Highlights from the Global Burden of Disease 2010 Studies

The following two tabs change content below.
Sanjay Basu
Sanjay Basu, MD PhD is a resident physician in the Department of Medicine at the University of California San Francisco. He blogs about the political economy of global health, epidemiology, sociology, economics and much more at EpiAnalysis. @sanjayb493

GlobalBurden_170This morning, The Lancet published the most comprehensive look at the Global Burden of Disease in over a decade. The “GBD 2010″ revealed major shifts in our understanding of global public health and what is causing disease worldwide. For those of you not planning to cure your insomnia tonight by reading all 196 pages of the text, here’s a quick run-down of the major results:

Longer, more disabled living

The GBD 2010 involved a Herculian effort of studies on population health produced over the last five years by 486 researchers at 302 institutions in 50 countries. The calculations of morbidity and mortality conducted by these researchers compare estimates for disease and disability in 1990 and 2010 among representative populations in nearly every area of the world–giving a broad picture of two decades of change in public health.


No doubt, those scientists involved in the project aged rather poorly given the chronic stress that must have resulted from conducting this study; their subjects aged significantly too, the results reveal: the major overall theme of the GBD is that the average person in the world today has a much improved chance of living to older age in most countries, but a much higher risk of spending a large portion of life disabled.

The GBD estimates reveal that for the first time in history, shortly after the year 2015, there will be more people above age 65 than below age 5. The causes of death have also changed among both men and women:


Our demographic futures are not just old; they are also chronically ill. For every year of life expectancy that has been added to the average length of life since the year 1990, GBD studies estimate that less than 10 months is healthy living, and the rest is spent in some degree of disability. This is particularly pronounced for people greater than age 50, who spend nearly half of their living time in a state of pain, immobility, or related incapacity that requires medical support.

Dual burdens

The emergence of chronic non-communicable diseases (NCDs) is prominent in the GBD study, but the results reveal that rather than a pure “epidemiological transition” in which infectious diseases dissipate while NCDs overtake the population, a “dual burden” seems apparent in numerous countries where infections are still prevalent, especially HIV and TB, while NCD rates rise. The two often play off one another, as diabetes and smoking for example increase the risk of TB, while HIV medications increase the risk of heart disease.

The prominent risk factors are nevertheless related to NCDs. Hypertension is now the leading risk factor for death in the world, followed by smoking, then alcohol use. Deaths from NCDs rose by nearly 8 million people between 1990 and 2010, meaning that they accounted for almost 2 out of every 3 deaths worldwide. Cancer deaths have risen 38% over the two decade, mostly those related to smoking (tracheal, bronchial, and lung cancers). Heart disease and strokes accounted for 1 in 4 deaths in 2010, up from 1 in 5 in 1990. And deaths from road traffic accidents increased as well.

HIV has not disappeared, it has just been ignored

While HIV overall peaked in the mid-2000′s, AIDS remains the leading cause of death in southern and eastern Africa, and of note also ranks number three in eastern Europe. Perhaps most importantly, it is still a leading cause of death among the young–killing more women between the ages of 15 and 49 years old than any other cause.

In fact, age-standardized death rates actually increased from HIV. And in some African and Caribbean countries, and Eastern Europe including Russia, people overall have been less healthy than before (see male, then female graphs below in terms of overall healthy life expectancy gained or lost from 1990 to 2010), with much of the burden due to HIV:


The stereotype of African exceptionalism…is partially true

Variations in these death rates between regions were extremely large. But no where are they more prominent than in comparisons between sub-Saharan Africa and the rest of the world. While the rest of the world increasingly dies from NCDs, the sub-Saharan African population still dies (76% of the time) from infectious, maternal, neonatal and nutritional causes of death.

Indeed, the years of life lost to disability in most of the world were due to lower back pain, depression, anemia, and emphysema. But in sub-Saharan Africa, they were due to neglected tropical diseases, HIV/AIDS, tuberculosis, and malaria. That being said, the trends in the Caribbean and Eastern Europe are concerning for also showing dramatic worsening of health among some populations.

Sociodemographic change is a leading driver of changes in health–except for children, where public health has also made a big difference

While population aging has dramatically changed the burden of illness in most of the world, precipitated by health improvements related to overall increases in income and smaller family sizes, one prominent reduction in deaths seems more directly attributable to public health interventions: the decline in child mortality.

While over 600,000 children died from measles in 1990, the 2010 estimates were only 125,000. Overall deaths among children have declined by nearly 60% since 1970, in a pattern that seems to correspond to major public health campaigns such as vaccination.



Challenges in data collection

No doubt there will be arguments about the GBD 2010 data. It is difficult to determine not only the burden of disease based on representative samples, but also contested metrics like disabiilty-adjusted life years lost or gained due to changes in disease, which are hard to calculate. Much of the argument will be about whether these burden estimates should prompt a reallocation of funding to different diseases.

The GBD2010 conducted an extensive and much-needed series of studies to further justify the measurements used to quantify disability. They used both a household and web-based survey that used paired comparison questions, in which respondents considered two hypothetical individuals with different, randomly selected health states and indicated which person they regarded as healthier. They then used these results to quantify the relative disability caused by conditions on a scale from 0 (implying no loss of health) to 1 (implying a health loss equivalent to death). Most cost-effectiveness studies seem to avoid such careful experimentally-driven estimates of disability or quality-adjusted life-years, using willy-nilly estimates of these quantities that can dramatically change results without any sense of external validity.



What does this mean for the future of public health?

In theory, the burden of disease amenable to funding (the burden of disease that’s preventable through better investments in public health interventions) might be an optimal metric to use for distributing funding – but who knows how to estimate that? At the very least, we can compare modifiable risk factors for disease, such as how much tobacco control investments provide returns in terms of lives saved.

But even these comparative risk assessments are fraught with methodological complexities. For example, many diseases are multifactorial and can’t be attributed to just one nutritional problem (a common issue with nutrition studies these days, as whole books are based on blaming just one nutritional component for the entire obesity epidemic). Christopher Murray, leader of the GBD, summarized some of the debates this way: “the potential for residual confounding of dietary risks, air pollution effects in smokers versus non-smokers, the effects of ambient air pollution on birth outcomes, maternal vitamin A deficiency on neonatal mortality, alcohol on tuberculosis, or intimate partner violence on HIV incidence… In each case, after lengthy and vigorous exchanges with the relevant experts, and when possible external experts, the core team—following the GBD protocol—convened and decided on whether the standard of evidence set for the study had been met.” (The graph below shows estimates of the contribution of different risk factors to disability-adjusted life years lost for (A) men, (B) women, and (C) both):


But no doubt there will be many fights about how to properly synthesize the raw data in the years to come. For some diseases like typhoid, natural history models based on very old parameters describing disease states had to be used to estimate the burden of disease. In other cases, disease pathogenesis itself remains elusive, such as with diabetes and its relationship to different nutritional components, since most of the variation in type 2 diabetes rates is not explained by obesity or genetics.

Regardless of the estimation procedures, another issue with such large-scale evidence is whether to paint the whole world in one color. Yes, hypertension, smoking and alcohol were overall the major killers. But there is much regional variation, and no doubt variation between smaller states and even towns, let alone variations by income.  The leading modifiable risk factor for death in Eastern Europe, most of Latin America, and southern sub-Saharan Africa in 2010 was alcohol use; in most of Asia, North Africa and Middle East, and central Europe it was high blood pressure. Obesity is the leading risk in Australasia and southern Latin America, and also ranks high in other high-income regions, North Africa and Middle East, and Oceania.

So the next steps with the GBD estimates will include piecing together a story of how to address some of these issues through global actions, and others through more regional or localized initiatives…

This post first appeared on the blog EpiAnalysis.

Post a Comment

Your email is never published nor shared. Required fields are marked *


You may use these HTML tags and attributes <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>