Greenspace redevelopment, pressure of displacement, and sleep quality among Black adults in Southwest Atlanta

The study addresses how greenspace redevelopment in historically disinvested, predominantly Black neighborhoods may interact with social environmental stressors linked to residential displacement to affect sleep quality among Black adults. While parks and greenspaces can confer health benefits, green redevelopment can catalyze "green gentrification," increasing property values and displacement pressures. Displacement is conceptualized to include economic, exclusionary, and displacement pressure (fear and actuality of neighborhood changes). Black communities face disproportionate environmental inequities and social stressors, including discrimination, vigilance in anticipation of discrimination, housing unaffordability, and financial strain—each of which can adversely affect sleep. Leveraging the Atlanta BeltLine redevelopment as a natural experiment, the study aims to examine whether these stressors are associated with poorer subjective sleep quality and whether associations differ between residents exposed versus unexposed to greenspace redevelopment. The authors hypothesized: (1) social environmental stressors (individually and combined) correlate with poorer sleep quality; (2) associations differ by exposure to greenspace redevelopment; and (3) exposed participants have poorer sleep quality.

The paper situates the research within evidence that greenspaces are linked to health benefits (more activity, lower stress, longer sleep duration) and environmental improvements. However, BIPOC and low-income neighborhoods have limited access to such amenities and are more exposed to pollutants due to residential segregation. Green infrastructure initiatives can spur "green gentrification," raising housing costs and displacing vulnerable residents. Social stressors—discrimination and social exclusion in new spaces, heightened vigilance to anticipated discrimination, housing unaffordability, and financial strain—may be amplified during redevelopment and are associated with poor sleep outcomes. Prior studies document links between financial stress, housing instability, discrimination, racism-related vigilance, and sleep difficulties, with Black individuals reporting higher vigilance and sleep difficulty compared to White individuals.

Design: Pilot, cross-sectional, quasi-experimental study (African American Sleep & Health [AASH] Study) comparing Black adults in neighborhoods exposed versus unexposed to greenspace redevelopment associated with the Atlanta BeltLine.

Exposure definition and site selection: A displacement risk index (developed using a structural racism lens) identified census block groups within the BeltLine target area with residents most susceptible to displacement. Composite scores were computed using z-scores for vulnerability and housing market indicators across 2000, 2007–2011, and 2012–2016, transformed to positive values, and categorized into quartiles (low to high displacement risk). Two block groups within the BeltLine area classified as medium to high risk (100% non-Hispanic Black residents; located along the 3-mile Westside trail built 2014–2017) were selected as exposed. Contextual evidence (rising median rents, higher cost burden, eviction, crime, vacancy; and observed juxtaposition of abandoned, rehabilitating, and newly constructed properties) supported displacement pressure.

Matching of comparison areas: Propensity score matching (PSM) was used to pair exposed block groups with unexposed City of Atlanta block groups not affected by the BeltLine. Indicators from the 2012–2016 ACS (economic, racial composition, population, and housing characteristics) informed matching. Propensity scores were estimated using percent below poverty as the outcome (sleep data unavailable at block-group level). Exclusion buffer: block groups within 1 mile of exposed areas were excluded to reduce spillover. Exposed block groups were matched 3:1 with replacement to unexposed block groups with propensity scores within ±0.01. Overlap was confirmed via box plots. Descriptive characteristics and PSM results are provided in supplementary materials.

Sampling: Two-stage cluster sampling was used within 4 block groups (2 exposed, 2 unexposed). Primary sampling units were blocks; secondary units were households. The sampling frame comprised 374 households (143 exposed; 231 unexposed). Differences in recruited households reflected larger blocks and fewer abandoned properties in unexposed areas; one unexposed site was predominantly apartment complexes (98% Black tenancy), so sampling included a fully occupied complex.

Household selection: For each selected block, one of four corners was randomly chosen as a start; recruiters proceeded clockwise, approaching every third household (or building in complexes). Apparent abandoned, for-sale, or Airbnb properties (identified by neighbors/tenants or keyless locks) were skipped and replaced with adjacent households.

Recruitment and eligibility: Up to five contact attempts per household. Initial advance letter and flyer included a website link with prescreening (Qualtrics) covering race, age, gender, prior sleep disorder diagnosis, and financial decision-maker status. On follow-up visits, in-person prescreening was administered when contact was made. Inclusion criteria: age ≥18, self-identified African American/Black, head of household or joint financial decision-maker; exclusion: prior diagnosed sleep disorder. Eligible participants provided written informed consent and completed an online survey (SurveyMonkey). Incentive: $25 gift card.

Data collection period: June–October 2019. IRB approvals: Florida State University (designated IRB) and Morehouse School of Medicine.

Measures: Social environmental stressors included everyday discrimination, heightened vigilance (racism-related vigilance), housing unaffordability (assessed among renters), and financial strain. Outcome: subjective sleep quality. Covariates included sociodemographic factors (details in survey; specifics not provided in the excerpt). Greenspace redevelopment exposure: residence in matched exposed vs unexposed block groups.

Statistical analysis: Linear regression models assessed associations between each stressor and sleep quality, testing effect modification by greenspace redevelopment status. A combined model included multiple stressors simultaneously. Listwise deletion was used for missing data (<3.03% for all variables except financial strain, which had 31.8% missing). Housing affordability analyses included renters only.

Sample: Final analytic sample N=66 Black adults (42 unexposed; 24 exposed), representing 18% of approached residents. Each participant received a unique survey ID to protect anonymity.

• Independent models: Associations between everyday discrimination, heightened vigilance, and housing unaffordability with subjective sleep quality were not modified by greenspace redevelopment exposure after adjustment.
• Financial strain: The association with subjective sleep quality differed by exposure status; exposed participants experiencing financial strain had poorer sleep quality compared with unexposed participants.
• Combined model: The financial strain–sleep quality association persisted when accounting for multiple stressors. The direction and magnitude varied across financial strain categories, with exposed participants sleeping poorer and/or better than unexposed participants depending on the strain category, indicating a nuanced pattern.
• Sample/context: N=66 (42 unexposed; 24 exposed); study area centered on Atlanta BeltLine redevelopment zones and matched unexposed neighborhoods.

Findings partially support the hypotheses. Social environmental stressors are relevant to sleep quality among Black adults living in neighborhoods affected by redevelopment, but greenspace redevelopment did not uniformly modify associations for discrimination, vigilance, or housing unaffordability. Financial strain emerged as the key stressor whose relationship with sleep differed by redevelopment exposure, suggesting that displacement pressure linked to redevelopment may exacerbate the sleep-related impact of financial hardship. The nuanced patterns across financial strain categories in the combined model point to heterogeneous experiences of redevelopment, possibly reflecting varying coping resources, stages of neighborhood change, or differences in cost burdens. Overall, results highlight that redevelopment initiatives can intersect with existing structural inequities, shaping stress pathways that influence sleep among Black residents.

This pilot quasi-experimental study contributes evidence that the relationship between social environmental stressors and sleep quality among Black adults varies with exposure to greenspace redevelopment, with financial strain showing differential associations by exposure. The results underscore the need for equity-focused planning in redevelopment efforts and for public health strategies that mitigate displacement pressures and associated stress.

Future research should employ larger, longitudinal designs across multiple redevelopment contexts to clarify causal pathways, explore mechanisms underlying the heterogeneous effects of financial strain, and evaluate policy interventions (e.g., affordable housing protections) that may buffer sleep-related harms during redevelopment.

• Pilot, cross-sectional design limits causal inference.
• Small sample size (N=66; 24 exposed, 42 unexposed) reduces statistical power and precision.
• Missing data handled via listwise deletion; financial strain had substantial missingness (31.8%), potentially introducing bias.
• Housing affordability items were asked only of renters, reducing sample size for those analyses and limiting generalizability to homeowners.
• Potential selection bias due to 18% participation rate.
• Exposure assignment at the block-group level with only two exposed block groups may limit generalizability and increase susceptibility to unmeasured neighborhood confounding.
• Self-reported measures of stressors and sleep quality may be subject to reporting bias.
• Study limited to specific neighborhoods in Atlanta; findings may not generalize to other settings.