, 2004, Doak et al., 2006, Flodmark et al., 2006, Hardeman et al., 2000, NHS Centre for Reviews and Dissemination, 2002, Sharma, 2006, Stice et al., 2006 and Summerbell et al., 2005), encompassing 70 studies. The participants were asked to consider their own intervention ideas and those presented, and prioritise potential elements of an intervention programme in three stages. Focus group schedules are shown in

Table 1. Sessions see more lasted1–2 h. Audio-recordings were transcribed verbatim. In order to explore perceptions of the causes of childhood obesity, we undertook a thematic analysis. Data were initially coded into emergent themes using NVivo7 computer package. An iterative inductive process was undertaken to identify relationships between themes and distill broad theoretical concepts (Spencer et al., 2003). All transcripts

were reviewed by the two moderators. Thematic coding was undertaken by one moderator, and emergent themes and relationships between them were reviewed by the second moderator. We convened 9 focus groups over 5 months in 2007. There was unavoidable heterogeneity within some groups, including one where a school governor was among a parent group. However, the flow of discussion was comparable to other parent sessions. In total there were 68 participants. The majority were female (60, 88%). PF-01367338 in vivo Of 55 participants disclosing ethnicity, 30 (55%) were from the three South Asian groups. (Table 2). The two overarching themes of influences on the development of childhood obesity to emerge are unhealthy food intake and lack of physical activity. These themes are consistent across a range of contexts which can be grouped into six areas; child, family, culture, school, local environment and macro-environment, although there is much fluidity between these. In terms of the wider environmental influences, most groups discussed the local environmental

context and professional participants explicitly articulated the wider societal view, particularly the influence of food marketing. Parents also implicitly alluded to societal influences through their stories. For example, reference science was made to media influences, the shift to sedentary lifestyles and the local abundance of fast food/takeaway shops. More proximal factors identified related to child and parental behaviours. For example, participants cited work commitments limiting parental time for food preparation and family activities, and unsafe local environments prompting parents to limit children’s physical activity. Whilst much data is widely applicable, some specific cultural contextual factors serve to explain particular health behaviours in South Asian communities. For example, extended families often live in one dwelling with hierarchical structures that give the grandmother control within the family and influence over the diets of the children.