causal pathway between early-life educational investment and adult CVH, a sensitivity analysis was performed utilizing all covariates except for child’s educational attainment (Supplemental Table 1). County-level descriptive variables, such as local tax rate (total tax revenue per capita divided by mean household income per capita of working age adults) [10] and prevalence of low household income (25th percentile or lower of the national household income distribution) were also evaluated [6,11].
Statistical Analysis. Participants’ clinical and sociodemographic characteristics were summarized by annual county school expenditure per student, using mean (standard deviation) for continuous variables and percentages for categorical variables. Multivariable linear regression models estimated the associations between annual school expenditure per student in a child’s county of residence at birth and CVH at year 22, adjusting for the covariates listed above. Statistical significance was set at α=0.05. Details regarding the handling of missing data and comparison to ineligible participants are provided in the Supplemental Methods section and Supplemental Table 2. All analyses were conducted in R version 4.3.0 [12].
3. Results
Study sample. Of 1089 participants, 54.6% were female and 51.1% were Black (Table 1). The mean age at the follow-up visit was 22.3 ± 0.7 years and 46.2% of the cohort lived in a low-income household. The mean school expenditure per student among all included counties was $6860±$1110 per year, and the range in each category of counties was: low (n = 410), $4020–6245; intermediate (n = 271), $6317–7392; high (n = 408), $7561–9902. By study year 22, 9.0% graduated college, 40.9% had some college education, and 38.5% graduated high school. The overall mean LE8 score was 69±14.
Table 1.
Baseline demographics and characteristics of participants.
| Characteristics | All Counties (Overall Participant Sample) | Low Expenditure Counties | Intermediate Expenditure Counties | High Expenditure Counties | p-value |
|---|---|---|---|---|---|
| Number of participants | 1089 | 410 | 271 | 408 | |
| Age at year 22 exam, years | 22.34 (0.67) | 22.61 (0.81) | 22.21 (0.58) | 22.14 (0.43) | <0.01 |
| Sex, Female | 595 (55%) | 212 (52%) | 168 (62%) | 215 (53%) | 0.02 |
| Low household income* | 504 (46%) | 178 (43%) | 126 (46%) | 200 (49%) | 0.30 |
| Race and ethnicity | <0.01 | ||||
| Black | 557 (51%) | 115 (28%) | 196 (72%) | 246 (60%) | |
| Hispanic | 284 (26%) | 196 (48%) | 22 (8.1%) | 66 (16%) | |
| White | 204 (19%) | 79 (19%) | 44 (16%) | 81 (20%) | |
| Another group | 44 (4.0%) | 20 (4.9%) | 9 (3.3%) | 15 (3.7%) | |
| Poverty rate per county, % | 15 (6) | 12 (4) | 19 (6) | 15 (5) | <0.01 |
| Household income per capita, $ | 39,763 (8905) | 41,511 (8700) | 35,760 (7230) | 40,666 (9311) | <0.01 |
| Local tax rate⁎⁎ | 1.02 (0.31) | 0.95 (0.17) | 1.12 (0.21) | 1.03 (0.43) | <0.01 |
| School expenditure per student ($/1000) | 6.86 (1.11) | 5.76 (0.34) | 6.76 (0.30) | 8.03 (0.72) | <0.01 |
| Baseline household income, $ | 33,934 (33,589) | 35,037 (33,491) | 34,760 (35,178) | 32,277 (32,610) | 0.40 |
| Mother’s age at baseline, years | 25.4 (6.0) | 25.3 (5.9) | 25.5 (6.3) | 25.4 (6.1) | >0.90 |
| Primary caregiver's educational attainment | 0.50 | ||||
| Less than high school | 178 (16%) | 74 (18%) | 39 (14%) | 65 (16%) | |
| High school or equivalent | 447 (41%) | 171 (42%) | 111 (41%) | 165 (40%) | |
| Some college or technical school | 234 (21%) | 91 (22%) | 55 (20%) | 88 (22%) | |
| College degree or higher | 230 (21%) | 74 (18%) | 66 (24%) | 90 (22%) | |
| Educational Status at Year 22 | 0.14 | ||||
| Less than high school | 126 (12%) | 44 (11%) | 30 (11%) | 52 (13%) | |
| High school or equivalent | 419 (39%) | 144 (35%) | 101 (37%) | 174 (43%) | |
| Some college or technical school | 445 (41%) | 187 (46%) | 111 (41%) | 147 (36%) | |
| College degree or higher | 98 (9.0%) | 35 (8.5%) | 29 (11%) | 34 (8.4%) | |
| Clinical variables | |||||
| Systolic blood pressure, mmHg | 116 (12) | 115 (12) | 115 (12) | 117 (12) | 0.10 |
| Non-HDL-C, mg/dL | 113 (34) | 118 (36) | 107 (32) | 111 (33) | <0.01 |
| HbA1c, % | 5.28 (0.51) | 5.27 (0.45) | 5.31 (0.63) | 5.26 (0.47) | 0.50 |
| Body mass index, kg/m2 | 29 (8) | 29 (8) | 28 (8) | 29 (8) | 0.70 |
| Overall LE8 CVH score | 69 (14) | 68 (14) | 69 (14) | 69 (15) | >0.90 |
| BMI LE8 Score | 64 (37) | 62 (36) | 66 (37) | 64 (37) | 0.40 |
| Non-HDL-C LE8 Score | 86 (23) | 83 (25) | 90 (21) | 87 (22) | <0.01 |
| Blood Pressure LE8 Score | 84 (24) | 84 (23) | 85 (24) | 82 (25) | 0.30 |
| Glucose LE8 Score | 95 (15) | 95 (15) | 94 (17) | 96 (13) | 0.20 |
| Physical Activity LE8 Score | 43 (48) | 45 (48) | 37 (47) | 45 (49) | 0.07 |
| Smoking LE8 Score | 68 (39) | 69 (39) | 66 (40) | 67 (40) | 0.60 |
| Diet LE8 Score | 40 (31) | 39 (31) | 41 (31) | 41 (32) | 0.60 |
Continuous variables are reported as mean and standard deviation (SD) while categorical variables are reported as counts (n) and frequencies %).
HDL-C: High-density Lipoprotein Cholesterol; HbA1c: Hemoglobin A1c; LE8: Life’s essential 8; CVH score: Cardiovascular health score.
Defined as 25th percentile or lower of the national household income distribution.
Defined as total tax revenue per capita divided by mean household income per capita among working age adults, unitless.
Compared to living in a county with high expenditure per student, living in a county with low expenditure was associated with 7.2 mg/dL higher non-HDL-C level (95% CI 2.6, 12.0) and 3.3-point lower LE8 cholesterol score (95% CI −6.5, −0.10) in young adulthood, and living in a county with intermediate school expenditure was associated with a 2.5-point lower LE8 glucose score (95% CI −4.6, −0.04) and an 8.1-point lower physical activity score (95% CI −15.0, −0.9) in young adulthood (full analyses shown on Table 2). Sensitivity analysis completed without adjustment for child’s educational attainment at year 22 yielded similar results, with the exception of a non-significant LE8 physical activity score among those from counties with intermediate spending.
Table 2.
Association of county-level school expenditure around time of birth and cardiovascular health in young adulthood1.
| County-level School Expenditure Around Time of Birth | Intermediate (versus High) Expenditure | Low (versus High) Expenditure | ||
|---|---|---|---|---|
| b (95% CI) | b (95% CI) | |||
| CVD Risk Factors | ||||
| Systolic BP, mmHg | −1.1 (−2.8, 0.6) | p = 0.20 | −1.5 (−3.0, 0.0) | p = 0.05 |
| Non-HDL-C, mg/dl | −3.5 (−8.6, 1.7) | p = 0.20 | 7.2 (2.6, 12.0)* | p = 0.02 |
| Hemoglobin A1c, % | 0.06 (−0.02, 0.1) | p = 0.14 | 0.02 (−0.05, 0.09) | p = 0.60 |
| Body mass index, kg/m2 | −0.4 (−1.6, 0.9) | p = 0.60 | 0.4 (−0.7, 1.5) | p = 0.50 |
| LE8 Cardiovascular Health Scores | ||||
| Overall LE8 score | −0.5 (−2.7, 1.6) | p = 0.60 | −1.0 (−2.9, 0.9) | p = 0.30 |
| LE8 BMI score | 1.6 (−4.0, 7.2) | p = 0.60 | −3.0 (−8.0, 2.0) | p = 0.20 |
| LE8 cholesterol score | 2.4 (−1.2, 5.9) | p = 0.20 | −3.3 (−6.5, −0.1)* | p = 0.04 |
| LE8 blood pressure score | 2.1 (−1.6, 5.7) | p = 0.30 | 2.3 (−1.0, 5.5) | p = 0.20 |
| LE8 glucose score | −2.5 (−4.8, −0.2)* | p = 0.03 | −1.1 (−3.2, 1.0) | p = 0.30 |
| LE8 physical activity score | −8.1 (−15.0, −0.9)* | p = 0.03 | −2.0 (−8.5, 4.4) | p = 0.50 |
| LE8 nicotine exposure score | −2.1 (−8.1, 3.8) | p = 0.50 | 0.9 (−4.5, 6.2) | p = 0.70 |
| LE8 diet score | −0.9 (−5.6, 3.9) | p = 0.70 | −2.2 (−6.5, 2.1) | p = 0.30 |
Represents statistically significant results
BP: Blood pressure; HDL-C: High-density Lipoprotein Cholesterol; HbA1c: Hemoglobin A1c; LE8: Life’s essential 8; CVD: Cardiovascular Disease.
Adjusted for sex, family income in childhood, childhood primary caregiver’s educational attainment, mother’s age at birth, and young adult educational attainment at age 22.
4. Discussion
In this cohort, lower annual school expenditure in the county of residence at birth was associated with some worse CVH factors in young adulthood, including non-HDL-C and LE8 cholesterol, glucose, and physical activity scores, even after adjusting for individual-level educational attainment. Further, the findings in low school expenditure counties were consistent between both the CVH factors and LE8 factors, likely reflecting a true, clinically relevant difference in non-HDL cholesterol values. Individual-level social factors in childhood – such as access to education and family income level – have consistently been linked to health status in adulthood [13,14]. There is also a large body of literature that investigates neighborhood-level variables and their influence on CVH [15,16]. Our findings further our understanding of area-level social determinants by demonstrating that community-level factors near birth may influence aspects of CVH into young adulthood. Many CVH factors were not associated with school expenditure at birth, which may be because young adulthood could be too early in life to detect changes in CVH components such as blood pressure or BMI.
U.S. counties that spend more on education per student tend to be wealthier [17] and have more available funding for education, which is related to prevailing public school education funding models that are based on local property tax revenue [18,19]. Wealthier counties also have fewer environmental hazards, greater access to healthier foods, and greater access to health care, which also influence health later in life [20,21]. These features of wealthier counties may potentially explain the lower LE8 physical activity and LE8 glucose scores seen in intermediate school-expenditure counties, as well as the worse cholesterol findings seen in low school-expenditure counties. Beyond the role of school expenditure in supporting individual educational attainment, our findings likely reflect the broader socioeconomic position of participants’ counties of residence at birth which may drive associations with better CVH factors in young adulthood. Because school expenditure may be indicative of the socioeconomic status of a county, the associations may operate through other unmeasured pathways that are not related to school expenditure and future work should investigate what these pathways may be.
However, our findings also demonstrated more pronounced differences in LE8 glucose and physical activity scores amongst intermediate school expenditure counties when compared to low school expenditure counties, which is unexpected given the framework of school expenditure as a reflection of county-level socioeconomic status. These seemingly discordant findings may be explained by the higher prevalence of food insecurity and physically labor-intensive occupations in lower-income counties, leading to higher blood sugar and physical activity LE8 scores than those in intermediate counties, who may have more access to regular meal consumption and higher rates of desk work [22]. Given that the sensitivity analysis yielded non-significant findings for LE8 physical activity score among those from intermediate counties, these findings should be interpreted with caution.
Further, it is important to note that despite the association with some worse LE8 component scores, overall LE8 score at year 22 was not significantly different amongst county-level school expenditure groups. LE8 is a measure of one’s CVH at a snapshot in time and is a useful tool to use when estimating population-level health, which is particularly useful in our analysis. Among our cohort of young adults at age 22, many of whom have yet to develop clinically apparent CVD, LE8 overall and component scores give insight on population-level risk of CVD development and the manner through which that risk presents itself.
Despite these important clarification, our results further underscore the potential importance of policy interventions that address social determinants to promote health. Policy programs that have addressed area-level determinants have positively influenced the health of their participants in the long-term [23]. For example, policies requiring chain food establishments to provide calorie counts for menu items, incorporation of bike lanes into neighborhood built environments, and ban on smoking in public areas all contributed to improvements in life expectancy [[24], [25], [26]].
Limitations. Our study has several limitations. First, the observational study design precludes causal inference. Second, the county-level school expenditure measure does not necessarily reflect a community’s values and norms related to education, differences in county-level fixed costs related to education such as transportation, or differences in student demographics such as the proportion with limited English proficiency, which may each influence the per-pupil expenditure [27]. Third, county-level school expenditure is measured at one point in time and does not reflect longitudinal changes or whether participants moved counties. Next, our analyses found that some LE8 factors were significantly worse in intermediate and low school expenditure groups despite no significant change in overall LE8, raising the possibility that these results could be due to chance and should be interpreted with caution. Additionally, data are not readily available describing the proportion of participants who attended school in close proximity to their place of birth, so these data do not account for those who may have moved away from their county of birth before starting school. Further, our analytic sample was different in some characteristics compared to those ineligible because they did not participate in the FF-CHAYA follow-up study. The potential of these differences to introduce bias in the results is acknowledged and must be considered when interpreting the findings. Lastly, county-level school expenditure is also associated with other area-level factors, such as educational quality, graduation rates, and test scores, that were not available for this analysis, which could be potential confounders for which we could not account.
In conclusion, lower area-level school expenditure around time of birth is associated with some worse CVH factors in young adulthood, including non-HDL-C and LE8 cholesterol, glucose, and physical activity scores, even after adjusting for individual-level educational attainment. Future examination of this relationship including using quasi-experimental methods are necessary to understand whether policy interventions related to increasing educational spending result in better health from childhood into adulthood.
CRediT authorship contribution statement
Mariam Ardehali: Writing – original draft, Visualization, Methodology, Investigation, Conceptualization. Abigail M. Gauen: Writing – original draft, Methodology, Investigation, Formal analysis. Kiarri N. Kershaw: Writing – review & editing. Veronica J. Zheng: Writing – original draft. Ikeoluwapo K. Bolakale-Rufai: Writing – original draft. Noreen Goldman: Writing – review & editing, Investigation, Funding acquisition. Daniel A. Notterman: Writing – review & editing, Investigation, Funding acquisition. Donald M. Lloyd-Jones: Writing – review & editing, Investigation, Funding acquisition. Norrina B. Allen: Writing – review & editing, Methodology, Investigation, Funding acquisition. Nilay S. Shah: Writing – review & editing, Supervision, Methodology, Investigation, Conceptualization.
Declaration of competing interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:
Nilay S. Shah reports financial support was provided by National Heart Lung and Blood Institute. Noreen Goldman reports financial support was provided by National Heart Lung and Blood Institute. Daniel Notterman reports financial support was provided by National Heart Lung and Blood Institute. Donald Lloyd-Jones reports financial support was provided by National Heart Lung and Blood Institute. Norrina Allen reports financial support was provided by National Heart Lung and Blood Institute. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
The FF-CHAYA Study is supported by grant R01HL149869 from the National Heart, Lung, and Blood Institute. This manuscript was also supported, in part, by American Heart Association grant 24CDA1266732, and National Institutes of Health grants K23HL157766 to NSS. Funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Footnotes
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.ajpc.2026.101673.
Appendix. Supplementary materials
References
- 1.Pool L.R., Aguayo L., Brzezinski M., et al. Childhood risk factors and adulthood cardiovascular disease: a systematic review. J Pediatr. 2021;232:118–126.e23. doi: 10.1016/j.jpeds.2021.01.053. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Magnani J.W., Ning H., Wilkins J.T., et al. Educational attainment and lifetime risk of cardiovascular disease. JAMA Cardiol. 2024;9(1):45. doi: 10.1001/jamacardio.2023.3990. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Reichman N.E .TJ., Garfinkel I., McLanahan S.S. Fragile families: sample and design. Child Youth V Rev. 2001;23(4–5):303–326. [Google Scholar]
- 4.Lam E.L., Gauen A.M., Kandula N.R., et al. Early childhood food insecurity and cardiovascular health in young adulthood. JAMA Cardiol. 2025;10(8):762. doi: 10.1001/jamacardio.2025.1062. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lloyd-Jones D.M., Allen N.B., Stein J., et al. Future of families: cardiovascular health among young adults cohort study: rationale, key questions, study design, and participant characteristics. J Am Heart Assoc. 2025;14(17) doi: 10.1161/JAHA.125.042030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Fragile families opportunity insights data on county intergenerational mobility restricted use data appendage. 2019. https://ffcws.princeton.edu/sites/g/files/toruqf4356/files/documents/ff_opim_b_9y_res1.pdf [Available from.
- 7.Johnson F.B., Hill J. In: CCD data file: school district financial survey fy 1997 final. Data C-CCo., editor. National Center for Education Statistics; 1996. editor-1997. [Google Scholar]
- 8.Lloyd-Jones D.M., Allen N.B., Anderson C.A.M., et al. Life's essential 8: updating and enhancing the American heart association's construct of cardiovascular health: a presidential advisory from the American heart association. Circulation. 2022;146(5):e18–e43. doi: 10.1161/CIR.0000000000001078. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Ning H., Krefman A., Zhao L., et al. Development and validation of a large synthetic cohort for the study of cardiovascular health across the life span. Am J Epidemiol. 2021;190(10):2208–2219. doi: 10.1093/aje/kwab137. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Federal, State. Local Governments . The United States Bureau of the Census; 1992. 1992 Census of governments - Finances. technical documentation. [Google Scholar]
- 11.Chetty R.H.N. The impacts of neighborhoods on intergenerational mobility II: county-level estimates. Q J Econ. 2017:1163–1228. [Google Scholar]
- 12.Team RC R: a language and environment for statistical computing_ vienna, austria: r foundation for statistical computing. 2023. https://www.R-project.org/ [Available from.
- 13.Kempel M.K., Winding T.N., Bottcher M., Andersen J.H. Evaluating the association between socioeconomic position and cardiometabolic risk markers in young adulthood by different life course models. BMC Public Health. 2022;22(1):694. doi: 10.1186/s12889-022-13158-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Soares S., Santos A.C., Peres F.S., et al. Early life socioeconomic circumstances and cardiometabolic health in childhood: evidence from the Generation XXI cohort. Prev Med. 2020;133 doi: 10.1016/j.ypmed.2020.106002. [DOI] [PubMed] [Google Scholar]
- 15.Patel S.A., Ali M.K., Narayan K.M., Mehta N.K. County-level variation in cardiovascular disease mortality in the United States in 2009-2013: comparative assessment of contributing factors. Am J Epidemiol. 2016;184(12):933–942. doi: 10.1093/aje/kww081. [DOI] [PubMed] [Google Scholar]
- 16.Son H., Zhang D., Shen Y., et al. Social determinants of cardiovascular health: a longitudinal analysis of cardiovascular disease mortality in US counties from 2009 to 2018. J Am Heart Assoc. 2023;12(2) doi: 10.1161/JAHA.122.026940. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Sherman J.D .GB., Poirier J.M. National Center for Education Statistics.; 2003. School district revenues for elementary and secondary education: 1997-98.https://nces.ed.gov/use-work/resource-library/report/statistical-analysis-report/school-district-revenues-elementary-and-secondary-education-1997-98 [Available from. [Google Scholar]
- 18.Hanson M.U.S. Public education spending statistics: education Data initiative. 2025. https://educationdata.org/public-education-spending-statistics [Available from.
- 19.Parrish T. Do rich and poor districts spend alike?: national center for education statistics. 1996. nces.ed.gov/pubs/web/97916.asp [Available from.
- 20.Loccoh E.C., Nguyen A., Kim G., Warraich H.J. Geospatial analysis of access to Health care and Internet services in the US. JAMA Netw Open. 2022;5(11) doi: 10.1001/jamanetworkopen.2022.43792. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Egen O., Beatty K., Blackley D.J., et al. Health and social conditions of the poorest versus wealthiest counties in the United States. Am J Public Health. 2017;107(1):130–135. doi: 10.2105/AJPH.2016.303515. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Beaulac J., Kristjansson E., Cummins S. A systematic review of food deserts, 1966-2007. Prev Chronic Dis. 2009;6(3) [PMC free article] [PubMed] [Google Scholar]
- 23.Knapper J.T., Ghasemzadeh N., Khayata M., et al. Time to change our focus: defining, promoting, and impacting cardiovascular population health. J Am Coll Cardiol. 2015;66(8):960–971. doi: 10.1016/j.jacc.2015.07.008. [DOI] [PubMed] [Google Scholar]
- 24.Alcorn T. Redefining public health in New York City. Lancet. 2012;379(9831):2037–2038. doi: 10.1016/s0140-6736(12)60879-4. [DOI] [PubMed] [Google Scholar]
- 25.Angell S.Y., Cobb L.K., Curtis C.J., et al. Change in trans fatty acid content of fast-food purchases associated with New York City's restaurant regulation: a pre-post study. Ann Intern Med. 2012;157(2):81–86. doi: 10.7326/0003-4819-157-2-201207170-00004. [DOI] [PubMed] [Google Scholar]
- 26.Kershaw K.N., Magnani J.W., Diez Roux A.V., et al. Neighborhoods and cardiovascular Health: a scientific statement from the American heart association. Circ Cardiovasc Qual Outcomes. 2024;17(1) doi: 10.1161/HCQ.0000000000000124. [DOI] [PubMed] [Google Scholar]
- 27.Sherman J.B., Luskin L. U.S. Department of Education. National Center for Education Statistics.; 1996. Assessment and analysis of school-level expenditures.https://nces.ed.gov/pubs/web/9619ch2.asp C. Working Paper No. 96-19[Available from. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.