The present study indicated that environmental factors had indirect effects on walking, moderate-intensity activity excluding walking, and vigorous-intensity activity through self-efficacy for exercise, social support for exercise and pros and cons for exercise among Japanese adults. A previous study  suggested that environmental factors indirectly affected walking, moderate-intensity activity, and vigorous-intensity activity through motivation and self-efficacy. Moreover, the availability of physical activity facilities directly affected walking and moderate-intensity activity; also, the quality of a neighborhood directly affected moderate physical activity and vigorous-intensity activity . Research results on direct and mediated effects of perceived equipment accessibility and neighborhood safety with regard to physical activity among adolescent girls indicated that these environmental factors were mediated by the effects of self-efficacy for overcoming barriers [14–16]. The present study investigated the environmental factors that affected moderate to vigorous-intensity activity in males and females aged from 20 to 79 years. While previous studies [14–16] had focused on different levels of intensity of physical activity and age groups, the present study supported the previous finding that environmental factors affect physical activity through psychological factors. The present study found that self-efficacy was the most influential factor that directly affected physical activity. This was consistent with a previous study, thus supporting the hypothesis that self-efficacy plays an important role as a moderate variable of behavior in social cognitive theory . In addition, social support influenced physical activity through self-efficacy. This again replicated the reported finding that social support affects physical activity through self-efficacy and motivation [13, 15, 41, 42].
Findings from the present study indicate that environmental factors directly affect the pros and cons for exercise. Pros for exercise directly affect self-efficacy and social support, and cons for exercise directly affect self-efficacy only. Previous studies had reported a direct relationship between pros and cons and engaging in physical activity . However, there have been no reports on how self-efficacy or social support affects the effects of pros and cons on physical activity. It also remains unclear how pros and cons affect the relationship between physical activity and environmental factors. The findings of the present study suggest that through self-efficacy and social support, pros and cons affect the impact of environmental factors on physical activity.
With regard to the indirect effect of environmental factors on physical activity, the path coefficient was highest from cons to self-efficacy. Perception of environmental factors affected low cons, which in turn affected high self-efficacy and promoted physical activity. This finding suggests that some environmental interventions could reduce cons associated with exercise and thereby promote physical activity in Japanese adults. For instance, these include maintaining neighborhood exercise facilities, improving access to available facilities, promoting better awareness of the facilities, and improving neighborhood safety.
The present study found differences in the model based on the types of physical activities. While a direct and positive effect from environmental factors was found with regard to walking, no such effect was found from vigorous and moderate-intensity activity excluding walking. Further research examining environmental factors associated with physical activity needs to consider these differences based on the types of physical activities when promoting physical activity. For instance, environmental factors conducive to walking can be seen to directly promote walking. On the other hand, for increasing vigorous-intensity activity, environmental factors should be able to reduce cons and increase pros and social support in order to gain high self-efficacy, which directly affects physical activity. Further study is needed to develop and evaluate the effects of specific interventions directed at the environmental factors that promote physical activity.
A few studies [13–16, 23–26] have examined the direct and indirect effects of built-environmental, psychological, and social factors on physical activity. Furthermore, no such study has been conducted in Japan, which has important cultural differences from other countries. For example, compared with individuals in Western countries, the prevalence of overweight participants (BMI ≥ 25.0 kg/m2) who are more likely to be more inactive people is low; 29.6% in males and 19.0% in females in Japan . Therefore, the present study contributes to the development of physical activity promotion strategies in Japan.
The present study also has a few limitations. First, the present study was conducted in an Internet setting. A potential limitation of Internet surveys is that respondents tend to be young, educated, and with a higher income . The participants in the study may have included a higher proportion of not employed females than those in the general Japanese population. Furthermore, previous studies [45, 46] have indicated that participants with high levels of leisure-time sedentary behavior, e.g., Internet and computer use, were often those associated with low levels of physical activity. The adjusted prevalence for age bracket of the Japanese population (according to the 2005 Census data ) indicated that 65.0% of the population engaged in ≥ 150 minutes per week of at least moderate-intensity activity.
In the present study, the percentage of individuals who engaged in physical activity was lower: 7.5%. Therefore, the present study participant may have underestimated the affects of environmental, psychological, and social factors on physical activity in comparison with the general Japanese population. Second, the present study used a self-administered questionnaire to examine physical activity; however, the reliability and validity of the scale was comprehensively examined and confirmed in a previous study . Therefore, the possibility of selection bias cannot be excluded. Moreover, in the present study, environmental factors were unable to confirm the existence of the path from moderate to vigorous-intensity activity. This could be because items for environmental factors may have been more heavily associated with walking than with moderate and vigorous-intensity activity. However, this scale for environmental factors included items such as the presence of facilities (walking trails, fitness clubs, etc.) for engaging in physical activity, as well as the availability of home fitness equipment (shoes, pedometers, dumbbells, etc.), which represent environmental factors that affect vigorous-intensity activity. Therefore, content validity appears to be appropriately retained. In addition, although outcome variables assessed items with reference to the physical activity, the psychological and social variables assessed items with reference to exercise in the present study. This may cause weak associations between these variables and physical activity. In future research, the outcome variables and the psychological and social variables should be matched.
Respondents of the present study were considered as a representative section of the population because the Internet research service organization that conducted the present study used randomly selected individuals from a pool of 1,150,000 so that an equal number of responses were obtained from each gender and age group between 20 to 79 years. Moreover, the respondents included those living in various regions with diverse occupations, and 71.6%, 76.6%, 75.9%, 69.5%, and 57.7% of the Japanese population in their 20's, 30's, 40's, 50's, and 60's-80's, respectively, used the Internet at least once per week ; this further strengthens the representativeness of the survey. Since no other study of the Japanese population has been conducted on this topic, the findings of the present study will be important for the development of intervention strategies for population-based health promotion in the future and will therefore contribute to the promotion of physical activity.