Estimates Of Adult Mortality Research Articles

Educational differentials in adult mortality in low- and middle-income countries

Résumé
 Dans les pays à haut revenu, il a été amplement démontré que la mortalité des adultes est plus faible au sein des catégories professionnelles favorisées ainsi que chez les personnes bénéficiant d’un niveau d’instruction élevé ou d’un revenu confortable. Dans les pays à revenu faible ou intermédiaire, les différentiels de mortalité aux âges adultes sont par contre nettement moins bien étudiés en raison du manque d'information sur les caractéristiques socio-économiques des personnes décédées. Cet article vise à évaluer si les informations fournies dans des enquêtes sur la survie des sœurs peuvent aider à combler cette lacune. Puisque les niveaux d’instruction sont très corrélés au sein des familles, les niveaux atteints par les fem­mes enquêtées peuvent servir à approximer le niveau d’instruction de leurs sœurs. Afin de quantifier les différences de mortalité, nous combinons toutes les Enquêtes Démographiques et de Santé dans un même modèle linéaire généralisé à effets mixtes. Dans la plupart des enquêtes, nous observons une relation inverse entre le niveau d’instruction et la mortalité adulte, en particulier dans les zones urbaines, mais des gradients opposés sont observés dans plusieurs pays durement frappés par l’épidémie du VIH/sida. L’hypothèse selon laquelle les femmes moins instruites auraient tendance à omettre disproportionnellement certains décès de leurs sœurs n’est pas vérifiée dans les données. Les données sur la survie des soeurs semblent plutôt bien refléter la complexité de la relation entre mortalité et éducation des adultes, en particulier dans les pays touchés par l’épidémie du SIDA. Dans l’ensem­ble, le recours à ces données ouvre de nouvelles voies pour mesurer les inégalités de mortalité aux âges adultes dans les pays à revenu faible ou intermédiaire.
 Abstract
 In high-income countries, there is extensive evidence showing that higher levels of educational attainment, higher income and higher occupational classes are correlated with lower mortality rates among adults. In low- and middle-income countries, far less is known on mortality differentials in this age group due to the lack of information on the socio-economic characteristics of the deceased. In this paper, we evaluate whether survey reports on the survival of sisters can help to fill this gap. We find that levels of educational attainment are correlated within families, and therefore characteristics of the deceased can be inferred from levels attained by their sisters responding to the survey. Because estimates based on sibling survival histories have large confidence intervals, we pool all Demographic and Health Surveys together and apply a generalized linear mixed-effects model to capture mortality differentials. In most surveys, higher education is indeed associated with lower risks of dying, especially in urban areas, but reverse gradients are observed in several countries experiencing generalized HIV epidemics. There is limited support for the hypothesis that women with less education are disproportionately omitting to report some deaths. Instead, we argue that sibling histories truly reflect the complexity of the relationship between adult mortality and education, especially in countries affected by the HIV/AIDS epidemic. Overall, the sibling approach offers new avenues for measuring inequalities in adult mortality in low- and middle-income countries.

Monitoring adult mortality by type of residence in the absence of death registration: a perspective from Burkina Faso

In the context of the post 2015 agenda, disaggregation of mortality indicators is needed to assess health inequalities within populations. However, producing sub-national estimates of adult mortality is notably difficult in the absence of death registration. Using Burkina Faso as a case study, this paper revisits the main avenues to quantify differences in adult mortality between the ages of 15 and 60 according to urban/rural residence. Estimates are based on reports on the survival of parents and siblings collected in surveys and in the 2006 census, and compared to levels inferred from recent household deaths or inferences based on child mortality. Results indicate that in Burkina Faso, adults living in urban areas still benefit from a health advantage compared to their rural counterparts. Thus, efforts made in reducing adult mortality in rural settings should be intensified. In terms of methods, this analysis shows the value of asking additional questions about the place of residence of close relatives to avoid misclassification errors. The approach adopted here could be implemented in other countries to facilitate the measurement of spatial inequalities in health indicators for all ages when monitoring Sustainable Development Goals(SDGs).

Método de estimação de grau de cobertura em pequenas áreas: uma aplicação nas microrregiões mineiras

O artigo tem como objetivo principal apresentar uma metodologia simples para estimar o grau de cobertura dos registros de óbitos e obter estimativas de mortalidade para pequenas áreas, usando Minas Gerais como exemplo. A metodologia proposta combina a padronização indireta da estrutura de mortalidade de áreas menores com base nas funções de mortalidade de áreas maiores, e depois a aplicação dos métodos de distribuição de mortes para obter estimativas mais robustas de mortalidade para áreas menores. Os resultados obtidos são robustos e adequados quando comparados com estimativas oficiais e usando outros métodos. As estimativas em dois estágios foram tão robustas que elas reduziram o sobre-registro de óbitos em alguns casos e melhorou as estimativas de mortalidade adulta em algumas áreas onde os dados são menos confiáveis.

Reporting Errors in Siblings’ Survival Histories and Their Impact on Adult Mortality Estimates: Results From a Record Linkage Study in Senegal

Estimates of adult mortality in countries with limited vital registration (e.g., sub-Saharan Africa) are often derived from information about the survival of a respondent's siblings. We evaluated the completeness and accuracy of such data through a record linkage study conducted in Bandafassi, located in southeastern Senegal. We linked at the individual level retrospective siblings' survival histories (SSH) reported by female respondents (n = 268) to prospective mortality data and genealogies collected through a health and demographic surveillance system (HDSS). Respondents often reported inaccurate lists of siblings. Additions to these lists were uncommon, but omissions were frequent: respondents omitted 3.8 % of their live sisters, 9.1 % of their deceased sisters, and 16.6 % of their sisters who had migrated out of the DSS area. Respondents underestimated the age at death of the siblings they reported during the interview, particularly among siblings who had died at older ages (≥45 years). Restricting SSH data to person-years and events having occurred during a recent reference period reduced list errors but not age and date errors. Overall, SSH data led to a 20 % underestimate of 45 q 15 relative to HDSS data. Our study suggests new quality improvement strategies for SSH data and demonstrates the potential use of HDSS data for the validation of "unconventional" demographic techniques.

On the Fast Track to Recovery: Island Foxes on the Northern Channel Islands

The island fox (Urocyon littoralis) represents an unusual case of a species that achieved virtual recovery a mere 15 years after population declines were first discovered. Island fox subspecies on San Miguel, Santa Rosa, and Santa Cruz islands declined precipitously in the mid-1990s due to predation by Golden Eagles (Aquila chrysaetos), which had not historically bred on the islands. In 2008, a 10-year period of recovery action implementation ended. The recovery program had included captive breeding and reintroduction of island foxes and capture and relocation of Golden Eagles. Free-ranging fox populations have been monitored to assess recovery of each subspecies and to detect potential threats of disease and predation. Monitoring included (1) annual grid trapping to allow estimation of annual population size via capture-mark-recapture methods and (2) systematic surveillance of radio-collared foxes to allow estimation of mortality rates and causes. A comprehensive demographic modeling effort produced a population recovery tool that uses adult mortality and population size estimates from the monitoring programs to estimate extinction risks for each fox population. The tool allows managers to assess when threats are sufficiently mitigated to consider populations acceptably safe from extinction. Population monitoring indicates that island foxes on the northern Channel Islands have increased up to 30-fold from population lows and that annual survival has been 90% or better in most years. The San Miguel and Santa Cruz subspecies have approached or reached predecline population levels, and application of the recovery tool indicates they will be biologically recovered by 2013. Biological recovery of the Santa Rosa subspecies, hindered by predation which caused lower survival in 2010, will occur by 2017.

Acaricidal activity of Cymbopogon winterianus, Vitex negundo and Withania somnifera against synthetic pyrethroid resistant Rhipicephalus (Boophilus) microplus

Detection of resistance levels against cypermethrin and deltamethrin, the most commonly used synthetic pyrethroids (SPs) against Rhipicephalus (Boophilus) microplus collected from Moga, Punjab (India) was carried out using larval packet test. Results indicated the presence of resistance of level I and III against cypermethrin (resistance factors (RF) = 4.67) and deltamethrin (RF = 34.2), respectively. Adult immersion test was used to assess the acaricidal activity of aqueous and ethanolic extracts of leaves of Cymbopogon winterianus, Vitex negundo, and Withania somnifera along with roots of V. negundo against the SP resistant engorged females of R. (B.) microplus. The efficacy of various extracts was assessed by estimation of percent adult mortality, reproductive index (RI), percent inhibition of oviposition (%IO), and hatching rate. A concentration dependent increase in tick mortality was recorded which was more marked with various ethanolic extracts, and highest mortality was recorded in ticks treated with ethanolic extract of leaves of C. winterianus. The LC50 values were determined by applying regression equation analysis to the probit transformed data of mortality for various aqueous and ethanolic extracts. Acaricidal property was recorded to be higher in ethanolic extracts, and high activity was found with the ethanolic extract of leaves of C. winterianus with LC50 (95% CL) values of 0.46% (0.35-0.59%), followed by W. somnifera as 5.21% (4.45-6.09%) and V. negundo as 7.02% (4.58-10.74%). The egg mass weight of the live ticks treated with different concentrations of the various extract was significantly (p < 0.01) lower than that of control ticks; consequently, the RI and the %IO value of the treated ticks were reduced. Further, complete inhibition of hatching was recorded in eggs laid by ticks treated with ethanolic extracts of leaves of V. negundo and aqueous extracts of leaves of W. somnifera. The results of the current study indicate that extracts of C. winterianus, V. negundo, and W. somnifera can be used for the control of SP resistant ticks.

Divergent mortality trends by ethnicity in Fiji.

To examine trends in infant mortality rate (IMR), adult mortality and life expectancy (LE) in the two major Fijian ethnic groups since 1975. Estimates of IMR, adult mortality (15-59 years) and LE by ethnicity are calculated from previously unreported Fiji Ministry of Health data and extracted from published sources. Over 1975-2008: IMR decreased from 33 to 20 deaths/1,000 live births in i-Taukei (Fiji Melanesians); and 38 to 18 in Fijians of Indian descent. Increased adult male mortality among i-Taukei and decline among Fijians of Indian descent led to an equal probability of dying in 2007 of 29%; while in female adults the probability trended upwards in i-Taukei to 25%, and declined in Fijians of Indian descent to 17%. Life expectancy in both ethnicities increased until 1985 (to 64 years for males; 68 for females) then forming a plateau in males of both ethnicities, and Fijian females of Indian descent, but declining in i-Taukei females to 66 years in 2007. Despite IMR declines over 1975-2008, LE for i-Taukei and Fijians of Indian descent has not increased since 1985, and has actually decreased in i-Taukei women, consistent with trends in adult mortality (15-59 years). Mortality analyses in Fiji that consider the entire population mask divergent trends in the major ethnic groups. This situation is most likely a consequence of non-communicable disease mortality, requiring further assessment and a strengthened response.

Adult mortality estimates for small areas in Brazil, 1980–2010: a methodological approach

BackgroundIn recent decades, life expectancy and health status have improved considerably in Brazil. Life expectancy at birth has risen from 48 years in 1960 to over 73 years in 2010. However, the country is marked by large regional and socioeconomic inequality. Due to data limitations, incomplete coverage of vital registration systems, and errors in age declaration for both population and death counts, little is known about the level and trends in adult mortality for sub-national areas. This study aimed to study the evolution of adult mortality in small areas in Brazil from 1980 to 2010, using 45q15 as a summary measure of adult mortality across small areas in Brazil. MethodsData were obtained from the mortality database available from Datasus, a vital statistics database organised by the Brazilian Ministry of Health that contains information on deaths by age, sex, and cause from 1979 to 2010. Population censuses were used to acquire the population by age and sex for 137 comparable small areas in Brazil, as defined by the National Statistics Office, for the years 1980 to 2010. We combined traditional formal demography methods, called death distribution methods, with indirect standardisation and Bayesian statistics to produce more reliable estimates of adult mortality for the 137 small areas. FindingsThe results show that data quality has improved considerably across Brazil during the period of analysis, but regional variation persists. On average, completeness of death counts increased from 80% (range 50–100) in 1980–91, to 95% (73–100) in 2000–10. Estimates of adult mortality show that more developed regions in the south and southeast parts of the country have lower mortality levels than the northeast, but there is a clear convergence process. InterpretationThis study indicates that the combination of traditional demographic methods to Bayesian statistics is a successful alternative to produce mortality estimates for small areas in Brazil. We also found a large improvement in the quality of mortality data in Brazil since 1980. Finally, we observe a convergence process on health status. FundingBrazilian National Counsel of Technological and Scientific Development (CNPq) and Minas Gerais Research Foundation (Fapemig).

Improving the Measurement of Maternal Mortality: The Sisterhood Method Revisited

BackgroundOver the past several decades the efforts to improve maternal survival and the consequent demand for accurate estimates of maternal mortality have increased. However, measuring maternal mortality remains a difficult task especially in developing countries with weak information systems. Sibling histories included in household surveys (most notably the Demographic and Health Surveys (DHS)) have emerged as an important source of maternal mortality data. Data have been mainly collected from women and have not been widely collected from men due to concerns about data quality. We assess data quality of histories obtained from men and the potential to improve the efficiency of surveys measuring maternal mortality by collecting such data.Methods and FindingsWe used data from 10 Demographic and Health Surveys (DHS) that have included a full sibling history in both their women’s and men’s questionnaires. We estimated adult and maternal mortality indicators from histories obtained from men and women. We assessed the completeness and accuracy of these histories using several indicators of data quality. Our study finds that mortality estimates based on sibling histories obtained from men do not systematically or significantly differ from those obtained from women. Quality indicators were similar when comparing data from men and women. Pooling data obtained from men and women produced narrower confidence intervals.ConclusionFrom experience across nine developing countries, sibling history data obtained from men appear to be a reliable source of information on adult and maternal mortality. Given that there are no significant differences between mortality estimates based on data obtained from men and women, data can be pooled to increase efficiency. This finding improves the feasibility for countries to generate robust empirical estimates of adult and maternal mortality from surveys. Further we recommend that male sibling histories be collected from all sample households rather than from a subsample.

Age-specific and sex-specific mortality in 187 countries, 1970–2010: a systematic analysis for the Global Burden of Disease Study 2010

Estimation of the number and rate of deaths by age and sex is a key first stage for calculation of the burden of disease in order to constrain estimates of cause-specific mortality and to measure premature mortality in populations. We aimed to estimate life tables and annual numbers of deaths for 187 countries from 1970 to 2010. We estimated trends in under-5 mortality rate (children aged 0-4 years) and probability of adult death (15-59 years) for each country with all available data. Death registration data were available for more than 100 countries and we corrected for undercount with improved death distribution methods. We applied refined methods to survey data on sibling survival that correct for survivor, zero-sibling, and recall bias. We separately estimated mortality from natural disasters and wars. We generated final estimates of under-5 mortality and adult mortality from the data with Gaussian process regression. We used these results as input parameters in a relational model life table system. We developed a model to extrapolate mortality to 110 years of age. All death rates and numbers have been estimated with 95% uncertainty intervals (95% UIs). From 1970 to 2010, global male life expectancy at birth increased from 56·4 years (95% UI 55·5-57·2) to 67·5 years (66·9-68·1) and global female life expectancy at birth increased from 61·2 years (60·2-62·0) to 73·3 years (72·8-73·8). Life expectancy at birth rose by 3-4 years every decade from 1970, apart from during the 1990s (increase in male life expectancy of 1·4 years and in female life expectancy of 1·6 years). Substantial reductions in mortality occurred in eastern and southern sub-Saharan Africa since 2004, coinciding with increased coverage of antiretroviral therapy and preventive measures against malaria. Sex-specific changes in life expectancy from 1970 to 2010 ranged from gains of 23-29 years in the Maldives and Bhutan to declines of 1-7 years in Belarus, Lesotho, Ukraine, and Zimbabwe. Globally, 52·8 million (95% UI 51·6-54·1 million) deaths occurred in 2010, which is about 13·5% more than occurred in 1990 (46·5 million [45·7-47·4 million]), and 21·9% more than occurred in 1970 (43·3 million [42·2-44·6 million]). Proportionally more deaths in 2010 occurred at age 70 years and older (42·8% in 2010 vs 33·1% in 1990), and 22·9% occurred at 80 years or older. Deaths in children younger than 5 years declined by almost 60% since 1970 (16·4 million [16·1-16·7 million] in 1970 vs 6·8 million [6·6-7·1 million] in 2010), especially at ages 1-59 months (10·8 million [10·4-11·1 million] in 1970 vs 4·0 million [3·8-4·2 million] in 2010). In all regions, including those most affected by HIV/AIDS, we noted increases in mean ages at death. Despite global and regional health crises, global life expectancy has increased continuously and substantially in the past 40 years. Yet substantial heterogeneity exists across age groups, among countries, and over different decades. 179 of 187 countries have had increases in life expectancy after the slowdown in progress in the 1990s. Efforts should be directed to reduce mortality in low-income and middle-income countries. Potential underestimation of achievement of the Millennium Development Goal 4 might result from limitations of demographic data on child mortality for the most recent time period. Improvement of civil registration system worldwide is crucial for better tracking of global mortality. Bill & Melinda Gates Foundation.

Adult Mortality From Sibling Survival Data: A Reappraisal of Selection Biases

Because of incomplete registration of deaths in most countries in sub-Saharan Africa, data on the survival of close relatives constitute the cornerstone of estimates of adult mortality. Since 1990, sibling histories have been widely collected in Demographic and Health Surveys and are increasingly being relied upon to estimate both general and maternal mortality. Until recently, the use of sibling histories was thought to lead to underestimates of mortality, but a more optimistic view in the literature emerged with the development by Gakidou and King (Demography 43:569-585, 2006) of corrections for selection biases. Based on microsimulations, this article shows that Gakidou and King's weighting scheme has been incorrectly applied to survey data, leading to overestimates of mortality, especially for males. The evidence for an association between mortality and sibship size in adulthood is reviewed. Female mortality appears to decline slightly with the number of surviving sisters, although this could be an artifact of severe recall errors in larger sibships or familial clustering of deaths. Under most circumstances, corrections for selection biases should have only a modest effect on sibling estimates.

Smoking-Attributable Mortality by Cause in the United States: Revising the CDC's Data and Estimates

Smoking is the leading cause of preventable mortality in the United States, but the methods and data used in the Centers for Disease Control and Prevention's (CDC) published estimates of adult smoking-attributable mortality have not been substantially revised since their introduction in the 1980s. We employed the CDC's general methodology for estimating smoking-attributable mortality but produced improved estimates by using recent, nationally representative relative risk data from the National Health Interview Survey-Linked Mortality Files and adjusting for confounding risk factors. We also produced estimates by smoking status and over time. Our use of more recent and nationally representative relative risks tended to decrease estimates of smoking deaths for men and increased estimates for women compared with the CDC's estimates. Adjustment for confounding factors further refined the estimates, particularly by smoking status. We estimated 200,000 smoking-attributable deaths for men and 180,000 smoking-attributable deaths for women in the United States in 2004. Estimated smoking-attributable mortality has finally begun to decline for both U.S. men and women. Our approach offers several substantive improvements in the estimation of smoking-attributable mortality by cause for the United States. Cigarette smoking remains a leading cause of preventable mortality in the United States, but we estimate that the number of smoking-attributable deaths has begun to decline.

Estimating Mortality Differences in Developed Countries From Survey Information on Maternal and Paternal Orphanhood

In general, the use of indirect methods is limited to developing countries. Developed countries are usually assumed to have no need to apply such methods because detailed demographic data exist. However, the potentialities of demographic analysis with direct methods are limited to the characteristics of available macro data on births, deaths, and migration. For instance, in many Western countries, official population statistics do not permit the estimation of mortality by socioeconomic status (SES) or migration background, or for estimating the relationship between parity and mortality. In order to overcome these shortcomings, I modify and extend the so-called orphanhood method for indirect estimation of adult mortality from survey information on maternal and paternal survival to allow its application to populations of developed countries. The method is demonstrated and tested with data from two independent Italian cross-sectional surveys by estimating overall and SES-specific life expectancy. The empirical applications reveal that the proposed method can be used successfully for estimating levels and trends of mortality differences in developed countries and thus offers new prospects for the analysis of mortality.

Differential Record Linkage by Hispanic Ethnicity and Age in Linked Mortality Studies

This study examines how the linkage of surveys to death records differs for Hispanics and non-Hispanic Whites and how such differences affect estimates of ethnic differences in U.S. adult mortality. I use data from the 1989-2000 National Health Interview Survey (NHIS) linked to the National Death Index (NDI) through 2002. Analyses assess how match score and match class vary by ethnicity, nativity, and age and whether mortality hazard ratios are sensitive to shifts in match criteria. Linkage quality is lower for Hispanic and foreign-born adults than for non-Hispanic White and U.S.-born adults. Modification of the linkage criteria determine whether the Hispanic mortality advantage is observed among middle-aged adults. The accuracy of adult mortality estimates depends on the quality of the linkage between surveys and death records.

Differences in life expectancy by education and occupation in Italy, 1980–94: Indirect estimates from maternal and paternal orphanhood

In the present study, we use the modified orphanhood method to analyse mortality differences by socio-economic status in Italy. This technique permits the indirect estimation of adult mortality from survey-based information on parents' survival in developed populations and helps to overcome several limitations of conventional studies on mortality differences by social class. We estimate a time series of life tables by education and occupation and analyse the differences in life expectancy by socio-economic status along with their changes between 1980–84, 1985–89, and 1990–94. Whereas mortality differences between the highest social class and the other socio-economic status groups increased among men, they decreased among women. We speculate about the reasons for these sex-specific trends and evaluate the application of indirect estimation techniques to the populations of developed countries.

A general model for mortality in adult tsetse (Glossina spp.)

Tsetse exhibit a U-shaped age-mortality curve, with high losses after eclosion and a well-marked ageing process, which is particularly dramatic in males. A three-parameter (k(1) -k(3) ) model for age-dependent adult instantaneous mortality rates was constructed using mark-recapture data for the tsetse fly Glossina morsitans morsitans Westwood (Diptera: Glossinidae). Mortality changed linearly with k(1) over all ages; k(2) affected only losses in roughly the first week of adult life, and k(3) controlled the ageing rate. Mortality pooled over age was twice as sensitive to changes in k(3) as in k(1) . Population growth rate was, however, similarly affected by these two parameters, reflecting the disproportionate effect of k(3) on mortality in the oldest flies that contribute least to the growth rate. Pooled-age mortality and growth rate were insensitive to changes in k(2) . The same model also provided good fits to data for laboratory colonies of female G. m. morsitans and Glossina austeni Newstead and should be applicable to all tsetse of both sexes. The new model for tsetse mortality should be incorporated into models of tsetse and trypanosome population dynamics; it will also inform the estimation of adult female mortality from ovarian dissection data.

Age, growth and mortality of black redhorse (Moxostoma duquesnei) and shorthead redhorse (M. macrolepidotum) in the Grand River, Ontario

Summary Despite comprising a large portion of riverine fish biomass, very little demographic information is available for redhorse sucker (Moxostoma spp.) populations in Ontario (Canada). Information is presented on size and age structure, growth, and mortality of black redhorse (M. duquesnei), a threatened fish species, and shorthead redhorse (M. macrolepidotum) in the Grand River, Ontario. Estimates of body condition, adult mortality, longevity, and maximum attained length were very similar for both species. However, the more abundant shorthead redhorse reached maturity at a smaller size and younger age and grew at a slower rate than the black redhorse. Compared to populations in more southerly parts of its range, Grand River black redhorse grew more slowly, matured later and at a larger size, reached a higher maximum length and age, and was characterized by lower adult mortality rates. Reflective of its central location in the species range, the age and length characteristics of the Grand River population of shorthead redhorse were intermediate of, and overlapped with, other sampled populations.

Adult mortality in a rural area of Senegal

This study provides original estimates of adult mortality in Mlomp a rural population of Senegal which has been monitored for twenty years. Causes of death are assessed through verbal autopsies which are completed by medical information. Between ages 15 and 60 male mortality is much higher than female mortality. Globally AIDS mortality does not have the tragic impact observed in other regions of Africa and maternal mortality is relatively low for a rural area unlike injuries which are common among men. In Mlomp non-communicable diseases especially cancers are predominant. In addition to behavioural factors infectious diseases may contribute to this situation. (authors)

Discrepancies between UN models and DHS survey estimates of maternal orphan prevalence: insights from analyses of survey data from Zimbabwe

Objectives:Model-based estimates of maternal (but not paternal) orphanhood are higher than those based on data from demographic and health surveys (DHS). We investigate the consistency of reporting of parental survival...

Adult HIV mortality in Zimbabwe

To compare alternative methods to vital registration systems for estimating adult mortality, and describe patterns of mortality in Manicaland, Zimbabwe, which has been severely affected by HIV. We compared estimates of adult mortality from (1) a single question on household mortality, (2) repeated household censuses, and (3) an adult cohort study with linked HIV testing from Manicaland, with a mathematical model fitted to local age-specific HIV prevalence (1998 -2000). The crude death rate from the single question (29 per 1000 person-years) was roughly consistent with that from the mathematical model (22 -25 per 1000 person-years), but much higher than that from the household censuses (12 per 1000 person-years). Adult mortality in the household censuses (males 0.65; females 0.51) was lower than in the cohort study (males 0.77; females 0.57), while mathematical models gave a much higher estimate, especially for females (males 0.80 -0.83; females 0.75 -0.80). The population attributable fraction of adult deaths due to HIV was 0.61 for men and 0.70 for women, with life expectancy estimated to be 34.3 years for males and 38.2 years for females. Each method for estimating adult mortality had limitations in terms of loss to follow-up (cohort study), under-ascertainment (household censuses), transparency of underlying processes (single question), and sensitivity to parameterization (mathematical model). However, these analyses make clear the advantages of longitudinal cohort data, which provide more complete ascertainment than household censuses, highlight possible inaccuracies in model assumptions, and allow direct quantification of the impact of HIV.