Factors Linking Neighborhood Variables and Stroke Recovery Remain Elusive

Modifiable factors underlying the association between neighborhood socioeconomic status (nSES) and stroke recovery remained elusive, as variables that were investigated to account for differences proved to be unrelated.

"Previous studies across multiple settings have consistently shown that the SES of where a stroke survivor lives is associated with 90-day stroke recovery outcomes," wrote Eric Stulberg, MD, MPH, of Thomas Jefferson University in Philadelphia, and co-authors in JAMA Network Open.

"In this cohort study of a bi-ethnic urban population with predominantly mild strokes, we did not find evidence that receipt of hyperacute treatment (mainly medical thrombolysis), density of home health services, density of rehabilitation clinics, density of recreation centers, transportation access, or walkability mediate the association between lower nSES and worse 90-day stroke recovery outcomes," Stulberg and colleagues reported.

"Aside from individual receipt of hyperacute treatment, our candidate mediators were primarily neighborhood-level factors, and we urge caution against interpreting to the individual level (the ecological fallacy)," they cautioned.

Neighborhood was defined as census tract in the study, and data were from the Brain Attack Surveillance in Corpus Christi (BASIC) cohort, with enrollment between 2009 and 2022 in Nueces County, Texas. The researchers evaluated information about 2,203 participants who had acute ischemic stroke.

The median age in the study was 66 years and 47.4% of participants were women. Most participants were described as Mexican American (59.3%); 33.5% were White, and 7.2% were other races and ethnicities. The median stroke severity in the study, based on National Institute of Health Stroke Scale scores, was 3 (NIHSS scores can range from 0, no deficit, to 42, severe impairment). The NIHSS score was 5 or less in 70.2% of the overall cohort.

The analysis focused on possible mediator associations with three 90-day outcomes: functional status, measured with questionnaires for activities of daily living (ADL) and instrumental ADL (IADL); depressive symptom burden, evaluated with the Patient Health Questionnaire (PHQ–8); and quality of life (QoL), evaluated with the Stroke-Specific QoL questionnaire (SS-QoL).

The researchers conducted causal mediation analyses to determine how variables like hyperacute treatment, postacute care resource density, transportation access, or walkability might mediate the association between lower nSES and worse 90-day stroke recovery.

Across all measures, a higher nSES was associated with better outcomes:

• PHQ-8 score: β = −1.21, 95% CI −1.86 to −0.56
• ADL-IADL score: β = −0.20, 95% CI −0.27 to −0.13
• SS-QoL score: β = 0.20, 95% CI 0.11-0.29

No variable had a significant effect on outcomes. "Shifting all mediator distributions from low-nSES to high-nSES neighborhoods modestly attenuated the association of nSES with PHQ-8 score by 14.1% (95% CI −36.3% to 64.5%) and accounted for 15.1% (95% CI −11.0% to 41.2%) of the association of nSES with ADL-IADL score and 5.6% (95% CI −25.4% to 36.6%) of the association of nSES with SS-QoL score," Stulberg and colleagues wrote.

"There was a potentially clinically meaningful but not statistically significant finding of rehabilitation clinic density and home health care service density acting as mediators between nSES and PHQ-8 and ADL-IADL score, respectively, as approximately 15% to 20% of the association between nSES with PHQ-8 and ADL-IADL scores were mediated by those respective variables," they added.

Factors not included in the study may have influenced poststroke recovery variation; this could include disposition after acute hospitalization; time to hyperacute treatment; timing, cost, and dose of poststroke therapy; nutritional status; and recurrent stroke, the researchers noted. "Caregiver and family-unit support may be patterned by nSES and exert a stronger influence than neighborhood-level factors," they observed.

"It remains unclear what factors could be targeted therapeutically to mitigate nSES-associated poststroke recovery differences," they stated. "Future studies should replicate our framework in a variety of populations and include other potential mediators to help inform interventions aimed at reducing differences in poststroke recovery by the SES of where someone lives."

Stulberg and colleagues pointed out prior work showing substantial variation in recovery outcomes by nSES among those receiving thrombectomy. A study published in 2025 showed that living in a more socioeconomically affluent neighborhood was associated with a more favorable functional outcome at 90 days. A 2026 study of patients in Spain showed that functional outcomes were less likely and onset-to-thrombectomy times were longer in more deprived areas.

Findings may not be generalizable outside of Nueces County, Texas, and that the definition of neighborhood-as-census-tract "may not fully represent the lived space for survivors of stroke in an urban locale," Stulberg and co-authors acknowledged.

Most people in the study had minor stroke, which may have reduced estimates of mediator effect since there is generally good recovery after minor stroke. "We urge replication in populations with higher rates of thrombectomy," the researchers suggested.

This study was funded by National Institute of Neurological Disorders and Stroke.

Stulberg reporrted no conflicts of interest.

Stulberg EL, et al. "Factors underlying stroke recovery variation by neighborhood socioeconomic status" JAMA Netw Open 2026; DOI: 10.1001/jamanetworkopen.2026.16362.