School‐based adult social support in segregated schools and late‐life cognitive function in the study of healthy aging in african americans (STAR).

Background: It is unclear how school segregation, a form of structural racism, affects late‐life cognition. Though segregated schools were often under‐resourced, they likely shielded Black students from interpersonal racism present at integrated schools. Social support may mediate and/or modify the relationship between school segregation and cognition. Method: 467 STAR participants aged 50+ reported if they attended a segregated school in 1st and 6th grade and received grade‐specific school‐based social support (feeling cared for by an adult at school). Baseline executive function (EF), semantic memory (SM), and verbal episodic memory (VEM) were measured using the Spanish and English Neuropsychological Assessment Scales. For each grade, linear regression models evaluated the association of segregated school attendance with each cognitive domain, sequentially adjusting for demographics, school‐based social support, and education (in pooled models). Effect modification by school‐based social support was examined through stratified models and interaction terms. Result: The mean age was 68.2 (Table 1). Participants attending a segregated school in either grade were more likely to report school‐based social support than those in integrated schools (both p‐values<0.05; Figure 1). For each grade, attending a segregated vs integrated school was not associated with EF or VEM but was associated with lower SM (β(95% CI): ‐0.28(‐0.48, ‐0.08); Figure 2); results were similar when adjusting for social support. For each grade, effect estimates of attending a segregated school on SM and EF were smaller among those with social support than those without, though differences were not statistically significant (interaction p‐values>0.05;Table 2). In first grade, attending a segregated school was associated with SM among those without social support (β(95% CI): ‐0.40(‐0.78, ‐0.02)) but not among those with social support (β(95% CI): ‐0.10(‐0.36, 0.16)). The effect of attending a segregated school on VEM was not lower among those with social support in 1st or 6th grade than their counterparts without it. Stratified results were similar when adjusting for education. Conclusion: Having attended a segregated school may have both harmful and protective effects on late‐life cognition. Our findings suggest that potential harmful effects of having attended segregated schools may be diminished by the increased likelihood of school‐based social support. [ABSTRACT FROM AUTHOR]