Advertisement

Gestational Diabetes Mellitus Is Associated with Altered Abundance of Exosomal MicroRNAs in Human Milk

      Highlights

      • Gestational Diabetes Mellitus is associated with reduced abundance of human milk exosomal microRNAs involved in important metabolic processes.
      • Highly abundant microRNAs in breast milk (miR-148a and miR-30b) are associated with infant growth and adiposity early in infancy.
      • Higher maternal diet quality is associated with increased abundance of each of measured miRNAs (miR-148a, miR-30n, miR-let-7a and miR-let-7d).
      • Human Milk miRNAs may be a therapeutic target to mitigate risk of metabolic outcomes in offspring of women with gestational diabetes mellitus.

      Abstract

      Purpose

      Human milk (HM) is a unique biological fluid that is enriched with a variety of factors, including microRNAs (miRNAs) that potentially provide both short- and long-term benefits to the infants. miRNAs are packaged within exosomes, making them bioavailable to infants. Gestational diabetes mellitus (GDM) may affect the abundance of exosomal miRNAs in HM, providing a mechanism for growth and adiposity variation in infants of mothers with GDM in early life. Therefore, the purposes of this study were to examine the impact of GDM on select miRNAs (miRNA-148a, miRNA-30b, miRNA-let-7a, and miRNA-let-7d) involved in metabolism and to examine the association of these miRNAs with measures of infant body composition in the first 6 months of life.

      Methods

      Milk samples were collected from a cohort of 94 mothers (62 mothers without GDM and 32 mothers with GDM) matched on body mass index strata at 1 month post partum. miRNA abundance was measured by real-time polymerase chain reaction. Linear regression models were used to examine potential differences in miRNA abundance in women with and without GDM, testing associations between miRNA abundance and infant growth and body composition measures from 1 to 6 months.

      Findings

      The abundances of miRNA-148a, miRNA-30b, miRNA-let-7a, and miRNA-let-7d were reduced in milk from mothers with GDM. Independent of GDM status, higher maternal diet quality was associated with increased abundance of each of the measured miRNAs. miRNA-148a was negatively associated with infant weight, percentage of body fat, and fat mass, whereas miRNA-30b was positively associated with infant weight and fat mass at 1 month of age. There was no association of milk miRNA-148a and miRNA-30b with infant weight at 1 month of age or with body composition measures at 3 months of age; however, miRNA-148a was negatively associated with infant weight at 6 months of age.

      Implications

      If supported by randomized dietary supplementation or other intervention trials, HM miRNAs may be a therapeutic target to mitigate risk of metabolic outcomes in offspring of women with GDM.

      Abbreviations:

      miRNA (microRNA), GDM (Gestational Diabetes Mellitus), HM (Human milk), NW (Normal weight), OW/OB (Overweight/Obese)

      Key words

      Introduction

      Nutrition in the early postnatal period plays an important role in infant growth and the developmental programming of metabolic disease susceptibility later in life.
      • Lucas A
      • Fewtrell MS
      • Cole TJ.
      Fetal origins of adult disease-the hypothesis revisited.
      ,
      • Plagemann A.
      A matter of insulin: developmental programming of body weight regulation.
      Breastfeeding is an ideal source of nutrients, playing a crucial role in nutritional programming in the growing infant.
      • Ellsworth L
      • Harman E
      • Padmanabhan V
      • Gregg B.
      Lactational programming of glucose homeostasis: a window of opportunity.
      The benefits of human milk (HM) may be mediated by a variety of bioactive factors, such as hormones, immunoglobulins, oligosaccharides, growth factors, and micro-RNAs (miRNAs),
      • Hamosh M.
      Bioactive factors in human milk.
      influencing infant growth and body composition in early life.
      • Fields DA
      • George B
      • Williams M
      • Whitaker K
      • Allison DB
      • Teague A
      • et al.
      Associations between human breast milk hormones and adipocytokines and infant growth and body composition in the first 6 months of life.
      • Fields DA
      • Schneider CR
      • Pavela G.
      A narrative review of the associations between six bioactive components in breast milk and infant adiposity.
      • Shah KB
      • Chernausek SD
      • Garman LD
      • Pezant NP
      • Plows JF
      • Kharoud HK
      • et al.
      Human milk exosomal microRNA: associations with maternal overweight/obesity and infant body composition at 1 month of life.
      miRNAs are small, noncoding RNAs that bind to complementary sequences within the 3’ untranslated region of messenger RNAs (mRNAs), modulating protein production.
      • Melnik BC
      • Schmitz G.
      MicroRNAs: Milk's epigenetic regulators.
      • Djuranovic S
      • Nahvi A
      • Green R.
      miRNA-mediated gene silencing by translational repression followed by mRNA deadenylation and decay.
      • Lewis BP
      • Burge CB
      • Bartel DP.
      Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets.
      miRNAs play a role in multiple biological processes, such as adipocyte differentiation, insulin signaling, and glucose homeostasis.
      • Mirra P
      • Nigro C
      • Prevenzano I
      • Leone A
      • Raciti GA
      • Formisano P
      • et al.
      The destiny of glucose from a microRNA perspective.
      HM is one of the richest sources of miRNAs.
      • Alsaweed M
      • Hartmann PE
      • Geddes DT
      • Kakulas F.
      MicroRNAs in breastmilk and the lactating breast: potential immunoprotectors and developmental regulators for the infant and the mother.
      The abundance of miRNAs in HM is affected by various maternal factors, such as nutrition, disease status, and specifically obesity during pregnancy. Our group recently found that the abundance of selected miRNAs involved in adipogenesis and insulin signaling is decreased in HM exosomes obtained from women with overweight or obesity.
      • Shah KB
      • Chernausek SD
      • Garman LD
      • Pezant NP
      • Plows JF
      • Kharoud HK
      • et al.
      Human milk exosomal microRNA: associations with maternal overweight/obesity and infant body composition at 1 month of life.
      However, to our knowledge, the impact of gestational diabetes mellitus (GDM) on HM exosomal miRNA composition and abundance has not been examined.
      Evidence indicates that exposure to GDM increases the risk of obesity and type 2 diabetes (T2DM) in offspring
      • Clausen TD
      • Mathiesen ER
      • Hansen T
      • Pedersen O
      • Jensen DM
      • Lauenborg J
      • et al.
      High prevalence of type 2 diabetes and pre-diabetes in adult offspring of women with gestational diabetes mellitus or type 1 diabetes: the role of intrauterine hyperglycemia.
      ,
      • Dabelea D
      • Hanson RL
      • Lindsay RS
      • Pettitt DJ
      • Imperatore G
      • Gabir MM
      • et al.
      Intrauterine exposure to diabetes conveys risks for type 2 diabetes and obesity: a study of discordant sibships.
      ; however, this risk is decreased by exclusive breastfeeding for at least the first 3 months of life.
      • Mayer-Davis EJ
      • Rifas-Shiman SL
      • Zhou L
      • Hu FB
      • Colditz GA
      • Gillman MW.
      Breast-feeding and risk for childhood obesity: does maternal diabetes or obesity status matter?.
      ,
      • Schaefer-Graf UM
      • Hartmann R
      • Pawliczak J
      • Passow D
      • Abou-Dakn M
      • Vetter K
      • et al.
      Association of breast-feeding and early childhood overweight in children from mothers with gestational diabetes mellitus.
      Given that infant formulas are deficient in exosomes and their miRNAs,
      • Melnik BC.
      Milk exosomal miRNAs: potential drivers of AMPK-to-mTORC1 switching in beta-cell de-differentiation of type 2 diabetes mellitus.
      miRNAs may be one of the mechanisms by which breastfeeding decreases future risk of obesity and T2DM in infants born to mothers with GDM. Accumulating evidence indicates that diabetes during pregnancy (both pregestational and GDM) can not only alter the composition of macronutrients and total energy content of HM but also alter the composition of bioactive factors, such as insulin, adiponectin, and ghrelin, in HM.
      • Peila C
      • Gazzolo D
      • Bertino E
      • Cresi F
      • Coscia A.
      Influence of diabetes during pregnancy on human milk composition.
      However, the impact of maternal GDM on exosomal miRNAs in HM and their subsequent association with infant growth is poorly understood. Studies have found that miRNA-148a, miRNA-30b, and miRNA-let-7a are highly abundant in HM and are involved in important metabolic pathways, such as insulin signaling and adipogenesis.
      • Shah KB
      • Chernausek SD
      • Garman LD
      • Pezant NP
      • Plows JF
      • Kharoud HK
      • et al.
      Human milk exosomal microRNA: associations with maternal overweight/obesity and infant body composition at 1 month of life.
      ,
      • Mirra P
      • Nigro C
      • Prevenzano I
      • Leone A
      • Raciti GA
      • Formisano P
      • et al.
      The destiny of glucose from a microRNA perspective.
      ,
      • Wagschal A
      • Najafi-Shoushtari SH
      • Wang L
      • Goedeke L
      • Sinha S
      • deLemos AS
      • et al.
      Genome-wide identification of microRNAs regulating cholesterol and triglyceride homeostasis.
      • Munch EM
      • Harris RA
      • Mohammad M
      • Benham AL
      • Pejerrey SM
      • Showalter L
      • et al.
      Transcriptome profiling of microRNA by Next-Gen deep sequencing reveals known and novel miRNA species in the lipid fraction of human breast milk.
      • Tingo L
      • Ahlberg E
      • Johansson L
      • Pedersen SA
      • Chawla K
      • Saetrom P
      • et al.
      Non-coding RNAs in human breast milk: a systematic review.
      We recently found that abundance of miRNA-148a, miRNA-30b, and miRNA-let-7d is lower in HM obtained from mothers with overweight or obesity.
      • Shah KB
      • Chernausek SD
      • Garman LD
      • Pezant NP
      • Plows JF
      • Kharoud HK
      • et al.
      Human milk exosomal microRNA: associations with maternal overweight/obesity and infant body composition at 1 month of life.
      Thus, in the present study, we chose these 4 miRNAs (miRNA-148a-3p, miRNA-30b-5p, miRNA-let-7a-5p, and miRNA-let-7d-5p) with the a priori hypothesis that GDM is associated with altered abundance of miRNAs that are involved in important metabolic pathways in HM. The objectives of the study were to compare the abundance of these exosomal miRNAs in HM collected at 1 month post partum from women with and without GDM, controlling for maternal prepregnancy body mass index (BMI) and other potential confounders, and to test associations between HM miRNA abundances at 1 month post partum and infant growth and body composition during the first 6 months of life.

      Materials and Methods

      Study Population

      All research procedures and protocols were approved by the institutional review board at the University of Minnesota. The participants in this study were a subset of a larger study entitled Maternal Metabolism, Milk, and the Microbiome (4M Study), which included 35 women with GDM and 150 women without GDM. The 4M Study included pregnant women with singleton pregnancy 21 to 45 years of age at delivery, with an intention to exclusively breastfeed for a minimum of 3 months, and, if multiparous, who achieved successful breastfeeding for at least three months. GDM was defined based on serum glucose laboratory results in the medical record following a 2-step GDM screening procedure. Specifically, at 26 to 28 weeks of gestation, a nonfasting, 1-hour, 50-g oral glucose challenge test was performed. If the serum glucose level was ≥130 mg/dL, a subsequent 3-hour, 100-g oral glucose tolerance test was performed, with GDM defined using the Carpenter-Coustan criteria (ie, if serum glucose exceeded ≥2 of the following levels: 95 mg/dL, 180 mg/dL, 155 mg/dL, and 140 mg/dL at the fasting, 1-hour, 2-hour, and 3-hour time points, respectively).
      • Carpenter MW
      • Coustan DR.
      Criteria for screening tests for gestational diabetes.
      Women were excluded if they used tobacco, consumed >1 alcoholic drink per week during pregnancy or lactation, or had history of preexisting diabetes. Other exclusion criteria included preterm or postterm birth (gestational age <37 or >42 weeks), an infant requiring neonatal intensive care unit admission for >48 hours, infant birth weight <2500 g, or a congenital defect in the infant that is known to impact feeding and growth. For the purposes of the present analysis, we first grouped participants by prepregnancy BMI group (obese [prepregnancy BMI ≥30 kg/m2], overweight [prepregnancy BMI of 25–<30 kg/m2], and normal weight [prepregnancy BMI of 18.5–<25 kg/m2]). For each of these BMI strata, we included all women with GDM and then randomly selected twice as many women without GDM from the same stratum, for an approximately 1:2 case-control ratio in each stratum. For statistical analysis, the 2 higher BMI groups were collapsed, for a total of 4 groups: normal weight non-GDM, overweight/obese non-GDM, normal weight GDM, and overweight/obese GDM.

      Maternal Characteristics and Dietary Components

      Clinical characteristics of maternal participants, including parity (0 or ≥1), mode of delivery (vaginal delivery or cesarean section), first recorded maternal weight and height in the medical record within 6 weeks from conception (dated using last menstrual period), gestational weight gain (weight at admission for delivery minus the first recorded maternal weight), maternal race (White or other), and maternal age (years), were collected from the electronic health records. Maternal dietary intake data were collected during the third trimester of pregnancy and at 1 and 3 months post partum using the Diet History Questionnaire II, a food frequency questionnaire, as described previously.
      • Tahir MJ
      • Haapala JL
      • Foster LP
      • Duncan KM
      • Teague AM
      • Kharbanda EO
      • et al.
      Higher maternal diet quality during pregnancy and lactation is associated with lower infant weight-for-length, body fat percent, and fat mass in early postnatal life.
      Maternal diet quality during pregnancy and lactation was estimated using the Healthy Eating Index.
      • Krebs-Smith SM
      • Pannucci TE
      • Subar AF
      • Kirkpatrick SI
      • Lerman JL
      • Tooze JA
      • et al.
      Update of the Healthy Eating Index: HEI-2015.

      Breast Milk Collection

      Mother-infant pairs reported to the study site at 1, 3, and 6 months (±5 days) post partum, between 8:00 and 10:00 am, and at least 1.5 hours since the last infant feeding. A prefeeding infant weight was obtained using a high-sensitivity scale (Seca 728, Seca, Birmingham, United Kingdom). Mothers then breastfed their infants ad libitum from both breasts. Two hours after feeding, a complete single-breast milk expression was obtained using a hospital-grade pump according to a standard protocol described previously.
      • Fields DA
      • George B
      • Williams M
      • Whitaker K
      • Allison DB
      • Teague A
      • et al.
      Associations between human breast milk hormones and adipocytokines and infant growth and body composition in the first 6 months of life.
      Collected breast milk was gently mixed, aliquoted, and stored at −80 °C within 20 minutes of collection. For this study, only the 1-month milk sample was assayed for miRNA abundance.

      Infant Growth and Body Composition

      At birth and 1, 3, and 6 months, infant length was measured using the Seca 416 infantometer. Infant naked weight was measured using the high-sensitivity scale embedded in the Pea Pod (COSMED USA Inc., Concord, California). Age- and sex-specific length for age z score (LAZ) and weight for age z score (WAZ) were subsequently calculated using the World Health Organization z score classification system for term infants.
      Group WHOMGRS
      WHO Child Growth Standards based on length/height, weight and age.
      Infant percentage body fat (BF), fat mass (FM), and fat-free mass (FFM) were assessed using Pea Pod at 1 and 3 months and using dual x-ray energy absorptiometry at 6 months, with the infant in a supine position, wearing only a disposable diaper, and swaddled in a light cotton blanket.
      • Mazess RB
      • Barden HS
      • Bisek JP
      • Hanson J.
      Dual-energy x-ray absorptiometry for total-body and regional bone-mineral and soft-tissue composition.

      Exosome Isolation and miRNA Analysis

      Samples (2 mL) of HM were thawed on ice and centrifuged twice (1200 × g, 4°C, 10 minutes) to remove fat, cells, and large debris. Defatted supernatant was centrifuged again (21,500 × g, 4°C, 50 minutes) to remove residual fat and casein and subsequently passed through 0.22-μm filters to remove residual cell debris, as previously described.
      • Xi Y
      • Jiang X
      • Li R
      • Chen M
      • Song W
      • Li X
      The levels of human milk microRNAs and their association with maternal weight characteristics.
      Exosomes were extracted from 250 μL of defatted filtered milk using ExoQuick Exosome Precipitation Solution (System Biosciences, Mountain View, California) according to the manufacturer's protocol. Total RNA was extracted from the isolated exosomes using the SeraMir Exosome RNA Isolation kit (System Biosciences) according to the manufacturer's instructions miRNA-specific primers were pooled according to the Applied Biosystems (Waltham, Massachusetts) protocol (publication 4465407). Reverse transcription was performed on 20 μL of total RNA using the TaqMan MicroRNA Reverse Transcription Kit (Life Technologies, Grand Island, New York). The full names, TaqMan assay identification numbers, and sequences of each of the selected miRNAs are available in Supplemental Table I. All quantitative polymerase chain reaction (PCR) reactions were performed in triplicate using TaqMan MicroRNA Assays and TaqMan Universal Master Mix II (no UNG) using the Bio-Rad (Hercules, California) CFX96 Touch Real-Time PCR Detection System.

      Statistical Analysis

      Univariate analyses, including participants’ demographic and anthropometric characteristics, were analyzed using GraphPad Prism statistical software, version 9.1.2 (GraphPad Software, San Diego, California) with a standard α of 0.05. Normality was determined using the Kolmogorov-Smirnov test. Between-group differences of normally distributed, continuous data were compared using the unpaired t test; nonnormally distributed data were analyzed by Mann-Whiney U test. Relative expression by quantitative PCR of individual miRNA species was calculated using the 2−ΔΔCt method
      • Livak KJ
      • Schmittgen TD.
      Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method.
      in which the geometric mean of all the miRNAs of the samples was used for normalization of the data.
      • Mestdagh P
      • Van Vlierberghe P
      • De Weer A
      • Muth D
      • Westermann F
      • Speleman F
      • et al.
      A novel and universal method for microRNA RT-qPCR data normalization.
      The normal weight non-GDM group was set as the reference group (ie, this group's fold change [FC] was set as 1.0), and the FC of each of the miRNAs for the other 3 groups (normal weight GDM, overweight/obese non-GDM, and overweight/obese GDM) was calculated as compared with the normal weight non-GDM group.
      Our stratified random sampling procedure was designed to achieve balance in BMI groups across the 2 GDM groups. However, residual confounding could exist, so linear regression models were constructed. Maternal age, maternal race, gestational weight gain, gestational age, maternal educational level, household income, quality of diet measured by the healthy eating index score and postpartum weight loss at 1 month, and infant sex were also included as covariates in our primary linear regression models. In the models, the FC of each miRNA served as the dependent variable, and GDM was the independent variable of primary interest. Stepwise regression was also performed, allowing independent variables to be added and/or removed from the model in a purely data-driven process until the model with the lowest Akaike information criterion confirmed whether GDM remained a significant term. The effect sizes, 95% confidence intervals (CIs), and P values were presented for all independent variables. To examine the effect of miRNA on infant size, we constructed 1-month infant growth (change in weight from birth to 1 month) and body composition at 1, 3, and 6 months of age linear regression models with weight, length, WAZ, WLZ, change in weight from birth to 1 month of age (Δ weight 0–1 month), percentage of BF, FM, and FFM as the dependent variables, with each miRNA FC as the primary independent variable. Models were controlled for maternal BMI, maternal age, infant sex, gestational age, birth weight, and birth length. Because these analyses were exploratory outcomes, the models were not controlled for multiple comparison tests. For the models that included Δ weight 0 to 1 month, infant birth weight and birth length were not included because Δ weight 0 to 1 month is affected by infant weight and length at birth.

      Results

      Clinical Characteristics of Mother-Infant Dyads

      Demographic and clinical information for the mother-infant dyads is presented in Table I and Table II. The WAZ and LAZ scores in Table I were calculated after adjusting for gestational age at birth. Given that our mothers with GDM were very well controlled based on their glycosylated hemoglobin level (data not shown) and did not have excessive gestational weight gain as defined by the American College of Obstetricians and Gynecologists, we were not surprised that infants born to mothers with GDM had lower WAZ scores (not statistically significant). Of 32 women with GDM, 20 received medication (insulin or an oral agent) and 8 women were diet controlled. Treatment information on 4 women was not available in the electronic medical records because of missing data. All our patients with GDM participate in educational sessions with certified diabetes educators. The patients send their blood glucose log to the practitioner on a weekly basis. In the present study, if closer follow-up to assess the level of glucose control was needed, this was done through the electronic record. When patients with GDM are taking medications such as insulin or oral agents, they are seen twice weekly for fetal surveillance and to review glucose logs. Thus, relatively lower birth weight for gestational age (not statistically significant) and lower weight at 1 month in the infants with GDM could be related to good glycemic control. Differences were observed in the racial distribution, gestational weight gain, and household income between the 2 groups. Infants born to mothers with GDM were born 1.7 weeks earlier than infants born to mothers in the non-GDM group, but birth weight z score differences did not reach statistical significance.
      Table IMaternal and infant birth characteristics.
      CharacteristicNon-GDM (n = 62)GDM (n = 32)P Value
      Weight and length for age z scores are adjusted for gestational age.
      Mothers
       Age, mean (SD), y32.58 (4.19)34.29 (4.25)0.071
       Prepregnancy BMI, mean (SD), kg/m227.98 (5.70)29.76 (7.39)0.245
       Gestational weight gain, mean (SD), kg12.97 (5.45)9.83 (5.39)0.010
       Excessive gestational weight gain, %50320.124
      Postpartum weight loss at 1 month, mean (SD), kg9.50 (3.30)10.34 (2.37)0.164
      Maternal race/ethnicity, % non-Hispanic White87690.003
      Maternal educational level, %0.123
       HS graduate, GED, or associate's degree1825
        Bachelor's degree3441
        Graduate degree4834
      Maternal income, %0.004
        <$60,0002722
        $60,000–$90,000249
        $90,000–$120,0004969
      Maternal Dietary Quality Score (HEI 2015), mean (SD)65.68 (10.0)62.65 (9.91)0.200
      Infants
       Sex, % male4645>0.999
       Birth weight, mean (SD), kg3.66 (0.48)3.34 (0.42)0.054
       Gestational age, mean (SD), wk39.83 (1.00)38.09 (1.92)<0.001
       Weight for age z score, mean (SD)
      Weight and length for age z scores are adjusted for gestational age.
      0.76 (0.97)0.14 (0.85)0.058
       Length for age z score, mean (SD)
      Weight and length for age z scores are adjusted for gestational age.
      1.39 (1.21)0.76 (1.13)0.120
       Weight for length z score, mean (SD)−0.64 (1.51)−0.90 (1.16)0.703
      BMI = body mass index; HEI = Healthy Eating Index; HS = high school; GDM = gestational diabetes mellitus; GED = General Educational Development.*P values were calculated using the 2-sided t test for continuous data and the Fisher exact test used for categorical data.
      Weight and length for age z scores are adjusted for gestational age.
      Table IIInfant characteristics at 1, 3, and 6 months.
      CharacteristicNon-GDM (n = 62)GDM (n = 32)P Value*
      1-Month infant characteristics
      The 1- and 3-month body compositions were assessed using air displacement plethysmography.
       Weight, mean (SD), kg4.58 (0.53)4.26 (0.54)0.045
       Weight for age z score, mean (SD)0.29 (0.82)−0.30 (1.17)0.048
       Δ Weight from 0–1 month, mean (SD), kg
      Δ Weight indicates change in weight for the respective ages at 1, 3, and 6 months
      0.89 (1.03)0.92 (0.34)0.860
       Length, mean (SD), cm54.71 (2.43)53.48 (1.86)0.022
       Length for age z score, mean (SD)0.20 (1.17)−0.51 (1.16)0.020
       Weight for length z score, mean (SD)0.21 (1.15)0.17 (0.91)0.952
       Body fat, mean (SD), %17.09 (4.17)16.47 (5.13)0.521
       Fat mass, mean (SD), kg0.79 (0.24)0.71 (0.29)0.300
       Fat-free mass, mean (SD), kg3.80 (0.39)3.54 (0.36)0.023
      3-Month infant characteristics
      The 1- and 3-month body compositions were assessed using air displacement plethysmography.
       Weight, mean (SD), kg6.31 (0.71)6.03 (0.81)0.474
       Weight for age z score, mean (SD)0.16 (0.75)−0.11 (1.00)0.514
       Δ Weight from 1–3 months, mean (SD), kg
      Δ Weight indicates change in weight for the respective ages at 1, 3, and 6 months
      1.98 (1.03)1.71 (0.54)0.534
       Length, mean (SD), cm61.19 (2.18)60.76 (2.18)0.993
       Length for age z score, mean (SD)0.19 (0.97)0.09 (0.92)0.822
       Weight for length z score, mean (SD)0.10 (0.94)−0.21 (1.16)0.263
       Body fat, mean (SD), %23.83 (4.12)24.22 (5.34)0.766
       Fat mass, mean (SD), kg1.51 (0.36)1.49 (0.50)0.929
       Fat-free mass, mean (SD), kg4.81 (0.51)4.57 (0.44)0.302
       Exclusively breastfed, %
      For exclusive breastfeeding status at 3 months, data were available from 24 of 32 infants in the GDM group and 61 of 62 infants in the non-GDM group.
      95.0883.330.095
      6-Month infant characteristics
      The 6-month body composition was assessed using dual energy x-ray absorptiometry.
       Weight, mean (SD), kg7.90 (0.87)7.49 (1.11)0.343
       Weight for age z score, mean (SD)0.21 (0.87)−0.23 (1.19)0.259
      Δ Weight from 1–6 months, mean (SD), kg
      Δ Weight indicates change in weight for the respective ages at 1, 3, and 6 months
      3.32 (0.33)3.23 (0.57)0.298
       Length, mean (SD), cm67.06 (2.22)67.34 (2.45)0.196
       Length for age z score, mean (SD)0.11 (0.99)0.26 (0.90)0.150
      Weight for length z score, mean (SD)0.30 (0.91)−0.44 (1.26)0.019
       Body fat, mean (SD), %33.40 (3.25)33.92 (3.54)0.898
       Fat mass, mean (SD), kg2.81 (0.49)2.73 (0.58)0.712
       Fat-free mass, mean (SD), kg5.58 (0.65)5.29 (0.64)0.568
      Exclusively breastfed, %
      For exclusive breastfeeding status at 6 months, data were available from 19 of 32 infants in the GDM group and 61 of 62 infants in the non-GDM group.
      90.1663.160.010
      (GDM = gestational diabetes mellitus.*P values were calculated using a 2-sided t test for continuous data and the Fisher exact test used for categorical data.
      The 1- and 3-month body compositions were assessed using air displacement plethysmography.
      Δ Weight indicates change in weight for the respective ages at 1, 3, and 6 months
      § The 6-month body composition was assessed using dual energy x-ray absorptiometry.
      || For exclusive breastfeeding status at 3 months, data were available from 24 of 32 infants in the GDM group and 61 of 62 infants in the non-GDM group.
      For exclusive breastfeeding status at 6 months, data were available from 19 of 32 infants in the GDM group and 61 of 62 infants in the non-GDM group.
      Infants born to mothers with GDM weighed less than infants born to mothers without GDM, and had lower fat-free mass at 1 month of age, but this difference was not present at 3- or 6-months (Table II). No difference in % body fat and fat mass was observed between groups at 1-, 3-, and 6 months of age. All the infants at 1 month of age were exclusively breast fed by study design. Exclusive breastfeeding at 6 months postpartum was more common in mothers without GDM than in mothers who had GDM.

      Abundance of HM miRNAs at 1 Month of Lactation in Milk of Mothers With GDM

      At 1 month postpartum, the abundance of each miRNA was lower in mothers with GDM than in mothers without GDM (Figure 1). miRNA-148a was 45% lower in the normal weight GDM group and 61% lower in the overweight/obese GDM group than in the normal weight non-GDM group. miRNA-30b was 65% lower in the normal weight GDM group and 64% lower in in the overweight/obese GDM group than in the normal weight non-GDM group. miRNA-let-7a was 35% lower in the normal weight GDM group; however, no difference was observed between the overweight/obese GDM group and normal weight non-GDM group. miRNA-let-7d was 48% lower in the normal weight GDM and 47% lower in the overweight/obese GDM than in the normal weight non-GDM group.
      Figure 1
      Figure 1Relative abundances measured as fold change of miRNA-148a, miRNA-30b, miRNA-let-7a, and miRNA-let-7d are reduced in milk of mothers with gestational diabetes mellitus (GDM). Error bars indicate SEM.
      (P < 0.05.
      ⁎⁎P < 0.01.
      ***P < 0.001.

      Abundance of HM miRNAs at 1 month of Lactation in Milk of Mothers With GDM

      GDM exposure was independently associated with a lower abundance of each miRNA (P < 0.05) (Table III). In addition, maternal age was inversely associated with the abundance of miRNA-148a (β = −0.044) and miRNA-let-7a (β = −0.041), whereas maternal diet quality was positively associated with abundance of each of the miRNAs (β = 0.023 for miRNA-148a, β = 0.024 for miRNA-30b, β = 0.018 for miRNA-let-7a, and β = 0.020 for miRNA-let-7d), suggesting an increase in miRNA abundance with better diet. Finally, GDM exposure remained independently associated with miRNA abundances in the final stepwise regression models (P < 0.037 for all) (Supplemental Table II).
      Table IIIRegression model using fold change for individual miRNAs in milk exosomes.
      Variableβ (95% CI)
      Each row gives the respective β (95% CI) values for the regression model, where each miRNA is a dependent variable and each of the other variables, such as GDM, prepregnancy BMI, maternal age, and diet quality, is a predictor variable. Models were controled for maternal prepregnancy BMI, maternal age, maternal race, household income, gestational weight gain, maternal dietary quality index score (Health Eating Index), maternal postpartum weight loss at 1 month, gestational age, and infant sex.
      miRNA-148amiRNA-30bmiRNA-let-7amiRNA-let-7d
      GDM−0.474
      P < 0.05.
      (−0.885 to −0.062)
      −0.562
      P < 0.05.
      (−1.109 to −0.016)
      −0.493
      P < 0.01.
      (−0.824 to −0.162)
      −0.667
      P < 0.01.
      (−1.113 to −0.221)
      Maternal prepregnancy BMI0.008 (−0.020 to 0.036)0.003 (−0.034 to 0.041)0.014 (−0.008 to 0.037)0.007 (−0.024 to 0.038)
      Maternal age−0.044
      P < 0.05.
      (−0.082 to −0.005)
      −0.041 (−0.092 to 0.009)−0.041
      P < 0.05.
      (−0.072 to −0.010)
      −0.031 (−0.073 to 0.010)
      Maternal race/ethnicity (non-Hispanic White)−0.674 (−2.215 to 0.482)−0.718 (−2.896 to 0.686)−0.296 (−1.360 to 0.812)−0.317 (−1.896 to 1.030)
      Household Income

      ($60-90K)
      0.171 (−0.288 to 0.632)0.504 (−0.107 to 1.115)−0.032 (−0.403 to 0.338)−0.159 (−0.459 to 0.451)
      Household income

      ($90–$120,000)
      0.309 (−0.110 to 0.730)0.660
      P < 0.05.
      (0.102 to 1.218)
      0.102 (−0.235 to 0.440)−0.003 (−0.459 to 0.451)
      Gestational weight gain (kg)−0.005 (−0.039 to 0.028)0.005 (−0.039 to 0.050)−0.003 (−0.030 to 0.023)−0.002 (−0.038 to 0.034)
      Gestational age (weeks)−0.020 (−0.116 to 0.076)−0.077 (−0.205 to 0.051)−0.041 (−0.119 to 0.036)−0.088 (−0.193 to 0.016)
      Infant Sex (females)0.125 (−0.156 to 0.407)0.032 (−0.341 to 0.406)0.052 (−0.174 to 0.279)0.139 (−0.166 to 0.445)
      Maternal Dietary Quality Score (HEI 2015)0.023
      P < 0.01.
      (0.006 to 0.040)
      0.024
      P < 0.05.
      (0.002 to 0.046)
      0.018
      P < 0.01.
      (0.004 to 0.031)
      0.020
      P < 0.05.
      (0.001 to 0.038)
      Postpartum weight loss at 1 month−0.016 (0.073 to 0.040)−0.050 (−0.126 to 0.026)−0.023 (−0.069 to 0.022)−0.007 (−0.069 to 0.054)
      BMI = body mass index; GDM = gestational diabetes mellitus; miRNA = microRNA.
      low asterisk Each row gives the respective β (95% CI) values for the regression model, where each miRNA is a dependent variable and each of the other variables, such as GDM, prepregnancy BMI, maternal age, and diet quality, is a predictor variable. Models were controled for maternal prepregnancy BMI, maternal age, maternal race, household income, gestational weight gain, maternal dietary quality index score (Health Eating Index), maternal postpartum weight loss at 1 month, gestational age, and infant sex.
      P < 0.05.
      P < 0.01.

      Association of HM miRNAs Sampled at 1 Month With Infant Weight and Adiposity at 1 Month

      miR-148a and miR-30b were both significantly associated with infant weight and body composition measures at 1 month, after controlling for gestational age, birth weight, and infant sex (Table IV). With every fold increase in miRNA-148a, 1-month infant weight, WAZ, percentage of BF, and FM decreased by 0.31 kg (P = 0.012), 0.58 (P = 0.032), 3.6% (P = 0.044), and 0.2 kg (P = 0.024), respectively. In contrast, the abundance of miRNA-30b was positively associated with infant anthropometrics at 1 month of age. For every fold increase in miRNA-30b, 1-month infant weight, WAZ, and FM increased by 0.25 kg (P = 0.003), 0.38 (P = 0.040), and 0.17 kg (P = 0.008), respectively. There was no association any miRNA FC with Δ weight from 0 to 1 month of age.
      Table IVAssociations between 1-month breast milk exosomal miRNA abundance and infant outcomes at 1, 3 and 6 months of age.
      Each row gives the respective β (95% CI) values for the regression model, where infant outcomes are dependent variables and the miRNA fold changes are independent variables. Models were controled for maternal prepregnancy body mass index, maternal age, infant sex, gestational age, birth weight, and birth length.
      miRNAβ (95% CI)
      WeightLengthWAZWLZBody fatFat massFat-free mass
      1-Month outcomes
      The 1- and 3-month body compositions were assessed using air displacement plethysmography.
      (n = 84 all infants)
      miRNA-148a−0.315
      P < 0.05.
      (−0.560 to 0.069)
      −0.347 (−1.372 to 0.676)−0.588
      P < 0.05.
      (−1.125 to 0.049)
      −0.007 (−1.186 to 0.440)−3.600
      P < 0.05.
      (−7.115 to 0.083)
      −0.222
      P < 0.05.
      (−0.415 to 0.029)
      −0.105 (−0.238 to 0.027)
      miRNA-30b0.253
      P < 0.05.
      (0.086, 0.419)
      0.420 (−0.336 to 1.176)0.382
      P < 0.05.
      (0.017 to 0.746)
      −0.066 (−0.283 to 0.817)2.960 (0.582 to 5.336)0.178
      P < 0.05.
      (0.047 to 0.308)
      0.087 (−0.002 to 0.177)
      miRNA-let-7a0.080 (−0.108 to 0.270)0.654 (−0.194 to 1.503)0.129 (−0.286 to 0.544)0.030 (−0.793 to 0.461)1.39 (−1.294 to 4.082)0.063 (−0.083 to 0.211)0.011 (−0.090 to 0.112)
      miRNA-let-7d−0.098 (−0.304 to 0.107)−0.545 (−1.470 to 0.380)−0.020 (−0.472 to 0.430)0.028 (−0.685 to 0.678)−1.37 (−4.302 to 1.555)−0.058 (−0.218 to 0.102)−0.027 (−0.138 to 0.083)
      3-Month outcomes
      The 1- and 3-month body compositions were assessed using air displacement plethysmography.
      (n = 76)
      miRNA-148a−0.027 (−0.392 to 0.336)−0.621 (−1.777 to 0.535)−0.262 (−0.571 to 0.390)−0.230 (−0.646 to 0.674)0.387 (−2.542 to 3.317)0.013 (−0.227 to 0.253)−0.050 (−0.280 to 0.179)
      miRNA-30b−0.158(−0.403 to 0.086)−0.014 (−0.852 to 0.823)0.132 (−0.455 to 0.175)0.075 (−0.538 to 0.328)−1.338 (−3.303 to 0.625)−0.108 (0.270 to 0.052)−0.061 (−0.214 to 0.091)
      miRNA-let-7a−0.126 (−0.400 to 0.148)0.751 (−0.229 to 1.144)−0.151 (−0.554 to 0.175)−0.537 (−1.016 to 0.000)−3.038
      P < 0.01.
      (−5.235 to 0.840)
      −0.214
      P < 0.05.
      (−0.392 to 0.035)
      0.100 (−0.073 to 0.274)
      miR-let-7d0.295 (0.010 to 0.580)−0.125 (−1.144 to 0.893)0.085 (−0.053 to 0.694)0.345 (−0.006 to 1.042)2.426
      P < 0.05.
      (0.137 to 4.715)
      0.211
      P < 0.05.
      (0.025 to 0.397)
      0.082 (−0.098 to 0.264)
      6-Month outcomes
      The 6-month body composition was assessed using dual energy x-ray absorptiometry.
      (n = 74)
      miRNA-148a−0.489
      P < 0.01.
      (−0.841 to 0.136)
      −1.669
      P < 0.01.
      (−2.557 to 0.780)
      −0.466 (−1.337 to 0.144)−0.244 (−0.969 to 0.399)−0.527 (−2.452 to 1.398)−0.195 (−0.451 to 0.061)−0.350
      P < 0.05.
      (−0.651 to 0.050)
      miRNA-30b0.039 (−0.207 to 0.286)0.871 (0.246 to 1.495)−0.001 (−0.393 to 0.451)−0.222 (−0.727 to 0.239)−0.865 (−2.284 to 0.552)−0.060 (−0.249 to 0.127)0.157 (−0.064 to 0.379)
      miRNA-let-7a0.114 (−0.228 to 0.457)0.620 (−0.262 to 1.504)−0.094 (−0.595 to 0.573)−0.592 (−1.164 to 0.176)1.007 (−0.867 to 2.883)0.134 (0.113 to 0.382)−0.037 (−0.329 to 0.254)
      miRNA-let-7d0.246 (−0.058 to 0.552)0.293 (−0.474 to 1.061)0.364 (−0.047 to 0.983)0.652 (0.082 to 1.251)-0.394 (−2.029 to 1.241)0.293 (−0.219 to 0.215)0.135 (−0.123 to 0.386)
      miRNA = microRNA; WAZ = weight for age z score; WLZ = weight for length z score.
      low asterisk Each row gives the respective β (95% CI) values for the regression model, where infant outcomes are dependent variables and the miRNA fold changes are independent variables. Models were controled for maternal prepregnancy body mass index, maternal age, infant sex, gestational age, birth weight, and birth length.
      The 1- and 3-month body compositions were assessed using air displacement plethysmography.
      The 6-month body composition was assessed using dual energy x-ray absorptiometry.
      § P < 0.05.
      || P < 0.01.
      The association of infant weight and percentage of BF with miRNA-148a and miRNA-30b did not persist at 3 months of age (Table IV). miRNA-let-7a and miRNA-let-7d were both significantly associated with percentage of BF and FM in the infant at 3 months of age, after controlling for gestational age, birth weight, and infant sex (Table IV). For every fold increase in miRNA-let-7a, percentage of BF and FM decreased by 3% (P = 0.007) and 0.21 kg (P = 0.019), respectively, at 3 months. In contrast, with every fold increase in miRNA-let-7d, percentage of BF and FM increased by 2.4% (P = 0.038) and 0.21 kg (P = 0.026), respectively, at 3 months.
      The associations observed between miR-148a abundance at 1 month and infant body composition at 1 month were observed again at 6 months of age. With every fold increase in miRNA-148a, infant weight, length, FFM, and total lean mass decreased by 0.5 kg (P = 0.007), 1.7 cm (P = 0.0003), 0.35 kg (P = 0.022), and 0.4 kg (P = 0.018), respectively (Table IV), at 6 months.

      Discussion

      Our findings indicate that exposure to diabetes during pregnancy is associated with altered abundance of HM miRNAs known to be involved in glucose homeostasis and adipogenesis, and these miRNAs are also associated with infant weight and adiposity in the first 6 months of life. The impact of GDM on HM miRNA abundance persists even after adjusting for a variety of potential confounders, such as maternal age, prepregnancy BMI, gestational weight gain, nutrition, gestational age, and infant sex. We also observed that higher maternal diet quality was associated with increased abundance of each of the measured miRNAs.
      The literature regarding the effects of maternal diabetes on HM miRNA abundance is sparse. Xi et al
      • Xi Y
      • Jiang X
      • Li R
      • Chen M
      • Song W
      • Li X
      The levels of human milk microRNAs and their association with maternal weight characteristics.
      examined relationships between gestational metabolic diseases (GDM and gestational hypertension) and HM miRNAs (miRNA-let-7a, miRNA-30b, and miRNA-378) involved in adipogenesis. Contrary to our findings, both miRNA-let-7a and miRNA-378 were increased in gestational metabolic diseases. However, after maternal prepregnancy BMI was controlled for, the relationship was no longer statistically significant, suggesting that the difference in miRNA abundance was related to obesity rather than gestational metabolic disease. The inclusion of mothers with gestational hypertension in the gestational metabolic disease group may account for the difference because women with chronic disease beyond GDM were excluded from the present study.
      Accumulating evidence suggests the systemic uptake of milk exosomes and their miRNAs in both animal models
      • Manca S
      • Upadhyaya B
      • Mutai E
      • Desaulniers AT
      • Cederberg RA
      • White BR
      • et al.
      Milk exosomes are bioavailable and distinct microRNA cargos have unique tissue distribution patterns.
      ,
      • Arntz OJ
      • Pieters BC
      • Oliveira MC
      • Broeren MG
      • Bennink MB
      • de Vries M
      • et al.
      Oral administration of bovine milk derived extracellular vesicles attenuates arthritis in two mouse models.
      and healthy human volunteers.
      • Baier SR
      • Nguyen C
      • Xie F
      • Wood JR
      • Zempleni J.
      MicroRNAs are absorbed in biologically meaningful amounts from nutritionally relevant doses of cow milk and affect gene expression in peripheral blood mononuclear cells, HEK-293 kidney cell cultures, and mouse livers.
      Milk exosomes and their miRNAs are absorbed in the systemic circulation in the infants and may impact epigenetic programming of metabolically active tissues, such as liver, pancreatic islets, and beige, brown, and white adipose tissues, in the early neonatal period and infancy.
      • Melnik BC
      • Stremmel W
      • Weiskirchen R
      • John SM
      • Schmitz G.
      Exosome-derived microRNAs of human milk and their effects on infant health and development.
      miRNA-148a-3p is the most abundant miRNA in human, bovine, and porcine milk.
      • Chen X
      • Gao C
      • Li H
      • Huang L
      • Sun Q
      • Dong Y
      • et al.
      Identification and characterization of microRNAs in raw milk during different periods of lactation, commercial fluid, and powdered milk products.
      • Izumi H
      • Kosaka N
      • Shimizu T
      • Sekine K
      • Ochiya T
      • Takase M.
      Bovine milk contains microRNA and messenger RNA that are stable under degradative conditions.
      • Gu Y
      • Li M
      • Wang T
      • Liang Y
      • Zhong Z
      • Wang X
      • et al.
      Lactation-related microRNA expression profiles of porcine breast milk exosomes.
      • Zhou Q
      • Li M
      • Wang X
      • Li Q
      • Wang T
      • Zhu Q
      • et al.
      Immune-related microRNAs are abundant in breast milk exosomes.
      In the present analysis, the abundance of miRNA-148a was lower in HM of mothers with GDM during pregnancy, a new finding. Furthermore, lower abundance of miRNA-148a was associated with higher infant weight, WAZ, and percentage of BF at 1 month of life. Evidence indicates that infants born to mothers with GDM have increased adiposity after controlling for adiposity at birth, despite good maternal glycemic control during the pregnancy and predominant breastfeeding during early infancy.
      • Logan KM
      • Emsley RJ
      • Jeffries S
      • Andrzejewska I
      • Hyde MJ
      • Gale C
      • et al.
      Development of early adiposity in infants of mothers with gestational diabetes mellitus.
      In addition to exposure to excess nutrients that may alter hypothalamic sensing of satiety, leading to alterations in appetite, lower miRNA-148a levels in HM of mothers with GDM might be 1 mechanism for the increased adiposity and weight gain observed in previous studies of infants exposed to diabetes in utero. We observed greater miRNA-148a abundance in mothers with higher diet quality. Thus, if miRNA-148a abundance can be increased in HM from mothers with GDM by improved diet, this might be a strategy to mitigate the increased risk of obesity and T2DM in infants born to mothers with GDM. miRNA-30b in GDM HM was lower in abundance and positively associated with infant weight and FM at 1 month of age. Overexpression of miRNA-30b stimulates adipogenesis, increases the size of lipid droplets, and upregulates lipogenic genes.
      • Zaragosi LE
      • Wdziekonski B
      • Brigand KL
      • Villageois P
      • Mari B
      • Waldmann R
      • et al.
      Small RNA sequencing reveals miR-642a-3p as a novel adipocyte-specific microRNA and miR-30 as a key regulator of human adipogenesis.
      These findings might explain the positive association of miRNA-30b with infant weight and FM at 1 month of life. An important finding is the opposite directionality of miRNA-148a and miRNA-30b with infant weight and FM at 1 month of age. It is possible that the abundance of these 2 miRNAs in HM indicate the fine-tuning by the miRNAs in HM to modulate early infant growth.
      Brown adipose tissue (BAT) and beige adipose tissue (BET) play a major role in energy expenditure, and impaired BAT function is associated with obesity and metabolic disorders.
      • Hu F
      • Wang M
      • Xiao T
      • Yin B
      • He L
      • Meng W
      • et al.
      miR-30 promotes thermogenesis and the development of beige fat by targeting RIP140.
      miRNA-30b/c is the main positive regulator of BAT and BET adipogenesis.
      • Hu F
      • Wang M
      • Xiao T
      • Yin B
      • He L
      • Meng W
      • et al.
      miR-30 promotes thermogenesis and the development of beige fat by targeting RIP140.
      miRNAs expressed in milk-derived exosomes may have an important developmental effect on the differentiation of adipose tissue, as hypothesized recently by Melnik et al.
      • Melnik BC
      • Stremmel W
      • Weiskirchen R
      • John SM
      • Schmitz G.
      Exosome-derived microRNAs of human milk and their effects on infant health and development.
      The protective role of breastfeeding in prevention of early obesity may be related to consumption of HM miRNA-30b.
      • Gharanei S
      • Shabir K
      • Brown JE
      • Weickert MO
      • Barber TM
      • Kyrou I
      • et al.
      Regulatory microRNAs in brown, brite and white adipose tissue.
      The increase of miRNA-30b–associated FM may be related to increased BAT, which would be a favorable effect. The ratio and kinetics of postnatal miRNA-148a/miRNA-30b/c expression may tune the type and amount of adipose tissue development and thus development of obesity later in life.
      Another important finding of our study is the positive association of maternal diet quality with the abundance of each of the measured miRNAs. Higher maternal diet quality during pregnancy and early lactation is associated with lower weight for length, percentage of BF, and FM in breastfed infants.
      • Tahir MJ
      • Haapala JL
      • Foster LP
      • Duncan KM
      • Teague AM
      • Kharbanda EO
      • et al.
      Higher maternal diet quality during pregnancy and lactation is associated with lower infant weight-for-length, body fat percent, and fat mass in early postnatal life.
      Current infant formulas are deficient in miRNAs,
      • Chen X
      • Gao C
      • Li H
      • Huang L
      • Sun Q
      • Dong Y
      • et al.
      Identification and characterization of microRNAs in raw milk during different periods of lactation, commercial fluid, and powdered milk products.
      ,
      • Golan-Gerstl R
      • Elbaum Shiff Y
      • Moshayoff V
      • Schecter D
      • Leshkowitz D
      • Reif S
      Characterization and biological function of milk-derived miRNAs.
      which may have negative effects on long-term metabolic health in the infant.
      • Melnik BC
      • Schmitz G.
      Milk's role as an epigenetic regulator in health and disease.
      Furthermore, formula-fed infants have more rapid weight gain out of proportion to linear growth than do breastfed infants. Formulas supplemented with particular exosomal miRNAs found to influence infant adiposity, such as miRNA-148a and miRNA-30b, might be a way to mitigate future obesity risk. Improved diet quality might lead to increased abundance of these miRNAs in HM, decreasing adiposity among infants born to mothers with GDM.
      Our study has several notable strengths, including a relatively large cohort of well-characterized lactating mothers. We used controlled and consistent sampling protocols for HM collection. Furthermore, we were able to account for a range of demographic, clinical, and lifestyle factors that may confound the association of miRNAs and offspring growth and/or body composition. Although ours is the first and 1 of the largest human studies to examine the impact of maternal GDM on specific miRNAs in HM, we acknowledge a few limitations. The first limitation is our inability to characterize miRNAs from other milk fractions, including cells and the lipid fraction of HM. However, HM exosomes contain the largest proportion of total HM miRNAs, and their miRNAs are the most bioavailable to the infant. Second, we used a targeted approach, selecting highly abundant miRNAs involved in important relevant metabolic pathways. This approach might miss other differentially abundant miRNAs in HM exosomes altered by maternal GDM but protects the study from type I error (incorrect rejection of the null hypothesis) caused by multiple comparisons. Larger studies are needed to assess the hundreds of known exosomal miRNAs. Third, we measured the abundance of miRNAs in a single HM sample at a single point in time. The extent to which a single milk sample may represent the mean output from a given individual women is unknown. Recent work by Raymond et al
      • Raymond F
      • Lefebvre G
      • Texari L
      • Pruvost S
      • Metairon S
      • Cottenet G
      • et al.
      Longitudinal human milk miRNA composition over the first 3 mo of lactation in a cohort of healthy mothers delivering term infants.
      found a dynamic change in HM miRNA composition from the second week to third month post partum. Fourth, we used 2 different methods to measure body composition: Pea Pod at 1 and 3 months and dual x-ray energy absorptiometry at 6 months of age. However, both of these methods are standard and validated to measure infant body composition.
      • Demerath EW
      • Fields DA
      Body composition assessment in the infant.
      Fifth, we measured infant body composition using Pea Pod and dual x-ray energy absorptiometry, which does not allow us to distinguish among the quantities of white adipose tissue, BAT, and BET. Although this is beyond the scope of the present study, this approach will be considered as part of our future studies. Furthermore, we do not have information on the maternal intake of ω3 fatty acids, which may have an impact on offspring body composition in childhood.
      • Moon RJ
      • Harvey NC
      • Robinson SM
      • Ntani G
      • Davies JH
      • Inskip HM
      • et al.
      Maternal plasma polyunsaturated fatty acid status in late pregnancy is associated with offspring body composition in childhood.
      Lastly, this is an observational cohort study with strong confounder adjustment, but nonetheless we cannot establish causality of the reported associations.

      Conclusions

      In summary, we found that the abundance of select miRNAs is reduced in HM of mothers with GDM early in lactation, and these miRNAs are associated with infant weight and adiposity from 1 to 6 months of age. miRNA-148a was associated with decreased infant weight and adiposity, and miRNA-30b was associated with increased infant weight and adiposity at 1 month of life, suggesting fine-tuning among different miRNAs to regulate early infant weight and body composition. Furthermore, we observed that better maternal diet quality was associated with increased abundance of each of the measured miRNAs. Increased risk of obesity and T2DM in infants born to mothers with GDM might be mitigated by improving the miRNA content of the milk from these mothers through improved diet quality or by supplementing infant formulas with these biologically active components. Caution should be exercised because this is an association; therefore, causality may not be inferred. Hence, randomized trials are required to understand the role of diet and maternal clinical factors on HM miRNA abundance and to define the functional role of HM exosomal miRNAs in infant growth.

      Disclosure

      The authors have indicated that they have no conflicts of interest regarding the content of this article.

      Acknowledgments

      We acknowledge and thank all the women and infants and health care practitioners who contributed to the Mothers and Infants Linked for Healthy Growth (MILk) study (NCT03301753), the Maternal Metabolism, Milk, and the Microbiome (4M) Study, and the study teams. We also thank Neely Miller and Kristin Sandness from the Center for Neurodevelopmental Behavior and Rebecca Hollister from the Center for Pediatric Obesity at the University of Minnesota for expert technical assistance and data collection support. The study could not be completed without the work of Lauren Asfaw from the University of Minnesota Department of Obstetrics and Gynecology and the Maternal-Fetal Medicine Clinic. Many thanks go to staff at the University of Oklahoma Health Sciences Center, especially Katy Duncan for her work in all aspects of the study, and to Elisabeth Seburg, Elanadora Sour, Abhilash Muthineni, and Prasanthi Kodhala at the HealthPartners Institute. A special thank you to Laurie Foster for her innovation and dedication to the MILk Study and 4M Study from the start. We also thank Lauren Asfaw at the Department of Obstetrics, Gynecology, and Women's Health, University of Minnesota. We also greatly acknowledge Kathy Kyler for assistance with technical and language editing and proofreading of the manuscript. We also acknowledge the Oklahoma Medical Research Foundation Quantitative Analysis Core for providing statistical analysis that was supported in part by Oklahoma Center of Biomedical Research Excellence (COBRE) grant 1 P30 GM110766-01. Author contributions are as follows: conceptualization: K.B. Shah, D.A. Fields, E.W. Demerath, and J.B. Tryggestad; methodology: K.B. Shah, D.A. Fields, S. Gulati, E.W. Demerath, and J.B. Tryggestad; formal analysis: K.B. Shah and N.P. Pezant; data curation: K.B. Shah, N.P. Pezant, H.K. Kharoud; writing-original draft preparation: K.B. Shah; writing, review and editing: K.B. Shah, D.A. Fields, N.P. Pezant, K. Jacobs, C.A. Gale, E.O. Kharbanda, E.M. Nagel, E.W. Demerath, and J.B. Tryggestad; supervision: D.A. Fields, E.W. Demerath, and J.B. Tryggestad; funding acquisition: K.B. Shah, D.A. Fields, K. Jacobs, and E.W. Demerath. All authors have read and agreed to the published version of the manuscript.

      Funding Sources

      This work was supported by Clinician Scientist Development Grant C5121701 from the Presbyterian Health Foundation (K.B. Shah), National Institutes of Health grant 2R01HD080444 (E.W. Demerath and D.A. Fields), and a Faculty Research Development grant from the University of Minnesota Office of Academic Clinical Affairs (K. Jacobs, C.A. Gale, and E.W. Demerath,). Emily Nagel is supported by National Institutes of Health grant T32DK083250.

      Appendix

      Supplemental Table S1miRNA Sequences
      miRNATaqMan Assay IDMature Sequence
      hsa-miR-148a-3p000470UCAGUGCACUACAGAACUUUGU
      hsa-miR-30b-5p000602UGUAAACAUCCUACACUCAGCU
      hsa-let-7a-5p000377UGAGGUAGUAGGUUGUAUAGUU
      hsa-let-7d-5p002283AGAGGUAGUAGGUUGCAUAGUU
      Supplemental Table S2Stepwise Regression Model
      miRNAStepwise Final Modelβp value
      FC miRNA-148aFC.148a ∼ Group + mat age + mat diet-0.3100.037*
      FC miRNA-30bFC.30b ∼ Group + mat age + PPWL1 + mat educ cat + income cat + mat diet-0.4660.026*
      FC miRNA-let-7aFC.let.7a ∼ Group + mat BMI + mat age + PPWL1 + mat educ cat + mat race + mat diet-0.3560.006**
      FC miRNA-let-7dFC.let.7d ∼ Group + mat age + GA + mat race + mat diet-0.5270.006**
      β-estimate for the stepwise regression model
      *=p <0.05, **=p<0.01; significant values are shown in bold
      FC-Fold change
      Mat-Maternal
      BMI-Maternal pre-pregnancy body mass index
      PPWL1-post partum weight loss at 1 month of lactation
      Edu cat-Education category
      GA-Gestational age

      References

        • Lucas A
        • Fewtrell MS
        • Cole TJ.
        Fetal origins of adult disease-the hypothesis revisited.
        BMJ. 1999; 319: 245-249
        • Plagemann A.
        A matter of insulin: developmental programming of body weight regulation.
        J Matern Fetal Neonatal Med. 2008; 21: 143-148
        • Ellsworth L
        • Harman E
        • Padmanabhan V
        • Gregg B.
        Lactational programming of glucose homeostasis: a window of opportunity.
        Reproduction. 2018; 156: R23-R42
        • Hamosh M.
        Bioactive factors in human milk.
        Pediatr Clin North Am. 2001; 48: 69-86
        • Fields DA
        • George B
        • Williams M
        • Whitaker K
        • Allison DB
        • Teague A
        • et al.
        Associations between human breast milk hormones and adipocytokines and infant growth and body composition in the first 6 months of life.
        Pediatr Obes. 2017; 12: 78-85
        • Fields DA
        • Schneider CR
        • Pavela G.
        A narrative review of the associations between six bioactive components in breast milk and infant adiposity.
        Obesity (Silver Spring). 2016; 24: 1213-1221
        • Shah KB
        • Chernausek SD
        • Garman LD
        • Pezant NP
        • Plows JF
        • Kharoud HK
        • et al.
        Human milk exosomal microRNA: associations with maternal overweight/obesity and infant body composition at 1 month of life.
        Nutrients. 2021; 13: 1091
        • Melnik BC
        • Schmitz G.
        MicroRNAs: Milk's epigenetic regulators.
        Best Pract Res Clin Endocrinol Metab. 2017; 31: 427-442
        • Djuranovic S
        • Nahvi A
        • Green R.
        miRNA-mediated gene silencing by translational repression followed by mRNA deadenylation and decay.
        Science. 2012; 336: 237-240
        • Lewis BP
        • Burge CB
        • Bartel DP.
        Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets.
        Cell. 2005; 120: 15-20
        • Mirra P
        • Nigro C
        • Prevenzano I
        • Leone A
        • Raciti GA
        • Formisano P
        • et al.
        The destiny of glucose from a microRNA perspective.
        Front Endocrinol (Lausanne). 2018; 9: 46
        • Alsaweed M
        • Hartmann PE
        • Geddes DT
        • Kakulas F.
        MicroRNAs in breastmilk and the lactating breast: potential immunoprotectors and developmental regulators for the infant and the mother.
        Int J Environ Res Public Health. 2015; 12: 13981-14020
        • Clausen TD
        • Mathiesen ER
        • Hansen T
        • Pedersen O
        • Jensen DM
        • Lauenborg J
        • et al.
        High prevalence of type 2 diabetes and pre-diabetes in adult offspring of women with gestational diabetes mellitus or type 1 diabetes: the role of intrauterine hyperglycemia.
        Diabetes Care. 2008; 31: 340-346
        • Dabelea D
        • Hanson RL
        • Lindsay RS
        • Pettitt DJ
        • Imperatore G
        • Gabir MM
        • et al.
        Intrauterine exposure to diabetes conveys risks for type 2 diabetes and obesity: a study of discordant sibships.
        Diabetes. 2000; 49: 2208-2211
        • Mayer-Davis EJ
        • Rifas-Shiman SL
        • Zhou L
        • Hu FB
        • Colditz GA
        • Gillman MW.
        Breast-feeding and risk for childhood obesity: does maternal diabetes or obesity status matter?.
        Diabetes Care. 2006; 29: 2231-2237
        • Schaefer-Graf UM
        • Hartmann R
        • Pawliczak J
        • Passow D
        • Abou-Dakn M
        • Vetter K
        • et al.
        Association of breast-feeding and early childhood overweight in children from mothers with gestational diabetes mellitus.
        Diabetes Care. 2006; 29: 1105-1107
        • Melnik BC.
        Milk exosomal miRNAs: potential drivers of AMPK-to-mTORC1 switching in beta-cell de-differentiation of type 2 diabetes mellitus.
        Nutr Metab (Lond). 2019; 16: 85
        • Peila C
        • Gazzolo D
        • Bertino E
        • Cresi F
        • Coscia A.
        Influence of diabetes during pregnancy on human milk composition.
        Nutrients. 2020; 12: 185
        • Wagschal A
        • Najafi-Shoushtari SH
        • Wang L
        • Goedeke L
        • Sinha S
        • deLemos AS
        • et al.
        Genome-wide identification of microRNAs regulating cholesterol and triglyceride homeostasis.
        Nat Med. 2015; 21: 1290-1297
        • Munch EM
        • Harris RA
        • Mohammad M
        • Benham AL
        • Pejerrey SM
        • Showalter L
        • et al.
        Transcriptome profiling of microRNA by Next-Gen deep sequencing reveals known and novel miRNA species in the lipid fraction of human breast milk.
        PLoS One. 2013; 8: e50564
        • Tingo L
        • Ahlberg E
        • Johansson L
        • Pedersen SA
        • Chawla K
        • Saetrom P
        • et al.
        Non-coding RNAs in human breast milk: a systematic review.
        Front Immunol. 2021; 12725323
        • Shah KB
        • Chernausek SD
        • Garman LD
        • Pezant NP
        • Plows JF
        • Kharoud HK
        • et al.
        Human milk exosomal microRNA: associations with maternal overweight/obesity and infant body composition at 1 month of life.
        Nutrients. 2021; 13: 1091
        • Carpenter MW
        • Coustan DR.
        Criteria for screening tests for gestational diabetes.
        Am J Obstet Gynecol. 1982; 144: 768-773
        • Tahir MJ
        • Haapala JL
        • Foster LP
        • Duncan KM
        • Teague AM
        • Kharbanda EO
        • et al.
        Higher maternal diet quality during pregnancy and lactation is associated with lower infant weight-for-length, body fat percent, and fat mass in early postnatal life.
        Nutrients. 2019; 11: 632
        • Krebs-Smith SM
        • Pannucci TE
        • Subar AF
        • Kirkpatrick SI
        • Lerman JL
        • Tooze JA
        • et al.
        Update of the Healthy Eating Index: HEI-2015.
        J Acad Nutr Diet. 2018; 118: 1591-1602
        • Group WHOMGRS
        WHO Child Growth Standards based on length/height, weight and age.
        Acta Paediatr Suppl. 2006; 450: 76-85
        • Mazess RB
        • Barden HS
        • Bisek JP
        • Hanson J.
        Dual-energy x-ray absorptiometry for total-body and regional bone-mineral and soft-tissue composition.
        Am J Clin Nutr. 1990; 51: 1106-1112
        • Xi Y
        • Jiang X
        • Li R
        • Chen M
        • Song W
        • Li X
        The levels of human milk microRNAs and their association with maternal weight characteristics.
        Eur J Clin Nutr. 2016; 70: 445-449
        • Livak KJ
        • Schmittgen TD.
        Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method.
        Methods. 2001; 25: 402-408
        • Mestdagh P
        • Van Vlierberghe P
        • De Weer A
        • Muth D
        • Westermann F
        • Speleman F
        • et al.
        A novel and universal method for microRNA RT-qPCR data normalization.
        Genome Biol. 2009; 10: R64
        • Xi Y
        • Jiang X
        • Li R
        • Chen M
        • Song W
        • Li X
        The levels of human milk microRNAs and their association with maternal weight characteristics.
        European Journal of Clinical Nutrition. 2016; 70: 445-449
        • Manca S
        • Upadhyaya B
        • Mutai E
        • Desaulniers AT
        • Cederberg RA
        • White BR
        • et al.
        Milk exosomes are bioavailable and distinct microRNA cargos have unique tissue distribution patterns.
        Sci Rep. 2018; 8: 11321
        • Arntz OJ
        • Pieters BC
        • Oliveira MC
        • Broeren MG
        • Bennink MB
        • de Vries M
        • et al.
        Oral administration of bovine milk derived extracellular vesicles attenuates arthritis in two mouse models.
        Mol Nutr Food Res. 2015; 59: 1701-1712
        • Baier SR
        • Nguyen C
        • Xie F
        • Wood JR
        • Zempleni J.
        MicroRNAs are absorbed in biologically meaningful amounts from nutritionally relevant doses of cow milk and affect gene expression in peripheral blood mononuclear cells, HEK-293 kidney cell cultures, and mouse livers.
        J Nutr. 2014; 144: 1495-1500
        • Melnik BC
        • Stremmel W
        • Weiskirchen R
        • John SM
        • Schmitz G.
        Exosome-derived microRNAs of human milk and their effects on infant health and development.
        Biomolecules. 2021; 11: 851
        • Chen X
        • Gao C
        • Li H
        • Huang L
        • Sun Q
        • Dong Y
        • et al.
        Identification and characterization of microRNAs in raw milk during different periods of lactation, commercial fluid, and powdered milk products.
        Cell Res. 2010; 20: 1128-1137
        • Izumi H
        • Kosaka N
        • Shimizu T
        • Sekine K
        • Ochiya T
        • Takase M.
        Bovine milk contains microRNA and messenger RNA that are stable under degradative conditions.
        J Dairy Sci. 2012; 95: 4831-4841
        • Gu Y
        • Li M
        • Wang T
        • Liang Y
        • Zhong Z
        • Wang X
        • et al.
        Lactation-related microRNA expression profiles of porcine breast milk exosomes.
        PLoS One. 2012; 7: e43691
        • Zhou Q
        • Li M
        • Wang X
        • Li Q
        • Wang T
        • Zhu Q
        • et al.
        Immune-related microRNAs are abundant in breast milk exosomes.
        Int J Biol Sci. 2012; 8: 118-123
        • Logan KM
        • Emsley RJ
        • Jeffries S
        • Andrzejewska I
        • Hyde MJ
        • Gale C
        • et al.
        Development of early adiposity in infants of mothers with gestational diabetes mellitus.
        Diabetes Care. 2016; 39: 1045-1051
        • Zaragosi LE
        • Wdziekonski B
        • Brigand KL
        • Villageois P
        • Mari B
        • Waldmann R
        • et al.
        Small RNA sequencing reveals miR-642a-3p as a novel adipocyte-specific microRNA and miR-30 as a key regulator of human adipogenesis.
        Genome Biol. 2011; 12: R64
        • Hu F
        • Wang M
        • Xiao T
        • Yin B
        • He L
        • Meng W
        • et al.
        miR-30 promotes thermogenesis and the development of beige fat by targeting RIP140.
        Diabetes. 2015; 64: 2056-2068
        • Gharanei S
        • Shabir K
        • Brown JE
        • Weickert MO
        • Barber TM
        • Kyrou I
        • et al.
        Regulatory microRNAs in brown, brite and white adipose tissue.
        Cells. 2020; 9: 2489
        • Golan-Gerstl R
        • Elbaum Shiff Y
        • Moshayoff V
        • Schecter D
        • Leshkowitz D
        • Reif S
        Characterization and biological function of milk-derived miRNAs.
        Mol Nutr Food Res. 2017; 611700009
        • Melnik BC
        • Schmitz G.
        Milk's role as an epigenetic regulator in health and disease.
        Diseases. 2017; 5: 12
        • Raymond F
        • Lefebvre G
        • Texari L
        • Pruvost S
        • Metairon S
        • Cottenet G
        • et al.
        Longitudinal human milk miRNA composition over the first 3 mo of lactation in a cohort of healthy mothers delivering term infants.
        J Nutr. 2021; 152: 94-106
        • Demerath EW
        • Fields DA
        Body composition assessment in the infant.
        Am J Hum Biol. 2014; 26: 291-304
        • Moon RJ
        • Harvey NC
        • Robinson SM
        • Ntani G
        • Davies JH
        • Inskip HM
        • et al.
        Maternal plasma polyunsaturated fatty acid status in late pregnancy is associated with offspring body composition in childhood.
        J Clin Endocrinol Metab. 2013; 98: 299-307

      Linked Article