National Institute for Public Health and the Environment (RIVM), Antonie van Leeuwenhoeklaan 9, 3721 MA Bilthoven, The Netherlands; Institute for Risk Assessment Sciences (IRAS), Utrecht University, Utrecht, The Netherlands.
Institute for Risk Assessment Sciences (IRAS), Utrecht University, Utrecht, The Netherlands.
Environ Res. 2018 Jan;160:531-540. doi: 10.1016/j.envres.2017.10.027. Epub 2017 Oct 26.
In epidemiological studies, exposure to green space is inconsistently associated with being overweight and physical activity, possibly because studies differ widely in their definition of green space exposure, inclusion of important confounders, study population and data analysis.
We evaluated whether the association of green space with being overweight and physical activity depended upon definition of greenspace.
We conducted a cross-sectional study using data from a Dutch national health survey of 387,195 adults. Distance to the nearest park entrance and surrounding green space, based on the Normalized Difference Vegetation Index (NDVI) or a detailed Dutch land-use database (TOP10NL), was calculated for each residential address. We used logistic regression analyses to study the association of green space exposure with being overweight and being moderately or vigorously physically active outdoors at least 150min/week (self-reported). To study the shape of the association, we specified natural splines and quintiles.
The distance to the nearest park entrance was not associated with being overweight or outdoor physical activity. Associations of surrounding green space with being overweight or outdoor physical activity were highly non-linear. For NDVI surrounding greenness, we observed significantly decreased odds of being overweight [300m buffer, odds ratio (OR) = 0.88; 95% CI: 0.86, 0.91] and increased odds for outdoor physical activity [300m buffer, OR = 1.14; 95% CI: 1.10, 1.17] in the highest quintile compared to the lowest quintile. For TOP10NL surrounding green space, associations were mostly non-significant. Associations were generally stronger for subjects living in less urban areas and for the smaller buffers.
Associations of green space with being overweight and outdoor physical activity differed considerably between different green space definitions. Associations were strongest for NDVI surrounding greenness.
在流行病学研究中，暴露于绿色空间与超重和身体活动的相关性不一致，这可能是因为研究在绿色空间暴露的定义、纳入重要混杂因素、研究人群和数据分析方面存在很大差异。
我们评估了绿色空间与超重和身体活动的相关性是否取决于绿色空间的定义。
我们使用来自荷兰一项全国性成年人健康调查的 387195 名成年人的数据进行了一项横断面研究。根据归一化差异植被指数（NDVI）或详细的荷兰土地利用数据库（TOP10NL），为每个居住地址计算了到最近公园入口的距离和周围绿色空间。我们使用逻辑回归分析研究了绿色空间暴露与超重和每周至少 150 分钟户外适度或剧烈身体活动的相关性（自我报告）。为了研究相关性的形状，我们指定了自然样条和五分位数。
到最近公园入口的距离与超重或户外身体活动无关。周围绿色空间与超重或户外身体活动的相关性高度非线性。对于 NDVI 周围的绿色度，我们观察到超重的几率显著降低[300m 缓冲区，优势比（OR）=0.88；95%置信区间（CI）：0.86，0.91]，而在最高五分位与最低五分位相比，户外身体活动的几率增加[300m 缓冲区，OR=1.14；95%CI：1.10，1.17]。对于 TOP10NL 周围的绿色空间，相关性大多不显著。对于居住在城市较少地区的受试者和较小缓冲区，相关性通常更强。
绿色空间与超重和户外身体活动的相关性因绿色空间的不同定义而有很大差异。与其他绿色空间定义相比，NDVI 周围的绿色空间相关性最强。
