The residential environment and blood pressure changes in adults: a longitudinal multi-cohort study

Abstract We aimed to study the associations between air pollution, noise and green spaces and longitudinal changes in systolic blood pressure (SBP) in three general population European cohorts: Constances, the Northern Finland Birth Cohort (NFBC1966) and the Dutch Famine Birth Cohort (DFBC). We included adults with no history of cardiovascular diseases and with at least two SBP measurements. We assessed air pollution exposure using yearly average PM2.5 and NO2 concentrations (in µg/m3); noise using daily average noise levels (Lden); and surrounding greenness using the Normalized Difference Vegetation Index (NDVI) within 300m from residence. For each participant we calculated the rate of change of SBP between baseline and follow-up, and we used linear regression to evaluate the associations with the residential environment in single-exposure models. We used weights to compensate for unavailable information at follow-up. Covariates were age, sex, smoking status, antihypertensive drugs, education, and occupation. We also included an interaction term with urban level (ie impervious density). We included 20,513 participants from Constances, 3,306 from NFBC1966, and 508 from DFBC, all with two SBP measures (mean age intervals 47-56, 31-47, and 50-58, respectively). SBP was overall higher at follow-up vs baseline. We found the highest mean levels of air pollution and noise in the DFBC (eg PM2.5: 30.6µg/m3±2.7), and of greenness in the NFBC1966 (0.58±0.11). After adjustment associations were overall close to the null and non-significant (eg in Constances, per 5 µg/m3 in PM2.5: β:0.06, 95%CI: -0.15,0.27; and per 0.1-unit NDVI: β:0.02, 95%CI: -0.07,0.10). However, in the NFBC1966, they were significant and in the opposite expected direction (eg PM2.5: β: -0.29, 95%CI: -0.43, -0.14). In a multi-cohort study of 3 EU countries the residential environment did not seem to play a major role in SBP changes. Future studies may use longer follow-ups and more SBP data to better capture changes. Key messages • Using three adult cohorts, we aimed to study if exposure to the residential living environment could slow down or speed up the rate of change in systolic blood pressure between baseline and follow-up. • Associations were overall close to the null and non-significant; we recommend that future studies use longer follow-ups to include more SBP measures and to better capture exposure trajectories.