Background and Purpose
China faces increases in noncommunicable diseases and in obesity, and deceases in walking, bicycling and physical activity. Nearly 25% of Chinese residents are overweight; 3% are obese.1 In the last two decades, physical activity declined by over 30% among Chinese adults.2 China is urbanizing rapidly; it will add another 350 million urban residents by 2025.3 Accommodating this population will require an estimated 170 new mass-transit systems, 5 billion square meters of road, and 40 billion square meters of new floor space.4 Chinese cities have specific built environment features, attitudes, and behaviors that may shape active living. Limited research examines active living in Chinese cities.

Objectives
This study identified associations between walking, bicycling, and physical activity and physical environment features for three districts in Shanghai.

Methods
The three districts included Xintiandi (hypothesized as highly walkable/bikeable), Zhongshan Park (hypothesized as moderately walkable/bikeable), and Lianhua Road (hypothesized as low walkable/bikeable).  (Hypotheses were based on the presence of features associated with physical activity in existing research.)  Each district was 1 km in area and included a subway station. We adapted the Irvine-Minnesota Inventory (IMI)5 environmental audit tool for use in China (now called the IMI-C).  Based on a literature review and interviews with Chinese urban developers, we expanded the IMI from 162 to 286 items.  Examples of new items include measures of obstruction of sidewalks by vendors or parked cars and visible air pollution. We translated the IMI-C into Chinese and developed it as an iPad application.  Working with colleagues from East China Normal University, we conducted the IMI-C on all street segments in each area (N=286), including 129 segments in Xintiandi, 60 segments in Zhongshan Park, and 97 segments in Lianhua Road. We conducted health surveys of residents in each district, using a modified version of the existing China Health and Nutrition Survey. 6 The survey assessed physical activity levels, Body Mass Index, health status, and aspects of residents’ neighborhood environment.  Respondents were randomly surveyed outside a grocery store or farmers’ market in each district. We conducted 633 surveys, including 129 in Xintiandi, 243 in Zhongshan Park, and 291 in Lianhua Road.

Results
Data from the environmental audit were analyzed using the “State of Place Index,” a proprietary algorithm that calculates an overall “walkability” score composed of eleven sub-scores that measure urban design dimensions empirically tied to walking and bicycling. As expected, we found that Xintiandi was more “walkable” than both Zhongshan Park and Lianhua Road in terms of having more of the built environment characteristics that are associated with walking and bicycling in existing research. Zhongshan Park and Lianhua Road did not differ from each other overall in terms of the walkability of their built environments, but there were important differences in their built environment dimensions that relate to walking and physical activity. We analyzed the data from the health survey to understand the relationship between the built environment and physical activity. Respondents in the three districts usee different travel modes, including primarily walking in Xintiandi, public transit in Zhongshan Park, and cars in Lianhua Road. Respondents in Xintiandi were the most physically active, spending the longest time on physical activity (for exercise and transportation). Respondents in Lianhua Road were the least active and spent the longest time (51 minutes more) on sedentary activities. Respondents in Zhongshan Park were between the other groups. Also Xintiandi, with the highest State of Place score, had the lowest reported BMI and highest reported rates of physical activity; followed by Zhongshan Park with the next highest State of Place score, next lowest BMI and next highest physical activity rates; and finally Lianhua Road with the lowest State of Place score, highest BMI, and lowest rates of physical activity.

Conclusions
Findings are consistent with expectations based on existing research in Western settings.   In this highly dense city, the most "urban" district was associated with greatest use of active travel modes and lowest reported BMI, and the most "suburban" district was associated with the greatest use of cars for travel and highest BMI.  We are currently collecting additional data on three districts in Hangzhou, to expand environmental variability and the robustness of our findings.

Implications for Practice and Policy
If findings remain consistent, they will bolster the assertion that future Chinese urban development should seek to adopt active design strategies to support higher rates of physical activity.

References

1. Wu, Y. (2006, August 17). Editorial. Overweight and obesity in China, British Medical Journal, 333(7564), 362–363.
2. Ng, S. W., Norton, E.C., & Popkin, B. M. (2009). Why have physical activity levels declined among Chinese adults? Findings from the 1991–2006 China Health and Nutrition Surveys, Social Science and Medicine, 68(7), 1305–1314.
3. Woetzel, J., et al. (2009). Preparing for China’s urban billion. McKinsey & Company, Shanghai.
4. The State of China’s Cities (2010). Better city, better life. China Science Center of International Eurasian Academy of Sciences. China Association of Mayors, and UN Habitat. United Nations Human Settlements Programme, Nairobi, Kenya.
5. Day, K., Boarnet, M., Alfonzo, M., & Forsyth, A. (2006). The Irvine–Minnesota Inventory to measure built environments. Development. American Journal of Preventive Medicine, 30(2), 144–152.
6. China Health and Nutrition Survey. (2012). At: http://www.cpc.unc.edu/projects/china.

Support / Funding Source
We are grateful for support from the Council on Tall Buildings and Urban Habitat and the New York University Global Health Research Fund.