Understanding non-communicable diseases (NCD) and the role of infection in Africa: building a partnership to generate big data

Lead Research Organisation: London Sch of Hygiene and Trop Medicine
Department Name: Epidemiology and Population Health

Abstract

Non-communicable diseases" (NCD) are increasing in Africa. They are a diverse group of conditions, not directly caused by infections and are most often long-term and lifestyle-related, such as diabetes, high blood pressure, cancers, asthma and kidney disease. The reasons for the increase in these diseases are complex but include increasing urbanisation, overweight/obesity and lifestyle risk factors, such as smoking tobacco, high alcohol intake and reduced physical activity. The increase in NCD, is occurring alongside a high level of infectious disease (such as HIV and tuberculosis) and under- or mal-nutrition, which often exacerbates these conditions and leads to very poor health outcomes and shortens lives. In Africa, deaths from NCD in adults now exceed deaths caused by infectious diseases. To truly understand the severity and extent of these diseases requires investigations using detailed information from a general population, referred to as population-based research. Population-based research (as opposed to hospital-based research) is particularly important in Africa as relatively few people use formal health services, especially in the early stages of illness. The unique and valuable data collected in repeated cycles of population-based research help us to determine the amount and causes of disease; the risk (or chance) of death from these conditions; how these diseases develop in individuals; and how the disease changes over time (for example, age and sex of people most commonly affected). To understand NCDs in Africa better, and to help inform African policy makers and researchers, we propose an African NCD Longitudinal Data Alliance (ANDLA). ANDLA will bring together population-based research data from multiple partners working in existing study sites across Africa to create large, harmonised datasets that are robust and powerful enough to address important research questions that individual study sites would struggle to answer on their own. ANDLA will provide a dynamic and sustainable platform of population-based data collected over time (longitudinal-data). The data will be standardised and made available in an easily usable form, and will focus on NCD and their risk factors. It will be a unique large-scale data resource for understanding NCD in Africa now and in future decades, to inform and shape effective policy in populations with limited resources and a large and growing amount of these disease.

Technical Summary

Global burden of disease estimates suggest NCD are overtaking infectious disease as the leading cause of adult mortality in Africa. Obesity, diabetes, hypertension and associated lifestyle risk factors are rising in urban and rural settings as populations age and urbanise. HIV-related cancers and those of infectious aetiology including liver and cervix, are highly prevalent, representing a higher fraction of total NCD burden in Africa than elsewhere. In future, cancers such as breast, lung and colon, will be increasingly important as countries undergo demographic transition. This is a critical time point to understand the epidemiology, phenotypes and determinants of NCD.
NCD data from LMIC are predominantly from small studies using differing methods. Evidence from adequately powered, "big" data platforms, of population-level data, is urgently needed to understand the burden, distribution and determinants of NCD and the intersection with HIV and other important infectious diseases. We propose to set up an African Non-communicable Disease Longitudinal data Alliance (ANDLA) to address this gap: working with policy makers and experienced research partners; bringing together partners with NCD data collected in demographic surveillance sites to address important questions relevant to our setting. This will enable a greater understanding of the current status of chronic NCD, provide evidence for future prevention and management strategies and build a platform for multi-site intervention development, monitoring and evaluation and health systems research. Longitudinal data are key to understanding disease progression, trends and contribution of risk factors to mortality.
We will analyse markers of cardio-metabolic mortality risk; develop and validate risk scores; and conduct analyses of HIV-related cancer mortality. Scientific advisory groups will inform future work which may include asthma, chronic kidney disease, disabilities, injuries, epilepsy and mental health.

Planned Impact

The aim of the African Non-communicable Disease Longitudinal data Alliance (ANDLA) is to provide policy relevant data and analyses to national, regional and international bodies. It is anticipated that our policy briefs on burden and risk factors of NCD conditions will be used by Ministries, and regional and national bodies such as the NCD Alliance and WHO. These groups are active in determining policy, creating guidelines for managing NCD in LIMC and lobbying internationally for an effective global response. Overall we expect that findings from ANDLA will have a substantive influence on the decisions made by stakeholders to address the health care priorities of these populations.
Each partner institution has close relationships with local policy makers, through National Research Institutes, Public Health Institutes, other Ministry of Health departments and other Ministries involved in the multi-sectoral response to NCD. These have the local remit to lobby parliamentarians and lawmakers to respond appropriately with legislation (and the political will to enforce this) in the areas most likely to have impact on future NCD. Our capacity to provide high quality research evidence to these bodies will facilitate the development of evidence-informed legislation in important areas including alcohol, tobacco, sugar and town-planning (to enable increased physical exercise in urban communities). It will support Ministries of Education and Nutrition to take the lead in mainstreaming NCD prevention messages into Personal Social and Health Education programmes in schools and colleges, and to the wider community.
As part of the foundation phase, ANDLA partners will engage with national stakeholders, to solicit input for the Alliance's research agenda and to ensure that policy-relevant questions are addressed. Stakeholders involved in co-ordinating the multi-sectoral approach to NCD prevention and scale-up of response will be particularly important in refining research questions. Host country stakeholders will be invited to the presentation days on each selected topic. Our findings will be communicated through dissemination in national meetings and policy briefings. We anticipate that the Alliance output will be highly influential in both the international and national fora where our high quality evidence will help direct the urgently-needed health care response.

Publications

10 25 50
 
Description Short Abstract
In the era of ART, HIV-positive individuals remain at higher risk of mortality compared with their HIV-negative counterparts. Understanding the causes of this HIV attributable mortality is important for directing interventions to prevent such deaths. We use verbal autopsy data from three demographic surveillance sites in Eastern and Southern Africa, with cause of death assigned using the InSilicoVA model, to describe the association between HIV and mortality attributable to non-communicable diseases (NCDs) in three broad groupings: all NCDs, cancer and cardiovascular disease. Preliminary results from two East African study sites suggested that at least 10% of NCD mortality there was attributable to HIV. There was no evidence for an association between HIV and all NCD mortality in the South African study, but HIV was associated with a doubling in the rate of cancer mortality in this setting. We observe inconsistent results of the effect of HIV on cardiovascular disease mortality, likely due to the small number of deaths attributed to this cause and the difficulty in assigning such deaths using these tools. The apparent increase in the rate of NCD mortality with HIV infection in some settings highlights the importance of understanding other chronic conditions with HIV, including ART side-effects.

Detailed analyses of mortality associated with individual conditions (hypertension, diabetes, obesity) and multimorbidity have been conducted and are being prepared for publication.

Detailed comparative analyses of cardiovascular risk scores have been conducted and are being prepared for publication.
Exploitation Route Our cardiovascular risk score work will have direct policy and clinical relevance in subSaharan Africa.
Sectors Healthcare

 
Description Preliminary findings have been presented at policy meetings in Malawi and at meetings of the Collaboration for Evidence Based Health Care in Africa meetings.
First Year Of Impact 2019
Sector Healthcare
Impact Types Policy & public services

 
Description Multimorbidity in Africa - Increasing Understanding of the Patient Experience and Epidemiology (MAfricaEE)
Amount £199,515 (GBP)
Funding ID MR/T037849/1 
Organisation Medical Research Council (MRC) 
Sector Public
Country United Kingdom
Start 04/2020 
End 09/2021
 
Title Extension of database to include socio-economic data and data from an additional two partners 
Description As described before, now including multiple SES indices and data from Rakai Health Institute and Africa Health Research Institute 
Type Of Material Database/Collection of data 
Year Produced 2019 
Provided To Others? Yes  
Impact Analyses in progress 
 
Title Pooled longitudinal data on non-communicable diseases from partners in sub-Saharan Africa 
Description Our partnership aims to develop the capacity to share and harmonise data through partnerships, expertise and standardised data-specifications. We currently consist of six demographic surveillance sites in SSA (Ifakara, Tanzania; Kisesa Tanzania; Karonga, Malawi; APHRC, Kenya; Agincourt, South Africa; Masaka, Uganda) but are open to new partners with relevant longitudinal data. All of the existing partners currently work together on HIV-related pooled analyses through the ALPHA network (http://alpha.lshtm.ac.uk/) and a few of the partners work together on NCD related projects (for example, a multisite study of Chronic Kidney Disease and as part of the Regional Expert group on NCD). We are an enthusiastic and experienced group of predominantly Africa-based researchers. All partners listed already collect data on cardio-metabolic and cancer risk factors, HIV, anthropometry and verbal autopsy. ANDLA builds into the ALPHA network's existing framework of harmonised datasets for the analysis of longitudinal HIV data with the addition of a harmonised data specification on lifestyle factors (alcohol, tobacco), family health history, anthropometry (weight, height MUAC) and biological measures (lipids, blood glucose, blood pressure). Each of the partner sites has been working with ALPHA data scientists for many years harmonising data on residency episodes (entry (due to birth or entry to DSS), exit (due to death or departure from DSS), HIV test results (not all sites), ART use (not all sites), education, fertility and cause of death (from verbal autopsy and physician review (not all sites)). Overall, ANDLA aims to generate a large harmonised longitudinal NCD database (in many cases, linked to stored samples available for future screening for additional infections and biochemical abnormalities) to enable comparison of key risk factors and determinants of NCDs across different African populations and to investigate rarer risk factors and outcomes. Activities 
Type Of Material Database/Collection of data 
Year Produced 2018 
Provided To Others? No  
Impact Pooled analyses on cancer deaths and attributable risk of HIV. Cardiovascular risk score developments underway. 
URL http://www.andla.org
 
Title Publication of Malawi NCD survey data 
Description N=30,000 Detailed survey and laboratory result data Cardiometabolic conditions Lifestyle factors Physical measurements 
Type Of Material Database/Collection of data 
Year Produced 2019 
Provided To Others? Yes  
Impact Used in multiple high profile analyses 
 
Description Partnership supported by award, as part of creation of ANDLA. 
Organisation African Population and Health Research Center
Country Kenya 
Sector Academic/University 
PI Contribution Support in data management and analysis of longitudinal NCD data
Collaborator Contribution Staff time to develop research questions, attend workshops and participate in analyses. Contribution of large longitudinal datasets
Impact Analyses and publications in progress
Start Year 2017
 
Description Partnership supported by award, as part of creation of ANDLA. 
Organisation Ifakara Health Institute
Country Tanzania, United Republic of 
Sector Charity/Non Profit 
PI Contribution Support in data management and analysis of longitudinal NCD data
Collaborator Contribution Staff time to develop research questions, attend workshops and participate in analyses. Contribution of large longitudinal datasets
Impact Analyses and publications in progress
Start Year 2017
 
Description Partnership supported by award, as part of creation of ANDLA. 
Organisation Medical Research Council (MRC)
Department Medical Research Council (MRC), MRC/UVRI Unit, Uganda
Country Uganda 
Sector Academic/University 
PI Contribution Support in data management and analysis of longitudinal NCD data
Collaborator Contribution Staff time to develop research questions, attend workshops and participate in analyses. Contribution of large longitudinal datasets
Impact Analyses and publications in progress
Start Year 2017
 
Description Partnership supported by award, as part of creation of ANDLA. 
Organisation National Institute for Medical Research, Tanzania
Country Tanzania, United Republic of 
Sector Public 
PI Contribution Support in data management and analysis of longitudinal NCD data
Collaborator Contribution Staff time to develop research questions, attend workshops and participate in analyses. Contribution of large longitudinal datasets
Impact Analyses and publications in progress
Start Year 2017
 
Description Partnership supported by award, as part of creation of ANDLA. 
Organisation Wits Health Consortium
Country South Africa 
Sector Private 
PI Contribution Support in data management and analysis of longitudinal NCD data
Collaborator Contribution Staff time to develop research questions, attend workshops and participate in analyses. Contribution of large longitudinal datasets
Impact Analyses and publications in progress
Start Year 2017
 
Description Planned creation of African Population Cohort Consortium 2020 
Organisation African Academy of Sciences
Country Kenya 
Sector Charity/Non Profit 
PI Contribution Through cohort work in Malawi and ANDLA network, invited to be part of Africa-wide consortium developing plans for region-wide initiative, supported by multiple funders and with multiple partners
Collaborator Contribution Experience in cohort establishment and data harmomonisation
Impact not yet
Start Year 2020
 
Description Planned creation of African Population Cohort Consortium 2020 
Organisation Bill and Melinda Gates Foundation
Country United States 
Sector Charity/Non Profit 
PI Contribution Through cohort work in Malawi and ANDLA network, invited to be part of Africa-wide consortium developing plans for region-wide initiative, supported by multiple funders and with multiple partners
Collaborator Contribution Experience in cohort establishment and data harmomonisation
Impact not yet
Start Year 2020
 
Description Planned creation of African Population Cohort Consortium 2020 
Organisation Medical Research Council (MRC)
Country United Kingdom 
Sector Public 
PI Contribution Through cohort work in Malawi and ANDLA network, invited to be part of Africa-wide consortium developing plans for region-wide initiative, supported by multiple funders and with multiple partners
Collaborator Contribution Experience in cohort establishment and data harmomonisation
Impact not yet
Start Year 2020
 
Description Planned creation of African Population Cohort Consortium 2020 
Organisation National Institutes of Health (NIH)
Country United States 
Sector Public 
PI Contribution Through cohort work in Malawi and ANDLA network, invited to be part of Africa-wide consortium developing plans for region-wide initiative, supported by multiple funders and with multiple partners
Collaborator Contribution Experience in cohort establishment and data harmomonisation
Impact not yet
Start Year 2020
 
Description Planned creation of African Population Cohort Consortium 2020 
Organisation Wellcome Trust
Country United Kingdom 
Sector Charity/Non Profit 
PI Contribution Through cohort work in Malawi and ANDLA network, invited to be part of Africa-wide consortium developing plans for region-wide initiative, supported by multiple funders and with multiple partners
Collaborator Contribution Experience in cohort establishment and data harmomonisation
Impact not yet
Start Year 2020
 
Description Subsequent partnership and data sharing collaboration 
Organisation Africa Health Research Institute
Country South Africa 
Sector Charity/Non Profit 
PI Contribution Development of analysis plans, support in analysis and publication of pooled and site-specific data
Collaborator Contribution Scientific engagements, attendance at analysis workshops, contribution of longitudinal NCD datasets
Impact In progress
Start Year 2018
 
Description Subsequent partnership and datasharing collaboration 
Organisation University of the Witwatersrand
Country South Africa 
Sector Academic/University 
PI Contribution We are exploring pooling of data between ANDLA and AWI-gen projects.
Collaborator Contribution We are exploring pooling of data between ANDLA and AWI-gen projects.
Impact No output yet
Start Year 2017
 
Description Analysis workshop on Cardiovascular risk scores in sub Saharan Africa 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Policymakers/politicians
Results and Impact Partner sites are currently preparing their NCD specifications, according to a standardised template, for our first workshop in Lilongwe, March 2018 (19-23rd) on cardiovascular risk scores. Each of the partner sites is accustomed to participating in ALPHA analytical workshops, where sites have the opportunity to send an epidemiologist and analyst to explore their site specific data using programmes (in the form of Stata do files) prepared by ALPHA analysts prior to the workshop. ANDLA will follow the same model.
Expected outcomes from the first ANDLA workshop n Lilongwe
The focus of the first workshop is to:
? ?To undertake multisite analyses exploring the mortality associated with cardiometabolic risk factors and development of an Africa cardiometabolic risk assessment tools.
? ?Generate multi-site and site-specific publications
? ?Generate framework for next priority analyses
Our workshop analyses will explore:
? ?The distribution of CVD risk in each site, comparing existing CVD risk scores that use blood measures to those that do not use blood measures
? ?Validation of existing CVD risk scores using mortality data with and without blood measures
? ?Determine the combination of risk factors which best predicts mortality over 1 year, 3 years, 5 years and longer (depending on availability of follow-up across the sites)
? ?In sites with HIV data, explore the extent to which CVD risk score performance differs by HIV status
List of participants for the March 2018 Workshop
Name of Participant Partner site and country/institution and country
Xavier Gomez-Olive Agincourt South Africa
Chodziwadziwa Kabudula Agincourt South Africa
Gershim Asiki APHRC Kenya
Marylene Wamukoya APHRC Kenya
James Beard Kisesa Tanzania
Coleman Kishamawe Kisesa Tanzania
Ramadhan Abdul Ifakara Tanzania
Grace Kyule Ifakara Tanzania
Eveline Geubbels Ifakara Tanzania
Innocent Erone Masaka Uganda
Robert Newton Masaka Uganda
Mia Crampin Karonga Malawi
Albert Dube Karonga Malawi
Estelle Mclean Karonga Malawi
Wisdom Nakanga Karonga Malawi
Alison Price Karonga Malawi
Joise Prynn Karonga Malawi
Pablo Perel WHO Geneva
Clara Calvert LSHTM London
David Prieto LSHTM London
Keith Tomlin LSHTM London
Kufre Okop CEPHA+ South Africa
Year(s) Of Engagement Activity 2018
 
Description Kick off meeting for ANDLA, with ALPHA network. 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact Kick off ANDLA meeting with parter representatives - clinicians, scientists, data managers, analyst, Nairobi March 2017 - Planning of workshops, analyses, sharing of ideas and resources
Year(s) Of Engagement Activity 2017
URL https://andla.org
 
Description Presentation on ANDLA network to INDEPTH Network 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Other audiences
Results and Impact Presentation on ANDLA aims, objectives and outputs to
Year(s) Of Engagement Activity 2017
URL http://www.indepth-network.org/news-events/events/newsindepth-agm-2017-dubai-uae
 
Description Presentation on Cancer and HIV 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact Presentation at IUSSP Cape Town November 2017Short Abstract
In the era of ART, HIV-positive individuals remain at higher risk of mortality compared with their HIV-negative counterparts. Understanding the causes of this HIV attributable mortality is important for directing interventions to prevent such deaths. We use verbal autopsy data from three demographic surveillance sites in Eastern and Southern Africa, with cause of death assigned using the InSilicoVA model, to describe the association between HIV and mortality attributable to non-communicable diseases (NCDs) in three broad groupings: all NCDs, cancer and cardiovascular disease. Preliminary results from two East African study sites suggested that at least 10% of NCD mortality there was attributable to HIV. There was no evidence for an association between HIV and all NCD mortality in the South African study, but HIV was associated with a doubling in the rate of cancer mortality in this setting. We observe inconsistent results of the effect of HIV on cardiovascular disease mortality, likely due to the small number of deaths attributed to this cause and the difficulty in assigning such deaths using these tools. The apparent increase in the rate of NCD mortality with HIV infection in some settings highlights the importance of understanding other chronic conditions with HIV, including ART side-effects.
Year(s) Of Engagement Activity 2017
URL https://iussp.confex.com/iussp/ipc2017/meetingapp.cgi/Paper/4246