Population determinants of urban health : episodes from Hong Kong

Abstract

During the 21st century, rapid urbanisation has taken place around the world. The emphasis of “healthy cities” as the continuous improvement of cities to enhance the holistic wellbeing of inhabitants is therefore increasingly important. Hong Kong – one of the most urbanized cities globally – has attained world-leading longevity yet reports the lowest meaning in life and has experienced poorer mental health and social support over the past decade. Hong Kong could therefore serve as a social laboratory to examine population determinants of the key domains of health, namely physical, mental and social wellbeing. In particular, unique characteristics of Hong Kong and new analytic methods were leveraged to advance our understanding of urban health.

With longevity as a key indicator of physical health, this thesis examined to what extent smoking, the largest cause of premature deaths, had contributed to life expectancy gaps between Hong Kong and high-income countries (HICs). I used the smoking impact ratio (SIR) method and the novel truncated cross-sectional length of life (TCAL) to estimate smoking-attributable survival advantages of Hong Kong males and females over their HIC counterparts. Between 1979 and 2016, Hong Kong accumulated a substantial survival advantage of 1.86 years (95% CI: 1.83, 1.89) for males and 2.50 years (2.47, 2.53) for females over HICs, of which the low smoking prevalence in Hong Kong contributed to 50.5% and 34.8% for males and females, respectively. Stringent tobacco control was therefore a fundamental reason for Hong Kong’s survival advantage over HICs.

Next, I examined prevailing and emerging environmental determinants of urban mental health. This included noise pollution and collective actions-related exposures, respectively, as both determinants are increasing worldwide and their mental health impact in highly urbanised regions has been sparsely documented. Using a large prospective cohort study, residential road traffic noise was shown associated with probable depression (OR = 1.17, 95% CI: 1.05, 1.31 for each increment of 10 dB(A) 24-hour average exposure) and poorer mental wellbeing (mean difference = -0.19, 95% CI: -0.31, -0.06), adjusting for individual- and neighbourhood-level attributes. Using marginal structural modelling, cumulative intrafamilial sociopolitical conflict over a 10-year period was associated with probable depression (OR = 1.45, 1.22, 1.74). Time spent on social media for sociopolitical news during protests was also associated with probable depression (OR = 2.17, 1.26, 3.72). Finally, the thesis focused on a key attribute of social wellbeing: family support. During and after major protests, a substantial and sustained decline in family support was detected. I showed that intrafamilial sociopolitical conflict was associated with poorer family support, which in turn was associated with increased risk of probable depression.

In sum, this thesis has identified important drivers for each domain of population health in Hong Kong. Notably, the determinants for Hong Kong’s world-leading longevity would not be sufficient to enhance population mental and social wellbeing. This is substantiated by the disconnect between advancements among domains of health in Hong Kong. As such, a multipronged and multisectoral approach would stand to comprehensively improve urban health.