Methods of prevalence-based models
| Model | Method of estimation | Prevalence rates used for projections | Validation against QOF data? | Model validation? | CIs? |
| Shaw et al21 | Logistic regression | Age and sex | No | No | No |
| Guariguata et al26 | Logistic regression | Age and sex, and urban/rural | No | No | No |
| Association of Public Health Observatories18 25 | Direct estimation from HSE for age, sex and IMD. Trend in obesity estimated by linear regression. | Age and sex, IMD (2004), Ethnicity and increases in obesity | Yes for 2008/2009 | No | Yes |
| PHE5 | Logistic regression | Age and sex, ethnicity, IMD 2015 | Yes for 2014/2015 | Yes: refitting model on 70% of data and assessing against remaining 30% | No |
HSE, Health Surveys for England; IMD, Index of Multiple Deprivation; PHE, Public Health England; QOF, Quality and Outcomes Framework.