Obesity and mortality
The association between obesity and mortality
There is a great deal of evidence which demonstrates that obesity measured either by BMI or waist circumference (as a measure of central adiposity), is a good predictor of an individual’s risk of death. A number of large scale prospective studies have demonstrated a ‘J-shaped’ association between BMI and risk of death (figure 1), with higher risks of death observed in the lower and upper BMI categories than in the middle categories [1,2]. However, whilst there is agreement in the published research that obese and underweight individuals have a greater risk of death than individuals of a healthy weight, the precise increase in mortality reported for a given level of obesity varies. This is likely to result from differences in the populations sampled, the other variables controlled for in the studies, and the methods employed.
Figure 1: Schematic illustration of the association between mortality and BMI for adults BMI
Existing estimates of the number of deaths attributable to obesity in England
The number of deaths that are attributable to obesity cannot be determined simply by counting the number of times obesity is recorded as a cause, or underlying cause, on death certificates. For a variety of reasons such an approach will tend to underestimate the proportion of obesity-related deaths. A recent study found that obesity was only mentioned on 0.23% of all death certificates issued in England during 2006, although this proportion has been increasing over time . In 2001, a report by the National Audit Office (NAO) ‘Tackling obesity in Britain’  estimated that approximately 6% of all deaths in England in 1998 were caused by obesity – a total of 30,000 excess deaths in that year. This compared to approximately 10% of deaths that were attributable to smoking, and 1% caused by road traffic incidents at that time.
The NAO analysis also showed that many of the deaths caused by obesity were premature – 9,000 of these deaths took place before state retirement age, and on average, each person whose death could be attributed to obesity lost nine years of life.
Other research using data from a similar time period suggested that the UK as a whole had the highest proportion of deaths attributable to obesity in Europe; 8.7%, compared to 7.7% of all deaths across the EU region .
It is difficult to assess how these figures may have changed since 1998. During this time the prevalence of obesity among English adults increased from 19.4% to 23.0% (an 18% increase in relative terms) . It therefore seems likely that more than 6% of all deaths will be attributable to obesity today. However, in 2010 there were fewer total deaths registered in England and Wales than in 1998 – 493,000 in 2010, compared to 553,000 in 1998 (an 11% reduction in relative terms) . It is not known whether the estimate of 30,000 excess deaths will have decreased in line with the total number of deaths, or whether it will have increased as a result of the higher obesity prevalence.
Why is it difficult to update these estimates?
Since the NAO estimates were produced, a number of publications have examined the methodology used to calculate obesity-attributable deaths. This work was prompted by the wide variety of different estimates of the number of obesity related deaths published for the US population. This work has shown that estimates vary considerably depending on the methodology used , and are also very sensitive to even small changes in the data used to produce the analysis [9,10]. In addition, a number of recent publications have provided updated and improved estimates of the association between obesity and mortality. The impact of obesity on mortality has been shown to vary by age and sex, and is also influenced by factors such as smoking status. Abdominal obesity and general obesity are also known to have independent effects on mortality risk. Ideally any estimates would take account of this variation, but this requires complex analysis and the necessary data are not always available.
At present we have very little information that can be used to quantify the effects of obesity on mortality within the English population. The estimates published for 1998 are likely to be out of date – both due to changes in mortality and obesity prevalence and improvements in methodology. Whilst it may be possible to produce quick updated estimates, sensitivity analysis based on US data suggests that any such figures would have a very large margin of error. As a result the potential usefulness of such figures is questionable. It is important however that we are able to produce an accurate assessment of the number or proportion deaths within England that are attributable to obesity. NOO will be looking to work with partner organisations to determine how such statistics can best be produced, and to quantify the margin of error around such figures.
- Pischon T, Boeing H, Hoffmann K, Bergmann M, Schulze MB, Overvad K, et al. General and abdominal adiposity and risk of death in Europe. New England Journal of Medicine 2008;359: 2105-20
- Berrington de Gonzalez A, Hartge P, Cerhan JR et al. Body-Mass Index and Mortality among 1.46 Million White Adults. New England Journal of Medicine 2010;363:2211-9
- Duncan M, Griffith M, Rutter H, Goldacre MJ. Certification of obesity as a cause of death in England 1979-2006. European Journal of Public Health 2010; 20(6):671-675
- National Audit Office. Tackling Obesity in Britain. London: NAO, 2001
- Banegas JR, López-García E, Gutiérrez-Fisac LJ, Guallar-Castillón P, Rodríguez-Artalejo F. A simple estimate of mortality attributable to excess weight in the European Union. European Journal of Clinical Nutrition 2003;57: 201–208
- Health Survey for England, 2009 Adult trend tables: http://www.ic.nhs.uk/pubs/hse09trends (accessed 15/11/2011)
Office of National Statistics. Death registrations by single year of age, England and Wales 2010: http://www.ons.gov.uk/ons/publications/rereference-tables.html?edition=tcm%3A77-229751 (accessed 15/11/2011)
Greenberg JA. Correcting Biases in Estimates of Mortality Attributable to Obesity. Obesity 2006; 14: 2071–2079
Flegal KM, Graubard BI, Williamson DF. Methods of calculating deaths attributable to obesity. American Journal of Epidemiology 2004; 160(4):331-8
Flegal KM, Williamson DF, Pamuk ER, Rosenberg HM. Estimating deaths attributable to obesity in the United States. American Journal of Public Health 2004; 94(9):1486-9