Glycemia Range and Offspring Weight and adiposity in response To Human milk (GROWTH) study: protocol for an observational cohort designed to study lactational programming

Abstract

Introduction

Maternal human milk feedings continue an offspring’s exposure to the programming stimuli of maternal metabolism during the postnatal period. While considerable research focuses on associations between in utero environments and offspring metabolic disease, few studies have been able to specifically measure how human milk composition modifies programming of children’s growth in conjunction with comprehensive measures of maternal glycaemia during pregnancy.

Methods and analysis

The Glycemia Range and Offspring Weight and adiposity in response To Human milk (GROWTH) Study is a longitudinal cohort enrolling women with a singleton pregnancy who (1) undergo serial testing of glycaemia during pregnancy and (2) are intending to provide their breast milk through direct breastfeeding or pumped milk as the primary nutrition for their infant. Enrolment started in October 2023 and is expected to be completed in December 2026. Key procedures include virtual lactation support visits, serial human milk sampling at three time points, maternal and infant blood sampling, serial maternal and child anthropometric measurements and diet assessment. After delivery, mother–child dyads are followed until children turn 2 years of age. The primary exposure variable is maternal glycaemia obtained from a fasting, 3 hour 100 g oral glucose tolerance test performed at 24–28 weeks of gestation, and the primary outcome measure is the composite of human milk linoleic and docosahexaenoic acid concentrations in milk samples collected at 1 month postpartum.

Ethics and dissemination

Lurie Children’s Hospital Institutional Review Board (IRB) provides central oversight of the GROWTH Study in conjunction with each participating centre’s IRB. The GROWTH Study data has the potential to inform perinatal health and future research in lactation and human milk science by providing comprehensive measures of human milk composition and early childhood growth and body composition parameters impacted by maternal metabolism in pregnancy.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• The Glycemia Range and Offspring Weight and adiposity in response To Human milk (GROWTH) study design will measure glycaemia during lactogenesis and associations with human milk composition.

• The study design allows for determination of cumulative and time-sensitive influences of both the maternal metabolic state and human milk nutrient concentrations on offspring growth and adiposity.

• Despite rigorous ascertainment of and adjustment for confounders, residual confounding is a limitation of all observational studies.

• Due to selection bias of enrolled participants, findings may not be generalisable to other populations.

Introduction

Maternal metabolism during pregnancy is a key contributor to the developmental origins of offspring metabolic disease. Hyperglycaemia in pregnancy, even below the diagnostic threshold of gestational diabetes mellitus (GDM), and overweight/obesity in pregnancy are associated with adverse metabolic outcomes in offspring including obesity and disorders of glucose metabolism.^{1 2} Importantly, as demonstrated in human epidemiologic and pre-clinical animal studies, offspring are exposed to maternal metabolic programming stimuli even after parturition through human milk (HM) feeding.³,⁶ While considerable research focuses on associations between in utero environments and offspring metabolic disease,^{2 7} less is known about how maternal metabolism influences child health through HM composition.

Bartol’s lactocrine hypothesis first suggested that factors in milk influence offspring postnatal development.⁸ Animal and human cohorts demonstrate lactational programming with associations between milk components and offspring metabolic phenotypes including postnatal growth rate, accrual of adipose tissue and later risk of insulin resistance, inflammation and metabolic syndrome.^{5 9} As the initial steps in the process of lactogenesis begin during the second half of gestation, the cellular machinery for milk biosynthesis by the mammary gland is susceptible to maternal metabolism during this period with implications on sufficiency of milk production and milk composition. Importantly, because lactation is a period for programming for the child, this also reveals it as a time window of opportunity for intervention to optimise offspring metabolic health.

Studies integrating longitudinal and detailed maternal metabolic characteristics in pregnancy, comprehensive HM profiling and maternal-child outcomes are sparse. Maternal glycaemia during pregnancy is commonly categorised as normal or abnormal based on testing at a single time point in mid-gestation, precluding a nuanced examination of maternal glycaemic exposures.¹⁰ Furthermore, few pre-birth cohorts with longitudinal HM profiling and later maternal-child outcome ascertainment exist given the costs and multi-disciplinary approach required to investigate health across obstetric, paediatric and endocrine perspectives.¹¹

Within the current clinical standards for defining GDM, both GDM and pre-pregnancy body mass index (BMI) categorised as overweight/obese have been associated with HM nutrient concentrations, including HM fatty acids.¹²,¹⁶ HM fatty acids are associated with child adiposity,¹⁷ although an ideal fatty acid profile in HM has yet to be determined. In preclinical animal models and observational human studies, we and others have shown a positive association between n-6 fatty acid concentrations in HM and adiposity markers in offspring.^{3 17 18} In contrast, maternal n-3 fatty acid supplementation leading to higher n-3 fatty acids in HM was associated with markers of adiposity in infants at some but not all times during the first year.¹⁹ Pregnancies affected by GDM and, separately, obesity have been associated with higher linoleic acid (LA; 18:2n-6) and lower docosahexaenoic acid (DHA; 22:6n-3) in HM.²⁰,²² High breastfeeding intensity after GDM exposure has been associated with offspring protection against cardiometabolic risk, yet breastfeeding after GDM has also been associated with higher offspring BMI.^{23 24} These studies did not adjust for maternal glycaemia during pregnancy, which could explain the contrasting findings. These findings suggest the need to further delineate independent associations between maternal glycaemia and offspring body composition while also accounting for the roles of HM nutrients. This would address the existing knowledge gap as to whether HM composition, specifically milk fatty acids but also other milk nutrients, can offset or augment offspring developmental pathways programmed in utero.

The Glycemia Range and Offspring Weight and adiposity in response To Human milk (GROWTH) Study is a prospective cohort study designed to address these gaps in understanding of lactational programming of offspring health in the context of metabolic exposures during pregnancy. The GROWTH Study is enrolling pregnant individuals with a stated intention to breastfeed or provide expressed milk for their infant’s nutrition. The study involves metabolic phenotyping during pregnancy with comprehensive, longitudinal glucose measurements, including oral glucose tolerance testing (OGTT) and continuous glucose monitoring (CGM), postpartum maternal metabolism and body composition, comprehensive analysis of expressed HM, and postnatal offspring growth assessment including direct measures of child adiposity through 2 years.

Objectives of the GROWTH Study

The GROWTH Study’s primary objective is to conduct prospective, detailed longitudinal profiling of HM at 1, 2 and 6 months postpartum to associate maternal glycaemia during and after pregnancy with HM factors that regulate early childhood adiposity. Data and biological specimen collection and analysis are designed to characterise the mother-milk-infant triad signalling pathways.²⁵ HM composition will also be evaluated in the context of other relevant programming exposures including pre-pregnancy BMI, maternal diet and markers of metabolism and insulin resistance into the postpartum period. The secondary objective is to measure growth trajectories and body composition in offspring at ages 1, 2 and 6 months and 2 years. The study will also measure maternal body composition through 2 years postpartum.

Primary hypothesis of the GROWTH Study

Glucose levels in pregnancy, even below the current GDM threshold, are associated with alterations in HM components that regulate offspring adiposity with a primary focus on HM fatty acids. Testing this hypothesis will address overarching questions as to whether maternal glycaemia and associated nutrition and metabolism during pregnancy alter mammary gland development to impact HM composition, which would in turn influence offspring metabolic programming as reflected in early childhood growth.

Methods and analysis

Human subjects research

Enrolment and participation in the GROWTH Study is occurring in two phases (figure 1). During Phase I, GROWTH leveraged and enrolled participants from the Glycaemic Observation and Metabolic Outcomes in Mothers and Offspring (GO MOMs) cohort. GO MOMs (NCT04860336) is an observational, multi-centre study that is conducting a longitudinal evaluation of glycaemia across the course of pregnancy with a primary aim to examine whether early pregnancy glycaemia is predictive of maternal and perinatal outcomes.²⁶ At the time that the GO MOMs study completed enrolment, the GROWTH Study had accomplished approximately one third of its enrolment target. Therefore, Phase II of the GROWTH Study adapted procedures to enrol pregnant individuals outside of the GO MOMs study. During both phases, a single Institutional Review Board (IRB) at Ann & Robert H. Lurie Children’s Hospital in Chicago, IL USA provides study oversight and approval prior to enrolment for any study procedures in conjunction with local site IRB approval. Participating parents provide signed, informed consent for themselves and their child to participate in the GROWTH Study. Enrolment began in October 2023 and is expected to be completed in December 2026 with estimated completion of all follow-up procedures to occur by December 2028.

Figure 1. GROWTH Study participant enrolment scheme during two phases. During Phase I, eligible participants expressing intention to breastfeed or provide expressed breast milk for infant feedings were identified from the cohort study GO MOMs. During Phase II, after GO MOMs completed enrolment, pregnant individuals were identified from routine prenatal appointments. After enrolment in both phases, procedures starting at 24–28 weeks of gestation through the child’s age of 2 years were similar. Two additional study visits at the child’s ages of 12 and 18 months are survey-only procedures. CGM, continuous glucose monitor; GO MOM, Glycemic Observation and Metabolic Outcomes in Mothers and Offspring; GROWTH, Glycemia Range and Offspring Weight and adiposity in response To Human milk; HM, human milk; OGTT, oral glucose tolerance test.

Study design and settings

The GROWTH Study is a multicentre, prospective, observational study for which eligibility criteria include: women with a singleton pregnancy, being a GO MOMs participant (Phase I only), age 18 years or older and intention to breastfeed or provide expressed breast milk for infant feedings. Exclusion criteria include: delivery prior to 35^0/7 weeks of gestation, non-English speaker (due to dietary recall procedures), and a lasting condition in the child known to impair feeding or growth. The complete lists of inclusion and exclusion criteria for Phases I and II are provided as online supplemental material. Three regional clinical research centres recruit and manage participants: Northwestern University/Lurie Children’s Hospital, University of Pittsburgh Medical Center Children’s Hospital of Pittsburgh/Magee Womens Hospital, Women & Infants Hospital Rhode Island (WIHRI). These three centres recruit healthy pregnant individuals without pre-existing diabetes who are either a part of GO MOMs (Phase I) or attending their routine prenatal care appointments (Phase II). Specific to Phase I, the WIHRI research centre enrolled and manages participants from Tufts Medical Center, Massachusetts General Hospital and WIHRI. The translational science centre for the study’s biorepository and detailed HM analysis, in addition to research services for maternal diet recall, occurs at the University of Michigan. Central data coordination and analysis for the entire study is managed by the Biostatistics Research Center at the Northwestern University Data Analysis and Coordinating Center in Chicago, IL.

Study schedule and data collection

Two tables provide an overview of study procedures, visit schedules and planned surveys that are completed during pregnancy (table 1) and postpartum through early childhood (table 2). The Northwestern University Data Analysis and Coordinating Centre manages database structure and storage and quality control for all participating sites primarily using the REDCap password-protected Health Insurance Portability and Accountability Act (HIPAA) compliant database system.

Table 1. GROWTH Study pregnancy-related study visits and procedures.

	Enrolment	24–28 weeks of gestation	32–36 weeks of gestation	Delivery and 30 days after
Procedures
Informed consent	✓
3 hour 100 g OGTT*		✓
CGM		✓	✓
Medical record review†				✓
Survey‡
Demographics	✓
Infant feeding intention	✓

^*Blood sampling for OGTT also includes measuring Hemoglobin A1c, insulin and lipid panel.

^†Medical record review includes data abstraction for details of pregnancy comorbidities, newborn growth measures as well as maternal or neonatal readmission during the first month.

^‡Surveys are staff administered.

CGM, continuous glucose monitor; GROWTH, Glycemia Range and Offspring Weight and adiposity in response To Human milk; OGTT, oral glucose tolerance test.

Table 2. GROWTH Study postpartum and early childhood study visits and procedures.

	Lactation visit(s)	1 month	2 months	6 months	12 months	18 months	2 years
Procedures
Virtual lactation support	✓
Human milk collection		✓	✓	✓
Maternal body composition*		✓	✓	✓			✓
Child anthropometry† and body composition*		✓	✓	✓			✓
Blood, maternal		✓
CGM, maternal		✓
Blood, child		✓					✓
Stool, child			✓	✓
Surveys‡
Return to work planning form	✓
BSES-SF	✓
24-hour diet recall, maternal		✓
Infant/child feeding and health form		✓	✓	✓	✓	✓	✓
EPDS		✓
PSS		✓		✓
BEBQ			✓	✓
CEBQ							✓
WBSS				✓
Toddler FFQ					✓	✓	✓

^*Body composition using BIA QuadScan 4000.

^†Child anthropometry: skinfold thicknesses with Harpenden callipers (triceps, subscapular, supra-iliac crest), weight, length, head circumference.

^‡Surveys are completed via REDCap or log in to online platform for FFQ.

BEBQ, Baby Eating Behaviour Questionnaire; BSES-SF, Breastfeeding Self-Efficacy Scale-Short Form; CEBQ, Child Eating Behavior Questionnaire; CGM, continuous glucose monitor; EPDS, Edinburgh Depression; FFQ, Food Frequency Questionnaire; GROWTH, Glycemia Range and Offspring Weight and adiposity in response To Human milk; PSS, Perceived Stress Scale; WBSS, Workplace Breastfeeding Support Scale.

Study procedures are designed to accomplish the following objectives in chronological order of study progress: (1) assess maternal metabolism during pregnancy with a focus on glycaemia, (2) obtain HM samples and determine HM composition, (3) obtain plasma markers of maternal and infant metabolism, (4) assess parent and child dietary intake and (5) perform standardised early childhood and maternal anthropometric measures including a direct assessment of body composition. These procedures are accomplished through on-site research visits during pregnancy, home or on-site visits by a study team after delivery, and online survey administration. Further details of procedures include:

Measuring glycaemia and metabolism during pregnancy

Two methods (OGTT and CGM) are used for measuring blood glucose excursions during pregnancy. In both Phases I and II, participants complete a fasting 100 g 3 hour OGTT at 24–28 weeks of gestation; haemoglobin A1c, insulin and lipid profiles are obtained during the OGTT. A CGM (DEXCOM G6 Pro) is placed at 24–28 weeks of gestation by study personnel and worn by the participant for 10 days. At 32–36 weeks, participants have a CGM placed again. Participants return the CGM by mail. Therefore, in both Phase I and II, monitoring by CGM occurs at 24–28 weeks and 32–36 weeks of gestation.

Virtual lactation support sessions and lactation experience assessments

Participants receive up to four virtual lactation support sessions, whether by phone or video chat based on the participant’s preference.²⁷ Coordination of visits begins approximately 48 hours after delivery, in anticipation of discharge from the post-partum hospitalisation. A minimum of one session is required by protocol, and the remainder can be spaced according to the participant’s support needs. The intention is to allow support during transition times such as back to work if applicable. Sessions last 30 min and content discussed is guided by the participant’s stated needs for support. Lactation support teams are local to each recruiting centre. Lactation support team members have certification in lactation support including Certified Lactation Specialist, Certified Breastfeeding Counsellor and International Board Certified Lactation Consultant. Some of the recruited women are at higher risk of lactation failure, thus additional support is targeted to also improve lactation success in mothers desiring to feed their infant breastmilk.

Home visits during early childhood

A two-person team provides an in-home research visit at the child’s age of 1, 2 and 6 months and 2 years. Biospecimen collection and anthropometric measures occur as specified below. If preferred by participants, on-site study visits can occur at the local study site’s medical campus.

HM sampling

Study teams collect 30 mL whole milk at the first three home visits, when the child is 1, 2 and 6 months old. Participants are instructed to provide 30 mL from a full breast expression between 08:00 and 10:00 on the morning of a visit. Detailed written instructions with diagrams are provided. Visits are coordinated to occur in the morning to ideally obtain milk within 1 hour of expression. When a study visit occurs later in the day, the timing of expressing milk for collection is shifted and participants are advised to express the sample 1 hour prior to team arrival. This prioritises reducing the interval between expression and processing. Participants document time of expression, expression mode (hand vs pump), time of most recent prior pump or breastfeed and side of collection (right vs left breast).²⁸ Expressed milk is kept refrigerated in the home and placed on ice during transport by the study team to the study centre. At each study site, aliquots of milk are created with RNAse inhibitor added to one aliquot before freezing at −80°C. These methods were designed to optimise HM for analysis of (1) fatty acid and untargeted lipidomics, (2) macronutrients and (3) mammary epithelial cell RNA/transcriptome. Residual milk aliquots are stored for future analysis. All specimens are frozen as whole milk at −80°C at the site until shipping and transport on dry ice for long term storage at −80°C in the biorepository at the University of Michigan. Biorepository freezers are equipped with a remote alarm system to ensure stable storage conditions. If lactation cessation occurs after the first month, study visits continue without the milk collection.

Maternal and infant blood sampling

At the first home visit, blood is collected from parents and infants using capillary-based blood sampling with the Tasso device (Seattle, Washington, USA) to obtain 500 μL whole blood in an EDTA microtainer.²⁹ The mother is asked to fast for at least 4 hours prior to the blood collection and the timing of the most recent meal is documented at the time of blood collection. Blood is kept on ice and on return to the research site, blood is processed to preserve maternal plasma to measure C-peptide and lipidomic profiles and infant plasma to measure lipidomic profiles. At ages 1 month and 2 years, for those who opt in, whole blood from children is preserved for future DNA extraction and analysis. Also at the first home visit, a blinded continuous glucose monitor (DEXCOM G6 Pro) is placed on the mother and worn for up to 10 days. This is returned by mail and data uploaded.

Maternal anthropometrics

Weight and height are obtained from the GO MOMs record during Phase I and measured during the first study visit at 24–28 weeks of gestation in Phase II. Weight is obtained with a calibrated scale at each post-partum visit. Bioimpedance analysis with the Bodystat Quadscan 4000 (Bodystat Ltd, Isle of Man, UK) is used to assess maternal body composition at all four home visits.

Early childhood anthropometrics

At all four home visits, portable equipment is used for measurements of weight, length, head circumference and skinfold thickness at supra-iliac crest, triceps and subscapular sites with Harpenden callipers (Baty International, UK). Bioimpedance measured by Bodystat Quadscan 4000 (Bodystat Ltd, Isle of Man, UK) assesses body composition. Study personnel undergo training and certification in obtaining study measurements using a standardised protocol across all research sites.

Maternal diet during lactation

In the 14-day period surrounding the first HM collection at the child’s age of 1 month, mothers provide two 24-hour recalls, 1 weekday and 1 weekend day. An experienced team of registered dietitians at the University of Michigan Nutrition Obesity Research Center NExT core administers recalls by phone using the US Department of Agriculture (USDA) 5-step multi-pass method.³⁰

Early childhood diet and behavioural feeding patterns

Early childhood dietary intake and feeding patterns are reported by parents. At infant age 1, 2, 6, 12 and 18 months, parents complete a modified Infant Feeding Practices Survey, third edition.³¹ Responses will be used to calculate a breastfeeding intensity score. At 12 and then 18 months and during the month prior to 2 years, parents complete a toddler food frequency questionnaire using the online platform Babyfeed.org.³² Parents complete the Baby Eating Behaviour Questionnaire (BEBQ) when the child is aged 2 and 6 months and the Child Eating Behaviour Questionnaire (CEBQ) at 2 years.^{33 34} The BEBQ is completed through an online survey, and the CEBQ is administered by research staff. Modified questions from the Infant Feeding Practices Survey, third edition, ask about the child’s recent health, hospitalisations and medication use.³¹

Surveys

Parents respond to online and staff-administered surveys throughout the study. These include reporting of baseline demographics, health habits and medication use during pregnancy. Online vs staff-administered surveys are balanced in consideration of participant burden for each type of survey and study visit duration. Surveys also assess breastfeeding intention and experiences of prenatal, immediate post-partum and ongoing lactation support. These use the Infant Feeding Intentions scale for baseline documentation.^{35 36} Subsequent surveys detail infant and toddler dietary intake and habits as previously described and as listed in table 2.

Infant stool specimens

At infant age 2 and 6 months, a stool specimen is collected by parents at home using a spatula to collect a specimen from the diaper and placed into a 2 mL flip-top Eppendorf tube. Parents keep the specimen frozen until the study team arrives for the visit and retrieves the specimen for storage at −80°C. Stool samples will be biobanked for use in ancillary studies.

Primary study exposures and outcomes

The study’s primary exposure is the OGTT sum of glucose z-scores, as a composite single glucose measure for each participant. The primary outcome considers two HM fatty acids together as a co-primary endpoint: LA and DHA in the HM sample at 1 month post-partum. These two outcomes were selected as they are modifiable through maternal diet and have been associated with child health outcomes.^{19 21 37 38} For offspring health, the primary outcome is child adiposity at age 2 years as fat mass calculated from bioelectrical impedance analysis (BIA) measures.

Secondary study exposures and outcomes

Secondary outcomes to be associated with maternal glycaemia during pregnancy include repeated measures of the HM fatty acid concentrations at 1, 2 and 6 months post-partum to assess the trajectory of HM metabolites. Additional analyses will measure associations between maternal exposures (eg, CGM excursions, maternal insulin sensitivity and secretion, plasma triglycerides, BMI, diet) and other relevant measures of HM composition (n-6:n-3 fatty acid ratio, and HM LA, palmitic acid, di-homo-gamma-linolenic acid, DHA, macronutrients, and lactose). Analysis will also test for associations between HM fatty acids and repeated measures of child adiposity from sum of skinfold thicknesses, BIA-reported body composition measures at 1, 2, 6 months and age 2 years, and BMI percentile at age 2 years.

Analysis

Power calculations for the GROWTH Study account for the coefficient of determination (RC2) between the OGTT sum of glucose z-scores, as a composite single glucose measure for each participant, and HM fatty acids. More specifically, these power considerations focused on detecting a clinically meaningful association between the maternal OGTT composite glucose at 24–28 weeks of gestation and HM LA and DHA concentrations. LA and DHA are considered as co-primary endpoints in analysis. Calculations also account for other model covariates. Preliminary data correlating OGTT and glucose challenge levels with fatty acid concentrations were used to inform sample size calculations. When n=375 (including loss to follow-up as specified below) and RC2=0.4, there is 80% power at two-sided 0.025 alpha level to detect a slope of 0.488 in the relationship between the predictor OGTT glucose sum of z-scores and the outcome HM LA. In other words, a glucose sum of z-scores higher by 1 unit would be associated with LA concentration higher by 0.488 units. When n=375 and RC2=0.4, there is 80% power at two-sided 0.025 alpha level to detect a slope of 0.068 in the relationship between the predictor OGTT glucose sum of z-scores and the outcome HM DHA. Sample size estimates assumed a conservative SD of 2 units for glucose sum of z-scores as this variable is the sum of 4 standardised normal variables. Calculations assumed a two-sided type I error rate of 0.025 to account for two primary outcomes (0.05/2). After inflating this sample size to allow for inclusion of potential confounders in multivariable modelling and loss to follow-up, the necessary sample size is 375. Thus, our target enrolment inclusive of Phases I and II will be 375 mother-infant pairs, anticipating and allowing for approximately 20% potential loss to follow-up.

Characteristics of the participant dyads will be summarised. Primary analyses will employ cross-sectional multivariable linear regression models. Secondary analyses will incorporate fixed effects for each exposure, measurement time, as well as their interaction, to assess whether the trajectory of HM concentrations differs by maternal pregnancy glycaemia. Individual-level random effects will also be considered. Each co-primary endpoint analysis will use a Bonferroni corrected two-sided type I error rate of 0.025 to control the overall error at 5%. Secondary analyses will employ a false discovery rate error control strategy to account for multiple testing.

Enhancing reproducibility of findings

The multi-centre aspect of the GROWTH Study will allow analyses to account for differences in HM composition which are known to vary by geographic location.³⁹ To further enhance interpretation of findings and address reproducibility of research findings, the GROWTH Study’s original design and funding will allow HM fatty acid and lipidomic profiling in 100 samples collected from participants in a separate observational cohort, the Maternal Obesity, Breast Milk Composition, and Infant Growth (MILK) Study.⁴⁰,⁴² The MILK Study enrolled participants through the University of Minnesota and University of Oklahoma Health Sciences Center. HM collection methods were similar to the GROWTH Study, including morning collections. Samples selected for analysis and interpretation through the GROWTH Study will enrich the number of participants with GDM.

Patient and public involvement

Participants of the GROWTH Study were not involved in study design. At the end of the GROWTH Study, all participants will receive a letter explaining the results of the study and where the results have been published.

Discussion

The impact of maternal metabolism on mammary gland function and HM composition is currently poorly defined. This pertains to glycaemia across all ranges detected during pregnancy. This is important to address because lactogenesis initiates mid-way through gestation, meaning that the mammary gland may be vulnerable to programming as it is transforming to prepare for the process of producing milk. Downstream from in utero programming is lactational programming, the metabolic programming of offspring that occurs specifically during breastfeeding or feeding of expressed HM and by HM composition.^{5 38} The evidence of prenatal and perinatal programming of childhood metabolic diseases including obesity and disorders of glucose metabolism is mounting, yet human studies in the field of lactational programming are sparse. Furthermore, results of studies examining the potential for breastfeeding after GDM to protect from the programming exposures that occurred in utero are sparse and conflicting. Studies designed to measure programming signals communicated across the mother-milk-infant triad are needed.^{11 25} The GROWTH Study will address this need by providing unique and comprehensive measures that start during gestation to better understand the independent effects of glycaemia during pregnancy and, subsequently, nutrient intake and feeding during the postnatal period on the programming of offspring adiposity. Research-based use of CGM during pregnancy is rapidly expanding with the goal of improving glycaemic control of diabetes in pregnancy.⁴³ However, the application to lactation-based research remains sparse.⁴⁴ Measurement of infant and early childhood body composition to evaluate adiposity gain is being performed with BIA. This tool has the advantage of being portable and non-invasive, yet has limitations. BIA is sensitive to individual hydration levels and there is a lack of validation studies in young children.⁴⁵ Fat mass will be calculated from BIA measures using the equation developed by Lingwood et al which has been validated against the PEA POD Air Displacement Plethysmography System.⁴⁶ Anthropometry with skinfold thicknesses is also being measured at each child study visit as an additional measure of adiposity.

While this report focuses on the GROWTH Study’s clinical procedures, translational studies in vitro will also be accomplished and test developmental regulation of adiposity. These studies will test ex-vivo exposures of a human infant pre-adipocyte strain to fatty acid profiles mimicking those measured in HM from GROWTH Study participants. This information will enhance understanding of how the lactation period can alter the trajectory of metabolic programming established in utero. Results of this study will advance the fields of nutrition and HM science as to how metabolic diseases impact HM composition and by identifying health-promoting components of HM. Opportunities for targeted interventions in high-risk dyads, perhaps dietary guidance to enhance HM fatty acid composition, may also be revealed through further establishment that metabolic programming of offspring is the result of an accumulation of signals during pregnancy and also lactation.

Ethics and dissemination

The IRB of Ann & Robert H. Lurie Children’s Hospital first approved this study protocol on December 27, 2022 under IRB # 2023–5748 and remains the central IRB. The GROWTH Study’s Lurie Children’s site is the coordinating centre and manages training and certification of staff across all participating centres for consistency in study procedures. IRB-approved communication platforms are used for communication with participants. De-identified biospecimens are stored at the GROWTH Study’s biorepository and translational science centre at the University of Michigan. Manuscripts will report outcomes for the primary and secondary objectives promptly after the study conclusion. After publication, a limited dataset will be available in the National Institute of Child Health and Human Development (NICHD) Data and Specimen Hub.