Summary

• Studies of the burden of disease provide a comprehensive assessment of the comparative magnitude of the premature mortality and disability resulting from different diseases and injuries and from important risk factors
• The first set of estimates of the global burden of disease (GBD) was produced for 1990. The most recent estimates are for 1990, 2005 and 2010
• The 2010 GBD study estimated the burden of 241 different diseases and injuries and of 43 risk factors. The 241 disorders are aggregated into three broad groupings: communicable, maternal, neonatal and nutritional disorders, non-communicable diseases and injuries
• The burden of each disorder is measured by the disability-adjusted life years (DALYs) lost to it. The measure combines the Years of Life Lost (YLL) to deaths from the disorder and the Years Lived with Disability (YLD) caused by it
• DALYs are an absolute measure that reflects the size and age structure of the population for which they are calculated
• YLL to a death at each age are calculated as life expectancy at that age in a low mortality reference life table in which e₀ = 86
• YLD from the impairments caused by a disorder are weighted by a disability weight of between 0 and 1 that reflects judgements by the public of the severity of the disablement caused by that impairment
• Burden of disease studies constrain the epidemiological estimates for all diseases combined to match an independent set of estimates of all-cause mortality
• Previous GBD studies raised the weight of DALYs lost by young adults relative to those lost by children and old people and discounted future DALYs at 3 per cent a year; the 2010 study assigns each DALY equal weight no matter when it is lost or who loses it
• By calculating the average health value in each age group of a population from the age-specific prevalence and disability weights of each of the impairments, one can estimate the health-adjusted life expectancy (HALE) of the population using Sullivan’s method
• The results of the GBD studies highlight the heavy burden of some diseases with low fatality, especially musculoskeletal disorders and mental health problems. They reveal that the epidemiologic transition is now well advanced in much of the world, but that sub-Saharan Africa is lagging behind.

Conducting a burden of disease study involves a mass of decisions about the statistical methods to be used to estimate YLL and YLD, the interpretation of the epidemiological evidence, the reference life table used to calculate YLL, the disability weights used to calculate YLD, whether to age weight and discount, and so on. Some of these decisions are technical ones; others are explicitly value judgements. All of them have implications for the findings of the study.