Generation Malawi: a family-based birth cohort investigating physical and mental health conditions in rural and urban Malawi – study protocol Authors: Wisdom P Nakanga, Robert C Stewart, Jullita Kenala Malava, Providence Nindi, Rebecca Nzawa-Soko, Innocent Nyanjagha, Angella Mainjeni, Thandile Gondwe, Eric Umar, Lucinda Manda-Taylor, Genesis Chorwe-Sungani, Melissa Gladstone, Marko Kerac, Angus MacBeth, Nadine Seward, Noel Patson, McNeil Ngongondo, Don Mathanga, Andrew M McIntosh, Amelia C Crampin, Generation Malawi Collaborator Group Published: 2026 ABSTRACT Abstract AbstractIntroductionMalawi is at a unique time point when infectious disease and undernutrition persist while modern diets, and changes in urbanisation, medical treatments and climate change, non-communicable long-term physical and mental health conditions are emerging. Malawi faces high socioeconomic hardships with 75% of families living in poverty with limited access to basic resources. With these demographic changes, very limited research has been undertaken into the determinants of physical and mental health-related trajectories over time, and in particular the role of prenatal and early life exposures in predetermining long-term conditions in later life.The study will establish a unique multigenerational birth cohort across contrasting rural and urban settings in Malawi to examine how pregnancy, early-life and intergenerational factors shape long-term mental and physical healthMethods and analysisGeneration Malawi is based within the open, health and demographic surveillance sites in rural (southern Karonga District) and urban (high-density sectors of Area 25, Lilongwe) Malawi. The study seeks to enrol up to 5000 pregnant women aged 16 years and above, and their spouses, during early pregnancy. Pregnant women and their spouses are followed up until the delivery date at which time their baby is recruited into the study. Children and their parents are then followed up at 1 week, 6 weeks, 4, 12, 24 and 36 months after delivery. At each point, we conduct detailed interviewer-led questionnaires, physical measurements and collection of bio-samples for immediate analysis and long-term storage.The data from this cohort will help us understand the causes of long-term ill health and, in the future, help develop effective interventions to prevent and promptly treat these conditions.Ethics and disseminationThis study protocol was approved by the National Health Science Research Committee in Malawi (number: 21/01/2653) and the Ethics Committee of the College of Medical, Veterinary and Life Sciences at the University of Glasgow, UK (protocol number: 200190183). The formative work for this study was granted ethical clearance by the College of Medicine Research Ethics Committee (protocol number P.11/19/2865). Results will be presented at relevant conferences and published in peer-reviewed journals.Keywords: Africa South of the Sahara, EPIDEMIOLOGY, Family, Antenatal, Longitudinal studies STRENGTHS AND LIMITATIONS OF THIS STUDY Longitudinal follow-up with detailed questionnaire data, physical assessment and biological sample collection from mothers, fathers and children enables investigation of biological, environmental, psychosocial determinants of physical and mental health and how this translates to offspring. Longitudinal follow-up with detailed questionnaire data, physical assessment and biological sample collection from mothers, fathers and children enables investigation of biological, environmental, psychosocial determinants of physical and mental health and how this translates to offspring. Inclusion of contrasting rural and urban populations enables assessment of the effects of urbanisation, socioeconomic adversity and lifestyle transitions on health outcomes. Inclusion of contrasting rural and urban populations enables assessment of the effects of urbanisation, socioeconomic adversity and lifestyle transitions on health outcomes. Loss to follow-up over time may introduce selection bias and reduce completeness of outcome data. Loss to follow-up over time may introduce selection bias and reduce completeness of outcome data. Lower participation and retention among fathers may limit the complete assessment of paternal and intergenerational influences. Lower participation and retention among fathers may limit the complete assessment of paternal and intergenerational influences. Introduction Malawi, similar to other low-income African countries (LI-AC), is at a unique time point when infectious diseases and undernutrition persist while non-communicable long-term physical and mental health conditions are also emerging. This situation is further compounded by rapid urbanisation and climate change, giving rise to increasingly complex health landscapes. In Malawi, 21.4% of the population are undernourished; 33.2% of children under 5 years of age are stunted (chronically undernourished) and 2.8% are wasted (acutely malnourished).1 At the same time 27% of adults have a high body mass index and 9% are obese.2 This coexistence of undernutrition and rising obesity highlights a dual burden of disease. Limited understanding of the determinants and consequences of long-term physical health conditions, such as cardiovascular disease and type 2 diabetes, as well as mental health conditions (eg, depression and anxiety) limits present opportunities for effective and timely treatment interventions. Early life exposures are increasingly recognised as important determinants of the development and outcomes of long-term conditions, alongside influences in mid and later life.3 Conceptualised in the ‘Developmental Origins of Health And Disease’ framework,4 exposure to undernutrition in utero and during infancy has been associated with increased risk of obesity, type 2 diabetes and cardiovascular disease.5 There is also evidence that exposures during the prenatal period and early infancy influence later risk of mental health conditions.6 For example, prenatal exposure to maternal depression may increase risk of mental health symptoms in childhood and adolescence, independent of shared genetic risk.7 Proposed biological pathways linking early life development to later life disease include fetal glucocorticoid overexposure,8 epigenetic modifications, inflammatory processes and alterations in the gut microbiome.9 Examining the roles of genetic (G) and environmental (E) effects and their interplay (G×E) is crucial for advancing mechanistic understanding.10 Despite the growing burden of long-term conditions in LI-AC there are major evidence gaps in understanding the impact of early environmental and genetic interplays on health and development and how this later affects life disease. Findings from high-income or other low- and middle-income countries (LMICs) cannot be assumed to apply to LI-AC where populations experience a high burden of overlapping nutritional, infectious, environmental and psychosocial exposures. To date, no study in LI-AC has integrated data on adult long-term physical and mental health across multiple interacting pathways (environmental, socioeconomic, psychological, nutritional, microbial and inflammatory determinants), from pre-conception through pregnancy and early childhood within a single birth cohort. The Generation Malawi (GM) study establishes a unique multigenerational family/birth cohort in two very diverse settings in rural and urban Malawi to study the longitudinal course and pregnancy, early-life and intergenerational effects for long-term mental and physical health conditions. The study is nested in health and demographic surveillance systems to facilitate long-term tracking and is linked to multiple other population studies. The study will generate detailed datasets and a biorepository; with both data and biosamples made accessible to external researchers. This paper describes the unique setting, methods, data and biosamples collected, and outlines the processes of data/sample access. Methodology Study design Generation Malawi is a multi-generational family/birth cohort nested within wider open population-based cohorts in rural and urban Malawi. Study setting The study is recruiting from a rural and an urban site in Malawi: the rural site (population 54,000) is an established Health and Demographic Surveillance Site (HDSS) in southern Karonga District (Northern Region, Malawi) with an area of 135 km2; the urban site (current population under surveillance 39 000) is a newly established HDSS in Area 25 of Lilongwe City (Central Region),11 with an area of less than 3 km2 (figure 1). These two geographically and socioeconomically distinct locations capture highly contrasting rural and urban health and exposures. The rural population in southern Karonga District predominantly consists of subsistence farmers, fisher people and traders. The densely populated and rapidly growing urban population in Lilongwe is socioeconomically mixed with seasonal casual workers, civil servants and professionals and those engaged in small and large businesses. Ethnicity in both sites is black African, almost exclusively Bantu; with the rural area predominantly Tumbuka and the urban population mixed, with Chewa the dominant tribal affiliation. MEIRU HDSS located in rural, Chilumba, Karonga district and urban, Area 25, Lilongwe. HDSS, Health and Demographic Surveillance Site; MEIRU, Malawi Epidemiology and Intervention Research Unit. Generation Malawi, the HDSS and other linked studies in both the rural and urban areas, are implemented by the Malawi Epidemiology and Intervention Research Unit (MEIRU), an independent unit formed as a partnership between the Malawi Ministry of Health, the Kamuzu University of Health Sciences, the London School of Hygiene and Tropical Medicine and the University of Glasgow. The University of Edinburgh leads the UK academic institutional engagement in Generation Malawi. The HDSS and associated population surveys of physical and mental long-term health conditions together form a project running parallel to Generation Malawi, known as Healthy Lives Malawi. Sample population All pregnant women (age 16 years or over) attending their initial antenatal clinic visit ‘booking’ at Chilumba Rural Hospital (CRH, rural site) or Area 25 Health Centre (HC, urban site) and who live within the geographical boundaries of the HDSS sites, are invited to participate, along with their spouses. Infants born to participating women are recruited to the study at birth. Main objective To establish a multigenerational family/birth cohort (nested within the Healthy Lives Malawi open, population-based cohorts) in rural (southern Karonga District) and urban (high-density sectors of Lilongwe Area 25) sites through which to study the longitudinal course, and pregnancy, early-life and intergenerational effects, for long-term mental and physical health conditions in Malawi. Specific objectives To measure the incidence, prevalence and prognosis of common mental disorders (depression, anxiety, post-traumatic stress disorder (PTSD), substance misuse) in women presenting to antenatal clinics, and their spouses, over the pregnancy/postnatal period in urban and rural populations. To measure the incidence, prevalence and prognosis of common mental disorders (depression, anxiety, post-traumatic stress disorder (PTSD), substance misuse) in women presenting to antenatal clinics, and their spouses, over the pregnancy/postnatal period in urban and rural populations. To identify associations of common mental disorders (depression, anxiety, PTSD, substance misuse) in women presenting to antenatal clinics, and their spouses, over the pregnancy/postnatal period in urban and rural populations, with socioeconomic status (SES), stressful life events (including intimate partner violence (IPV)), nutritional status and physical health. To identify associations of common mental disorders (depression, anxiety, PTSD, substance misuse) in women presenting to antenatal clinics, and their spouses, over the pregnancy/postnatal period in urban and rural populations, with socioeconomic status (SES), stressful life events (including intimate partner violence (IPV)), nutritional status and physical health. To describe the longitudinal patterns of growth, nutrition and development from birth in infants born to these women in rural and urban populations (including the prevalence of stunting, underweight and severe developmental delay/disability/congenital abnormality). To describe the longitudinal patterns of growth, nutrition and development from birth in infants born to these women in rural and urban populations (including the prevalence of stunting, underweight and severe developmental delay/disability/congenital abnormality). To determine the impact of parental mental health on pregnancy and birth outcomes and early child growth and development in urban and rural populations. To determine the impact of parental mental health on pregnancy and birth outcomes and early child growth and development in urban and rural populations. To determine the effects of covariates (as above) on any association between parental mental health and pregnancy and birth outcomes and early child growth and development in urban and rural populations. To determine the effects of covariates (as above) on any association between parental mental health and pregnancy and birth outcomes and early child growth and development in urban and rural populations. To collect a range of biosamples (including blood, mucosal swabs, faeces, urine, hair and breastmilk) from parents and offspring for storage in a biorepository to enable study of the role of genomics, epigenetics, microbiome, stress hormones, inflammation, metabolomics in trajectories and pregnancy, early-life and intergenerational effects, of mental and physical health conditions. To collect a range of biosamples (including blood, mucosal swabs, faeces, urine, hair and breastmilk) from parents and offspring for storage in a biorepository to enable study of the role of genomics, epigenetics, microbiome, stress hormones, inflammation, metabolomics in trajectories and pregnancy, early-life and intergenerational effects, of mental and physical health conditions. To provide a platform for Malawian researchers from MEIRU and the wider research community to extend follow-up of the families and nest substudies further exploring the longitudinal course, and pregnancy, early-life and intergenerational effects, for long-term mental and physical health conditions. To provide a platform for Malawian researchers from MEIRU and the wider research community to extend follow-up of the families and nest substudies further exploring the longitudinal course, and pregnancy, early-life and intergenerational effects, for long-term mental and physical health conditions. Sample size determination We plan to recruit 5000 pregnant women attending their first antenatal clinic (ANC) visit along with their spouses. Accounting for an expected 15% loss before delivery (including 4% multiple births), the resulting birth cohort will be around 4500 infants. The goal is to develop a resource usable for examining various conditions and subgroups; thus, a larger sample enhances future applications. This size is sufficient to accurately estimate disease burden and effects across the entire population and key subgroups such as sex and rural/urban areas. The study has over 90% power to detect an association between depression (prevalence 10%) and a risk factor with an OR of 1.2 in 5000 mothers. Study period Formative activities for the study (see below) commenced in August 2019. Enrolment in the rural site study began in February 2022 and in the urban site in November 2022. Rural recruitment paused in December 2023 but restarted in 2025. The anticipated end of participant enrolment is December 2027. Between 2020 and 2024, progress of the study was severely disrupted by the COVID-19 pandemic, economic challenges facing Malawi (eg, inflation, disrupted fuel availability, foreign exchange restrictions) and other health (eg, cholera outbreaks) and environmental (eg, cyclones) emergencies, with some of these constituting ongoing challenges in Malawi. Formative work Patient and public involvement Before study implementation, a national launch of the programme was conducted to ensure engagement at all levels of the health service, the research community, district authorities, town councils and traditional authorities. Residents were engaged at both an individual and community level in all aspects of the programme. These interactions have been published elsewhere.12 13 Development of measures We translated, adapted and piloted a core battery of self-report, interview and observational measures for adult mental and physical health, psychosocial risk factors and child health, growth and development in the two predominant languages spoken in the study sites (Chichewa in Lilongwe and Chitumbuka in Karonga). This process involved translation, expert consensus meetings, focus group discussions and piloting to establish the measures’ acceptability, reliability and validity. We conducted criterion validation of the mental health measures (9-item Patient Health Questionnaire (PHQ-9), 7-item Generalised Anxiety Disorder Scale (GAD-7), Primary Care PTSD Screen for Diagnostic and Statistical Manual of Mental Disorders (DSM)-5 (PC-PTSD-5)) against gold standard diagnostic interviews (Manuscript under review). Establishing clinics and referral pathways MEIRU has been running non-communicable disease (NCD) clinics for managing diabetes and hypertension, embedded within health facilities at the two study sites (Area 25 HC in Lilongwe and CRH in Karonga) since 2013. Referral pathways to these NCD clinics were integrated into the Generation Malawi procedures. We also set up new mental health clinics at the two sites and established referral pathways for participants identified as having possible mental health or substance use conditions, suicide risk or risk of harm through IPV. The mental health clinics provide assessment, diagnosis, signposting and treatment including ‘Friendship Bench’ Problem Solving Therapy,14 counselling for alcohol and substance use problems and prescription of psychotropic medication. When needed, both the NCD and mental health clinics refer to secondary/tertiary care in nearby government hospitals. Protocols for identifying and referring children identified as requiring medical assessment (eg, any anthropometry indicating severe malnutrition, acute illness or developmental disabilities) were also established. Study procedures Study contacts Participants are assessed for eligibility (table 1), recruited and have baseline assessment in the second trimester of pregnancy (or in third trimester if ‘late booker’ for antenatal care). Subsequent study contacts take place in third trimester, at delivery (when infants are recruited to the study) and postnatally at 1 week, 6 weeks, 16 weeks, 12 months, 24 months and 36 months. The schedule of study contacts is shown in table 2. Table 1Inclusion and exclusion criteriaInclusion criteria Age 16 or over.Ultrasound confirmed pregnancy.Resident of Lilongwe Area 25 or southern Karonga HDSS catchment areas.Signed informed consent (for 16–17 year olds who do not meet criteria for being an emancipated minor, consent of parent/guardian required). Exclusion criteria Not fluent in Chichewa, Chitumbuka or English.Not planning to remain living in the HDSS areas for at least the next 6 months.Too physically or mentally unwell to give informed consent. HDSS, Health and Demographic Surveillance Site. Inclusion and exclusion criteria Age 16 or over. Age 16 or over. Ultrasound confirmed pregnancy. Ultrasound confirmed pregnancy. Resident of Lilongwe Area 25 or southern Karonga HDSS catchment areas. Resident of Lilongwe Area 25 or southern Karonga HDSS catchment areas. Signed informed consent (for 16–17 year olds who do not meet criteria for being an emancipated minor, consent of parent/guardian required). Signed informed consent (for 16–17 year olds who do not meet criteria for being an emancipated minor, consent of parent/guardian required). Not fluent in Chichewa, Chitumbuka or English. Not fluent in Chichewa, Chitumbuka or English. Not planning to remain living in the HDSS areas for at least the next 6 months. Not planning to remain living in the HDSS areas for at least the next 6 months. Too physically or mentally unwell to give informed consent. Too physically or mentally unwell to give informed consent. HDSS, Health and Demographic Surveillance Site. Table 2Summary of contactsAntenatal contactsPostnatal contactsEligibilityBaseline3rd trimesterDeliveryWeek 1Week 6Week 16Month 12Month 24Month 36Contact timingFirst ANC visit (at which fetal USS done)Soon (usually <1 week) after eligibility contact34–39 weeks gestationWithin 24 hours of delivery.At baby check visit (7–9 days postnatal)At 6-week vaccinationBetween 16 and 24 weeks postnatalBetween 12 and 15 months postnatalBetween 24 and 30 months postnatalBetween 36 and 42 months postnatalContact placeFacility - ANC(Eligibility for spouses may be done at baseline home visit instead if not present at ANC)Home (may be seen at place of work or elsewhere if needed)Facility - ANCFacility - place of deliveryFacility - postnatal check clinicFacility - vaccination clinicHome (may be seen at place of work or elsewhere if needed)HomeHome (may be seen at place of work or elsewhere if needed)Home (may be seen at place of work or elsewhere if needed)Index womenFetal USS (routine)Eligibility Agreement to home visitEnrolmentConsent interviewAnthropometry BPBiosamplesInterviewAnthropometry BPBiosamplesDelivery recordInterviewAnthropometry BPBiosamplesInterview BiosamplesInterviewAnthropometry BPInterview Anthropometry BPBiosamplesInterview Anthropometry BPBiosamplesSpouseIf present - eligibility, agreement to home visitEligibility, enrolment/consentInterviewAnthropometry BPBiosamplesInterview Anthropometry BPBiosamplesInterview Anthropometry BPBiosamplesInterview Anthropometry BPBiosamplesChildEnrolmentMaternal interviewAnthropometryBody compositionBiosamplesMaternal interview Anthropometry Body composition BiosamplesMaternal interview Anthropometry Body composition BiosamplesMaternal interview AnthropometryBody compositionBiosamplesMaternal interview Anthropometry BiosamplesMaternal interview AnthropometryBiosamplesMaternal interview AnthropometryBiosamplesANC, antenatal clinic; BP, blood pressure; USS, ultrasound scan. Summary of contacts ANC, antenatal clinic; BP, blood pressure; USS, ultrasound scan. Eligibility assessment at antenatal clinic Pregnant women attending for their first antenatal appointment (‘booking’) are identified at the Ministry of Health (MOH) Antenatal Clinics in Area 25 HC and CRH. Research staff conduct a short group briefing about the study in the ANC waiting area. The women then proceed through routine MOH ANC procedures including fetal ultrasound scan (USS) dating. (USS scanning was introduced in the clinics as part of the National Institute for Health Research (Global Health Research (GHR) Project: 17/63/08 DIPLOMATIC collaboration).15 Women with a confirmed pregnancy and USS determined expected delivery date (EDD) are then seen individually by research staff in a private room and assessed for eligibility using a tablet-based questionnaire programmed using Open Data Kit. Any spouses attending alongside the women are also assessed for eligibility. Eligible individuals who are interested to take part are given a scheduled appointment for a baseline home visit, to take place usually within a week. Women who wish to discuss the study with their spouse at home are given the opportunity to do so, and visit arrangements are subsequently made by phone. Women can choose not to involve their spouse and still participate in the study. Current enrolment of the household in the HDSS is checked and details of potential participants whose household is not yet registered in the HDSS but within the geographical area covered by the HDSS are referred to the HDSS team who make a registration visit (usually within 2 days). Baseline contact at home: consenting, enrolment and baseline data collection (woman and spouse) Prearranged home visits are made by research fieldworkers working in pairs. Information sheets are read out independently to the woman and spouse (if present at the home) in the language of their choice (English, Chichewa or Chitumbuka), and consent sought. Participants who are willing and able to give informed consent, sign or provide a witnessed thumbprint on the consent form. Participants must provide consent for all procedures to be recruited, including consent for providing baseline biospecimens. They are able to withdraw consent for biosampling at later contacts and still remain in the study. Any 16–17 year-old assessed as fulfilling criteria for being an ‘emancipated minor’ (defined as living outside family home or having other significant reason not to involve parent/guardian) can give their own informed consent. Those not considered emancipated give assent and require informed consent from parent/legal guardian, as per Malawi research ethics guidelines. Following consenting, the fieldworker administers the questionnaire in as private a setting as possible, and conducts anthropometry and other physical measurements. A separate biosampler attends the household and conducts biosampling on participating individuals. Baseline information and samples collected are summarised in tables 3 and 4 and online supplemental table 1 and described in more detail below. Table 3Adult measuresDataMeasure/formsContactDemographics Personal details (age, sex, marital status)Standard MEIRU questionnaireBLIndividual SES Education, occupation, parental educationStandard MEIRU questionnaireBL Employment status, personal incomeStandard MEIRU questionnaireBL, M12Obstetrical/reproductive Gestational age/EDD by USSPregnancy USSE Current pregnancy/postnatal care infoPhoto of health passport pagesBL, T3, D, W16, M12, M24, M36 Obstetrical and childbirth historyStandard MEIRU questionnaireBL Pregnancy intention/family planning controlAdapted London Measure of Unplanned Pregnancy /DHS questionnairesBL, (M12, M24, M36 if pregnant) Delivery recordPhoto of delivery+midwife formDGeneral health and disability History of diagnosed physical and mental disorders. Current medications.Standard MEIRU questionnairesBL HIV and ARTStandard MEIRU questionnairesBL, M24 DisabilityWHODAS (12 item)BL, M24Mental health Depression, anxiety, somatic symptoms - currentPHQ-9; GAD7; three items from PHQ-15BL, T3, W16, M24, M36 Lifetime major traumatic event+PTSD screen (current)Adapted PC-PTSDBL Childbirth specific trauma and+PTSD screen (current) (women only)Adapted PC-PTSD and questions from City Birth Trauma ScaleW16, M24 Lifetime depressive and manic symptoms self report (including if occurred in perinatal period)Adapted PHQ-2+functioning Q for lifetime, 4 questions adapted from SCID, PHQBL Alcohol (current and past)AUDITBL Substance use (lifetime)ASSISTBL Substance use (3 months) including alcoholASSISTBL, T3, W16, M24, M36Cognitive testing TMB Digit Symbol, Matrix Reasoning, Visual Paired AssociatesBLBonding Antenatal bonding (women only)PAIT3 Bonding with infantMAIW16Risk and resilience NeuroticismEPQ-SF Neuroticism 6 itemBL Food insecurityHFIASBL, T3, W16, M24, M36 Social support (confiding relationship)MSPSS+identity of ‘significant other’BL, T3, W16, M24, M36 Life eventsAdapted SHARP questionnaires+ELSQ/CPASBL, T3, W16, M24, M36 Intimate partner violence (women only)Adapted VAWI - WHO multicountry studyBL, W16, M24, M36 Coercive control (women only)Adapted VAWI - WHO multicountry studyBL, M24, M36 Adverse childhood experiencesAdapted ACE-IQW16Physical measures HR, BPBL, W16, M24 Height, weight and body compositionBL, W16, M24 Mid upper arm circumferenceBL, W16, M12, M24ACE-IQ, Adverse Childhood Experiences - International Questionnaire; ANC, antenatal clinic; ART, antiretroviral therapy; ASSIST, Alcohol, Smoking and Substance Involvement Screening Test; AUDIT, Alcohol Use Disorders Identification Test; BL, baseline/enrolment (home); BP, blood pressure; CPAS, Creative Process Assessment Scale; D, delivery (labour ward - facility); DHS, Demographic and Health Survey; E, eligibility contact (ANC - facility); EDD, expected delivery date; ELSQ, Early Life Stress Questionnaire; EPQ-SF, Eysenck Personality Questionnaire - short form; GAD-7, 7-item Generalised Anxiety Disorder Scale; HFIAS, Household Food Insecurity Access Scale; HR, heart rate; M12, month 12 (home); M24, month 24 (home); M36, month 36 (home); MAI, Maternal Attachment Inventory; MEIRU, Malawi Epidemiology and Intervention Research Unit; MSPSS, Multidimensional Scale of Perceived Social Support; PAI, Prenatal Attachment Inventory; PC, primary care; PHQ-9, 9-item Patient Health Questionnaire; PTSD, post-traumatic stress disorder; SCID, Structured Clinical Interview for DSM ; SES, socioeconomic status; SHARP, sub-Saharan Africa Regional Partnership for Mental Health Capacity Building; T3, third trimester (ANC - facility); TMB, Test My Brain; USS, ultrasound scan; VAWI, Violence Against Women Instrument; W16, week 16 (home); WHODAS, WHO Disability Assessment Schedule. Adult measures ACE-IQ, Adverse Childhood Experiences - International Questionnaire; ANC, antenatal clinic; ART, antiretroviral therapy; ASSIST, Alcohol, Smoking and Substance Involvement Screening Test; AUDIT, Alcohol Use Disorders Identification Test; BL, baseline/enrolment (home); BP, blood pressure; CPAS, Creative Process Assessment Scale; D, delivery (labour ward - facility); DHS, Demographic and Health Survey; E, eligibility contact (ANC - facility); EDD, expected delivery date; ELSQ, Early Life Stress Questionnaire; EPQ-SF, Eysenck Personality Questionnaire - short form; GAD-7, 7-item Generalised Anxiety Disorder Scale; HFIAS, Household Food Insecurity Access Scale; HR, heart rate; M12, month 12 (home); M24, month 24 (home); M36, month 36 (home); MAI, Maternal Attachment Inventory; MEIRU, Malawi Epidemiology and Intervention Research Unit; MSPSS, Multidimensional Scale of Perceived Social Support; PAI, Prenatal Attachment Inventory; PC, primary care; PHQ-9, 9-item Patient Health Questionnaire; PTSD, post-traumatic stress disorder; SCID, Structured Clinical Interview for DSM ; SES, socioeconomic status; SHARP, sub-Saharan Africa Regional Partnership for Mental Health Capacity Building; T3, third trimester (ANC - facility); TMB, Test My Brain; USS, ultrasound scan; VAWI, Violence Against Women Instrument; W16, week 16 (home); WHODAS, WHO Disability Assessment Schedule. Table 4Biosamples collected in the Generation Malawi study at the different contactsSampleDetailsContact collectedAdultBlood serum· Same day tests: full blood count and random blood glucose.· Serum, plasma and cell pellets aliquots stored at −80°C.BL, T3, W16, M12, M24, M36Blood plasmaFaeces (or rectal swab)· Diluted, aliquoted and stored at −80°C.T3Swabs - nasopharyngeal, oropharynx, skin, vagina· Collected from women only using appropriate swabs and stored at −80°C.T3Hair· Stored in tin foil at room temperature.T3, W6, W16Breast milk· Expressed into collection tube, for centrifuge and storage at −80°C.W6, W16Saliva· Women and spouses will collect saliva into sampling tubes stored at −80°C.T3, W6, W16Placenta· Four placenta biopsy samples are collected approximately 4 cm from the umbilical cord. These samples are stored in formalin and in RNA protect at −80°C.DInfantsUmbilical cord blood serum· 16 mL of cord blood will be collected into EDTA and serum test tubes, then taken to the laboratory for centrifuging, dividing serum 5, plasma 3 and cell pellets stored at −80°C.DUmbilical cord blood plasmaSerum blood· Collected via heel prick into serum tube and on dry blood spots.· Samples aliquoted and frozen at −80°C.W1, W6, W16Faeces or rectal swab· Collected from infants from diapers using a faeces scoop.· Diluted, aliquoted and stored at −80°C.D, W1, W6, W16Swabs - nasopharyngeal, oropharynx, skin· Collected and stored at −80°C.D, W6, W16Urine· Samples are collected in urine bags, pipetted into tubes, then stored at −80 °C.W1, W6, W16Hair· Samples will be stored in tin foil at room temperature.W6, W16Saliva· Infant saliva is collected with salivates placed in the mouth for 1 min, then centrifuged, aliquoted and stored −80°C.W6, W16ANC, antenatal clinic; BL, baseline/enrolment (home); D, delivery (labour ward - facility); M12, month 12 (home); M24, month 24 (home); M36, month 36 (home); T3, third trimester (ANC - facility); W1, week 1 (post-natal check - facility); W6, week 6 (under-5 clinic - facility); W16, week 16 (home). Biosamples collected in the Generation Malawi study at the different contacts ANC, antenatal clinic; BL, baseline/enrolment (home); D, delivery (labour ward - facility); M12, month 12 (home); M24, month 24 (home); M36, month 36 (home); T3, third trimester (ANC - facility); W1, week 1 (post-natal check - facility); W6, week 6 (under-5 clinic - facility); W16, week 16 (home). Third trimester contact at antenatal clinic (woman only) Participating women attending a routine ANC visit between 34+0 to 37+6 weeks gestation (from March 2025 this was adjusted to 34+0 weeks up to 39+6 weeks) are identified and invited to participate in a second study contact. In the ANC waiting area, women enrolled in the study are identified from GM study stickers in their ‘health passports’ (patient held medical record booklet) and from a participant list. Information and samples collected are summarised in tables 3 and 4 and described in more detail below. Women who do not attend a given third trimester contact (3T) ANC appointment as expected are contacted by phone and encouraged to attend on another date. Non-participation at 3T ANC contact does not preclude ongoing participation in the study. Delivery contact at facility: delivery details and enrolment of child In the rural site, most women enrolled in the study deliver at CRH, with some deliveries taking place at Karonga District Hospital some 80 km distant (either electively for primiparous women, multiple births or prior obstetrical complications or following emergency referral from CRH). In the urban site, most women deliver at Area 25 HC where full obstetrical services including instrumental deliveries and caesarean section are available. Maternity staff at all these facilities have been briefed about the study and trained in initial study delivery procedures. The maternity staff identify women in labour who are in the study (from the health passport) and record key delivery information on a paper form including date/time, APGAR scores and birth weight and take cord blood and placental swabs/samples (and meconium if present). Maternity staff then alert the GM research team who conduct the rest of the delivery contact procedures within 24 hours. These procedures include enrolling the child in the study, collecting questionnaire data, child anthropometry (including body composition using PEAPOD air displacement plethysmography) and biosamples (tables 4 and 5). Women who have not delivered by their EDD are contacted by phone. If they have already delivered at a different facility, then the delivery questionnaire is administered by phone (or a home visit arranged) and available information collected. As long as the mother consents to enrolment of the child, the family remains in the study. Table 5Child measuresDataMeasure/formContactChildhood illness and vaccination Episodes of childhood illnessDirect questionnaire and health passportW1, W6, W16, M12, M24, M36 Vaccination historyHealth passportW1, W6, W16, M12, M24, M36 HIV/ARTStandard MEIRU questionnairesW1, W6, W16, M12, M24, M36 Current IMCI illness concern and referralStandard MEIRU questionnairesW6, W16, M12, M24, M36Development Motor disabilityPrechtls videoW16 DisabilityWashington group questionnaireM24, 36 General developmentMalawi development assessment toolW16, M12, M24, M36 Sleep historyStandard MEIRU questionnairesW6, W16, M12, M24, M36 FeedingStandard MEIRU questionnairesW1, W6, W16, M12, M24, M36 StimulationFamily care indicatorsW16Language MacArthur Bates QuestionnaireW16, M24 BehaviourStrengths and difficulties, STARTM36 CognitionBabyscreenM24 CognitionTangerine EF TouchM36Mother-child interaction and adversity Family caregivingFamily care indicatorsM12, M240 on the PHQ-9 suicidal ideation item.27 Mental health history and neuroticism Self-report of lifetime possible depressive episode (three items adapted from PHQ-9) and whether this occurred perinatally (women only). Self-report of lifetime symptoms of possible mania (three items adapted from Structured Clinical Interview for DSM (SCID)28 and PHQ-9) and whether this occurred postnatally (women only). Neuroticism, the tendency to negative emotional states including worry, is an established risk factor for depressive and anxiety disorders.29 It is measured using the six neuroticism questions from the revised abbreviated Eysenck Personality Questionnaire.30 Recent and lifetime alcohol and substance use The Alcohol Use Disorders Identification Test, a 10-item WHO developed screening tool, is used to measure consumption and harmful use of alcohol. This tool has been adapted and used in several studies in Malawi.31 32 We conducted further adaptation including development of a fieldworker tool to evaluate consumption of locally brewed drinks. Use and harmful use of other substances is screened using two questions from the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST), a screening tool developed by the WHO (accessed at: https://www.who.int/substance_abuse/activities/assist_test/en/). Cognition Cognitive testing in LI-AC settings such as rural Malawi requires the use of tests that are culturally neutral and can be completed by participants of varying literacy levels. We adapted measures developed by The Many Brains Project (www.ManyBrains.net): Matrix Reasoning (attention, perception, cognitive control, working memory); Digit Symbol Matching (processing speed, visual short-term memory); and Visual Paired Associates (perception, declarative memory, working memory). Tests are self-administered using a tablet computer after a brief orientation by the fieldworker. Bonding: antenatal and postnatal The antenatal bond of the mother with her unborn child is measured with the Prenatal Attachment Inventory.33 We adapted and validated a 19 item version (Manuscript in progress). Maternal and paternal postnatal bonding with the infant is measured with the Maternal Attachment Inventory.34 Psychosocial risk and resilience: social support, food insecurity, intimate partner violence, adverse childhood experiences Perceived social support is measured using the Multidimensional Scale of Perceived Social Support, a brief measure of the respondent’s perception of the adequacy of support he/she receives from family, friends and ‘significant other’.35 A Chichewa version has been previously validated in Malawi.36 Food insecurity is measured with the Household Food Insecurity Access Scale.37 IPV is measured using the WHO Violence Against Women Instrument. This measures women’s experience of IPV in three domains: psychological, physical and sexual violence.38 In addition, questions on coercive control are included at baseline. The questions are administered via an audio tablet-assisted self-interview (ACASI) to ensure confidentiality.39 The Adverse Childhood Experiences - International Questionnaire (ACE-IQ) is administered to women and men using an ACASI. It asks about experience of childhood adversity in 13 domains, including exposure to violence, neglect, sexual abuse and family dysfunction. The ACE-IQ had been previously translated into Chichewa and validated and used in Malawi among adolescents.40 Stressful life events (including bereavement, severe illness, being a victim of crime and loss of employment) are recorded using questions adapted from the SHARP Study.27 Physical measurements Adult weight and body fat compositions are measured using OMRON body composition monitor scales. Standing height is measured with a SECA Stadiometer to the nearest 0.1 cm. Mid-upper arm circumference is measured using insertion tapes. Blood pressure is measured using calibrated electronic blood pressure monitors to a standard protocol. Biological samples Biological samples from adult participants include prenatal venepuncture samples and mucosal swabs and postnatal placental sampling, saliva, hair and breast milk. They are listed in table 4, with further details described below. Child measures Delivery Twin or singleton, gestational age at delivery, delivery place, delivery type, APGAR scores, immediate birth weight. Routine delivery record photographed. Child health and nutrition Vaccination history, episodes of childhood illness, HIV and use of antiretroviral medications. Feeding questionnaire adapted from DHS questionnaire.41 Sleep Infant Sleep questionnaire adapted from Brief Infant Sleep Questionnaire – Revised.42 General development Malawi Development Assessment Tool assessing the domains of gross motor, fine motor, language and social development43 assessed at week 16 (W16), month 24 (M24) and month 36 (M36). Neuromotor disability risk Assessed at W16 using Prechtl General Movement Assessment.44 This involves video recording of the infant lying supine to detect complex ‘fidgety’ movement, absence of which indicates that the infant is at risk of neuromotor disability. Stimulation in the home This is assessed using the UNICEF-developed Family Care Indicators Questionnaire.45 Care outside the home General information on care outside the home is assessed using the adapted Nursery School Attendance questionnaire. Mother–child interaction At W16 contact, a videoed, semi-structured mother–child interaction procedure is used46 that allows coding against the Qualitative Scales of the Observational Ratings of Mother-Child Interaction of the National Institute of Child Health and Human Development scales. Mother and child are recorded interacting face to face for 2 min without any toys or other stimulations (direct interaction). Then, the mother is asked to complete a caregiving activity. Lastly, the mother and child play with a set of simple, age-appropriate toys provided by the researcher for 7 min (free play). Recording is done at the participant’s home using a tablet attached to a tripod. At W16, M24 and M36 contacts, the Observation of Mother–Child Interaction (OMCI) procedure is used. OMCI is a short observational tool created for the evaluation of parent–child interactions.47 Coding is completed by the fieldworker while watching a live 5-minute session of parents interacting with their child over a simple book or another play interaction. OMCI showed validity and reliability in LMIC settings including in sub-Saharan Africa.48 It has been adapted and used in Malawi in the neuRocog Exposed inFants ImmuNE (REFINE) study (Gladstone, personal communication). Although coded live, we also video the procedure.49 Language and communication At M24 and M36 contacts, child language and communication is assessed using the MacArthur-Bates Communicative Development Inventories: Word and Gesture, a systematic interview to the parent (or caregiver) designed for children 8–18 months of age but which can be extended to 37 months.50 In Malawi, the Communicative Development Inventory has demonstrated high sensitivity in detecting developmental changes in children from 12 to 30 months of age.51 Physical measurements Infant weight (kg) is measured using Tanita Baby weighing scales. Infant length is measured using length boards. Standing height is measured with a SECA Stadiometer to the nearest 0.1 cm. Infant baby head circumference and mid-upper arm circumference are measured using insertion tapes. The PEAPOD infant body composition system is used to assess infant body composition. This device measures infant body composition, fat and fat-free mass through air displacement plethysmography.52 Biological samples Biological samples from children include capillary blood sampling, stool, urine and mucosal swabs. They are itemised in table 3 and further details are described below. Biosample collection, processing and storage Biosampling is conducted by trained personnel operating in the health facilities and travelling to participants’ homes. Biosamples at delivery (placental biopsies/swabs, cord blood and meconium if present) are collected by the maternity staff responsible for the delivery. All samples are collected to Standard Operating Procedures (SOPs) with appropriate training. Samples are transported in field cool boxes with icepacks to the laboratory within 4 hours of collection or are stored in monitored facility fridges until transport to laboratory. The SOP for biosample collection and transport (including transport in cool boxes) has been used at MEIRU in these settings for many years, for a wide range of laboratory investigations, including immunological assays, microbiological investigations, DNA and RNA extraction as well as biochemistry and haematology and the quality of the samples is preserved. On arrival at the laboratory, the biosamples are registered and processed to SOP. Real-time assays (full blood counts and glucose) are conducted, and these results are fed back to participants by field teams. Multiple aliquots are stored in monitored and alarmed −80°C freezers and locations (box, rack and freezer) for each aliquot are documented and tracked in FreezerWorks (rural) and a customised database (urban). Duplicate sets of aliquots are transported to the other (ie, urban or rural) laboratories for secure storage to insure against catastrophic loss in either site. The laboratories have generator back-up and an on-call rota of administrative and laboratory staff to respond to emergencies. Management, governance and sharing of data and samples MEIRU data are held securely to best practices as outlined in Medical Research Council policy and elsewhere.53 MEIRU data team ensure that highly detailed longitudinal datasets can be released publicly with minimal disclosure risk. MEIRU already makes large datasets available https://www.meiru.info/data-sets/ to ensure that Malawi populations are fully represented. We share specific datasets to individual researchers and institutions on request and rarely refuse applications. Access is governed through two cohort access committees. Linked data MEIRU stores linked data on several million contacts with hundreds of thousands of participants and associated biosamples from Karonga district dating from 1979 onwards. Later data include an NCD survey 2013–2016 (rural and urban, 28 000 participants, detailed data on cardiometabolic conditions and behavioural risk factors, examinations, biosampling, aliquots stored and accessible). A long-term conditions survey within the Healthy Lives - Malawi study is currently accruing data 2021–2025 (rural and urban, 50 000 participants). These datasets include parents (preconception measures) and grandparents of Generation Malawi infants. Nested studies include kidney disease, stroke, liver disease and more, https://www.meiru.info/data-sets/. All the work is nested in rural and urban health and demographic surveillance sites, with monthly data on births, deaths, verbal autopsies and annual migration and socioeconomic individual and household data. As of 2024, the data and sample collections are sufficient to promote analysis and sharing. The Data and Biorepository Access Committees, convened as part of the Malawi-led governance of the resource, will make decisions on the availability and security levels for each tranche of data and to consider custom requests for datasets and access to biosamples according to defined criteria. In future requests for data or sample access will be processed through the MEIRU data repository platforms, prior to that launch requests can be made by contacting info@meiru.mw Data sharing Data will be Findable, Accessible, Interoperable and Reusable. Current data sharing procedures involve the recipient signing a MEIRU data use agreement and being provided access to a de-identified unlinkable analysis dataset. In future, we will capitalise on the African Population Cohort Consortium Trusted-Research-Environment-in-a-box platform currently under development, to enable permitted researchers to access data and analytical tools in a ‘safe haven’ without ability to download datasets. Sample sharing The nature and volumes of biosamples collected are balanced between storage of adequate volume for repeated access relating to different studies and avenues of research, either internally or externally generated. The samples constitute a valuable, depletable resource and access is governed by a Biopository Access Committee. Analysis of Generation Malawi samples requires reciprocal sharing of laboratory assay data generated, which then becomes sharable in turn, to maximise the value of the resource. Statistical analysis Baseline characteristics will be summarised using standard descriptive statistics. Continuous variables will be reported as means with SD for normally distributed data, or medians with IQRs for skewed data. Categorical variables will be presented as frequencies and percentages. Unadjusted associations between exposures (eg, SES, rural/urban residency, education) and outcomes (depression, anxiety, PTSD, birth outcomes and infant growth) will be assessed using Pearson’s χ2 for categorical variables. Associations between exposures and continuous outcomes will be compared across groups using independent samples t-tests or one-way analysis of variance, as appropriate. For multivariable analyses, to formalise structural assumptions and mitigate confounding Directed Acyclic Graphs (DAGs) will be developed a priori using expert knowledge and existing literature to map the assumed causal structures of exposures, covariates, mediators and outcomes. DAGs will be used to identify the minimally sufficient adjustment sets required to estimate the total effects of our primary exposures, ensuring adjustment for true confounders while avoiding conditioning on colliders or mediators. To evaluate the effect of exposures on longitudinal outcomes, while accounting for the correlation of repeated measures within individuals over time, mixed-effects modelling will be applied. Linear mixed-effects models will be used to analyse continuous outcomes, such as infant growth, over follow-up waves. For binary or count outcomes, such as the clinical presence of depression or PTSD, generalised linear mixed models will be applied. All longitudinal models will incorporate exposure-by-time interaction terms to assess whether the rates of change in the outcomes differ according to exposure status and will include random intercepts to account for baseline clustering. To identify mediating pathways between exposures and outcomes, causal mediation analyses will be conducted. To overcome the limitations of traditional mediation approaches, which are biased in the presence of exposure-induced mediator-outcome confounding, and strict assumptions surrounding the causal order of the mediators and linearity assumptions, we will estimate interventional direct and indirect effects.54 Further, we will adapt this approach to account for both time-varying confounders and time-varying mediators across the follow-up waves.55 Patterns of missing data will be assessed. If appropriate we will employ multiple imputation by chained equations to generate imputed datasets assuming data is missing at random. Primary multivariable analyses will be conducted on the imputed data, and pooled results will be reported using Rubin’s rules. To assess the robustness of our findings and evaluate potential bias introduced by attrition or non-participation, sensitivity analyses will be performed comparing the imputed results with complete-case analyses. Analyses will be performed using Stata V.18.0/R V.4.3.0. Discussion Strengths and limitations The Generation Malawi study offers a unique and valuable resource for understanding intergenerational and life-course influences on health in a low-income African context, where such data are scarce. Several key strengths enhance the scientific value and sustainability of this cohort. The study is implemented in both urban and rural settings, allowing comparative analyses across diverse socio-environmental contexts. It is nested within an established HDSS, enabling accurate linkage to demographic, morbidity and mortality data, as well as to other ongoing studies. The broad consent for biological sample storage and future use provides opportunities for a wide range of analyses, including genetic, epigenetic and biomarker studies. The study has several potential limitations. Loss to follow-up over time may introduce selection bias and reduce the completeness of outcome data, particularly for longitudinal analyses. Lower participation and retention among fathers may limit the ability to fully assess paternal contributions and intergenerational influences on child health outcomes. Additionally, recruitment is based on antenatal care attendance and residence within the HDSS catchment areas, which may limit generalisability and introduce selection bias by excluding women who do not access antenatal care or populations living outside the surveillance areas. Future strategy and sustainability Funds permitting, the Generation Malawi has been established with the intention that follow-up should continue for as long as new questions can be addressed and cohort retention is adequate to provide a good and representative sample size of infants (immediate), young children (within the scope of this project) and school-age children and adolescents (within the scope of future applications). These follow-up assessments should be conducted at least annually to ensure ongoing engagement with cohort participants. Shifting cohort data collection to routine data linkage and low-level cohort maintenance is highly unlikely in the next decade. (1) Infrastructure is lacking: Routine data (births, deaths, health service contacts) in Malawi are at best patchy and often completely absent, owing to under-resourced human resources and infrastructure. (2) Data are of poor quality: Vital registration is incomplete and often excludes the most vulnerable. Furthermore, as we have shown, individuals with long-term conditions are predominantly undiagnosed, with little indication of change. The future African Population Cohorts Consortium is likely to characterise this cohort as highly valuable, increasing the likelihood of supplementary funding through an African Population Cohorts Consortium (APCC) funder network. Given the interest to date, we predict multiple further requests for associate research to contribute to operational costs (other MEIRU Longitudinal Population Study contributions from awards to date range from 10% to 50% of their cost). Plans to collaborate MEIRU actively shares data and samples across several research networks and is active in a work package within the African Population Cohorts Consortium to develop standardised tools and harmonised datasets. We collaborate with partners in Malawi and the UK on studies exploring the early origins of cardiometabolic risk, the genomics of depression and anxiety, the effects of maternal malaria on infant immunity, the role of gut pathogens in child growth and the role of early-life malnutrition in later-life NCD risk (CHild malnutrition and Adult NCD: Generating Evidence on mechanistic links to inform future policy/practice (CHANGE) project).56 MEIRU also contributes to a Wellcome working group on longitudinal mental health data and works closely with international cohorts, including Avon Longitudinal Study of Parents and Children (ALSPAC), Born in Bradford, the Dutch Birth Cohort, Kula Leap, C-Gull (Liverpool) and emerging Ethiopian birth cohorts, to strengthen future collaborations. Ethics and dissemination This study protocol was approved by the National Health Science Research Committee in Malawi (number: 21/01/2653) and the Ethics Committee of the College of Medical Veterinary and Life Sciences at the University of Glasgow, UK (protocol number: 200190183). The formative work for this study was granted ethical clearance by the College of Medicine Research Ethics Committee (protocol number P.11/19/2865). Findings and interpretations from this study will be made public through policy briefs, seminars, local and international conferences and publications in peer-reviewed open-access journals. As part of the dissemination plans, the findings will be made available locally to different channels, including regional and national meetings of Preventive Health and NCDs and Mental Health at the Ministry of Health. Through a Wellcome Trust funded public engagement award, we have created a documentary ‘Growing Up in Malawi’ featuring 20 families participating in Generation Malawi. Interviews, film of birth cohort activities and home-life, and family-shot footage of daily life and key events during pregnancy and early infancy are interwoven to the experience of participating in a birth cohort as well as the experience of expecting a new child in rural and urban Malawi (https://www.youtube.com/watch?v=igzj8ixC3ao). In collaboration with the African Alliance of Maternal Mental Health, we have also created maternal mental health advocacy videos featuring women from the birth cohort experiencing mental health conditions who have agreed to share their stories to help others (https://www.youtube.com/watch?v=oiTYNg6hieI). Supplementary material Footnotes Funding: This work is supported by the UK Research and Innovation (UKRI) Medical Research Council (MRC) GCRF award to AM (MR/S035818/1), Wellcome Trust awards to ACC (217073/Z/19/Z) and AM (223165) for the genetic analysis and UKRI MRC LPS award to ACC (APP41245). Prepublication history and additional supplemental material for this paper are available online. To view these files, please visit the journal online (https://doi.org/10.1136/bmjopen-2026-122132). Provenance and peer review: Not commissioned; externally peer reviewed. Patient consent for publication: Not applicable. Collaborators: Generation Malawi collaborator group: Abena Amoah, Andrew Matchado, Balthazar Mtenga, Charles Mwansambo, Charlotte Hanlon, Chikondi Chiweza, Debbie Lawlor, Debby Bogaert, Deborah Nyirenda, Elizabeth Chodzaza, Effie Chipeta, Emmie Mbale, Estelle McLean, Joseph Mkandawire, Kazione Kulisewa, Linda Nyondo-Mipando, Louis Banda, Owen Nkoka, Albert Dube, Louis Gadama, Luis Gadama, Martyn Pickersgill, Moffat Nyirenda, Mphatso Chisale, Myness NdamboIain, Iain McInnes, Rebecca Reynolds, Victor Mwapasa, Paul Kambiya. Map disclaimer: The depiction of boundaries on this map does not imply the expression of any opinion whatsoever on the part of BMJ (or any member of its group) concerning the legal status of any country, territory, jurisdiction or area or of its authorities. This map is provided without any warranty of any kind, either express or implied. Patient and public involvement: Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details. Contributor Information Generation Malawi Collaborator Group: Abena Amoah, Andrew Matchado, Balthazar Mtenga, Charles Mwansambo, Charlotte Hanlon, Chikondi Chiweza, Debbie Lawlor, Debby Bogaert, Deborah Nyirenda, Elizabeth Chodzaza, Effie Chipeta, Emmie Mbale, Estelle McLean, Joseph Mkandawire, Kazione Kulisewa, Linda Nyondo-Mipando, Louis Banda, Owen Nkoka, Albert Dube, Louis Gadama, Luis Gadama, Martyn Pickersgill, Moffat Nyirenda, Mphatso Chisale, Myness NdamboIain, Iain McInnes, Victor Mwapasa, Paul Kambiya, and Rebecca Reynolds References Supplementary Materials