5.1 Estimating the burden of disease

Chapter 1 argues that, as with a physician taking X-rays from various angles, the complex relation between health and environment must be looked at from different perspectives to better understand how best to intervene to improve human health. A conceptual framework was proposed to guide the viewing process. This framework took as its starting point those driving forces which, through pressures of one kind or another, significantly alter the state of the environment. The dynamics of these relationships were explored in Chapters 2 and 3.

Chapter 4 went on to consider specifically how lowered environmental quality can lead to increased levels of exposure and risk to human health. Different degraded environments were shown to be associated with specific health outcomes. For example, the role of air pollutants in provoking respiratory diseases and cancers, and the role of contaminated water and food in causing diarrhoeal disease were discussed. However, many environmental health hazards are associated with more than one health problem (Table 4.23) and certain hazards interact, making quantification of environmental health impact difficult.

In this chapter we look at environmental health problems from the viewpoint of the burden of death, disease and disability, and analyse the relative importance of the different environmental factors discussed earlier.

Needless to say, this exercise is fraught with uncertainty. Not only is the state of death, sickness and disability reporting very incomplete and often even arbitrary, but rarely is any effort made to report contributing factors, be these environmental, nutritional or otherwise. Statistics on the "cause of death" and the classification of diseases focus on the body organ in which disease is found and the pathological processes within that organ. Thus the causal role of any factor cannot be derived directly from routinely-recorded mortality and morbidity statistics. Instead, approximations of the role of different factors must be sought which are consistent with the known epidemiology of disease and disability.

In addition, as stated in the WHO Constitution, health is not only the absence of disease, but a state of complete physical and mental wellbeing (WHO, 1946). Many of the environment and development problems described in previous chapters impact on human well-being beyond actual "disease". The deprivations of a life in poverty, of being a refugee or war victim, or of living and working in an insecure or polluted environment have been highlighted. But the limitations of "health status" data mean that this chapter must focus on death, disease and disability. Regrettably, in so doing, the real impact of environmental factors on health is likely to be Underestimated.

Table 5.1 shows two different estimates for the total number of deaths for the major causes of death for 1990-1993. The numbers in the right-hand column have been taken from the World health report (WHO, 1995a), and are based primarily on routinely-recorded mortality statistics. In the left-hand column, numbers from the Global burden of disease study (Murray & Lopez, 1996b) are presented. Some of these numbers have been adjusted as described below. The specific diseases in Table 5.1 account for nearly 83% of all deaths. The biggest "killers" (cancer, heart disease and cardiovascular diseases) mainly affect elderly people and are the most common causes of death in developed countries. Many of these deaths can be considered "natural"; in many countries "senility" rather than a specific disease would be ascribed as cause of death. Deaths from infectious diseases, such as acute respiratory infections (ARI), diarrhoeal diseases and vaccine-preventable infections (Table 5.1) mainly affect children in developing countries. Important conditions in the group of "other identified diseases" include digestive diseases, congenital malformations, maternal conditions and different tropical diseases.

Table 5.1
Estimated global number of deaths =1990-1993

Sources: A. deaths in 1990 according to Murray and Lopez, 1996b; B. deaths in 1993 affording to WHO, 1995a.

Each of the two estimation methods used to produce Table 5.1 involves judgements about the best sources of mortality data, which explains the differences between the two columns for certain disease categories. The estimates by Murray & Lopez (1996b) in the left-hand column are based on the assumption that all of the deaths from "unknown causes" in older age groups are in fact from cardiovascular and other non-communicable diseases.

Murray & Lopez (1996b) have also analysed the different estimates of global deaths for specific diseases in order to avoid "double counting" of deaths. For instance, the sum of all the estimates of early childhood deaths for specific diseases (such as malaria) should not exceed the total number of deaths in this age group. The numbers in the left-hand column of Table 5.1 have been adjusted to take this into account, which explains the lower estimated number of deaths for malaria, tuberculosis and perinatal conditions in the left-hand column.

To clarify the relative importance of different causative factors the analysis must take into account the age at which death occurs, or even better, the number of years of life lost (YLL). Furthermore, not all negative health outcomes are expressed only in terms of mortality. Morbidity and disability are important outcomes that ideally should be taken into account as well. In recent years much attention has been given to comparing different approaches that seek to incorporate years lost due to death, as well as years lost due to morbidity or disability. One such approach which has generated a comprehensive picture of the "burden of disease" covering all major disease and injury categories is the disability-adjusted life years (DALYs) concept (Murray & Lopez, 1996b). We will use estimates of burden of disease based on this approach in the following sections, as these estimates include all diseases of interest to our analysis and all regions of the world. Furthermore, the estimates relate to what is considered "preventable".

Any estimate of the burden of disease that purports to combine the health lost due to a combination of death, disease and disability is going to be very approximate due to the incompleteness of data and lack of agreement on measurement methodologies (WHO, 1997a). Murray & Lopez (1996b) made every effort to overcome these problems and to produce representative estimates for different diseases and regions. For the broad descriptions and comparisons intended in this book, the DALY approach is likely to give the current best estimate of the "true" burden of disease.

Each DALY indicates the loss of a year of healthy life - that is, time lived with a disability or time lost through premature death. The number of DALYs in different regions provides a guide to the relative distribution of disease burden: the higher the DALYs, the greater the burden. For example, the number of DALYs per 1000 people in sub-Saharan Africa in 1990 was about five times greater than in the established market economies, indicating how much greater the burden of disease is in least developed countries.

To calculate disease burden in DALYs, data on premature mortality and disability are combined. The number of YLL is assessed as the difference between the actual age at death and the age at which the person could have been expected to die, given the average age of mortality of an advanced developed country (82.5 years for women and 80 years for men). Next, the incidence of disability due to disease or injury is estimated on the basis of available information pertaining to each community. Different weights are assigned to different disability conditions, according to severity. Finally, discounting and age-weighting systems are incorporated because this methodology assumes that future years of life lost contribute less to the burden of disease than current ones. In total, 500 different conditions or disease sequelae have been separately evaluated. These have been grouped into 96 detailed causes and a variety of cause groups or clusters (Murray & Lopez, 1996b).

Table 5.2
Global YLL and DALYs for major health conditions - 1990

Source: based on data in Murray & Lopez, 1996b.

While most of the DALY burden is due to premature death, the disability component is important for chronic illnesses which are present over many years. This is seen in Table 5.2 where the early death component of DALYs (i.e. years of life lost - YLL) is presented along with the DALYs for selected causes. The DALY for diarrhoeal diseases, for instance, is only 5% greater than the diarrhoeal YLL (Table 5.2), but for chronic respiratory diseases, the DALY is more than double the YLL. The difference is even more dramatic for mental health conditions: DALYs exceed YLL by a factor of nearly 14. This is not surprising given that mental disease patients often suffer long periods of illness before they die.

Some diseases, such as acute respiratory infections (ARI) and diarrhoeal diseases, are of particular importance to children's health. Thirty percent of the estimated number of deaths for all diseases occur before 15 years of age (Table 5.3), but for ARI and diarrhoeal diseases specifically, the percentage is 67% and 88% respectively. (Conversely, only 3% of the deaths from cardiovascular diseases (CVD) and 3% of the deaths from cancer occur in the 0-14 years age group (Table 5.3). The child proportions of DALYs are greater: 48% for all diseases and more than 90% for ARI and diarrhoeal diseases (Table 5.3). This is because the estimated years of healthy life lost are greater when a child dies than when an adult dies. As this chapter will show, the diseases that particularly affect children are also influenced by environmental quality.

Table 5.3
Proportion of global burden of deaths and diseases occurring in children under age 15 - 1990

* % of all deaths in this disease category
** % of all DALYs in this disease category

Source: Murray & Lopez, 1996b.

In the following sections the different health conditions are discussed in the order in which they appear in Table 5.2, with the aim of further clarifying the importance of specific environmental conditions to each category. The conditions discussed are responsible for about 70% of the total global DALY burden and are those for which the environment is a major factor (see Section 4.10).