Evaluating segregated school attendance as a mediator of geographic inequities in late‐life cognitive function: evidence from the Study of Healthy Aging in African Americans (STAR)

Abstract

INTRODUCTION

Birth in the Southern United States is associated with poorer late‐life cognitive health, especially among Black Americans, yet the role of school segregation is unclear.

METHODS

Utilizing decomposition methods, we estimated the total effect, natural direct effect (NDE), and natural indirect effect (NIE) of Southern birth on domain‐specific cognition among 727 older Black adults, adjusting for early‐life covariates. We also estimated the proportion of the total effect mediated by self‐reported segregated school attendance.

RESULTS

Southern birth was associated with lower late‐life executive function and semantic memory; estimates were negative but not significant for verbal episodic memory. The direct effect of Southern birth was negative but not significant for all domains. Attending a segregated school mediated 35% and 49% of the total association between Southern birth and executive function (NIE: −0.07, 95% confidence interval [CI]: [−0.18, 0.02]) and semantic memory (NIE:−0.17, 95% CI: [−0.29, −0.06]).

DISCUSSION

School segregation may partially drive geographic inequities in late‐life cognition in the United States.

Highlights

Southern birth has been linked to poorer cognitive health in later life. Segregated schooling may partially explain geographic disparities in brain aging. Black adults born in the South had lower cognitive function and were more likely to attend segregated schools. Segregated schooling accounted for 35% to 49% of the association between Southern birth and cognition.

BACKGROUND

Older Black adults are approximately twice as likely to experience dementia compared to older White adults, even after accounting for well-established dementia risk factors.¹ Older Black adults born in the South, which accounts for around 80% of older Black adults in the United States,² have greater dementia incidence and lower late-life cognitive function, compared to their counterparts born outside the South.^3,4 Even for those who moved out of the South in adulthood, early-life experiences such as educational environments may have persistent effects on later-life health outcomes, including dementia risk.^4–7 The distribution of the Black population in the United States reflects a legacy of slavery and internal migration patterns, such as the Great Migration and White flight, which align with the geographic patterns of systemic disinvestment in communities of color over generations.⁸ While inequities in educational investments have been and continue to be common across the United States, they may be more pronounced in the South due to the history of both de jure and de facto segregated schools during the Jim Crow era.⁹ The role of segregated schooling in driving geographic inequities in cognitive health in later life remains underexplored. To our knowledge, this study is the first to use causal mediation techniques to examine the impact of school segregation on the relationship between Southern birth and cognitive function among older Black Americans.

Education plays a crucial role in brain development and is a well-established risk factor for dementia.^10–15 Schooling experiences vary by geography (e.g., shorter term lengths and funding in the Southern United States)^8,9,16 and across racial groups (e.g., disinvestment in schools attended by a majority of students racialized as Black).⁸ Historical racial school segregation may increase dementia risk through multiple, and potentially conflicting, pathways.^17–20 Structural racism underpinned racially segregated schools, producing disparities in funding, resources, and educational quality.⁸ Schools serving predominantly Black students often experienced resource limitations, shorter school durations, and reduced educational quality, largely due to inequitable funding allocations,^16,21–23 which may result in lower educational attainment.²⁴ Reduced educational opportunities can constrain adult socioeconomic status (SES) through restricted income and occupational mobility,^16,25–27 and low SES is associated with a greater risk of chronic health conditions and limited access to preventative care and treatment – both risk factors for dementia.²⁸ At the same time, Black students who attended desegregated schools often experienced high levels of interpersonal racial discrimination within the school setting.^29,30 This was further compounded by the displacement of Black teachers after Brown, which disrupted supportive educational environments and mentorship networks that had previously buffered the impacts of racial segregation.^31,32 Such experiences of interpersonal racism can trigger chronic stress responses through repeated activation of the hypothalamic–pituitary–adrenal axis and elevated glucocorticoid levels.^33–37 Over time, these prolonged stress processes contribute to neurodegeneration, cardiovascular disease, and poorer mental health,^38–40 all of which heighten dementia risk. The effects of the Jim Crow era persist in the form of policies such as, but not limited to, discriminatory zoning laws and home loan practices,^41,42 which has an impact on access to quality education (e.g., inequitable funding) and may be particularly relevant for brain health due to heightened neuroplasticity occurring during childhood.⁴³ Studies have demonstrated that segregated schooling is associated with lower cognitive function in older Black adults^{17–20,32,44}; however, the extent to which these differential schooling experiences contribute to the geographic disparities in late-life cognition has not been fully established.

We utilized causal mediation methods^45,46 to estimate the direct (not mediated through segregated school attendance) and indirect (mediated through segregated school attendance) effects of Southern birth. Drawing on early childhood development research,⁴³ we hypothesized that segregated schooling would account for a significant portion of the variation in the relationship between Southern birth and late-life cognitive performance. However, we also expected that segregated schooling would not fully explain this association, as the health impacts of geography extend beyond the schooling system to include factors ranging from policies around healthcare access and quality, residential segregation, environmental exposures, early childhood nutrition, and beyond.

METHODS

Study population

Data came from the Study of Healthy Aging in African Americans (STAR) cohort, composed of community-dwelling older adults residing in California's San Francisco Bay area, predominantly in Oakland and Richmond. The STAR cohort is an epidemiological study of Kaiser Permanente Northern California (KPNC) members who previously participated in Multiphasic Health Checkup (MHC) exams between 1964 and 1985, identified as Blacks or African Americans, and were aged 50 years or older on January 1, 2018. The MHC exams yield prospectively collected clinical, lifestyle, and behavioral information. Sampling was randomly stratified by age and educational attainment to recruit approximately equal proportions of participants ages 50 to 64 and 65 and older and with varied levels of educational attainment. The exclusion criteria included electronic medical record diagnosis of dementia or other neurodegenerative diseases (frontotemporal dementia, Lewy body disease, Pick's disease, Parkinson's disease with dementia, and Huntington's disease) and the presence of health conditions that would impede participation in study interviews (hospice activity in the last year, history of end-stage renal disease or dialysis in the previous year, history of severe chronic obstructive pulmonary disease in the last 6 months, or congestive heart failure hospitalizations in the previous 6 months). A total of 764 individuals aged 50+ were enrolled (mean age 68, range 53 to 94; 69% female) at baseline (2017 to 2018). The study obtained approval from the KPNC Institutional Review Board, and all participants provided informed consent.

Our analytic sample excluded participants who were not born in the United States (n = 11), those without information on which US state they were born in (n = 21), and those with no information on whether their school was segregated during any grade attended (n = 1) (Figure S1). We additionally excluded participants who were missing any of the domain-specific baseline cognitive scores and anyone missing covariate information (n = 4). The final analytical sample included 727 participants.

Measures

Southern birth

At baseline, STAR participants reported their state of birth, which was then coded as a Southern versus other region using the U.S. Census Bureau regional designations.⁴⁷ This included Alabama, Arkansas, Delaware, District of Columbia, Florida, Georgia, Kentucky, Louisiana, Maryland, Mississippi, North Carolina, Oklahoma, South Carolina, Tennessee, Texas, Virginia, and West Virginia.

RESEARCH IN CONTEXT

• Systematic review: The literature was reviewed using academic databases (e.g., PubMed, ERIC). Few studies have examined the associations between Southern birth, school segregation, and cognitive outcomes in Black adults; these studies informed our hypotheses and are cited accordingly. While most research examining birth region implications on cognitive health disparities focuses on cardiovascular mechanisms, none evaluated the mediating role of segregated schooling between Southern birth and cognitive function.
• Interpretation: Our findings suggest that interventions aimed at eliminating school segregation, or investing in segregated schools, could partially reduce adverse cognitive outcomes in later life for Black adults who were born in the South. However, factors beyond segregated schooling also contribute to these disparities.
• Future directions: Additional studies in larger, more diverse samples are needed to examine life course factors driving racial disparities in late-life cognition, with more detailed assessments of segregated schooling experiences and dementia risk among older Black adults.

Segregated school attendance

At baseline, participants were first asked if they attended school in grades 1, 6, 9, and 12. For each grade, participants who reported having attended school were asked, “Was this a segregated school?” Responses were either recorded as “yes” or “no.” We created a binary variable indicating whether participants self-reported if they ever attended a segregated school at any of the given grades (1 = attended a segregated school for at least one grade; 0 = never attended a segregated school).

Cognitive function

Baseline cognitive function was assessed during in-person interviews with the Spanish and English Neuropsychological Assessment Scales (SENAS), a battery of cognitive tests developed with item response theory methodology to enable valid comparisons of cognition and cognitive change across racially, ethnically, and linguistically diverse groups.^48,49 We calculated domain-specific scores for verbal episodic memory, semantic memory, and executive function. Briefly, the semantic memory domain evaluates long-term memory, which involves knowledge of facts, concepts, and general information about the world. The semantic memory component was assessed through verbal (object naming) and non-verbal (picture association) tasks. The verbal episodic memory domain measures the ability to recall specific past events or experiences, such as conversations or stories. The verbal episodic memory score was derived from a multitrial word-list-learning task with a delayed recall component. Lastly, the executive function domain reflects higher-level cognitive processes related to planning, decision-making, problem-solving, and managing tasks. The executive function composite was based on tasks assessing category fluency (e.g., vegetables and fruits), phonemic (letter) fluency, and working memory (e.g., digit-span backward, list sorting). Each cognitive domain score was standardized using the baseline sample to produce domain-specific z-scores.

Confounders

We selected confounders that would have preceded place of birth and experiences of school segregation (see Appendix S1 for conceptual framework). Due to the proximity in time between the exposure (Southern birth) and the mediator (segregated schooling), we considered the same set of confounders for both the exposure–mediator and mediator–outcome relationships. These included baseline age (continuous), sex/gender (male/female), and parental education. Parental education was defined using the highest level of education achieved by either the participant's mother or father figure, whichever was higher or non-missing. Consistent with prior work in this cohort,⁷ we operationalized parental education as less than high school versus high school completion or more. We also included a missing indicator category for those without information on parental education for both parents, as this could reflect socioeconomic disadvantage, given that the missingness may be patterned by household structure and related childhood financial and human capital.⁵⁰

Statistical analysis

Descriptive analyses compared segregated school attendance and covariates by Southern birth. Bivariate analysis was subsequently conducted to assess statistically significant differences by Southern birth, using chi-squared tests for categorical variables and t tests for continuous variables (Table 1). We first estimated the total effect of Southern birth on domain-specific cognitive performance z-scores, adjusting for age, sex, and parental education through a series of linear regressions. We next evaluated whether Southern birth was associated with segregated schooling utilizing a Poisson regression with robust standard errors. This step is necessary for establishing the plausibility of segregated schooling as a potential mediator of the Southern birth-cognition association.

TABLE 1 Baseline descriptive characteristics (mean [SD] or percentage) of the analytical sample of participants from the Study of Healthy Aging in African Americans (STAR) aged 50 years or older (n = 727) by birth in the Southern United States.^a

	Overall (n = 727)	Southern birth (n = 279)	Non-Southern birth (n = 448)
Respondent demographic characteristics
Age, years*	68.8 (8.8)	74.9 (8.4)	65.0 (6.6)
Sex
Male	30.4%	32.3%	29.2%
Female	69.6%	67.7%	70.8%
Highest level of parental educational attainment*
Less than high school completion	24.2%	38.4%	15.4%
High school completion or more	66.2%	45.9%	78.8%
Missing for both parents	9.6%	15.8%	5.8%
Attended a segregated school*	38.7%	71.7%	18.1%

We used a counterfactual framework⁵¹ to determine whether geographic inequities of Southern birth on late-life cognitive function were mediated through segregated schooling, through the estimation of natural direct effects (NDEs) and natural indirect effects (NIEs). Estimating these effects required the specification of two models. The first model regressed segregated schooling on Southern birth and confounders of the exposure–mediator relationship (age, sex, and parental education). Predictions from this mediator model were used to estimate the counterfactual mediator values for each respondent across exposure values. The second model regressed cognition on Southern birth, the segregated school attendance mediator, and mediator–outcome confounders (age, sex, and parental education). We additionally included cross-product terms between Southern birth and segregated school attendance in the outcome models given evidence of multiplicative interaction between Southern birth and segregated school attendance on late-life cognition. We used the “mediate” command in Stata,⁵² which combined estimates from the mediator and outcome models to estimate the NDEs and NIEs utilizing Imai et al.’s potential outcomes framework.⁵³ Percentile bootstrap 95% confidence intervals were estimated by repeating the analysis on 1000 resamples with replacement. The mediated proportion was computed as the ratio of the NIEs divided by the total effect. For valid estimation, our models require the assumption that there are no unmeasured exposure–outcome, mediator–outcome, or exposure–mediator confounders and no mediator–outcome confounders affected by exposure. See Appendix S1 for further details on assumptions.

As a sensitivity analysis, we restricted our sample to participants with segregated schooling information available for all grades to ensure greater consistency in the exposure definition. In contrast, the primary analysis included participants with information on segregated schooling at any grade, rather than all grades – thereby inherently restricting the sample to those who attended at least 12th grade – to avoid conditioning on educational attainment, a possible downstream outcome of our mediator. However, the majority of participants in the primary sample participants (90%) had information on segregated schooling across all grades. To address potential birth cohort differences, we conducted stratified analyses using age 70 as the cutoff. Finally, we additionally accounted for rural status at birth, given its association with educational and health outcomes,^8,54 treating it as a covariate to evaluate the robustness of our findings. Analyses were performed using Stata 18.0 (StataCorp, College Station, TX).

RESULTS

Descriptive findings

Participants averaged 68.8 (8.8 ± standard deviation [SD]) years of age, and 70% were women (Table 1). Over one-third attended a school that was segregated, and over half had at least one parent who completed high school. Additionally, 38% of participants were born in the South. Compared to those who were born in a non-Southern state, participants who were born in the South were older (e.g., 74.9 ± 8.8 years vs 65.0 ± 6.6 years), more likely to have attended a segregated school (e.g., 72% vs 18%), and had less educated parents (e.g., 46% vs 79% had at least one parent who completed high school).

Total effect of Southern birth on cognition

Being born in the South (vs other regions) was associated with lower executive function (β = −0.20; 95% CI: −0.35, −0.05) and semantic memory z-scores (β = −0.35; 95% CI: −0.50, −0.19) (Figure 1; Table S1). Average verbal episodic memory z-scores were similar for participants born in the South compared to other regions (β = −0.07; 95% CI: −0.23, 0.08).

[IMAGE OMITTED. SEE PDF]

Exposure–mediator relationship between US Southern birth and school segregation

Participants who were born in the South were 3.29 times as likely to have attended a segregated school (RR = 3.29; 95% CI: 2.60, 4.15), compared to those born in other parts of the United States.

Mediation analysis

The estimated direct effect of being born in the South, not mediated by segregated school attendance, was −0.14 (95% CI: −0.31, 0.04) for executive function z-scores, although the lower bound of the confidence interval crossed the null. Segregated school attendance mediated 35% (95% CI: −16% to 84%) of the association between Southern birth and executive function z-scores (NIE: β = −0.07; 95% CI: −0.18, 0.02). For semantic memory z-scores, the direct effect of being born in the South, not mediated by segregated school attendance, was −0.19 (95% CI: −0.39, 0.01). Segregated school attendance mediated 49% (95% CI: 10% to 85%) of the association between Southern birth and semantic memory z-scores (NIE: β = −0.17; 95% CI: −0.29, −0.06).

Although the total association between being born in the South and verbal episodic memory z-scores was null, the estimated NDEs and NIEs from the mediation analysis may be informative if direct and indirect effects offset one another. Our results suggest no significant associations between Southern birth and verbal episodic memory z-scores, either through direct associations (direct effect: β = −0.11; 95% CI: −0.28, 0.05) or pathways mediated by segregated school attendance (indirect effect: β = 0.04; 95% CI: −0.06, 0.15) (Figure 1; Table S1).

Sensitivity analysis

Results were generally consistent when the sample was restricted to participants with information on their segregated schooling attendance for all grades (Table S2), with the exception of slightly larger total (β = −0.38; 95% CI: −0.54, −0.21) and NDE (β = −0.21; 95% CI: −0.41, −0.01) estimates for the semantic memory domain (45% mediated; 95% CI: 7% to 84%). When examining potential birth cohort differences, we observed the expected overall attenuation in our estimates, likely reflecting modest sample sizes and limited precision. No significant differences were observed between age groups, and results were consistent with our primary findings, as indicated by overlapping 95% confidence intervals (Table S3). These findings should nonetheless be replicated in larger studies, as our ability to detect cohort-related differences may be limited. When we additionally adjusted for rural status at birth, estimates remained comparable to those in the primary analyses (Table S4).

DISCUSSION

In a cohort of older Black adults residing in Northern California, we found that, consistent with prior research,^5,7 being born in the South was associated with lower executive function and semantic memory z-scores and that, as hypothesized, individuals born in the South were more likely to have attended segregated schools. Segregated school attendance mediated 38% to 49% of the overall association between Southern birth and late-life executive function and semantic memory z-scores. There was no evidence of an overall association between Southern birth and verbal episodic memory z-scores; direct and indirect estimates were also null. Our findings extend a growing body of research pointing to the role of Southern birth and early-life geography more generally in shaping later-life health outcomes. Interventions targeting aspects of the schooling environment offer a potential avenue for ameliorating geographic inequities in health. This is the first study, to our knowledge, to formally evaluate segregated school attendance as a mediator of the relationship between Southern birth and later-life outcomes.

While direct comparisons with previous studies are challenging due to differences in geographic classifications, study samples, measurement of outcome variables, and modeling techniques, our findings align with research suggesting a harmful impact of Southern birth on dementia-related outcomes. For instance, Farina et al.⁴ found in a nationally representative sample that Black adults who were born in the South and currently resided there had the highest dementia risk, followed by Black adults born in the South but residing in other geographical regions. Similar to our findings, prior work among a different group of KPNC members found an association between Stroke Belt birth and higher dementia incidence in older Black Americans who left the South by midlife.⁵ Given the historical context of entrenched racial segregation across various sectors, especially in the South, our findings highlight the importance of considering the role of geography on late-life cognitive outcomes. They also underscore the significance of early-life conditions, suggesting that older Black Americans born in the South carry a heightened dementia risk with them regardless of geographic relocation, given that all participants were currently residing in Northern California.

George et al.⁷ observed an association between Stroke Belt birth and lower late-life executive function and semantic memory among Black Americans within the STAR cohort. Importantly, the continued association with Stroke Belt birth, even after informal covariate adjustment of education, suggests that factors beyond years of schooling and other aspects of socioeconomic status may be at play. Our study advances this work by explicitly incorporating measures of school segregation to better capture the historical and social contexts that shaped Black Americans’ schooling experiences.^8,55 Furthermore, by examining how structural educational inequities, a modifiable mediator, contribute to geographic cognitive differences, this study more closely approximates the effects of possible educational reforms and offers insights into intervention pathways that could mitigate dementia risk.

Similarly, Lamar et al.¹⁸ found that older Black adults in Chicago who resided in the South at age 12 had lower cognitive scores. When examining effect modification by segregated school status, point estimates were lower for those who resided in the South and attended a desegregated school compared to their counterparts who resided in the South and attended a legally segregated school or resided in the Northeast/Midwest; however, confidence intervals overlapped. While moderation identifies populations at greater risk, it does not clarify the underlying mechanisms driving disparities. In contrast, our mediation-focused approach allowed us to interrogate geographic cognitive disparities and quantify the extent to which school segregation explained these differences. The divergences between our findings and those by Lamar et al. may also reflect varying implications from the experience of attending racially integrated schools and then migrating to Chicago versus Northern California. There were different patterns of migration from the South, and it is plausible that the selection processes and timing driving migration to Chicago versus Northern California led to differences in the cognitive health of older adults.⁵⁶

Interpretation of these findings should also consider birth cohort variation, as STAR participants (born 1920s to 1960s) were likely exposed to shifting educational policies and school quality shaped by state and local contexts.^8,57–59 However, desegregation efforts did not accelerate until the late 1960s,⁵⁵ suggesting relatively consistent experiences of segregation among Southern-born participants. These dynamics highlight the importance of considering the intersection of time and place in understanding how educational reforms may differentially shape late-life cognitive outcomes.

Our study has several limitations. First, due to the small number of participants who were born outside of the South had attended only segregated schools and the resulting potential collinearity issues, the operationalization of segregated school attendance was limited to a binary measure – defined as ever versus never attended a segregated school. Prior studies have shown the importance of more nuanced measures of segregated school attendance (e.g., grade-specific exposures) in shaping later-life cognitive outcomes, as well as varying implications of de jure versus de facto segregation.^19,32 The lack of a more detailed gradient in the mediator definition likely resulted in an underestimation of the indirect effect of Southern birth through segregated school attendance. Similarly, due to data sparsity and potential positivity violations, we were unable to expand the exposure definition to explore potential cognitive differences between Southern-born individuals and those born in other specific regions of the United States (e.g., Northeast). Further categorization of the geographical exposure may help uncover patterns of cognitive inequities in other regions of the United States.

Our sample consisted of older Black/African American adults currently living in California; thus, all individuals born in the South in our sample migrated from the South by midlife or earlier – a feature that provides a unique lens on this population. Despite a potentially shorter period of exposure to the South, we still found evidence of its long-lasting impacts on cognition, both through experiences of segregated schooling and through other pathways. Prior research showed the importance of considering both Southern birth and continued residence in the South in shaping late-life outcomes (e.g., Farina).⁴ In addition, individuals may have migrated from the South for varying reasons, which could differentially relate to early-life factors associated with dementia risk. Though we were also unable to account for the duration of time participants spent in the South before relocating to Northern California, our focus on participants currently residing in the same region could mitigate some of the potential influence of later-life environments on cognitive outcomes. Finally, our analytic sample likely reflects a selective group of Southern-born Black older adults – those who migrated to Northern California, survived, and enrolled in STAR. Compared to Southern-born adults nationwide, Southern-born participants were more likely to have migrated to California (e.g., 38% vs 32% of Black California residents (50+) were born in the South – authors’ calculations).⁶⁰ While we hypothesize that our participants may exhibit selective migration patterns (e.g., healthy migrant bias as supported by prior evidence of SES and mortality differentials from the Great Migration),^8,61 our findings may be understating true differences associated with Southern birth, though we cannot make this assumption. This work should be replicated within nationally representative studies to enhance generalizability and to allow for considering heterogeneity by place of residence beyond early life. Studies in larger samples would also support studying additional sources of heterogeneity (e.g., more granular regional comparisons).

Despite these limitations, our study advances evidence that the relationship between Southern birth and cognitive function among older Black adults may be partly driven by differences in educational experiences in segregated schools. This suggests that interventions aimed at reducing persistent school segregation⁶² may help ameliorate geographic inequities in late-life cognition as younger birth cohorts begin to age, although extension to nationally representative cohorts are needed. Structural inequities in current educational environments mirror those present during the historical desegregation era, with racial segregation now compounded by economic segregation,⁶³ possibly jointly impacting cognitive outcomes linked to Southern birth, as evidenced by younger cohorts.^64,65 Current evidence-based interventions – such as equitable school resource allocation, culturally responsive teaching, and policies promoting racial and socioeconomic integration^62,66–68 – are critical for school desegregation efforts in order to mitigate these disparities and support cognitive health into older adulthood. Since segregated schooling did not fully account for the geographic differences in late-life semantic memory z-scores, the relationship between Southern birth and late-life cognition appeared to also be driven by mechanisms other than attending a segregated school. This suggests that other early-life factors, such as the impacts of residential segregation, as well as downstream mechanisms in mid- and late-life (e.g., cardiovascular disease), may also contribute to cognitive aging across the life course. These factors should be the focus of future efforts at formally quantifying mediators of the relationship between Southern birth and later-life cognition.

ACKNOWLEDGMENTS

The Study of Healthy Aging in African Americans is funded by the National Institute on Aging (NIA) (R01AG050782 MPIs: Whitmer and Gilsanz). Dr. Gutierrez was supported by the NIA (F31AG081071; Gutierrez, PI and T32AG049663; Torres, PI). The funding agencies played no role in study design, data collection, analysis, and interpretation, in manuscript writing, or in the decision to submit the article for publication.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest. Author disclosures are available in the Supporting Information.

CONSENT STATEMENT

The study was approved by the KPNC and University of California Davis Institutional Review Boards and all enrolled participants provided written informed consent.

References

Alzheimer's Association. 2021 Alzheimer's disease facts and figures. Alzheimers Dement J Alzheimers Assoc. 2021;17(3):327‐406. doi:10.1002/alz.12328

Ruggles S, Flood S, Goeken R, et al. IPUMS USA: Version 10.0. Published online 2020. doi:10.18128/d010.v10.0

Kornblith E, Bahorik A, Boscardin WJ, Xia F, Barnes DE, Yaffe K. Association of race and ethnicity with incidence of dementia among older adults. JAMA. 2022;327(15):1488‐1495. doi:10.1001/jama.2022.3550

Farina MP, Crimmins EM, Hayward MD. The role of Southern context in shaping life course exposures linked to dementia incidence for Black and White older adults in the United States. BMC Geriatr. 2024;24(1):927. doi:10.1186/s12877‐024‐05508‐x

Gilsanz P, Mayeda ER, Glymour MM, Quesenberry CP, Whitmer RA. Association between birth in a high stroke mortality state, race, and risk of dementia. JAMA Neurol. 2017;74(9):1056‐1062. doi:10.1001/jamaneurol.2017.1553

Glymour MM, Kosheleva A, Wadley VG, Weiss C, Manly JJ. The geographic distribution of dementia mortality: elevated mortality rates for black and white americans by place of birth. Alzheimer Dis Assoc Disord. 2011;25(3):196‐202. doi:10.1097/WAD.0b013e31820905e7

George KM, Peterson RL, Gilsanz P, et al. Stroke Belt birth state and late‐life cognition in the Study of Healthy Aging in African Americans (STAR). Ann Epidemiol. 2021;64:26‐32. doi:10.1016/j.annepidem.2021.09.001

Glymour MM, Manly JJ. Lifecourse social conditions and racial and ethnic patterns of cognitive aging. Neuropsychol Rev. 2008;18(3):223‐254. doi:10.1007/s11065‐008‐9064‐z

Card D, Krueger AB. School resources and student outcomes: an overview of the literature and new evidence from North and South Carolina. J Econ Perspect. 1996;10(4):31‐50. doi:10.1257/jep.10.4.31

Marden JR, Tchetgen Tchetgen EJ, Kawachi I, Glymour MM. Contribution of socioeconomic status at 3 life‐course periods to late‐life memory function and decline: early and late predictors of dementia risk. Am J Epidemiol. 2017;186(7):805‐814. doi:10.1093/aje/kwx155

Whalley LJ, Dick FD, McNeill G. A life‐course approach to the aetiology of late‐onset dementias. Lancet Neurol. 2006;5(1):87‐96. doi:10.1016/S1474‐4422(05)70286‐6

De Ronchi D, Fratiglioni L, Rucci P, Paternicò A, Graziani S, Dalmonte E. The effect of education on dementia occurrence in an Italian population with middle to high socioeconomic status. Neurology. 1998;50(5):1231‐1238. doi:10.1212/wnl.50.5.1231

Brayne C, Calloway P. The association of education and socioeconomic status with the Mini Mental State Examination and the clinical diagnosis of dementia in elderly people. Age Ageing. 1990;19(2):91‐96. doi:10.1093/ageing/19.2.91

Fratiglioni L, Grut M, Forsell Y, et al. Prevalence of Alzheimer's disease and other dementias in an elderly urban population: relationship with age, sex, and education. Neurology. 1991;41(12):1886‐1892. doi:10.1212/wnl.41.12.1886

Gatz M, Svedberg P, Pedersen NL, Mortimer JA, Berg S, Johansson B. Education and the risk of Alzheimer's disease: findings from the study of dementia in Swedish twins. J Gerontol B Psychol Sci Soc Sci. 2001;56(5):P292‐P300. doi:10.1093/geronb/56.5.p292

Card D, Krueger AB. Does school quality matter? Returns to education and the characteristics of public schools in the united states. J Polit Econ. 1992;100(1):1‐40. doi:10.1086/261805

Aiken‐Morgan AT, Gamaldo AA, Sims RC, Allaire JC, Whitfield KE. Education desegregation and cognitive change in African American older adults. J Gerontol B Psychol Sci Soc Sci. 2015;70(3):348‐356. doi:10.1093/geronb/gbu153

Lamar M, Lerner AJ, James BD, et al. Relationship of early‐life residence and educational experience to level and change in cognitive functioning: results of the minority aging research study. J Gerontol B Psychol Sci Soc Sci. 2020;75(7):e81‐e92. doi:10.1093/geronb/gbz031

Peterson RL, George KM, Barnes LL, et al. Association of timing of school desegregation in the United States with late‐life cognition in the Study of Healthy Aging in African Americans (STAR) cohort. JAMA Netw Open. 2021;4(10):e2129052. doi:10.1001/jamanetworkopen.2021.29052

Whitfield KE, Wiggins SA. The impact of desegregation on cognition among older African Americans. J Black Psychol. 2003;29(3):275‐291.

Walsemann KM, Ureña S, Farina MP, Ailshire JA. Race inequity in school attendance across the Jim Crow South and its implications for Black‐White disparities in trajectories of cognitive function among older adults. J Gerontol B Psychol Sci Soc Sci. 2022;77(8):1467‐1477. doi:10.1093/geronb/gbac026

Crowe M, Clay OJ, Martin RC, et al. Indicators of childhood quality of education in relation to cognitive function in older adulthood. J Gerontol A Biol Sci Med Sci. 2013;68(2):198‐204. doi:10.1093/gerona/gls122

Liu SY, Manly JJ, Capistrant BD, Glymour MM. Historical differences in school term length and measured blood pressure: contributions to persistent racial disparities among US‐Born adults. PloS One. 2015;10(6):e0129673. doi:10.1371/journal.pone.0129673

Frisvold D, Golberstein E. School quality and the education–health relationship: evidence from Blacks in segregated schools. J Health Econ. 2011;30(6):1232‐1245. doi:10.1016/j.jhealeco.2011.08.003

Rothstein R. The racial achievement gap, segregated schools, and segregated neighborhoods: a constitutional insult. Race Soc Probl. 2015;7(1):21‐30. doi:10.1007/s12552‐014‐9134‐1

Hyman J. Does money matter in the long run? Effects of school spending on educational attainment. Am Econ J Econ Policy. 2017;9(4):256‐280. doi:10.1257/pol.20150249

Jackson CK, Johnson RC, Persico C. The effects of school spending on educational and economic outcomes: evidence from school finance reforms *. Q J Econ. 2016;131(1):157‐218. doi:10.1093/qje/qjv036

Leigh JP. Direct and indirect effects of education on health. Soc Sci Med. 1983;17(4):227‐234. doi:10.1016/0277‐9536(83)90120‐X

Rodgers HR, Bullock CS. School desegregation: a policy analysis. J Black Stud. 1972;2(4):409‐437.

United States Commision on Civil Rights. Southern School Desegregation, 1966‐1967.; 1967. Accessed January 30, 2024. https://www2.law.umaryland.edu/marshall/usccr/documents/pressrel67.pdf

Thompson O. School desegregation and black teacher employment. Rev Econ Stat. 2022;104(5):962‐980. doi:10.1162/rest_a_00984

Gutierrez S, Whitmer RA, Soh Y, et al. School‐based racial segregation, social support, and late‐life cognitive function in the Study of Healthy Aging in African Americans (STAR). Alzheimers Dement. 2024;20(9):6257‐6267. doi:10.1002/alz.14112

Myers HF. Ethnicity‐ and socio‐economic status‐related stresses in context: an integrative review and conceptual model. J Behav Med. 2009;32(1):9‐19. doi:10.1007/s10865‐008‐9181‐4

McEwen BS. Stress, adaptation, and disease. Allostasis and allostatic load. Ann N Y Acad Sci. 1998;840:33‐44. doi:10.1111/j.1749‐6632.1998.tb09546.x

Lupien SJ, McEwen BS, Gunnar MR, Heim C. Effects of stress throughout the lifespan on the brain, behaviour and cognition. Nat Rev Neurosci. 2009;10(6):434‐445. doi:10.1038/nrn2639

Brody GH, Lei MK, Chae DH, Yu T, Kogan SM, Beach SRH. Perceived discrimination among African American adolescents and allostatic load: a longitudinal analysis with buffering effects. Child Dev. 2014;85(3):989‐1002. doi:10.1111/cdev.12213

Skinner ML, Shirtcliff EA, Haggerty KP, Coe CL, Catalano RF. Allostasis model facilitates understanding race differences in the diurnal cortisol rhythm. Dev Psychopathol. 2011;23(4):1167‐1186. doi:10.1017/S095457941100054X

Vyas S, Rodrigues AJ, Silva JM, et al. Chronic stress and glucocorticoids: from neuronal plasticity to neurodegeneration. Neural Plast. 2016:6391686. doi:10.1155/2016/6391686 2016

Bauer ME. Stress, glucocorticoids and ageing of the immune system. Stress Amst Neth. 2005;8(1):69‐83. doi:10.1080/10253890500100240

Brotman DJ, Golden SH, Wittstein IS. The cardiovascular toll of stress. Lancet Lond Engl. 2007;370(9592):1089‐1100. doi:10.1016/S0140‐6736(07)61305‐1

Logan JR. The persistence of segregation in the 21st Century Metropolis. City Community. 2013;12(2). doi:10.1111/cico.12021

Logan JR, Minca E, Adar S. The geography of inequality: why separate means unequal in American public schools. Sociol Educ. 2012;85. doi:10.1177/0038040711431588

Knudsen EI. Sensitive periods in the development of the brain and behavior. J Cogn Neurosci. 2004;16(8):1412‐1425. doi:10.1162/0898929042304796

Walsemann KM, Kerr EM, Ailshire JA, Herd P. Black‐White variation in the relationship between early educational experiences and trajectories of cognitive function among US‐born older adults. SSM—Popul Health. 2022;19:101184. doi:10.1016/j.ssmph.2022.101184

VanderWeele TJ. Mediation analysis: a practitioner's guide. Annu Rev Public Health. 2016;37:17‐32. doi:10.1146/annurev‐publhealth‐032315‐021402

Hernán MA, Robins JM. Causal Inference: What If. Chapman & Hall/CRC Press; 2024. https://www.hsph.harvard.edu/miguel‐hernan/wp‐content/uploads/sites/1268/2024/01/hernanrobins_WhatIf_2jan24.pdf

US Census Bureau. Census Regions and Divisions of the United States. Census.gov. 2010. Accessed March 18, 2022. https://www.census.gov/geographies/reference‐maps/2010/geo/2010‐census‐regions‐and‐divisions‐of‐the‐united‐states.html

Mungas D, Reed BR, Haan MN, González H. Spanish and English neuropsychological assessment scales: relationship to demographics, language, cognition, and independent function. Neuropsychology. 2005;19(4):466‐475. doi:10.1037/0894‐4105.19.4.466

Mungas D, Reed BR, Crane PK, Haan MN, González H. Spanish and English Neuropsychological Assessment Scales (SENAS): further development and psychometric characteristics. Psychol Assess. 2004;16(4):347‐359. doi:10.1037/1040‐3590.16.4.347

Vable AM, Gilsanz P, Nguyen TT, Kawachi I, Glymour MM. Validation of a theoretically motivated approach to measuring childhood socioeconomic circumstances in the Health and Retirement Study. PLoS ONE. 2017;12(10):e0185898. doi:10.1371/journal.pone.0185898

Pearl J. The causal mediation formula–a guide to the assessment of pathways and mechanisms. Prev Sci Off J Soc Prev Res. 2012;13(4):426‐436. doi:10.1007/s11121‐011‐0270‐1

StataCorp LLC. mediate. Accessed 2024. https://www.stata.com/manuals/causalmediate.pdf

Imai K, Keele L, Tingley D. A general approach to causal mediation analysis. Psychol Methods. 2010;15(4):309‐334. doi:10.1037/a0020761

Hall KS, Gao S, Unverzagt FW, Hendrie HC. Low education and childhood rural residence: risk for Alzheimer's disease in African Americans. Neurology. 2000;54(1):95‐99. doi:10.1212/wnl.54.1.95

Walsemann KM, Fisk CE, Farina MP, Abbruzzi E, Ailshire JA. Race, gender, and cohort differences in the educational experiences of Black and White Americans. Popul Res Policy Rev. 2023;42(6):86. doi:10.1007/s11113‐023‐09831‐w

Wilkerson I. The Warmth of Other Suns: The Epic Story of America's Great Migration. Knopf Doubleday Publishing Group; 2010.

Gutierrez S, Whitmer RA, Thomas MD, et al. Evaluating administrative measures of school quality as mediators of the relationship between attending a segregated school and cognitive function among older Black individuals: the STAR Study. Am J Epidemiol. 2025;194(11):3238‐3247. doi:10.1093/aje/kwaf150 Published online published online ahead of print

Soh Y, Whitmer RA, Mayeda ER, et al. State‐Level indicators of childhood educational quality and incident dementia in older Black and White Adults. JAMA Neurol. 2023;80(4):352‐359. doi:10.1001/jamaneurol.2022.5337

Walsemann KM, Hair NL, Farina MP, Tyagi P, Jackson H, Ailshire JA. State‐level desegregation in the U.S. South and mid‐life cognitive function among Black and White adults. Soc Sci Med. 1982;338:116319. doi:10.1016/j.socscimed.2023.116319 2023

Ruggles S, Flood S, Sobek M, et al. IPUMS USA: Version 16.0 [2021 ACS 1‐year]. Published online 2025. doi:10.18128/D010.V16.0

Black DA, Sanders SG, Taylor EJ, Taylor LJ. The impact of the Great Migration on mortality of African Americans: evidence from the Deep South. Am Econ Rev. 2015;105(2):477‐503. doi:10.1257/aer.20120642

Reardon SF, Grewal ET, Kalogrides D, Greenberg E. Brown fades: the end of court‐ordered school desegregation and the resegregation of American public schools. J Policy Anal Manage. 2012;31(4):876‐904. doi:10.1002/pam.21649

Orfield G, Frankenberg E, Ee J, Kucsera J, Brown at 60: Great Progress, a Long Retreat and an Uncertain Future. Civ Rights Proj‐Proy Derechos Civ UCLA. Published online. 2014. http://escholarship.org/uc/item/0dn6z3vc

Wang G, Schwartz GL, Kim MH, et al. School racial segregation and the health of black children. Pediatrics. 2022;149(5):e2021055952. doi:10.1542/peds.2021‐055952

Kim MH, Schwartz GL, White JS, et al. School racial segregation and long‐term cardiovascular health among Black adults in the US: a quasi‐experimental study. PLOS Med. 2022;19(6):e1004031. doi:10.1371/journal.pmed.1004031

Johnson RC. Long‐run Impacts of school desegregation & school quality on adult attainments. Natl Bur Econ Res Work Pap Ser. 2011. doi:10.3386/w16664 No. 16664

Himmelstein KEW, Lawrence JA, Jahn JL, et al. Association between racial wealth inequities and racial disparities in longevity among US adults and role of reparations payments. JAMA Netw Open. 2022;5(11):e2240519. doi:10.1001/jamanetworkopen.2022.40519 1992 to 2018

Bottiani JH, McDaniel HL, Henderson L, Castillo JE, Bradshaw CP. Buffering effects of racial discrimination on school engagement: the role of culturally responsive teachers and caring school police. J Sch Health. 2020;90(12):1019‐1029. doi:10.1111/josh.12967