Obesity is an emerging epidemic worldwide. In the last few decades, rapid urbanisation and economic growth in China have led to increasing prevalence of overweight and obesity [1]. China is in an epidemiological transition defined by positive energy balance and mounting prevalence of non-communicable diseases [2, 3]. In 2000, 25.6% of the urban and 17.3% of the rural populations were overweight or obese, compared to 12.2% and 7.7% in 1989 [1].

In addition to a dramatic increase in energy intake [4, 5], China has experienced a noticeable decrease in energy expenditure in both urban and rural areas [6]. Decreases in physical activity have been reported for several domains, including occupation, transportation and household activity [7]. In China, occupation has been a major source of physical activity [8]. According to the China Health and Nutrition Survey (CHNS), between 1991 and 2000, energy expenditure from occupational physical activity decreased by 22% among men and 24% among women, and the decrease contributed to the increase in body weight [9]. During the same nine-year period, energy expenditure from household physical activity decreased by 57% among men and 51% among women [9]. A similar trend was found with transportation physical activity [6]. As the Chinese society moves toward modern inactive lifestyles [7], leisure-time physical activity (LTPA) remains low [10]. Population survey data from 2006 found that only 13% of Chinese men and 8% of women engaged in any LTPA [6]. Although population data have shown a slight increase in LTPA, the increase has not been sufficient to compensate for declines in occupational and household activity [6].

Sedentary behaviour (too much sitting) is distinct from too little physical activity. Multiple studies have concluded that sedentary time is an independent risk factor for metabolic risk and chronic diseases [11, 12]. TV viewing particularly has been associated with increased risks for overweight/obesity [13, 14], diabetes [13, 15], and metabolic syndrome [16, 17]. Sedentary behaviours are understudied in China. One survey study in urban Qingdao found that adults spent an average of 3.2 sedentary hours per day watching TV, reading, or using a computer during non-working time [18]. Excess sedentary time is a public health concern in China given high rates of TV and personal computer ownership, particularly among urban residents [1]. No study could be found that examined sedentary behaviours in rural China.

In China, rural residents are currently more physically active and less obese than their urban counterparts [19]. However, a national study found a greater decrease in physical activity in rural areas than in urban areas over a recent ten-year period [20]. The current study was conducted in Suixi County, Guangdong, China, an area that has experienced farming-to-business shifts and some degree of economic modernisation, in common with many rural areas in China. Study objectives are: to assess levels of physical activity and sedentary behaviours of farming and non-farming adults in rural Suixi; to describe different activity patterns during farming and non-farming seasons; and to examine factors associated with LTPA and TV viewing.

This study in rural Suixi found differences in physical activity and sedentary behaviour by occupation. Farmers had significantly more occupational physical activity compared to their non-farming counterparts, however, they appeared to substitute farming activities with sedentary behaviours (primarily TV viewing) during non-farming seasons. Based on this pattern, one may expect further decreases in occupational physical activity as a result of fewer farming-related occupations, less labour-intensive farm work, and shorter farming seasons. Without compensating forms of physical activity or reduction of TV time, this pattern suggests future increases in overweight/obesity and associated chronic diseases.

The prevalence of LTPA was low in the study sample, which was consistent with findings from previous studies [6, 10, 19]. Correlates of LTPA were explored. After adjusting for other demographic characteristics and environmental variables, those in our study were more likely to participate in LTPA if they were unmarried, living in a smaller household, having sports equipment in the home and if their family, friends, or neighbours participated in LTPA.

The strongest correlate of LTPA was social modelling. During the "member checking" phase of the study, informants strongly agreed with this finding. A woman informant stated, "We are in a very collective society and nobody exercises alone." Other informants confirmed that families, friends, and neighbours in their communities were closely connected (i.e., "The social circle is small here. If one person goes out for a walk, immediately others will notice and probably will follow. They can chat while walking.") Studies in western countries have suggested that seeing other people exercise was associated with individuals' own participation in LTPA, possibly through modelling and prompting [26–28]. In the context of rural China, exercising with others may provide additional and possibly essential social reinforcement. Social support and group activity interventions are effective in Western countries [29], and they may be even more effective in China, but this remains to be demonstrated.

Another finding was that those who had sports equipment in the home had twice the odds of engaging in LTPA. In a cross-sectional study, such a finding should be interpreted with caution, since it is possible that those who had already been active would be more likely to obtain such equipment. Informants suggested that home equipment was only half of the equation if access to exercise facilities and localities was limited. For example, a young man stated "It is true that I am more active because I have a basketball, but I play in streets where it is not safe. I would have played a lot more if there was a basketball court around." Although the Chinese government has implemented the "Sports for All" program to expand relevant facilities and infrastructure [30], rural residents still have limited access to such facilities, as demonstrated by one informant's comment, "Sometimes we have to travel for hours to a bigger town for sports. We do not have anything here in our village." In this study sample, adults who were married or who were living in a larger household were less active. This may be a result of related lifestyles and competing priorities of other household activities. An informant said: "We have to travel so far for any activity. If one is married and has children (like me), how can s/he have time for any recreation?"

TV viewing was the major sedentary behaviour. The median TV time was 12 hours per week, and 39% watched TV for 14 hours or more. This is comparable to TV viewing time reported by adults in the US [31, 32]. A woman informant described the "common" lifestyle in her village: "People in my village tend to turn on TV immediately after meals, and then they rest (sitting or lying down) for one hour after lunch, and another hour after dinner." In the current sample, 96% of participants had a TV set in the home, indicating substantial influences from modernisation. Although computers were not as common (23% participants had a computer at home), it is reasonable to postulate that as computer ownership increases, screen time sedentary behaviour may further increase in rural China.

None of the correlates of LTPA was significantly associated with TV time. This finding was similar to a previous study, in which environmental variables that facilitated physical activity were not related to TV time [33]. Furthermore, no association was found between TV time and LTPA in the current sample (Table 2). This suggests that TV viewing may be a behaviour that is independent of physical activity in rural Suixi [11].

A major strength of the study was using community participation to enhance the design, survey administration and results interpretation. These procedures improved the cultural relevance of research questions, qualitatively validated research findings, and provided interpretations that extended quantitative results. Although a small sample from one county in southern China limits generalizability, and a cross-sectional study cannot establish causality, these findings provide insights to guide future larger-scale and prospective studies that employ objective measures of the environment, physical activity, multiple sedentary behaviours (including sleeping) [34], and health outcomes. Future studies should assess additional aspects of the built and social environments and explore variables related to sedentary behaviours, especially TV viewing.

The current study, together with previous studies on the consequences of modernisation in China [6, 9, 10, 35], suggests that if no effective public health actions or social and environmental initiatives are implemented, further decreases in physical activity and increases in sedentary behaviours leading to higher risk of major chronic diseases are likely, as a by-product of the economic development, modernisation, and urbanisation in China.

In rapid social and epidemiological transitions, Chinese central and local governments are in the position of making health-promoting changes and taking preventive measures through city planning initiatives, infrastructure building, and land-use development to provide activity environments [36]. Physical activity programs should be implemented in rural China targeting multiple levels of influences [37]. Interventions to improve accessibility of recreational facilities and to promote physical activity as a social behaviour may be culturally tailored.