Street-view and traditional greenness metrics with adults' sitting time in high-density living in Hong Kong: Comparing associations, air pollution and noise roles, and population heterogeneity

Sci Total Environ. 2023 Apr 20:870:161778. doi: 10.1016/j.scitotenv.2023.161778. Epub 2023 Jan 28.

Abstract

Background: Emerging evidence suggests neighborhood greenness is associated with physical activity; however, the sitting-specific associations with multi-source greenness metrics remain unclear, especially in high-density cities.

Objectives: This cross-sectional study examined: 1) the associations of street-view greenness (SVG) and traditional greenness metrics (i.e., Normalized Difference Vegetable Index (NDVI) and park density) with sitting time; 2) the potential moderating/mediating roles of objective/perceived air pollution and perceived roadside noise; and 3) how the associations vary by demographics and socioeconomic status.

Methods: Interview survey data of 1977 adults in Hong Kong from 2014 and 2015 was linked to environmental data. Using an object-based image classification algorithm, SVG was derived from Google Street View images, capturing human-viewed street-level greenery. NDVI was derived from Landsat 8 satellite images using the normalized difference between the near-infrared and red bands. Park density was calculated by point density. In the main analyses including regressions, parallel mediation, interaction, and stratified models, the environmental metrics were measured within a 1000-m Euclidean buffer of residence.

Results: SVG and park density were negatively associated with sitting time after adjusting for covariates including physical activity while NDVI was not significantly associated with sitting time, and results were robust with 800-1800 m Euclidean and 1400-1800 m network distance. Greenness-sitting associations were not moderated/mediated by perceived air pollution/roadside noise while SVG-sitting associations were moderated by objective NO2, O3, and PM2.5 and mediated by O3. SVG-sitting associations differed by age, having under-school-aged children, birthplace, education, and occupation type while associations between traditional greenness metrics and prolonged sitting showed no significant population heterogeneity.

Conclusions: SVG appears to be more accurate in estimating exposure than traditional metrics to reflect greenness-sitting associations, objective air pollution moderating and mediating roles, and population heterogeneity, which emphasizes the importance of street-level greenness planning for health promotion in terms of reducing sitting time.

Keywords: Age; Air pollution; Neighborhood greenness; Prolonged sitting; Socioeconomic status.

MeSH terms

  • Adult
  • Air Pollutants* / analysis
  • Air Pollution* / analysis
  • Child
  • Cross-Sectional Studies
  • Environmental Exposure / analysis
  • Hong Kong
  • Humans
  • Noise
  • Particulate Matter / analysis
  • Sitting Position
  • Vegetables

Substances

  • Air Pollutants
  • Particulate Matter