DIABETICMedicine DOI: 10.1111/dme.12298
Research: Pathophysiology Maternal insulin resistance, triglycerides and cord blood insulin in relation to post-natal weight trajectories and body composition in the offspring up to 2 years S. Brunner1, D. Schmid1, K. Hu¨ttinger1, D. Much1, E. Heimberg2, E.-M. Sedlmeier3,4, M. Bru¨derl5, J. Kratzsch6, B. L. Bader3,4, U. Amann-Gassner1 and H. Hauner1,3 1 Else Kro¨ner-Fresenius-Center for Nutritional Medicine, Klinikum rechts der Isar, Technische Universita¨t Mu¨nchen, Munich, 2Department of Pediatrics, Universita¨tsklinikum Tu¨bingen, 3ZIEL—Research Center for Nutrition and Food Sciences, Nutritional Medicine Unit, 4ZIEL—PhD Graduate School, ‘Epigenetics, Imprinting and Nutrition’, Technische Universita¨t Mu¨nchen, Freising, 5Institute for Medical Statistics and Epidemiology, Klinikum rechts der Isar, Technische Universita¨t Mu¨nchen, Munich and 6Institute of Laboratory Medicine, Clinical Chemistry and Molecular Diagnostics, University Hospital Leipzig, Leipzig, Germany
Accepted 30 July 2013
Abstract Aims The intrauterine metabolic environment might have a programming effect on offspring body composition. We aimed to explore associations of maternal variables of glucose and lipid metabolism during pregnancy, as well as cord blood insulin, with infant growth and body composition up to 2 years post-partum. Methods Data of pregnant women and their infants came from a randomized controlled trial designed to investigate the impact of nutritional fatty acids on adipose tissue development in the offspring. Of the 208 pregnant women enrolled, 118 infants were examined at 2 years. In the present analysis, maternal fasting plasma insulin, homeostasis model assessment of insulin resistance and serum triglycerides measured during pregnancy, as well as insulin in umbilical cord plasma, were related to infant growth and body composition assessed by skinfold thickness measurements and abdominal ultrasonography up to 2 years of age.
Maternal homeostasis model assessment of insulin resistance at the 32nd week of gestation was significantly inversely associated with infant lean body mass at birth, whereas the change in serum triglycerides during pregnancy was positively associated with ponderal index at 4 months, but not at later time points. Cord plasma insulin correlated positively with birthweight and neonatal fat mass and was inversely associated with body weight gain up to 2 years after multiple adjustments. Subsequent stratification by gender revealed that this relationship with weight gain was stronger, and significant only in girls.
Results
Conclusions Cord blood insulin is inversely associated with subsequent infant weight gain up to 2 years and this seems to be more pronounced in girls.
Diabet. Med. 30, 1500–1507 (2013)
Introduction Because of the increasing global prevalence of childhood overweight and obesity [1], the identification of early determinants for adiposity development has received considerable attention. Several studies suggest a role of the metabolic milieu during pregnancy for fetal and later infant growth or body composition. The most convincing evidence comes from studies in the field of gestational diabetes showing higher rates of macrosomia and increased neonatal Correspondence to: Hans Hauner. E-mail [email protected] (Clinical Trials Registry No; NCT 00362089)
1500
fat mass in diabetic pregnancies [2], as well as a continuous association between maternal glycaemia and neonatal adiposity, even in the non-diabetic range [3]. Maternal insulin resistance during pregnancy might further translate into higher weight gain and adiposity in infancy [4]. Besides, maternal variables of lipid metabolism, including triglycerides, have been shown to be related to fetal growth and newborn fat mass [5–7], although this association may vary depending on maternal glucose tolerance status [8] or BMI category [9]. However, longer-term associations of maternal lipids with infant growth variables or adiposity beyond birth are largely unstudied.
ª 2013 The Authors. Diabetic Medicine ª 2013 Diabetes UK
Research article
What’s new? • Associations of maternal metabolic variables and cord blood insulin with infant anthropometry beyond birth have not been adequately addressed. • We investigated the relationship of maternal insulin, homeostasis model assessment of insulin resistance, triglycerides and cord blood insulin with infant weight gain and body composition up to 2 years. • Maternal metabolic variables were transiently related to infant growth and body composition outcomes. • Cord blood insulin was inversely associated with weight gain up to 2 years, and this relationship was stronger and significant only in girls. • Fetal insulin exposure might differentially affect subsequent weight gain up to 2 years in both sexes.
Likewise, insulin in umbilical cord blood, produced by the fetus in response to maternal glycaemia, has been widely investigated in relation to neonatal anthropometry and measures of adiposity [10–12], but data on its relationship with weight gain and body composition later in infancy are scarce [11,13,14]. Recent data indicate that girls are already intrinsically more insulin-resistant than boys at birth [15]. Later on, fetal insulin exposure might have differential implications for post-natal weight development between the sexes, with girls being potentially programmed towards slower growth rates by fetal hyperinsulinemia [14]. We aimed to investigate the relationship of maternal metabolic variables [insulin, insulin resistance expressed by homeostasis model assessment (HOMA-IR) and triglycerides] measured in late pregnancy (32nd week of gestation) and cord blood insulin with infant growth and body composition from birth up to 2 years of age and to explore potential sex-specific differences.
Subjects and methods Data came from the Impact of Nutritional Fatty Acids on Infant Adipose Tissue Development (INFAT) study, an open-label randomized controlled trial originally designed to examine the effect of reducing the maternal dietary n-6: n-3 fatty acid ratio during pregnancy and lactation on infant adipose tissue development. The study design and the clinical results on infant fat mass up to 1 year of age were previously described [16,17]. In brief, 208 healthy pregnant women with singleton pregnancies and a pre-pregnancy BMI between 18 and 30 kg/m2 were enrolled and randomly assigned to either an intervention (n = 104) or a control group (n = 104) from
ª 2013 The Authors. Diabetic Medicine ª 2013 Diabetes UK
DIABETICMedicine
the 15th week of gestation until 4 months post-partum. Women in the intervention group received (1) a dietary supplement containing 1200 mg n-3 long-chain polyunsaturated fatty acids per day and (2) nutritional counselling to normalize the consumption of arachidonic acid to a moderate level of intake. In contrast, women in the control group were advised to keep to a healthy diet according to current recommendations of the German Nutrition Society. The study protocol was approved by the ethical committee of the Technische Universit€at M€ unchen (no. 1479/06/2006/2/21) and all participants gave written informed consent.
Collection of samples
Maternal blood was collected at the 15th and 32nd week of gestation in the morning after an overnight fast. Immediately after delivery, umbilical vein ethylenediaminetetraacetic acid (EDTA) blood samples were collected. Umbilical cord blood was analysed from 137 newborns. Plasma was prepared by centrifugation at 2000 g at 4 °C for 10 min and subsequently aliquoted and stored at 80 °C.
Maternal characteristics and infant anthropometry
Maternal pre-pregnancy weight and height were retrieved from the maternity card, a booklet containing medical information related to antenatal and natal care provided to expecting mothers by their gynaecologist. Maternal glucose tolerance status was defined based on clinical diagnosis by the women’s gynaecologist. Clinically diagnosed gestational diabetes was mostly confirmed by a standardized oral glucose tolerance test and defined based on thresholds according to Coustan et al. [18]. In only one case, the diagnosis was based on elevated glucose at random. The therapy regimen (diet or insulin treatment) of women with gestational diabetes was retrieved from their medical record. The infants were examined at birth (for skinfolds: 3–5 days post-partum), at 6 weeks, 4 months, 1 and 2 years post-partum. Birthweight, length, sex and gestational age of the newborn were retrieved from the medical record. Anthropometric measurements of the infants were taken by trained investigators according to standardized procedures [16]. Skinfolds were measured in triplicate with a Holtain caliper (Holtain Ltd, Crymych, UK) at the left body axis at four sites (triceps, biceps, subscapular and suprailiac). Body fat percentage was calculated via predictive skinfold equations according to Weststrate et al. [19]. Fat mass was calculated as percentage of body fat multiplied by the current body weight of the infant. Lean body mass was calculated as body weight minus fat mass. Abdominal ultrasonography to estimate subcutaneous abdominal and preperitoneal fat was performed by two well-trained paediatricians at 6 weeks, 4 months, 1 and 2 years post-partum according to Holzhauer et al. [20].
1501
DIABETICMedicine
Cord blood insulin in relation to offspring weight gain S. Brunner et al.
Laboratory analyses
Plasma insulin was measured by a chemiluminescence immunoassay using the LIAISON analyser (Diasorin, Saluggia, Italy). Inter- and intra-assay coefficients of variation were both < 4%. Fasting plasma glucose and serum triglyceride levels were determined by an approved external laboratory (Synlab Labordienstleistungen, Munich, Germany). HOMA-IR was calculated from fasting glucose and insulin according to Matthews et al. [21].
Statistical analysis
Statistical analyses were performed with the R software package (version 2.8.1; R Foundation for Statistical Computing, http://www.r-project.org) and Predictive Analytics SoftWare (PASW) software (version 18.0; SPSS Inc., Chicago, IL, USA). To determine differences in the maternal variables between the randomized groups, multiple linear regression models [F-test, analysis of covariance (ANCOVA)] adjusting for age and pre-pregnancy BMI as additional independent variables were employed. To analyse differences in cord blood insulin between the groups or sexes, respectively, Mann– Whitney tests were used. To assess associations of the various maternal metabolic variables at the 32nd week of gestation, and cord blood insulin with infant clinical outcomes up to 2 years of age, data of both groups were pooled. Multiple linear regression models, including the covariates maternal pre-pregnancy BMI, gestational weight gain, maternal glucose tolerance status, pregnancy duration, sex and group allocation for the data at birth, and, additionally, ponderal index at birth and mode of infant feeding at the later time points, were performed. A two-sided P-value < 0.05 was considered statistically significant.
Results We examined 188 newborns at birth, 180 infants at 6 weeks, 174 at 4 months, 170 at 1 year and 118 at 2 years of age. Maternal, newborn and infant characteristics up to 2 years are presented in Table 1. Individuals who completed all assessments did not significantly differ from the rest of the study population in the major socio-demographic and clinical variables (data not shown). Maternal fasting insulin and HOMA-IR at the 32nd week of gestation did not significantly differ between the randomized groups (data not shown). In contrast, triglyceride levels increased significantly less from baseline (15th week of gestation) until 32nd week of pregnancy in the intervention compared with the control group (D intervention 76.2 53.5 mg/dl vs. control 103.6 58.1 mg/dl, P < 0.001), resulting in significantly higher levels in the control group compared with the intervention group at the 32nd week of gestation (P < 0.001). Cord plasma insulin concentrations
1502
did not differ by group (data not shown). Girls displayed significantly higher cord plasma insulin levels compared with boys (P = 0.037, Table 1), and also when extreme outliers with very high insulin levels (> 3 9 interquartile range, n = 7, P = 0.022) were excluded.
Maternal insulin, HOMA-IR and triglycerides at the 32nd week of gestation in relation to infant clinical outcomes up to 2 years post-partum
Maternal insulin, HOMA-IR and triglyceride levels at the 32nd week of gestation were found to be largely unrelated to infant growth and body composition outcomes up to 2 years post-partum. However, HOMA-IR was significantly inversely associated with lean body mass at birth [adjusted regression coefficient (badj) (95% CI) 54.94 ( 99.23 to 10.64) g, P = 0.016] in the analysis controlling for maternal pre-pregnancy BMI, gestational weight
Table 1 Maternal, newborn and 2-year infant characteristics Maternal variables (n = 208 at baseline) Age at enrollment (years) 31.8 4.7 Pre-pregnancy BMI (kg/m2) 22.3 3.0 Parity (primiparae) 58.5 Gestational diabetes 9.0 Dietary treatment 76.5 Insulin treatment 23.5 Gestational weight gain (kg) 15.6 4.9 Education (≥ 12 years at school) 69.1 Energy intake at 32nd week of 2079 415 gestation (kcal/day) (n = 166) Insulin at 32nd week of 76.7 43.1 gestation (pmol/l) (n = 182) HOMA-IR at 32nd week of 2.2 1.2 gestation (n = 175) Triglycerides at 32nd week of 197.0 66.2 gestation (mg/dl) (n = 187) Newborn variables (n = 188) Sex (girls) 47.9 Gestational age (weeks) 39.6 1.5 Birthweight (g) 3443 520 Ponderal index (kg/m3) 24.8 2.4 Sum of four skinfolds (mm) (n = 168) 16.0 2.6 Body fat (%) (n = 168) 13.8 2.7 Cord blood insulin (pmol/l) (n = 137) 28.30 (36.40) Girls (n = 63) 31.90 (37.10)* Boys (n = 74) 25.85 (24.53)* Mode of infant feeding (4 months post-partum) Exclusively breastfed 64.6 Partially breastfed 15.2 Formula fed 20.2 Infant variables at 2 years (n = 118) Body weight (g) 12393 1395 Ponderal index (kg/m3) 18.8 1.8 Sum of four skinfolds (mm) (n = 110) 23.7 3.4 Body fat (%) (n = 110) 19.1 2.3 Weight gain (birth to 2 years) (g) 8919 1306 Data are presented as mean SD, median (interquartile range) (for cord blood insulin) or percentage. *Significant difference between the sexes (P = 0.037), Mann– Whitney test.
ª 2013 The Authors. Diabetic Medicine ª 2013 Diabetes UK
Research article
gain, maternal glucose tolerance status, pregnancy duration, group and sex (see also Supporting Information, Table S1). Moreover, the change in maternal serum triglyceride concentration between the 15th and 32nd week of gestation was weakly, but significantly associated with infant ponderal index at 4 months post-partum [badj: 0.01 (0 to 0.01) kg/m3, P = 0.020], but not with any of the other growth or body composition outcomes up to 2 years post-partum. No significant relationships were found for absolute triglyceride levels (see also Supporting Information, Table S2).
Cord plasma insulin in relation to infant clinical outcomes up to 2 years post-partum
In the analysis comprising all available data at birth, cord blood insulin was significantly positively related to birthweight [b: 1.68 (0.03 to 3.33) g, P = 0.048], the sum of four skinfolds [b: 0.02 (0.01 to 0.02) mm, P = 0.001], percentage body fat [b: 0.02 (0.01 to 0.03)%, P = 0.001], as well as the absolute amount of fat mass [b: 0.81 (0.31 to 1.31) g, P = 0.002] in the newborns. These relationships remained significant after adjustment for maternal pre-pregnancy BMI, gestational weight gain, maternal glucose tolerance status, pregnancy duration, group and sex (Table 2). In contrast, significant inverse relationships of cord blood insulin were found with infant weight, BMI, fat and lean body mass at 2 years, as well as with weight gain from birth up to 2 years in the unadjusted analysis (Table 2). However, in the adjusted model controlling for maternal pre-pregnancy BMI, gestational weight gain, maternal glucose tolerance status, pregnancy duration, group, sex, ponderal index at birth and mode of infant feeding at 4 months post-partum, only the relationship with weight gain remained significant [badj: 6.90 ( 13.13 to 0.67) g, P = 0.030] (Table 2). When the analysis was restricted to participants completing all study visits up to 2 years (n = 118, “completers”), significant inverse associations with infant anthropometrics were already apparent at 1 year of age (BMI: P = 0.015; sum of four skinfolds: P = 0.036; percentage body fat: P = 0.027; fat mass: P = 0.024; weight gain up to 1 year: P = 0.010 in the adjusted model) (Table 2). Testing for sex interaction revealed no significant interaction term for the association with weight gain up to 2 years (P = 0.710). Nevertheless, the relationship of cord blood insulin with weight gain up to 2 years turned out to be much stronger in girls compared with boys [girls: badj 7.63 ( 15.29 to 0.03) g, P = 0.051; boys: badj 2.82 ( 15.76 to 10.11) g, P = 0.662; Fig. 1]. No significant associations were found with the data set on infant fat mass assessed by ultrasonography, neither for maternal variables nor for cord blood insulin (data not shown).
ª 2013 The Authors. Diabetic Medicine ª 2013 Diabetes UK
DIABETICMedicine
Discussion In this study, we investigated associations of various maternal markers of glucose and lipid metabolism and cord blood insulin with subsequent infant growth and body composition over the first 2 years of life. The major finding of our study was that cord blood insulin was inversely associated with anthropometric measures from the first year of life onwards, but only the association with weight gain persisted up to 2 years after adjustment for covariates. This association was significant only in girls, which is consistent with recent findings showing cord C-peptide as a proxy for fetal insulin to be inversely associated with infant weight development over the 1st year of life in girls, but not in boys [14]. In accordance with other studies [14,15], we observed significantly higher cord plasma insulin concentrations in girls compared with boys, although this was not entirely consistent across the literature [10–12]. In view of the lower birthweights but higher cord blood insulin levels in girls, it was suggested that girls might be inherently more insulin-resistant in utero and around birth compared with boys [15]. This ‘gender insulin hypothesis’ proposes that gender-specific genes affecting insulin sensitivity might account for the gender difference in birthweight [22]. Our data and previous findings thus might suggest a programming effect of fetal insulin exposure for early weight gain, for which girls may be more susceptible than boys [14]. The consequences of such an effect with regard to later risk for metabolic diseases remain to be clarified. It has been reported that girls are also more insulin-resistant compared with boys in childhood and that Type 2 diabetes is more common in girls [23]. Furthermore, observational studies have related slower early growth patterns with insulin resistance or diabetes in later life [24,25]. Slower weight gain has also been reported in infants born to mothers with gestational diabetes [26], which could be interpreted as a programming effect of higher fetal insulin concentrations as well. In contrast to fetal insulin, maternal variables of glucose metabolism were largely unrelated to the infant clinical outcomes, except that higher maternal HOMA-IR at the 32nd week of gestation was associated with lower lean body mass at birth. Although our data showed no explicit relationship with fat mass, this finding might to some extent be compatible with data showing increased body fat in newborns from mothers with gestational diabetes compared with women with normal glucose tolerance [2] and the direct association of maternal glucose homeostasis with neonatal adiposity across the full range of maternal glycaemia, and not restricted to manifest diabetes [3]. However, in the follow-up of the Hyperglycaemia and Adverse Pregnancy Outcome (HAPO) study, no such relationship could be found with adiposity at 2 years of age within a group of pregnant women without diabetes [13].
1503
1504 0.00 0.73 2.39 4.65
123 123 123 127
0.02 to 0.01) 4.65 to 0.07) 10.00 to 1.21) 14.46 to 2.85)
0.01 2.36 5.60 8.66
90 90 90 93
( ( ( (
7.60 ( 13.79 to 1.41) 0.01 ( 0.01 to 0.00) 0.01 ( 0.01 to 0.00)
93 93 90
to 0.01) to 0.99) to 0.37) to 0.92)
3.00 ( 7.06 to 1.06) 0.01 ( 0.01 to 0) 0.00 ( 0.02 to 0.02)
127 127 127 0.01 2.44 5.15 8.39
0.02 (0.01 to 0.03) 0.81 (0.31 to 1.31) 0.89 ( 0.39 to 2.17)
130 130 130
( ( ( (
1.68 0.00 0.00 0.02
137 137 137 130
(0.03 to 3.33) ( 0.01 to 0.01) (0 to 0.01) (0.01 to 0.02)
b (95% CI)†
0.269 0.043* 0.013* 0.004*
0.017* 0.021* 0.064
0.733 0.408 0.092 0.016*
0.150 0.053 0.930
0.001* 0.002* 0.176
0.048* 0.875 0.434 0.001*
P
(0.27 to 3.47) ( 0.01 to 0.01) ( 0.00 to 0.01) (0.01 to 0.03)
( ( ( (
0.01 2.18 3.78 7.02
to to to to
0.01) 1.33) 1.44) 0.28)
0.01 2.02 3.80 6.90
( ( ( (
0.02 to 0.01) 4.48 to 0.43) 8.45 to 0.85) 13.13 to 0.67)
5.50 ( 12.12 to 1.13) 0.01 ( 0.01 to 0.00) 0.01 ( 0.03 to 0.01)
0.00 0.43 1.17 3.36
1.58 ( 5.48 to 2.32) 0.00 ( 0.01 to 0.00) 0.00 ( 0.02 to 0.02)
0.02 (0.01 to 0.03) 0.80 (0.28 to 1.32) 1.01 ( 0.21 to 2.24)
1.87 0.00 0.00 0.02
b (95% CI)†
P
0.249 0.105 0.108 0.030*
0.103 0.115 0.226
0.765 0.630 0.375 0.071
0.424 0.121 0.901
0.002* 0.003* 0.105
0.023* 0.803 0.423 0.003*
90 90 90 93
93 93 90
89 89 89
93 93 93 89
n
( 0.92 to 2.87) ( 0.02 to 0.01) ( 0.01 to 0.01) (0.00 to 0.03)
0.01 2.30 2.36 5.77
( ( ( (
0.03 to 0.00) 4.30 to 0.31) 5.45 to 0.73) 10.15 to 1.39)
4.37 ( 9.04 to 0.29) 0.01 ( 0.01 to 0.00) 0.02 ( 0.04 to 0.00)
0.01 (0.00 to 0.03) 0.63 (0.01 to 1.26) 0.41 ( 1.02 to 1.85)
0.98 0.01 0.00 0.01
b (95% CI)†
0.027* 0.024* 0.133 0.010*
0.066 0.015* 0.036*
0.025* 0.045* 0.567
0.308 0.382 0.833 0.031*
P
*P < 0.05. †Data are presented as the regression coefficient (b) along with the 95% confidence interval. ‡Analysis 1 was based on all available data at the respective time points. §Analysis 2 was based on the cohort with full data from birth until 2 years post-partum. ¶At the time point of birth, results were corrected for maternal pre-pregnancy BMI, gestational weight gain, pregnancy duration, group, sex and maternal glucose tolerance status in the adjusted model. At the later time points, results were additionally adjusted for ponderal index at birth and breastfeeding status (fully breastfed, partially breastfed or formula) at 4 months post-partum, respectively.
Birth Birthweight (g) Ponderal index (kg/m3) BMI (kg/m2) Sum of four skinfold thicknesses (mm) Body fat (%) Fat mass (g) Lean body mass (g) 1 year post-partum Weight (g) BMI (kg/m2) Sum of four skinfold thicknesses (mm) Body fat (%) Fat mass (g) Lean body mass (g) Weight gain (g) (birth–1 year post-partum) 2 years post-partum Weight (g) BMI (kg/m2) Sum of four skinfold thicknesses (mm) Body fat (%) Fat mass (g) Lean body mass (g) Weight gain (g) (birth–2 years post-partum)
n
Adjusted model¶
Unadjusted model
Adjusted model¶
Analysis 2 ‘completers only’§
Analysis 1 ‘all available data’‡
Table 2 Umbilical cord plasma insulin (pmol/l) in relation to infant outcomes at birth, 1 year and at 2 years post-partum
DIABETICMedicine Cord blood insulin in relation to offspring weight gain S. Brunner et al.
ª 2013 The Authors. Diabetic Medicine ª 2013 Diabetes UK
Research article
DIABETICMedicine
Girls (n = 39) b: –8.99 [–15.65; –2.33] g, P = 0.010 badj: –7.63 [–15.29; 0.03] g, P = 0.051
Infant sex Girls Boys Girls Boys
Boys (n = 54) b: –6.47 [–18.60; 5.66] g, P = 0.289 badj: –2.82 [–15.76; 10.11] g, P = 0.662
FIGURE 1 Association of cord plasma insulin with infant weight gain up to 2 years post-partum stratified by gender. b, regression coefficient (95% confidence interval) from linear regression analysis; badj, data adjusted for maternal pre-pregnancy BMI, gestational weight gain, pregnancy duration, group, maternal glucose tolerance status, ponderal index at birth and mode of infant feeding at 4 months post-partum (fully breastfed, partially breastfed or formula).
In another recent study, maternal insulin resistance during pregnancy, irrespective of glucose tolerance status, emerged as a significant independent predictor of infant weight gain and adiposity at 1 year of age measured by skinfold thicknesses [4]. However, these findings might not be directly comparable with our study because of differing methodological approaches. It could thus be plausible that measuring insulin resistance only in the fasting state as HOMA-IR is not sensitive enough to detect associations with infant adiposity or weight gain, compared with a more complex assessment of insulin resistance under a glucose challenge. Besides markers of glucose metabolism, we further assessed whether maternal triglyceride levels in later pregnancy are related to infant outcomes from birth until 2 years post-partum. Increasing triglyceride concentrations over the course of pregnancy are a characteristic feature of normal pregnancy. Higher triglyceride levels in the maternal circulation may enhance the concentration gradient across the placenta, resulting in accelerated transport and deposition of lipids in fetal tissues [27]. Against this theory, we could not find any significant associations of maternal triglycerides with neonatal body composition, apart from a weak relationship between the change of maternal serum triglyceride levels from 15th until the 32nd week of gestation and subsequent ponderal index at 4 months. However, this slight effect seems clinically irrelevant or might even have occurred by chance with regard to the numerous correlations explored. Thus, we could not confirm previous findings of a small pilot study showing the increase in triglycerides from ª 2013 The Authors. Diabetic Medicine ª 2013 Diabetes UK
early to late pregnancy to be highly predictive for neonatal adiposity [28]. Other studies reported associations between absolute maternal triglycerides levels and birthweight or large-for-gestational-age infants [5,9,29]. Different sampling time points, weight and health status of the women might explain the discrepancy to our findings. The strengths of our study are the extensive longitudinal assessment of infant body composition over multiple time points by two complementary methods from birth and through the first 2 years of life and the collection of biosamples both from the maternal and the fetal side. In addition, we assessed a large set of maternal and child factors related to infant adiposity to adjust for in the multivariate analysis. However, although not the focus of this study, other factors such as lifestyle in pregnancy, including physical activity and diet, might influence the investigated associations through their effect on gestational weight gain or birthweight [30]. A major limitation of our study is the considerable loss to follow-up beyond the 1st year of life. Although the individuals lost to follow-up were comparable with the completers in their main characteristics, our results might suggest a certain selection bias. Nevertheless, despite the relatively small study population, it is intriguing that we could still confirm sex-specific effects with regard to the associations of fetal insulin with post-natal weight gain. However, as no adjustments for multiple testing were made within this rather exploratory approach, all significant findings should be interpreted with caution. It should also be acknowledged, that neither skinfold measurements nor abdominal sonography represent direct methods for the
1505
DIABETICMedicine
Cord blood insulin in relation to offspring weight gain S. Brunner et al.
assessment of fat mass. Another limitation is that our study protocol did not include a standardized glucose tolerance test to assess insulin resistance under a glucose challenge, so that we had to rely on measures of insulin resistance in the fasting state. Furthermore, we did not measure other variables related to fetal insulin secretion, such as C-peptide or proinsulin, or lipids in umbilical cord blood, which might also come into consideration as potential determinants of fetal growth [7]. In conclusion, maternal insulin resistance and triglycerides in the last trimester of pregnancy were only transiently related to newborn/early post-natal infant growth and body composition and did not emerge as major determinants for infant weight development up to 2 years. In contrast, cord blood insulin was highly correlated with birthweight and newborn fat mass, and significantly inversely associated with weight gain over the first 2 years of life in girls. This might suggest a role for fetal insulin in programming energy homeostasis in early life, which takes effect differently between the sexes.
Funding sources
The study was funded by grants from the Else Kr€ oner-Fresenius Foundation, Bad Homburg, Germany; the International Unilever Foundation, Hamburg, Germany; the EU-funded EARNEST consortium (FOOD-CT-2005-007036); and the German Ministry of Education and Research via the Competence Network on Obesity (Kompetenznetz Adipositas, 01GI0842). There was no intervention from any sponsor with any of the research aspects of the study, including the study design, intervention, data collection, analysis and interpretation, or writing of the manuscript.
Competing interests
HH is on the Advisory Board for Weight Watchers International and has received grants from Riemser and Weight Watchers for clinical trials and payment for lectures from Novartis, Roche Germany, and Sanofi Aventis. All authors declare that there is no duality of interest associated with this manuscript.
Acknowledgements
We thank all the families who participated in the study. We would like to thank the technical staff of the Institute of Laboratory Medicine, Clinical Chemistry and Molecular Diagnostics,University Hospital Leipzig, Leipzig, Germany for assistance in performing the laboratory analyses. We thank Petra Wolf (Institute for Medical Statistics and Epidemiology, Klinikum rechts der Isar, Technische Universit€at M€ unchen, Munich, Germany) for statistical support.
1506
References 1 de Onis M, Blossner M, Borghi E. Global prevalence and trends of overweight and obesity among preschool children. Am J Clin Nutr 2010; 92: 1257–1264. 2 Catalano PM, Thomas A, Huston-Presley L, Amini SB. Increased fetal adiposity: a very sensitive marker of abnormal in utero development. Am J Obstet Gynecol 2003; 189: 1698–1704. 3 HAPO Study Group. Hyperglycemia and Adverse Pregnancy Outcome (HAPO) Study: associations with neonatal anthropometrics. Diabetes 2009; 58: 453–459. 4 Hamilton JK, Odrobina E, Yin J, Hanley AJ, Zinman B, Retnakaran R. Maternal insulin sensitivity during pregnancy predicts infant weight gain and adiposity at 1 year of age. Obesity (Silver Spring) 2010; 18: 340–346. 5 Di Cianni G, Miccoli R, Volpe L, Lencioni C, Ghio A, Giovannitti MG et al. Maternal triglyceride levels and newborn weight in pregnant women with normal glucose tolerance. Diabet Med 2005; 22: 21–25. 6 Harmon KA, Gerard L, Jensen DR, Kealey EH, Hernandez TL, Reece MS et al. Continuous glucose profiles in obese and normal-weight pregnant women on a controlled diet: metabolic determinants of fetal growth. Diabetes Care 2011; 34: 2198– 2204. 7 Schaefer-Graf UM, Graf K, Kulbacka I, Kjos SL, Dudenhausen J, Vetter K et al. Maternal lipids as strong determinants of fetal environment and growth in pregnancies with gestational diabetes mellitus. Diabetes Care 2008; 31: 1858–1863. 8 Schaefer-Graf UM, Meitzner K, Ortega-Senovilla H, Graf K, Vetter K, Abou-Dakn M et al. Differences in the implications of maternal lipids on fetal metabolism and growth between gestational diabetes mellitus and control pregnancies. Diabet Med 2011; 28: 1053–1059. 9 Misra VK, Trudeau S, Perni U. Maternal serum lipids during pregnancy and infant birthweight: the influence of prepregnancy BMI. Obesity (Silver Spring) 2011; 19: 1476–1481. 10 Ong K, Kratzsch J, Kiess W, Costello M, Scott C, Dunger D. Size at birth and cord blood levels of insulin, insulin-like growth factor I (IGF-I), IGF-II, IGF-binding protein-1 (IGFBP-1), IGFBP-3, and the soluble IGF-II/mannose-6-phosphate receptor in term human infants. The ALSPAC Study Team. Avon Longitudinal Study of Pregnancy and Childhood. J Clin Endocrinol Metab 2000; 85: 4266–4269. 11 Ong KK, Ahmed ML, Sherriff A, Woods KA, Watts A, Golding J et al. Cord blood leptin is associated with size at birth and predicts infancy weight gain in humans. ALSPAC Study Team. Avon Longitudinal Study of Pregnancy and Childhood. J Clin Endocrinol Metab 1999; 84: 1145–1148. 12 Tsai PJ, Yu CH, Hsu SP, Lee YH, Chiou CH, Hsu YW et al. Cord plasma concentrations of adiponectin and leptin in healthy term neonates: positive correlation with birthweight and neonatal adiposity. Clin Endocrinol (Oxf) 2004; 61: 88–93. 13 Pettitt DJ, McKenna S, McLaughlin C, Patterson CC, Hadden DR, McCance DR. Maternal glucose at 28 weeks of gestation is not associated with obesity in 2-year-old offspring: the Belfast Hyperglycemia and Adverse Pregnancy Outcome (HAPO) family study. Diabetes Care 2010; 33: 1219–1223. 14 Regnault N, Botton J, Heude B, Forhan A, Hankard R, Foliguet B et al. Higher cord C-peptide concentrations are associated with slower growth rate in the 1st year of life in girls but not in boys. Diabetes 2011; 60: 2152–2159. 15 Shields BM, Knight B, Hopper H, Hill A, Powell RJ, Hattersley AT et al. Measurement of cord insulin and insulin-related peptides suggests that girls are more insulin resistant than boys at birth. Diabetes Care 2007; 30: 2661–2666.
ª 2013 The Authors. Diabetic Medicine ª 2013 Diabetes UK
Research article
16 Hauner H, Much D, Vollhardt C, Brunner S, Schmid D, Sedlmeier EM et al. Effect of reducing the n-6:n-3 long-chain PUFA ratio during pregnancy and lactation on infant adipose tissue growth within the first year of life: an open-label randomized controlled trial. Am J Clin Nutr 2012; 95: 383–394. 17 Hauner H, Vollhardt C, Schneider KT, Zimmermann A, Schuster T, Amann-Gassner U. The impact of nutritional fatty acids during pregnancy and lactation on early human adipose tissue development. Rationale and design of the INFAT study. Ann Nutr Metab 2009; 54: 97–103. 18 Coustan DR, Lowe LP, Metzger BE, Dyer AR. International Association of Diabetes and Pregnancy Study Groups. The Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study: paving the way for new diagnostic criteria for gestational diabetes mellitus. Am J Obstet Gynecol 2010; 202: e651–e656. 19 Weststrate JA, Deurenberg P. Body composition in children: proposal for a method for calculating body fat percentage from total body density or skinfold-thickness measurements. Am J Clin Nutr 1989; 50: 1104–1115. 20 Holzhauer S, Zwijsen RM, Jaddoe VW, Boehm G, Moll HA, Mulder PG et al. Sonographic assessment of abdominal fat distribution in infancy. Eur J Epidemiol 2009; 24: 521–529. 21 Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985; 28: 412–419. 22 Wilkin TJ, Murphy MJ. The gender insulin hypothesis: why girls are born lighter than boys, and the implications for insulin resistance. Int J Obes (Lond) 2006; 30: 1056–1061. 23 Murphy MJ, Metcalf BS, Voss LD, Jeffery AN, Kirkby J, Mallam KM et al. Girls at five are intrinsically more insulin resistant than boys: The Programming Hypotheses Revisited—The EarlyBird Study (EarlyBird 6). Pediatrics 2004; 113: 82–86. 24 Bhargava SK, Sachdev HS, Fall CH, Osmond C, Lakshmy R, Barker DJ et al. Relation of serial changes in childhood body mass index to impaired glucose tolerance in young adulthood. N Engl J Med 2004; 350: 865–875.
ª 2013 The Authors. Diabetic Medicine ª 2013 Diabetes UK
DIABETICMedicine
25 Barker DJ, Osmond C, Forsen TJ, Kajantie E, Eriksson JG. Trajectories of growth among children who have coronary events as adults. N Engl J Med 2005; 353: 1802–1809. 26 Parker M, Rifas-Shiman SL, Belfort MB, Taveras EM, Oken E, Mantzoros C et al. Gestational glucose tolerance and cord blood leptin levels predict slower weight gain in early infancy. J Pediatr 2011; 158: 227–233. 27 Heerwagen MJ, Miller MR, Barbour LA, Friedman JE. Maternal obesity and fetal metabolic programming: a fertile epigenetic soil. Am J Physiol Regul Integr Comp Physiol 2010; 299: R711–722. 28 Barbour LA, Hernandez TL, Reece MS, Horton TJ, Kahn BF, Nadeau KJ et al. Change in fasting triglycerides from early to late gestation is highly predictive of neonatal adiposity. Diabetes 2009; 58: A84. 29 Vrijkotte TG, Krukziener N, Hutten BA, Vollebregt KC, van Eijsden M, Twickler MB. Maternal lipid profile during early pregnancy and pregnancy complications and outcomes: The ABCD Study. J Clin Endocrinol Metab 2012; 97: 3917–3925. 30 Thangaratinam S, Rogozinska E, Jolly K, Glinkowski S, Roseboom T, Tomlinson JW et al. Effects of interventions in pregnancy on maternal weight and obstetric outcomes: meta-analysis of randomised evidence. Br Med J 2012; 344: e2088.
Supporting Information Additional Supporting Information may be found in the online version of this article: Table S1. Maternal insulin and HOMA-IR measured at the 32nd week of gestation in relation to infant outcomes up to 2 years post-partum. Table S2. Maternal triglycerides at the 32nd week of gestation and change in triglycerides from the 15th week until the 32nd week of gestation in relation to infant outcomes up to 2 years post-partum.
1507
Research: Pathophysiology Maternal insulin resistance, triglycerides and cord blood insulin in relation to post-natal weight trajectories and body composition in the offspring up to 2 years S. Brunner1, D. Schmid1, K. Hu¨ttinger1, D. Much1, E. Heimberg2, E.-M. Sedlmeier3,4, M. Bru¨derl5, J. Kratzsch6, B. L. Bader3,4, U. Amann-Gassner1 and H. Hauner1,3 1 Else Kro¨ner-Fresenius-Center for Nutritional Medicine, Klinikum rechts der Isar, Technische Universita¨t Mu¨nchen, Munich, 2Department of Pediatrics, Universita¨tsklinikum Tu¨bingen, 3ZIEL—Research Center for Nutrition and Food Sciences, Nutritional Medicine Unit, 4ZIEL—PhD Graduate School, ‘Epigenetics, Imprinting and Nutrition’, Technische Universita¨t Mu¨nchen, Freising, 5Institute for Medical Statistics and Epidemiology, Klinikum rechts der Isar, Technische Universita¨t Mu¨nchen, Munich and 6Institute of Laboratory Medicine, Clinical Chemistry and Molecular Diagnostics, University Hospital Leipzig, Leipzig, Germany
Accepted 30 July 2013
Abstract Aims The intrauterine metabolic environment might have a programming effect on offspring body composition. We aimed to explore associations of maternal variables of glucose and lipid metabolism during pregnancy, as well as cord blood insulin, with infant growth and body composition up to 2 years post-partum. Methods Data of pregnant women and their infants came from a randomized controlled trial designed to investigate the impact of nutritional fatty acids on adipose tissue development in the offspring. Of the 208 pregnant women enrolled, 118 infants were examined at 2 years. In the present analysis, maternal fasting plasma insulin, homeostasis model assessment of insulin resistance and serum triglycerides measured during pregnancy, as well as insulin in umbilical cord plasma, were related to infant growth and body composition assessed by skinfold thickness measurements and abdominal ultrasonography up to 2 years of age.
Maternal homeostasis model assessment of insulin resistance at the 32nd week of gestation was significantly inversely associated with infant lean body mass at birth, whereas the change in serum triglycerides during pregnancy was positively associated with ponderal index at 4 months, but not at later time points. Cord plasma insulin correlated positively with birthweight and neonatal fat mass and was inversely associated with body weight gain up to 2 years after multiple adjustments. Subsequent stratification by gender revealed that this relationship with weight gain was stronger, and significant only in girls.
Results
Conclusions Cord blood insulin is inversely associated with subsequent infant weight gain up to 2 years and this seems to be more pronounced in girls.
Diabet. Med. 30, 1500–1507 (2013)
Introduction Because of the increasing global prevalence of childhood overweight and obesity [1], the identification of early determinants for adiposity development has received considerable attention. Several studies suggest a role of the metabolic milieu during pregnancy for fetal and later infant growth or body composition. The most convincing evidence comes from studies in the field of gestational diabetes showing higher rates of macrosomia and increased neonatal Correspondence to: Hans Hauner. E-mail [email protected] (Clinical Trials Registry No; NCT 00362089)
1500
fat mass in diabetic pregnancies [2], as well as a continuous association between maternal glycaemia and neonatal adiposity, even in the non-diabetic range [3]. Maternal insulin resistance during pregnancy might further translate into higher weight gain and adiposity in infancy [4]. Besides, maternal variables of lipid metabolism, including triglycerides, have been shown to be related to fetal growth and newborn fat mass [5–7], although this association may vary depending on maternal glucose tolerance status [8] or BMI category [9]. However, longer-term associations of maternal lipids with infant growth variables or adiposity beyond birth are largely unstudied.
ª 2013 The Authors. Diabetic Medicine ª 2013 Diabetes UK
Research article
What’s new? • Associations of maternal metabolic variables and cord blood insulin with infant anthropometry beyond birth have not been adequately addressed. • We investigated the relationship of maternal insulin, homeostasis model assessment of insulin resistance, triglycerides and cord blood insulin with infant weight gain and body composition up to 2 years. • Maternal metabolic variables were transiently related to infant growth and body composition outcomes. • Cord blood insulin was inversely associated with weight gain up to 2 years, and this relationship was stronger and significant only in girls. • Fetal insulin exposure might differentially affect subsequent weight gain up to 2 years in both sexes.
Likewise, insulin in umbilical cord blood, produced by the fetus in response to maternal glycaemia, has been widely investigated in relation to neonatal anthropometry and measures of adiposity [10–12], but data on its relationship with weight gain and body composition later in infancy are scarce [11,13,14]. Recent data indicate that girls are already intrinsically more insulin-resistant than boys at birth [15]. Later on, fetal insulin exposure might have differential implications for post-natal weight development between the sexes, with girls being potentially programmed towards slower growth rates by fetal hyperinsulinemia [14]. We aimed to investigate the relationship of maternal metabolic variables [insulin, insulin resistance expressed by homeostasis model assessment (HOMA-IR) and triglycerides] measured in late pregnancy (32nd week of gestation) and cord blood insulin with infant growth and body composition from birth up to 2 years of age and to explore potential sex-specific differences.
Subjects and methods Data came from the Impact of Nutritional Fatty Acids on Infant Adipose Tissue Development (INFAT) study, an open-label randomized controlled trial originally designed to examine the effect of reducing the maternal dietary n-6: n-3 fatty acid ratio during pregnancy and lactation on infant adipose tissue development. The study design and the clinical results on infant fat mass up to 1 year of age were previously described [16,17]. In brief, 208 healthy pregnant women with singleton pregnancies and a pre-pregnancy BMI between 18 and 30 kg/m2 were enrolled and randomly assigned to either an intervention (n = 104) or a control group (n = 104) from
ª 2013 The Authors. Diabetic Medicine ª 2013 Diabetes UK
DIABETICMedicine
the 15th week of gestation until 4 months post-partum. Women in the intervention group received (1) a dietary supplement containing 1200 mg n-3 long-chain polyunsaturated fatty acids per day and (2) nutritional counselling to normalize the consumption of arachidonic acid to a moderate level of intake. In contrast, women in the control group were advised to keep to a healthy diet according to current recommendations of the German Nutrition Society. The study protocol was approved by the ethical committee of the Technische Universit€at M€ unchen (no. 1479/06/2006/2/21) and all participants gave written informed consent.
Collection of samples
Maternal blood was collected at the 15th and 32nd week of gestation in the morning after an overnight fast. Immediately after delivery, umbilical vein ethylenediaminetetraacetic acid (EDTA) blood samples were collected. Umbilical cord blood was analysed from 137 newborns. Plasma was prepared by centrifugation at 2000 g at 4 °C for 10 min and subsequently aliquoted and stored at 80 °C.
Maternal characteristics and infant anthropometry
Maternal pre-pregnancy weight and height were retrieved from the maternity card, a booklet containing medical information related to antenatal and natal care provided to expecting mothers by their gynaecologist. Maternal glucose tolerance status was defined based on clinical diagnosis by the women’s gynaecologist. Clinically diagnosed gestational diabetes was mostly confirmed by a standardized oral glucose tolerance test and defined based on thresholds according to Coustan et al. [18]. In only one case, the diagnosis was based on elevated glucose at random. The therapy regimen (diet or insulin treatment) of women with gestational diabetes was retrieved from their medical record. The infants were examined at birth (for skinfolds: 3–5 days post-partum), at 6 weeks, 4 months, 1 and 2 years post-partum. Birthweight, length, sex and gestational age of the newborn were retrieved from the medical record. Anthropometric measurements of the infants were taken by trained investigators according to standardized procedures [16]. Skinfolds were measured in triplicate with a Holtain caliper (Holtain Ltd, Crymych, UK) at the left body axis at four sites (triceps, biceps, subscapular and suprailiac). Body fat percentage was calculated via predictive skinfold equations according to Weststrate et al. [19]. Fat mass was calculated as percentage of body fat multiplied by the current body weight of the infant. Lean body mass was calculated as body weight minus fat mass. Abdominal ultrasonography to estimate subcutaneous abdominal and preperitoneal fat was performed by two well-trained paediatricians at 6 weeks, 4 months, 1 and 2 years post-partum according to Holzhauer et al. [20].
1501
DIABETICMedicine
Cord blood insulin in relation to offspring weight gain S. Brunner et al.
Laboratory analyses
Plasma insulin was measured by a chemiluminescence immunoassay using the LIAISON analyser (Diasorin, Saluggia, Italy). Inter- and intra-assay coefficients of variation were both < 4%. Fasting plasma glucose and serum triglyceride levels were determined by an approved external laboratory (Synlab Labordienstleistungen, Munich, Germany). HOMA-IR was calculated from fasting glucose and insulin according to Matthews et al. [21].
Statistical analysis
Statistical analyses were performed with the R software package (version 2.8.1; R Foundation for Statistical Computing, http://www.r-project.org) and Predictive Analytics SoftWare (PASW) software (version 18.0; SPSS Inc., Chicago, IL, USA). To determine differences in the maternal variables between the randomized groups, multiple linear regression models [F-test, analysis of covariance (ANCOVA)] adjusting for age and pre-pregnancy BMI as additional independent variables were employed. To analyse differences in cord blood insulin between the groups or sexes, respectively, Mann– Whitney tests were used. To assess associations of the various maternal metabolic variables at the 32nd week of gestation, and cord blood insulin with infant clinical outcomes up to 2 years of age, data of both groups were pooled. Multiple linear regression models, including the covariates maternal pre-pregnancy BMI, gestational weight gain, maternal glucose tolerance status, pregnancy duration, sex and group allocation for the data at birth, and, additionally, ponderal index at birth and mode of infant feeding at the later time points, were performed. A two-sided P-value < 0.05 was considered statistically significant.
Results We examined 188 newborns at birth, 180 infants at 6 weeks, 174 at 4 months, 170 at 1 year and 118 at 2 years of age. Maternal, newborn and infant characteristics up to 2 years are presented in Table 1. Individuals who completed all assessments did not significantly differ from the rest of the study population in the major socio-demographic and clinical variables (data not shown). Maternal fasting insulin and HOMA-IR at the 32nd week of gestation did not significantly differ between the randomized groups (data not shown). In contrast, triglyceride levels increased significantly less from baseline (15th week of gestation) until 32nd week of pregnancy in the intervention compared with the control group (D intervention 76.2 53.5 mg/dl vs. control 103.6 58.1 mg/dl, P < 0.001), resulting in significantly higher levels in the control group compared with the intervention group at the 32nd week of gestation (P < 0.001). Cord plasma insulin concentrations
1502
did not differ by group (data not shown). Girls displayed significantly higher cord plasma insulin levels compared with boys (P = 0.037, Table 1), and also when extreme outliers with very high insulin levels (> 3 9 interquartile range, n = 7, P = 0.022) were excluded.
Maternal insulin, HOMA-IR and triglycerides at the 32nd week of gestation in relation to infant clinical outcomes up to 2 years post-partum
Maternal insulin, HOMA-IR and triglyceride levels at the 32nd week of gestation were found to be largely unrelated to infant growth and body composition outcomes up to 2 years post-partum. However, HOMA-IR was significantly inversely associated with lean body mass at birth [adjusted regression coefficient (badj) (95% CI) 54.94 ( 99.23 to 10.64) g, P = 0.016] in the analysis controlling for maternal pre-pregnancy BMI, gestational weight
Table 1 Maternal, newborn and 2-year infant characteristics Maternal variables (n = 208 at baseline) Age at enrollment (years) 31.8 4.7 Pre-pregnancy BMI (kg/m2) 22.3 3.0 Parity (primiparae) 58.5 Gestational diabetes 9.0 Dietary treatment 76.5 Insulin treatment 23.5 Gestational weight gain (kg) 15.6 4.9 Education (≥ 12 years at school) 69.1 Energy intake at 32nd week of 2079 415 gestation (kcal/day) (n = 166) Insulin at 32nd week of 76.7 43.1 gestation (pmol/l) (n = 182) HOMA-IR at 32nd week of 2.2 1.2 gestation (n = 175) Triglycerides at 32nd week of 197.0 66.2 gestation (mg/dl) (n = 187) Newborn variables (n = 188) Sex (girls) 47.9 Gestational age (weeks) 39.6 1.5 Birthweight (g) 3443 520 Ponderal index (kg/m3) 24.8 2.4 Sum of four skinfolds (mm) (n = 168) 16.0 2.6 Body fat (%) (n = 168) 13.8 2.7 Cord blood insulin (pmol/l) (n = 137) 28.30 (36.40) Girls (n = 63) 31.90 (37.10)* Boys (n = 74) 25.85 (24.53)* Mode of infant feeding (4 months post-partum) Exclusively breastfed 64.6 Partially breastfed 15.2 Formula fed 20.2 Infant variables at 2 years (n = 118) Body weight (g) 12393 1395 Ponderal index (kg/m3) 18.8 1.8 Sum of four skinfolds (mm) (n = 110) 23.7 3.4 Body fat (%) (n = 110) 19.1 2.3 Weight gain (birth to 2 years) (g) 8919 1306 Data are presented as mean SD, median (interquartile range) (for cord blood insulin) or percentage. *Significant difference between the sexes (P = 0.037), Mann– Whitney test.
ª 2013 The Authors. Diabetic Medicine ª 2013 Diabetes UK
Research article
gain, maternal glucose tolerance status, pregnancy duration, group and sex (see also Supporting Information, Table S1). Moreover, the change in maternal serum triglyceride concentration between the 15th and 32nd week of gestation was weakly, but significantly associated with infant ponderal index at 4 months post-partum [badj: 0.01 (0 to 0.01) kg/m3, P = 0.020], but not with any of the other growth or body composition outcomes up to 2 years post-partum. No significant relationships were found for absolute triglyceride levels (see also Supporting Information, Table S2).
Cord plasma insulin in relation to infant clinical outcomes up to 2 years post-partum
In the analysis comprising all available data at birth, cord blood insulin was significantly positively related to birthweight [b: 1.68 (0.03 to 3.33) g, P = 0.048], the sum of four skinfolds [b: 0.02 (0.01 to 0.02) mm, P = 0.001], percentage body fat [b: 0.02 (0.01 to 0.03)%, P = 0.001], as well as the absolute amount of fat mass [b: 0.81 (0.31 to 1.31) g, P = 0.002] in the newborns. These relationships remained significant after adjustment for maternal pre-pregnancy BMI, gestational weight gain, maternal glucose tolerance status, pregnancy duration, group and sex (Table 2). In contrast, significant inverse relationships of cord blood insulin were found with infant weight, BMI, fat and lean body mass at 2 years, as well as with weight gain from birth up to 2 years in the unadjusted analysis (Table 2). However, in the adjusted model controlling for maternal pre-pregnancy BMI, gestational weight gain, maternal glucose tolerance status, pregnancy duration, group, sex, ponderal index at birth and mode of infant feeding at 4 months post-partum, only the relationship with weight gain remained significant [badj: 6.90 ( 13.13 to 0.67) g, P = 0.030] (Table 2). When the analysis was restricted to participants completing all study visits up to 2 years (n = 118, “completers”), significant inverse associations with infant anthropometrics were already apparent at 1 year of age (BMI: P = 0.015; sum of four skinfolds: P = 0.036; percentage body fat: P = 0.027; fat mass: P = 0.024; weight gain up to 1 year: P = 0.010 in the adjusted model) (Table 2). Testing for sex interaction revealed no significant interaction term for the association with weight gain up to 2 years (P = 0.710). Nevertheless, the relationship of cord blood insulin with weight gain up to 2 years turned out to be much stronger in girls compared with boys [girls: badj 7.63 ( 15.29 to 0.03) g, P = 0.051; boys: badj 2.82 ( 15.76 to 10.11) g, P = 0.662; Fig. 1]. No significant associations were found with the data set on infant fat mass assessed by ultrasonography, neither for maternal variables nor for cord blood insulin (data not shown).
ª 2013 The Authors. Diabetic Medicine ª 2013 Diabetes UK
DIABETICMedicine
Discussion In this study, we investigated associations of various maternal markers of glucose and lipid metabolism and cord blood insulin with subsequent infant growth and body composition over the first 2 years of life. The major finding of our study was that cord blood insulin was inversely associated with anthropometric measures from the first year of life onwards, but only the association with weight gain persisted up to 2 years after adjustment for covariates. This association was significant only in girls, which is consistent with recent findings showing cord C-peptide as a proxy for fetal insulin to be inversely associated with infant weight development over the 1st year of life in girls, but not in boys [14]. In accordance with other studies [14,15], we observed significantly higher cord plasma insulin concentrations in girls compared with boys, although this was not entirely consistent across the literature [10–12]. In view of the lower birthweights but higher cord blood insulin levels in girls, it was suggested that girls might be inherently more insulin-resistant in utero and around birth compared with boys [15]. This ‘gender insulin hypothesis’ proposes that gender-specific genes affecting insulin sensitivity might account for the gender difference in birthweight [22]. Our data and previous findings thus might suggest a programming effect of fetal insulin exposure for early weight gain, for which girls may be more susceptible than boys [14]. The consequences of such an effect with regard to later risk for metabolic diseases remain to be clarified. It has been reported that girls are also more insulin-resistant compared with boys in childhood and that Type 2 diabetes is more common in girls [23]. Furthermore, observational studies have related slower early growth patterns with insulin resistance or diabetes in later life [24,25]. Slower weight gain has also been reported in infants born to mothers with gestational diabetes [26], which could be interpreted as a programming effect of higher fetal insulin concentrations as well. In contrast to fetal insulin, maternal variables of glucose metabolism were largely unrelated to the infant clinical outcomes, except that higher maternal HOMA-IR at the 32nd week of gestation was associated with lower lean body mass at birth. Although our data showed no explicit relationship with fat mass, this finding might to some extent be compatible with data showing increased body fat in newborns from mothers with gestational diabetes compared with women with normal glucose tolerance [2] and the direct association of maternal glucose homeostasis with neonatal adiposity across the full range of maternal glycaemia, and not restricted to manifest diabetes [3]. However, in the follow-up of the Hyperglycaemia and Adverse Pregnancy Outcome (HAPO) study, no such relationship could be found with adiposity at 2 years of age within a group of pregnant women without diabetes [13].
1503
1504 0.00 0.73 2.39 4.65
123 123 123 127
0.02 to 0.01) 4.65 to 0.07) 10.00 to 1.21) 14.46 to 2.85)
0.01 2.36 5.60 8.66
90 90 90 93
( ( ( (
7.60 ( 13.79 to 1.41) 0.01 ( 0.01 to 0.00) 0.01 ( 0.01 to 0.00)
93 93 90
to 0.01) to 0.99) to 0.37) to 0.92)
3.00 ( 7.06 to 1.06) 0.01 ( 0.01 to 0) 0.00 ( 0.02 to 0.02)
127 127 127 0.01 2.44 5.15 8.39
0.02 (0.01 to 0.03) 0.81 (0.31 to 1.31) 0.89 ( 0.39 to 2.17)
130 130 130
( ( ( (
1.68 0.00 0.00 0.02
137 137 137 130
(0.03 to 3.33) ( 0.01 to 0.01) (0 to 0.01) (0.01 to 0.02)
b (95% CI)†
0.269 0.043* 0.013* 0.004*
0.017* 0.021* 0.064
0.733 0.408 0.092 0.016*
0.150 0.053 0.930
0.001* 0.002* 0.176
0.048* 0.875 0.434 0.001*
P
(0.27 to 3.47) ( 0.01 to 0.01) ( 0.00 to 0.01) (0.01 to 0.03)
( ( ( (
0.01 2.18 3.78 7.02
to to to to
0.01) 1.33) 1.44) 0.28)
0.01 2.02 3.80 6.90
( ( ( (
0.02 to 0.01) 4.48 to 0.43) 8.45 to 0.85) 13.13 to 0.67)
5.50 ( 12.12 to 1.13) 0.01 ( 0.01 to 0.00) 0.01 ( 0.03 to 0.01)
0.00 0.43 1.17 3.36
1.58 ( 5.48 to 2.32) 0.00 ( 0.01 to 0.00) 0.00 ( 0.02 to 0.02)
0.02 (0.01 to 0.03) 0.80 (0.28 to 1.32) 1.01 ( 0.21 to 2.24)
1.87 0.00 0.00 0.02
b (95% CI)†
P
0.249 0.105 0.108 0.030*
0.103 0.115 0.226
0.765 0.630 0.375 0.071
0.424 0.121 0.901
0.002* 0.003* 0.105
0.023* 0.803 0.423 0.003*
90 90 90 93
93 93 90
89 89 89
93 93 93 89
n
( 0.92 to 2.87) ( 0.02 to 0.01) ( 0.01 to 0.01) (0.00 to 0.03)
0.01 2.30 2.36 5.77
( ( ( (
0.03 to 0.00) 4.30 to 0.31) 5.45 to 0.73) 10.15 to 1.39)
4.37 ( 9.04 to 0.29) 0.01 ( 0.01 to 0.00) 0.02 ( 0.04 to 0.00)
0.01 (0.00 to 0.03) 0.63 (0.01 to 1.26) 0.41 ( 1.02 to 1.85)
0.98 0.01 0.00 0.01
b (95% CI)†
0.027* 0.024* 0.133 0.010*
0.066 0.015* 0.036*
0.025* 0.045* 0.567
0.308 0.382 0.833 0.031*
P
*P < 0.05. †Data are presented as the regression coefficient (b) along with the 95% confidence interval. ‡Analysis 1 was based on all available data at the respective time points. §Analysis 2 was based on the cohort with full data from birth until 2 years post-partum. ¶At the time point of birth, results were corrected for maternal pre-pregnancy BMI, gestational weight gain, pregnancy duration, group, sex and maternal glucose tolerance status in the adjusted model. At the later time points, results were additionally adjusted for ponderal index at birth and breastfeeding status (fully breastfed, partially breastfed or formula) at 4 months post-partum, respectively.
Birth Birthweight (g) Ponderal index (kg/m3) BMI (kg/m2) Sum of four skinfold thicknesses (mm) Body fat (%) Fat mass (g) Lean body mass (g) 1 year post-partum Weight (g) BMI (kg/m2) Sum of four skinfold thicknesses (mm) Body fat (%) Fat mass (g) Lean body mass (g) Weight gain (g) (birth–1 year post-partum) 2 years post-partum Weight (g) BMI (kg/m2) Sum of four skinfold thicknesses (mm) Body fat (%) Fat mass (g) Lean body mass (g) Weight gain (g) (birth–2 years post-partum)
n
Adjusted model¶
Unadjusted model
Adjusted model¶
Analysis 2 ‘completers only’§
Analysis 1 ‘all available data’‡
Table 2 Umbilical cord plasma insulin (pmol/l) in relation to infant outcomes at birth, 1 year and at 2 years post-partum
DIABETICMedicine Cord blood insulin in relation to offspring weight gain S. Brunner et al.
ª 2013 The Authors. Diabetic Medicine ª 2013 Diabetes UK
Research article
DIABETICMedicine
Girls (n = 39) b: –8.99 [–15.65; –2.33] g, P = 0.010 badj: –7.63 [–15.29; 0.03] g, P = 0.051
Infant sex Girls Boys Girls Boys
Boys (n = 54) b: –6.47 [–18.60; 5.66] g, P = 0.289 badj: –2.82 [–15.76; 10.11] g, P = 0.662
FIGURE 1 Association of cord plasma insulin with infant weight gain up to 2 years post-partum stratified by gender. b, regression coefficient (95% confidence interval) from linear regression analysis; badj, data adjusted for maternal pre-pregnancy BMI, gestational weight gain, pregnancy duration, group, maternal glucose tolerance status, ponderal index at birth and mode of infant feeding at 4 months post-partum (fully breastfed, partially breastfed or formula).
In another recent study, maternal insulin resistance during pregnancy, irrespective of glucose tolerance status, emerged as a significant independent predictor of infant weight gain and adiposity at 1 year of age measured by skinfold thicknesses [4]. However, these findings might not be directly comparable with our study because of differing methodological approaches. It could thus be plausible that measuring insulin resistance only in the fasting state as HOMA-IR is not sensitive enough to detect associations with infant adiposity or weight gain, compared with a more complex assessment of insulin resistance under a glucose challenge. Besides markers of glucose metabolism, we further assessed whether maternal triglyceride levels in later pregnancy are related to infant outcomes from birth until 2 years post-partum. Increasing triglyceride concentrations over the course of pregnancy are a characteristic feature of normal pregnancy. Higher triglyceride levels in the maternal circulation may enhance the concentration gradient across the placenta, resulting in accelerated transport and deposition of lipids in fetal tissues [27]. Against this theory, we could not find any significant associations of maternal triglycerides with neonatal body composition, apart from a weak relationship between the change of maternal serum triglyceride levels from 15th until the 32nd week of gestation and subsequent ponderal index at 4 months. However, this slight effect seems clinically irrelevant or might even have occurred by chance with regard to the numerous correlations explored. Thus, we could not confirm previous findings of a small pilot study showing the increase in triglycerides from ª 2013 The Authors. Diabetic Medicine ª 2013 Diabetes UK
early to late pregnancy to be highly predictive for neonatal adiposity [28]. Other studies reported associations between absolute maternal triglycerides levels and birthweight or large-for-gestational-age infants [5,9,29]. Different sampling time points, weight and health status of the women might explain the discrepancy to our findings. The strengths of our study are the extensive longitudinal assessment of infant body composition over multiple time points by two complementary methods from birth and through the first 2 years of life and the collection of biosamples both from the maternal and the fetal side. In addition, we assessed a large set of maternal and child factors related to infant adiposity to adjust for in the multivariate analysis. However, although not the focus of this study, other factors such as lifestyle in pregnancy, including physical activity and diet, might influence the investigated associations through their effect on gestational weight gain or birthweight [30]. A major limitation of our study is the considerable loss to follow-up beyond the 1st year of life. Although the individuals lost to follow-up were comparable with the completers in their main characteristics, our results might suggest a certain selection bias. Nevertheless, despite the relatively small study population, it is intriguing that we could still confirm sex-specific effects with regard to the associations of fetal insulin with post-natal weight gain. However, as no adjustments for multiple testing were made within this rather exploratory approach, all significant findings should be interpreted with caution. It should also be acknowledged, that neither skinfold measurements nor abdominal sonography represent direct methods for the
1505
DIABETICMedicine
Cord blood insulin in relation to offspring weight gain S. Brunner et al.
assessment of fat mass. Another limitation is that our study protocol did not include a standardized glucose tolerance test to assess insulin resistance under a glucose challenge, so that we had to rely on measures of insulin resistance in the fasting state. Furthermore, we did not measure other variables related to fetal insulin secretion, such as C-peptide or proinsulin, or lipids in umbilical cord blood, which might also come into consideration as potential determinants of fetal growth [7]. In conclusion, maternal insulin resistance and triglycerides in the last trimester of pregnancy were only transiently related to newborn/early post-natal infant growth and body composition and did not emerge as major determinants for infant weight development up to 2 years. In contrast, cord blood insulin was highly correlated with birthweight and newborn fat mass, and significantly inversely associated with weight gain over the first 2 years of life in girls. This might suggest a role for fetal insulin in programming energy homeostasis in early life, which takes effect differently between the sexes.
Funding sources
The study was funded by grants from the Else Kr€ oner-Fresenius Foundation, Bad Homburg, Germany; the International Unilever Foundation, Hamburg, Germany; the EU-funded EARNEST consortium (FOOD-CT-2005-007036); and the German Ministry of Education and Research via the Competence Network on Obesity (Kompetenznetz Adipositas, 01GI0842). There was no intervention from any sponsor with any of the research aspects of the study, including the study design, intervention, data collection, analysis and interpretation, or writing of the manuscript.
Competing interests
HH is on the Advisory Board for Weight Watchers International and has received grants from Riemser and Weight Watchers for clinical trials and payment for lectures from Novartis, Roche Germany, and Sanofi Aventis. All authors declare that there is no duality of interest associated with this manuscript.
Acknowledgements
We thank all the families who participated in the study. We would like to thank the technical staff of the Institute of Laboratory Medicine, Clinical Chemistry and Molecular Diagnostics,University Hospital Leipzig, Leipzig, Germany for assistance in performing the laboratory analyses. We thank Petra Wolf (Institute for Medical Statistics and Epidemiology, Klinikum rechts der Isar, Technische Universit€at M€ unchen, Munich, Germany) for statistical support.
1506
References 1 de Onis M, Blossner M, Borghi E. Global prevalence and trends of overweight and obesity among preschool children. Am J Clin Nutr 2010; 92: 1257–1264. 2 Catalano PM, Thomas A, Huston-Presley L, Amini SB. Increased fetal adiposity: a very sensitive marker of abnormal in utero development. Am J Obstet Gynecol 2003; 189: 1698–1704. 3 HAPO Study Group. Hyperglycemia and Adverse Pregnancy Outcome (HAPO) Study: associations with neonatal anthropometrics. Diabetes 2009; 58: 453–459. 4 Hamilton JK, Odrobina E, Yin J, Hanley AJ, Zinman B, Retnakaran R. Maternal insulin sensitivity during pregnancy predicts infant weight gain and adiposity at 1 year of age. Obesity (Silver Spring) 2010; 18: 340–346. 5 Di Cianni G, Miccoli R, Volpe L, Lencioni C, Ghio A, Giovannitti MG et al. Maternal triglyceride levels and newborn weight in pregnant women with normal glucose tolerance. Diabet Med 2005; 22: 21–25. 6 Harmon KA, Gerard L, Jensen DR, Kealey EH, Hernandez TL, Reece MS et al. Continuous glucose profiles in obese and normal-weight pregnant women on a controlled diet: metabolic determinants of fetal growth. Diabetes Care 2011; 34: 2198– 2204. 7 Schaefer-Graf UM, Graf K, Kulbacka I, Kjos SL, Dudenhausen J, Vetter K et al. Maternal lipids as strong determinants of fetal environment and growth in pregnancies with gestational diabetes mellitus. Diabetes Care 2008; 31: 1858–1863. 8 Schaefer-Graf UM, Meitzner K, Ortega-Senovilla H, Graf K, Vetter K, Abou-Dakn M et al. Differences in the implications of maternal lipids on fetal metabolism and growth between gestational diabetes mellitus and control pregnancies. Diabet Med 2011; 28: 1053–1059. 9 Misra VK, Trudeau S, Perni U. Maternal serum lipids during pregnancy and infant birthweight: the influence of prepregnancy BMI. Obesity (Silver Spring) 2011; 19: 1476–1481. 10 Ong K, Kratzsch J, Kiess W, Costello M, Scott C, Dunger D. Size at birth and cord blood levels of insulin, insulin-like growth factor I (IGF-I), IGF-II, IGF-binding protein-1 (IGFBP-1), IGFBP-3, and the soluble IGF-II/mannose-6-phosphate receptor in term human infants. The ALSPAC Study Team. Avon Longitudinal Study of Pregnancy and Childhood. J Clin Endocrinol Metab 2000; 85: 4266–4269. 11 Ong KK, Ahmed ML, Sherriff A, Woods KA, Watts A, Golding J et al. Cord blood leptin is associated with size at birth and predicts infancy weight gain in humans. ALSPAC Study Team. Avon Longitudinal Study of Pregnancy and Childhood. J Clin Endocrinol Metab 1999; 84: 1145–1148. 12 Tsai PJ, Yu CH, Hsu SP, Lee YH, Chiou CH, Hsu YW et al. Cord plasma concentrations of adiponectin and leptin in healthy term neonates: positive correlation with birthweight and neonatal adiposity. Clin Endocrinol (Oxf) 2004; 61: 88–93. 13 Pettitt DJ, McKenna S, McLaughlin C, Patterson CC, Hadden DR, McCance DR. Maternal glucose at 28 weeks of gestation is not associated with obesity in 2-year-old offspring: the Belfast Hyperglycemia and Adverse Pregnancy Outcome (HAPO) family study. Diabetes Care 2010; 33: 1219–1223. 14 Regnault N, Botton J, Heude B, Forhan A, Hankard R, Foliguet B et al. Higher cord C-peptide concentrations are associated with slower growth rate in the 1st year of life in girls but not in boys. Diabetes 2011; 60: 2152–2159. 15 Shields BM, Knight B, Hopper H, Hill A, Powell RJ, Hattersley AT et al. Measurement of cord insulin and insulin-related peptides suggests that girls are more insulin resistant than boys at birth. Diabetes Care 2007; 30: 2661–2666.
ª 2013 The Authors. Diabetic Medicine ª 2013 Diabetes UK
Research article
16 Hauner H, Much D, Vollhardt C, Brunner S, Schmid D, Sedlmeier EM et al. Effect of reducing the n-6:n-3 long-chain PUFA ratio during pregnancy and lactation on infant adipose tissue growth within the first year of life: an open-label randomized controlled trial. Am J Clin Nutr 2012; 95: 383–394. 17 Hauner H, Vollhardt C, Schneider KT, Zimmermann A, Schuster T, Amann-Gassner U. The impact of nutritional fatty acids during pregnancy and lactation on early human adipose tissue development. Rationale and design of the INFAT study. Ann Nutr Metab 2009; 54: 97–103. 18 Coustan DR, Lowe LP, Metzger BE, Dyer AR. International Association of Diabetes and Pregnancy Study Groups. The Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study: paving the way for new diagnostic criteria for gestational diabetes mellitus. Am J Obstet Gynecol 2010; 202: e651–e656. 19 Weststrate JA, Deurenberg P. Body composition in children: proposal for a method for calculating body fat percentage from total body density or skinfold-thickness measurements. Am J Clin Nutr 1989; 50: 1104–1115. 20 Holzhauer S, Zwijsen RM, Jaddoe VW, Boehm G, Moll HA, Mulder PG et al. Sonographic assessment of abdominal fat distribution in infancy. Eur J Epidemiol 2009; 24: 521–529. 21 Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985; 28: 412–419. 22 Wilkin TJ, Murphy MJ. The gender insulin hypothesis: why girls are born lighter than boys, and the implications for insulin resistance. Int J Obes (Lond) 2006; 30: 1056–1061. 23 Murphy MJ, Metcalf BS, Voss LD, Jeffery AN, Kirkby J, Mallam KM et al. Girls at five are intrinsically more insulin resistant than boys: The Programming Hypotheses Revisited—The EarlyBird Study (EarlyBird 6). Pediatrics 2004; 113: 82–86. 24 Bhargava SK, Sachdev HS, Fall CH, Osmond C, Lakshmy R, Barker DJ et al. Relation of serial changes in childhood body mass index to impaired glucose tolerance in young adulthood. N Engl J Med 2004; 350: 865–875.
ª 2013 The Authors. Diabetic Medicine ª 2013 Diabetes UK
DIABETICMedicine
25 Barker DJ, Osmond C, Forsen TJ, Kajantie E, Eriksson JG. Trajectories of growth among children who have coronary events as adults. N Engl J Med 2005; 353: 1802–1809. 26 Parker M, Rifas-Shiman SL, Belfort MB, Taveras EM, Oken E, Mantzoros C et al. Gestational glucose tolerance and cord blood leptin levels predict slower weight gain in early infancy. J Pediatr 2011; 158: 227–233. 27 Heerwagen MJ, Miller MR, Barbour LA, Friedman JE. Maternal obesity and fetal metabolic programming: a fertile epigenetic soil. Am J Physiol Regul Integr Comp Physiol 2010; 299: R711–722. 28 Barbour LA, Hernandez TL, Reece MS, Horton TJ, Kahn BF, Nadeau KJ et al. Change in fasting triglycerides from early to late gestation is highly predictive of neonatal adiposity. Diabetes 2009; 58: A84. 29 Vrijkotte TG, Krukziener N, Hutten BA, Vollebregt KC, van Eijsden M, Twickler MB. Maternal lipid profile during early pregnancy and pregnancy complications and outcomes: The ABCD Study. J Clin Endocrinol Metab 2012; 97: 3917–3925. 30 Thangaratinam S, Rogozinska E, Jolly K, Glinkowski S, Roseboom T, Tomlinson JW et al. Effects of interventions in pregnancy on maternal weight and obstetric outcomes: meta-analysis of randomised evidence. Br Med J 2012; 344: e2088.
Supporting Information Additional Supporting Information may be found in the online version of this article: Table S1. Maternal insulin and HOMA-IR measured at the 32nd week of gestation in relation to infant outcomes up to 2 years post-partum. Table S2. Maternal triglycerides at the 32nd week of gestation and change in triglycerides from the 15th week until the 32nd week of gestation in relation to infant outcomes up to 2 years post-partum.
1507
