Childhood Obesity and Neighbourhood Environments: integrating MEDMI and SAIL

Evidence is accumulating that neighbourhood green/blue space may promote good health and well-being through various mechanisms, including supporting and promoting physical activity.

This pilot project investigated the opportunity to combine the strengths of MEDMI and SAIL (The Secure Anonymised information Linkage Dataset at Swansea University) to link environmental, health and socio economic data.  Specifically the project looked at the association between neighbourhood characteristics (especially greenspace and socio-economic deprivation), environmental phenomena (rainfall) and childhood obesity.

The analysis involved extracting National Child Measurement Programme data from SAIL. These data are highly protected, meaning that in order to merge them geographically with environmental data from MEDMI, a reasonable proportion needed to be suppressed. The loss of data is non-random (since suppression is related to prevalence and area population), meaning considerable potential bias was introduced into the analysis. In future, it is proposed that the environmental data is extracted and linked to the health data securely within SAIL, where suppression will not be required.

The project also provided some useful lessons about the ease of accessing MEDMI data and has contributed to plans for the development of a new web-interface for accessing MEDMI data. Although not suitable for peer reviewed publication, this research will lead to future big data collaborations.