A practical guide to cross-cultural and multi-sited data collection in the biological and behavioural sciences.

Researchers in the biological and behavioural sciences are increasingly conducting collaborative, multi-sited projects to address how phenomena vary across ecologies. These types of projects, however, pose additional workflow challenges beyond those typically encountered in single-sited projects. Through specific attention to cross-cultural research projects, we highlight four key aspects of multi-sited projects that must be considered during the design phase to ensure success: (1) project and team management; (2) protocol and instrument development; (3) data management and documentation; and (4) equitable and collaborative practices. Our recommendations are supported by examples from our experiences collaborating on the Evolutionary Demography of Religion project, a mixed-methods project collecting data across five countries in collaboration with research partners in each host country. To existing discourse, we contribute new recommendations around team and project management, introduce practical recommendations for exploring the validity of instruments through qualitative techniques during piloting, highlight the importance of good documentation at all steps of the project, and demonstrate how data management workflows can be strengthened through open science practices. While this project was rooted in cross-cultural human behavioural ecology and evolutionary anthropology, lessons learned from this project are applicable to multi-sited research across the biological and behavioural sciences.

Genome-wide association study of leprosy in Malawi and Mali.

Leprosy is a chronic infection of the skin and peripheral nerves caused by Mycobacterium leprae. Despite recent improvements in disease control, leprosy remains an important cause of infectious disability globally. Large-scale genetic association studies in Chinese, Vietnamese and Indian populations have identified over 30 susceptibility loci for leprosy. There is a significant burden of leprosy in Africa, however it is uncertain whether the findings of published genetic association studies are generalizable to African populations. To address this, we conducted a genome-wide association study (GWAS) of leprosy in Malawian (327 cases, 436 controls) and Malian (247 cases, 368 controls) individuals. In that analysis, we replicated four risk loci previously reported in China, Vietnam and India; MHC Class I and II, LACC1 and SLC29A3. We further identified a novel leprosy susceptibility locus at 10q24 (rs2015583; combined p = 8.81 × 10-9; OR = 0.51 [95% CI 0.40 - 0.64]). Using publicly-available data we characterise regulatory activity at this locus, identifying ACTR1A as a candidate mediator of leprosy risk. This locus shows evidence of recent positive selection and demonstrates pleiotropy with established risk loci for inflammatory bowel disease and childhood-onset asthma. A shared genetic architecture for leprosy and inflammatory bowel disease has been previously described. We expand on this, strengthening the hypothesis that selection pressure driven by leprosy has shaped the evolution of autoimmune and atopic disease in modern populations. More broadly, our data highlights the importance of defining the genetic architecture of disease across genetically diverse populations, and that disease insights derived from GWAS in one population may not translate to all affected populations.

Discovery and fine-mapping of kidney function loci in first genome-wide association study in Africans.

Genome-wide association studies (GWAS) of kidney function have uncovered hundreds of loci, primarily in populations of European ancestry. We have undertaken the first continental African GWAS of estimated glomerular filtration rate (eGFR), a measure of kidney function used to define chronic kidney disease (CKD). We conducted GWAS of eGFR in 3288 East Africans from the Uganda General Population Cohort (GPC) and replicated in 8224 African Americans from the Women's Health Initiative. Loci attaining genome-wide significant evidence for association (P < 5 × 10-8) were followed up with Bayesian fine-mapping to localize potential causal variants. The predictive power of a genetic risk score (GRS) constructed from previously reported trans-ancestry eGFR lead single nucleotide polymorphism (SNPs) was evaluated in the Uganda GPC. We identified and validated two eGFR loci. At the glycine amidinotransferase (GATM) locus, the association signal (lead SNP rs2433603, P = 1.0 × 10-8) in the Uganda GPC GWAS was distinct from previously reported signals at this locus. At the haemoglobin beta (HBB) locus, the association signal (lead SNP rs141845179, P = 3.0 × 10-8) has been previously reported. The lead SNP at the HBB locus accounted for 88% of the posterior probability of causality after fine-mapping, but did not colocalise with kidney expression quantitative trait loci. The trans-ancestry GRS of eGFR was not significantly predictive into the Ugandan population. In the first GWAS of eGFR in continental Africa, we validated two previously reported loci at GATM and HBB. At the GATM locus, the association signal was distinct from that previously reported. These results demonstrate the value of performing GWAS in continental Africans, providing a rich genomic resource to larger consortia for further discovery and fine-mapping. The study emphasizes that additional large-scale efforts in Africa are warranted to gain further insight into the genetic architecture of CKD.

Multi-country collaborative citizen science projects to co-design cardiovascular disease prevention strategies and advocacy: findings from Ethiopia, Malawi, Rwanda, and South Africa.

BACKGROUND: Cardiovascular diseases (CVD) were responsible for 20.5 million annual deaths globally in 2021, with a disproportionally high burden in sub-Saharan Africa (SSA). There is growing evidence of the use of citizen science and co-design approaches in developing interventions in different fields, but less so in the context of CVD prevention interventions in SSA. This paper reports on the collaborative multi-country project that employed citizen science and a co-design approach to (i) explore CVD risk perceptions, (ii) develop tailored prevention strategies, and (iii) support advocacy in different low-income settings in SSA. METHODS: This is a participatory citizen science study with a co-design component. Data was collected from 205 participants aged 18 to 75 years in rural and urban communities in Malawi, Ethiopia and Rwanda, and urban South Africa. Fifty-one trained citizen scientists used a mobile app-based (EpiCollect) semi-structured survey questionnaire to collect data on CVD risk perceptions from participants purposively selected from two communities per country. Data collected per community included 100-150 photographs and 150-240 voice recordings on CVD risk perceptions, communication and health-seeking intentions. Thematic and comparative analysis were undertaken with the citizen scientists and the results were used to support citizen scientists-led stakeholder advocacy workshops. Findings are presented using bubble graphs based on weighted proportions of key risk factors indicated. RESULTS: Nearly three in every five of the participants interviewed reported having a relative with CVD. The main perceived causes of CVD in all communities were substance use, food-related factors, and litter, followed by physical inactivity, emotional factors, poverty, crime, and violence. The perceived positive factors for cardiovascular health were nutrition, physical activity, green space, and clean/peaceful communities. Multi-level stakeholders (45-84 persons/country) including key decision makers participated in advocacy workshops and supported the identification and prioritization of community-specific CVD prevention strategies and implementation actions. Citizen science-informed CVD risk screening and referral to care interventions were piloted in six communities in three countries with about 4795 adults screened and those at risk referred for care. Health sector stakeholders indicated their support for utilising a citizen-engaged approach in national NCDs prevention programmes. The citizen scientists were excited by the opportunity to lead research and advocacy. CONCLUSION: The collaborative engagement, participatory learning, and co-designing activities enhanced active engagement between citizen scientists, researchers, and stakeholders. This, in turn, provided context-specific insights on CVD prevention in the different SSA settings.

Handgrip strength as a simple measure for screening prediabetes and type 2 diabetes mellitus risk among adults in Malawi: A cross-sectional study.

OBJECTIVE: Handgrip strength, a simple measure of muscle strength, has been reported as a predictor of both prediabetes and type 2 diabetes mellitus (T2DM) and has been suggested for screening prediabetes and T2DM risk. This study examined the relationship of handgrip strength with prediabetes and T2DM among rural- and urban-dwelling adults in Malawi. METHODS: This was a cross-sectional study nested in a follow-up study of prediabetic and prehypertensive individuals identified during an extensive noncommunicable disease survey in Malawi. A total of 261 participants (women: 64%) were recruited. Univariate and multivariate binary logistic regression analyses were performed to examine the association of prediabetes and T2DM with relative handgrip strength. RESULTS: The mean (SD) age of the participants was 49.7 (13.6) years, and 54.0% were between the ages of 40 and 59 years. The mean (SD) absolute handgrip strength and relative handgrip strength were 28.8 (7.3) kg and 1.16 (0.40) kg/BMI, respectively, and the mean relative handgrip strength differed significantly (p < 0.001) by T2DM status. In unadjusted model, the odds ratio (OR) of prediabetes and T2DM per unit increase in relative handgrip strength was 0.12 [95% CI; 0.04-0.33]. The result remained significant after adjusting for age (continuous), sex, place of study, hypertension, dyslipidaemia and level of education (aOR [95% CI]; 0.19 [0.03-0.95]). CONCLUSIONS: The findings suggest that handgrip strength could be a relatively inexpensive, noninvasive measure for contributing to risk scores to identify high-risk individuals for screening diabetes in SSA.

Intervention development of a brief messaging intervention for a randomised controlled trial to improve diabetes treatment adherence in sub-Saharan Africa.

BACKGROUND: Brief messaging interventions, including Short Message Service (SMS) text-messages, delivered via mobile device platforms, show promise to support and improve treatment adherence. To understand how these interventions work, and to facilitate transparency, we need clear descriptions of the intervention development process. METHOD: We describe and reflect on the process of designing and pretesting an evidence- and theory-informed brief messaging intervention, to improve diabetes treatment adherence in sub-Saharan Africa. We followed the stepwise approach recommended by the Medical Research Council, United Kingdom (MRC UK) Framework for Development and Evaluation of Complex Health Interventions and guidance for mobile health intervention development. RESULTS: We used a four-phase, iterative approach that first generated primary and secondary evidence on the lived experience of diabetes, diabetes treatment services and mobile-phone use. Second, we designed a type 2 diabetes-specific, brief text-message library, building on our previous hypertension text-message library, as well as drawing on the primary and secondary data from phase one, and on expert opinion. We then mapped the brief text-messages onto behaviour change (COM-B) theoretical constructs. Third, we refined and finalised the newly developed brief text-message library through stakeholder consultation and translated it into three local languages. Finally, we piloted the intervention by pre-testing the automated delivery of the brief text-messages in the trial sites in Malawi and South Africa. The final SMS text Adherence suppoRt for people with type 2 diabetes (StAR2D) intervention was tested in a randomised controlled trial in Malawi and South Africa (trial registration: ISRCTN70768808 ). CONCLUSION: The complexity of public health interventions requires that we give more attention to intervention development work. Our documentation and reflection on the StAR2D intervention development process promotes transparency, replicability, assessment of intervention quality, and comparison with other studies.

Effect of Acute Illness on Contact Patterns, Malawi, 2017.

The way persons interact when ill could profoundly affect transmission of infectious agents. To obtain data on these patterns in Africa, we recorded self-reported named contacts and opportunities for casual contact in rural northern Malawi. We interviewed 384 patients and 257 caregivers about contacts over three 24-hour periods: day of the clinic visit for acute illness, the next day, and 2 weeks later when well. For participants of all ages, the number of adult contacts and the proportion using public transportation was higher on the day of the clinic visit than later when well. Compared with the day after the clinic visit, well participants (2 weeks later) named a mean of 0.4 extra contacts; the increase was larger for indoor or prolonged contacts. When well, participants were more likely to visit other houses and congregate settings. When ill, they had more visitors at home. These findings could help refine models of infection spread.

A comparison of the associations between adiposity and lipids in Malawi and the United Kingdom.

BACKGROUND: The prevalence of excess adiposity, as measured by elevated body mass index (BMI) and waist-hip ratio (WHR), is increasing in sub-Saharan African (SSA) populations. This could add a considerable burden of cardiovascular and metabolic diseases for which these populations are currently ill-prepared. Evidence from white, European origin populations shows that higher adiposity leads to an adverse lipid profile; whether these associations are similar in all SSA populations requires further exploration. This study compared the association of BMI and WHR with lipid profile in urban Malawi with a contemporary cohort with contrasting socioeconomic, demographic, and ethnic characteristics in the United Kingdom (UK). METHODS: We used data from 1248 adolescents (mean 18.7 years) and 2277 Malawian adults (mean 49.8 years), all urban-dwelling, and from 3201 adolescents (mean 17.8 years) and 6323 adults (mean 49.7 years) resident in the UK. Adiposity measures and fasting lipids were assessed in both settings, and the associations of BMI and WHR with total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C) and triglycerides (TG) were assessed by sex and age groups in both studies. RESULTS: Malawian female adults were more adipose and had more adverse lipid profiles than their UK counterparts. In contrast, Malawian adolescent and adult males were leaner and had more favourable lipid profiles than in the UK. Higher BMI and WHR were associated with increased TC, LDL-C and TG and reduced HDL-C in both settings. The magnitude of the associations of BMI and WHR with lipids was mostly similar or slightly weaker in the Malawian compared with the UK cohort in both adolescents and adults. One exception was the stronger association between increasing adiposity and elevated TC and LDL-C in Malawian compared to UK men. CONCLUSIONS: Malawian adult women have greater adiposity and more adverse lipid profiles compared with their UK counterparts. Similar associations of adiposity with adverse lipid profiles were observed for Malawian and UK adults in most age and sex groups studied. Sustained efforts are urgently needed to address the excess adiposity and adverse lipid profiles in Malawi to mitigate a future epidemic of cardio-metabolic disease among the poorest populations.

Prevalence and risk factors for chronic kidney disease of unknown cause in Malawi: a cross-sectional analysis in a rural and urban population.

BACKGROUND: An epidemic of chronic kidney disease of unknown cause (CKDu) is occurring in rural communities in tropical regions of low-and middle-income countries in South America and India. Little information is available from Southern African countries which have similar climatic and occupational characteristics to CKDu-endemic countries. We investigated whether CKDu is prevalent in Malawi and identified its potential risk factors in this setting. METHODS: We conducted a cross-sectional study from January-August 2018 collecting bio samples and anthropometric data in two Malawian populations. The sample comprised adults > 18 years (n = 821) without diabetes, hypertension, and proteinuria. Estimates of glomerular filtration rate (eGFR) were calculated using the CKD-EPI equation. Linear and logistic regression models were applied with potential risk factors, to estimate risk of reduced eGFR. RESULTS: The mean eGFR was 117.1 ± 16.0 ml/min per 1.73m2 and the mean participant age was 33.5 ± 12.7 years. The prevalence of eGFR< 60 was 0.2% (95% confidence interval (95% CI) 0.1, 0.9); the prevalence of eGFR< 90 was 5% (95% CI =3.2, 6.3). We observed a higher prevalence in the rural population (5% (3.6, 7.8)), versus urban (3% (1.4, 6.7)). Age and BMI were associated with reduced eGFR< 90 [Odds ratio (OR) (95%CI) =3.59 (2.58, 5.21) per ten-year increment]; [OR (95%CI) =2.01 (1.27, 3.43) per 5 kg/m2 increment] respectively. No increased risk of eGFR < 90 was observed for rural participants [OR (95%CI) =1.75 (0.50, 6.30)]. CONCLUSIONS: Reduced kidney function consistent with the definition of CKDu is not common in the areas of Malawi sampled, compared to that observed in other tropical or sub-tropical countries in Central America and South Asia. Reduced eGFR< 90 was related to age, BMI, and was more common in rural areas. These findings are important as they contradict some current hypothesis that CKDu is endemic across tropical and sub-tropical countries. This study has enabled standardized comparisons of impaired kidney function between and within tropical/subtropical regions of the world and will help form the basis for further etiological research, surveillance strategies, and the implementation and evaluation of interventions.

How to estimate glomerular filtration rate in sub-Saharan Africa: design and methods of the African Research into Kidney Diseases (ARK) study.

BACKGROUND: Chronic kidney disease (CKD) is a substantial cause of morbidity and mortality worldwide with disproportionate effects in sub-Saharan Africa (SSA). The optimal methods to estimate glomerular filtration rate (GFR) and therefore to determine the presence of CKD in SSA are uncertain. We plan to measure iohexol excretion to accurately determine GFR in Malawi, South Africa and Uganda. We will then assess the performance of existing equations to estimate GFR and determine whether a modified equation can better improve estimation of GFR in sub-Saharan Africa. METHODS: The African Research on Kidney Disease (ARK) study is a three-country study embedded within existing cohorts. We seek to enrol 3000 adults > 18 years based on baseline serum creatinine. Study procedures include questionnaires on socio-demographics and established risk factors for kidney disease along with anthropometry, body composition, blood pressure, blood chemistry and urine microscopy and albuminuria. We will measure GFR (mGFR) by plasma clearance of iohexol at 120, 180 and 240 min. We will compare eGFR determined by established equations with mGFR using Bland-Altman plots. We will use regression methods to estimate GFR and compare the newly derived model with existing equations. DISCUSSION: Through the ARK study, we aim to establish the optimal approach to estimate GFR in SSA. The study has the advantage of drawing participants from three countries, which will increase the applicability of the findings across the region. It is also embedded within established cohorts that have longitudinal information and serial measures that can be used to characterize kidney disease over a period of time. This will help to overcome the limitations of previous research, including small numbers, selected population sub-groups, and lack of data on proteinuria. The ARK collaboration provides an opportunity for close working partnerships across different centres, using standardized protocols and measurements, and shared bio-repositories. We plan to build on the collaboration for this study for future work on kidney disease in sub-Saharan Africa, and welcome additional partners from across the continent.

Assessing the costs and efficiency of HIV testing and treatment services in rural Malawi: implications for future "test and start" strategies.

BACKGROUND: Reaching the 90-90-90 targets requires efficient resource use to deliver HIV testing and treatment services. We investigated the costs and efficiency of HIV services in relation to HIV testing yield in rural Karonga District, Malawi. METHODS: Costs of HIV services were measured over 12 months to September 2017 in five health facilities, drawing on recognised health costing principles. Financial and economic costs were collected in Malawi Kwacha and United States Dollars (US$). Costs were calculated using a provider perspective to estimate average annual costs (2017 US$) per HIV testing episode, per HIV-positive case diagnosed, and per patient-year on antiretroviral therapy (ART), by facility. Costs were assessed in relation to scale of operation and facility-level annual HIV positivity rate. A one-way sensitivity analysis was undertaken to understand how staffing levels and the HIV positivity rate affected HIV testing costs. RESULTS: HIV testing episodes per day and per full-time equivalent HIV health worker averaged 3.3 (range 2.0 to 5.7). The HIV positivity rate averaged 2.4% (range 1.9 to 3.7%). The average cost per testing episode was US$2.85 (range US$1.95 to US$8.55), and the average cost per HIV diagnosis was US$116.35 (range US$77.42 to US$234.11), with the highest costs found in facilities with the lowest daily number of tests and lowest HIV yield respectively. The mean facility-level cost per patient-year on ART was approximately US$100 (range US$90.67 to US$115.42). ART drugs were the largest cost component averaging 71% (range 55 to 76%). The cost per patient-year of viral load tests averaged US$4.50 (range US$0.52 to US$7.00) with cost variation reflecting differences in the tests to ART patient ratio across facilities. CONCLUSION: Greater efficiencies in HIV service delivery are possible in Karonga through increasing daily testing episodes among existing health workers or allocating health workers to tasks in addition to testing. Costs per diagnosis will increase as yields decline, and therefore, encouraging targeted testing strategies that increase yield will be more efficient. Given the contribution of drug costs to per patient-year treatment costs, it is critical to preserve the life-span of first-line ART regimens, underlining the need for continuing adherence support and regular viral load monitoring.

Community-Level Knowledge and Perceptions of Stroke in Rural Malawi.

Background and Purpose- The incidence of stroke in Malawi is unknown but major risk factors, including hypertension, obesity, and diabetes mellitus, are highly prevalent. We sought to understand community-level knowledge about stroke. Methods- A population-based cross-sectional study was conducted in rural Malawi (2016-2017). Adults aged ≥15 years were randomly selected and interviewed about their knowledge and perceptions of stroke symptoms, risk factors, and prevention. Logistic regression was used to investigate sociodemographic factors associated with stroke knowledge. Results- Of 812 selected, 739 (91% response rate) were seen and consented; 57% were female, and the median age was 52.0 years. Knowledge of stroke was poor: 71% knew no (correct) risk factors. Witchcraft (20.6%) was mentioned as frequently as hypertension (19.8%) as a cause. Knowledge of stroke was greatest in the most educated and wealthy and lowest in men, the never married, and the youngest age group. HIV-positive individuals had higher knowledge of prevention (odds ratio, 2.91; 95% CI, 1.21-7.03) than HIV negative individuals. Conclusions- Knowledge about stroke is very low in this community, particularly among the least educated and poor. Programs to support prevention, early recognition, and timely hospital presentation after a stroke are needed.

Lusting, learning and lasting in school: sexual debut, school performance and dropout among adolescents in primary schools in Karonga district, northern Malawi.

Age at sexual debut is known to have implications for future sexual behaviours and health outcomes, including HIV infection, early pregnancy and maternal mortality, but may also influence educational outcomes. Longitudinal data on schooling and sexual behaviour from a demographic surveillance site in Karonga district, northern Malawi, were analysed for 3153 respondents between the ages of 12 and 25 years to examine the association between sexual debut and primary school dropout, and the role of prior school performance. Time to dropout was modelled using the Fine and Gray survival model to account for the competing event of primary school completion. To deal with the time-varying nature of age at sexual debut and school performance, models were fitted using landmark analyses. Sexual debut was found to be associated with a five-fold increase in rate of subsequent dropout for girls and a two-fold increase in dropout rate for boys (adjusted hazard ratio [aHR] of 5.27, CI 4.22-6.57, and 2.19, CI 1.77-2.7, respectively). For girls who were sexually active by age 16, only 16% ultimately completed primary schooling, compared with 70% aged 18 or older at sexual debut. Prior to sexual debut, girls had primary completion levels similar to those of boys. The association between sexual debut and school dropout could not be explained by prior poor school performance: the effect of sexual debut on dropout was as strong among those who were not behind in school as among those who were overage for their school grade. Girls who were sexually active were more likely to repeat a grade, with no effect being seen for boys. Pathways to dropout are complex and may differ for boys and girls. Interventions are needed to improve school progression so children complete primary school before sexual debut, and to improve sex education and contraception provision.

Identifying mixed Mycobacterium tuberculosis infections from whole genome sequence data.

BACKGROUND: Mixed, polyclonal Mycobacterium tuberculosis infection occurs in natural populations. Developing an effective method for detecting such cases is important in measuring the success of treatment and reconstruction of transmission between patients. Using whole genome sequence (WGS) data, we assess two methods for detecting mixed infection: (i) a combination of the number of heterozygous sites and the proportion of heterozygous sites to total SNPs, and (ii) Bayesian model-based clustering of allele frequencies from sequencing reads at heterozygous sites. RESULTS: In silico and in vitro artificially mixed and known pure M. tuberculosis samples were analysed to determine the specificity and sensitivity of each method. We found that both approaches were effective in distinguishing between pure strains and mixed infection where there was relatively high (> 10%) proportion of a minor strain in the mixture. A large dataset of clinical isolates (n = 1963) from the Karonga Prevention Study in Northern Malawi was tested to examine correlations with patient characteristics and outcomes with mixed infection. The frequency of mixed infection in the population was found to be around 10%, with an association with year of diagnosis, but no association with age, sex, HIV status or previous tuberculosis. CONCLUSIONS: Mixed Mycobacterium tuberculosis infection was identified in silico using whole genome sequence data. The methods presented here can be applied to population-wide analyses of tuberculosis to estimate the frequency of mixed infection, and to identify individual cases of mixed infections. These cases are important when considering the evolution and transmission of the disease, and in patient treatment.

A cleaner burning biomass-fuelled cookstove intervention to prevent pneumonia in children under 5 years old in rural Malawi (the Cooking and Pneumonia Study): a cluster randomised controlled trial.

BACKGROUND: WHO estimates exposure to air pollution from cooking with solid fuels is associated with over 4 million premature deaths worldwide every year including half a million children under the age of 5 years from pneumonia. We hypothesised that replacing open fires with cleaner burning biomass-fuelled cookstoves would reduce pneumonia incidence in young children. METHODS: We did a community-level open cluster randomised controlled trial to compare the effects of a cleaner burning biomass-fuelled cookstove intervention to continuation of open fire cooking on pneumonia in children living in two rural districts, Chikhwawa and Karonga, of Malawi. Clusters were randomly allocated to intervention and control groups using a computer-generated randomisation schedule with stratification by site, distance from health centre, and size of cluster. Within clusters, households with a child under the age of 4·5 years were eligible. Intervention households received two biomass-fuelled cookstoves and a solar panel. The primary outcome was WHO Integrated Management of Childhood Illness (IMCI)-defined pneumonia episodes in children under 5 years of age. Efficacy and safety analyses were by intention to treat. The trial is registered with ISRCTN, number ISRCTN59448623. FINDINGS: We enrolled 10 750 children from 8626 households across 150 clusters between Dec 9, 2013, and Feb 28, 2016. 10 543 children from 8470 households contributed 15 991 child-years of follow-up data to the intention-to-treat analysis. The IMCI pneumonia incidence rate in the intervention group was 15·76 (95% CI 14·89-16·63) per 100 child-years and in the control group 15·58 (95% CI 14·72-16·45) per 100 child-years, with an intervention versus control incidence rate ratio (IRR) of 1·01 (95% CI 0·91-1·13; p=0·80). Cooking-related serious adverse events (burns) were seen in 19 children; nine in the intervention and ten (one death) in the control group (IRR 0·91 [95% CI 0·37-2·23]; p=0·83). INTERPRETATION: We found no evidence that an intervention comprising cleaner burning biomass-fuelled cookstoves reduced the risk of pneumonia in young children in rural Malawi. Effective strategies to reduce the adverse health effects of household air pollution are needed. FUNDING: Medical Research Council, UK Department for International Development, and Wellcome Trust.

Challenges in the Estimation of the Annual Risk of Mycobacterium tuberculosis Infection in Children Aged Less Than 5 Years.

Accurate estimates of Mycobacterium tuberculosis infection in young children provide a critical indicator of ongoing community transmission of M. tuberculosis. Cross-reactions due to infection with environmental mycobacteria and/or bacille Calmette-Guérin (BCG) vaccination compromise the estimates derived from population-level tuberculin skin-test surveys using traditional cutoff methods. Newer statistical approaches are prone to failure of model convergence, especially in settings where the prevalence of M. tuberculosis infection is low and environmental sensitization is high. We conducted a tuberculin skin-test survey in 5,119 preschool children in the general population and among household contacts of tuberculosis cases in 2012-2014 in a district in northern Malawi where sensitization to environmental mycobacteria is common and almost all children are BCG-vaccinated. We compared different proposed methods of estimating M. tuberculosis prevalence, including a method described by Rust and Thomas more than 40 years ago. With the different methods, estimated prevalence in the general population was 0.7%-11.5% at ages <2 years and 0.8%-3.3% at ages 2-4 years. The Rust and Thomas method was the only method to give a lower estimate in the younger age group (0.7% vs 0.8%), suggesting that it was the only method that adjusted appropriately for the marked effect of BCG-attributable induration in the very young.

A blood RNA signature for tuberculosis disease risk: a prospective cohort study.

BACKGROUND: Identification of blood biomarkers that prospectively predict progression of Mycobacterium tuberculosis infection to tuberculosis disease might lead to interventions that combat the tuberculosis epidemic. We aimed to assess whether global gene expression measured in whole blood of healthy people allowed identification of prospective signatures of risk of active tuberculosis disease. METHODS: In this prospective cohort study, we followed up healthy, South African adolescents aged 12-18 years from the adolescent cohort study (ACS) who were infected with M tuberculosis for 2 years. We collected blood samples from study participants every 6 months and monitored the adolescents for progression to tuberculosis disease. A prospective signature of risk was derived from whole blood RNA sequencing data by comparing participants who developed active tuberculosis disease (progressors) with those who remained healthy (matched controls). After adaptation to multiplex quantitative real-time PCR (qRT-PCR), the signature was used to predict tuberculosis disease in untouched adolescent samples and in samples from independent cohorts of South African and Gambian adult progressors and controls. Participants of the independent cohorts were household contacts of adults with active pulmonary tuberculosis disease. FINDINGS: Between July 6, 2005, and April 23, 2007, we enrolled 6363 participants from the ACS study and 4466 from independent South African and Gambian cohorts. 46 progressors and 107 matched controls were identified in the ACS cohort. A 16 gene signature of risk was identified. The signature predicted tuberculosis progression with a sensitivity of 66·1% (95% CI 63·2-68·9) and a specificity of 80·6% (79·2-82·0) in the 12 months preceding tuberculosis diagnosis. The risk signature was validated in an untouched group of adolescents (p=0·018 for RNA sequencing and p=0·0095 for qRT-PCR) and in the independent South African and Gambian cohorts (p values <0·0001 by qRT-PCR) with a sensitivity of 53·7% (42·6-64·3) and a specificity of 82·8% (76·7-86) in the 12 months preceding tuberculosis. INTERPRETATION: The whole blood tuberculosis risk signature prospectively identified people at risk of developing active tuberculosis, opening the possibility for targeted intervention to prevent the disease. FUNDING: Bill & Melinda Gates Foundation, the National Institutes of Health, Aeras, the European Union, and the South African Medical Research Council.

Diagnosis of childhood tuberculosis and host RNA expression in Africa.

BACKGROUND: Improved diagnostic tests for tuberculosis in children are needed. We hypothesized that transcriptional signatures of host blood could be used to distinguish tuberculosis from other diseases in African children who either were or were not infected with the human immunodeficiency virus (HIV). METHODS: The study population comprised prospective cohorts of children who were undergoing evaluation for suspected tuberculosis in South Africa (655 children), Malawi (701 children), and Kenya (1599 children). Patients were assigned to groups according to whether the diagnosis was culture-confirmed tuberculosis, culture-negative tuberculosis, diseases other than tuberculosis, or latent tuberculosis infection. Diagnostic signatures distinguishing tuberculosis from other diseases and from latent tuberculosis infection were identified from genomewide analysis of RNA expression in host blood. RESULTS: We identified a 51-transcript signature distinguishing tuberculosis from other diseases in the South African and Malawian children (the discovery cohort). In the Kenyan children (the validation cohort), a risk score based on the signature for tuberculosis and for diseases other than tuberculosis showed a sensitivity of 82.9% (95% confidence interval [CI], 68.6 to 94.3) and a specificity of 83.6% (95% CI, 74.6 to 92.7) for the diagnosis of culture-confirmed tuberculosis. Among patients with cultures negative for Mycobacterium tuberculosis who were treated for tuberculosis (those with highly probable, probable, or possible cases of tuberculosis), the estimated sensitivity was 62.5 to 82.3%, 42.1 to 80.8%, and 35.3 to 79.6%, respectively, for different estimates of actual tuberculosis in the groups. In comparison, the sensitivity of the Xpert MTB/RIF assay for molecular detection of M. tuberculosis DNA in cases of culture-confirmed tuberculosis was 54.3% (95% CI, 37.1 to 68.6), and the sensitivity in highly probable, probable, or possible cases was an estimated 25.0 to 35.7%, 5.3 to 13.3%, and 0%, respectively; the specificity of the assay was 100%. CONCLUSIONS: RNA expression signatures provided data that helped distinguish tuberculosis from other diseases in African children with and those without HIV infection. (Funded by the European Union Action for Diseases of Poverty Program and others).