Biological Impacts of Rising Temperatures on Maternal, Fetal, and Newborn Health: A Cohort Study (BIRTH Cohort)

Background Anthropogenic climate change has led to a significant rise in global temperatures, resulting in more frequent, intense, and prolonged heat events. This has profound health implications for vulnerable populations, particularly in low-resource settings. Pregnant women are particularly susceptible, as heat exposure can disrupt maternal thermoregulation and fetal development, potentially leading to complications such as preeclampsia, gestational diabetes, hypertensive disorders, cardiovascular events, and increased risks of preterm birth, low birth weight (LBW), small-for-gestational-age (SGA) births, and stillbirths. Studies have indicated a 1.04-fold increase in preterm birth odds per 1°C rise in temperature, with even greater risks during heatwaves. Additionally, associations have been observed between heat exposure and stillbirth, congenital anomalies, and gestational diabetes. Despite growing epidemiological evidence, there is limited data on the biological mechanisms driving these associations, particularly in low- and middle-income countries (LMICs). Several biological pathways have been proposed to explain the impact of heat on pregnancy and birth outcomes. These include heat-induced hyperthermia in early pregnancy associated with neural tube defects. Exposure to high temperatures may lead to placental insufficiency due to altered blood flow. Chronic uteroplacental insufficiency may cause fetal hypoxia and intrauterine growth retardation (IUGR) and small for gestational age (SGA) newborns. Other heat stress-related pathways include oxidative stress, inflammatory responses, and prostaglandin-mediated change. In late pregnancy, heat exposure may induce early labor and preterm birth through release of oxytocin and prostaglandins. Increased maternal core temperature can impair fetal oxygenation and growth by redistributing blood flow away from the uterus. Study Rationale Despite growing concern over the impact of extreme heat on maternal and fetal health, evidence on the biological pathways in humans remains sparse and inconsistent, largely due to challenges in isolating heat stress as an independent exposure during pregnancy. While animal studies and limited human experimental research have identified physiological changes under heat stress, the specific pathophysiological mechanisms linking heat exposure to adverse pregnancy and birth outcomes in humans remain poorly understood. Notably, most available evidence stems from high-income countries, despite the fact that the burden of preterm birth, stillbirth, and fetal growth restriction is disproportionately higher in low- and middle-income countries (LMICs). This study aims to address these gaps by exploring the potential biological pathways linking heat exposure to adverse maternal and birth outcomes in a high-risk LMIC setting. In the Sindh province of Pakistan, where this study is based, pregnant women are frequently exposed to extreme temperatures both outdoors and indoors. They often engage in manual agricultural work, care for livestock, and perform household chores like cooking over open fires, often without access to cooling mechanisms. These intersecting vulnerabilities-physical exertion, prolonged heat exposure, and limited structural protection-contribute to their heightened risk. Study Objectives The primary objectives of the study are: * To assess the acute and sustained effects of environmental heat exposure on pregnancy and birth outcomes in rural and peri-urban areas of Pakistan. * To assess the bio-physiological pathways through which heat stress affects maternal, fetal, and infant health across different gestational ages in rural and peri-urban areas of Pakistan. The secondary objectives are: • To evaluate how gestational age, socio-demographic, maternal and nutritional factors modify the association between heat exposure and adverse pregnancy outcomes in rural and peri-urban areas of Pakistan. METHODS Study Design and Study Sites A prospective cohort study will be conducted among pregnant women residing in three districts of Sindh province of Pakistan: Matiari, Tando Muhammad Khan (TMK), and Tharparkar. These regions experience extreme summer temperatures and face multiple socio-economic challenges, including poor infrastructure, food insecurity, and high rates of maternal and neonatal health complications. Participants will be enrolled in their first trimester and followed through delivery and up to 12 months postpartum. The study will integrate environmental measurements, wearable sensors, imaging studies and biological samples to assess physiological, behavioral, and environmental responses to heat stress. Sample Size To detect a 15% increase in the risk of LBW among pregnant women exposed to heat stress compared to an unexposed cohort (1:2 ratio), and assuming 80% power, a 5% significance level, and a 20% attrition rate, 6,000 participants will be required. This calculation is based on a baseline LBW prevalence of 26.2% reported in the 2018 National Nutrition Survey for the rural population. The sample size for SVN under the same assumptions is 2,106 based on a prevalence of 46%. For the detailed sub-study, sample size calculations were based on various key physiological parameters assessed in controlled settings, including changes in skin temperature, maternal heart rate, blood pressure, cardiac output, and fetal heart rate. To detect a 10% change in these parameters, the maximum required sample size is 200 participants, assuming 90% power, a 5% significance level, and a 20% attrition rate. However, due to the limited scope of previous studies conducted under controlled conditions and the anticipated higher attrition rate, we have chosen a sub-study sample size of 1,000 women. Participant Recruitment and Enrollment Pregnant women will be identified through community-based outreach in coordination with Lady Health Workers (LHWs). Field teams will conduct site mapping and community engagement to identify women who are newly married, within 1-2 years of their last pregnancy, or planning conception. Pregnancy status will be confirmed through rapid urine tests, followed by ultrasound verification conducted by study physicians at designated study offices. After obtaining informed consent, a recruitment form will be completed. At study offices, antenatal assessment, ultrasound imaging, and biological sample collection will be performed. For the sub-study on physiological monitoring, a subset of participants will be enrolled based on their consent to wear wearable devices and allow skin temperature assessment. Transportation assistance will be provided to facilitate participation in study activities. Data Collection Procedures Baseline Assessments At enrollment, gestational age will be estimated using fetal crown-rump length via ultrasound, based on the INTERGROWTH-21st protocol (±7 days accuracy). Baseline data will include sociodemographic information, medical and obstetric history, housing characteristics, occupational exposure, and anthropometric measurements. A standardized 24-hour dietary recall will be conducted at each antenatal care (ANC) visit, capturing information on food intake, frequency, meal composition, and seasonal variations. Follow-Up Assessments Serial ultrasounds will be conducted at 12-14, 18-22, 28-32, and 36-37 weeks' gestation, including Doppler and fetal echocardiography. Maternal blood and urine samples will be collected to assess markers such as complete blood count (CBC), hemoglobin, ferritin, glucose tolerance, dehydration status, and biological stress indicators. Samples will be handled using standardized protocols, transported in temperature-controlled containers (2-8°C), and stored at the Aga Khan University Nutrition Research Laboratory (NRL), Karachi, for analysis. Environmental Exposure Monitoring Ambient indoor temperature will be recorded continuously using wearable temperature/humidity loggers. Wet Bulb Globe Thermometers will provide outdoor environmental indices, accounting for multiple heat stress factors (e.g., radiant temperature, humidity, wind speed, radiation and clothing) and to calculate WBGT for each participant throughout their pregnancy. A subset of participants will wear smartwatches to monitor heart rate, sleep patterns, and physical activity throughout pregnancy. Skin temperature will be recorded using i-Button loggers worn in customized abdominal belts. A field team will oversee device function and data transfer fortnightly. Delivery and Postnatal Follow-Up: At delivery, newborn weight, length, mid-upper arm circumference (MUAC), and head circumference will be recorded by trained staff within 48 hours of birth. A delivery notification system and routine calls when the participant is near term will ensure timely outcome capture. For the sub-cohort, placental tissue and cord blood will be collected within 10-15 minutes post-delivery, snap-frozen, and stored at -80°C. Additional bio-specimens (stool and sputum) will be archived for future analyses, including epigenetic and molecular analyses. Breastmilk sampling will also be done for the sub-cohort to understand the effect of high temperatures on breastmilk quality and composition. The mother-infant dyad will be followed for twelve months post-delivery, with monthly assessments of anthropometry, feeding practices, immunization, morbidity, and neurodevelopment. Key maternal and newborn health outcomes will be tracked, including pregnancy complications, gestational age at delivery, birth weight, infant growth, and maternal and neonatal mortality. Training of Study Staff and Quality Assurance Experienced local personnel, including physicians, sonologists, nurses/ midwives, and data collectors, will be hired. Study doctors will have prior obstetrics and gynecology training, and experience in obstetric ultrasound. They will receive a 3-month training in ultrasound and maternal echocardiography at Aga Khan University (AKU), alongside completion of the International Society of Ultrasound in Obstetrics and Gynecology (ISUOG) ultrasound course. Each study site will have three teams (two data collectors and one supervisor per team). Centralized training will include orientation to the study protocols, a question-by-question tool review, and anthropometry training per Food and Nutrition Technical Assistance III (FANTA III) guidelines. Field practice and pilot study will be conducted. Personnel operating the wearable devices will receive additional training in handling the devices and regular data retrieval. Maternal and infant anthropometry will follow INTERGROWTH-21st standards. Maternal weight will be measured with Seca 874 U scales (precision: 0.1 kg). Newborn weight will be recorded within 48 hours using Seca 354 infant scales. Infant recumbent length will be measured with an Infantometer; head circumference and MUAC will be recorded to the nearest millimeter. All equipment will be calibrated biweekly. Maternal weight will be recorded twice by the same anthropometrist, and infant measurements will be taken independently by two blinded anthropometrists. Statistical Analysis Ambient heat exposure during pregnancy for each participant will be assessed using Mean Heat Index calculated from wearable temperature and humidity loggers using the Rothfusz regression equation \[67\]. Exposure will be examined across gestational windows relevant to fetal development for exploratory analyses. Descriptive statistics will summarize maternal characteristics, environmental exposures, and birth outcomes. For primary analyses, associations between heat exposure and adverse birth outcomes low birth weight (LBW) and small vulnerable newborns (SVN) will be assessed using multivariable logistic regression for binary outcomes and multivariable linear regression for continuous birth weight. Heat exposure will be modelled primarily as a continuous variable. All models will adjust for maternal age, socioeconomic status, gestational age at delivery, parity, and season. Effect modification by maternal characteristics and season will be examined. Exploratory analyses will include trimester-specific models to identify sensitive exposure windows. Sensitivity analyses excluding multiple births will assess the robustness of findings among singletons. Alternative exposure definitions will be tested for consistency, including (i) number of days with daily maximum temperature \>40°C \[68\], and (ii) the 90th percentile of the temperature distribution \[69\]. Exploratory analyses will evaluate additional exposure definitions to determine which metrics demonstrate the strongest associations with outcomes and could be recommended for future use or validation. Distributed lag and time-varying exposure models will be explored to assess delayed or cumulative effects across gestation. Potential non-linear exposure-response relationships will be examined using Generalized Additive Models (GAMs), with results reported as supplementary to the primary linear models. Given multiple exposure definitions in exploratory analyses, findings will be interpreted cautiously, focusing on effect sizes, confidence intervals, and biological plausibility rather than p-values alone. For a subset of the study participants, mediation analysis will be conducted to investigate potential biological pathways linking heat exposure to fetal growth restriction, focusing on maternal stress response, inflammation, and placental dysfunction. All mediation models will adjust for key confounders. Statistical analyses will be performed using Stata version 18. Data Quality Assurance Data will be collected using handheld devices equipped with a dedicated data collection application, enabling real-time capture and secure storage of participant information at enrollment and follow-up. At the end of each day, data will be transmitted via a secure internet connection to the Aga Khan University (AKU) server. In areas where internet access is unavailable, the team leader will export the data manually to a password-protected device to ensure no data is lost. Environmental data from wearable and fixed devices will be retrieved periodically by designated staff and uploaded to the server. All data will be archived in a secure repository with access restricted through AKU LAN authentication, and user access will be managed according to assigned roles. A remote backup system, synchronized daily, will safeguard data against loss or corruption. To ensure data quality, automated validation checks will be integrated into the data entry platform to flag out-of-range values and inconsistencies, particularly in anthropometric measurements and dietary data. A dedicated sonologist will centrally oversee field-based imaging activities and conduct daily quality assessments of ultrasound, Doppler, and echocardiography data to ensure image clarity and diagnostic accuracy. Standard Operating Procedures (SOPs) will be disseminated to all study teams and strictly followed across all stages, including participant recruitment, data collection, data management, and analysis. Biochemical assays will also adhere to standardized protocols with appropriate internal quality control measures. Ethical Considerations The study protocol has been reviewed and approved by the Ethical Review Committee (ERC) of the Aga Khan University and National Bioethics Committee (NBC). Informed consent will be obtained from all participants prior to enrollment. Data confidentiality and participant privacy will be strictly maintained throughout the study. At each study and follow-up visit, participants will be screened by study physicians for any health issues, including high blood pressure, dehydration, or pregnancy-related complications. Those requiring care will be referred to appropriate healthcare facilities with transportation support. Referral pathways will be in place for managing complications and emergencies at healthcare facilities. Incidental findings, such as cardiac or fetal abnormalities detected during assessments, will be disclosed to participants and referred for specialized care. Pregnancy outcome data will be collected only with participant consent, and all information will be kept confidential.