Residents of deprived neighbourhoods have consistently been found to be less physically active than residents of non-deprived neighbourhoods, independent of their individual socio-economic status [1–6]. In the past decade, numerous area-based initiatives (ABIs) have been implemented in deprived neighbourhoods across Western-Europe . These large-scale initiatives aim to tackle the multitude of socio-economic and environmental problems in these neighbourhoods that might influence physical activity (PA) behaviour, including employment, income, housing, crime, and social cohesion. There is little robust evidence of their effect on PA [8–10]. Quasi-experimental evaluations of natural experiments may be useful to assess their effectiveness [10–13]. However, in the field of PA, this type of evaluations is still in its infancy .
ABIs may affect PA behaviour via different pathways. Better economic position may improve access to social and material resources for PA . Stronger community bonds may enlarge social support and companionship for PA, which have consistently been associated with higher levels of PA [15, 16]. Stronger community bonds may also reinforce positive social norms for healthy behaviours such as PA . Improvements with respect to housing and the physical environment may improve neighbourhood aesthetics, pedestrian infrastructure and recreational facilities, which have all been consistently associated with PA [16–22]. A safer neighbourhood may reduce the fear of outdoor activities, although evidence for an association between neighbourhood safety and PA is less consistent .
The English New Deal for Communities (NDC) is one of the few ABIs that has been used as a natural experiment to explore its impact on PA [24, 25]. The NDC aimed to improve the socio-economic and environmental situation in England’s most deprived areas. At 4-year and 6-year follow-up, a flat post-intervention trend in PA was found for NDC areas and control areas [24, 25]. The authors concluded that the NDC had no impact on PA.
However, the impact of the NDC on PA may have been underestimated, as no pre-intervention trends were included in the evaluations. Moreover, previous research has suggested that the impact might depend on the ability of interventions to influence the outcome and on the quality of their implementation [9, 26]. Interventions that are aimed at environmental problems instead of socio-economic problems may be more likely to cause district-wide changes in LTPA within a relatively short period of time because of a wider reach and shorter lag-times of effect. Studies are needed that assess changes in PA trends before and during the intervention and that focus on areas where intensive environmental interventions have been implemented.
An opportunity for such a study arose in 2008 with the implementation of a Dutch ABI called the District Approach. The District Approach aims to alleviate problems of employment, education, housing and the physical environment, safety, and social integration in 40 of the most deprived districts of the Netherlands. Districts have been selected based on their accumulation of economic, physical, and social problems, judged on statistics and survey data. Each district developed its own mix of socio-economic and environmental interventions, resulting in large between-district variations in the intensity of interventions .
This study aimed to evaluate the short-term impact of the District Approach on trends in leisure-time PA (walking, cycling, and sports) using a quasi-experimental interrupted time-series design. First, PA trends before and during the intervention were assessed in all deprived target districts and in various control groups. Next, trends were assessed in those deprived target districts where environmental interventions have been implemented most intensively. We expected to find a more positive PA trend change in deprived target districts than in the control groups, especially in deprived target districts with intensive environmental interventions.