Global and regional cause-of-death patterns in 1990.

Global and regional cause-of-death patterns in 1990 C.J.L. Murray1 & A.D. Lopez2 Demographic estimation techniques suggest that worldwide about 50 million deaths occur each year, of which about 39 million are in the developing countries. In countries with adequate registration of vital statistics, the age at death and the cause can be reliably determined. Only about 30-35% of all deaths are captured by vital registration (excluding sample registration schemes); for the remainder, cause-of- death estimation procedures are required. Indirect methods which model the cause-of-death structure as a function of the level of mortality can provide reasonable estimates for broad cause-of-death groups. Such methods are generally unreliable for more specific causes. In this case, estimates can be constructed from community-level mortality surveillance systems or from epidemiological evidence on specific diseases. Some check on the plausibility of the estimates is possible in view of the hierarchical structure of cause-of-death lists and the well-known age-specific patterns of diseases and injuries. The results of applying these methods to estimate the causes of death for over 120 diseases or injuries, by age, sex and region, are described. The estimates have been derived in order to calculate the years of life lost due to premature death, one of the two components of overall disability-adjusted life years (DALYs) calculated for the 1993 World development report. Previous attempts at cause-of- death estimation have been limited to a few diseases only, with little age-specific detail. The estimates reported in detail here should serve as a useful reference for further public health research to support the determination of health sector priorities. Introduction Reliable information on global and regional deaths by cause are an essential input to planning, managing and evaluating the performance of the health sector in developing countries. The numbers of deaths by cause influence the manner in which resources are al- located to different service programmes and research activities. An accurate assessment of current death rates by cause in different regions also forms the baseline against which new health programmes must be evaluated. Without a reasonable baseline, we shall not, in 5 or 10 years from now, be able to assess what has worked and what has failed. In addition, reliable information on deaths by cause is an essen- tial input to the assessment of the cost-effectiveness of new technologies for disease control and health promotion. The epidemiological transition, where the cause-of-death structure shifts profoundly from Assistant Professor of International Health Economics, Har- vard Center for Population and Development Studies, 9 Bow Street, Cambridge MA 02138, USA. Requests for reprints should be sent to this author. 2 Scientist, Tobacco or Health Programme, World Health Organ- ization, Geneva, Switzerland. Reprint No. 5495 infectious to chronic diseases, is under way in most middle-income countries and among the richer com- munities in low-income countries (1, 2). Estimating the cause-of-death pattem for 1990 will also provide a quantification of the extent to which this pattern has changed in different regions and for the devel- oping world as a whole. While data on deaths by cause are required for objective planning and evaluation of the health sec- tor, the available datasets are wholly inadequate. A decade ago, claims conceming child and adult mor- tality by disease-specific programmes at the World Health Organization and by individual disease experts exceeded the total deaths in each age group by two- to threefold. Through the efforts of the World Bank and WHO, more consistent estimates of mortality, by cause, under age 5 have been devel- oped although these still remain uncertain. More- over, no plausible or consistent estimates for death over age 5 exist despite their increasing importance in the context of the epidemiological transition. The purpose of the present analysis is to redress these gaps in critical information. We present summary results from the Global Burden of Disease study, which involved over 100 disease experts (see Acknowledgements) and was a basic input into the World Bank's 1993 World development report: Investing in health (3). Bulletin of the World Health Organization, 1994, 72 (3): 447-480 © World Health Organization 1994 447 C.J.L. Murray & A.D. Lopez Attempts to estimate global mortality by cause are not new. The basic problem is that reliable vital registration data, with the cause of death coded by a physician, are available for only a small number of mostly developed countries (4). Estimates of mortal- ity by cause need to be based on other approaches for virtually all of the populations of sub-Saharan Africa, most of Asia, the Middle East and North Africa. Indirect techniques to do this were first developed by Preston to model the relationship between total mortality and cause-specific mortality for broad groups of causes, based on an analysis of historical vital registration data for the developed and a few developing countries (5). In particular, cause-specific mortality was postulated to be a linear function of total mortality. Preston's work has formed the basis of nearly all subsequent approaches to estimating causes of death in regions without vital registration. Several others have refined the approach by estimating equations for specific age groups, incorporating more recent data or examining more detailed lists (6-9).a It is axiomatic that models only capture the rela- tionship between cause-specific mortality and total mortality present in the countries with data. As few developing countries (particularly those with higher mortality) have good vital registration data, these model estimates are largely based on the experience of developed countries with a low mortality. Even the historical data for moderate levels of mortality in developed countries included in Preston's analysis tend to underestimate the mortality from communi- cable diseases. The relevance of models built on his- torical data is also affected by problems of diagnos- tic quality, revisions to the Intemational Classi- fication of Diseases, and basic differences in disease epidemiology in different regions. Furthermore, changes in cause-specific mortality with respect to total mortality are difficult to interpret in view of the parallel decline in the proportion of deaths coded to senility and ill-defined conditions. Our premise is that model estimates of cause-specific mortality are reasonable for large groups of causes. Model-based estimates for detailed causes, however, are probably not valid. Recent studies have employed altemative pro- cedures for estimating specific causes of death based on a review of data from disease-specific surveil- lance systems and the epidemiological literature (10, 11). Estimates of mortality from a particular cause such as malaria can be built up from epide- a Bulatao RA, Stephens PW. Estimates and projections of mor- tality by cause: a global overview, 1970-2015. Unpublished manuscript prepared for the World Bank, 1991. miological data on incidence, remission, and case- fatality rates. The major limitation of the epidemi- ological approach to cause-of-death estimation is the lack of data for many diseases in many regions. There is also a tendency for the epidemiological approach to yield higher estimates than vital registra- tion or model-based estimates, possibly because disease-specific analyses tend to be more inclusive than exclusive. The estimates presented in this report use a com- bination of data sources and approaches, exploiting vital registration data where available, using models of the epidemiological transition to estimate broad causes, and supplementing these with a distillation of disease-specific data sources. Not only have esti- mates been derived for very many more causes than previous attempts at global estimation, but they have also been presented for specific age groups. In con- sidering several competing and exhaustive causes simultaneously, we have also been constrained by independent estimates of total mortality by age. This is clearly not relevant for disease-specific estimates carried out in isolation and is a major reason why the estimates reported here for several diseases are lower than previously claimed. Methods Cause-of-death estimates for the developing and developed world depend first on estimates of the total mortality by age and sex. These deaths can then be attributed to particular causes. The analysis in this study is based on the following eight geographical regions as given in the World development report 1993 (3): the Established Market Economies (EME), the Former Socialist Economies of Europe (FSE), Latin America and the Caribbean (LAC), China (CHN), India (IND), Other Asia and Islands (OAI), Middle Eastem Crescent (which includes North Africa, the Middle East, Pakistan and the Central Asian Republics of the Former Soviet Union) (MEC), and Sub-Saharan Africa (SSA). The demographic esti- mates of mortality by age and sex have been devel- oped by the World Bank. The database for estimating child mortality is unquestionably much better devel- oped than that for adult mortality (12). Indeed, there is considerable controversy among demographers over the levels of adult mortality in some developing regions without good vital registration systems, where mortality was estimated indirectly from census and survey data. For example, the United Nations Population Division and the World Bank estimates of adult mortality by age and sex can differ by as much as 50%, but in general the differences are smaller. 448 WHO Bulletin OMS. Vol 721994 Global and regional cause-of-death patterns The cause-of-death groupings and detailed caus- es examined were initially developed by Murray et al. (9). Their system was extensively modified for the Global Burden of Disease study, preserving how- ever the division of mortality into three large groups: communicable, maternal and perinatal (I); noncom- municable (II); and injuries (III). These are then fur- ther subdivided into several more specific causes. A list of causes selected for the study is given in the Annex. Clearly, this list, by being selective, has omitted some causes which, with further analysis, may justifiably be included in the future. Our approach to attributing death to one single cause is based on the principles of the international classification of diseases: each death is coded to the underlying cause that initiated the sequence of events leading to death. For example, a patient with lung cancer who dies from respiratory failure from a post- obstructive pneumonia is coded to lung cancer. In the case of young children, where several causes may contribute significantly to death, the underlying cause has been selected as the primary cause of death, based on expert opinion about the nature of disease interactions at these ages. One disease can also be a risk factor for another. A patient with cir- rhosis who dies from an oesophageal variceal bleed will be coded to cirrhosis despite the fact that his or her cirrhosis may have been caused by hepatitis B infection at an early age. Finally, deaths attributed to senility and ill-defined causes have been proportion- ately allocated either (if under age 5) to the commu- nicable, maternal and perinatal causes, or (if older) to noncommunicable causes. Our estimates of mortality by cause have been constructed from three types of estimates. First, for regions or parts of regions with good vital registra- tion data, we have used deaths coded by the vital registration system according to the ninth revision of the International Classification of Diseases. This includes all deaths in the Established Market Econo- mies and Former Socialist Economies, 61% of deaths in Latin America and the Caribbean, 23% of those in the Middle Eastern Crescent, and 11% of those in Other Asia and Islands (Table 1). As China does not have a complete vital registration system, a random sample of the population was monitored through the Disease Surveillance Points (DSP) system (13), in which teams (including a physician) review hospital records or interview the family to determine the cause for each death. However, not all deaths in the surveillance sites are captured by the DSP system. Underreporting is estimated at 10.8% in urban areas and 15.3% in rural areas. According to other meth- ods for assessing completeness such as the Brass "growth-balance" method (14), the World Bank has estimated that underregistration of deaths in China is closer to 30%. Hence for China, an adjustment for underreporting was first made using the World Bank-estimated underreporting ratio, and then dis- tributed across urban and rural areas on the basis of the information from the DSP system. For India, the Survey of Causes of Death (rural) provides useful information on lay-reported causes of death. This system collects information via a "verbal autopsy" on about one-half of 1% of all rural deaths in India, based on about 1300 primary health care centres Table 1: Methods for estimating causes of death and percentage of deaths registered, by region Percentage of Non-registered deaths deaths Sample Groups Detailed Adjustment Regiona registered registrationb 1, 11 and IlIl causesc algorithmd EME 99 FSE 99 CHN DSP EPI LAC 61 Models EPI/m X OAI 11 Models EPI/m X MEC 23 Models EPI/m X IND SCD (R) SCD (R) EPI SSA Models EPI/m X a EME, Established Market Economies; FSE, Former Socialist Economies of Europe; CHN, China; LAC, Latin America and the Caribbean; OAI, Other Asia and Islands; MEC, Middle Eastern Crescent; IND, India; SSA, Sub-Saharan Africa. b DSP = Disease Surveillance Points system in China. SCD (R) = Survey of Causes of Death (rural) in India. c EPI = epidemiological estimates. EPI/m = epidemiological estimates and model estimates. d X indicates that an adjustment algorithm was used. WHO Bulletin OMS. Vol 72 1994 449 C.J.L. Murray & A.D. Lopez spread throughout the country. This dataset has been used to establish the size of Groups I, II and III and for some finer information on injuries (Table 1). The second main source is model-derived esti- mates of cause-of-death patterns based on total age- specific mortality. Building on the original work of Preston, we examined the relationship between Groups I, II and III mortality and total mortality for each age group. Data for the latest available year and a year from the 1950s from all countries, assessed as having complete and reliable vital registration, were included to expand the number of countries with moderate or high mortality in the sample. The rela- tionship between Groups I and II and total mortality is non-linear; at higher mortality rates, Group I begins to increase faster. As a result, linear regres- sion equations will tend to underestimate Group I mortality, particularly at higher mortality levels. To address this bias, natural log regression equations were used; the predicted mortality for Groups I, II and III from the equations was then adjusted to equal the total mortality. With a few exceptions, the coeffi- cients for total mortality and the intercept were sta- tistically significant (<0.05). Group III (injuries) had non-significant intercepts at ages 0-4, 15-29, 60-69 and 70+ years. It is important to emphasize, how- ever, that because many of the slopes and intercepts in the age group 70+ were not significant, we have less confidence in predicting even the highest level disaggregration of mortality into the causes for this age group. The third source of estimates is built up from studies by disease experts on the regional epidemiol- ogy of specific diseases. Specialists in diseases or injuries contributed their assessment of incidence, prevalence, remission and case-fatality rates based on review of existing data for each disease. These estimates were carefully evaluated for internal con- sistency utilizing a competing-risks computer model. Any internal inconsistencies were reviewed with dis- ease experts and revised. The age pattern of predict- ed mortality by cause was also carefully reviewed for plausibility. Estimates for Groups I, II and III totals, by age and sex, for the EME and FSE regions were based on vital registration data. For MEC, OAI and LAC regions, vital registration data for the subcomponent of the respective region with good registration were combined with model estimates for the residual parts. In China, the adjusted data from the Disease Surveillance Points system were applied to the total urban and rural population. In India, the Survey of Causes of Death (rural) system was used (15). For Sub-Saharan Africa, the Groups I, II and III totals were based solely on model estimates. All those regions or subregions with model estimates were subject to revision as discussed below. The following approach was used for the detailed causes. For EME and FSE regions, only minor adjustments were made to the vital registration data for deaths due to HIV infection. For LAC, vital registration data were used for the subregion with good registrations and this age-sex-specific distribu- tion of causes (within Groups I, II, and III) was then used in the other parts of that region. For example, the percent of Group II mortality due to lung cancer within any age-sex group would be taken to be approximately similar in the areas with and without registration. Adjustments in both subregions were made for some Group I causes, (e.g., vaccine- preventable diseases) based on specific epidemio- logical data. For China, the age-specific proportions suggested by the DSP system were adopted with minor adjustments. For the majority of OAI, MEC, India, and Sub-Saharan Africa, most of the estimates were based on the epidemiological approach. Alterna- tive methods were developed for clusters of causes including cancers, neuropsychiatric disorders, cardio- vascular diseases, chronic respiratory diseases, and injuries as outlined below. Total cancer deaths for OAI, MEC, IND, and SSA regions by age groups were based on the mod- els. These were then distributed by site as follows. Firsly, the distribution of deaths by site was obtained by multiplying the incidence recorded at IARC (International Agency for Research on Cancer) regional cancer registries and the case-fatality rates recorded in cancer registries with the best follow-up by IARC (16). This distribution of deaths was then applied to the total number of cancer deaths suggest- ed by the models to yield the estimated deaths by site. For these same four regions, the total deaths by age and sex for neuropsychiatric causes were based on the models, whereas the estimates for detailed causes were based on the average percent distribu- tion of mortality within each age-sex group recorded by vital registration systems in the other four regions (i.e., EME, FSE, LAC and CHN). A similar approach was taken for estimating asthma and chronic obstruc- tive pulmonary disease (COPD) from total chronic respiratory mortality, based on vital registration data disaggregrated into the ICD 3- and 4-digit codes for about 20 countries. A large number of deaths in China coded to cor pulmonale were transferred to COPD, based on discussions with those familiar with the cause-of-death coding applied in the DSP system. Estimates of total cardiovascular mortality for OAI, MEC, IND, and SSA regions were based on the models. Four detailed causes, or groups of caus- es, were identified: rheumatic heart disease, ischaem- WHO Bulletin OMS. Vol 72 1994450 Global and regional cause-of-death patterns ic heart disease, cerebrovascular disease, and a new category entitled "inflammatory heart disease" (peri- carditis, endocarditis, myocarditis and cardiomyopa- thies). Small autopsy series, clinical case studies and limited survey data suggest that this last category is an important cause of death in high mortality popula- tions. For each of these four causes, local studies have been used to qualitatively score the relative risk of each cause in each region. Using actual data from EME, FSE, LAC and CHN, these qualitative scores have been converted into percent distributions by cause for each age and sex group. (This is an obvi- ously crude approach given the importance of cardio- vascular causes of death, but it does provide prelimi- nary estimates of the possible mortality due to cardiopathologies in high-mortality populations). Because the pattern of injury mortality by detailed cause is highly variable across regions and within regions, no satisfactory method is available to predict local injury patterns; in China, for example, suicides and drownings predominate. We have there- fore used the average percent distribution of injuries by age and sex for EME, FSE, LAC and CHN to esti- mate the pattern of detailed injuries for the other four regions, supplemented by the available epidemiologi- cal information on injury patterns. Initial estimates of Group I mortality (communi- cable, maternal or perinatal), based on the epidemi- ological approach (i.e., summation of estimates by experts in specific diseases), considerably exceeded the total Group I mortality by 200-300% for men in some age groups in MEC, OAI, IND, and SSA. In these regions, therefore, the reduction in cause- specific Group I numbers to equal Group I deaths produces particularly important differences between the present and previous estimates. As all epidemio- logical estimates were subject to the same critical review, they were all considered equally plausible. Hence, an algorithm was developed to proportionate- ly reduce all Group I causes equally. Essentially, the algorithm was as follows: if the overestimation was less than 10%, then all causes were equally reduced to sum to the total available Group I mortality for that age-sex group. Otherwise an alternative Group I total was defined for each age- sex group using the constant (intercept) in the regres- sion equation, plus 1 standard deviation (Fig. 1). Adjustments were then also made so that Groups I, II and III summed to the total mortality. This new equation defined the Group I upper bound. Where Group I was overestimated by more than 10%, each cause was automatically reduced by 10%. If the resulting estimate was outside of the Group I upper bound, the upper-bound estimate of Group I was used instead and the detailed causes were proportion- ately reduced to equal the new Group I total. If the Fig. 1. Predicted Group I versus total standardized death rates, plus and minus 1 standard error (top and bottom curves), females age 45-59 years. cQ 350 2 30 -- =25 .~~~~~~~~- ---------------;"------- -- C 20 .;; ; " ' 15.X .;;;;;---- _-- e)cn AC:> ar-A Q0 1 nsrnzDU 45U bDU O3U Total death rate I uou resulting estimate, after a 10% reduction, was small- er than the Group I upper bound, then the estimates were used without further modification. Fig. 1 illus- trates how the width of the confidence band increas- es dramatically with rising overall mortality, based on the estimates for women aged 45-59 years. Results Fig. 2 illustrates the profound differences in cause- of-death structure between developed and developing regions due to differences both in population age structure and, more importantly, in the age-specific mortality rates for different groups of diseases. Group I (communicable, maternal or perinatal causes) accounts for 40% of the deaths in developing regions but only 5% of deaths in developed regions. Group III (injuries) causes roughly the same proportion (8-9%) of deaths in both regions but are twice as common among males than females. For the developing regions as a whole, one in every two deaths now occurs from noncommuni- cable diseases. Indeed, the ratio of deaths from Group II to Group I causes, which is a rough indica- tor of the epidemiological transition, is about 5 in China, and 2 in Latin America, compared with about 17 in developed regions and unity elsewhere in the developing world, except for Sub-Saharan Africa where Group I causes are still 2-3 times more com- mon than Group II. The relative importance of inju- ries (Group III) in the cause-of-death structure is least in the industrialized countries and India (6-6.5% of all deaths), rising to just under 10% in Eastem Europe and Latin America, and to almost 12% in China. Detailed mortality estimates for 120 causes by age and sex, separately for developed and developing regions, are given in the Annex; to facilitate the WHO Bulletin OMS. Vol 72 1994 451 C.J.L. Murray & A.D. Lopez Fig. 2. Probabilities of dying from three groups of causes for males and females, by age group and region, 1990. (EME, Established Market Economies; FSE, Former Socialist Economies; CHN, China; LAC, Latin America and Caribbean; OAI, Other Asia and Islands; MEC, Middle Eastern Crescent; IND, India; SSA, Sub-Saharan Africa). Males 0-14 years .... ... ..... ..... +.es........ . j.. j.... .......... ............. . v 24.3 e............. m. . . . . 14.7 12.2 12.8 8.0--: 5.0 1.4 2i EME FSE CHN LAC OAI MEC IND SSA 40 0 40 30 20 10 0 40 30 20 10 0 Females L.J__ _ _M EME FSE CHN LAC OAI MEC IND SSA 15-59 years 32.2 22.9 11.2 150 :3 1. .6- EME FSE CHN LAO OAI MEC IND SSA 60-69 years aQ a 0 Group I: Communicable Group II: Noncommunicable _ Group Illi: Injuries computation of rates, this tabulation also gives the estimated populations at risk in 1990, by age and sex, for the two broad regions. It is important to note, however, that there is substantial epidemiological diversity among both developed and developing countries. Some popula- tions in Latin America and East Asia have patterns of mortality similar to those in industrialized coun- tries and very different from those in other develop- ing regions. Death rates are considerably higher in Eastern Europe than in other developed countries. These differences are clear from Fig. 2 which shows the risks of dying for three age groups (0-14, 15-59, and 60-69 years) from the three broad groups of causes in each region. On average, a newborn child in the Established Market Economies has only a 1% chance of dying before reaching adulthood (age 15 years), which is markedly lower than the 20-25% risk in Sub-Saha- ran Africa. Almost all of this excess mortality arises from Group I causes. The risk of childhood death is also comparatively low in China, with the probability WHO Bulletin OMS. Vol 72 1994 40 30 20 10 0 40 30 20 10 0 40 0-14 years 30 .. .......... wz...................... .... ... 21.7 20 ------------- 162 12.5 ~ 110.1 1.1 2.0 452 Global and regional cause-of-death patterns of dying being roughly similar for boys and girls. The only other region with a relatively small differ- ence in survival chances between males and females is MEC; otherwise, male death rates are considerably higher than those for females and are due almost entirely to the higher death rates from Group I causes. During adulthood (15-59 years) the risk of death in all regions is largely determined by the noncom- municable diseases, although Group I causes still contribute significantly in Sub-Saharan Africa and, to a lesser extent, in India. The risk of death from injuries is a major public health problem among males in all regions, but is particularly high (one- third of the overall risk) in FSE. Interestingly, the risk of death in adulthood (15-59 years) from non- communicable diseases is higher for both men and women in all developing regions compared with the Established Market Economies, something which is not widely appreciated but is consistent with the pat- tem observed in the industrialized countries earlier this century. The excess is less marked for males, however, owing to the impact of tobacco (17). Smoking-related mortality is one of the principal fac- tors underlying the comparatively high risk of death (18.5%) from noncommunicable diseases among adult males in Eastem Europe, typically accounting for about 40% of male deaths at these ages. At older ages (60 to 69 years), mortality is dominated by the noncommunicable diseases, although even at these ages a substantial risk of death from communicable diseases in Sub-Saharan Africa remains. At these ages, the comparative advantage in survival prob- ability enjoyed by the EME countries is considerably less (the excess compared with SSA and IND being 1.5-2), which reflects the progressive convergence of mortality risks with advancing age. Developed regions (11 million deaths) Comparatively few deaths in developed countries are now due to communicable, matemal or perinatal causes (see Table 2). Of half a million such deaths, a quarter of a million involve respiratory infections in middle age or particularly old age, and should per- haps be considered together with the noncommuni- cable respiratory deaths. The remainder include peri- natal conditions (90 000 deaths, virtually all occurring in the first few weeks of life), and HIV infection (40 000 deaths in 1990, but with substantial increas- es expected). Where communicable diseases are rare, the large majority of the deaths are due to noncommunicable diseases: cardiovascular diseases alone cause 5.3 million deaths a year (half of them from ischaemic heart disease and a quarter from stroke), malignant neoplasms caused 2.4 million deaths (0.5 million from lung cancer, plus 0.25 million from other tobacco-related cancers, such as mouth, pharynx, oesophagus, pancreas and bladder), and the most important remaining category is respiratory disease, which causes 0.5 million deaths (chiefly from chron- ic obstructive pulmonary disease). The importance of diabetes is underestimated in these Tables, because it can also cause death indirectly, by increasing the incidence of heart disease and stroke. Table 2: Distribution of deaths from three groups of causes, by region, 1990 No. of deaths (x 1000) attributed to: I. Communicable, maternal and perinatal 11. Noncommunicable Regiona causes causes Ill. Injuries Total EME 439 (6.2)b 6 238 (87.6) 445 (6.2) 7 121 FSE 136 (3.6) 3 264 (86.8) 362 (9.6) 3 762 CHN 1 343 (15.1) 6519 (73.4) 1 023 (11.5) 8885 LAC 966 (32.3) 1 733 (57.9) 293 (9.8) 2 992 OAI 2 306 (41.8) 2 736 (49.6) 477 (8.6) 5 519 MEC 2 026 (46.2) 1 966 (44.8) 392 (8.9) 4 384 IND 4 060 (43.3) 4 700 (50.2) 611 (6.5) 9 371 SSA 5 415 (68.2) 1 898 (23.9) 624 (7.9) 7 937 World 16 690 (33.4) 29 055 (58.1) 4 227 (8.5) 49 971 a EME, Established Market Economies; FSE, Former Socialist Economies; CHN, China; LAC, Latin America and the Caribbean; OAI, Other Asia and Islands; MEC, Middle Eastern Crescent; IND, India; SSA, Sub- Saharan Africa. b Figures in parentheses are percentages. WHO Bulletin OMS. Vol 72 1994 453 C.J.L. Murray & A.D. Lopez Developing regions (39 million deaths) Age 0-14 years (15 million deaths). At present, about 13 million children aged 0-4 years die each year in the developing regions, the three chief causes being conditions arising in the perinatal period, diar- rhoea, and acute respiratory infections. Other leading causes include measles, malaria, tetanus and pertus- sis. There are about another 2.2 million deaths each year at ages 5-14 in the developing regions, of which 1.3 million involve communicable diseases (primarily the same ones that dominate the under-5 pattern, but with the notable addition of 150 000 deaths from tuberculosis). Injuries (most notably from drowning or motor vehicles) are a major cause of death throughout childhood, accounting for almost one million deaths a year. Age 15-59 years (10 million deaths). One in 5 male deaths and, significantly, one in 3 female deaths, among adults are due to Group I conditions. Tuber- culosis is a leading killer of young adults, claiming over 1.2 million lives each year, as are maternal causes, acute respiratory infections, HIV infection, malaria, diarrhoeal diseases, and syphilis. Signifi- cantly, the noncommunicable diseases dominate the cause of death structure, accounting for half of all deaths in this age group. The principal causes of pre- mature adult mortality are similar for the developing and developed regions. In addition, several sites of cancer are major causes of adult death in developing countries, most notably liver cancer, oesophageal cancer and oropharyngeal cancer, with the mortality in each case being higher for males than females. Lung cancer is already a major cause of adult male mortality in developing countries. Pericarditis, endo- carditis, myocarditis and cardiomyopathies, as well as rheumatic heart disease, are significant causes of death among adults. Among the injuries, in addition to motor-vehicle accidents and drownings, occupational injuries claim the lives of about 86 000 male workers and some 27 000 female workers at ages 15-59 years. There is also a very substantial mortality among adults due to suicide (480 000 deaths, 45% of whom were women) and war (about 160 000 males and 70 000 females). Comparative mortality is perhaps better assessed by examining the risks of death. The probability of adult death from selected causes, as well as broad groups of causes, is shown in Table 3. What is most striking about this Table is the comparatively poor survival chances of adult males in Eastern Europe. Almost 3 in 10 males reaching age 15 can expect to die before age 60, twice the level in the industrial- ized countries and higher than anywhere else in the world, except Sub-Saharan Africa. Much of this excess is attributable to higher mortality from inju- ries, heart disease, stroke and lung cancer. The risk of female mortality varies from 7% in the industrial- ized countries to 31% in Sub-Saharan Africa. Among African men, the risk of death from tuberculosis is currently four times that from AIDS. Age 60. (14 million deaths). Only about 10% of deaths in old age are due to Group I causes, primari- ly tuberculosis and acute respiratory infections; the vast majority of deaths at older ages, as in developed countries, are from noncommunicable diseases. The leading causes of death are chronic diseases similar to those identified at earlier ages of adulthood; details of the numbers of deaths estimated in 1990 can be ascertained from the Annex. Discussion The estimates presented here agree reasonably well with those previously reported for specific diseases or groups of diseases. They do not agree exactly, however, and we believe it is important to emphasize why. More than 100 diseases and injuries have been analysed, substantially more than the one to five conditions which have generally been estimated in specific disease reviews. This fact alone will tend to lead to lower mortality estimates simply on the basis of competing causes. Unlike many previous discus- sions of specific causes of death, we have been constrained to make the cause-specific deaths by age and sex sum to the total mortality for each age and sex group. Awareness of competing causes of child mortality has naturally led over the past five years to estimates that are more consistent with each other. The same cannot be said of estimates of the adult causes of death which have not been compared with each other in a consistent fashion. The approach followed in this analysis is to try to exploit all existing data sources, based on a pru- dent assessment of their completeness, reliability and relevance. Public health research has led to a very substantial amount of data from community surveys, registration systems and the like which can provide important insights into cause-of-death patterns. Vital registration, or sample vital registration, or surveil- lance systems (where complete registration is not feasible) will always be the method of choice to pro- vide continuous mortality data, provided that the cause of death is reliably certified. The goal of ad- equate registration of vital events is still beyond the means of many developing countries; in the mean- time, much can be learned from other sources includ- ing hospital-based data, partial registration, and com- munity studies that inform public health priorities. Neither the accuracy, nor the inaccuracy of the estimation procedures employed in this study should WHO Bulletin OMS. Vol 72 1994454 Global and regional cause-of-death patterns CD CD C/)co coN to uCD.N, 0 .0~~~~C CDN CD C)( 0)0o) co NC\ ')0 N~~~~c 0 Cu NL LI- ~~~~~LCML E LA) Q\ 0o ~CO Eu C~ 0 <u\Y 0 co CUL coN ci 0 - CJ C 0 0)CEw co U- 0 0O 0 N 0) C~~~~~~C 0 co0 WHO Bulletin OMS. Vol 72 19 00qc 1' -* q '- LO 0) q* q0OCD '- C') C') --CD CD "I C') M o .6 .C. .6.6 .N co 0) C')CM) C') CD LA C') CM LA CM. OCD 0) LA) LA6- 0 C\D 0 0) LA) 0 C') N 0 CD N LA) N 0 co N CD 6 07) 0 N 0 coC') N 67 6~ 0 'DR 0 07) 0 C'I)C') 0 N- N 0 0) 0 LO 0 CM 0 LA 6~ C') C') CM CD6- c'o 6 I I 6 C>I Ic, 6 6 r~I I'm 6) 6 N C') N 0r) 0) 0z CD 0 0 C'O N C')! 0) Nm LA CO CD C')LA N CD LA C' LA 'C? C') LA 0) 0) C') CD CD C') N C') C') 0 co (D CM C-C C > 5 C/ 0 0 -ia 0 0 -5 0) 0 LO 0 N CD CD 0 0)LA CD CD 6 0 C.) 0 C') LA 0) C') 0 N) 0 Cu 0) .C Cu -J'o C'I) C'I) CD cD 0 LOC') 0 0') C') C') 0 CD? C') CD) C')6 0 C') 6Y CO CD -C 0 E 0a) T- 0) CM C\D LA00) C'I) CD? CM CD 0 LO C') C') C') CD? LO 0) Cu CM CA co 0 co 0 Cu CY) N C\) CD 0 LA C') CD 0) CD4 LO C\D 0) C'f) 0~ C') CY) Cu CZ -5) .5) 0a cu$ 0) cc 0 01 N 0 co C'I) 0 0 N 0 0) 0 LA)C') 6 C') 0 LO CD LA) 0 0 0 CM CD CM) 0) CD) C') LO 0 LA 0) N 0 01) 0 0 CuE Cus C') C') CDC' 0 CD C') 0 CDC') 0 LO 0 0 C') 0 N~ 0 CY) 0 C') N\ 0 C') 0 0) 0 LA) 0 0 6) co 6 CD 67 CD 6 6o N7 0 6- N 6> co 0i 0 C') CD co 0 C\D C') 6 0 0LA 0 C'I) 00 CD? Nt 0 0 LCD 0 CD N6 N-6l LA C\D C') N C')tCo4 C~D L6 11C') Lrj N CD C\j 06 CDr C'1 CM CD 0) N 0 N 0) 0) It 0) c66 LA CD co0 CD 0 Cm CCDi 00 66c LO0 N1 C') N N> C') 0 LA 0 NC') NM C'I) o o 03) LA) cD N: 11-0 LA 0 C')- o C') T-0 C'MN- N- Ni C') N- N 6~ > 0 c cuo .co0 E0 E 0nCZC 0 E *0 C%S 0 Cu CD Cu 0 0) Cu E 0 cou C E *0 0 Cu a) n co > co 0D Cr Cu 0 -~ . co a. 00 0Q 455 C.J.L. Murray & A.D. Lopez be exaggerated. Vital registration is likely to provide a good and, in many cases, a very good basis for determining the causes of death. Nevertheless, even these returns are often faulty, suffering from poor medical certification practices, diagnostic biases, sociocultural influences and so on. In the absence of vital registration, well-focused, well-maintained dis- ease surveillance systems in defined populations can be expected to yield good epidemiological data on the causes of death. Their utility, as for vital registra- tion, depend clearly on good coverage of events, reli- able certification, and internationally comparable diagnostic procedures. Lay-reporting systems are likely to be only of use for determining broad cause- of-death categories, but even these "real" data may be preferable to indirect estimates from models of the mortality transition. Such models, as used in this study, are predicated on the average experience of the more developed countries at earlier stages of epi- demiological transition and thus inherently assume the same cause-level relationships for contemporary developing countries. This is clearly highly conten- tious and we have therefore tried to adjust the model-based estimates, wherever possible, using direct epidemiological evidence. In some cases (such as for cancer), we have preferred to use the indirect estimates of mortality, suggested by the models, rather than estimates based on cancer incidence because the level of underreporting by registries in some areas (e.g., India) seemed to us to be sufficient- ly high to invalidate this approach. Communicable, maternal and perinatal causes remain an important unfinished agenda in the devel- oping world, much more so in Sub-Saharan Africa, India, the Middle-Eastern Crescent, and Other Asia and Islands. Some causes of mortality are now well recognized but still remain major public health chal- lenges including acute respiratory infections, diar- rhoeal diseases, measles, tetanus, HIV infection, malaria, and maternal causes. Other Group I causes such as syphilis, meningitis, and especially tubercu- losis are underappreciated as causes of mortality. Models of the epidemiological transition imply a steady decline in communicable, maternal and peri- natal mortality through time. It would be foolhardy, however, to take such a complacent attitude towards Group I mortality in light of their continued high levels in many parts of the world and increasing trends for some diseases, e.g., HIV infection and possibly malaria. While recognizing the need for continued vigi- lance over communicable, maternal and perinatal mortality, it is important to realize that globally the noncommunicable diseases have already emerged as the leading causes of death in developing regions. Even in poor countries, the epidemiological transi- tion is under way with profound implications on the demand for health care to address the burden of chronic diseases. Moreover, with a number of cost- effective interventions targeted to communicable dis- ease mortality in children, it is reasonable to expect the proportion of mortality due to noncommunicable diseases to increase. At present, the risk of death from noncommunicable diseases during adulthood (15-60 years) is considerably higher in the develop- ing world than in the Established Market Economies, suggesting that the future, in effect, has already arrived. This article does not provide information on the time trend in noncommunicable diseases. The histor- ical, statistical returns for developed countries and data for those developing countries with long series of vital registration data suggest that age-specific rates of noncommunicable diseases have declined throughout most of this century. Specific trends in noncommunicable diseases as well as the overall trend are hard to predict, as most of our causal mod- els do not explain the sustained decline observed over the past century. While this decline is likely to continue, behavioural change in many developing regions may slow or alter this trend. Something like 50-60% of adult males in developing countries are regular smokers, and on current trends an epidemic of smoking-related mortality is inevitable (18). As a consequence, increases in lung cancer, other cancers causally related to smoking, chronic lung diseases, and cardiovascular diseases may be expected. The estimates reported here are the result of applying several different approaches, analysing numerous data sources (some of which were not pre- viously available), and making a large number of decisions about the relative validity of one estimate versus another. We hope that by publishing these estimates, despite their imprecision, we will encour- age not only the further development of methods and data sources to improve on them, but also provide a reference to guide policy in assessing the major pub- lic health issues contributing to the disease burden at the end of this twentieth century. We have empha- sized the importance of developing a baseline esti- mate of the cause-of-death pattern in all age groups in developing and developed regions for the pur- poses of planning, managing and evaluating health sector investments. Equally important for a compre- hensive assessment of population health status are non-fatal health outcomes, which have not been addressed here. Mortality data, despite their limita- tions, are more widely available than comparable information for morbidity or disability; hence the estimation of cause-of-death patterns, despite the substantial uncertainty involved, is more reliably informed by empirical data. Nevertheless, there is WHO Bulletin OMS. Vol 72 1994456 Global and regional cause-of-death patterns clearly an urgent need for comprehensive and com- parable data on non-fatal health outcomes. The broader issue of measuring both mortality and dis- ability is addressed by the Global Burden of Disease study which provides a new integrative framework for assessing and monitoring the health of popula- tions (19-21). Acknowledgements We should like to acknowledge the financial and/or techni- cal support of the Edna McConnel Clark Foundation, the Rockefeller Foundation, the World Bank, and the World Health Organization. We would particularly like to thank Richard Peto for his comments on earlier drafts of this article and Caroline Cook for her tireless efforts. The follow- ing individuals generously contributed time and technical advice to this undertaking: C. Abou-Zahr, M. Adams, M. Adrian, P. Arthur, R. Ashley, A. Ashworth-Hill, K. Bailey, D. Barmes, L. Barnes, R. Beaglehole, M. Belsey, R. Berkelman, S. Berkley, S. Berman, P. Blake, B. Bloom, M. Blossner, J.L. Bobadilla, L. Brabin, U. Brinkman, J. Broomberg, C. Broome, R. Bumgarner, D. Bundy, A. Burton, J. Campbell, P. Carlevaro, P. Cattand, J. Cattani, M. Chamie, L.C. Chen, C-M. Chen, E. Cooper, P. Cowley, D. Daumerie, P. Desjeux, H. Emblad, R. Etzel, J. Ferlay, P. Fine, J. Fomey, J.C. Funck, A. Galazka, M. Garcia, M. Garenne, S. Gillespie, T. Godal, A. Goerdt, J. Gorstein, S. Gove, R. Govindaraj, M. Grant, R. Guidotti, W. Gulbinat, 1. Gyarfas, F. Hamers, H.R. Hapsara, J. Harris, L. Heise, L. Heligman, P. Heller, J. Hempel, K. Hill, A. Hill, C.J. Hong, H. Jamai, D. Jamison, J.P. Jardel, E. Jimenez, F. Kaferstein, A. Kalache, M. Kane, P. Kenya, N. Khaltaev, D. Kilpatrick, H. King, B. Kirkwood, A. Kochi, J. Kumaresan, M.H. Leclerq, N. Lee, L. Lloyd, J. Lob-Levyt, L. Lopez Bravo, D. Mabey, A. Mann, P. Mahapatra, 1. Martin, G. Mayberly, P. McKeigue, A. Measham, G. Medley, J. Menchaca, M. Mercier, T. Mertens, E. Michael, M. Michaud, A. Moncayo, R. Morrow, Y. Motarjemi, K. Mott, S. Nadeen, D. Negrel, W. Newbrender, M. Noel, G. Oakley, M. Orzeszyna, M. Parkin, D. Peterson, A. Pio, P. Pisani, A. Preker, J. Pronczuk, E. Pupulin, X. Qiao, G. Quinke, C. Ramachandran, R. Rannan-Eliya, H. Remme, J.M. Robine, C.J. Romer, M. Rosenberg, R. Rothenberg, P. Sandiford, N. Sartorius, A. Schapira, E. Sherwin, A. Silman, G. Smith, P. Smith, B. Smutharaks, J. Stjernsward, R. Stoneburner, T. Studwick, M. Subramanian, D. Symmons, M. Thuriaux, B. Thylefors, 1. Timaeus, A. Tomkins, C. Torel, J. Tulloch, P. Vaughn, S. Vidwans, F. Vinicor, R. Waldman, G. Walker, D. Weil, J. Wenger, W. Whang, E. Wheeler, R. Wilkins, G. Yang, R. Yip, Z. Yusef, and A. Zwi. However, the views expressed in the article are entirely those of the authors. Resume Repartition mondiale et regionale des causes de d6ces en 1990 Depuis longtemps, les renseignements sur les causes de d6ces sont la base statistique utilis6e pour faciliter la d6termination des priorit6s de sant6 et surveiller l'efficacit6 des interventions. Les donn6es fiables sur les causes de d6ces sont toutefois loin d'etre largement disponibles. Le pre- sent article donne pour la premiere fois des esti- mations mondiales et regionales completes des causes de deces pour les huit regions du monde d6finies par la Banque mondiale dans son Rap- port sur le d6veloppement dans le monde de 1993, separement pour les hommes et pour les femmes et pour plus de 100 causes de d6ces. Ces estimations ont ete 6tablies pour les classes d'age suivantes: 0-4, 5-14, 15-44, 45-59, 60-69 et 70 ans et plus. Les tentatives ant6rieures pour donner des estimations mondiales et r6gionales des causes de deces se sont limit6es A une seule cause de d6ces, tout au plus A quelques-unes, et A une tranche d'age particuliere (en g6n6ral, le nourrisson et le jeune enfant), ou bien ont donn6 des estimations pour des categories de causes de d6ces tres larges, d'une utilit6 douteuse pour d6fi- nir le besoin en strat6gies d'intervention de sant6 cibl6es. Plusieurs sources de donn6es ont ete utili- s6es pour proc6der aux estimations. L'enregistre- ment des statistiques de vie, soit pour la popula- tion entiere d'un pays (lorsque ces donn6es existent), soit pour un 6chantillon de sites d'enre- gistrement, a ete utilis6 lorsqu'il 6tait disponible, en ajustant sur la sous-d6claration et les erreurs de declaration de la cause initiale de d6ces. Pour les populations pour lesquelles ces donn6es n'existaient pas, on a mis au point des modeles de relation niveau-cause de d6ces, d'apres ce que l'on sait dans les pays ou les donn6es sur les causes de d6ces sont fiables depuis une quaran- taine d'ann6es. Ces modeles ont fourni une esti- mation preliminaire de la structure des causes de d6ces, qui a par la suite ete r6vis6e en s'ap- puyant sur des etudes au niveau de la commu- naute et les statistiques hospitalieres. Les mo- deles (6tablis en prenant ± un 6cart type de la constante de r6gression) ont 6galement produit un algorithme statistique servant A determiner les limites de plausibilit6 des estimations. D'apres ces m6thodes, les maladies non transmissibles, et principalement les maladies car- dio-vasculaires, les cancers et les pneumopathies chroniques, sont les premieres causes de d6ces dans le monde et sont maintenant devenues des causes majeures de d6ces dans les pays en developpement. Ainsi, en 1990, d'apres nos esti- mations, 14,3 millions de personnes sont dec6- d6es de maladies cardio-vasculaires (5,1 millions de cardiopathie ischemique, 4,6 millions d'acci- dent c6r6bro-vasculaire) et 6,1 millions des suites d'un cancer (dont pres d'un million d'un cancer du WHO Bulletin OMS. Vol 72 1994 457 C.J.L. Murray & A.D. Lopez poumon). Si la proportion de deces par maladie chronique majeure est plus faible dans les pays en d6veloppement que dans les regions d6velop- pees, les risques de d6ces sont sup6rieurs pour de nombreuses causes, en particulier chez les hommes. II apparalt donc qu'il est urgent d'interve- nir contre les maladies non transmissibles, et en particulier contre les habitudes nuisibles pour la sante sous-jacentes a leur 6tiologie, et ce quel que soit le niveau de d6veloppement sanitaire des pays, mais plus encore dans les pays en d6velop- pement, qui comme la Chine ont fait des progres considerables dans la lutte contre les maladies infectieuses. II reste toutefois que la liste est encore longue des mesures a prendre pour diminuer consid6ra- blement la mortalite pr6maturee par les maladies infectieuses majeures; en effet, dans leur ensem- ble, ces maladies tuent environ 16,7 millions de personnes chaque annee qui, a un demi-million pres, vivent toutes dans des pays en d6veloppe- ment. Les premieres causes de d6ces en 1990 par cette cat6gorie de maladie sont notamment les infections respiratoires aigues (4,3 millions de d6ces), les maladies diarrh6iques (2,9 millions), les causes p6rinatales (2,5 millions), la tuberculo- se (2,0 millions), la rougeole et le paludisme (pres d'l million de d6ces pour chacune de ces mala- dies). Les traumatismes (intentionnels ou autres) sont des causes tres importantes de d6ces a tous les niveaux de developpement sanitaire, le nom- bre r6sultant de d6ces etant estim6 a 4,3 millions en 1990, dont 850 000 par accident de la route, 800 000 par suicide, 380 000 par noyade et 300 000 par homicide et faits de guerre. Ces estimations devraient etre periodique- ment remises a jour au fur et a mesure que des donn6es plus recentes et plus fiables seront dis- ponibles. Elles devraient fournir en attendant une base d'information utile permettant de fixer des priorit6s sanitaires dans le monde. References 1. Omran AR. The epidemiological transition: a theory of the epidemiology of population change. Milbank Memorial Fund quarterly, 1971, 49: 509-538. 2. Frenk J et al. Elements for a theory of the health transition. Health transition review, 1991, 1: 21-38. 3. The World Bank. World development report 1993. Investing in health. New York, Oxford University Press, 1993. 4. Ruzicka LT, Lopez AD. The use of cause-of-death statistics for health situation assessment: national and international experiences. World health statistic quarterly, 1990, 43: 249-259. 5. Preston SH. Mortality patterns in national popula- tions. New York, Academic Press, 1976. 6. Hull TH et al. A framework for estimating causes of death in Indonesia. Majalah demografi Indonesia, 1981, 15: 77-125. 7. Lopez AD, Hull TH. A note on estimating the cause of death structure in high mortality populations. Popu- lation bulletin of the United Nations, 1983, 14: 66-70. 8. Hakulinen T et al. Global and regional mortality patterns by cause of death in 1980. International journal of epidemiology, 1986, 15: 226-233. 9. Murray CJL, Yang G, Qiao X. Adult mortality: lev- els, patterns and causes. In: Feachem, RGS et al., The health of adults in the developing world. Oxford, Oxford University Press (for the World Bank) 1993: 23-111. 10. Lopez AD. Causes of death in the industrialized and developing countries: estimates for 1985-1990. In: Jamison, DT et al., eds. Disease control priorities in developing countries. Oxford, Oxford University Press (for the World Bank), 1993: 35-50. 11. Pisani P, Parkin DM, Ferlay J. Estimates of the worldwide mortality from eighteen major cancers in 1985: implications for prevention and projections of future burden. International journal of cancer, 1993, 55: 891-903. 12. United Nations. Child mortality since the 1960s: a database for developing countries. New York, United Nations, 1992. 13. Department of Hygiene and Immunization, Minis- try of Public Health and Chinese Academy of Preventive Medicine. 1990 annual report on dis- ease surveillance in China. Diseases surveillance report 1. Beijing, Hua Xia Publishing House, 1992. 14. Brass W. Methods for estimating fertility and mortal- ity from limited and defective data. Chapel Hill, NC, Carolina Population Centre, Laboratories for Popula- tion Studies, 1975. 15. Registrar-General, Government of India. Survey of causes of death (rural). Annual report 1988. New Delhi, Government of India, 1990. 16. Parkin DM, Pisani P, Ferlay J. Estimates of the worldwide incidence of eighteen major cancers in 1985. Intemational joumal of cancer, 1993, 54: 594-606. 17. Peto R et al. Mortality from tobacco in developed countries: indirect estimates from national vital sta- tistics. Lancet, 1992, 339: 1268-1278. 18. Peto R, Lopez AD. Worldwide mortality from cur- rent smoking pattern. In: Durston B, Jamrozik K., eds. Tobacco and health 1990: the global war. Pro- ceedings of the Seventh World Conference on Tobacco or Health. Perth, Health Department of Western Australia, 1990: 66-68. 19. Murray CJL. Quantifying the burden of disease: the technical basis for disability-adjusted life years. Bul- letin of the World Health Organization, 1994, 72: 429-445. 20. Murray CJL, Lopez AD. Quantifying disability: data, methods and results. Bulletin of the World Health Organization, 1994, 72: 481-494. 21. Murray CJL, Lopez AD, Jamison DT. The global burden of disease in 1990: summary results, sensiti- vity analysis and future directions. Bulletin of the World Health Organization, 1994, 72: 495-509. 458 WHO Bulletin OMS. Vol 72 1994 Global and regional cause-of-death patterns cn _ E Cu 0 X 0) .2 cE U co- ) ~ ~ 0 ;- c^~~~~~~ eo WHOcoB. WHO Bulletin OMS. Vol 72 1994 a)c U) a) N tf N tf lq0 ID I, Cl£) 0) a CO V a) U)0a) V I c~oe(V o r-'LO 6c V LAO (0(0C') 0)L LA Ct) N NC0 Lei C)C)r- -- 0 NN~ \iN'\- C\)N' co CI) (0NC C'I) 0) N-' c( CO ~ ~ ~ ~ ~ ~ ~ ) -O0C) 0(~ (0 '- (D D _' r-: 0 ( C\i N' c-_ r- q0~(>N (R ci c? _: cl f, (O 0o C4 13cm a _ N CMia) t (D - OD CO _ *tC\i CYi 0) N\ C') N\ NO Ci -~ ' i .h U) ' co' ''0 t O't)0)(D' 0) CY)co'~ C') CMN'-0 c C-ic '-0 LON qt NC C')'- C C o0(0 (0N '' A '''''''ND8&|@@|@@@@@@@|ClA ' ' ' ''' 1- U)(DL ' N 0' NC (D C') CM NN -C'- ( --- - CY) COC'f) CY) Li CO) CL C't C\)ui cli clC') N Ico L) cliC' (0 clil O 2 ,.. l~~ ~~~ lCO C\ N~-. (0.'N N LO ,7-: 4 CO r- C i L_ c) t cur ;rQ CI U)U) -2 co 7 a)) ODQ) * J 0 >'°iD (D o3 E0 ( CDX _ v a) _ <t > c «i o 0 C)6 cliC') C') Nt OD0CC) CM Co - 00DC _' LOC')'- N 0 N'-t-N 0 url LO C -- M u- W M C)M r- . . 0 - 0) - OD C" (3 - O CM P- CM cs 't-c m -T "gr N_ ) __*'ILAN -) C LLOAN0'C_ LA C')'-r-~ C' C) (D0 CM . ~ ~ c.s.DCq LI CM-CM CM CM) '' ) A) ''-0 N Nr0'-N' ' -N LA \J_C\ -(0 (CD 0L (0 N-L) _00 ))O X 0LOo LOv a) _ co c 'T 0- u(D LO CM \j 0)en cJ c ~~~~c)D a) S co >o Ca lL.0 )0 o .D Cuo O O OO < cs z < < < << m mMm -E Cu5~C 0 CE :3 ct% co I= E -Du a) C COCu U0 c CZ -'c uM C/) < a) a: a. u- 0 -L ..-...-C) 0000000 < <<« << o 0i a) co 0 0 0 ec Ul) a) U) a1) C. .C 4c 459 (j) a1) (aw CD CD to 't 't CZ) (a)CD CD co CD CY) a) (D LO 't 6 C> a) E L)U- CD a) (aX enR Cucaco 0 -c r- CIO a) 0 z 0 0 0 0 0. 0 0) 0) a- 0)C- U) 0 l- enX en n (0 LAi LA LO C'I) C'I) 0") CY) C') (0 (Dco q LO '-0 N\ LA C'f)- LA N7P-~LA (0 (0 C') N\ Nl-~ csiCM (0 C')"N 0Y) NM (0(0 N\- 0o ) 0)00) (0 CY) N\(o Us N- NC\'- LO LA CY) N\ C'I) (00)- 0c 0) 0) 0- 0) 0- N\ - 1 (D o c'i c c6 T CM N- C.J.L. Murray & c0a)cm 0 0 0 0 0 a 0 0) 0) *- 6 *0) 0 .0 co n U) X en 0 c c - 0 0 - 0 0- co 46 46 U) 0 .-c 0) 0 z In E a) U- 0) a, 0) cc a,K 0 C') a,) N~ LO LO 0 U)a) 0) 0 r- 0: a,co 6 0 A c 0 :6 CO co~ aCCM O - t o *_ oa A.D. Lopez cc NC' ccCD ':N 0CO _- t o4 cnt *'i N a C_ _ Co N_cc CrCC)0 0 CY)0CY) )t- 0C)c ccua, a, _- N 00o N(D C) a) c)c ( C') N N N "t coo0 a, rN ) C')CM C'_ C oC 0 C CM co CM Cn - CO ' N1- N c0 CM _ ,1 1190 ClcN '-NQ N N N 0) NN 0-NCY C',)N- LO - C') N ,It CI) 1f coc C\) U. - - 0,~)C) -cco) CY) a c0 a ,7 C c LO C' 'it -C)C') CO cq Cli cc V: cc N 0 Nqcc cc N C) a ccd-C') 0C)OtNcc C c C' - - CYcc 0N N CM c ND0 C' cc 0 cc c cco l-cc~ P,-a,c cfD _ CM CM cl- P- O) MO- - CII C\j CY) ceN _U)c cc CM U') CM0 LO O 0 'NU a, C')@@ C' C') CM) C _q cn ck - CM O t 1-L - rl CV LOCY) cc ) c) NN O - co 0 NM .Nc cc N cc U-( r -~ I) IN 0C- a, aM 0 N cc a, aM N N 0 C _(Du-c a, a, N cc rN 0c - cc NN c Ct)Cot _- L _- CM Nal 0,OD CO_CY-C)a CN T 0 c C _ ~~~a, c' 0c_c CV - 9O C (D O <P C U) lq O O (D 'd q Ul (D_p C ul r, OR r, OR (31 C CV Cur CO CY) Cl P- CO fl _ C) Ut) _ t _ CY) Co a) Co u) LO 00 _ CO N_ CMJ _ - CI CD0 t) _ 0 - 0 t)1LO>_) CIJ _ _ CO 0l C) CD aO Le) _ t Cl 00 C) O t- _ C14 Cl qt C\i 03 C\i 1- qt CMN C0 O O 0 C4l CV) _ 0 ) a)t 11 cl) OD 001 0 0u t N U) Ch sM rlCq_ s CM It "t _ CM _ Cp(0 0 lCO ~'! OROp'o'R'R _D00 '_0 U) U) CN C-. C\i _ U) _ T_ N-_ r- IstC\ _ lst --NN CcY c- Nc,-N--c'' CN ... N'CM ' s1- C ) cco C)' C)' C C')~~~~~~~~~~C r'aO )Co 0 a)cc 0 0c o o c) tN oN -t_ O __ 0 -C\i r, 6 (D r- r- 6 6 CO O 6 h U) 0 06 06 ( C\i 6 P- CO a) oOcU)LC C') ' C' _-C ccCC0DNO CM- C') CM 0 't - C')N U) C')NCm N C\l - 'ccD C\ C')A O WI Cc 0'O0R,O( N( 0O rNoc0C(DO Ccc_(c a ,a0o o CV) a) _ r- 00 r- _t LOc _(CO CO CV) COi O) CMiCY r C\J - U) CO \ O C a,o'-c ccccN- OD CY CM-Cl) LOoo |NCM co o 0 r- Cf c)_)LC CY-C LO C -,-c-__- cc CO '\(P l a)0Nr 0~iUt cqi PI: ( Dcl OR WtV)-: P) C)OR 0.-: )cl C_ a) C a)C-DC) coLC CO CON cs 0 Ir o0 _) _~t-0CVcO C 00CY) co CM - ~~~- (D) _ CO LO U) CM COc7s cq _~ P,i csi w cxi0P cii ur (n _- _: _Q 0~ cs c cqi _,~(~(O~~ ~(3 co_)CI) r M C e0 aMMM _ C\i C\i _s _ _ CEi CDi COl LC\ _ N cl (P lp C' '1 r0j|||tr, oq cl C' Li cp -7 W§ l C'c 'CMc' 'I N* CM C'CM 'OC - - '- -'C C 'CMN- CO'- CM NNCM N- CMNNM-NC NM - N\ - \N C\i _L_eiCM c00 * r-~O 6 6 06 6 O ) r_ r_~0 c6 r_ I c 6 r- _ 6 r\i-0 0 c6 6O r-~O o6r- 4 r._ J c6 6 r,0 N 0 oC c) LO u t aNNC U) O N a '_1< r_ - CJ_ _CM) C')(V)CM t 1. CcO lCC') CM ct c0a, C') CO "It CUJ C\ILO C\J CV) ND _tC- t - - a0 0~~~~~~~~~~ xE 0Co@g 3 a O 8 E X g e, 2 t @ g U)y a_ CM co t U) (0 s0 o _ _ _ _ _ . -CM . _ CM CO w ) (D._so ,cnl.C0 c ~~~~~~( c G) )C_ < < 6 < < 6< 6 < < < <: < <: m O O O D CxD Cm C u) co uJ I I _ jE co OMS o U cC) U)E(D 0 cca CL cj Co 0 10 (D0 CO)0~ co,EC 0 .2 00Cu0co.QE aUU) ) )' C,uCucD a) 0C ~~~~C~~~~~) ~ ~ ~ ~ ~ ~ c%_ (nL -? .0 0) UC'i-_(ao E'0 --'WCu0 C -0 o a, Cu ~~~~~~( n E 0 Cu -0 C ) : ~~0'NC~~)~ct0Z 0.Z . 0 .. 0 *~~~~~~~~~~~~~~~~oC l .0 N '-C' cc . cc') '-N '- WHO Bulletin OMS. Vol 72 1994 cl ct 2 Global and regional cause-of-death patterns CD- CDi N - N 00 ItN0 * N C0)DCDDLA 0) N A-L tNC LO 0)Cco C)o v -- '- 'CO '-NN7- N-CDi CD LA)0) 'C OCA ' NCY CY ) ~- C') -CY~~~~~~~~~~~~~~~~~~~~~~~' N' N ' V CO Nl V N D ' N LCD N- N O-t - N-C')'-_ OO0 OC)O)--0 s00_c oO0_(-hC \ O CC tOC0 0)0)LA( c')Nr 1z N'-0 C'Id NN N 0 X "'' N0 0 _ CY? aCIO '_ r1-0_ N'-' N0- a Y - v C V V l D0- r- CMCDC 0' 0 OLO CDC ) 0 (D VN N _- NN,-N NN V CL -O3 0) t 0 CD_ 04 LC CY'' 0 19) CD CIl V_ 4N I V) V V)0 -_ I V CD V) eD CY -e tt@t ,t Id CR NO CY NRr N - C N Oq C q_~N p V: lt V 4N O_NR OD CYtV0 10 0 ) 0 CY0l0 -0 0-'3 CM Cl CY M CY< VU)0 c 1 -X OD 0D CY CY t- r- CYNNXN(1 L aid '-0)'-'-'-000~~~~~~~~606 N6C c-i 0) CO OCO~c NC\! '-'- N C' C'I)0 CD- ' Ca ) V: r- - Nc- N'- OD CM C4)LACDN C') - - - C')N'CM 0 '-- COl '-' 'IT coCO) O))A ' t0C)0)' C'f) C') N LA'N-)D DcO' 0)DL' 'C O O N-N''''' U C')'- N- N-CD 'N N- 00)CY) N0) 0N- N 'N RT- 0'0 C')'-C 0N- '-N C L N-\- COL C')'L- r U) cmcNCMCDC')NC')CN-CL LACD M n NC~--N C') )C' 0')- (OC OCDN -C')C '- NCON- C')-'CO- D NLAL N D LACON'-O OC C)NC'0)N00) -O ' CD0)'-~ N\- '-N Cf) N0 Ch~~~~~~~~~~~~~~~~~~~~C LOCM 0C u.. ~ ~ ~ ~ ~ -C 0C,uC -C 0 *)00't l *OM )a)qq CD )\i Cu\ co S,0 ~ ~ u uJCiC E .CC o~~~~~~~~~~ 0~~~~~~~~~~ U) oCot ) )U CL -0 ILC)C- .C C~~ a m 0C :.Cu0 co C 0 0 >'f lco o~~~~~~~~~~~~~~~~~~~~~~c rul' -~ ICI(camIQ< 0>. .0 r-_ C C F_ CO - - .~~~~~~~~~~~~~~~~~~~~~~ .\ .j .a:.. a. CtU) i 0 UD (6 N: 06 _t_ ) CM ) 'tU)'CO''' CD- N mr C) 0)'-CQ cx c w r z0 t uc LA N LA N- _ N C' CO LA'_ N N 0 LA LA 0 CO 'COCO N- )N N-C D0) C0()O MC)LA OD C') CM CD) 'CD ',- ~ CM'\ Nl OR _q Oe a?U) lq- _i Cl 0q LA N- _-CDNN CO C' CD CN 0 OD C') N U)- N-0NDN" - 0)C''.-.C. '-.C'. aNC' C' No0c o wc CO C ' _N COC LO t- 00 CM<D CD 0)OD tC)Cv C')NN0CO'-NN- C'0 CD (DN- CM _0 L) _ C\ CI 0) Cl CM CW r t-CM -CM CY) o ) CM _) CDX x s_ a) cs Lc) _C _4r- a CM co _: N N Lq0 :O N -RC CO CMi C\i CD _CMi r- _ CD cl) o) CM __ _CM CM _ CD 0 N- C) N- N-0'- D CO C CD0 c) t N- NC 0 - 0)00 C\' Ct CM CD0C 00) C C LA CO N- '0- CD N- CO-COCCM -CtCO' "DcsUM0 Oa)utq o co ur ui -CO U) 0'- L' ) t U) C)CY 0 C\ - C') CM lD P- 0q a (D Ul Cl? lD _ (D 0 N CO) CD0-u) C C'oC r- ODCO0D_o _ _CD 0N o_L CM - 0) C')T- LA LAO NNCDCD ') CO0o ) c NC uz It0 0)LOL U) CM_- CM 0q _ O N C4 O Cl? r, P,- U) 0q '1CD '-0N N0w C' N- N- LA CDcD LAN-N- 0)CODN- C)'- '- C') COC-CD __ )CMA C\ 0 a) a) Noceq r x cf m ) (0 C')ur0 U)Lo 'N_ (D U) CM )A cO COCCLAC\L q cD rl. (D Ci Cy) lp C lpcm C? 0) cD N C OC') r0OD N O C e _ C\ _ ) Cl- _ CM CM CM _: P, C0 0 C4 U) IQ Cli '1 cD t. 0 CON N DCNLA L C') 0CD'-vCDCN C') C') C') CO C0 * CD CN C') Nf VC'N - Cl--eD Of CO CYC ~~~0 a) -o ' cXX- Cu- r -L c coXEEcc 0C a: cL) (D-*c- C\ Eo0c:ur mm WHO Bulletin OMS. Vol 72 1994 -S a L) CM q OD N 0 U)) aV) CD CD ef en r- (O a) coCO) Mco "t C) ar) CM "0nc U) 0 co -c 0 V z Ce 0 0) Cu LO + 0 L- ') CD Co ( a') 0) C) 0) Cu C') co NCM N N- CM0 N CMN o CD 0 OD CMi CM - CO 0 0I) CD LA) N-c CD C') CD aC) OD .2 - 0 Cu0 0 = a.< U) c 0 .i-0 0) a;- U) n U) U C0 0) U) *0 v- 6 C *) la 0 U) U) la) U) U) LU a ~c c)00 a) 0 03cus acl co 0) '7 0 0 0 - c - Cu 0 -o Cu XCu I<'A 461 a C.J.L. Murray & A.D. Lopez 0)0 -, NA6 (0rlN 6444 4666Ci46CiC l l 6r- ~66c lc Y l ) O F MCl Lo 00-c, U) -ItC 0 -C)L 4 1 1 -rC\! )O --A\0LAl(0 \0LA'- W0(00) - - LOALA(0(C < NL --'''''-N'' N LANi -(_LAO- L - ' + p"c qr- ~ l :qlq 9qN'd :qqN l R"I Pc PU I:e qa it0 0000 D ( 'N-t )c q -rl Oc M C 7 r DC t( D000 l e 00CN LO'- l (0C')'U'- '-'-(00LO --( N\ qW-NO 'NMrN\l LA) rNO CMcLOLA 19i C\i 1-:~~~~~~~ Q) ')O'-0)N-COLAOOqq N OqP,~qLIQA(0 '(00)OCaqoq qJNlpN NOOLAq6C1-Co iClC 0 q MCY iC ojoi-L 6tC M W \ Mr- tC CjCl C MjCY)66 - - CD CY)C)cr)lq L Cl V V)ClMCIJ CMCD0(0 LA)( LO ')LACf)C l6 D ')'MC) M C') - - - E Cu - -"tWCr(O -C MCIJC'0 CDLOL0) NN-l C -NC 'LOA CO 0 OC0C O LAT 0)-M N'I a'I o , lM MC CM)(CMNNN N C')N- (0 NN'C- Cf) LAO C Y DMU)C Cl Clr WC N-'Wa- '-N- N N'-aM0 CM (0( NC.. (DN(0N-) C')( (0 COC)'- C r -C U')CN-'- 0 0 cONL ~ L i0 v64N(0-C' LAC\iN 0 C')66(0 C M \ CM N\\ CM 0 Cu Nco NNC')M COC') '- L- C')V'C)L NC C0 -()N o N- 0(0~ ~~ ~ ~~~~ 'n -C n11 '- N (0CDN 0LCA -c N * N~~~ C o N CuLCL W --'tC OC OM N-D- qtLN N-- N-C O L \ Cu \ 6 u)r co ' --N- '- -n cn '- N'-(0 1-(0N' CD N-CY)N'-t CDL0 OL '- N C) ' 0- '\JN\'- N\ U) '- o'-'- WOOW NW (~ -r- 6 1c'- (0 L (0 -N'-(0M'-C')LAO'N -NC) NN 6 r- 'q -'-T It N- '- W1C' 'C' '-NC NNC I O6 0)' LO(0'-C (0 ')CM cm - - N- 4T N~0 cm LO c)N.C. A. AL.A. .C')CO -n (CO O- C')L CO LOAICON-CONNL 'O N N-(') NC' .0 LA (O-C Or o - r cic')OL O D% -WWM --N( DW - -MaN -MC U) *0~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C Cu Cu~~~~~~~~~~~~~a00 ~~~~~~~~~~~~~~~~C uCu C 0 C u) VCu(I)C Cu~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~aCuucuo- Cuu E cCu ~~~~~~~~~~~~~~~~~~~~~~~~~~-Cu ~~~~~~~~~~~cuCu coV> u C , xuQ EE coo C 4) 0 Cu~~~~~~ .~~~ VCuD- cotn c uM 4)Cu~~~~~~~~~~~~~~~~~~~*0 O.~~~~~~~~~~~~~~)(~~~:Cu C E ) a ED Cu M 0 E :C"Cu3~C CCD>uC . Cu uC:EuECUCO-CuCu Cu ~~ ~~~~~~~~~~~~~~~ ~~~0CuECEE> '0CL E '~~~ ~ C~~>2~~UE~~0 06ocoCu- -~~...54c~ ~~~~j m e O Z . .Z.0.1 ICE. *0 *.(D.~~~~~~~~~~~~~~~~~~~~~~~0 . cU) Cu.'-NC')'ULA(0N-(0C)'-'-'-'-'- '-N .'-NC')'~~~~~~~~~~~~~~~~~~~~iLA(0 .'-NC')~~~~~~ '-N'E-NUw 0 ~~~~~~~~~«« < ««««<~~~~~~~~~~~nCDOO OO OO OW>(! OOI-- WHO Bulletin OMS. Vol 72 1994462 Global and regional cause-of-death patterns c o0) c Ic '.c(o'C ''O 'I I ' ' ' ' ' ' L) N'- (6iO 16_-_ CN C0 c 0a £CM _ c') 0 4 10 (0 N It 10 N o C') co url cx LO 17 a) Cl) CD CM OD (0 0)(0cyi co 10 CM ur C') CM Ct - co C') r) CY) (0 (o 10 LC) 0) cq CD) 10 LO N 0 0) r-o N (0 r-o (0 (0 co (0(D co C) C') C') (0 N~ CM Ni CY) N 0 Le )CQ -CMIt (D Cl) 10 N LO CO CM- CM CN 0) clf o c csN CM') C-M CM aD r C\j ON< CM Cl? CM (D CM CM rN N (D ) N 0 N ND 00 - c N 1-' C') L0 c6 10 LA (l ( 0~ rs ( I. _ - N C\l C) aiCM 0) Nl CM I. .I c'J ... ... C') (D(0(0 D .1 . I... l CM CM NM MN La(0 (0 C') N C' 0) CY) (0 M M U) coo_ D OCLO Ct) C'C') CO0(0'-A'- (0N ci .q .i LO CY) lq coiLOL oD 0) _oUOD CY) L C (0NDc 0CC'0 00N L C! ... "it 0' 0 o' (0_ o 0l C t0 o O cN11 o0 X C) 0) 0) ) C) o 0 CM Cm_ I_OM; U) ' _ _ _ C') N\o _D ol N LO _l C a(Y 1,0 O C ')'- '-- - ~~~. .~CC.P- .:C \ A CDU) C\ CM_ _ N -m C" w 9 q ci ll o') c) N0 _') C') ND ) 0) I C4 )0N'-C\ N'-NN0)6C C C 0L LA'- I t C')- cu ~~~~~~~~~~~~~~CU C U~ a c Co 0) -cn0 0 "tsC (C C -S?CL CUE cU C0C U .C CU Ct J .Y C DCUCU Co~ E' *CU 0 UCD-cnC C U Vo Co Un CCf CU C: 3 r c0n --InC Fz. C Ua)C,)0<C)C DCCL Di 0 & 0 .'-NC')~~ co. .0- N C )5LA(CO0.,-NC').CC~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. Q~~~~~~~~~~~~~ .E:~~~~~~~~~~~~~~~~~~~~ .-~ac.uCD -roCDo ooo : 000( WHO Bulletin OMS. Vol 72 1994 -s a n CUc co 0 ._C CU co 0 z CD 0r)LO CD + zC ON 0) 6(0 0) cm C') 0) tn co CO aw 't U6 x10 Co 0) CU + 0 N 0) (06 C') -CU: ~CU U) E 0 c 0 U w LU ._ - 1t0) 0)9 0)* ' U) x a) 0) 0 U) 0 m0 la 0 *) 6 U) n m cola c E ._ n 0 a) 0a) cn co QO ') 0 00 V-C CIO CU CU 0 -V r- (ACIO 'a 463 C.J.L. Murray & A.D. Lopez 0)0)N00OCD 0 LN 0 0O clc'i lCq7l a) N C') 00O+ CD0-I -D cO'D - c o o (9 6o -e'joNo) 6 U) - C-J C\ CD o C'JN-ao ' CDO N 3 oItt .,. In InL tc mc E C')0) C')CD -~cl CD. -'.'' 6 co 10_ 0t CON-_ J 0 O O 0 _ 10'-D 0 00- N _ eCM 00)0 O st 00 C' 10 ' C) 0) o0 cm C' U COI COY _ _i QD C\l CO CDC C1M')0000 C') ' 1 00 0 N I C') 0000 0 ) 0)1 01z C') 0) C) CD N1 d N '-0 0) N100I) CO '- Cl 100 C') - 0) N CD 0 N C) 00 r-N X0C' ) N tC' 0 CN CN C') N -_C 0)10U - C') 10 _ coC'J 0 XOC_)0 10_- OD-COCO 00-t _ C CO t Cl- v- cqD NT - - 1rN 10000) N- N 106 cM 4 <c r, c-i c6 o (670CM - (D CM ODC) t-_JcDrwC% 0 CM 00 "itita Ncs q Nr''c "t m) L10N C "IC'rI: C00 C .4 7 C'* ''0 ''- __ C'j LQ Nr0 l r, ORP,eUN cp P,O Ns X C' '0)0NC'0))0t)N0- a)C Le') 1 -C CMO 0 CD'-, 1'P 'N_ (P0 U)NU r 'L)0'-1 C D NON 10 N C') _ r0 -t _ t0 )1 cI C'O C')M -- 1 IC t - C Clf) ,f o~~~~~~~~~~~~0 ooirM-,oi C') aDC) 100-C') CM CY)cqt 0)N '-0CM CM 0) CD CM 0)C '-' C')- flN01 10 '1 CD '0 CD CDO- 0CMJ '-10C 0) C') 0C) N- 10N CM0 0)' 9 \ CDC LON D M -t C _ 0 MO MO a0 a 0 _-UCO _C'0 r0a 0C0- 0) C')Ct C) C\ 0) -C) - 0) 0)- t LO -CD -NrLC) t N C')c0'- 'R0 P "I*:O U) 0l 0q CM I' U) 'pCD Ci CO 0 q0U ) CO Ci 'Id Ui (D CN 0 0~ 0 cl) c\ 0) ur) 'IT _co ct_lCl (D o) CII) CQi 0) U) _ r- U) Cl 0 a) Cl r- ur N C\ - -C')- U1 C\- 0 - C') )- P- - - "t C '' 0U lD01 '0 'I0 0) 'O CDq( _C') U)_0 00C'_ O D' L -b Co co ci ul) Cl c14 "It _ \Gi _ 0X 0) 0r 0) c\i P- c\i c 00 CO _0) N w C') 04 C') 0\ ' 000 0 C\ C' O C') ''n0 ' N C0 C0C)N 0)ttC 0 0 ' Cf C\ 1 -tN-C) C\i inC\ -W _ _ Nt '-CD-N_ _(CDC') C 10 M0) '- C') 01 )CVCM CM C\J C~C - '.CD It) CM r-I I- (D CMCO (o IICM CM dbt - I_ C\M U)i uS (6Dvc) C\ C a)- CD - - (D - - - _ . ~ ~ ~ ~. 't U-) CO.~~~~~~~~~~~~~~~~~~~cM~~~~~~~~~~~~~~~~~0 N 00 )C N L d C ) _ (D O 0 'I 01 N0 D O DW C')WN _- 0 0 0C' 01Nc C')C'- 1 'r ION C') C\iC 019u o C\dCr) r-' u-ia1\"toM (O co qt wc0 _0 P- a) - 0) ur LC ( -O _ a) c\i r c\i co a) 4 4 c\j o m c\i r- cII co eC\ D'- C MC C' ')- a) CJ *C In O0 ~~~~co .0 ')E - C°- x ~~~~~~~~~~~Ec c a)-) .) oNIn<In'~'*~<: In:<:6< <<<:< mOOO ] aDDuJlJ sLLu I L In _ j) E co mIn a) -W> O S In4C Ea 0) a)a) 0 c 0DnI )p )aco m In 0 n C - I n -Q 0 C D0 IIa) a) 0):3 U) 0)~~~~~W 7M5 C .' a).r co 0 z - Cn 0 * CU co Cla 0 -m co Cl) U) n 0) 0) CU C.) cn a) (a a1) 0) In- CD 0O CD ') C') 10 .- U - ) 0 ma) F 0 0 a) 0 co co CU Cla 0 CU 01 *0 (0 464 Global and regional cause-of-death patterns 0) cD cO t 0) o a cN CY C) a)C -_L- C- ' 0 0) I' ' 'N ' ' ' ' ' CAN ' cli o' OD 0) CD q v o~~~~~~~~~ C') NM C') N co N cMoMc N 6lf)00 N v- N CDN'-) CM) ci I CM NP CD N q. OD NC CDC C,,) - C LOCM''M '0) '' '' C D ''' \N '' C'' C') N co CD N ~ A ~ N'C C.) CO CO~~~~LAC' CO coT))U 0. Cu 0v o)O t 0 CO ) co co C - 0)n- -Cu Ca., E,C Cu Cu, ,C Cu,X co 0 (D a) - co cUU)D)CCuo -N 0 m 'D 0 l @B-Co co cs~~~~~~~~~~~~ 0 ct C *S 75u) (D CuCCu-CO Cn 0E I..0Cu~220si - 0 C Co E 0 - . LE ) N ~ ~ ~ -rn C 0mC 3 ' a, oOC 0 C ~~~~~~~~~~~~~~~~~~~~~~~~~r cL co n - =C CL (C C1 CCC (J(DC L 4 C Q C~L fl0 C a. 2 2. 0) - 43 a - \ c-2 .a7aD r i( , 6ci. . . mC)" oc 0 o a) o C\i c\i t _ urlrv NN CD - - - CM - r - 0)UL CD '' ( ) 2AC) 0 L t q _ (P 7 N CM 00 CM __ CV N N (D C' Ci (D r- _- Cl C" (D lt _- C')0-CD 0 '- DC 0 C(D _ U')C) CM cD LA C'I) C'I) 0')Ns'DN '0 '- 00') lqN CcN 0_ oCD CO NC')_ CO LO N C\j- ~ N' _) CD) C') q-C )N N ) C DCLN N' 0- 0- N N '0) CD CD W N 0 00CD ) CiJ Cl C)t_j it C! CC!t C! C! C\! r, l rs _ (D (D coa Cn _) LO s C) 0 L0 N NN N C\)D Nao C\i_ cp l)o:c 00CDLACD'00 N-ONNC LC) Nt CM a,)c a- )0 -CY r-1-cIJ0 t -U 0 00(D ) C aCuO a -0-0owo=Cu .(D-0 r F 6 *. O0Cu @ C - NCM C5 LO D CMNCm< < < < C WHO Bulletin OMS. Vol 72 1994 ax CD Cn 0 0) co 0 CD 0) L) 0 CD) a)LO C) 0 (aU) E 0 U. 'O c co C 0 c cn U) z (D C') 0 0 Ca) 0 0 C') co C') CD N lq g CD aI) CY CQUI) N UV) N N C') N C') CDt "t 03) C') 0) N 0) 0 0r cli co N LO U) U) CMI Ci 0) 0~ a 01)a) 0) CD CD 0F) LA a, LA I') CV) 0) Cu (A0 E 0 c0 0-co U' 0 0 0L._ m 0 I- E 0 U. 0 x0 0 4;' CA 0 0 0 00 0 0 0 0 0 0 LU la E ._ 0 0) 0 c0 C Cu E G .20C -Cu0 0 = a. (c X co a) 'e 000 . - Cl U) co C.) r- '0 CnCuCu '0 465 C.J.L. Murray & A.D. Lopez CMco ) N 0) CD M N0 LA 0) 0) C') C') c') 0D CJ w c) 0 LA w N c'J 0) (0 t) C') N CD LI) t 0 C'J 0 0 C\J'- C\! CY) CM C) P- uz-)t OC0ocoC') C\CN 'LA N toC 0co CM -o C) t CY) O (D co 00 Ie C\j CV) N ~ 0c N C'0s) CD0oCcD C NC C) c 'C\I CO rC0D 'D v ccic\ a; Ci 'OCM ' 'N CN C\j ''-LANLA' 'CD C\I LAC "<DLOC0 t DO- LO (D vtoC CO CM CD "It Cl? C\!O ON-C'I)O CM I'C' a) LO (L CI CM 0 LO CM ICMCIC C\ C' O "It l LO O lC\ r,- CO 't LO r- (O 't 00 r- u n q CM 04 CM COD CD'- lx 9 C' ' .l C 00 o Ct) CM LO CO@ @| CCi 0) " 0) CM C') N CD N CD LA Cci CtOR u CC C) 0 CO 0 n t 0)c-(D CD O r- - 'I t -s- - C\ CD C') C\) LO '0 .C C\i 0) C CM OD 0 CCI D CDO0CD CD -CCLA C ) 0LAC t CCI t CD 0D LO DC U)C OtU C M000C CCI CD ur i ~06 Lc t 06 - cli Lci \ \ r-- cl co L CM I- C) C OM 0) LC) C CM 0M O C) (D W WLC0DU)O' r- 6CO U)C6 0i CO\i Uc) CC N c O C M T - r- CO O' U) 06 6 6 6 LC cI cui ci CD C\J CM . . ~ ~. 'T CCO ( O'- C')C '- CCI I C) CY) fC'O Cr-~,,~c 'INLALA"I:CDCDC ORI"IC-NNCDCD CY LC) o 0cst M uruN oM s 0 'ItO O0 - CM N 0 -(D CM a) tCM r- ma)u r- o 0 0uz ot o o C\") 6 6 NcC, )C) C\i co Cl CO LO C") CM - 06 C\i o cm 06 - a ) C cirucui \ C\ CD CM C\I CCI (p wt ,t 0cq N o3q m a cm L cna)C\i CM ta) (.0d r- - c* r 4 *\ ID 0)C) C CM C' .C'?O O l .' 0) CM"(IC'DCC C'0- CM C\- CO) CM ')' LA C )' .' CMl C\! 0 C')0 CD C) LA C' 0)0) - 0) C\! 0o CC CD CC)Lo (o CmCC-AOu) o u)ICMCA0 )Lc u) a) Cco(D LA)CM ' 'It CO C\'C\-'-T CM CM '\DCCI U) U) 0~~~~~~CC,) 0~~~~~C CU C U~~~~C) )* a) Eo C/ C ).i~~~0 a) U O-0 D ( D Ua) >nCDc. L0~CD c0.7I cE CZ. *'c '~ a ) .(D 'c >, ol7C19 U) CZ E a)IL (1n(1 CZ 2 'ac U) 0 O cco z U) a) cts a)(D OCD "t 't 6CO CMc6 cu O rn (O 0) L9 OC 't LO 't cO6 a) U) () 0x 0 C U) ma) :it co C)0b n 0 E 0 0 ._ w cn m 00. E U- 0) aE 0 x 0 0 0 ._ m * 0 0 n la0 *0 0 0 EC U)fw 46 Global and regional cause-of-death patterns - O Ctc)- ' O O-( CO )CM - CDMON qtOcc)) U)LOCDc) 0 tI MV)( n0 Cu - 0) aOC W)CMCO'I-cc ~ CM~~ ~ N CM) ' NCM) Lt) -CM C - It ) - t a0a,- ,N' c-N CMNN < C4J-I. N- c''- C (9 COO)N\NN\i- CiCcr-l- o '- CD t)' CD It) *DLO r CO r- c6 CMc~-c6~ '-C N - N- E o 't cDco0 O- cmr-. a)I)L ntc co' a) ) a) cmNN-c o. ) ot)0 o c-oCDcmCDC cc vicc c6c 6 c6 N~ N\NC C\') iC 6 NI:-ifl C06 6c6C c' 6c U)n ' 0 LOt)N '-N-NCM ' N-Lt) cmc o a)) 0 cc) -t rN- )n0)cm -U)N -LocmLOCM)CO ' 'N- O-N ' i) oi4'.I'~--6 : 66- ic')iOi 6-'C "C' 0 6 a) 0) cmN C'N ncmc LOIt) C N~~~~C Cu~~ N- 0 OICO NCDOO )C 0) 'CO - CON N-- '''''' 0 aI) Oc) Lot) .. 0 U LO LO~-'CY)CDU)C')COMN ' - '- co N) 't 'c' :E (q~C C')CID C') --C') N1 E -c U) ,i 0 CM it - - '0) "tc) CD 11~t -It U)c) COD) 0) CO 0U.) N-'0 lCOCO mtN-'CD M g Lccc --4 \ oNc; i6 cj6i- P. c6c c') cj v 4c6 c6 i oo6c 4 Cu C') N- UnL)It)0- 0 -NCMCD C')C') N-C')- N\ lqIt) LO )c)O'-NCM '-0 'cC CO 0' \N CM Lt) C'f ) CM) z + 0 )N It6' r' ) ... co cm ncc N- ic6 -acNU) N- NN 0) Ln ~ a) n I -0 )CON' '' ' ''R''' L ) '' ' ' CON t'') N N CDN- -- cU) 0 ) '-NCDq 0'co ' 'N CO' 'LO ' C O ''''' 0I)N-T )cO'D- qt- 'LO- '- C> r q-0C D-) OC)'' U) ' 'NU'lU'CM -NL D- CIO) CO) -C') "~ - :( ci cui ci\ir-.c coco--Ncj c6' 6 6c4 ic\ii c6ioc6 6 iuc- co 'n C')n'-'- M0)CDCMC x 0 '- '~-CDCj M0 -cc'ni6c o oc',c'j to 6) ~- r--c nj,- '-Ctc' 1) 0 U') C')0 - N\N '-\ C') CD LO o)0 r LU)OcC') c0N-CY) - 1C D 0) D- O C\'cv ) coUn - m Lo)U)C o Lt)Lt) N-T -U)Cn)-cD~ It ) c')C' cc c')ccD C')-- - )CIO') - 0) c) CO -LO U')C OCO - CDCo - -N c')CD CD ciD )CMC N LO LO (-DU) 6ciciN -I t) CO iCD Cu 0 I N- LO coOU)N \IIqIN ~CM oN- co oCM '-C) c)NC)N'-mc '-N W 20'-'- -U nco oN-( Coo C')C)0' toN-'-C~TC \ 0' to(D )- 0 co C) C 0.~~~~~~C ,~~~~~~~ CU)CDu - - (C)U 0 co 0FA CD coCu C 0 CuaEi co -- cn~~~~C~CO) a) 0 .- cn 0) -T *L Cu Ec C u : . c 0C * 0 ;~~~~- .0 Cu o C C w uc E o( oC r-C~~~~~~0~ C).00~~~~~~c .0(Cu'- co 'a 0 -(D~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ t C0*a CoI I."- E o fi < 0 ocL - ~ ~~0.o o~~~~.... ..0,-N~~~~~CIO r-cL rco Cu C :0= 'N ' tc N cc)'-'-'-'-E'- E'- C) ()D .- 0.-C)'t)c 'N 'W o a~~~~~' E -«<< < « coooo o «« ccccc WHulei-OMS. Vol 72 1994 467C(n0 Cz m o C.J.L. Murray & cD0) Cu co 0)c6 0)C) C- n U) *0 x e) a) (U cm * (U 0la 0 0 U) 0 *0 U) 0 uJ 468 V co 0 .S C 'b 6 z co E U- 0 co (D6 coCD a) U) U) 0LO C') 0) cMCD 6 ur Cu) 0) co 0 0)CD 0 co CY) Cu) U) C') 0) C) CY) ~Cu C) 0 U) 0 A.D. Lopez _7 lp _ r,, '1 0 cl Nq0,- 9O C) U) CD- 0 CD U) t N 0C') t N CM ') M C _ csi N 0) CO 0 C'D C') N 0C 0 C' C' CO (CD U) N U) O IN 0 U) o CO CD0)NCDCD-'-'-NC') 'CD0 CM coC' CD CM In w cm CD0 <D o - CM CO CR N _ CRCD cv6o 0) N 0 U) 0 N- C' 0 lq CDU)0UtN'- "C N Vw 0) N CD U) (6 U)' NP C') r- N~ 0) 0) N ) C') 0) 0)0a) Ul) cD 11 cD 1- N1 co C' DCDN- C' )- N\ C') NI C') 19 Nj 'V) co~~ ~~~~c N- 0) C')'-'- - U) (O 0) C') N ''- M N'- o C')r _ D i C C') N 0 CM 'O N U) D0 0) CD N 0) CD D U CD)0 )Co(D N ON 0) U) 0 N D U) U) 0o N00 -N(D0)U)~N C')~- XL OCM - - NN D _C DCO0N ''N O ' -'- 0000 tC')C0 -'-0O.O.0 -l-:) ~Ci1 4Ci L- c 6cjo _~a c cc cq _ic ( c~~JC'J~~~~C'J U)'- '-~~~~CM C '>-)0NW) CM CM) ) C U CD CN cm * oC0Da)Oa) DCO ' ' '0)' N'-c_u) NCDrN0)CU_NNC0 '(D ) t N_ ) CM t u)I-C NCM_ 00 C 0 CMN-0))) aU) '' LO))__ N - C') _ o 'It Ur) - cY) U0 u) D0)a ) ClC_co to It CD CD o CD'N( C) '- M ' -OCD C')N -CD(6 _ o *, oD'UoCDC')0'N'C')' N cDu)CMM o')C')N - '0 CMi t_It _CM It Cr OD LOU CY U) CMi CD CV) CMi - CW C') cDO )C) C'C C' N 'D- a) C')'-C) CM '- CY)N ) 0 ) c o,~ N 9 cn~ ciN 0-N N_ cs -D _ t 00 - r- o CD t_ &a scCf) CO l CD- It 0PI_Q( C?9 N co CO_lqP~ClV O ) (.DCY) OC'- rs |cI cv 0ur ce c N_ C)t ,t Cli _ CM _CM t - U) 0 C U) C' Uo) U C' tN CD U) ) U) "_ CD'-o N 0)C N CD U) C ( 6 0 aC C' -_ _ _(6D 00- '-_ O_N C_tt N WU)C _- C0N0)N- 0) N 'I CM CM - CN _ MU)C NC _UU) Ct M _ (D DNO CD N NU)O0)_- _)-_ (D 0 d C') - N'- U) CN - N - - '-t MrCNO - N _ CM _ t _ q (O r- C M C\l _C') 'C)NX ) t LS (D 'Id P- _~ r' P, _N (D 0) Ch C\J - O~q 0 OD' CMJ C14 LO (O 0) CMi Cl) C\i CMi _ C\ O -19t dR ) cs cv a) ao s) _ 0 - - CM - ODNCO - _ U) C')' CM 8@@ X b t N o o t LO N 0 LOur ur _ c _t (O W 0 r- O rl I-t __C\i Lei CM_ CO_r c6 CO 4 ao ur zo 6_ o c6 cm _U) - CM C) - ' C)U_ )'N'- )aD 'T'O' '' O I'_ CD) 'C ')CD t ' -LO o C' '-C') N'-'-CD ~~~~~ '- U) C')CD c6 C\i U\c\ Y _ )CD(- _- '- (6 _ __ _ _ 'd-N'-I 0 C'- C') C')C- '- '-_rN N C Dq o 7 _t 3 V: LU) 0 CO _ C CM CV)- 4t Cf- CV) CV) _ CM C\ co 0 CD - N 'I D N- C' "t N 000 'I U) N U) U) CD C' 0)- N CD CD " (60 C') N N U) 'I CD C C' CD N, U) (6 M' 6C )NC)C)0)NO00( " - C 0 DC' CMU) N 0) It O 'DNItt '-(60 CDI N U)'OD-C')0 N -NONC')L) '-UC) r- N OD N, N'- '-'- CO') LC6 )N'I- C' 'IT C4 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~cu 0 (D~~~~~~~~~~~~~~~~~~* CuD 0 CZ~ 0Cun 0D Cucn E c r - 4)~~~~~~~~~CD CcolvbCO Cu cu CuCu CL co~~~~~Cu .- C u 0 Cu 0 (DEC Cu1 Cuu.~ C Ua) CucoC.c a CucoCus Cu cu -o C am c C~CuCu7 .)~~CCo) ~~~~~~ E~~~osC.)co 0 O n u uCr _ Cu ~ ~ Cu).,CucoVCu o _ FnCu= PCuoo( x >.cuVQ CuDa) E CLo 7 E CDcoCl)cD EL- 0 u U cn> co a o N co 75 72 ~~~~~~HO ulltinOMS Voa72199 II Global and regional cause-of-death patterns 0 0 t LA) L UL) 0 CD CM LA LA C) 0V) CD N- cq N N Lt LO r- r- N cli ur 0) C) co 0 0 0 0 0 a) co CD CV) CM 9 CV)(D 9 UC) CD LC) 0 CM 0 0 0 0 0 0 0O 0 (D It Ur) - U) - C (DC0 C C - N N-q 0 0)0 co C'uic C') NQ 0)~ CD C') CO')- N'- 0- 0r) N- CD C') N -CD cD D N CD C') NM C-0) N-N CD(D - N-) N- N- LA) co N- LA C') C') N\ CD N- C')C') C'I) N\ 00- 0 C')' CD 0l 0 0 Nl)co CI C')CD C') C') L)AC'I) CDO.tC')LA 06- CD C 4 0)0 t C') LO - c 66 N - N _ N- a) IT c N _ , I . D- 0 )CD I ct a)oi _i C\J C\l 60 N NC LA4v LCD ''''''00)0) ''''' C'l *~~~~~~~C CY) CY) .... ~~N-CDOC- ''''''0'''''' CD C- ~LAC')N-C') LA6)0)C CD LA ' AI, -0)CD C LA 0CDM -C * . .C? C.? a)0N- 'C') Ul) 1- _ b_,c( LA-'.(O _,) -: .~- N - C ) LA 't CN _ N _ r- ~ 0 _ 0 Ch '' ' N' ' '' '' ' '' 0) C'CY) _N- 'LA '0 'O''-L' '' ''- ''''0 OCD (D co _') CY) U) P COL)a0) - LOAICv)')tLO qt0)I- 0) ''t CI' ') 'N- C') OD - ' '0) LA C') CD ~'q- LO 0)LA')- - - NNiC-N- CDc Ci~)'a)- CD C')CD C') NtC co Q)~~~~~CU 0 2 utcofl a U) U)~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~U '0 Cu O~~~ ~ ~~ouU MI) -Cu Cui 5C 0 cuU~ U)CuCu-- C a,0~~~~~~~~~~~~~~~~~~~C .'a) a(U ~~~~~~~ ~c )0-cu 0co- a Cu c - >'-OflC~~~~CL-n ~~ ~~Oo.S .6~~~~~~~~cEcu -SCC3 .DO Cu 0 = .'-NC')~~~~~~co LA D 'C0 Co---'.-Co )CLCLii 0 a (~~~~~0 An««««rOOCoOODOc WHO Bulletin OMS. Vol 72 1994~~~~~~~0 C') CD C' t 0D L aD A N- '- CD N DN _-N0 0co' ,Kr cli CMa Oc qt m CO qtD 't- 0- CDO' '_ C)N CD' LRAN-C')NP'- r N-'- a) c-M Ncqt _ ls 00 - C') CM CM 04 0) CM) D ') ') CD) C D) O O U) C4 0@|O Cl 0) N-) C) r-_N N- CD Cl C') ) CN- C CmNCM)C') lo: o: N'( 0) DC O CM0NO) CD NLO a) CY)o-' N-N-cicsci)0 a 0) CD N- 0)N 0C 0 o N N-. 0 N1 CD N q0) CD C'O - CD CD N\ C) N 0) O OD00) 0) C) C_) '-)- O'C) C')UlA 0 C\ D LO 0) LO O OD -CDC o-N-CM0)CDM N LA0 'L- CD 0) C' N- D U)_ g Kr1t )P CM ODi LO C\CM_O0t D CDLA-: NNt a) o a N- - (D0))N- N NCD CDi LA U) 0 CM ItUD w - CDN (D O M C O C) CM N La)C 0 CD 0) ci L LAO)tN Ns_ e C'? 'DR CD' 'N-' LAuLACr) 0o CD CM_ U)A \0'CY) NC NCU 0 CD C!ls \ )s S pN0-CM1,u N-0M 0) 0 N-Nr- C'I) C') 0C '-00YC) NM NY C) CD CD Nl 't N\ 0) It LA O)(D - CuC, Cl) 0 > C.) cO C_) 0 comi .(°- t5 a)>° VE . _ c-o oO0mm -.NC\' ) qLAC)D '-NC') co 0 (D a CD cn Cr Ln co CD r- ur 19 Cf7 CD cx u? LO I' Un a) E a) U- CU -i) Culc Cu co U) a) 0 - co zo ci.-C Cu00 C a Cu ._- U) U1) a) U) au C U) 469 C.J.L. Murray & A.D. Lopez 00 CDU)NN1-1-W 00 OC')C')LOCM00LOLO C' U)N000 UMCMC)' ) N-(0000CM U) )N 0 CM') '~C') 0NCj 0'-0)UCM - C')'-~ C CO') 000''UN 0)'0000P,~c')00G 'c 7c'0 U)'00 00' C\JLO 00 0) 00q 0) N C') C') 0)CD U00MN N0 "t0'- U) 00 C-M CM C') CD N~ 0) 0)0)' 0'- C co U) ( Nt 1- UY) CY CD 00 '-) c C'-') U 0) 0)( CY)00)T-C COD'\ 00 cn CD C) C') CY) C!0(PCN7'-U) U) 0) NC l 00)OD C CMU0C)C 0 U) CN 0)00 C').. CY)N N '- CO) '-CD'- N\ CMj c'I C\j I'I''''0)'CIU)C')00N'-~I CMj lqtCO q- CM - CDtLO) N NDC\) CMDLO C')( '- C') a)' 00 C' CD M- CD a0 CU) U) N0-V- N1- C' N N q "g N CD'- - - - - N- '-C') LO)q U')'- C') C')' C'I) CM *' ' ''-0000NCDC')CD-NCD0) *'' CDCD000- CD'CDV 0) 0C LO CO') ' LOU) Nq '-'-v'- C) '- - CM C't) 'N 'U0 CD N 0-'- O))-00 N)))I) 0 C) N~CD CD '-0 N6 00)6 U) 00 C' -' N\ 00 4 -'000 N N- C')C\ ) CD N6 ''C'N N'- Uf) - N\lCDC -ItCl-CM)'-T LO C t- LO-VI) NCMD 00NN-C '000I)NWMNNTowlU) I'N N\ 6CD NN0)U'- '-'14Ci '0IC) CY)NN00 'C' U) U)'Lo- ''''''CDCI DNCC L6N\i-: '0CDN. () L6 5ei~ Lci 600-N ' 000000)\lrlN00LCD U)CDNY0fC') C')'- UC) 00'- 0 00 U)'- C') CY) '- N'l- 'C'? PN~U)tqCDNN "'t0)NC\! CD- '-N\ CD N- 00'- 0 C')N - N, '-0)0)-C') 0) NCDO L )- CI)'- C\I') NY o l '0)U)qc')1000N 0)'-U)ORNCD'Pc') r L6 DC\DNLC ~-6CaIC'jc6 6 CY) C\NJ-- C D '- - - N-C') N -'- N- N W ,* U)0q M 00U)0N U)N 0 )C')N 0)N 0CD LO N 0a CO ODC') CD000NC') NC N'\J- CMNCD UL)rNMCMN0) '- CD CDI* N-0)OD04NU) 00 OD C) C') C) U)q '-C') NM N LO CM L CY) 1- CO C~~~~~~~~~) CI) (1~)0 0(j c CZ COC~~~~( 0n ) 2 C a)cn~~~~~E CQ)~~~~~~U)~~~~U)VVVn >DE 0) -CO -~~~~~~~~~~~EC. ~~ m CL cn 0 (n - 0 .0)cc a U- *- )W A0 - m CD E CCOC CU) E ) > )ts 0 U)c CO (D cnE o m 0 CES -o COU) 0 C 01) z 0) 0) CD) 6 0) CD LO + CD CD a) L6 0) Ur) C) 0 C) co 0) -C 0l) 0) CZ C.) 0 0) 0) I- w U) Cu x U) U) U 0) co .0 U) la cu 0 U) 470 c c c E c u c c c -9 Global and regional cause-of-death patterns CD 0 C')NC') CDNO'- COU N''r '' NC')NLO t- N cts V) U) 0 umI _ 0 ac(D C) - LO CJ CI 0 I CD CIJVi C't ) - N- C') + .u 9 Lp LP O °N 0 Cr- 0 co O CDCDN---~ C N CDN- cDCD a) C O COD qtCD .CD .C 4 4 oi c6 c-i In|gtC D'''''-O O '-0 -- C') C0C' ) 00 C C.4 _ U) L- L6 c6 Lc u6 C3i C\JC\i 0l CC OC6\ LCO C') N 't _C CMN en-C4 E O D C-D CD NCY) CD - 0 M N N 0 0 6o 2 co) co_ mc 0 +o) 0Z0. NN CDC '' ' ' 0 ' ' CoO0)CDC'_)C')C')Ne% s~CC CD C/CO _ C N- Co 0. CoC0 C'J . C O NO c ui'NO(6 ' -C' '0)'- O'-C LOa)ur £ ) OD r'I Cn .C1 't M O CM CD0 n O P t Lo .D or-~- o CO r-'CiCiMN o o . .D N O N - .-.. Ci N. . CU ° U U)U U) ( 00o CD') co C\ 'N c C N ''''M'N o )u6 N oN'- O~CoNCC0 N X- Nc cC\ N 0) a C1) 06t It <- cn rCO CD C\j -S It CM O)t ''U0 _ w a~~~~~Nw C\--( wir " wcsts co 0 cj xuic6 - - o' u- c6 o _r- r-- c6 C cm 1t 0I 'I cn aM a ) CM N - CD _I C\ L _ X) )o 0 0 N CD M CM D '''3N.''3' N co V < 0 N n N C.) 0 r- r 1tC NnC\N- C') NN'CM CO CD 0) 0 CODC' CO WH 6e-OMS. r-Vol 721 o 0) '-0 N- It ' 'CDOC O.....c U) c6 c 6cC \DO N C') CD U) CO C' 0) oDcoDN C\JoDCO co') LON I- CM)CNO CD 0) ') 'CO ) .CJ .CC ~ aiu-)C 4-O 4f' N- LuO 0) 0 Co) C') U) '- 0) -tN-CO a)CLO 'N 0)Ncm- ItcD CM - N0r-co NC) cm)' 0) NC') '- 0)~ Dc )c n ...c N NL -NN' U) cn) c )(C0-C') C\)C\-C\NOC' C\CDCDC')x~ NOD )OC\!CLI) C')C'CD ')NCDC')N 0) CDNCC'I)0 O LCDC C) -C') qN- ' CD c CY) U) cts~~~~~~~~U U) U)~~~cnCD( 00)-c CD ) CO- U) .~~~~~~~~~~~~~~C ).-U) 0S-U ccCu 0 C C~~~~~~~aUC ) CU U)U) U) Coos0 c0 0 .9 ~ ~ ~~~~~~~~CDa ' aU) U)'COU)C. = 5 ) a UC-C C Ur U)C co EA E E a)co0 <U)5 a. CU)OF- 0 U)1)UCU,C0 r :3 ,-, . .O NCM M CtmU) CU O'~~~ ~.NC')l4 Lic6r: jaCO D -C )- - - - - .-NCM)"COi4UiCD' uJ 0. <<««< <«<«<«< <O<O<O<D WHO Bulletin OMS. Vol 72 1994 N a00) O 0a0(0 C) '- _ 0) O~~~1DcO McxON NONN NC '-Ct 0 CN0-CM CM - Nod- C\J 0O rl (D C\!M _ N CM c - _P 0~ Ci Ci f1. ~( LONCD ' CD IV' 00cq _ C' ) C N_ COCOto-U - r-CM- N- cl 00 LO 'CDLoOC ) C0 0 CMJ CV) CMNM'- ' 0) C') 0 u C' N C'?) '- a D 0) 0 C CDco (D r- It LO CM U)OCO(DCO) C') -O CDC) C'I) NM U NO ' 0' ' 0 C ') C) _C) N CY'CY) CC)Ot 'D' t C') - t o ) It 0 ) qNc- C\!qocM's si N 00 cs M nC) 1 - M C'O N -O_) 0'- O a) a) (o (P cm aw cl _ cx o 0 CD MCDOC%J )-0 qC C') -C- 0O0Co N t CMN 0oD 0) 0 CM - N 0 '0) C CD lqr O -- - -0)09 N ICO N C') CO 0) C' N ND1t CML C)N C\) 0 )D 0 CO CO CDN t NCN N -C CD T Cl)~~~~~~~~C C C 0~~~~co c co V0 C a)C CIf CYCU5a .NC')Q C XC D- C') O««C DCCX D 0D0D CO°> ta (DC *._N X) V I) CD. CY -<< < < <:< < mMmmCm (A -c U) 0 C) -c cn U) U) U) co C) -C a) co - 471 C.J.L. Murray & 0) CD 0 N- 0: 0 I-l- 0 0 0 c x 0 U'i .r 0m la 00 a)0 0 0 QO n LU 4o 0 la -W E LU 472 en U) 0 U1) 1-- 0) 0 6 z 0CD E 0 U- 0) C0 6 CO 0) U, lt I 0 C'I) 0Cq ltN 0) U) C) + 0) C) N- 0) 6 (0 0) inLO N a) 0 0 0) "cM U) 0li 0 (na)co A.D. Lopez 0 (0 NQ C) w N- C') _-000 CMCM N _ N\ CON 0)C) CO N U,) 0r 0) 0) r 'R -N '-'L-'- WqtL U) n q loq0,a~cx 0 lfc_oq 7cm u 6 00N- N- CD'-N U cmNC C 0C CMl C') CO 0n N 0)N- U, C')0' )N- 0 o oquz r-a Nur r 0C-O d0 C') 0) N LO6 o Nu r c N co cma 'aD-'C')0 'AiCi CD N- U, N Ca N 0 0)00)0 U, U, U,0 C NIt O 1< N- 0' 0 NO )- _- CM C-) 0) t 0(0' C') NMN U, coutC')U, C') 0) U, -P O: CON- 0'- 'I N *N 00O '0am 0N-U 0NNw N-0C'NN m0 ON-Ccq ax 0 co cm 4 6 cm rz -_ cm ov 6 4 c\i 6 6a) cq o r c\ico cuN0 ') N '- N '_- _ c \ r ca)U(D (DN0 D0-C M t C\M CO)ONN04-0CCC) 0 N N C'CM'_ tN CM 0 04 ) - N '- N C 0 c')C) c' 0 It--oCM cl ( C') ( N( cr U, ) 0) N U, C Cl- 't M co O C') 6 Ncq 110 r~C - - 0 r- '---N:c>M oc C\ C. . , C') C') -C') N'-'0 CO N Ut, LO OD CVLC CM) Nl C') N~ Nl- 04 Nl N N- C CO'0 00 C) N- N- N- U, t0 w(00 0 p- C Y t X _ w a) cq(D7! cp q a) 0 o q r,. "I N : :6o o N-04 -t ()£- cf) L 't - W 11- _- Its1 o ocst qt 0) C'U)C 'N r-CMiu) C) cwo - i C- LC) 0 CCM co CMcr)_II cV) aw Cq C! C'O _- w _ CiC_ _ CfM_Tc1r LO lt * *r CD'0(U( 0N00'-C' ' 6 C')(D CO O _C') NO NU), 0) (0ONU,0 U 0' - - (U U, (0 '-U_,(0 '- N N'- NL(0 C')'- '0 C0 '00))0 (N 00 N 0) 0)0M (0C C\i LO)OC\i CltC\_ C\j1 < \Cr) - N _OONo')cs0 N Na)r cq cm N'- C) N0) O C) C) N C) N\CDD0 O _ ) - _ - ' )'U ,'0 '_ -U ,(0 N 0 U),'C ')N ,C N\ ' LC. LO -,N'- C\) N - tC6U, ' 6 ' 'J 0 ' c m CO N' NM CO C\l _ \ c P r '-'N _') LOUt_ c w *@ |t 'NCM N-NO'C' )' '(0 'N0 D ) 0_O_D_CO~~~~~~~~~~f 0 (0NJ 00 C c) 00)0)0 0t t000 w000N 0 N ') (00z) (0 U,N ( (C) _0 U)( 0'- O(0 _ O 0(0 -ur 6 tj_ cai r-- r,0 6aw 4aaw co o 6a) r c\n ic-; c,, ii6 ai N u\i r- L6-noj ai 6 6 c c6 r-:COi O6 cj 4 6X)CY lq coCM LO CCO U)) rCM CM _ _ CM_ 0 00 CM C \j _ CM U )O 00U CM C -- O CM CO- LO 0 N-U, _- C'I) LOC'c N CM M~~~~~~~~~~~~ co 0 U) 0 C U) cc An 0 0 a) _NX0 00 . - . x~~ ~ ~ ~ ~ ~ -E "' )U 'a t::00c I ) I* ) j 0EU)W)0 oullico VoU) U) 0E co co ~~~~c 0~ U) 0 'm .~0 U U)E 0D a) o a0c0 -~0 0 0 -:Q 0 .oCO 11 0'- > (, ct o (D a> 0 =ao 0 rEU). 0U) 0 CDVo.c0 ~ (D ts E0 c 00Ccmcn >oK -> o '3 0-0- 0 00):3co- 00-CE$ ( C -U0- U) -.C U)Cc)" 00--75(- 0 J - CO 0 '(O 0W< O U 0 00<'E 00 g~ .'-NC')~U,(0N 00)'-'-'-'-'- '-N .'-NC')U,(0.'-NC7 )'2-N ' 4-N cQ WHO Bulletin OMS. Vol 72 1994 Q a co Global and regional cause-of-death patterns U) 00)co + 0 N- a)0) co CD 0)(D Cu E0) t 0 a')LL 1106 cna) C) 0 0 c6 CDj CD CM CM Cf) N- co CD 0) co N- CD (D aV) 0 (D 0 LC) 0 0co 0 0M 0 0lf N N-t N '0)N- CM D N-- 0C0)-c '-'' '' r-coOD NM1'- \ N- 00CD C C "itN - N )P l N-CM CDN NMp 3 l ' )0 -C CD C N C")- N- CD - 'CM a) - Na (D)U)CM t N- -CD- N0 C-DN NN C'. .- O C.'. 0 NNN-)_'_Cc DrN ro)N- 0)N ( O DC'- C')-' '-NheDC wc |w a) N N lt N- 0 N- CD N -0 C) CM- -cMO) LO CD) C'I) (P CR ,: ,q CD -LO- - 'D 'N D N CD tN Cl (3D _~( lL qC R( _ r- CWLCDO U Cli UZ 0 -C'I) LC*'- C') 'C ' '' N c CD) 0DLO(C(V - U)o~~~~~~c .C Z + CDi CJ CO 0 MLO'C'' ''C''')C i'C 0C\I U) CO0) cZ r- o aw _ 00 o) (. 'T P-N tl a-*c) _M -cm -CMa) c n _ cw _M - 0 3) cII CM cli It c tnuCM 0 V '- N'N CM Z+ U- o c'O a) Ca) C) .. X, (9 N-DC CD N '-a)N ac 0n C's D CDD N P l _\- CD O U) C \0N .D' C'C'.C. C) CO ..) 0 LONC' 0LO cm w cmCO- U) C 0 0 0D D)o.tLO o CDoCM _ ' C CO - L N '\ CD _ 'N£NN66' - 'cDe U) 0O C'N-M__CO _ 0 0 NON0 t LO 0'' ' OC(D-'O C' - co6 i c\i4 -)c\i a) t - cl _~ cio XC LOCDCD'-NC0CM -_ - _ CYU) c n NC M _C )_C > ~ ~~~0 0CMO a cm co 0 CD C> CDU) (CM C)O ))' '- |0|@CXO 6 -'C6iU) :U c-iu i 4 C\O r--CC CC\_UU C co LO CD 0) 0)co N- CMN t -a x 4) ~~~cr o rs ur ur V0~_ pcs 0 0 NurQ cl- 1. . . . . .M .M . ) Oc U) C) o r- 't LOr )C CD i )N' uNC_ M _OO- _ ur c u _0 rto u\CC\C _ LO - ic_ oCDN C') '-c') co C') NCD N nCCL)~C)--'N 'C DC C.DC').O_M< C 'I CY) LO_C\-C N-- CD cD C\t eN 0)0 ) 0O NLOM O N'C-Il Nc)P -I C)') CM- C')CMMItN0 .0 WHO Bulei OMS. VolCCCC72O 1994C)C C'O U) 0~~~~~~~~~~~~~~~~~~~~~( Cu~C c co0)- 0 n o la 0CM)U) 0-c* 0 .2 0~- CuCu~U)a 0 --- la -C U0 0.S' i CuEco c u coE W D02-0utoCu -'8 ~ 0 0 MM )C0Co Cu Cui5 ocuoU)Cun L- o c U U ) ) U) co 0. o~~~.0- C')'C6 . 0-a LU = CJM ttjcr'-C''i CC CDNq-CD)- - --- - .NC\)iCDCD.ui (L< < <<<««<<<<««<< << < 0000'U WHO Bulletin OMS. Vol 72 1994 N N- - CD _ 0 0 CD CD N CD "'d .4 N- C') CD N- CD C' N N-rC) N- CD N- CDN '- CD'- CD rs lo cq _ 0 _ (D NO)L -C\') r CM)'- C 'L L N-l:Ncr 0~L ~. 4 C'P) CN N4 ON-O)''NM'C')''-N O0NaD ~ CD_ 'd-C C') C') C D 0 co '0 'cC_LC) CD N Co CD 0 a CD N- N C0) NNM- 0) ''''''t( NC ')' 0Mqt V M - - 2"-0 _Cl _C\ CM N CD N 0 _ C' 0) CD O CD CY CM cq "I Co rs 0~ 0 q "I*: rls o pl 0) CM CM Nt C\j CO (D r- rl O aw o) Ul) CV - _ CM CD oo 0 N- 0) '0 CD 0 N- 0-00 CM C CDN-N- 0O LO D-)0O NCDN CDN- C')C dw N- 0 '~ '-N- 'N- 0) CD C') C 0 N '-' N- _ _)UD C- - N- N N- d 0) N- CD ' '0O CDv _CDN'-C)CD N- 1JD CDU) C CD C') 0 ) st O C' )CD O00 CD C') CD C') 0)0 C') CD CD CD CD'- -N C)' _- N _- '- CD CD d CD C Ci -O 0~~~ t .3 o U @ o S @ . E2 ( _ co- st1i o ._co - u<:< zr < <:- 0m mm .92 00'D0 (AU) 3n'=-' z o '0-0o8c= C uD( a: a.LL LLO0 cbIx -.'-NCu)CD Cy'- ') <.«<'(<< < < C Cu~0 co encia DD 0 0 0 co U) 0 c CD co 'a 473 I II LI C.J.L. Murray & A.D. Lopez C' C _'- t WrCN Ct N 0 _ 0) N Cd uCD NC t N 6 -_- N CY) N 00 - NcD LO C') CC) CD N 0)CM'- NC- C' CD co C') NC I)0 N CD NO-~ ''N CD' CDC') C'P)- '- CM Nv'-l c CM sc o _c co LON 7 .. cs-Cl) 0) 0 c CD- 0) 'CD - -'- NI 0)'- CD UCD 'N'NN ' C')- CD CD CD 0 N o L (D ~cD 'Al'I ''C')'', 0) CC) Ct) CDCM N ( '7 CO-C\ D tO (ot\ tN(D C 4 N C' C C ') C) '- CD t - - CV) C') LCD CDi I) *13 (C CD CD '* C') C(D fCD CM 04 l co cD| CiN N~rl~(D h- L ' CD CM 0)0) c) ' CDO'-Ca)D 0 N I C C CD CD N 0 N 0) CM N 0- ul' '00)D0 ' C\ 't' CMN\ C C')- C'I) C-i cai cs Ci qI i C) o- CM CM N N- CM C') C\( lr cXi Li I- I- P - 'CCD''C') CM - r - N 0) t N 0 N 0 t '- N 0a Ct 0) CD CDw 0 N - lt N C')M -- UC) CM-- C 'q' lo IQ VO O U)c qbU)CCM(P U) CM oic t _' - NC 0 0 _ CM 0 CM cJCD - C') C') t N_ @ |_ )Iq - - o -D cs)1OD C_ ur cso c C_ N0_ CD -N N CD 0) C')0)'_ ) CM(P V:co LQ (P cp LQo) a) CoC\ w _- t (D u" ClJcn 0_ _cJa wc CM- coCCD _UD 0'UOCDCD )0) N'_ '''C0D' ' '0'CDO')( ' 't' ''N N LOCD- CD CDC') aCDCD 0C') _- CD NCD C D CD C oo CM 0C" 0 o _ r" urc cn7V_ lur ~9 O CD''OCD C CDCr' LO 0CDC'C' ) CCD OD _ N OC'CM cfC)ITCMDONrCDltC )i q0'- C)N OD CM CM CM - CD D-N0 CON _CN CM C'f) v- CM0L c')' c'c'_ 66 V .Q *_ CM)tC NN( CM C N''N' 'N'-D)'V' 'N LO '0 C\ P M -L0' N C-: C\i pl. C")- (Dc\s CD o @ | @ u~~~~~~~~~~~~~L) CDC') 03) C ')NCM o~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~c co - ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ o~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~c a) ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 0 (1) z~~~l 0o ur cm _ O a~~~~~~~~~~~~~~~~~~~~~D _0oa a_a_cxsc c 0ou_0_(DC2a D_Cte Mcn D- 0 o a ) 0 CD CO a) co -CU (D c a)ocm s _ t__cM_cM0CM0oo0co o r_DC-Or _ co *o _ - C - 0 >, - n0 yQ0 8 co3 X y > 5 z g ° ig E D S E e '° a X " * i | E Ca 0~~~~~~~~~~~~~~~~~~~~~~H Bulei OMS VDCaol 7219 Cl)0 CD Ec a,t a,D co C l0 cn a5 V'a> CU) >*D E0, 'DB. ~EO ~~~~~~~~~~~~~~ Ii E 2 0 0 o (D~-2 0 a) 0 C) N C") Z cc .~~W 7 *0 a cts 0 - C) co a) 0 z en0) co 6 CD a) CD co 0 CD 0) CY) aL)O CU E 0CM L) C) Ch 0 oa 0 ) Ch 0) 0 C.) .0 C* co0. co COC0) 0) 0) U) 0 x0 0) 0) >1 .0 0 o 0 0*W 0 - E *_ 0 474 LI 1 Global and regional cause-of-death patterns C D) N C * _ C')CD 0 0 CD ~C%J '-N 0)0_ u''O '_c '0 C D 'o Le)UDt) Ch _ C') _ - C _ C\! ur ci q tlu |c C'f)00 D C'' ' CM ) r,. 9 l C COC)N N: .- .o .U o O 00 CY CM CM CV) CM-- - Lef) CMl -N -'l 'O LO_ 'C _t4 CY) CM Lrt(O 04 CM o N 0) Ns CD w 0)n_ (00( 0) 'DcDu ClDN'O CM'O N CY ') CM (ocD n N' - 1660)- C'q'-CY') C') CM - r-ClCD CDN 0'' '' - 'CD0 0 'C '- '. N C') C Y _C CD CD)N_ LO - 4t _- 0 _ C') C')_ N -CD C') CN__ '- N (00(w O Nur*c')N- ||N'-NI ) -_ C\i-NCCM 0 r,-_ P _ _cp loq uq_ 0 t,cxiQ _ LO0 N p- CO _ C0--CO_0CDNNN_N-N- C')a ,0a) D-'-NN)N'-r- N- NVcDcn_c w qcnw cM_ cm cY) l4 __ ' 0) 0 OD o.C 'q CcY) CD N) rC CM c' C) LO o. P cs CM Co cn C-i 4 C\i L6 i CM Cli (n li C; r-w Cx coCQ cs4 ui .4 0)N 0) C(D CD- v- co a 0) - c' cn cY) o> CD (D .) I0 q G rl. 'C D' CD N L C'MN - LC - NM co 'NC CM UC) 4n - CD )cD _)I.C') ')'-r 3DOU C l IQCDC' C')C')C' '''''0)CD0 o~~~ 60)_c'-c o -<^-< -C 0tcn 0 Cl? 114 C N CO CM '10 cn r,w_ Cl W o) cn_ cn) _ 1- rll rl- Ui C') N cn 0 cD I~~ ~ IC t U') Cl li 06 (D C\i lt cCM U) CD cni v- cli. . C6C'J V)0)NNY | @ @ 0o0c0cn c')ocn c '-C C') (0'- '- C' 0) 0 cnN urwc n'- _ '_r --t I " C C%j CI _ * * .- clc C") _ _ 0 _ O C') C') (0N C')-' ~~~~~~~~. 0a C l CD'0) N_ ec cc- r? tn w 04 LO V)W cn itCDDN C\) N @~~~~~~ ll*noeDe " 'A "Id:CCD tCM C') C c'itcsiNN cai . . n. . cN o:"Id: n wcq ur1 cmcs1"Id "Io0 s rolc9 P, 0 cco ur N q cs1 N Ct c0n 0O 00 OD-_ 0_u_ _ C N c _ cM cu (U co co a5 C. > OQ~~ ~~U =) (D-CQ Ca - C wn°<wc c ow-O _c c 0-: c: Co CCu coC C---Cu0 C~-0 coc - ). - 4 -S?co 5A0) o r ( cn0 E cj0 " U) Cu-o ~~~~CU CU~~~~~r-co Cu- Co Cu- 0r -CU IDCuCu4mCt ECO .~cOCuCuC Cu Cu 0C-L E '0,8Cu 3 w - r C CouCu U) Cu Cu CuCc CU ~ ~ ~ ~ ~ ~ ~ ~ ~ ~' U 0u C C:CO) o: Cu Co Coa C"C CDCUCL la CL c0Cu ~~~~~~0 00 -,m z rCUC Cu C/) CU 7-cUs ECc>- a. 'C O I).i 0 COLIOM Cu(D a ' f L 04 CL o C;C\'10 Cc a. .-.... 0 = .'-NCM M)-16CDNC D0)'-I-I-I-I .-N1:C\)'UC DCD( ' .'-CM "OU' )CD( '-C')cl a0<<<<<< < <«< ««««DOoooooQo ~«<"-CDcDcDCDm ui - coa) 0 0 0 C CU -a cu cu Cu WHO Bulletin OMS. Vol 72 1994 V C c Cv(A az 0 -E -o 6 z (u(D 0) 0) N I)CD 0 LO CD CD CU Cu w LOt co 0 C') CD Cu U) Cu + 0 cn 0) CD CD 0A CD 0) Cu Cu N 0 N CD NC0 CM o C) aN o 0) (D 0C) LCD N cm 0 Cy)O CM N 0) 0 N 0 C') CD N 0 CD CDN 0. N 0 CD 0 N C. CM O- N1 CM CD CD 0 N 0 Nx N 0 0 00 c L CD CD s0. CD N Nx 0. CM C NO 0 "tO C')M 0) & a) co 0 - n .c C. U) n Cu U (a 0) 0) Cu UF) 0) Cu .0 *0 Cu co 0 *0 la Cu U) uJ a- 0 c 0 D 0 0)Cu Cu co I II LiI I I i 475 C.J.L. Murray & A.D. Lopez CO CM COT 110 r N CM V 7- LO V N CC-N -) oo _ N C CC) _- CD I( __6 U-i0) U) CC) CC)U) ' N ( en _ t Cf) UC) NC C) 'CD U)N C~i NNL C C CM CC U) -O-D (D_ 0 OR COD U) OD 0)(D'C) OD tOCO 0- U)U)C0 N)O_ U) (0C C'f) U) CDt CM CDN N OD CM LO )_ )CM0 0(0 ND CM CrDN 0 0 'D 0 Cq CC OD OU) CC) (D0 U) QN - ,C-J 0 u -: N CD U)- - - 0) - . ( I- '(I 0N'. N N q N D CC))U) N 0N Cm U) 0(t _ 0-, '0) C' CY 0 N CM _ ' s1' CU) ' 'CD Id:) (C\! LP CO CO CM _ Cf)l? lIQ CMC)l N ',-CC)' 'U) NN N CM CM c") r- _CM U)(0CD '0 '' LCU)N0C) U') C') CO) -C NC)- NNCDU,C,)0)ON- N- NCcm c16 _ cw r-, 0 u c6 6 r-s r- cq o CM CM CC)CC)- C') -:P,o N 9VM No "II:M C4 'C_ CMC'CC C C''M_ L'f) C-CD C0 CD- 0) CC) C_ ) t CC )Ns' tCD' ' U) U) _--- NNCM tD $ * cr |U)t ' ' U) NCM (O ''''''.C.D ''.CR N - It -NM C') CD ) - C) rN co N rN 1' CC (0 (D N (D a cNi C 0i \i _- U) U) CN U) U) ) CC) U) _- CC C,)Ct- N NCM C) - OD_ 0 -C) Cu) N CC) CD q D (U) CD oO C q- No0 - ) CDN'r~-C C0C 6CDN0 PN0 0) N - U) )N )'ML-N04- CCl) @ ~ ~ D| 0) NC. . Ds v(DCDNU)C0.CD 0. NcoDC'J'r- N _ co s co tv rl (D CM CJ - Cli _ CM 6 co cM o i 67- (Cs C) N NN m ' D ONO U)'_0O C C)N'-_url CQ CM C@*|c)'_-U)scDCDCC'-N LfN C-o _ _ t N U) ~~~C\i ( D Ul) qtU)NLCM (0 NCC)N N CV l .qCR9 N .( co CM C\ CM C) C\'- I* |sb C l? C\! .M c CI) C\C)C C\'C O CCs_ ) 0 oN N CD N O -_ (0 \ DNtO N 0 CDCC) N-tC OC UC)CY)Nr 0 OC) CY) CM0) It0o CD t OD IN'- N CC) ' CN CN CM'N -C \D CY_ \ CMO_ - U) 0) NU)C)Nrcw c ''C )' ' '0' _DU)CCco C\ C) U) r LOC ' 6) CY -cUl_ CY) _ 0 Nr-0tU)C) N'- - C\ ' C\ _' CM _ _C\ CD O - r- CC) CD O NCM N ' ' ' ''C D ' CC )0)Na)' 'N'N@r-C c)N' CDL. _N _-_-CC)'_ __ '' ' ' '0O' ' ''U)0)C )''- ' '' 'CC) ~~. 9.C~C 7C d-'ll CM OD CC)_ 0) n) 0O)'C CO OD 0) t - 0 r- -rN _ o C) D0oCMCD (D CM C\ _ D _- T _ U) C _ tN CM CM_ C6 co 0 a) 0 ) Cu ) CCo 4 a)oa ~CDC uaC~~~~~~~~~~~~~~~o CZ D CuE cor aE cuo CU C). Co~~coB ji .~~~ 0 0 Cu V~~~~c aC u00) CDC~)a Ec U. a) co .0 ED5 COE LU 4) -- >- -= U)a. -D '= CM E Cu- a)aOcL ..j3 0 co 0~) ~c C5 Co coC (D)0 Co - =U) >I. (D~~~~~~~~~W tn7M VD cCu 0 .r Cu a) a zo u -S cu + 0 N 0) co 0 (0 0) to 0 0) N 'C LO 0 a) 0t Cu 0 aw CD N 0) Ln C) co0 LO (0 0) DU) 0 0) $U) U) Cu a) Co 0 _ a) 0 0 a) Cu 0 c m 0 * J .0 co (U) -) a)0 3 C) (c x E ~1 0 0) 0) 6(U (U x 60)(U .0 (U co(U 0 C *- c (A n *- co V la la E (U 476 Global and regional cause-of-death patterns C\i ci 06 CC\0cl( -) N-CV)N 1- 900 O CD CD 10CDCO N-)0cNo CM CN0)'0\' 'INur cvNLO CD _ _ sN _) UN CD CO CM LC) CM _ lql a) lq uz Cf) r- .- r- uC) _ _ co _ cN 0! CRo) coc (D C\J CMl Le) r, O O N P- CV) CM _ D)0) N st t 0 ' CC)N U) CX D -N CD- - NO 0 0 cOooi £ a CDco) C a N C)ltU)t CO C-0 NCf U-)CD) 10- C') _ C') _' D ' CON (00C NN' '-CMaOD( c 0 ca "t . ~NC')0C c6D lbC)-: CD NCD01 C' 1010n rCM 66 s coCM -: 1 C' CU')1 t N C' N N N)CD C' C')0 CD' -C')' U CD C') N M -- OD00 0 -C') - 10 C' v- N - 0 cNi '''''10 ) 6 CD CNC\ - N 10-1 0 0 N-C') '- 10 u - N4CD cm N-C) CMuN1 C CM CDCM01)N^- N C\ 0) o a) 0CV')- CM Co (D CMC') C'I) C') Nqc\N'- N L9. ... oir-c' N'-CM 0 C '1 r ... ) Uz CY) CM aw cmc c N_u NNCM o C') 0) N M _CD- r- 1CD1C CD 10 aC r6 i o I-cm 16) LOCD D 0,)rN E DCt ' 0) NMN-C) C\j t CN CD"C\JCO' )'ONcDOO NNN-'- 0) C')NCM- - - --' 0) -6 lqCI: V- LOc<< :<< <mc JOOO CM CU C 0C 'us0 -0D C0 03 Co ad 0 ' 00 0 8 r 75 D 0co 0 E c 0 t22 o sc)wZc00 nZ:0Co6)5j)*U) 5 <0- 1 c C cc$ C' C')' (D . 0C«««««CO 0 00 0 0C 0) 0) CD C') a)'- ''' oi4 -: Cl? c o7 CD0)rCD co c CD' -t NC C') 0NCVC\ c N- 0)0 CD N - N 100)TC-.ll Lq 00 t- 0 D 'U N CtOD-CD N C)MU) CM N 0)- CDC) N 0) 0) 10 N N C'.'. N N11 rN cMC')-NrM-LOL \ ClC) rN'- cl cl 0w 0a ON - \-N C\ 0) NCD '' ''''NC D ''-0.P~c - _rC)1 ' C0 _ 00 0 10 0 CD D CU) C\ C\C\l - U) - U\ C'C 'U) a) LO'- (OCDN- C) CN (0NO(D 10 '' CD CD 0) C) t oo C')- LO C C 0CN! C4 U) N cD O). 'N.'. .0) NCD)a) _ 0 0 LOCDCY) C O aD Ci (P C\! rIl N0M'CO O W 0 N 0 n 1' C') N CN t C CN 0CD00) N C') C')0)- N- - c') N C\- C )')C~~~~0 ~~~0 C.) 0 V co a 0) C.0 co.c CD- ') 10CD2 N 0 -o cot- 0 a. U- U- ._ C\l C,) IV LO Q C\j C- WHO Bulletin OMS. Vol 72 1994 0 am a) 0CD + urN 0Y) CD6 co a) 10 CY) U) 1t 0 ,tl 0 'EE a) IL CZ -C 'D 0 0 o -U0 a) co -z CD N N 0 Cq0 0o U) N(Dlt C') 0.N) 0Dcli CM a)N DO 0 _ 0 CD O O0 0 CD CO) C'I) 07)OD CD NM NO C) Ns 0 'C4 C') 0) N C\D CM CM C') 0 C\! N O CN N- 04O O N N C') 0_ OD Cf) NO 0. C D Ul C') 0) 0 CD 0 O C O 00 . -)0: 0D = ns 0 0)CD 0 zrl N 0) CD 0 CD 0) LO 0 m 6 0 10 6 C') NI, 10 0 0 CD00 0U 0) 0 A- b-0 0-0)co 0 0) 0) 0) U) C) x 0 0) 4 0) 0-.0 0) 0 C 0 0 *0 0la 0 0 0) 0 -C co a) 'O 0 0 -C 0 co 0 0 0 CD r- CO0 VD 477 C.J.L. Murray & A.D. Lopez c' N ') Nl- N- N N LO C') LA Lu Ul) CO_ _ _O_ 0) _ N I Q1N U) CJ ) N- ) L N LA L '0) N-LA U) r- Cf t XCM Cf7"It N- C') LA LA w 0A_ C') t C') 0) 0AL C') C'_) I' O CCO 00U) '''''C' _- 0) - ( l .. LA C') 0O- 0Cocf LA N N LA 0) LA LA 0 N L 0) 0'- LA 4 - '0 - CMN'-0'- '-L Nt- A N-) N N 0 oo(D r U C') C'_O )LA N 0 N- 0 UL N- UL UL C' C' UL 0) N N LA N U.- . LO C LO - Lt C N -A' ,N~'0 C- r~6 6666oio -'lqtC') - CY) '-L0 00 ) N_ '- 0_ -C_ C\j NN N *C ) *_ - - - CM N C') '' C C _l- C') C')N- C') 0A - L) 0) c' 0L 0) C' 0) 0 LR LA NCOCN CY) 'LALALA'-'-L 'NC t )C>J C J0 6\l( '''LA'- '0) 0)' 'LA C\J0 ''''''OC')L' 'N- _-_-C') C'i '''''''NL ' '0O '''''') t' ' C)' 'N- (.LA).. O OD CY) CO 0) C' t 0-_q L LA N LAO- N- N LA L 0 N- N- L _A'-N\-LN-(N0Nc-4rOAc % J\)L oON LAC b U U C o _ CM _ t LA - - a) N) 0 N- N- _- 0 C') t t NL 0 LA C' 0)6 6D(C\J st O Cf) C\J C\i stC\i 0 0_ C\ CJN_ C') A) '-N _L 0A - 'L N NC\L AU) ' 0 )C M0) CN- OC -N- - N(C'-(D _C\ _ 'N' CM OC C\' '00 N C\ '0 0) ' \jN C' C)NNM 'N ' C\ ' 0C)f)'' 0 _'' N '- A ''- I '- LA C\j CO - D LO' '-N- ''_-ONCLANLAN0)L0LA AD AD LA0)LA@N-o LA LO _ t LA coAC') N- CMCo-Nt 0~ ur Ut q cl '1:rv cl_ t c t -c -' ''''''NNL _a D_uONN ''- )0 'AN CM _ N N-U)LANt-L C'M C) CM LO lt r- oo CM c\ C\ - CM ri6Ci6N 00 A'-\LC L - ~~~. LO .1 .D 00 .) N C\ -NLA Nli 0C') C') C)C N 00 04 CM - - 7 c,:qa) _ cl9 Ct cl aw ucl '''0 LA' N--N '' '0 0 -C\i co CO (D CM 0u- t C\i - CM L)( C\ _ co _ CM UC) C\j Cl CY) CY) C\ q Lu r _Ncs a "II 9t @o _(D C") ULO CM CM _~~~~~~~~~~~~~~~4 _ 'I 'I N 0 t L 0 N- 'I 0 0 'OC C ') N\I LC C0-(DC) NCM NL\A'\I(D -C)'-C')L ADC')'-L c0) C') 0) NIJ C ') U)~~~~~~~~~a 0 a) :5~~~a 0o :30) 4) 2( 0o (a)Uc (U~~~~c 9) cn r- 0~~~~~) a))KiU 0)oE 0) E0 co a)-n (t5) 0 ~~~-C C'S co~~ 'U ~~~~~~ ( 0 00(UCM 'a CLM c.Q )a C(U~~ ~~~~c a ;C(U(U a 0 1- au E E 0 E a).- t.0 E U)> 0 co (D C's 0 2 m~~~~H Cl (U co0 G a) - zo C) 0 LO E N- a)(0 0r) CA LO (U 0 C') r) LO CD a) c0 0X cn 0) 0 CZ (0 a) 0r C')r ~a) 0 a) _ 0 a) (UO I-*0 0 0 Lu 0 x 0 0)0) 0 5> 0 U) 0 x la' .0 0 CU 0 0 0c0 U) 0 0 E 478 LI 41 1 Global and regional cause-of-death patterns N 0l) cli (0 I0 al) CM N CM(0 co co ur CI) co. Cl) OR CM r- CD CM cM 10 N N1 1- 'Rt C' 0 C' '- 0 - 000 (0(0( 010,Mr Wr - 1 1001010 Nl- N- '- C'It00C' D-N NI000t co csi~~~~~~~~~~~~~~~~~~~~~~ (010(0- N'- 10' N\N'J -'-N 0 - NNliC'l) '- C'f)O CD N 10 -LO N-D C') N'- C')100Cl0C'It)O)Cl-CMN'''ION ) , CD',-C\l-N(0C')t-N o10' o) 0) u')- C')CM' NM C') CDltC')- - - N ' 'qtCDCDC I CNC '-C'f) lqtNq r N'-MN \lC')(0C') - N\ o,Ct'"f C)1--LOC \lI Ir6NN CY) CM- 'ITCM C')N'- ,it-(0 0(0D C NNN-'N- N'-C 6' r c P O'a- - -oa('-OaN C') co 0(0 NC D'I-' N C' C'I) 0 LO010'- NOWm '-(0'N\(0 NC' 'N '- N T- a)~~~~~~~~~~~~~~~~( oo N')- C') '-1q00) (0 \ - 10) 'I C'f) a)IC C') NM C') C- 0C00)N'- 10 C CY)NNM- 'I-NNC N(0(0C')0NN C'I) a C) (0'(0~~N N- 0)? (00)o C' 0) N: N~ NI 'N 0. 'N . 'IT-0 ON MWL'N0-10N'- '- NOD'1 1 '-'0'''' ur C'U0 0t- cM~~~~0 C((0 0C\ CD It C\ C') C'- ) 10NCM 1D (-t (w U) 0 0 _ C C' 0) (0 C 0)0'C '- C'C- (w0 100) t No Co (0 C') C' 10' 10(0 N N 010 (0 C\j - CD (D q U) (D OR C4 C4 ai o Ci (D 00 ') N N C' a "it C') 0)0LO V- '-_cO N 10C_')C') C' NDswU N) CD0(N 0N C (D '0 ) 0' (0 - 00 CM CN 10t 0) N0) C') CM (0 C t0Tc1( a7) 07 cl) _ Co o_c\ c 0 co comt 0 r Nca LOur CM 0L N _1CC N'- N (' N (D _ 'I CN0)R C- 0( LO t 00) Co0C't o) C') C) (CCM _ N'- 1 0 csC)R(0 C')-OR - 0 (01 "t CY) c\j __ _CM C\ _ o0-' '- (010O_- N N100c N0)010 N)u 0) N C) C') N - rN 0)' (00)O 0 0Y)C) ItC'-01 LC 10 (01Nr0) 10C N o10 0CN 00 (0- 0C-) 0 C,) ) CoCLe CN - Lq '1 C) 0R q _ cD O _ O (D LI! o 00 C) CY CDO 0 CM CM '-( 0 N CO _- C') '-N N N cmNpur o~q al ur O~ur O rl O~ 0 (D aw o cm _f- c-w Oc00 CD CY Y \ \ M0 \ atw \ P- C C,, Fl- _ r- 0) _4 00 ) (O LC a) (D V) CY C\l UD Co CV) (0 UC) C\l UC) C) r C\j CM rl IT (D UC) a) a) (D0 LO _r- CM CM CM Co CM lt lt a) aw Co LO CM Co _ 0V P- LOIs0 (D CY 10) CM 1 CD r- V) Co Co CY)_ cm CMJ CO OD C) Cv) O () C) 1- CD CD CM C,) - --D CO -l CM CM CM r- - r) uzT 00 r- CII LO aw N w o0 OD't - - (D r 0 "t r- (D C\l (D t- lt C) UC) N 'It 00 r- OD -mc\l oII a 1 _0)'I Cj - Dr(-Nt ai a6 o 4 6O 4 c , ci 0 C\i CO N (0 '-C) r Nt a o T c\ LO cLI) C f ) CN l tC) t CO N)c\ C X) ) cN C\! a! (D V! O 0 00 LO 00 -T c) 0 O 04 CY) O C) q _t C\l 0 _ C\J - o_It OD cIJ CI u10) (0co 0)N N'-\ C\ 'Zi - cU en C Ino* 0) 0 no 0 -I o L c -ON0a ___ .*_c ot(CO -- U) (D _ CM CCI<:_) (DU:(a Co co<<< c<<z :< OOOOa >< :< coCo E 00ean 'D Ci co 0 0 s -C *n CM 2 tS co co ~ ~ ~ 1I 0C EI In -In~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ L c a0> O- a I )0~C00 - . co N C75 >C. n.Z0Z 0oL : .'- CC) 0- < <««<(000000000 «<< < < < WHO Bulletin OMS. Vol 72 1994 r C-' N- c1 C) C-co 0) ci (D O 7 co 0) C\) url Uh'v a(n 0 Co 00 a z CD 0 LO 0) In + 0 N 0)(D 0 co 0) 10 1 0) N 1) 0 CoMC 0 0) In + 0 0) 0) 0) ~a) or (00 0M 0 _c 0) 0 0 0a In 0 3: a) * " 0) U) cn UU) x U) U) D 0) ._ci CU) U) E U) 0 U) U) la cni In co -C 0 V- a 00 Cl Un a) In .-6C. -CIn Inco 'a c c c E c u (A 0 O -M .9 479 C' ,- 0 CD qt IN CD N- CD -i 0 qlt AC)C D' D O )~ 0 0L ) A' DN N N NL C') 0 - CM0')NCOtCf) CO) -O CO)CD)CY Le) 1-0N\Nl 0 N'-CM v--CO 14* D CD()0()0t0 0)' '-CD C C CM'- Cv)CM- C') CO - CO '- C'LLO OCD CM D Nl NC'NOC CO) "it C\l ~~~~ ~~~~R-C\iC\i1j N- N-J N-t No LA 4 LA LA N-0LO C ')CM )CM') 00) CM LA LA) '-) - CM NII CD 0 NOe)-CDONLA V00 )N C') 'l- CD fl,c-i6--o -L NCt-o I- ) OC)- CDC '- NV C-N 0 C D0l- C')0)LO)1 CM 0) N~ L CO)CODLAL LA 0L CO0) CO))0LO CLOCDCDL CMNN)C 00C)N CM-C')' C J0NI w-0)-CM)CLOCO)'CO '-CDNCM-N 0 - -N0)NT-CDt-' CD '--CD 0') NCDDNl C') '- -- - q-CMNCM C-'N N 0N lAqC)M N - c') C-CD CM '-NCM CO)CM '\-CD COC' C'I)00)LUACN40) LO C') ' - gCOCDO '-CD-00v -C)LCD0-'-qDCD 'LA) - NM C') OD- CMN- NM NC-C'I) N'- N~~~~~~~~~~~~~~~~~~~~~~CjC NLA '- N'-C NM N- NM (6ODC c r- CM6oiCY) r- 6' -CD CD 0)C) qtN'- C'CCDt-NC-N CMNCO - C) CD CO)N 0 LA)N- - NM 0)ylLANNCm-NCm'C N NCM --'CN 1 OD7- CDONLA'- 0'-tC'NC CO) 11 C'f CD Ci \ -: 1 NCM DC)0)Nr- Nc')N C" O C Y)C 0) LOA0 - (O() -NCMCD I N D CDNC CD - C') 1- It N C) C) LONCDNCMLC CD C') C')N0\c)N LA) V'~C')CO CDI0) -'-C') -CD CO)LAO '- N CD CDA' - C')N MC LOOD 'C- q )L) C)N -( M, O r ) L)C P-0 ~ N-'--NM0CN C 1'C -O f l C l OCCCDDDC)') ,000OCD ' aLA0CM --0N0CM)'CLAC CD'- C') c') OD LO '--'- N' C') '-C') - C NNiNMC NCM '-O N') CD C')N- CD C- ) N N~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\ N -L NIt)It 0M ALALN C'N0) )0N CD CD N N 00' C')0 C')0 0) CDJ C') 0' CD 07) CD Cqt LAi - NNCCNM 0) 0 co~~~~ x ~~~~~~0 E Cu C0A~~~c(D 0D co CoCc C.) a)~ co)4) 0 _ Ec EC 0 n 0Cuto ODCcu Q 00 0CuC ~c (D C COCD0 C 0C : uC X 0* ' 0*30 ~ C CuD E OD 0orCM0 E c-02 > 000Eco Cu0 u0 CDJJ -u~ , o'E .0 C0to Ct Lo a CL a o I- ~~ N'0A -- 'F- E0~~W C.J.L. Murray & A.D. Lopez 'u C_ co -c 0 -o S5 zeco 1z U 003) 0)N co 6 CD a) LO a6 a) a)E LX O CD CD 0) co 0F) CD O (0 N 0) (0 0 CD 0) 0) co CO ) CMg e 0 0) 0 -50) Cu Cu 0) la m 0 0) or U) U- U U) M0) 0) la c co *0 0 U) U) 40