A systematic review of causal pathways of socioeconomic inequalities in stroke

Socioeconomic status (SES), often measured by education, income, occupation, or area-level deprivation, impacts stroke incidence and outcomes, yet the underlying mechanisms remain unclear. This review synthesizes causal analyses quantifying drivers of these inequalities.

We conducted a systematic review (PROSPERO CRD42024554285) and reported following the PRISMA-2020 guidelines. Observational studies applying causal mediation analysis between SES and stroke risk, disability, or mortality were included from PubMed, Embase, Scopus, and Google Scholar. SES indicators, outcomes, mediators, and decompositions into natural direct effect (NDE) and natural indirect effect (NIE) were extracted. Risk of bias and certainty of evidence were assessed using ROBINS-E and GRADE. A narrative synthesis was undertaken, and findings were illustrated in causal diagrams.

Of 12,034 records, 19 studies (15 in high-income countries) were included. Lower SES increased stroke incidence through hypertension (NIE 14-21% of the total effect, moderate certainty), although one study restricted to women reported smaller effects (2-4%). Smoking (6-19.9%, very low certainty). At 3 months post-stroke, the combined outcome of death or disability was higher due to severe strokes (38.5% for ischemic, 57-94% for hemorrhagic, moderate certainty). One study found that hypertension, atrial fibrillation, and smoking together mediated 28.5% of the SES effect on stroke severity (low certainty). Reduced access to thrombolysis and stroke units mediated 2.7% of 3-month disability/mortality (very low certainty), while greater distance to specialized centers explained 48% of inequalities in thrombectomy access (low certainty). Long-term mortality (⩾6 months) was mediated by comorbidities (18%) and healthcare coverage (24-55%), both with low certainty.

Hypertension, smoking, and differential stroke severity at presentation are the main pathways through which low SES increases stroke risk and causes worse outcomes. Targeting these may reduce inequalities, though evidence from low-income settings and emerging mediators (e.g. early-life SES, environmental exposures, care quality) is lacking.