Abstract
OBJECTIVES: To develop and validate the Anhui Environmental Exposure Questionnaire (AEEQ) and evaluate whether its composite score independently predicts incident cerebral haemorrhage in older adults. METHODS: In a prospective cohort (January 2022-April 2024) from Hefei, we enrolled 460 participants aged 60-79 years with ≥5 years' residence and MMSE ≥24, excluding prior stroke/ICH/TBI, heavy-industry/mining workers, lifetime smoking ≥5 pack-years, and recent (≤6 months) antihypertensive changes. The 42-item AEEQ spans six domains. Content validity used Delphi procedures; construct validity used split-sample EFA/CFA; reliability used Cronbach's α, split-half, and 14 ± 3-day ICCs; criterion validity correlated domains with annual-mean residential PM₂.₅ (calendar year prior to baseline) and PSQI. Incident haemorrhage was ascertained over ≈24 months; Cox models adjusted for age, sex, hypertension, alcohol, and anticoagulation tested the AEEQ (Z-score), with a prespecified antihypertensive interaction. RESULTS: Content validity was high (S-CVI/Ave 0.96). Factorability was adequate (KMO 0.91); EFA supported six factors (66.1% variance). CFA fit was excellent (χ(2)/df 2.19; CFI 0.965; TLI 0.958; RMSEA 0.041; SRMR 0.052). Reliability was strong (total α 0.90; domain α 0.78-0.86; total ICC 0.86). Criterion validity was consistent with hypotheses (air-pollution domain vs. PM₂.₅ r = 0.62; sleep domain vs. PSQI ρ = -0.56). Sixteen haemorrhages occurred (3.5%; ~920 person-years), with monotonic incidence across AEEQ quartiles (0.9, 1.7, 2.6, 8.7%; p-trend = 0.0047). Each 1-SD higher AEEQ predicted greater risk (HR 1.47, 95% CI 1.12-1.93) and improved discrimination beyond clinical covariates (C-statistic 0.72 → 0.80; NRI 0.15, p = 0.045); effects were stronger in participants not using antihypertensives (interaction p = 0.048). Equity profiling showed higher AEEQ scores in lower education and renters. CONCLUSION: The AEEQ reliably quantifies chronic environmental burden and adds independent, dose-responsive prognostic value for cerebral haemorrhage, supporting its use in community screening and targeted prevention.