Does the built environment moderate the impact of mass media campaigns for physical activity?

[Truncated] INTRODUCTION
The built environment is associated with physical activity yet little is known about how the effects of physical activity campaigns might vary across differing neighbourhood environments. Mass media campaigns are widely used in Australia, the US, the UK and Canada to promote physical activity. The purpose of this research was to conduct a series of studies to learn if environmental features moderate the effects of physical activity promotion mass media communication strategies. Three mass media-led campaigns were examined: Find Thirty every day a Western Australian state-wide physical activity campaign and WV Walks (West Virginia) and BC Walks (Broome County, New York), two walking campaigns conducted in North America. The overall aim of this thesis was to investigate the extent to which the built environment moderates the impact of mass media campaigns designed to increase physical activity amongst adults. The underlying hypothesis was that the environment positively moderates cognitive campaign impacts, behavioural impacts, and the relationship between cognitive and behavioural impacts.
METHODS
Perceptions of the neighbourhood environment were examined for moderating effects of all three campaigns and objectively measured environmental data were linked with survey data in Find Thirty every day. In Find Thirty, awareness, a range of other hierarchical cognitive campaign effects and behavioural effects were examined for moderation in both a cross-sectional and cohort study design. In WV Walks and BC Walks awareness and behavioural effects were examined in a cohort study design.
RESULTS
In Find Thirty every day, contrary to expectations, significant interactions were found suggesting that the mix of destinations negatively moderated some cognitive campaign impacts. Those with a lower mix of destinations were more likely to accept the campaign message and take action than those with a greater mix. Environmental perceptions were not associated with cognitive campaign effects. In terms of behaviour, again significant negative interactions were found whereby the mix of destinations negatively moderated the likelihood of achieving sufficient levels of overall walking (but no other behavioural effects): i.e., people living in neighbourhoods with a greater mix of destinations, did less walking. Neither objectively measured neighbourhood walkability nor neighbourhood perceptions moderated behavioural impacts.
In WV Walks, no significant associations with campaign awareness were identified. For walking behaviour among the intervention community, those with a greater mix of destinations increased their days walking significantly more than those with a lower mix of destinations. While the interaction term testing this difference did not reach statistical significance, no such association was observed in the comparison community. The results therefore suggested a positive moderating effect of the environment on behavioural campaign impact, which was consistent with the hypothesis but in contrast to the Find Thirty findings. No other environmental perceptions moderated this relationship in the WV Walks campaign, nor was there any evidence that the environment moderated campaign awareness or the relationship between awareness and behavioural impacts. In the BC Walks campaign, only a limited number of environmental variables were available to examine and there was no evidence of any moderating effects.