Background: Internationally, the prevalence of adults with obesity is a major public health concern. Few studies investigate the explanatory pathways between fast-food outlets and body mass index (BMI). We use structural equation modelling (SEM) to explore an alternative hypothesis to existing research, using area-level deprivation as the predictor of BMI and fast-food outlets and diet quality as mediators.

Methods: Adults (n=7,544) from wave two of the Yorkshire Health Study provided self-reported diet, height and weight (used to calculate BMI). Diet quality was based on sugary drinks, wholemeal (whole grain) bread, and portions of fruit and vegetables. Fast-food outlets were mapped using the Ordnance Survey Points of Interest (PoI) within 2km radial buffers around home postcode which were summed to indicate availability. Age (years), gender (female/male) and longstanding health conditions (yes/no) were included as covariates.

Results: There was little evidence linking fast-food outlets to diet or BMI. An independent association between fast-food outlet availability and BMI operated counterintuitively and was small in effect. There was also little evidence of mediation between fast-food outlet availability and BMI. However, there was more evidence that area-level deprivation was associated with increased BMI, both as an independent effect and through poorer diet quality.

Conclusion: This exploratory study offers a first step for considering complexity and pathways linking fast-food outlets, area-level deprivation, diet quality and BMI. Research should respond to and build on the hypothesised pathways and our simple framework presented within our study.