Associations between neighborhood built environment and blood pressure levels among residents and joint exposure profiles of built environment factors
- VernacularTitle:建成环境与居民血压水平的关联及其联合暴露特征
- Author:
Huiqin WANG
1
;
Kangkang GU
1
;
Peihong LI
1
;
Haoliang WANG
1
;
Yucheng HUANG
1
Author Information
- Publication Type:Investigation
- Keywords: built environment; blood pressure; multi-model comparison; joint exposure; machine learning
- From: Journal of Environmental and Occupational Medicine 2026;43(9):1123-1131
- CountryChina
- Language:Chinese
- Abstract: Background Chronic non-communicable diseases, including hypertension, have become major threats to urban public health. The built environment has been linked to blood pressure. Objective To examine joint exposure patterns of built environment factors in relation to residents' blood pressure status and to identify environmental profiles of neighborhoods that may be favorable or unfavorable to hypertension prevention. Methods (1) A cross-sectional study was conducted among 1910 residents in Longzihu District, Bengbu City. Demographic characteristics, clinical indicators (including blood pressure), and lifestyle behavioral data were collected. Nine built environment indicators were derived within a 500-meter buffer zone surrounding each participant's residential address. (2) Using elevated blood pressure status as a binary outcome (all residents) and grade of elevated blood pressure as an ordinal multi-class outcome (residents with elevated blood pressure), the predictive performance of four binary classification models and nine multi-class classification models was compared, respectively. Feature importance was ranked based on the best-performing model. Multi-variable logistic regression with stepwise adjustment was then used to identify key environmental factors. (3) Using partial dependence plots, a joint exposure analysis was conducted in the target population of residents with elevated blood pressure (high-normal blood pressure and hypertension grades 1–3), with the 25th and 75th percentiles of key environmental factors as the low- and high-level thresholds, respectively. The 95% confidence intervals were estimated using the Bootstrap method. Results (1) Among the study participants, 863 had elevated blood pressure: 349 (40.4%) had high-normal blood pressure, and 398 (46.1%), 93 (10.8%), and 23 (2.7%) had hypertension grades 1, 2, and 3, respectively. (2) Logistic regression showed the best performance in analyzing the association between built environment factors and elevated blood pressure status (n=1910), whereas random forest showed the best performance in analyzing the association between built environment factors and grade of elevated blood pressure (n=863). Road density, bus stop density, and hospital/clinic density were key environmental factors. (3) Pairwise joint exposure patterns were observed among the three key environmental factors: When any two were both below the low-level thresholds (road ≤0.54 km·km−2, bus stop ≤0.51 points·km−2, hospital ≤5.99 points·km−2), the risk of a higher grade of elevated blood pressure was greater than the overall sample average (relative predicted odds 1.29–1.31, 95% confidence interval lower bound >1); when any two were both above the high-level thresholds (road ≥0.75 km·km−2, bus stop ≥1.25 points·km−2, hospital ≥15.00 points·km−2), the risk was lower than the overall sample average (relative predicted odds 0.54–0.61, 95% confidence interval upper bound <1); low-high and high-low combinations did not differ significantly from the overall sample average (P > 0.05). Conclusion Joint exposure profiles characterized by road density, bus stop density, and hospital/clinic density are associated with elevated blood pressure among residents: Simultaneously low levels of any two of these environmental indicators are associated with a higher predicted risk of a more severe grade, whereas simultaneously high levels are associated with a lower predicted risk. These findings suggest that coordinated configurations of road networks, public transit accessibility, and medical service facilities may represent important environmental feature of neighborhoods favorable to hypertension prevention and control.
