Green space and physical activity
Results from this study suggest that the amount of green space in people's living environment has little influence on people's level of physical activity. No significant relations were found between the percentage of green space in the living environment and whether or not people meet the Dutch public health recommendations for physical activity, sports and walking for commuting purposes.
We found a negative relation between the amount of green space and walking and cycling during leisure time. People in greener living environments undertake these activities less often. These results are in accordance with the Dutch study by Den Hertog et al. (2006)  in which different neighbourhoods in the city of Amsterdam were compared and in which a negative relation between green space and walking and cycling was found as well.
The finding that people with more green space in their living environment less often walk or cycle is probably due to the fact that in greener living environments, facilities such as shops are further away and people more often use a car to reach facilities. Furthermore, greener living environments in more urban areas are often set out more spaciously, reducing the facility density and increasing the possibility of parking a car near one's home. The study by Den Hertog et al. (2006)  performed in the Netherlands showed that – within an urban environment – both the density of facilities and parking possibilities were important determinants for the amount of physical activity undertaken, especially walking and cycling. In neighbourhoods with a high density of facilities and without private parking spaces, people more often choose to walk or cycle .
Our results concerning walking and cycling during leisure time contrast with studies which find rather strong indications for a relationship between attractive streetscapes and the amount of walking and cycling in Australia, the United States and the U.K. . A reason for the differences found could be that our data on green space did not provide specific information on the attractiveness of the streetscape. We were not able to investigate the influence of small green areas, like for instance trees along the roads. Furthermore, the differences found could be due to the walking and cycling culture in the Netherlands, which gives citizens of the Netherlands lots of opportunities to walk and cycle safely elsewhere, even when there is no green space in the direct vicinity of their homes.
We did find a positive relation between green space and gardening and cycling for commuting purposes. Especially the amount of agricultural green space influenced these types of physical activity positively. People with more agricultural green space in their living environment garden more often and for a longer duration, and if they cycle for commuting purposes, they spend more time on it. The fact that people with more agricultural green space in their living environment garden more often and spend more time on it, is most probably due to the fact that people in areas with more agricultural green space own larger size gardens.
An explanation for the fact that people who cycle for commuting purposes spend more time on this – a result which was also found in the study of Wendel-Vos et al. (2004)  – is that living environments with more agricultural green space are often located further away from cities, which is where most jobs are available. Therefore, people in the areas with more agricultural green space have to cycle more minutes to reach their work or school. Or as Wendel-Vos et al. (2004)  explains, 'the result reflects the fact that people in outskirts of town spend more time on bicycling to the city'.
Regarding the relationship between green space and physical activity in different levels of urbanicity, the relation appeared to be stronger in the more rural areas than in the urban areas. The strongest relation was found in slightly urban areas.
Concerning the subgroup analysis, the link between physical activity and green space was strongest for people aged under 25 and for elderly, lower educated people and people with a low income. This is in line with our hypothesis that children, elderly and lower socio-economic groups, spend more time in the vicinity of their homes and are therefore likely to be more affected by the design of their direct living environment.
Physical activity as an explanation for the relationship between green space and health
The fact that people spend more time on cycling for commuting purposes and on gardening could not explain the relation between green space and health. Therefore, we can conclude that physical activity is not a likely mechanism behind the relation between green space in people's direct living environment and health that was found in previous studies.
In analyses in which only people were included who garden or people who spend time on cycling for commuting purposes, no relationship was found between green space and self-perceived health. This can be explained by the fact that in these small subgroups the variation in green space is smaller. Additionally, people who spend time on gardening already spend time in green space and the extra benefit of green space outside their homes might not be discernable. Furthermore, people who garden and people who cycle for commuting purposes are probably healthier.
However, it is important to note that although people with greener living environments do not more often meet the Dutch public health recommendations for physical activity, it is possible that they more often undertake physical activity in a green environment. Different studies have shown that people with more green space in their living environment more often use green space . Because we did not have any data on where people were physically active, we were not able to find out whether people with greener living environments more often exercise in green spaces. A study by De Vries et al. (2004)  showed that the local green space supply does not determine how often people engage in recreation, but it does determine where people engage in recreation. The findings of this study also suggest that if there is no green space available people seek alternatives in other environments. Undertaking more physical activity in a green environment as opposed to an urban environment could have health benefits in the form of reduced stress symptoms [41, 42].
Furthermore, it is possible that the lack of an relation between the level of physical activity and the amount of green space is due to the high density of sports facilities and safe cycle tracks and footpaths almost anywhere in the Netherlands. Under these circumstances, the availability of green space is not a necessary condition to be physically active.
Strengths and limitations
This is one of the first studies to investigate whether the amount of physical activity undertaken can contribute to the explanation for the relation between green space and health found in previous studies. Where most studies only investigate the relationship between physical activity and green space or the relationship between green space and health, we investigated both the relationship between physical activity and green space as well as the relationship between green space, physical activity and health. Furthermore, unlike other studies performed in the Netherlands or in other countries, this study specifically investigates the relation between different types of green space and different types of physical activity for different subgroups and levels of urbanicity.
The data on health (and physical activity) and land use were derived from various databases; consequently, there is no single source bias.
In our study we used objective environmental measures. Objective environmental measures reduce the risks of respondent bias. However, subjective environmental measures can also provide important information. People's perception of green spaces may, in fact, motivate their behaviour more than the actual amount of available green space. Green spaces or green spaces that are considered unsafe or of poor quality tend to be avoided. Thus, supplementing objective measures with measures of an individual's perception will improve our understanding of how the green environment affects physical activity level. We used a self-report measure for physical activity which is the most commonly used measure for assessing physical activity [7, 33]. Using a self-report measure for physical activity has the advantage that it is easy to administer and generally acceptable to participants, and can measure a wide range of values . Self-report measures have the disadvantage of incomplete recall and exaggeration of the amount of activity . For this study there are no direct consequences in this respect, because we are interested in the relationship between green space and physical activity and it is not likely that people living in greener living environments will exaggerate more or less than people in less green living environments.
The measure used for physical activity, the SQUASH questionnaire, was not validated for each of the specific physical activities which are distinguished in this paper.
A limitation of our study is its cross-sectional design. The study does not inform us about the direction of causation. A second limitation is that we did not know where people were physically active. Other studies have found a significant correlation between the availability of green space and the use of green space . Our study shows that the absence of green space does not necessarily lead to less physical activity in general, but that people probably compensate for the lack of green space by being physically active elsewhere. Future research should include questions on where people are physically active.
Furthermore, some potentially important control variables could not be taken into account. It would, for instance, have been interesting to see whether the density of (sports) facilities in different living environments has an effect on the level of physical activity. In addition, ownership of a dog, which has been proved to influence the level of physical activity, could not be taken into account . Research has shown that there are rather strong indications for a relation between attractive streetscapes and the amount of walking and cycling in Australia, United States and U.K. . Unfortunately, we were not able to investigate whether this relationship can also be found in the Netherlands, because we did not have detailed information on the greenness of the streetscape.