In pregnant women with type 1 diabetes mellitus. Higher Cord Blood Levels of Fatty Acids in Pregnant Women With Type 1 Diabetes Mellitus

Transcription

1 CLINICAL RESEARCH ARTICLE Higher Cord Blood Levels of Fatty Acids in Pregnant Women With Type 1 Diabetes Mellitus Josip Djelmis, 1 Marina Ivanišević, 1 Gernot Desoye, 2 Mireille van Poppel, 3 Edina Berberović, 4 Dragan Soldo, 5 and Slavko Oreskovic 1 1 Referral Center for Diabetes in Pregnancy, Ministry of Health Republic of Croatia, Clinical Department of Obstetrics and Gynecology, Zagreb University Hospital Center, School of Medicine, University of Zagreb, HR Zagreb, Croatia; 2 Department of Obstetrics and Gynecology, Medical University of Graz, A-8036 Graz, Austria; 3 Institute of Sport Science, University of Graz, A-8036 Graz, Austria; 4 Clinical Department of Obstetrics and Gynecology, Holy Spirit University Hospital, HR Zagreb, Croatia; and 5 Department of Obstetrics and Gynecology, Mostar Clinical Hospital Center, School of Medicine, University of Mostar, BIH Mostar, Bosnia and Herzegovina Context: Type 1 diabetes mellitus (T1DM) is associated with a disturbance of carbohydrate and lipid metabolism. Objective: To determine whether T1DM alters maternal and neonatal fatty acid (FA) levels. Design: Observational study. Setting: Academic hospital. Patients: Sixty pregnant women (30 women with T1DM with good glycemic control and 30 healthy women) were included in the study. Maternal blood, umbilical vein, and artery blood samples were collected immediately upon delivery. Following lipid extraction, the FA profiles of the total FA pool of maternal serum and umbilical vein and artery serum were determined by gas chromatography. Results: Total FA concentration in maternal serum did not differ between the study groups; it was significantly higher in umbilical vein serum of the T1DM group compared with that in the control group [median (interquartile range)]: T1DM ( to ) and control (657.5 to ; P, 0.001), and in umbilical artery vein serum: T1DM ( to ) and control (643.3 to ; P, 0.001). Composition of FAs in umbilical vein serum showed significantly higher concentrations of saturated, monounsaturated, and polyunsaturated FAs (SFAs, MUFAs, and PUFAs, respectively) in the T1DM group than compared with those in the control group (P = 0.001). Furthermore, cord blood levels of leptin (P, 0.001), C-peptide (P, 0.001), and insulin resistance (P = 0.015) were higher in the T1DM group compared with controls. Conclusion: The neonates born to mothers with T1DM had higher concentrations of total FAs, SFAs and MUFAs, as well as PUFAs, compared with control newborns. (J Clin Endocrinol Metab 103: , 2018) In pregnant women with type 1 diabetes mellitus (T1DM), endogenous insulin production is absent or minimal, thus requiring exogenous insulin for glycemic control and prevention of ketoacidosis (1). In addition to disturbances of carbohydrate metabolism, T1DM is associated with changes in the content and composition of maternal lipids (2). Whether T1DM also influences content and composition of fetal lipids in the umbilical circulation is currently unknown. ISSN Print X ISSN Online Printed in USA Copyright 2018 Endocrine Society Received 2 February Accepted 25 April First Published Online 1 May 2018 Abbreviations: AA, arachidonic acid; BMI, body mass index; CS, Caesarean section; DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; FA, fatty acid; GDM, gestational diabetes; HOMA, homeostasis model assessment; LC-PUFA, long-chain polyunsaturated fatty acid; MUFA, monounsaturated fatty acid; PUFA, polyunsaturated fatty acid; r, Pearson correlation coefficient; rs, Spearman correlation coefficient; SFA, saturated fatty acid; T1/2DM, type 1/2 diabetes mellitus J Clin Endocrinol Metab, July 2018, 103(7): doi: /jc

2 doi: /jc Maternal sources of fetal lipids include lipoproteinassociated triacylglycerols, phospholipids, and cholesterol esters, as well as free fatty acids (FAs). The current understanding of maternal lipid supply for the fetus involves the transfer of free FAs and free cholesterol, after which, both moieties are esterified in the fetal liver and packaged into lipoproteins (2). Despite the presence of FA transporters in the syncytiotrophoblast (3), free FAs are likely to cross this membrane by diffusion (4, 5). Nonessential FAs can also be formed de novo from carbohydrates and acetate in the fetus under the control of fetal insulin (6, 7). In normal circumstances, free FAs have a minor role in energy provision; however, they generally can serve as energy deposits in the neonatal period (6). Particularly important for the fetus are the essential FAs, i.e., long-chain polyunsaturated FA (LC-PUFAs) (8). One of those LC-PUFAs, docosahexaenoic acid (DHA), has a major role in the development of nervous system cells, and its cord blood concentrations are associated with psychomotor development at 6 months of age (9, 10). In pregnant women, there is an increased maternalfetal transfer of essential over nonessential FAs (11, 12). The aim of the study was to investigate the impact of T1DM on materno-fetal levels of FAs. Specifically, we tested the hypothesis of whether T1DM alters the FA profiles of maternal blood, the umbilical vein, and artery blood. Materials and Methods Study population This prospective study included a total of 60 pregnant women with singleton pregnancies who had either T1DM (T1DM group, n = 30) or were healthy (control group, n = 30). Pregnant women with T1DM were followed up at the Department of Obstetrics and Gynecology, Zagreb University Hospital Center, from 2012 to Study women had a T1DM duration of 5 to 10 years, were without diabetes complications, had a normal course of pregnancy, and delivered healthy, eutrophic newborns within the 25th to 75th birthweight percentile, according to weeks of pregnancy, sex, and parity (13). Pregnant women with T1DM were on intensified insulin analogs therapy and had well-controlled glycemia. Healthy women underwent gestational diabetes (GDM) screening, according to International Association of Diabetes and Pregnancy Study Groups criteria, with 75 g oral glucose tolerance test between gestational weeks 24 and 34 (14). Women with preterm delivery, multiple pregnancies, diabetic complications (retinopathy, nephropathy, neuropathy, and chronic hypertension), fetal chromosomal anomalies, or malformations were excluded from the study. Additional exclusion criteria were smoking by self-report, n-3 and n-6 FA supplementation, gestational hypertension/pre-eclampsia, and GDM. Gestational age was determined by the first day of last menstruation and verified by ultrasound examination between 6 and 10 weeks of gestation. All pregnancies were terminated by elective Caesarean section (CS) between gestational weeks 39 and 40. The mothers fasted for at least 10 hours before the CS. Maternal blood glucose during delivery was maintained in the range of 4 to 5 mm by IV infusion of 500 ml 5% glucose with insulin. Pregnancies in mothers with T1DM were terminated by CS as a standard procedure, whereas in the control group, indications for CS were breach presentation, narrow pelvis, post-cs state, or CS on mother s demand. Maternal venous blood samples (5 ml) were collected during delivery. Umbilical cord blood was obtained during the CS, immediately after delivery but before the placenta was removed through puncture of the umbilical vessels. Blood was centrifuged immediately after collection, and the sera were stored at 280 C until analysis. C-Peptide and leptin concentrations and HbA 1c percentage were determined in maternal and umbilical vein serum. Free FA concentrations were quantified in maternal and umbilical vein and arterial serum. Ethical statement The study was approved by the Ethics Committee of the Clinical Department of Obstetrics and Gynecology, Zagreb University Hospital Center, School of Medicine, University of Zagreb (No /117A-2012). All women included in the study provided written, informed consent for themselves and their newborns. Data collection The following parameters were recorded: maternal height (centimeters) and weight (kilograms) measured before pregnancy; gestational weight gain as the difference between weight before pregnancy (self-reported) and at time of delivery; and prepregnancy body mass index (BMI; kilograms per square meter; BMI), calculated from prepregnancy values. Maternal HbA 1c was determined in each trimester in the T1DM group and between the 36th and 38th weeks of pregnancy in the control group. Neonatal birth weight (grams), birth length (centimeters), and 1 and 5 minute Apgar scores were measured. Blood sample analyses FA quantification Total lipids were extracted by a mixture of chloroform/ methanol solvent, a method of increasing polarity, according to Folch et al. (15). Heptadecanoic acid (C17:0) was used as the internal standard. FAs from lipid extracts were converted to methyl esters by trans-esterification with methanolic HCl (International Organization for Standardization standard). The FA profile was determined by gas chromatography-mass spectrometry on a Varian 3400 (Varian, Palo Alto, CA), equipped with a Saturn II ion trap mass spectrometer operating in the electron impact mode, as previously described (12). The concentration of each FA (micrograms per milliliter) was quantified by comparing the area of the internal standard peak (C17:0) with the peak area surface of a single FA. The results are expressed as percent (micrograms per 100 mg FAs) of arachidonic acid (AA) and DHA. Glucose, HbA 1c, C-peptide, and leptin quantification Glucose levels were quantified by the hexokinase method on a Cobas C301 analyzer, and HbA 1c levels in whole blood

3 2622 Djelmis et al T1DM Alters Neonatal Fatty Acids J Clin Endocrinol Metab, July 2018, 103(7): were measured by the Turbidimetric Inhibition Immunoassay on a Cobas C501 analyzer with reagents from the manufacturer (Roche, Basel, Switzerland). C-peptide concentration was determined by ELISA (C-peptide ELISA; Mercodia, Uppsala, Sweden). Serum leptin concentration was determined by ELISA (Quantikine Human Leptin Immunoassay; R&D Systems, Minneapolis, MN). Neonatal insulin resistance was calculated according to homeostasis model assessment 2 (HOMA2) (16), using software available online ( Statistical analysis Statistical analyses were performed with SPSS statistical package (release by SPSS version 24 software; IBM, Chicago, IL). Continuous variables are expressed as the means 6 SD or medians (25th to 75th percentile) for a skewed distribution, and qualitative variables are presented as frequencies and percentages. Between-group differences in normally distributed continuous variables were assessed with a t test. The Mann-Whitney U test was used for variables with a skewed distribution, and a x 2 -test was used for proportions. For repeated measurements of continuous data, the Wilcoxon signed-rank test was used. Pearson correlation coefficient (r) was used to assess linear dependence between normally distributed variables. Linear regression was performed between the concentration of total FAs in maternal and umbilical vein serum and between docosahexaenoic and arachidonic FAs. Data that were not normally distributed were log transformed before analyses. Regression with Spearman correlation coefficient (rs) was performed for non-normality data. Statistical tests were two sided. Statistical analyses were considered significant when P, Results Maternal and neonatal characteristics Women with T1DM and controls did not differ (P. 0.05) in maternal age, BMI, or gestational weight gain (Table 1). Neonates did not differ in gestational age at delivery, neonatal birth weight, or Apgar scores at 1 and 5 minutes. Neonates born to mothers with T1DM were shorter (P = 0.001) and thus, had a higher ponderal index (P, 0.001) compared with control group neonates. The HbA 1c values in the T1DM group dropped from 6.6% to 5.9% from the first to the third trimester (P = 0.022) but were still higher (P, 0.001) than in the control group (4.8%; Table 1). Glucose, C-peptide, leptin, glucose levels, and insulin resistance in maternal and neonatal blood Maternal C-peptide concentration was higher (P = 0.007) in the control group than in the T1DM group (Table 1). Women with T1DM had median C-peptide levels of (interquartile range (94 to 554.8) pm, indicating that their b-cell function and endogenous Table 1. Maternal and Neonatal Characteristics T1DM (n = 30) Control (n = 30) P Maternal characteristics Age, y Prepregnancy BMI, kg/m Gestational weight gain, kg HbA 1C first trimester, % a NA HbA 1C second trimester, % NA HbA 1C third trimester, % a Gestational age at delivery, wk Maternal serum measurements C-Peptide, pm ( ) ( ) Glucose, mm 4.0 ( ) 3.5 ( ) Leptin, ng/ml 25.0 ( ) 19.7 ( ) Neonatal characteristics Birth weight, g Birth length, cm Ponderal index, g, 3 100/(cm) ( ) 2.6 ( ) <0.001 Apgar score at 1 min Apgar score at 5 min Sex ratio: male/female, n (%) b 14/16 (46.7/53.3) 20/10 (66.7/33.3) Umbilical vein blood measurements C-Peptide, pm ( ) ( ) <0.001 Insulin resistance HOMA2 1.5 ( ) 0.7 ( ) Glucose, mm 3.2 ( ) 2.9 ( ) Leptin, ng/ml 15.8 ( ) 6.8 ( ) <0.001 Data are means (SD) unless otherwise indicated; data are median (first and third quartiles); t test was used to test statistical difference between two groups or Mann-Whitney U test. P values in bold are significant. Abbreviation: NA, not available. a Wilcoxon signed-rank test; HbA 1c between first and third trimester in T1DM; P = b The x 2 test was used on group comparison.

4 doi: /jc insulin production were, in part, preserved. Maternal plasma glucose and leptin concentrations were similar in both groups. In the neonates, the umbilical cord vein concentrations of C-peptide (P, 0.001) and leptin (P, 0.001), as well as HOMA2 insulin resistance (P = 0.015), were higher in the T1DM group, whereas plasma glucose concentrations were similar in both groups (Table 1). FAs in maternal serum The total FA concentration in serum of women with T1DM did not differ significantly (P = 0.074) when compared with the control group (Table 2). Although the concentration of saturated FAs (SFAs) in mothers with T1DM was not significantly different from controls (P = 0.071), the percentage of SFAs was significantly lower in T1DM than in the control group (34.8% vs 37.6%, P = 0.006). The concentration of both monounsaturated FAs (MUFAs) and PUFAs was significantly higher in serum of women with T1DM compared with controls (P = and P = 0.027,respectively). However, the percentage of MUFAs and PUFAs was not significantly different between the groups (Table 2). FAs in umbilical vein serum In umbilical vein serum, the concentration of total FAs, SFAs, MUFAs, and PUFAs was higher in T1DM compared with controls (Table 2). The percentage of SFAs in umbilical vein was lower in the T1DM group (34.9% vs 44.3%, P = 0.002). In contrast, the percentage of MUFAs was higher in T1DM (31.9% vs 19.9%, P = 0.002), whereas the percentage of PUFAs did not differ (32.8% vs 31.4%, P = 0.53; Table 2). FAs in umbilical artery serum In umbilical artery serum, the concentration of total FAs, MUFAs, and PUFAs was higher in T1DM when compared with the controls (Table 2). Compared with controls, the concentration of SFAs and the percentage of SFAs were lower in T1DM (201.8 vs 229.9, P ; 34.7% vs 44.0%, P, 0.001). The percentage of MUFAs did not differ between the two groups, whereas the percentage of PUFAs was higher in T1DM than in the control group (31.3% vs 28.8%, P = 0.008). Comparisons of FAs between maternal and umbilical vein serum The concentrations of total FAs, SFAs, MUFAs, and PUFAs were higher in maternal serum of both T1DM and control women when compared the corresponding umbilical vein serum (all P, 0.001; Table 2). In controls, maternal serum was less rich in SFAs compared with umbilical vein serum (37.6% vs 44.3%, P = 0.004), but this difference was not found in T1DM (34.8% vs 34.9%, P = 0.230). A lower percentage of MUFAs was found in maternal serum of T1DM (21.3% vs 31.9%) when compared with umbilical vein serum (P, 0.001), whereas in the control group, the percentage MUFAs was higher in maternal (20.3%) than in umbilical vein serum (19.9%, P, 0.001). A higher percentage of PUFAs was found in maternal serum than in umbilical vein serum in both groups (T1DM: 43.3% vs 32.8%; controls: 41.7% vs 31.4%). Comparisons of FAs between umbilical vein and artery serum In T1DM, the concentrations of total FAs, SFAs, MUFAs, and PUFAs were all significantly higher in umbilical vein compared with the corresponding umbilical artery serum, but the percentages were not different (Table 2). In controls, total FA concentration was higher in umbilical vein serum compared with artery serum. No differences in the concentration of SFAs and MUFAs were found between umbilical vein and artery serum of controls. The PUFA concentration was lower in umbilical vein than in artery serum. The percentage of SFAs was not different between venous and arterial serum in controls, whereas the percentage of MUFAs and PUFAs was lower and higher, respectively, in umbilical vein than in artery serum. FA profile in maternal, umbilical vein, and umbilical artery serum In maternal vein serum, the concentrations of all particular FAs tended to be higher in T1DM, but only the concentrations of oleic (C18:1n-9), DHA (C22:6n-3), and linoleic acid (C18:3n-6) were significantly different from controls (Table 3). In the umbilical vein serum, the concentrations of all particular FAs, except myristic acid (C14:0), were significantly higher in the T1DM group compared with the control group. In the umbilical artery serum, differences between T1DM and controls were smaller, and only concentrations for palmitoleic (C16:1n-7), oleic (C18:1n-9), linoleic acid (C18:3n-6), and AA (C20:4n-6) were higher in the T1DM group compared with the control group. The FA concentrations not higher in T1DM were palmitic acid (C16:0) and stearic acid (C18:0) in umbilical artery serum, and in addition, the percentage DHA was significantly lower in T1DM compared with controls (2.7% vs 3.4%, P = 0.02). Correlation between maternal serum and umbilical vein serum of total FAs and among neonatal ponderal index, glucose, and insulin resistance (HOMA2) Maternal and umbilical vein total FAs were correlated with Pearson correlation coefficient (r)(r = 0.685; P, 0.001;

5 2624 Djelmis et al T1DM Alters Neonatal Fatty Acids J Clin Endocrinol Metab, July 2018, 103(7): Table 2. Concentration of Total, SFA, MUFA, and PUFA and Percent of FAs (micrograms per 100 mg FAs) in Maternal and Umbilical Vein and Artery Blood Maternal Serum FAs T1DM (n = 30) Control (n = 30) P Total FAs, mg/ml ( ) a ( ) b SFAs, mg/ml ( ) e ( ) f Percent SFAs, mg/100 mg FA 34.8 ( ) i 37.6 ( ) j MUFAs, mg/ml ( ) m ( ) n Percent MUFAs, mg/100 mg FA 21.3 ( ) q 20.3 ( ) r PUFAs (n-3 + n-6 FAs), mg/ml ( ) u ( ) v Percent of PUFA (n-3 + n-6 FAs), mg/100 mg FA 43.3 ( ) y 41.7 ( ) z (Continued) Fig. 1). Umbilical artery and umbilical vein total FAs were strongly correlated (rs = 0.883; P, 0.001). Maternal glucose concentrations were strongly correlated with glucose in umbilical vein (rs = 0.763; P, 0.001; (Table 4). Longer-term, higher maternal glucose levels, reflected in higher HbA 1c (percentage) levels, were correlated with neonatal ponderal index (rs = 0.507; P = 0.003). Ponderal index was positively correlated with both leptin (rs = 0.263; P = 0.043) and insulin resistance (HOMA2; rs = 0.526; P = 0.005) in umbilical cord blood. Both C-peptide (rs = 0.379; P = 0.003) and insulin resistance (HOMA2; rs = 0.456; P = 0.02) in umbilical cord blood were correlated to total FAs in umbilical artery serum. Discussion The current study compared FA concentrations and the FA profile in maternal and umbilical venous and arterial serum between well-controlled T1DM and control mothers. Adiposity, which is a confounder of lipid and FAs levels, was comparable in the mothers, as there was no between-group difference in BMI and gestational weight gain. Similar to other studies (17), also, leptin levels did not differ between the two study groups. The most important finding was that well-controlled T1DM affects the FA levels, mostly in the neonates, in which the levels of total FAs, as well as most of the SFAs, MUFAs and PUFAs, were elevated compared with healthy controls. Importantly, also, the cord blood concentrations of the essential FAs, and in particular of DHA, were higher in the T1DM neonates than in the controls. Total FA concentrations were not significantly different between pregnant T1DM with good metabolic control and healthy, nondiabetic women. Large variation in the data may have precluded significance, as the concentrations of some FAs and FA classes were significantly elevated in maternal serum in the women with T1DM compared with control women. Consistent with the present results, also in GDM pregnancies, maternal lipids did not differ from control pregnancies (18). T1DM effects on neonatal anthropometrics and hormones The cord blood levels of C-peptide, leptin, and insulin resistance were also significantly higher in the T1DM group compared with the control group. T1DM, similar to GDM and T2DM, is a known factor that predisposes women to have hypertrophic newborns through an enhanced fetal glucose-insulin axis. This eventually leads to more fetal adipose tissue, as reflected by higher leptin levels (19, 20). Although the placenta also produces leptin, almost all leptin is released into the maternal circulation (21), thus making its contribution to the circulating leptin pool in the fetus small. Insulin and leptin in the cord blood are correlated, especially in neonates in the highest birthweight category (22). Leptin increases with the amount of adipose tissue and regulates fetal weight. Hence, children born to diabetic mothers have higher cord blood leptin levels than those born to nondiabetic mothers. Our leptin results are consistent with previous studies (23). Although we have not directly measured neonatal fat mass or relative body fat, the positive correlation (rs = 0.263; P = 0.043) between leptin and ponderal index may suggest an increase in neonatal fat in the T1DM pregnancies. The higher insulin resistance HOMA2 (P = 0.015) in neonates born to mothers with T1DM compared with control mothers is an important finding. Insulin resistance (HOMA2) correlated with the neonatal ponderal index. The higher insulin resistance in T1DM neonates is likely the result of the higher maternal glucose levels, reflected by higher HbA 1c values in the third trimester than in the controls, but effects of early hyperglycemia cannot be ruled out (24). Interestingly, despite neonatal hyperinsulinemia in the T1DM neonates, their birthweight was not different from control neonates. However, the shorter stature in T1DM neonates, which is another important and unexpected finding, may be associated with reduced lean mass. Thus, T1DM appears to affect the lean and fat compartments

6 doi: /jc Table 2. Concentration of Total, SFA, MUFA, and PUFA and Percent of FAs (micrograms per 100 mg FAs) in Maternal and Umbilical Vein and Artery Blood (Continued) Umbilical Vein Umbilical Artery T1DM (n = 30) Control (n = 30) P T1DM (n = 30) Control (n = 30) P ( ) a,c ( ) b,d < ( ) c ( ) d ( ) e,g ( ) f,h ( ) g ( ) h ( ) i,k 44.3 ( ) j,l ( ) k 44.0 ( ) l < ( ) m,o ( ) n,p ( ) o ( ) p ( ) q,s 19.9 ( ) r,t ( ) s 30.7 ( ) t ( ) u,w ( ) v,x ( ) w 186,1 ( ) x ( ) y,aa 31.4 ( ) z,bb ( ) aa 28.8 ( ) bb Data presented are median and first and third quartiles. The Mann-Whitney U test was used to test significant differences between the two groups. Wilcoxon signed-rank test (see footnotes). P values in bold are significant. Abbreviations: MUFAs, monounsaturated FAs; SFAs, saturated FAs. a Maternal and umbilical vein serum concentration of total FAs in T1DM, P, b Maternal and umbilical vein serum concentration of total FAs in control, P, c Umbilical vein and umbilical artery serum concentration of total FAs in T1DM, P, d Umbilical vein and umbilical artery serum concentration of total FAs in control, P = e Maternal and umbilical vein serum concentration of SFAs in T1DM, P, f Maternal and umbilical vein serum concentration of SFAs in control, P, g Umbilical vein and umbilical artery serum concentration of SFAs in T1DM, P, h Umbilical vein and umbilical artery serum concentration of SFAs in control, P = i Maternal and umbilical vein serum percentage of SFAs in T1DM, P = j Maternal and umbilical vein serum percentage of SFAs in control, P = k Umbilical vein and umbilical artery percentage of SFAs in T1DM, P = l Umbilical vein and umbilical artery percentage of SFAs in control, P = m Maternal and umbilical vein serum concentration of MUFAs in T1DM, P, 0.001). n Maternal and umbilical vein serum concentration of MUFAs in control, P, o Umbilical vein and umbilical artery serum concentration of MUFAs in T1DM, P, p Umbilical vein and umbilical artery serum concentration of MUFAs in control, P = q Maternal and umbilical vein serum percentage of MUFAs in T1DM, P, r Maternal and umbilical vein serum percentage of MUFAs in control, P, s Umbilical vein and umbilical artery percentage of MUFAs in T1DM, P = t Umbilical vein and umbilical artery serum percentage of MUFAs in control, P = u Maternal and umbilical vein serum concentration of PUFAs in T1DM, P, v Maternal and umbilical vein serum concentration of PUFAs in control, P, w Umbilical vein and umbilical artery serum concentration of PUFAs in T1DM, P, x Umbilical vein and umbilical artery serum concentration of PUFAs in control, P, y Maternal and umbilical vein serum percentage of PUFAs in T1DM, P, z Maternal and umbilical vein serum percentage of PUFAs in control, P, aa Umbilical vein and umbilical artery serum percentage of PUFAs in T1DM, P = bb Umbilical vein and umbilical artery serum percentage of PUFAs in control, P = disproportionally, although this must be verified by more direct measurements of body composition. A shorter length of neonates born to mothers of diabetic pregnancies was already found in GDM (25). In that study, the interpretation was offered that fetal FAs, through the transcription factor peroxisome proliferator-activated receptor-g, may drive mesenchymal stem cell differentiation to a more adipogenic than osteoblastic phenotype. Whereas this is speculative, the elevated cord blood levels of FAs, found here in T1DM, may support this notion, which certainly warrants further studies. T1DM effect on neonatal FAs In well-controlled T1DM, maternal FA levels were higher than in controls, although this did not reach statistical significance for each species. This was accompanied by higher concentrations of total FAs, SFAs, MUFAs, and PUFAs in the neonate. Steady-state levels are determined by the following: transplacental

7 2626 Djelmis et al T1DM Alters Neonatal Fatty Acids J Clin Endocrinol Metab, July 2018, 103(7): Table 3. FA Profile in Maternal Vein Serum, Umbilical Vein, and Umbilical Artery Serum transfer, de novo synthesis out of glucose, uptake into tissue, and lipolysis. The study design does not allow for distinguishing among these possibilities. Our study confirmed earlier findings (4, 26) of correlations in total FA concentration between maternal Figure 1. Linear regression between total FAs in maternal and umbilical vein serum (r = 0.685; P, 0.001). Maternal Serum FAs, mg/ml T1DM (n = 30) Control (n = 30) P Myristic acid (C14:0) 41.8 ( ) 8.8 ( ) Pentadecanoic acid (C15:0) 7.4 ( ) 1.9 ( ) Palmitic acid (C16:0) ( ) ( ) Stearic acid (C18:0) ( ) 62.7 ( ) Palmitoleic acid (C16:1n-7) 48.9 ( ) 35.3 ( ) Oleic acid (C18:1n-9) ( ) ( ) Vaccenic acid (C18:1n-7) 6.7 ( ) 0.0 ( a-linoleic acid (C18:3n-3) 10.2 ( ) 4.4 ( ) Eicosapentanoic acid (C20:5n-3) 4.0 ( ) 0.3 ( ) DHA (C22:6n-3) ( ) 70.7 ( ) Linoleic acid (C18:3n-6) 12.9 ( ) 0.8 ( ) AA (C20:4n-6) ( ) ( ) AA:(DHA + EPA) 2.8 ( ) 2.4 ( ) Percent AA, mg/100 mg FA 5.8 ( ) 5.9 ( ) Percent DHA, mg/100 mg FA 2.0 ( ) 1.8 ( ) (Continued) and umbilical vein serum and extends this to mothers with T1DM. Essential FAs are important for normal fetal growth and development (27). Transplacental transport of DHA, eicosapentaenoic acid (EPA), and AA occurs in several steps, from cell transmembrane transport through intracellular transport modulated by protein carriers to further passage through the cell membrane toward the fetus (8, 28). In our previous study, we demonstrated that there was no difference in LC-PUFA n-3 percentage between T1DM and control groups of pregnant women and their fetuses (12), which is in line with the results presented here. Our results differ from an earlier study that found lower percentage of AA and DHA in neonates born to mothers with T1DM (29). Here, the levels of AA and DHA were higher in women with T1DM and significantly higher in their fetuses compared with their control counterparts (Table 3). The major difference between the two studies is that we measured FA percentage in serum, whereas the lower AA and DHA percentage was found in neonatal phospholipids. This would suggest that T1DM modifies the incorporation of AA and DHA into phospholipids. Whether DHA concentrations in cord blood serum or in the phospholipid fraction are more relevant for neonatal development

8 doi: /jc Table 3. FA Profile in Maternal Vein Serum, Umbilical Vein, and Umbilical Artery Serum (Continued) Umbilical Vein Umbilical Artery T1DM (n 5 30) Control (n 5 30) P T1DM (n 5 30) Control (n 5 30) P 8.8 ( ) 2.2 ( ) ( ) 4.2 ( ) ( ) 0.5 ( ) ( ) 1.0 ( ) ( ) ( ) ( ) ( ) ( ) 36.7 ( ) ( ) ( ) ( ) 8.1 ( ) ( ) 28.6 ( ) ( ) 65.9 ( ) < ( ) ( ) < ( ) 0.0 ( ) ( ) 12.8 ( ) ( ) 0.0 ( ) ND 0.0 ( ) 0.0 ( ) ND 0.2 ( ) 0.0 ( ) NA 0.0 ( ) 0.0 ( ) NA 45.9 ( ) 19.9 ( ) ( ) 36.2 ( ) ( ) 0.3 ( ) < ( ) 1.1 ( ) ( ) 59.2 ( ) < ( ) ( ) ( ) 4.0 ( ) ( ) 3.7 ( ) ( ) 12.8 ( ) ( ) 13.5 ( ) ( ) 3.1 ( ) ( ) 3.4 ( ) Data presented are median and first and third quartiles. Mann-Whitney U test was used to test statistical difference between two groups. P values in bold are significant. Abbreviations: EPA, eicosapentaenoic acid; NA, not available; ND, not detected. remains to be established, but it is pertinent that the brain DHA transporter major facilitator superfamily domaincontaining protein 2 transfers DHA in phospholipids across the blood brain barrier (30). The placenta is important for selective AA and DHA canalization from the mother to the fetus; the evidence is a high coefficient of correlation between AA and DHA in the maternal and the umbilical vein blood (Fig. 2). The median maternal percent of AA (5.8%) and DHA (2.0%) in serum of mothers with T1DM was similar to the control group (AA, 5.9%; DHA, 1.8%). AA serves as a precursor of proinflammatory eicosanoids, whereas Table 4. Significant Spearman Correlations Between Umbilical Vein and Umbilical Artery Serum Total FAs and Among Neonatal Ponderal Index, Glucose, C-Peptide, Leptin, HbA 1c,andInsulin Resistance (HOMA2) Correlations rs P Glucose concentration between maternal 0.763,0.001 and umbilical vein Ponderal index and insulin resistance (HOMA2) C-Peptide and total FAs in umbilical vein Leptin and ponderal index Total FAs between umbilical artery and 0.883,0.001 umbilical vein AA and DHA in umbilical vein 0.513,0.001 C-Peptide and total FAs in umbilical artery Insulin resistance (HOMA2) and total FAs in umbilical artery Leptin and C-peptide in umbilical vein serum All P values are significant. DHA is a precursor of anti-inflammatory resolvins and protectins (31); this could suggest a proinflammatory environment in T1DM neonates that is perhaps associated with hyperleptinemia. In a recent randomized controlled trial, we found that DHA supplementation stimulated the production of endogenous insulin in women with T1DM and was accompanied by lower C-peptide levels in the newborn (32). In our study, neonates born by mothers with T1DM had an elevated AA/DHA + EPA ratio (median 4.1:1). This imbalance may be an indication of their increased risk to develop obesity later in life (33). It may well be that a beneficial effect of DHA supplementation may lie in improving the neonatal glucose-insulin axis and perhaps indirectly also, in DHA uptake into the brain across the blood brain barrier. This may also indirectly decrease the risk for childhood obesity. Strengths of the study This study investigates FA profiles and FA concentration in T1DM in arterial and venous cord blood. We have combined the FA measurements with those of key hormones known to influence lipid and FA levels, i.e., insulin and leptin. Cord blood was obtained with the placenta still in situ and the neonate not separated, reducing the influences of the postpartum period and allowing the results to represent the in vivo situation as closely as possible. As all women delivered by elective CS and were denied access to food, the levels of FAs and other parameters measured in umbilical serum were not influenced by mode of delivery or nutritional status of

9 2628 Djelmis et al T1DM Alters Neonatal Fatty Acids J Clin Endocrinol Metab, July 2018, 103(7): Figure 2. Linear regression between DHA and AA in umbilical vein serum (r = 0.924; P, 0.001). the women. The measurement of the arterial and venous cord blood is another strength. There has been only one study in GDM that also determined the FA profile in both arms of the umbilical circulation (34). The combination of these measurements with concentrations of hormonal regulators in the umbilical cord allows conclusions about potential mechanisms that determine the neonatal FA profile, although these must be confirmed in larger studies and other populations. We must acknowledge some limitations: The small sample size precluded significance of higher maternal FA levels in mothers with T1DM. The study included only well-controlled women with T1DM. The results may be different, especially for neonatal outcomes, in poorly controlled T1DM or in women without residual b-cell function, because adaptive and protective placental mechanisms may become exhausted under more extreme metabolic conditions (35). Future studies should measure DHA levels in serum, as well as in the phospholipids of newborn cord blood, and test which of these two fractions is more important for brain development and cognitive function in the children born to T1DM pregnancies. Conclusion In conclusion, the levels of leptin, C-peptide, insulin resistance, total FAs, SFAs, MUFAs, PUFAs, DHA, and AA were significantly higher in T1DM umbilical vein serum compared with those in the control group. Acknowledgments Financial Support: The study was part of the scientific project approved by the Ministry of Science, Education and Technology of the Republic of Croatia, entitled Metabolic and Endocrine Changes in Pregnant Patients with Diabetes (No ). Author Contributions: J.D. designed the study, developed the statistical analysis plan, wrote the manuscript, is the guarantor of this work, and takes responsibility for the integrity of the data and the accuracy of the data analysis. M.I., S.O., and D.S. collected data. E.B. extracted lipids and determined the fattyacidprofilebygaschromatography-mass spectrometry. G.D. and M.v.P. reviewed and edited the manuscript and contributed to the discussion. Correspondence and Reprint Requests: Josip Djelmis, MD, Clinical Department of Obstetrics and Gynecology, Zagreb University Hospital Center, Petrova 13, HR Zagreb, Croatia. josip. djelmis@zg.t-com.hr. Disclosure Summary: The authors have nothing to disclose. References 1. Atkinson MA, Eisenbarth GS. Type 1 diabetes: new perspectives on disease pathogenesis and treatment. Lancet. 2001;358(9277): Herrera E. Metabolic changes in diabetic pregnancy. In: Djelmis J, Desoye G, Ivanisevic M, eds. Diabetology of Pregnancy. Basel: Karger; 2005: Campbell FM, Taffesse S, Gordon MJ. Dutta-Roy AK. Plasma membrane fatty-acid-binding protein in human placenta: identification and characterization. Biochem Biophys Res Commun. 1995; 209(3): Hendrickse W, Stammers JP, Hull D. The transfer of free fatty acids across the human placenta. Br J Obstet Gynaecol. 1985;92(9): Desoye G, Gauster M, Wadsack C. Placental transport in pregnancy pathologies. Am J Clin Nutr. 2011;94(Suppl 6): 1896S 1902S. 6. Herrera E, Desoye G. Maternal and fetal lipid metabolism under normal and gestational diabetic conditions. Horm Mol Biol Clin Investig. 2016;26(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(6):

10 doi: /jc Hanebutt FL, Demmelmair H, Schiessl B, Larqué E, Koletzko B. Long-chain polyunsaturated fatty acid (LC-PUFA) transfer across the placenta. Clin Nutr. 2008;27(5): Neuringer M, Reisbick S, Janowsky J. The role of n-3 fatty acids in visual and cognitive development: current evidence and methods of assessment. J Pediatr. 1994;125(5):S39 S Zornoza-Moreno M, Fuentes-Hernández S, Carrión V, Alcántara- López MV, Madrid JA, López-Soler C, Sánchez-Solís M, Larqué E. Is low docosahexaenoic acid associated with disturbed rhythms and neurodevelopment in offsprings of diabetic mothers? Eur J Clin Nutr. 2014;68(8): Larqué E, Demmelmair H, Berger B, Hasbargen U, Koletzko B. In vivo investigation of the placental transfer of (13)C-labeled fatty acids in humans. J Lipid Res. 2003;44(1): Berberovic E, Ivanisevic M, Juras J, Horvaticek M, Delas I, Djelmis J. Arachidonic and docosahexaenoic acid in the blood of a mother and umbilical vein in diabetic pregnant women. J Matern Fetal Neonatal Med. 2013;26(13): Kolcić I, Polašek O, Pfeifer D, Smolej-Narancić N, Ilijić M,Bljajić D, Biloglav Z, Ivanisević M, Delmis J. Birth weight of healthy newborns in Zagreb area, Croatia. Coll Antropol. 2005;29(1): International Association of Diabetes and Pregnancy Study Groups Consensus Panel, Metzger BE, Gabbe SG, Persson B, Buchanan TA, Catalano PA, Damm P, Dyer AR, Leiva A, Hod M, Kitzmiler JL, Lowe LP, McIntyre HD, Oats JJ, Omori Y, Schmidt MI. International association of diabetes and pregnancy study groups recommendations on the diagnosis and classification of hyperglycemia in pregnancy. Diabetes Care. 2010;33(3): Folch J, Lees M, Sloane Stanley GH. A simple method for the isolation and purification of total lipides from animal tissues. J Biol Chem. 1957;226(1): Levy JC, Matthews DR, Hermans MP. Correct homeostasis model assessment (HOMA) evaluation uses the computer program. Diabetes Care. 1998;21(12): Stock SM, Bremme KA. Elevation of plasma leptin levels during pregnancy in normal and diabetic women. Metabolism. 1998; 47(7): Schaefer-Graf UM, Meitzner K, Ortega-Senovilla H, Graf K, Vetter K, Abou-Dakn M, Herrera E. Differences in the implications of maternal lipids on fetal metabolism and growth between gestational diabetes mellitus and control pregnancies. Diabet Med. 2011;28(9): Nielsen LR, Rehfeld JF, Pedersen-Bjergaard U, Damm P, Mathiesen ER. Pregnancy-induced rise in serum C-peptide concentrations in women with type 1 diabetes. Diabetes Care. 2009;32(6): Maffei M, Volpe L, Di Cianni G, Bertacca A, Ferdeghini M, Murru S, Teti G, Casadidio I, Cecchetti P, Navalesi R, Benzi L. Plasma leptin levels in newborns from normal and diabetic mothers. Horm Metab Res. 1998;30(9): Linnemann K, Malek A, Sager R, Blum WF, Schneider H, Fusch C. Leptin production and release in the dually in vitro perfused human placenta. J Clin Endocrinol Metab. 2000;85(11): Wolf HJ, Ebenbichler CF, Huter O, Bodner J, Lechleitner M, Föger B, Patsch JR, Desoye G. Fetal leptin and insulin levels only correlate inlarge-for-gestational age infants. Eur J Endocrinol. 2000;142(6): Persson B, Westgren M, Celsi G, Nord E, Ortqvist E. Leptin concentrations in cord blood in normal newborn infants and offspring of diabetic mothers. Horm Metab Res. 1999;31(8): Desoye G, Nolan CJ. The fetal glucose steal: an underappreciated phenomenon in diabetic pregnancy. Diabetologia. 2016;59(6): Lampl M, Jeanty P. Exposure to maternal diabetes is associated with altered fetal growth patterns: A hypothesis regarding metabolic allocation to growth under hyperglycemic-hypoxemic conditions. Am J Hum Biol. 2004;16(3): Wolfe MD, Chuang LT, Rayburn WF, Wen PC, VanderJagt DJ, Glew RH. Low fatty acid concentrations in neonatal cord serum correlate with maternal serum. J Matern Fetal Neonatal Med. 2012;25(8): Innis SM. Essential fatty acids in growth and development. Prog Lipid Res. 1991;30(1): Campbell FM, Dutta-Roy AK. Plasma membrane fatty acidbinding protein (FABPpm) is exclusively located in the maternal facing membranes of the human placenta. FEBS Lett. 1995;375(3): Ghebremeskel K, Thomas B, Lowy C, Min Y, Crawford MA. Type 1 diabetes compromises plasma arachidonic and docosahexaenoic acids in newborn babies. Lipids. 2004;39(4): Nguyen LN, Ma D, Shui G, Wong P, Cazenave-Gassiot A, Zhang X, Wenk MR, Goh EL, Silver DL. Mfsd2a is a transporter for the essential omega-3 fatty acid docosahexaenoic acid. Nature. 2014; 509(7501): Serhan CN, Chiang N, Dalli J, Levy BD. Lipid mediators in the resolution of inflammation. Cold Spring Harb Perspect Biol. 2014; 7(2):a Horvaticek M, Djelmis J, Ivanisevic M, Oreskovic S, Herman M. Effect of eicosapentaenoic acid and docosahexaenoic acid supplementation on C-peptide preservation in pregnant women with type-1 diabetes: randomized placebo controlled clinical trial. Eur J Clin Nutr. 2017;71(8): Simopoulos AP. An increase in the omega-6/omega-3 fatty acid ratio increases the risk for obesity. Nutrients. 2016;8(3): Ortega-Senovilla H, Alvino G, Taricco E, Cetin I, Herrera E. Gestational diabetes mellitus upsets the proportion of fatty acids in umbilical arterial but not venous plasma. Diabetes Care. 2009; 32(1): Gauster M, Hiden U, van Poppel M, Frank S, Wadsack C, Hauguel-de Mouzon S, Desoye G. Dysregulation of placental endothelial lipase in obese women with gestational diabetes mellitus. Diabetes. 2011;60(10):