TL;DR
This study investigates how housing and community conditions impact multidimensional health in Hong Kong, particularly for middle- and low-income groups. It reveals that these environmental factors significantly affect various health outcomes, with lower-income populations being disproportionately affected by poor living conditions. The findings highlight the need for targeted policies to improve urban environments for vulnerable populations.
What Problem Does This Paper Address?
Urbanization, while bringing progress, often leads to deteriorating housing and community conditions, such as poor ventilation and a lack of public spaces. These factors are increasingly recognized as critical social determinants of health. However, much of the existing research has focused on unidimensional health outcomes and has not adequately explored how these environmental issues affect different income groups. This paper aims to fill this gap by examining the multidimensional health impacts of housing and community problems across various income strata in Hong Kong, a highly urbanized city with significant income inequality and housing challenges.
Methods
The study employed a quantitative approach using cross-sectional survey data. To identify significant relationships between numerous housing/community problems and multiple health outcomes, the researchers utilized a Lasso regression for initial variable selection. Lasso is a statistical technique that performs feature selection by shrinking the coefficients of less important predictors to zero, effectively identifying the most relevant variables.
Following the Lasso selection, Ordinary Least Squares (OLS) and logistic regressions were used to model the relationships. Linear regression was applied for continuous health indicators like PCS and MCS scores, while logistic regression was used for binary outcomes such as self-reported health status, chronic disease, and sleep problems. To account for potential multicollinearity among the explanatory variables, a generalized Lasso linear regression was also employed. The analysis was conducted separately for the entire population and for a sub-sample of middle- to low-income individuals to compare their distinct experiences.
Datasets
The study utilized data from the first wave of the Strategic Public Policy Research (SPPR) project called "Trends and Implications of Poverty and Social Disadvantages in Hong Kong: A Multi-disciplinary and Longitudinal Study." This dataset was collected through face-to-face interviews between May 2014 and July 2015. The final sample for analysis comprised 1785 observations. The data included information on housing conditions, community problems, and various health indicators.
Key Insights and Ideas
- Multidimensionality Matters: A unidimensional approach to health is insufficient. Different housing and community problems affect various health dimensions (physical, mental, chronic disease, sleep) in distinct ways, and a holistic view is crucial for understanding the full impact.
- Income Disparity Amplifies Impact: The middle- to low-income group is not only more exposed to poor living conditions but also experiences a more severe health impact from these conditions compared to the general population. Even a single reported problem can have a larger negative effect on their health.
- Targeted Interventions Needed: A "one-size-fits-all" policy approach to improving urban environments is likely to be inefficient. Identifying specific, significant problems for vulnerable groups, like the middle- to low-income population, allows for more precise and impactful policy interventions.
- Built Environment as a Social Determinant: The study reinforces the idea that the built environment is a powerful social determinant of health, influencing well-being beyond individual choices or behaviors.
Key Results
- Whole Population: Various housing conditions like "rot in window frames or floors" and "poor ventilation" were linked to increased DASS-Anxiety and DASS-Stress scores. Community issues such as "traffic or business noise" and "drunk or rowdy people in the street/park" also correlated with higher anxiety and stress levels. Shortage of space and air pollution were associated with poorer self-reported health.
- Middle- to Low-Income Group: This group showed a stronger sensitivity to specific conditions. For instance, they were more likely to experience chronic diseases when exposed to being "too hot in the summer/too cold in the winter" and "noisy neighbors or loud parties," whereas the general population showed no such link.
- Severity of Impact: For the middle- to low-income group, the absolute values of the significant coefficients were larger, indicating that each identified problem had a more profound impact on their health outcomes compared to the general population.
- Differential Sensitivity: While the middle- to low-income group was more severely affected, they were sensitive to fewer community/housing conditions overall than the general population, suggesting a more concentrated impact of specific issues on their health.
Limitations and Future Work
The authors acknowledge several limitations:
- Cross-sectional Data: The study relies on cross-sectional data, meaning it cannot establish causality or temporal relationships between housing conditions and health outcomes. Future longitudinal studies are recommended to clarify these causal pathways.
- Self-Reported Data: Both health measures and housing/community conditions were self-reported, which can introduce bias. Future research could validate these self-reports with objective measures.
- Focus on Poor Conditions: The study primarily focused on negative housing and community conditions. Including the positive impacts of well-maintained environments could provide a more complete picture.
- Limited Built Environment Measures: Certain conventional built environment measures, such as accessibility to green space and walkability, were not included.
- Dataset Overlap: The comparison between the whole population and the middle- to low-income group datasets, while informative, might present statistical challenges due to their overlap.
Future work should involve longitudinal studies, objective data collection, incorporating positive environmental factors, and exploring a broader range of built environment indicators to gain a more comprehensive understanding of the complex interplay between living conditions and health.
Summary updated