Background Commencement of lifelong highly active antiretroviral therapy (HAART) immediately after HIV diagnosis (option B+), for treatment of human immunodeficiency virus (HIV), has greatly improved maternal-infant health in sub Saharan Africa (SSA). However, this development has also dramatically increased the number of maternally HAART/HIV-exposed-uninfected (HEU) infants in areas of high HIV prevalence. Compared to their HIV-unexposed uninfected (HUU) counterparts, HEU infants show increased mortality, higher rates of adverse birth outcomes, infectious and non-communicable diseases and impaired growth, immune responses and neurodevelopment. Adverse clinical outcomes and their respective risk factors alongside associated biomarkers of HEU infants in SSA have been insufficiently characterized. Early exposure to HAART and HIV might be risk factors for the adverse outcomes in HEU infants but other potential risk factors and biomarkers remain understudied.

Methods The University of Zimbabwe-College of Health Science birth cohort is a prospective cohort study of perinatal HIV and in utero HAART exposure throughout the breastfeeding period in the era of option B+. 600 HIV infected and 600 HIV uninfected pregnant women ≥20 weeks of gestation are being enrolled from four primary health centres in poor high-density residential areas of Harare. Clinical, socio-demographic/economic, nutritional and environmental data and bio-samples including maternal urine, stool, plasma, milk, cord blood, amniotic fluid as well as infant serum, dried blood spots and stool are being collected at enrolment, delivery and longitudinal follow-ups as mother-infant pairs from delivery, week(s) 1, 6, 10, 14, 24, 36, 48, 72 and 96 after birth. Infants are being assessed for congenital transmission of HIV, hepatitis B/C viruses, cytomegalovirus, syphilis, and growth, neurodevelopment, and immune-dysregulation. Sub-studies are addressing maternal-infant immunometabolomics, latent tuberculosis infection, dysbiosis of the gut microbiome and the effect of maternal stress thereof. The primary end point of this study is infant mortality until two years of age in HEU versus HUU infants. Secondary outcomes include HEU morbidity.

Conclusion Our study will provide a comprehensive assessment of risk factors and associated biomarkers for adverse clinical outcomes for HEU infants and ultimately help developing strategies to mitigate effects of HIV, comorbidities and early life HAART exposure on pregnancy outcome and infant health.

Trial registration number, date Key words: HIV, Option B+ highly active antiretroviral therapy (HAART), in utero exposure, breastfeeding, antenatal co-morbidities, immune dysfunction, microbiota, genomics, pregnancy outcomes, neurodevelopment infant health.

Over one million infants are born to HIV-1-infected women every year in Sub Saharan Africa. In Zimbabwe, annual births stand at 379,000 with approximately 48,000 infants being born every year to HIV-1-infected women on lifelong highly active antiretroviral therapy (HAART). Concerns have been expressed on the general health and developmental outcomes of HIV-1 exposed but uninfected (HEU) infants born to HIV-1-infected women. In spite of being HIV negative, HEU children have 3.9-and 2.0-fold higher mortality than HIV unexposed and uninfected (HUU) infants during the first and second year(s) of life, respectively. Compared to their HUU counterparts, HEU children may have adverse birth outcomes, impaired growth, development, and other health deficits. This is a significant public health concern as in adulthood, infants born with low birth weight (LBW) are more likely to die from non-communicable diseases (NCDs) such as hypertension, stroke and type 2 diabetes. In Zimbabwe, HEU infants have been well-studied during the pre-HAART era; however, contemporary evidence is lacking. In addition, previous studies have not characterized women in pregnancy and the outcome has not been controlled for other common antenatal infective pathogens including additional risk factors such as preterm birth, LBW, duration of breastfeeding, poverty and maternal psychosocial stressors. Therefore, an in-depth and comprehensive characterization of HEU infants is needed not only to avoid HIV-infection but also to improve short-and long-term health outcomes of the exposed.

Risks of HAART in pregnancy In Zimbabwe, the current standard of care for HIV infected women to prevent mother to child transmission (MTCT) of HIV consists of Tenofovir, Lamivudine and Efavirenz (TENOLAM-E). However, antiretroviral drugs can accumulate in the amniotic fluid, thus posing a risk during delicate embryonic and foetal developments. Timing of HAART might also be important, since a South African study showed that mothers who initiated zidovudine based HAART before conception had twice the odds of preterm delivery compared to mothers who started HAART after conception. Toxic mechanisms of anti-HIV drugs include mitochondrial toxicity and mitochondrial DNA mutations in mothers and possibly also in infants. Mitochondrial dysfunction will in turn increase the production of reactive oxygen species and activation of IL-1 production. Interference with lipid metabolism is another mechanism for HAART toxicity. Early exposure to Efavirenz-based HAART has been associated with neurodevelopmental and socio-emotional challenges which may lead to behavioural deficits. In addition, persistent infections, detrimental effects of socio-economic status and maternal dietary contaminants such as dioxins or chlorinated pesticides may also play a role.

The reasons for the observed adverse health outcomes of HEU infants could be related to an altered immune system state. However, immune profiles beyond one year of HEU children remain largely unexplored. In Zimbabwe, vaccination campaigns for active immunization of infants are promoted via the expanded program of immunization (EPI). Even though peak antibody titres following some EPI vaccines are normal in HEU infants, there have been no studies of the longevity of these immune responses. In utero HIV-exposure increases pro-inflammatory cytokine levels produced by placental cells. Co-infections seem to be important since Kenyan HEU neonates (with a high number of maternal co-infections) had higher levels of inflammatory plasma markers than HEU controls from studies in the USA. Further, HEU infants have higher CMV loads, which may be driven partly by maternal viremia and may contribute to immune activation. Thus, HEU infants continue to have baseline inflammation which can be linked to poor short- and long-term outcomes and anti-inflammatory interventions might improve infant health. Interventional studies with HAART treatment in pregnant women are ethically challenging; however, the impact of HAART on neuro/immune development could be assessed in studies like the UZ-CHS birth cohort, where more often, due to economic challenges mothers' book late for their antenatals, and some being HAART naïve in their third trimester. Evidence regarding the risk of HAART exposure in the option B+ era for the short and long-term outcomes of HEU infants remains limited. However, to the best of our knowledge, no study has correlated HAART levels in maternal amniotic fluid, urine, plasma and longitudinal breast milk and related to adverse pregnancy outcomes, infant growth, metabolism and (immune/neuro) development.

Methods The University of Zimbabwe College of Health Science (UZ-CHS) Birth Cohort study is a prospective observational cohort study comparing infants born to HIV infected and HIV uninfected women in a low-resource setting of high-density western suburbs of Harare primary health polyclinics. 1,200 pregnant women (600 HIV infected mothers and 600 HIV uninfected controls) in their third trimester were enrolled from January 2016 to June 2019. Participants are being followed as mother-baby pairs at birth, within 10 days of life 6, 10, 14, 24, 48, 72 and 96 weeks of age. At each visit clinical examinations are being used to assess health and the impact of environmental factors. In addition, questionnaires are administered and bio-samples for laboratory tests. The design of the study is non-interventional cohort but participants are being offered advice regarding health and hygiene.

Aim of the study This study aims to examine the impact of maternal HIV status, cumulative HAART exposure in utero and throughout breastfeeding, immune suppression and immune activation and environmental variables on short-and long-term outcomes of infants in a low resource environment. Our study is focusing on mortality and morbidity of HEU infants, testing established risk factors from previous cohorts. Data and bio-materials are being collected prospectively from mothers and infants from pregnancy to 2 years of age and compare HEU to HUU infants from the same community. For a comprehensive assessment, the role of maternal socio economic status, household factors, comorbidities including co-infections with persistent viruses such as cytomegalovirus (CMV), hepatitis b/c (HBV)/(HCV), bacterial diseases such as syphilis and tuberculosis (TB), intestinal helminths, emerging NCDs, such as gestational diabetes mellitus (GDM) and hypertensive disorders and maternal nutritional status on pregnancy outcomes, infant mortality and development, immunity and health will be tested. The UZ-CHS offers an opportunity to establish references ranges that are currently not available, for various subgroups of pregnant women with a favourable outcome of pregnancy for clinical use for future patients.

Study sites and local municipal engagement The municipality of Harare has established 12 polyclinics each comprising units for primary health care, maternity, post-natal care and a family health services. Medical services include HIV counselling and rapid testing and initiation and administration of HAART. Monitoring and follow-up for other chronic conditions including GDM and arterial hypertension are also offered. A physician comes twice a week per site to attend to complicated cases. Maternity services comprise antenatal care, follow-up visits and HIV and syphilis screening and treatment. It costs a mother US$25 to access antenatal services until 10 days after delivery whilst health services are free of charge for the infant until 5 years of age. Mothers are encouraged to book early during pregnancy (at week 4-12) and to test for HIV together with their spouses. Currently, due to social-economic challenges, most pregnant women go for antenatal registration at the polyclinics when their pregnancies are at advanced stages, a situation that puts their unborn babies at risk for HIV transmission, perpetuating the vicious cycle of poverty and HIV infection. Following a pre-study site feasibility assessment, out of the 12 polyclinics, Kuwadzana, Dzivaresekwa (Rujeko), Glenview and Budiriro were selected based on the higher volumes of maternal and child health services, frequency of HIV positivity and absence of competing researches targeting the same population. The catchment areas of the selected four polyclinics have relatively stable communities that do not frequently change their rented accommodation. Municipal nurses and midwifes at these facilities assist with study procedures such as data collection during delivery that occur late at night. All participants reside in South-western high-density areas of Harare. Residents of this community are of relatively poor socio-economic status with high unemployment rates. Most individuals generally cannot afford 3 appropriate meals per day. Most families (72%) live on less than 1 dollar per day and supplement their diet through small scale gardening around their houses. Families live in houses of similar structure and size (200 square metres). An average of 2-6 families live in these relatively small houses and in some extreme situations up to 5 family members share a single room. Water supply is erratic, necessitating the use of communal boreholes at 200-1,500 metres distance. This manually fetched water is used for drinking, cooking, bathing and toilet flushing. Sewage bursts are common and may occur close to boreholes. Contaminated boreholes in two of our four study sites, Budiriro and Glenview, have been epicentres of the recent 2018 cholera epidemic. Residents of these high-density areas share a disproportionate burden of both infectious diseases and NCDs relative to those living in low-density areas in the same city.

Sample size and statistical power The primary end point of our study is the comparison of infant mortality until two years of age in HEU versus HUU infants. In Zimbabwe, 50 infant deaths in the first year and 5 infant deaths in the second year of life are expected. In the pre-HAART era, a 3.2-3.9-fold higher mortality in the first year of life and a 2-fold higher mortality in the second year of life was expected. Mortality of HEU in the HAART era with more efficient HIV treatment might now be improved and in a conservative estimation, we assumed a 2-fold increase in mortality in HEU infants within the first two years of life. We therefore expected 5.5% mortality in HUU infants and 11% in HEU infants. According to our power analysis using G*Power3 a sample size of 1,198 women was required to detect such a difference with a power of 90% and a significance level of 0.05 in a two-sided analysis. We therefore opted for a sample size of 1,200 mothers, 600 with and 600 without HIV infection. Potential participants were identified during routine ante-natal care (ANC) visits at any of the four City of Harare Polyclinics. The potential participants were briefed on the study and those who verbally agreed to participate were given the participant information sheet written either in English or translated to vernacular Shona language to ensure full comprehension of the study aims and activities using the language that the participant understands best. Participation is voluntary and women are given the opportunity to withdraw at any time during the course of the study. Literacy is nearly universal in Zimbabwe and all potential participants should be able to read and comprehend the informed consent form.

Mothers' questionnaires procedures and measurements At enrolment, all mothers answered a structured questionnaire aiming at a comprehensive clinical, behavioural and environmental characterization. Specific questions will address sexual behaviour and reproductive health issues including sexually transmitted diseases and contraception use. General health, health seeking behaviour, maternal stress and drug/herbal use will be assessed. In HIV positive mothers, HAART use and duration, co-morbidities and issues related to HIV status disclosure and stigma will be recorded.

To address hygiene, specific questions on sanitation (type and number of toilets, sewage system), drinking water, type of energy used (indoor pollution), type of house, number of rooms used for sleeping by the family and the number of individuals sharing a single room will be asked. Economic information comprising employment situation, family monthly income, food security, diet, money set aside for food, healthcare, cooking fuel and savings will be assessed. Ownership of agricultural land or household assets including radios, televisions, mobile/ fixed telephones, refrigerators, bicycles, motorcycles/ scooters, and cars will provide additional information regarding social economic status. Further questions will address maternal life style such as alcohol use, smoking, sleeping habits, physical activities, domestic violence and support from partner including knowledge, beliefs and practices.

Physical examination and point of care testing At enrolment, the study midwife performed a full physical examination, blood pressure checks, assessment of oedema and anthropometry. WHO clinical staging will be done for all HIV positive women. Assessment of nutritional status will be done using standard anthropometric indices including body mass index (BMI). Mid upper arm circumference (MUAC) will be used as an indicator for malnutrition or obesity. The MUAC will be measured at the mid-point distance between the (acromion and olecranon. A South African study demonstrated that a MUAC of <23 cm, predicted a pre-pregnancy BMI of <18.5 kg/m2. For practical reasons, a maternal MUAC cut-offs of ≤25 cm will be used as a marker for malnutrition.

Blood pressure (BP) will be measured after allowing the mother at least 5 minutes of rest. A Tenso MED BP machine with a normal or large cuff covering at least half of the arm length is being used. Hypertension in pregnancy is defined by a blood pressure of ≥140/90 mmHg on at least 2 separate measurements 4 hours apart, anytime during the pregnancy. Mild to moderate hypertension as systolic BP of 140-159 mmHg or diastolic BP of 90-109 mmHg with severe hypertension refers to systolic blood pressure of at least 160 mmHg or diastolic blood pressure at least 110 mmHg. Chronic hypertension is defined by hypertension, diagnosed either before 20 weeks gestation or first diagnosed during pregnancy and persisting beyond 12 weeks postpartum. All mothers with proteinuria based on urine dipsticks are referred to the hospital for investigation of pre-eclampsia. Diagnosis of pre-eclampsia will follow established criteria.

Maternal and infant follow-up visits Mothers are reminded of their upcoming follow-up visits by phone call or through text message. Data can still be collected in cases of missed visits provided the participant turns up within a seven day window period of the scheduled visits. If a mother does not turn up for their scheduled follow-up appointments, attempts to contact her or her next of kin by phone and/or home visits by the research nurse are made strictly following mother's preferences documented at enrolment. Sick infants will be seen by the study Pediatrician. Information on place, date and mode of delivery, longitudinal infant feeding practices and comprehensive information on health, development and environment will be collected. Extensive questionnaires similar to enrolment will be re-administered at 6, 12, 18 and 24 months. A physical examination of mothers and their infants will be performed for mothers until week 6 and for babies at every visit. Study procedures also include extensive bio-sampling of blood, breast milk and stool. At each visit, transport reimbursement for clinic attendance is provided along with some refreshments. The mother may be asked to discontinue if she consecutively fails to turn up for at least two successive scheduled study visits without any reason(s) or communication.

Infant mortality The key outcome of this study is the composite endpoint "infant death", defined as stillbirth (gestational age at birth ≥20 weeks) and infants who die during the first two years of life. Neonatal deaths (0-364 days) will further be divided into three subgroups: early neonatal death (0-6 days), late neonatal death (7-27 days), and post-neonatal death (28-364 days). Late deaths (365-730 days) will also be recorded. Mortality will be assessed separately in HIV exposed and infected (HEI), HEU and HUU infants.

Adverse outcomes at delivery include preterm birth. Gestational age will be divided into early preterm ≤31 weeks, late preterm (32-36 weeks), term (at least 37 weeks) and post term birth (>42 weeks of gestation). Adverse birth outcomes include need for resuscitation at birth because of birth asphyxia or respiratory distress syndrome. Birth asphyxia will be defined as any medical condition resulting from decreased or discontinued supply of oxygen to a new-born infant before, during or soon after birth. Additional adverse outcomes are low Apgar at 5 minutes score (<7), small for gestational age i.e. weight/length below the 10th percentile for gestational age, LBW (<2,500g, weighed within the first hour of life, very LBW (<1,000g), foetal macrosomia (birth weight > 4,000g, microcephaly <2 SD from the mean, birth defects (malformations, extra digits) and multiple births.

Infant growth Growth will be expressed as Z scores according to WHO definitions: Weight-for-age (WAZ), height-for-age (HAZ), weight-for-height (WHZ) and head-circumference-for-age (HCAZ). Growth outcomes will be evaluated as continuous variables (attained Z-score and change in Z-score between visits). In addition, categorical outcomes; moderate wasting, stunting, severe stunting, underweight and microcephaly will be assessed as HAZ <-2, HAZ <-3, WAZ <-2; and HCAZ <-2, respectively. As a biological marker for infant growth, serum levels of plasma insulin-like growth factor-1 will be determined at 6 weeks of age using ELISAs, R&D Systems, Minneapolis.

Infant MUAC is being measured from 6 weeks of age and sensitivity, specificity, positive predictive value, and negative predictive values of the MUAC for stunting, underweight, wasting, anaemia and adverse neurodevelopment in HEI, HEU and HUU infants aged 0-<6 months, 6-<12 months and 12-24 months will be determined. WHZ will be used as the gold standard for wasting.

Assessment of infant morbidity Morbidity diaries are issued to mothers at delivery. Medical professionals document medical reports in the morbidity diary during sick visits or hospitalizations. These will include: neonatal hypoglycaemia, neonatal jaundice, infant anaemia, skin rashes, congenital infections, vertical transmission of infection over the 2 years, any hospitalisation (frequency and duration), sick visits (dates and duration), documented local inpatient treatments, stunting, wasting and abnormal neurodevelopment. After episodes of illnesses, follow up notes will be recorded in the diary. Mothers will also continuously subjectively assess infant well-being and encouraged to document any child sickness in the diary for comprehensive documentation. The diary will be submitted to the study on exit, at 24 months.

Infant development Neurodevelopmental outcomes will be evaluated from six weeks until two years of age using the Denver II screening tool. This tool assesses gross and fine motor abilities, development of language, social skills, attention span, general compliance and fearfulness. In addition, infant developmental milestones (e.g. time of teething, first sitting, walking or talking) will be recorded.

Maternal bio-samples Biological specimens including whole blood, plasma and serum, urine, and breast milk will be collected at all visits; stool, cord blood, amniotic fluid and placenta are collected at delivery. Sensitive samples (stool, milk, urine) will be placed at 4°C for no longer than 4 hours before processing. All samples will be stored in a bio-bank at -80°C within a maximum of 6 hours after acquisition.

Maternal stool samples will be screened for intestinal protozoa and helminthes trophozoites and ova using direct wet mount microscopy and the formal-ether concentration technique.

Determination of maternal full blood counts For assessment of full blood counts (FBC), venous blood will be collected in 5ml Ethylenediaminetetraacetic acid (EDTA) tubes by a vacutainer system and analysed using a Mindray© Haematology 3-part differential, 16 parameters BC3600 Analyser, China using whole blood. The WHO definition for diagnosis of anaemia in pregnancy will be used (haemoglobin concentration <11.0 g/dl.) Infant blood samples Infant venous blood will be collected whenever possible at 7-10 days for FBC. Whenever possible, plasma samples will be stored. Volume of blood collected will be within recommended amounts depending on the weight of the infant (i.e. ≤1-5% of total blood volume over 24 hours and ≤ 10% over 8 weeks). Blood will not be collected from severely sick infants and those with suspected severe anaemia according to clinical assessment. Dried blood spots will be collected at every visit using whole blood samples or a needle prick puncture of the heel. 3-5 drops of blood will be collected and dried on Whatman filter paper and stored at -80◦ Celsius.

Maternal blood glucose levels Screening for gestational diabetes using serum non-fasting random blood sugar levels will be done for a subgroup of women at enrolment using the Mindray BS200 Chemistry Analyser (Mindray, China), including glycosylated haemoglobulin (HbA1c) levels. A cut off for HbA1c of 6.5% will identify undiagnosed diabetes whilst an HbA1c ≥5.6% will mark individuals with an increased risk for future diabetes according to the WHO guidelines.

Maternal kidney and liver function profiles Creatinine, urea and electrolytes (sodium, potassium and chloride) will be done on mothers' serum samples using the Beckman Coulter AU680 chemistry analyser, Germany. Creatinine levels will be used to estimate the maternal glomerular filtration rate using the Modified for Diet in Renal Disease equation. Liver function and cholestasis will be assessed using serum albumin, fibrinogen, bilirubin, lactate dehydrogenase (LDH), Alanine transaminase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP) and gamma glutamyl transferase (GGT).

Infant liver and kidney function tests will be assessed in all infants at 6 and 24 months provided sufficient material was available.

Maternal lipid and bone profiles High density lipoprotein cholesterol (HDL-C), low density lipoprotein cholesterol (LDL-C), total cholesterol (TC) and TC/HDL-C ratio at baseline will be determined using serum at enrolment and at 24 months follow-up. TC will be used to calculate LDL-C when concentrations are below 4.4 mmol/L. For maternal bone profiles, longitudinal measurements of serum ionized calcium, phosphorus, albumin and ALP levels in pregnancy and at weaning will be performed.

Maternal diagnosis of HIV Maternal screening for HIV infection will be done using the serial testing algorithm according to the Zimbabwe National Guidelines for HIV Testing and Counselling. Qualitative rapid immunochromatographic assays, Standard Diagnostics (SD) Bioline HIV-1/2 3.0 (SD Inc., Kyonggi-Do, South Korea) is being used for initial screening. Positive tests are confirmed using the Alere DetermineTM HIV-1/2 kit (Abbott Diagnostics, USA. Western blotting is being used for indeterminate test results. HIV testing of HIV uninfected mothers will be repeated at 6 months intervals to identify sero-converters.

Diagnosis of infant HIV In all HIV exposed infants, early detection of HIV is being done using the COBAS1 AmpliPrep/COBAS1 TaqMan1 HIV-1 Qualitative Test (Roche Molecular System) according to manufacturer's instructions. Tests are done at every visit from delivery until weaning or diagnosis of HIV infection.

HIV-1 RNA load quantification Maternal and infant plasma samples will be quantified for HIV-1 RNA load at enrolment and 24 months, respectively using an automated TaqMan Cobas AmpliPrep Roche Diagnostics, Branchburg NJ), according to the manufacturer's instructions. The lower detection limit of the assay is 20 copies/mL.

Assessment of maternal-infant immune status Maternal absolute CD4+ T-lymphocyte counts are enumerated in EDTA blood samples for all HIV positives using Partec Cyflow counter, Germany at enrolment and at 12 months intervals. Analyses will be performed within 6 hours of blood collection.

HAART levels Levels of Efavirenz, Lamivudine, and Tenofovir in different maternal compartments including plasma, maternal milk, urine and amniotic fluid will be measured using high performance liquid chromatography.

Assessment of maternal co-infections HBV screening of all mothers and all exposed infants will be done by measuring HBV surface antigen (HBsAg), anti-HBsAg and anti-core antibodies will be detected using immunochromatographic, Pointecare Diagnostics, USA kits according to manufacturer's instructions. Maternal serum CMV DNA loads and congenital CMV will be measured using the RealStar CMV PCR kit v1.0 Germany, following isolation of viral DNA using the QIAamp MinElute Virus Spin Kit Germany.

Infants' dried blood spots will be evaluated for infection markers, CMV, HBV, HCV and HIV for all exposed infants. Anti-Treponema pallidum antibody isotypes will be used for maternal syphilis profiling using the SD Bioline 3.0 qualitative immunochromatographic kit. Rapid Plasma Reagin (RPR) test will be used for detection of Venereal disease research laboratory (VDRL) antigen containing micro-particulate carbon. Syphilis serology will be done on infant plasma at 24 months in exposed infants using the SD Bioline 3.0 qualitative immunochromatographic kit.

Immunogenetics of susceptibility to HIV and co-infections Human leukocyte antigen (HLA) and killer cell immunoglobulin-like receptors (KIRs) typing will be carried out on maternal and infant DNA samples using DNA sequencing and sequence specific primer polymerase chain reaction (PCR) as previously described previously. HLA and KIR variants will be correlated to maternal virological status, immune status and clinical outcomes.

Infant humoural immune responses to EPI vaccines Correlates of immunity against infectious diseases among HEU infants remain poorly understood. Antibody titres to measles, rubella, rotavirus, poliovirus, hepatitis B virus (HBV), diphtheria, tetanus toxoid, pertussis, pneumococcus and Haemophilus influenza type b (Hib), will be measured using commercial ELISA-based methods. Measurements will be performed after vaccination at 12, 18 and/ or 24 months. Serum titres correlating with protection (e.g. anti-rotavirus immunoglobulin A ≥20 U/mL will be assessed in HUU, HEI and HEU infants.

Assessment of infant atopic manifestations and sensitization Causes and timing of atopic dermatitis remain poorly described in SSA. Maternal allergic conditions in the third trimester will be assessed by a detailed questionnaire regarding maternal diet including food craving, maternal drug use and allergic and intestinal symptoms. Starting at delivery, the research midwife or study nurse will be physically examining all infants for the presence of atopic dermatitis defined by the presence of an itchy rash and/ or visible flexural dermatitis. Diaper rash, rash around the eyes and/ or scalp scaling will not considered diagnostic for eczema. Total and specific serum immunoglobulin E (IgE) in selected infant blood samples will be measured on an Immuno-assay analyser (Thermofisher, Freiburg, Germany) to support the clinical diagnosis of an atopic condition as previously described. Specifically, IgE antibodies against hen's egg, cow's milk, peanuts, aero-allergens, grass pollen, cat, dog dander and house dust mite will be assessed. Infants with specific serum IgE levels >0.3 IU/ml against one or more of the tested foods or inhalant allergens will be considered sensitised. Food allergies will be correlated with infant morbidity, antenatal drug use and nutrition including the food the mothers craved for in pregnancy.

Summary of sub-studies

1. Gut microbiota sub-study This sub-study aims to assess how HIV infection and HAART use affect gut microbiota maturation to understand the pathogenesis of environmental enteric dysfunction (EED. Sequential mother-infant stool samples from pregnancy to 24 months will be used to perform 16 small-subunit (S) ribosomal ribonucleic acid (rRNA) sequencing for assessment of phylum and operational taxonomic unit composition with ≥97% sequence identity, as well as microbial diversity. In selected samples bacterial full genome metagenomics shotgun sequencing to identify the microbiota metabolic potential will be done. Further, mass spectrometry will be performed for metabolome including assessment of markers of microbial translocation across the damaged gut wall such as lipopolysaccharides (LPS) and endotoxin core antibody (EndoCAb) and intestinal fatty acid binding protein (I-FABP) as a marker for enterocyte damage. Microbiota and metabolite markers will be correlated to infant growth, development and susceptibility to infectious diseases.
2. Immunometabolomics sub-study This sub-study aims to investigate at least 24 plasma inflammatory immune activation markers, using ELISA-based multiplex testing by Luminex technology, as risk factors for adverse pregnancy outcomes in HIV-infected women and subsequent impaired infant growth and development. For this study-study we will select ≥50 HIV-infected women with long-term (>7 months), medium-term (1-7 months) and no/ short-term (no or <1 months) HAART exposure and uninfected women as controls. To assess the metabolome and mitochondrial dysfunction, mass spectrometry techniques will be used to measure a wide range of metabolites including amino acids, reactive oxygen species, triglycerides, diacylglycerides, cholesterol, phospholipids, fatty acids, peroxidized lipids and oxylipids.
3. Cell-mediated immunity sub-study We hypothesise that alterations in number and function in NK cells might occur in HEU infants and predispose them to severe infections and malignancies associated with persistent viruses. In infants with sufficient blood volume of ≥2 mL, we will isolate peripheral blood derived mononuclear cells from HUU and HEU infants with short-, medium- and long-term maternal HAART exposure at 24 months to determine numbers and function of T cells and NK cells as previously described.
4. Latent TB co-infection sub-study We hypothesize that undiagnosed latent TB can partially explain increased infant mortality in infants in low-resource settings. TST have been used to screen for latent TB; however, due to an interaction with BCG vaccine which is given routinely in Zimbabwe, TST specificity may be greatly reduced in our study population. TB, IGRAs) will be used to test latent TB in a subgroup women in pregnancy and at 6 weeks after delivery and evaluated against TST. Further, placental samples will be tested for histological lesions suggestive for TB inflammation. IGRAs measurements and size of TST lesions with be correlated with adverse maternal and infant outcomes.
5. Maternal stress sub-study This will assess prevalence, risk factors and coping mechanisms of maternal stress using the perceived stress questionnaire at enrolment and 6 weeks after delivery.
6. HIV/ HAART exposure and neurodevelopmental outcomes This sub-study is evaluating potential long-term neurodevelopmental sequelae related to HIV infection and extended HAART exposure using the Mullen tool at 12 and 24 months.

Future studies might follow infants into adolescence. A parallel rural observational cohort is desirable to control for household and environmental confounders associated with the current urban based study.

Inclusion Criteria:

Consenting pregnant woman of Bantu origin of ≥15 years of age, at least 20 weeks of gestation at enrolment and planning to deliver at any of the 4 study sites, Kuwadzana, Rujeko, Glenview or Budiriro. Mothers should be willing to be followed together with their babies from delivery, and willing to provide the required data and biological specimens in follow-up visits for two years.

Exclusion Criteria:

• Presence of severe maternal mental disorders.

All pregnant women with a documented positive HIV status were encouraged to enrol in the study. For every HIV positive pregnant woman recruited, the 10th HIV negative women was recruited, accounting for the 12% HIV prevalence within this population. The pregnancy was dated by the standard way of using menstrual history together with first trimester ultrasound scan where available.

Only anonymous data will be stored at local sites and transferred between sites for collaborative research efforts. All original data will be stored locally at the UZ-CHS for at least 10 years on suitable electronic devices. Data transfer with collaborating partners will be encrypted. using EGAcryptor. With our European partners we will follow the current European General Data Protection Regulation effective May 25, 2018.

within after 10 years after completion of the study

Results will be published in journals with scientific quality assurance, and priority will be given to open access journals whenever possible. Published data will be made available to other researchers upon request and verification that additional use of data that goes beyond the studies proposed here is covered by the original approved study protocols and informed consent. The need for well -defined management and handling of research data is a priority both for national and international research organizations and for science may not over-emphasised