Gut microbiome profiling of a rural and urban South African cohort reveals biomarkers of a population in lifestyle transition

BACKGROUND

Comparisons of traditional hunter-gatherers and pre-agricultural communities in Africa with urban and suburban Western North American and European cohorts have clearly shown that diet, lifestyle and environment are associated with gut microbiome composition. Yet, little is known about the gut microbiome composition of most communities in the very diverse African continent. South Africa comprises a richly diverse ethnolinguistic population that is experiencing an ongoing epidemiological transition and concurrent spike in the prevalence of obesity, largely attributed to a shift towards more Westernized diets and increasingly inactive lifestyle practices. To characterize the microbiome of African adults living in more mainstream lifestyle settings and investigate associations between the microbiome and obesity, we conducted a pilot study, designed collaboratively with community leaders, in two South African cohorts representative of urban and transitioning rural populations. As the rate of overweight and obesity is particularly high in women, we collected single time-point stool samples from 170 HIV-negative women (51 at Soweto; 119 at Bushbuckridge), performed 16S rRNA gene sequencing on these samples and compared the data to concurrently collected anthropometric data.

RESULTS

We found the overall gut microbiome of our cohorts to be reflective of their ongoing epidemiological transition. Specifically, we find that geographical location was more important for sample clustering than lean/obese status and observed a relatively higher abundance of the Melainabacteria, Vampirovibrio, a predatory bacterium, in Bushbuckridge. Also, Prevotella, despite its generally high prevalence in the cohorts, showed an association with obesity. In comparisons with benchmarked datasets representative of non-Western populations, relatively higher abundance values were observed in our dataset for Barnesiella (log2fold change (FC) = 4.5), Alistipes (log2FC = 3.9), Bacteroides (log2FC = 4.2), Parabacteroides (log2FC = 3.1) and Treponema (log2FC = 1.6), with the exception of Prevotella (log2FC = - 4.7).

CONCLUSIONS

Altogether, this work identifies putative microbial features associated with host health in a historically understudied community undergoing an epidemiological transition. Furthermore, we note the crucial role of community engagement to the success of a study in an African setting, the importance of more population-specific studies to inform targeted interventions as well as present a basic foundation for future research.

83 A Deficit Accumulation Frailty Index Predicts Mortality in Older South Africans: Findings from the HAALSI Study

Introduction

Few studies have investigated frailty in older people in sub-Saharan Africa, yet such information is vital to prepare responses to rapid population ageing. We aimed to derive and test a cumulative deficit frailty index in a population of older people from rural South Africa.

Methods

We analysed data from the Health and Ageing in Africa: Longitudinal Studies of an INDEPTH Community (HAALSI) study, which enrolled participants aged 40 years and older nested within the Agincourt Health and Demographic Survey Site, South Africa. We created a 32-variable cumulative deficit frailty index using questionnaire (illnesses, symptoms and activities of daily living), physical performance and physiological indices, and blood test results. Each variable was dichotomised to 1 (deficit) or 0 (no deficit). The frailty index for each individual was calculated as the mean of all frailty variables. Frailty categories were defined using cut-offs from the UK electronic frailty index: 0-0.12 (non-frail), >0.12-0.24 (mild frailty), >0.24-0.36 (moderate frailty) and >0.36 (severe frailty). Cox proportional hazards models, both unadjusted and adjusted for age and sex, were fitted to test the association between frailty status and all-cause mortality.

Results

We analysed data from 3989 participants, mean age 61 years (SD 13); 2175 (54.5%) were female. The mean follow-up period was 17 months; 1464 (36.7%) were non-frail, 2059 (51.6%) had mild frailty, 402 (10.1%) had moderate frailty and 64 (1.6%) had severe frailty. A total of 135 (3.4%) died. Adjusted Cox models showed worse frailty category was associated with higher risk of death compared with non-frail individuals: hazard ratios 1.94 (95% CI 1.23, 3.07) for mild frailty, 3.25 (95% CI 1.86, 5.68) for moderate frailty, and 5.50 (95% CI 2.44, 12.40) for severe frailty.

Conclusions

Frailty measured by a cumulative deficits index is common and predicts mortality in a rural population of older South Africans.

Blood pressure-measuring devices in rural South Africa: an audit conducted by the SASPI team in the Agincourt field site

BACKGROUND

Cardiovascular disease is an important cause of morbidity and mortality in South Africa. The Southern Africa Stroke Prevention Initiative (SASPI) found a high prevalence of stroke in the rural Agincourt subdistrict, Limpopo province. Hypertension is the commonest vascular risk factor in our population and it is essential that primary care services be adequately equipped to detect and treat hypertension. The aim of this study was to assess the number, accuracy and working condition of blood pressure measuring devices (BMD) in the clinics that serve the field site, and to assess the clinic sisters' perceptions of the availability of antihypertensive medication and aspirin.

METHODS

In each of the clinics serving the site we assessed the BMDs and cuffs using the following criteria: general condition, bladder size, state of rubber components, operation of the inlet valve and control of valve operation. The legibility of the gauge, level and condition of the mercury, and the condition of the glass tube were checked when relevant. The performance of the BMD was then assessed both with the cuff used in the clinic and with a new functioning cuff, against an accurate mercury sphygmomanometer. By interviewing the clinic sister we could assess the availability of antihypertensive medication and aspirin, as well as the state of the drug delivery system.

RESULTS

All BMDs were mercury sphygmomanometers. Four clinics had one BMD each, one clinic had two, and one clinic had four. In one clinic the device was not functional at all until the study cuff was used. None of the clinics had spare cuffs and only one clinic had access to a large cuff. Nine out of 10 (90%) cuffs tested had unsatisfactory valve function, and none was of the size recommended by the guidelines. Although the condition of the mercury was only considered satisfactory in 40% of BMDs, once a new cuff had been fitted to the BMDs all of them were accurate to within 4 mmHg between 50 and 250 mmHg. Fifty per cent of clinic sisters felt they always had sufficient stock of hydrochlorothiazide and alpha-methyldopa, but the supply of more expensive medication was less reliable. Only one clinic always had sufficient aspirin.

CONCLUSION

Although none of the primary care clinics had fully functioning BMDs, almost all the defects related to malfunctioning and inappropriately sized cuffs, which would be inexpensive to repair or replace. A procedure for routine servicing or replacement of both BMDs and cuffs is needed, as well as optimisation of medication delivery to remote areas.

Undiagnosed pulmonary tuberculosis deaths in rural South Africa

This research explores death from pulmonary tuberculosis (PTB) using a verbal autopsy (VA) tool within the established Agincourt Health and Demographic Surveillance System site in South Africa's rural northeast. Previous work on active case finding in the area highlighted a modest burden of undiagnosed PTB in the community. This VA research confirms the existence of undiagnosed PTB deaths, with 13 (46%) of 28 PTB deaths among the permanent adult population (n = 38,251) going undetected by the health service. There was a median duration of coughing in the community of 16 weeks among these undiagnosed PTB deaths. As most undiagnosed cases present to the health service at some point during their illness, intervention strategies to support early diagnosis at this level can only be re-emphasised by this work.

Assessing health seeking behaviour among tuberculosis patients in rural South Africa

SETTING

South Africa's rural Northern Province.

OBJECTIVES

To examine patterns of health seeking behaviour among hospitalised tuberculosis patients.

DESIGN

Information on personal characteristics, health seeking behaviour and delays to presentation and hospitalisation was collected from hospitalised TB patients. Analysis of rates was used to investigate factors associated with delay.

RESULTS

Among 298 patients, median total delay to hospitalisation was 10 weeks, with patient delay contributing a greater proportion than service provider delay. Patients more often presented initially to public hospitals (41%) or clinics (31 %) than to spiritual/traditional healers (15%) or private GPs (13%). Total delay was shorter amongst those presenting to hospitals than those presenting to clinics (rate ratio 1.33, 95%CI 1.13-1.85), with a significantly smaller proportion of the total delay attributable to the health service provider (18% vs. 42%). Those exhibiting a conventional risk profile for TB (migrants, alcohol drinkers, history of TB) were diagnosed most quickly by health services, while women remained undiagnosed for longer.

CONCLUSION

Considerable delay exists between symptom onset and treatment initiation among pulmonary tuberculosis patients. While a substantial delay was attributable to late patient presentation, an important, preventable period of infectiousness was caused by the failure of recognised clinical services to diagnose tuberculosis among symptomatic individuals.

Active case finding: understanding the burden of tuberculosis in rural South Africa

SETTING

The Agincourt demographic and health surveillance site in South Africa's rural Northern Province.

OBJECTIVES

To accurately assess the true burden of tuberculosis in a rural sub-district with a known high prevalence of human immunodeficiency virus.

DESIGN

Data on hospital registrations of tuberculosis were combined with data from an ongoing demographic health and surveillance system to accurately describe the burden of tuberculosis in a well-defined community. Undiagnosed active cases of sputum-positive disease in the community were detected among chronic coughers identified by heads of household during a single-pass census interview.

RESULTS

The incidence of hospitalised tuberculosis among the permanently resident population (n = 56 566) was 212/100,000 person-years during 1999. The average point prevalence of detected tuberculosis (all forms) among patients aged over 10 years was 133/100,000, and 81/100,000 for sputum-positive pulmonary disease. This compares with a point prevalence of 16/100,000 cases of sputum-positive disease detected through active case finding.

CONCLUSION

For every nine cases of sputum positive pulmonary tuberculosis being treated at any one time, there are two cases of undiagnosed disease in the community. This study demonstrates a modest burden of undiagnosed tuberculosis among residents in a rural sub-district in South Africa.

Validation and application of verbal autopsies in a rural area of South Africa

OBJECTIVE

To validate the causes of death determined with a single verbal autopsy instrument covering all age groups in the Agincourt subdistrict of rural South Africa.

METHODS

Verbal autopsies (VAs) were conducted on all deaths recorded during annual demographic and health surveillance over a 3-year period (1992-95) in a population of about 63 000 people. Trained fieldworkers elicited signs and symptoms of the terminal illness from a close caregiver, using a comprehensive questionnaire written in the local language. Questionnaires were assessed blind by three clinicians who assigned a probable cause of death using a stepwise consensus process. Validation involved comparison of VA diagnoses with hospital reference diagnoses obtained for those who died in a district hospital; and calculation of sensitivity, specificity and positive predictive value (PPV) for children under 5 years, and adults 15 years and older.

RESULTS

A total of 127 hospital diagnoses satisfied the criteria for inclusion as reference diagnoses. For communicable diseases, sensitivity of VA diagnoses among children was 69%, specificity 96%, and PPV 90%; among adults the values were 89, 93 and 76%. Lower values were found for non-communicable diseases: 75, 91 and 86% among children; and 64, 50 and 80% among adults. Most misclassification occurred within the category itself. For deaths due to accidents or violence, sensitivity was 100%, specificity 97%, and PPV 80% among children; and 75, 98 and 60% among adults. Since causes of death were largely age-specific, few differences in sensitivity, specificity and PPV were found for adults and children. The frequency distribution of causes of death based on VAs closely approximated that of the hospital records used for validation.

CONCLUSION

VA findings need to be validated before they can be applied to district health planning. In Agincourt, a single verbal autopsy instrument provided a reasonable estimate of the frequency of causes of death among adults and children. Findings can be reliably used to inform local health planning and evaluation.

Health system reform and the role of field sites based upon demographic and health surveillance

Field sites for demographic and health surveillance have made well-recognized contributions to the evaluation of new or untested interventions, largely through efficacy trials involving new technologies or the delivery of selected services, e.g. vaccines, oral rehydration therapy and alternative contraceptive methods. Their role in health system reform, whether national or international, has, however, proved considerably more limited. The present article explores the characteristics and defining features of such field sites in low-income and middle-income countries and argues that many currently active sites have a largely untapped potential for contributing substantially to national and subnational health development. Since the populations covered by these sites often correspond with the boundaries of districts or subdistricts, the strategic use of information generated by demographic surveillance can inform the decentralization efforts of national and provincial health authorities. Among the areas of particular importance are the following: making population-based information available and providing an information resource; evaluating programmes and interventions; and developing applications to policy and practice. The question is posed as to whether their potential contribution to health system reform justifies arguing for adaptations to these field sites and expanded investment in them.

The Agincourt demographic and health study--site description, baseline findings and implications

OBJECTIVES

To present baseline results from first-phase demographic and health surveillance in the Agincourt field site, Bushbuckridge. To contrast findings with international data, and comment on their relevance to health development.

DESIGN

Multi-round, prospective community-based study. Baseline census, 1992.

SETTING

A subdistrict in South Africa's rural north-east, adjacent to the border with Mozambique.

SUBJECTS

Entire population of the Agincourt subdistrict.

OUTCOME MEASURES

Baseline variables for each resident included age, sex, months spent at home during 1991, mother alive/dead, highest educational standard achieved, and refugee status. A full demographic profile was constructed.

RESULTS

In 1992 the subdistrict contained 57,609 persons, 26.4% of whom were Mozambican, with a population density of 148 persons per km2. Forty-four per cent were under 15 years of age, and the dependency ratio was 93%. Fertility was declining, along with a moderate decline in adult female mortality. The approximate total fertility rate was 4.0; teenage parenting was common and almost 40% of 19-year-olds had at least one child. Over 60% of men and 14% of women in the 30-49-year age group were migrants, resulting in a high proportion of single-parent households. Despite improvements, most children experienced delays in reaching primary school, and less than half made the transition to secondary school. Educational levels among Mozambican children were lower than those of local children.

CONCLUSIONS

Agincourt contains a complex mix of communities comprising migrant workers, Mozambicans and a more stable permanent population. The area shares features with sub-Saharan Africa, although the mortality and fertility transitions have progressed further. Reliable, local information is essential for managing decentralised health systems.

The Agincourt field site--evolution and current status

OBJECTIVE

To describe the evolution and current status of the Agincourt field site, Bushbuckridge, in South Africa's rural north-east.

SETTING

A defined subdistrict, with a population of some 60,000 people including Mozambican refugees, in a former 'homeland' area with substantial labour migration.

APPROACH

Three phases are described: origins and establishment of the field site; a programme of health systems research underpinned by multi-round (prospective) demographic and health surveillance; and contributions to the University of the North's Dikgale field site.

COMMENT

Knowledge of trends in population health (e.g. mortality) is important when shaping the skill base and organisational framework of a district health system. There are valuable opportunities for collaborative research with field sites elsewhere in sub-Saharan Africa, and within South Africa itself. This will be facilitated by a common data model. A country-wide network of field sites, with surveillance capability, would complement the occasional national census and demographic and health survey, and strengthen South Africa's embryonic health information system.

Reversal in mortality trends: evidence from the Agincourt field site, South Africa, 1992-1995

OBJECTIVE

To examine changes in mortality in rural South Africa over the period 1992-1995 by age, sex and cause of death.

DESIGN

As with much of sub-Saharan Africa, South Africa lacks effective vital registration and information on mortality is lacking. The Agincourt demographic and health surveillance system was established to inform health policy and practice with regard to rural subdistrict populations.

METHODS

Prospective community-based study involving annual update of a household census with enquiry into all birth, death and migration events. All reported deaths (n = 1001) are the subject of a verbal autopsy.

RESULTS

An increasing trend in overall mortality relative to general population growth in the study area is apparent. There is evidence for a reversal in the previously declining trend in mortality among women 20-44 years. A comparison of 1992-1993 with 1994-1995 shows that most of the increase in mortality is concentrated in the younger adult (20-49 year) age group. AIDS and related diseases, particularly tuberculosis, appear primarily responsible. Injuries and violence (especially homicide) and circulatory disease are important, under-recognized causes of death, although their levels have remained constant over the period.

CONCLUSIONS

Mortality from AIDS and related diseases appears responsible for the probable reversal in mortality emerging in South Africa's rural northeast. Findings carry implications for the emerging system of decentralized health care.

Who dies from what? Determining cause of death in South Africa's rural north-east

Information on cause of death is essential for rational public health planning, yet mortality data in South Africa is limited. In the Agincourt subdistrict, verbal autopsies (VA) have been used to determine cause of death. A VA is conducted on all deaths recorded during annual demographic and health surveillance. Trained lay fieldworkers interview a close caregiver to elicit signs and symptoms of the terminal illness. Each questionnaire is reviewed by three medical practitioners blind to each other's assessment, who assign a 'probable cause of death' where possible. Of 1001 deaths of adults and children identified between 1992 and 1995, 932 VAs were completed. The profile of deaths reflects a mixed picture: the 'unfinished agenda' of communicable disease and malnutrition (diarrhoea and kwashiorkor predominantly) are responsible for over half of deaths in under-fives, accidents are prominent in the 5-14 age-group, while the 'emerging agenda' of violence and chronic degenerative disease (particularly circulatory disease) is pronounced among the middle-aged and elderly. This profile shows the social and demographic transition to be well underway within a rural, underdeveloped population. Validation of VA findings demonstrate that the cause of death profile derived from VA can be used with confidence for planning purposes. Findings of note include the high death rates from kwashiorkor and violence, emerging AIDS and pulmonary tuberculosis, and circulatory deaths in the middle-aged and young elderly. A deeper understanding of the causal factors underlying these critical health problems is needed to strengthen policy and better target interventions.

Stroke in rural South Africa--contributing to the little known about a big problem

OBJECTIVES

To describe the extent of mortality from cerebrovascular accident (CVA) in a rural South African population.

DESIGN

Annual demographic and health surveillance with verbal autopsy of all deaths, 1992-1995.

OUTCOME MEASURES

Stroke mortality rate by age and sex.

RESULTS

Stroke mortality increased with age and is higher in men than women over age 35. Proportionate mortality ratio from CVA: 10.3% of deaths in the 35-64-year age group.

CONCLUSION

Cerebrovascular disease is an important cause of death in South Africa's rural north-east. Community-based research is needed to inform policy and practice.

The Pholela Health Centre--the origins of community-oriented primary health care (COPC). An appreciation of the work of Sidney and Emily Kark

The health centre practice pioneered by Sidney and Emily Kark and their colleagues at Pholela during the 1940s was a forerunner of and direct contributor to what later emerged as 'the primary health care approach'. This article gives a detailed account of the context, work and methodologies used at the Pholela Health Centre, emphasising the development of concepts that are now well recognised and described as community-oriented primary health care (COPC). COPC remains highly relevant to health service development in South Africa today.

Developing district health systems in the rural Transvaal. Issues arising from the Tintswalo/Bushbuckridge experience

District health systems are increasingly acknowledged as a foundation for national health services based on primary health care. Initial efforts to institute a demonstration district health system in the Bushbuckridge area of the eastern Transvaal are described. These include efforts to overcome the organisational and administrative fragmentation caused by homeland and provincial boundaries. Close attention needs to be given to district-level health management, the complementary roles of district and regional health authorities, working relationships and accountability among professional staff from different disciplines, involvement of the community in a district health authority and the district health system as an element of local government.

Public health initiatives in South Africa in the 1940s and 1950s: lessons for a post-apartheid era

Inspiration drawn from South African public health initiatives in the 1940s played an important role in the development of the network of community and migrant health centers in the United States. The first such center at Pholela in Natal emphasized the need for a comprehensive (preventive and curative) service that based its practices on empirical data derived from epidemiological and anthropological research. In addition, community consultation preceded the introduction of new service or research initiatives. The Institute of Family and Community Health in Durban pioneered community-based multidisciplinary training and developed Pholela and other sites as centers for service, teaching, and research. Several important lessons for South African health professionals emerge from the Pholela experience. First, public health models of the past need to be reintroduced locally; second, the training of public health professionals needs to be upgraded and reoriented; third, appropriate research programs need to respond to community needs and address service demands; fourth, community involvement strategies need to be implemented early on; and fifth, funding sources for innovation in health service provision should be sought.

Planning professional education at schools of public health

OBJECTIVES

Professional education in public health should equip graduates with adequate knowledge and skills to manage diverse and complex problems. How best to address this challenge is widely debated. We describe the Harvard School of Public Health's self-evaluation and development of a practice-oriented program.

METHODS

As part of Harvard's schoolwide review of the master of public health (MPH), self-administered questionnaires were distributed to all MPH students, 1987 to 1989, and international and US alumni, 1979 to 1986. Extensive discussions were conducted with relevant student and faculty groups.

RESULTS

Survey results provided a basis for educational policy and curricular changes that culminated in a revised MPH that targets key areas of public health practice. Examples from the Harvard experience are provided.

CONCLUSIONS

Information derived from student and alumni surveys can be highly effective in the process of guiding curricular change at schools of public health. This should be coupled with a strategic approach to gain faculty support for proposed innovations. Ongoing monitoring and modification of the new curriculum is essential.

Vaccination coverage in Mhala and Elim in 1990--a health systems approach

This paper assesses the implementation of the 1986 Gazankulu policy on immunisation in 2 of the homeland's 6 health wards, Mhala and Elim. Vaccination coverage was estimated using the cluster sampling technique recommended by the Expanded Programme on Immunisation of the World Health Organisation. Vaccination coverage of children aged 12-23 months who have received eight valid doses (as stipulated in the programme) is estimated at 25% in Mhala and 53% in Elim (P < 0.001). While both figures can be substantially improved, the analysis concentrates on factors that help explain the differences between 2 districts that fall under the same central administration. As well as demographic differences, key factors include the presence of community-based health facilities and availability of staff (including programme leaders). The need for regional planners to have detailed and comprehensive knowledge of the health system at district level, and to make allowance for this, is put forward as vital to effective health planning. The failure of the Gazankulu policy to attain its objectives also reflects the often unrecognised consequences of the fragmentation of health services.