Centre for Population Studies
DFID Reproductive Health Work Programme
Without abstracts Ian TimaeusComparison of survey and model-based estimates of mortality and orphan numbers in sub-Saharan AfricaGrassly, N.;Lewis, J.;Mahy, M.;Walker, N.;Timaeus, I.M.(2004)Objectives: To validate UNAIDS/WHO projections of AIDS and other cause mortality in the countries of sub-Saharan Africa. Design: Comparison of UNAIDS/WHO model-based estimates of mortality and orphans due to AIDS and other causes with estimates from Demographic and Health Surveys (DHS), Multiple Indicator Cluster Surveys (MICS) and vital registration. Methods Regression of recent DHS and MICS estimates of orphan numbers in 36 countries against estimates based on UNAIDS/WHO projections of adult mortality. Comparison of estimates of mortality in Zimbabwe based on corrected vital registration data with UNAIDS/WHO projections by age and through time. Results: UNAIDS/WHO model-based estimates of the fraction of children whose mother has died are consistently higher than estimates based on surveys with on average an additional 1.9% of children orphaned (p < 0.001). This difference is independent of the level of orphanhood, HIV prevalence or survey type. Model-based estimates of the fraction of children whose father has died are also significantly higher (p = 0.002), although the difference disappears for higher prevalence countries. A comparison of UNAIDS/WHO projections with independent mortality estimates in Zimbabwe suggest the former significantly overestimate the number of AIDS and non-AIDS deaths in 1995. Conclusions: The discrepancy between projections and survey estimates of orphan numbers particularly in countries with low HIV prevalence suggests that adult mortality due to causes other than AIDS is overestimated in UNAIDS/WHO projections of orphan numbers. Under-enumeration of orphans in surveys and/or inaccurate fertility estimates may also contribute, although independent adult mortality data suggest this is unlikely to be a major factor. After correcting projections of orphans due to causes other than AIDS, UNAIDS/WHO projections and survey estimates of orphans due to all causes show remarkable consistency. In Zimbabwe, more recent surveillance data suggests the level of the epidemic has been over-estimated. These results validate the current approach of UNAIDS/WHO to estimating AIDS mortality and orphans. Population Studies (Forthcoming) Household composition and dynamics in KwaZulu Natal, South Africa: mirroring social reality in longitudinal data collectionHosegood, V.;Timaeus, I.M.(2004)This paper describes new approaches used by the Africa Centre Demographic Information System (ACDIS) to conceptualise households and collect data on them prospectively in a rural district of KwaZulu Natal, South Africa. These approaches include self-definition of household membership, the inclusion of both resident and non-resident household members in the study population, linkage of individuals to multiple households, and distinguishing between different types of household member. Individual and household data from the first year (2000) of ACDIS are presented. They reveal that 23% of the population are non-resident and that 3% of them are concurrently members of more than one household within the surveillance area. The ACDIS data are compared with the 1996 national census data for the same district to illustrate how differences in the way that households are defined affect statistics on household size and composition. According to ACDIS, only 28% of households are headed by women, compared with 55% in the census. This difference can be attributed largely to the exclusion of non-resident male heads from the census household rosters. The demographic surveillance system finds that 19% of all male household heads are non-resident. Unlike many demographic phenomena, the household is a social construct with no biological base. Households are defined by their members and enumerations of them should be grounded in those self-definitions. Improved understanding of rural South African demographic, economic and health issues requires data collection methods that capture the social reality of fluid household composition, high levels of individual and household mobility, non-resident household members, and multiple household memberships. Other large-scale household surveys and prospective community studies could collect data that do more justice to the complexity of households and permit the study of these social dynamics. African households: an Exploration of Census Data (Forthcoming) Levels and causes of adult mortality in rural South AfricaHosegood, V.;Vanneste, A-M.;Timaeus, I.M.(2004)Background: South Africa has experienced one of the world's most rapidly progressing HIV epidemics. Data on causes of death in rural communities are scarce and of poor quality. Methods: Demographic surveillance was used to enumerate the population and all adult deaths (n=1021) in 2000 in a study area in northern KwaZulu Natal. We conducted verbal autopsy interviews with the caregivers of those who died to identify the causes of adult death in the community. Results: Mortality in the study area rose sharply in the late-1990s. By 2000 the probability of dying between ages 15 and 60 was 58% for women and 75% for men. AIDS, with or without tuberculosis, is the leading cause of death in adulthood (48%). Injuries, most of which stem from road traffic accidents or violence, cause 20% of deaths of men aged 15-44. In the age group 60 years or more, non-communicable diseases account for 76% and 71% of deaths of women and men respectively. Discussion: This population has experienced a sudden and massive rise in adult mortality. This can be accounted for by AIDS deaths. Mortality from non-communicable disease and (among men) injuries is also high. Antenatal HIV seroprevalence continued to rise in rural KwaZulu Natal in the late 1990s, reaching 40% in some clinics in this district. Adult mortality will continue to rise during this decade unless effective treatment interventions are introduced. AIDS Vol: 18 Pages: 663-671 Impact on mortality of the AIDS epidemic in Northern Namibia assessed using parish registersNotkola, V.;Timaeus, I.M. Siiskonen, H.(2004)Background: HIV spread rapidly in Namibia in the 1990s. As in most parts of Africa, however, few data exist to document the impact of AIDS on mortality. Such data are important for improving knowledge of the epidemiology of HIV infection and for the development of programmes to mitigate the impact of the AIDS epidemic. Data and Methods: The study analyses death records from the registers of six Evangelical Lutheran parishes in northern Namibia. The dataset covers the experience between 1980 and 2000 of more than 4000 couples marrying between 1956 and 1995 and their children. We examine the trend in post-neonatal mortality and the age-standardised death rates at 20-64 of both men and women and use Poisson regression for rates to smooth the data and test for statistically significant discontinuities in the trend. The detailed age patterns of mortality in 1980-1993 and 1994-2000 are contrasted. Results: Post-neonatal mortality rose more than sevenfold between 1991 and 2000. Adult women's mortality rose more than threefold and men's mortality more than twofold between 1993 and 2000. The rise in men's mortality was concentrated at ages 30 to 54 and the rise in women's mortality at ages 25 to 49. Discussion: The pattern of mortality increase by age is consistent with the hypothesis that it is entirely due to AIDS. Other parish registers probably exist in Africa that could be exploited to study the impact of AIDS on mortality and demographic trends more generally. AIDS Vol: 18 Pages: 1061-1065 Mortality in South Africa 1985-1996: from apartheid to AIDSTimaeus, I.M.;Dorrington, R.E.;Nannan, N.;Bradshaw, D.(2004)Background: The HIV/AIDS epidemic is producing a substantial decline in life expectancy in many African countries. Nevertheless, the registration of deaths in Africa is inadequate and limited empirical evidence exists quantifying the impact on mortality of deaths from AIDS in either national populations or localised study sites. Such data are important for calibration of the models of the epidemic that are used for forecasting. They can also inform policies intended to mitigate the social and economic impact of AIDS. Data from annual surveys of women attending antenatal clinics in South Africa reveal that the proportion of them infected with HIV rose from less than 1 per cent to over 22 per cent during the course of the 1990s. The same period saw the final collapse of the Apartheid system, the establishment of a democracy in 1994 and an intensive period of restructuring of all public sector institutions and services. This study synthesises all available sources of data for South Africa to examine what has happened to adult mortality in the country over this eventful period in its history. Methods: The study analyses civil registration statistics and the retrospective data on mortality collected in the 1996 Census, the 1998 South Africa Demographic and Health Survey and other recent national surveys. The registration-based data series extends (in different forms) to May 2000. These different sources produce fairly consistent estimates of adult women's mortality in South Africa for the mid-1980s and for the mid-1990s but somewhat less consistent estimates for the intervening period and for men. Civil registration of deaths is incomplete for the black African population but has improved rapidly since the so-called independent homelands were reintegrated into the country. As the proportion of the population that is black varies by age, so does the completeness of the registration statistics. Thus, as race was not collected on death certificates for most of the 1990s, normal methods of evaluation of recent deaths data such as the Growth-Balance Equation are inappropriate and a more complex process of adjustment is required. This is explained in the paper. Various internal and external checks suggest it yields rather good estimates of adult mortality. The paper examines both the trend in overall adult mortality (45q15) in South Africa and changes in the detailed age pattern of mortality of each sex. Results: The mortality of adults in South Africa was more-or-less stagnant in the 1980s and early 1990s. By 1992, the probability of dying between ages 15 and 60 was about 22 per cent for women and 38 per cent for men. Until 1993, the age pattern of mortality of women in South Africa was broadly similar to that in Princeton West model life tables. However, this pattern has been transformed by the recent rise in mortality, which is concentrated at the ages at which AIDS deaths are most common. Women's mortality nearly tripled at ages 25-34 between 1993 and 1999. Men in South Africa had an unusual age pattern of mortality even in the 1980s. The death rates of men aged 20-54 were 1.5 to two times higher than one would expect based on Princeton West models fitted to data on women or on older men. The subsequent rise in mortality has intensified this pattern: the mortality of working age men has risen while that of elderly men has changed hardly at all. However, while the mortality of men aged 15-24 rose in the early 1990s, it has since fallen back and it is the death rates of men in their thirties that have since been rising most. By 1999-2000, the probability of dying between ages 15 and 60 had risen to about 31 per cent for women and 46 per cent for men. Discussion: The rapid rise in the mortality of women in South Africa since the early 1990s has been concentrated among women in their late twenties and thirties. This is consistent with what one would expect in a population afflicted by a generalized epidemic of HIV. It broadly fits the rise predicted by the ASSA model of the epidemic when it is fitted to HIV prevalence data from the surveys of antenatal clinics (Dorrington, 1998). One can infer from the age pattern of mortality among men, that they already had unusually high mortality from violence, injuries and other occupationally-related causes in 1985. The conflicts surrounding the final years of Apartheid saw an increase in deaths from violence, especially among young men. Men's mortality has risen more slowly than women's mortality since the 1994 elections. This trend reflects the net impact of two developments: falling mortality from violence, from which 15-24 year old men have benefited most, and rising mortality from AIDS, which has had most impact on the 30-44 age group. An epidemic of violent deaths is being replaced by an even worse AIDS epidemic. Burden of Disease Research Unit technical report (Forthcoming) Adult Mortality in Sub-Saharan Africa: Evidence from the Demographic and Health SurveysTimaeus, I.M.;Jasseh, M.(2004)Objective To investigate levels, trends and age patterns of adult mortality from AIDS and other causes in sub-Saharan Africa from retrospective data collected in the Demographic and Health Survey (DHS) programme of household surveys. Data The analysis uses the sibling histories collected from adult women in 24 nationally representative DHS surveys conducted in 22 mainland sub-Saharan African countries between 1992 and 2000. These provide information on almost 7 million sibling-years of exposure at ages 5-69 during the 9 years before each survey and on 35,560 deaths at these ages during that period. Second, data on the orphanhood of children aged 5-14 collected on the DHS household schedules are also used to estimate adult mortality Methods Deaths of siblings are modelled using Poisson regression. The mortality rates in each country are assumed to follow one log-linear trend until 4 years after adult HIV prevalence reached 1 per cent. The explanatory variables include an external standard that serves to smooth these observed rates. After this date, adult mortality in each country is allowed to follow a divergent trend. The age pattern of mortality increase in countries with generalized AIDS epidemics is estimated for each sex. To improve the precision of these estimates, we combine the data from different surveys. The orphanhood data are adjusted for biases arising from the reduced fertility of HIV positive women and the mortality of vertically-infected children and then analysed using standard methods. Results The sibling history estimates are generally confirmed by those from orphanhood. In 16 of the 18 countries in which the sibling histories cover the relevant period, they record a significant upward discontinuity in the trend in adult mortality about 4 years after the country developed a generalized AIDS epidemic. The exceptions are Ethiopia and Nigeria. The fastest rises in mortality have occurred in South Africa, Zambia, Zimbabwe, Uganda, and Cameroon. Adult mortality has risen relatively slowly in Kenya, Tanzania and has hardly changed at all in Burkina Faso. In the early years of the AIDS epidemic the rise in mortality was dispersed across the entire adult age range. As the epidemic matures, however, excess mortality becomes more concentrated among women aged 25-39 and men aged 30-44. These data suggest that about 20 per cent more men than women die of AIDS early in the epidemic, dropping to 10 per cent excess of male AIDS deaths as the epidemic matures. Conclusion The DHS sibling histories confirm that a sharp increase in adult mortality has occurred in Africa since HIV has become prevalent. The size and speed of the rise in mortality in different countries and evolving age pattern of mortality impact are broadly in line with predictions from epidemiological data and models. However, in contradiction to conventional wisdom, they suggest that more men than women are dying of AIDS in Africa. Some of the discrepant findings undoubtedly result from reporting errors in the sibling histories. Others merit further investigation. It is time for a systematic attempt to reconcile the demographic and epidemiological evidence concerning AIDS in Africa. Vol: Pages: Initial burden of disease estimates for South Africa, 2000Bradshaw, D.;Groenewald, P.;Laubscher, R.;Nannan, N.;Nojilana, B.;Norman, R.;Pieterse, D.;Schneider, M.;Bourne, D.E.;Timaeus, I.M.;Dorrington, R.;Johnson, L.(2003)BACKGROUND: This paper describes the first national burden of disease study for South Africa. The main focus is the burden due to premature mortality, i.e. years of life lost (YLLs). In addition, estimates of the burden contributed by morbidity, i.e. the years lived with disability (YLDs), are obtained to calculate disability-adjusted life years (DALYs); and the impact of AIDS on premature mortality in the year 2010 is assessed. METHOD: Owing to the rapid mortality transition and the lack of timely data, a modelling approach has been adopted. The total mortality for the year 2000 is estimated using a demographic and AIDS model. The non-AIDS cause-of-death profile is estimated using three sources of data: Statistics South Africa, the National Department of Home Affairs, and the National Injury Mortality Surveillance System. A ratio method is used to estimate the YLDs from the YLL estimates. RESULTS: The top single cause of mortality burden was HIV/AIDS followed by homicide, tuberculosis, road traffic accidents and diarrhoea. HIV/AIDS accounted for 38% of total YLLs, which is proportionately higher for females (47%) than for males (33%). Pre-transitional diseases, usually associated with poverty and underdevelopment, accounted for 25%, non-communicable diseases 21% and injuries 16% of YLLs. The DALY estimates highlight the fact that mortality alone underestimates the burden of disease, especially with regard to unintentional injuries, respiratory disease, and nervous system, mental and sense organ disorders. The impact of HIV/AIDS is expected to more than double the burden of premature mortality by the year 2010. CONCLUSION: This study has drawn together data from a range of sources to develop coherent estimates of premature mortality by cause. South Africa is experiencing a quadruple burden of disease comprising the pre-transitional diseases, the emerging chronic diseases, injuries, and HIV/AIDS. Unless interventions that reduce morbidity and delay morbidity become widely available, the burden due to HIV/AIDS can be expected to grow very rapidly in the next few years. An improved base of information is needed to assess the morbidity impact more accurately. South African Medical Journal Vol: 93 (9) Pages: 682-688 The South African fertility decline: Evidence from two censuses and a Demographic and Health SurveyMoultrie, T.;Timaeus, I.(2003)Inadequate data and apartheid policies have meant that, until recently, most demographers have not had the opportunity to investigate the level of, and trend in, the fertility of South African women. The 1996 South Africa Census and the 1998 Demographic and Health Survey provide the first widely available and nationally representative demographic data on South Africa since 1970. Using these data, this paper describes the South African fertility decline from 1955 to 1996. Having identified and adjusted for several errors in the 1996 Census data, the paper argues that total fertility at that time was 3.2 children per woman nationally, and 3.5 children per woman for African South Africans. These levels are lower than in any other sub-Saharan African country. We show also that fertility in South Africa has been falling since the 1960s. Thus, fertility transition predates the establishment of a family planning programme in the country in 1974. Population Studies Vol: 57 (3) Pages: 265-283 HIV/AIDS data in South AfricaDorrington, R.;D. Bourne;Bradshaw, D.;Laubscher, R.;Timaeus, I.M.(2002)Letter Lancet Vol: 360 (9340) Pages: 1177-1177 Trends in South African fertility between 1970 and 1998: an analysis of the 1996 Census and 1998 Demographic and Health SurveyMoultrie, T.A.;Timaeus, I.M.(2002)[an error occurred while processing this directive] Burden of Disease Research Unit Technical Report x+81 Some implications of HIV/AIDS on adult mortality in South AfricaR.E. Dorrington;D. Bourne;Bradshaw, D.;Laubscher, R.;Timaeus, I.M.(2002)AIDS Analysis Africa Vol: 12 (5) Pages: 3-6 Improved methods and assumptions for estimation of the HIV/AIDS epidemic and its impact: Recommendations of the UNAIDS Reference Group on Estimates, Modelling and ProjectionsUNAIDS Reference Group on Estimates, Modelling and Projections; (includes Zaba, B. & Timaeus, I)(2002)UNAIDS and WHO produce biannual country-specific estimates of HIV/AIDS and its impact. These estimates are based on methods and assumptions that reflect the best understanding of HIV epidemiology and demography at the time. Where significant advances are made in epidemiological and demographic research, the methods and assumptions must evolve to match these advances. UNAIDS established an Epidemiology Reference Group in 1999 to advise them and other organisations on HIV epidemiology and methods for making estimates and projections of HIV/AIDS. During the meeting of the reference group in 2001, four priority areas were identified where methods and assumptions should be reviewed and perhaps modified: a) models of the HIV epidemic, b) survival of adults with HIV-1 in low and middle income countries, c) survival of children with HIV-1 in low and middle income countries, and d) methods to estimate numbers of AIDS orphans. Research and literature reviews were carried out by Reference Group members and invited specialists, prior to meetings held during 2001-2. Recommendations reflecting the consensus of the meeting participants on the four priority areas were determined at each meeting. These recommendations were followed in UNAIDS and WHO development of country-specific estimates of HIV/AIDS and its impact for end of 2001. AIDS Vol: 16 (9) Pages: W1-14 Fertility decline in NepalCollumbien, M.;Timaeus, I.M.;Acharya, L.(2001)This paper investigates fertility in Nepal using the measures of parity progression proposed by Brass and Juarez (1983) to detect the onset of fertility decline. The analysis is based largely on the 1991 Nepal Fertility, Family Planning and Health Survey. Evaluation of the birth history data collected in this survey indicates that they are sufficiently reliable to determine fertility trends. The sample size allows analysis at sub-national level. Supporting evidence as to the pattern of decline is provided by the 1991 Census and by earlier surveys. A rapid fall in fertility occurred in Nepal in the 1980s. By 1991, total fertility had fallen from at least 6 children to a little under 5. A small part of this decline is probably a temporary period effect, stemming from a rise in women's ages at marriage. Most of the decline can be explained by limitation of family size. Parity progression ratios from the third birth onwards all decrease consistently from cohort to cohort. Measures of parity progression calculated from the 1991 data agree well with those calculated for equivalent cohorts using the 1976 Nepal Fertility Survey data. They indicate that progression between middle-order births probably began to decrease as long ago as the early 1970s. When looking at differentials in fertility levels and trends it is clear that fertility has fallen throughout Nepal. Analysis of the regional pattern of decline reveals a complex pattern. Broadly speaking, fertility decline has been more dramatic in the eastern half of Nepal, and also stronger in the Terai and hills than in the mountains. Fertility has fallen furthest in the urban areas and among women who ever attended school. The middle-order parity progression ratios of younger women have fallen dramatically. This development is likely to presage a steep decline in fertility in Nepal during the 1990s. The paper concludes with a short discussion of the implications of its findings for the interpretation of fertility transition in Nepal. It suggests that substantial demand to limit family size is well-established in Nepal. Fertility declined as access to modern methods of contraception spread. Fertility Transition in South Asia 99-120 The impact of HIV/AIDS on adult mortality in South AfricaDorrington, R.;D. Bourne;Bradshaw, D.;Laubscher, R.;Timaeus, I.M.(2001)[an error occurred while processing this directive] Burden of Disease Unit Technical Paper iii, 56 Age Patterns of Childhood Mortality in Sub-Saharan Africa: Recent evidence from the Demographic and Health SurveysJasseh, M.; Timaeus, I.M.(2001)This paper uses the birth histories collected in Demographic and Health Surveys to examine age patterns of mortality in childhood across sub-national regions of 24 sub-Saharan African countries. The death rates are modelled to determine the impact on the age pattern of mortality of location, date, level of mortality, and seasonality. Few African countries have a regionally homogeneous pattern of mortality in childhood and in two-thirds of the sub-national populations considered mortality at 1 to 4 is higher, relative to infancy, than in any family of Princeton model life tables. This distinctive pattern has become less marked with time. Extreme forms of it are now confined to eastern Mali, Niger and parts of northern Ghana. A more moderate form of the pattern is found in many other parts of West Africa and along a belt that stretches north-eastwards from central Zimbabwe and southern Zambia, through to northern Tanzania. Childhood mortality is rising in Zambia and Zimbabwe, especially at ages 1 and 5. This can be linked to the spread of HIV/AIDS. Presented at: IUSSP 24th General Population Conference (18-24th August, 2001) Fertility and living arrangements in South AfricaMoultrie, T.A.;Timaeus, I.M.(2001)This paper investigates fertility among African women in South Africa. Variation in fertility levels is influenced by such factors as rural or urban residence, and level of education and household income. Differential fertility between women of different language groups is accounted for largely by underlying socio-economic factors. A further factor investigated by this paper is the impact of household structure on fertility in South Africa using the 1993 South Africa Living Standards and Development Study. Household structure is examined from the perspective of women. We focus on whether women live with a husband, or with relatives of their parents' generation, or with relatives of their own generation. The analysis concentrates on women aged twenty or over who are already mothers. For those women, we hypothesise their living arrangements mediate between their socio-economic and cultural characteristics and the number of children that they have borne. Living with relatives from the previous generation is found to have a negligible net impact on the lifetime fertility of mothers. However, women who live with relatives from their own generation have borne about a fifth fewer children than other women of the same age after controlling for the impact of household income, the woman's schooling, regional differences and urban-rural residence. Women from Nguni language groups have relatively large families. While this largely reflects economic and educational disadvantage, it is also conditional on their living arrangements. Unmarried and separated mothers have about a fifth fewer children than married mothers of the same age. It is within the domestic context that the influence of other characteristics is transmitted into differences in numbers of children. Women's living arrangements have become more diverse over the past thirty years in South Africa. They both modify and mediate the effects of other factors on fertility. Journal of Southern African Studies Vol: 27 (2) Pages: 207-223 Adult mortality in the less developed worldTimaeus, I.M.(2001)The term adult mortality may describe either the mortality of everyone except children or the mortality of young and middle-aged adults but not the old. Statistics on adult mortality in the less developed world are unreliable but show that mortality among young and middle-aged adults fell markedly during the twentieth century. The non-communicable diseases as well as the communicable diseases contributed to this. Adult mortality is now lower in some poor countries than in parts of the industrialized world. Elsewhere progress has been slow. Moreover, adult mortality has risen rapidly since the 1980s in those parts of Africa that have developed a generalized epidemic of HIV/AIDS. Even where adult mortality is still falling, deaths of adults represent a growing proportion of all deaths. This reflects population aging and the relatively rapid drop in child mortality. In response, agencies and researchers have become more interested in developing and promoting appropriate health policy for adults in poor countries. Several key interventions, such as fiscal measures to reduce smoking, fall outside the area of responsibility of Ministries of Health. International Encyclopedia of the Social and Behavioral Sciences 1 142-147 Commentary: Mortality decline in SenegalTimaeus, I.M.(2001)Commentary International Journal of Epidemiology Vol: 30 (6) Pages: 1294-1295 Estimation of adult mortality from data on adult siblingsTimaeus, I.M.;Zaba, B.;Ali, M.M.;(2001)This research developed a simple regression-based method for estimating adult survivorship from data on the survival of a respondent's adult siblings. The data required to implement this method of estimation are straightforward and can be collected in a single-round household survey. Only two questions are required: 'How many living siblings do you have aged 15 years or more?' and 'How many of your siblings died after surviving to age 15 years or more'. Such data on the survival of adult siblings already exist for many populations as a by-product of efforts to measure maternal mortality by the sisterhood method. They can now be exploited to measure all-cause mortality. Our initial appraisal of the performance of this adult sibling method suggests that it has several advantages over existing methods for measuring adult mortality in countries with limited and defective data. First, the relationship between the proportion of adult siblings alive and life table survivorship is a close one that varies little between populations with different patterns of ages at childbirth and of mortality. If the data are accurate, they should yield rather precise estimates. Second, the estimates are efficient in terms of the statistical precision of the proportions because data will typically be collected on considerably more siblings than respondents. Third, the adult sibling method should perform better in populations with significant mortality from HIV/AIDS than existing indirect techniques for the estimation of adult mortality. On the one hand, any selection bias resulting from shared risks of infection should be small, while, on the other, the estimation procedure is relatively unaffected by the unusual age pattern of mortality that develops in populations where HIV infection is prevalent. Initial applications of the adult sibling method confirm that it is a useful new way of measuring adult mortality. Like any indirect method, the results do not yield the detailed information on time trends and age patterns of adult mortality that are provided by accurate direct data. On the other hand, collecting accurate direct data on adult mortality has proved a difficult challenge (Timaeus, 1991c). In particular, collecting sibling histories, as has been done in several Demographic and Health Surveys involves a lengthy and, therefore, expensive series of questions. It is unlikely to be successful in enquiries conducted with less well-trained field staff or less experienced professional staff. The simpler questions required to apply the indirect adult sibling method can be used more widely. Brass Tacks: Essays in Medical Demography 43-66 Adult mortality in Africa in the era of AIDSTimaeus, I.M.(1999)This paper examines the impact of HIV/AIDS on adult mortality in Africa. A detailed analysis is carried out of data collected from four HIV endemic countries - Malawi, Tanzania, Uganda and Zimbabwe - and a non-endemic country - Senegal. Using different data sources as well as various estimation techniques, the paper demonstrates the impact of HIV/AIDS on adult mortality by showing a clear distinction between the recent mortality experience of the HIV endemic and non-endemic countries. In Senegal (the non-endemic country), the sibling history data collected in a recent Demographic and Health Survey (DHS) provide evidence of continuing decline in adult mortality since the 1970s but in the four HIV endemic countries, the DHS data show that the initial downward trend in adult mortality has been reversed. Adult mortality has been on the increase since the middle 1980s. Because of their high value, particularly to the estimation of the impact on mortality of AIDS, the paper calls for the collection of sibling histories in all future fertility surveys in sub-Saharan Africa. The African Population in the 21st Century: Third African Population Conference, Durban South Africa, 6-10 December 1999 2 377-395 Mortality in sub-Saharan AfricaTimaeus, I.M.(1999)Mainly as a result of the Demographic and Health Surveys (DHS) programme, the availability of data for the study of mortality in sub-Saharan Africa improved during the 1990s. This paper updates earlier reviews of levels and trends in under-five mortality using data from 19 African censuses conducted during the period 1986 to 1992 and 20 DHS surveys conducted in the 1990s. It updates earlier reviews of the statistics on adults using the information on recent deaths and orphanhood that was collected in many of these recent censuses and the sibling history data collected in 11 of the DHS surveys. Africa remains a high mortality region. It also continues to be characterised by very diverse mortality conditions, making it difficult to summarize the situation in the continent. Mortality decline slowed or stopped in a number of African countries during the 1980s and adverse trends developed too early in several of them for AIDS deaths to be the full explanation. Even in its initial stages though, the HIV epidemic has had a dramatic impact on adult mortality in much of Eastern Africa. For example, the probability of dying between ages 15 and 60 in Zimbabwe in the mid-1980s was less than 20 per cent; ten years later it was more than 40 per cent. Such sharp rises in mortality emphasize the importance of improving the scope and quality of the data collected about adult mortality in Africa. Questions on recent deaths and orphanhood should be asked in all future censuses and sibling histories retained in the questionnaire for future fertility surveys. Health and Mortality: Issues of global concern 110-131 Impact of the HIV epidemic on mortality in sub-Saharan Africa: evidence from national surveys and censusesTimaeus, I.M.(1998)Objective: To measure recent trends in all-cause child and adult mortality in national populations in sub-Saharan Africa. Design: Secondary analysis of data collected in national household surveys and censuses. Methods: The index of infant and child mortality is the probability of dying before age 5. For adult mortality, it is the probability of dying between ages 15 and 60. Mortality trends are assessed in three ways: by comparison of data collected in the 1990s with those from the 1980s; using the retrospective reports of the survival of women's children and siblings collected by Demographic and Health Survey inquiries; and by comparing the latter estimates with estimates from data on orphanhood. Results: Under-five mortality is stagnant or rising in several African countries. In some of them, however, adverse trends developed too early in the 1980s to be attributable to HIV. In most countries, the three approaches to monitoring adult mortality yield consistent results. Adult death rates doubled or tripled between the 1980s and mid-1990s in Uganda, Zambia, and Zimbabwe. Mortality also rose substantially elsewhere in Eastern and Middle Africa but not in Western Africa. Increases in mortality are concentrated among young adults. In general, men are worst affected but in Uganda the rise in women's mortality is greater. Conclusions: Data can be collected in national household surveys and censuses to monitor the mortality impact of HIV in Africa. Such data have begun to document the differential impact of the epidemic. In those countries with data in which HIV became prevalent by the late-1980s, massive rises in adult mortality occurred by the mid-1990s. AIDS Vol: 12 (suppl 1) Pages: S15-27 Measurement of adult mortality in populations affected by AIDS: an assessment of the orphanhood methodTimaeus, I.M.;Nunn, A.J.(1997)This paper demonstrates that orphanhood data can be used to estimate adult women's mortality in populations experiencing an epidemic of AIDS. It develops both a correction for selection bias in reports of orphanhood and a revised procedure for estimating life table survivorship for use in populations with significant AIDS mortality. These new methods yield mortality estimates for a Ugandan population that are consistent with those obtained by prospective surveillance. Countries that lack effective death registration systems should ask about the survival of mothers in the census and surveys in order to monitor the impact on mortality of the AIDS epidemic. Health Transition Review Vol: 7 (suppl. 2) Pages: 23-43 Fertility Decline in ZimbabweMuhwava, W.;Timaeus, I.M.(1996)The extent to which fertility has declined in Zimbabwe has been hotly debated. This study set out to resolve this controversy by conducting a comprehensive analysis of all the fertility data available from national censuses and surveys. This included the first in-depth analysis of the 1994 Demographic and Health Survey data and the first combined analysis of all inquiries since 1969. As well as examining summary measures of total fertility, the study presents estimates of parity progression for each cohort interviewed in the two DHS surveys using the method proposed by Brass and Juarez to adjust for truncation bias. In addition, we check our fertility estimates against the Census enumerations by carrying out an intercensal population projection based on them. The results suggest that fertility fell slightly during the civil war of the 1970s but may have risen briefly immediately after independence. A sustained fertility decline then began in the mid-1980s that has continued into the 1990s. We agree with those that claim that the two DHS surveys in Zimbabwe underestimate current fertility. However, so do earlier inquiries. Thus, adjustment of the data leaves unaltered the conclusion that total period fertility has fallen by about a third. The total fertility rate in 1994 in Zimbabwe was about 4.7 children per woman. Ages at the onset of childbearing have been rising in Zimbabwe. This accounts for part of the decline in period fertility. Most of the decline, however, is due to decreases in parity progression. These began at high parities early in the 1980s but spread rapidly down to lower-order births. This contrasts with pattern observed in most Asian countries. By 1994, progression to fourth and higher-order births had fallen by more than 25 per cent. Fertility in Zimbabwe is incontrovertibly in transition. Fertility began to fall in Zimbabwe at about the time that family planning services were made available to the whole population and use of contraception has been the main proximate determinant of the decline. Equally, fertility preferences have fallen across the population. Supply and demand side factors acted in tandem to produce a fertility transition. Educated and working women led in contraceptive adoption and reproductive change, as did women in commercial and farming areas. However, contraceptive prevalence has now risen in all population subgroups, although it remains lower in Matabeleland than other parts of the country. The family planning programme relies heavily on the pill for both limiting and spacing. While met need has been rising, especially for limiting, substantial unmet need remains, especially for spacing. Nevertheless, access factors do not play a major role in determining reproductive behaviour. Stationary facilities are fairly uniformly available across the country and the positive impact of the community-based programme for the distribution of pills is clear. Centre for Population Studies Research Paper 96-1 iv, 53 The impact of HIV-1 infection on child health in sub-Saharan Africa: the burden on the health servicesWalraven, G.;Nicoll, A.;Njau, M.;Timaeus, I.(1996)HIV-1 infection in sub-Saharan Africa is resulting in substantial child mortality and an increase in the number of sick children presenting to health services. Many of the sick children come to health centres and hospitals, inflating numbers on paediatric wards. The presentations of childhood HIV-1 infection are many and varied so that HIV-1 infection is the new 'great imitator' of other conditions. Some other infections are more severe in HIV-1 infected children (specifically bacterial infections and measles). However, there is no clear evidence of consequent rises in the incidence of other childhood infections, though this is likely to be the case for tuberculosis. HIV-1 infected children with other infections often respond to locally available anti-microbials, but may require longer courses. Treatment is problematic because of the impossibility of distinguishing infected from uninfected children and because of shortages of medicines, which are being intensified further by the child and adult HIV-1 epidemics. Severe HIV disease in adult family members is adding to child morbidity and creating substantial orphanhood. Staff fear nosocomial infection, while simultaneously experiencing falling personal incomes and lacking resources to care for their patients. Substantial numbers of trained staff are being lost because of HIV-1 caused disease and death. The reality of HIV-1 infection through breast-feeding is not yet appreciated. When this becomes generally apparent, there is a risk that a lethal increase in bottle feeding could occur in some areas. Reduction in the number of new paediatric HIV-1 infections in sub-Saharan Africa can be achieved only by ameliorating the adult HIV-1 epidemic, reducing unnecessary blood transfusions and ensuring a safe blood supply. Tropical Medicine and International Health Vol: 1 (1) Pages: 3-14 Fertility change in Sub-Saharan Africa: a review of the evidenceCleland, J.;Onuoha, N.;Timaeus, I.(1994)The Onset of Fertility Transition in Sub-Saharan Africa 1-20 Environment and Health in Developing Countries: an Analysis of Intra-Urban Differentials Using Existing DataStephens, C.;Timaeus, I.;Akerman, M.;Avle, S.;Maia, P.B.;Campanario, P.;Doe, B.;Lush, L.;Tetteh, D.;Harpham, T.(1994) Why do women use contraception?Timaeus, I.M.;Moultrie, T.A.()Much of the demographic literature presumes that women use contraception to either limit family size or space births. This paper argues that women also use contraception to postpone pregnancy. Postponement is not synonymous with spacing but arises when women delay their next birth for reasons unrelated to the age of their youngest child. We demonstrate that postponement has a distinctive impact on the shape of birth interval distributions that differs from those of family size limitation, birth spacing, and a mixture of the two behaviours. Some populations, such as South Africa, have developed fertility regimes characterized by birth intervals far longer than can be accounted for by birth spacing. Postponement of births is most likely to become widespread in countries characterized by social unrest and dysfunctional institutions and those developed countries with particularly high opportunity costs of childbearing for women. Vol: Pages: |
