The Information Centre (The IC) commissioned the National Centre for Social Research to produce model based estimates of healthy lifestyle behaviours, using information from the Health Survey for England (HSfE). These estimates were produced to help meet users' requirements for more up to date information at the local area level, and will be published on the Neighbourhood Statistics (NeSS) website which is managed by the Office for National Statistics (ONS) on Monday 17 December 2007.
Please visit www.neighbourhood.statistics.gov.uk to get access to the model based estimates of healthy lifestyle behaviours. The instructions to get to the NeSS Healthy Lifestyle Behaviours: Model Based Estimates, 2003-2005 are as follows:
- go to the NeSS Home Page - www.neighbourhood.statistics.gov.uk
- click on 'I want to view or download data by topic'
- click on Health and Care under 'Neighbourhood Statistics' topics
- select the Healthy Lifestyle Behaviours: Model Based Estimates, 2003-2005 data set.
A model-based approach to producing healthy lifestyle prevalence estimates for each Middle Super Output Area (MSOA) and Local Authority (LA) in England was used because the sample size of national surveys such as the HSfE was too small to provide reliable estimates at a small area level. Model-based estimates and 95 per cent confidence intervals have been produced using 2003-2005 data from the Health Survey for England covering the prevalence of the following healthy lifestyle indicators for adults aged 16 or over:
- smoking among adults
- binge drinking for adults
- obesity among adults
- consumption of 5 or more portions of fruit and vegetables a day among adults.
Model-based estimates for children's fruit and vegetable consumption have not been released at either MSOA or LA level. A relative lack of precision (shown by the wide confidence intervals) indicated a poor fit of the models to the data.
The 2003-2005 estimates are the second set of model-based healthy lifestyle prevalence estimates to be published on NeSS. Differences in geographical boundaries, modeling methodologies and data sources, however, mean that they are not comparable to the preceding estimates for 2000-2002.