Association between Maternal and Foetal Erythrocyte Fatty Acid Profiles and Birth Weight

Regular foetal development is crucial for assuring good health status in the offspring. The quality and quantity of maternal dietary fatty acids (FAs) can affect growth. The study aimed to: (1) investigate the association of maternal/foetal lipid profiles with birth weight (BW); and (2) compare these profiles in small, appropriate, and large for gestational age (SGA, AGA, and LGA) infants. FAs were measured in erythrocyte membranes using gas chromatography analysis in 607 mother–infant pairs (316 males, 52.1%). In the quantile regression, a significant association between BW and levels of maternal linoleic acid (LA; C18:2, n-6; coefficient: 18.66; p = 0.010), arachidonic acid (AA; C20:4, n-6; coefficient: 11.35; p = 0.007), docosahexaenoic acid (DHA; C22:6, n-3; coefficient: 29.73; p = 0.007), polyunsaturated FAs (coefficient: 8.55; p = 0.001), foetal DHA (coefficient: −22.82; p = 0.037), and saturated FAs (coefficient: −65.41; p = 0.002) was found. Myristic (C14:0) and pentadecanoic acids (C15:0), both maternal (p = 0.000; p = 0.017) and foetal (p = 0.009; p = 0.002), and maternal erucic acid (C22:1, n-9; p = 0.026) were found at higher levels in SGA infants as compared to AGA ones. Conversely, maternal LA, AA, and omega 6 FAs levels were higher in AGA infants (p = 0.037; p = 0.003; p = 0.026, respectively). Maternal and foetal polyunsaturated and omega 6 FAs levels are positively related to BW, while a lipid profile rich in saturated FAs and erucic acid may influence the risk of SGA.


Introduction
Regular foetal development during pregnancy is pivotal for assuring good short and long-term health status of the offspring. Both higher and lower intrauterine growth are not only related to The "Feeding" study was approved by the Ethical Committees of the "Ospedale Pediatrico Bambino Gesù" (OPBG) and the San Camillo Forlanini Hospital (SCH), in full agreement with the national and international regulations and the Declaration of Helsinki (2000). All the participants signed an informed consent form.

Subjects
From the initial cohort of 1000 pregnant women enrolled, 153 (15.3%) mothers withdrew from the study (6 genetic diagnoses, 24 childbirth complications not allowing blood collection, 32 personal reasons, 8 miscarriages, and 83 deliveries in different hospitals). No differences were found in age, anthropometrics, and SES of women who participated or withdrew from the study (data not reported). A complete data set of FAs profiles was available for 694 mother-infant pairs out of 847 (81.9%). For the current analysis, we excluded 41 (5.9%) women with history of diabetes (including gestational diabetes) and 46 (6.9%) mother-infant pairs because of missing data.
A final sample of 607 mother-infant pairs (316 males, 52.1%) was available for the analysis. Most of the women were of normal weight before pregnancy (427, 70.3%) while GWG was inadequate, adequate, or excessive in 27.4%, 40.0%, and 32.6% of women, respectively. All the infants were born at term and 483 (80.5%) of them were classified as AGA, while 49 (8.2%) and 686 (11.3%) were SGA and LGA, respectively. Table 1 shows maternal and foetal characteristics of the sample. Data are expressed as median and interquartile range (IQR) 1 or as number and percentage (%) 2 ; BL, birth crown-heel length; BW, birth weight; GWG, gestational weight gain; HC, head circumference; LGA, large for gestational age; SDS, standard deviation score; SGA, small for gestational age. Anthropometric measures were taken at study enrolment for mothers, and at birth for infants.

Maternal and
No differences was found in the correlation between maternal and foetal fatty acids in the three group, except for erucic acid (r = 0.06, p = 0.702) and total SFAs (r = 0.06, p = 0.665), for which no correlation was found in the SGA group; and trans 16:1 n-7 for which a correlation was found only in the AGA group (r = 0.26, p = 0.000).

Discussion
Our results indicate the following: (1) maternal LC-PUFAs levels are positively associated with birth weight; (2) foetal SFAs are negatively associated with birth weight; (3) both maternal and foetal SFAs levels are higher in the SGA group as compared to the AGA and LGA groups; and (4) maternal erucic acid levels are higher in the SGA group as compared to AGA and LGA groups. Therefore, a maternal lipid profile rich in saturated fatty acid and erucic acid seems to be able to predict giving birth to a SGA infant.
It is difficult to compare the results of the present investigation with previous ones. Indeed, three studies investigated plasma levels of FAs [7,8,11,14]. The four studies performed on erythrocyte membranes [9,10,12,13] basically found overlapping levels of FAs, even though two of them [10,13], compared NBW and LBW infants instead of the SGA and AGA classes, one study focused on few FAs [9], and the remaining study provided comparable data only for cord blood [13].
We found maternal LA, AA, DHA, and total LC-PUFA levels to be positively associated with offspring weight at delivery. Both maternal and foetal LC-PUFA levels were found to be higher in the AGA group when compared to the SGA group. These findings confirm what was previously found on LC-PUFAs with respect to their indispensable role in foetus and infant growth and development [9,10,12,13]. Conversely, DHA was inversely associated with birth weight and, exclusively in the SGA group, levels were found to be higher in infants when compared to mothers. We hypothesise a preferential transfer of this polyunsaturated fatty acid through the placenta due to the increased needs of the foetus. The possible mechanism could be through an up-regulation of mRNA expression in placental fatty acid transporters as a compensatory mechanism in SGA foetuses [20].
A previous study by Bobinski et al., compared the fatty acid profiles in maternal and cord blood in AGA (n = 54) and SGA (n = 239) infants born at term. No difference was found in maternal lipid profile. Conversely, they found foetal lauric acid (12:0) levels to be higher in SGA infants. They hypothesized the increase level of lauric acid to be a response to increased energy requirements of infants belonging to the SGA group [14]. Recently, Meher et al. investigated the association between maternal fatty acid profile across gestation and cord blood lipid profiles in 46 LBW and 52 NBW infants. They found higher levels of maternal erythrocyte SFAs in women delivering LBW babies and attributed this to the inadequate transfer of these fatty acids through the placenta, contributing to inadequate foetal growth [13]. Our results suggest a negative association between both maternal and foetal SFAs and birth weight. Maternal myristic (C14:0) and pentadecanoic (C15:0) acids, and foetal C14:0, C15:0, and eptadecanoic acid (C17:0) were all detected to be higher in SGA infants when compared to AGA ones. When we examined the SGA group alone we found no statistical difference between foetal and maternal SFAs, in contrast to the AGA and LGA group where maternal SFA levels appeared to be higher. The correlation was positive. Hence, we do not assume decreased placental transfer of these FAs during pregnancy, or their role as energy supply, but, conversely, we hypothesise their over-representation in mothers to reflect the higher level in the foetus and their possible adverse effects on development [21].
In respect to trans fatty acids, we found trans elaidic acid (trans 18:1 n-9) levels to be higher in SGA infants when compared to the AGA ones. Trans FAs are described to be inversely associated to LC-PUFAs in pregnant women and their new-borns and may interfere with metabolism and trans-placental transfer [22]. Previous studies on plasma concentration suggested possible important effects of trans FAs on foetal growth [7].
Finally, even if no association was found between erucic acid and birth weight, we found maternal erucic acid levels to be higher in the SGA group when compared to the AGA and LGA ones. No difference was found for foetal erucic acid. When we examined the SGA group alone, erucic acid levels resulted higher in the mother, and no correlation between the mother and foetus was found.
While erythrocyte FAs may better reflect FAs intake than dietary recalls [23], in mothers they do not necessary reflect diet since SFAs can be synthesized, while PUFAs can be elongated. These mechanisms may be variable between mothers for numerous reasons, including genetic variations. In any case, our previous work showed a strong correlation between maternal and foetal lipid profile, so we assumed maternal erythrocytes FAs to reflect infants ones, and both to be a possible factor influencing infant size at birth [15].
The mechanisms linking maternal and foetal overexpression of SFAs and the delivery of SGA infants are not well established. Evidence from previous studies on animal models (mice, dams, rabbits, and swine) showed opposing results. Some investigations support early effects of unbalanced high fat (HF) diets on offspring development, causing impaired intrauterine growth and low birth-weight offspring [24][25][26]; conversely, others found foetal overgrowth [27] or no difference with the control group, in response to a maternal HF diet [28].
In general, FAs interact with the human placenta and initiate several cascade events, differing with respect to their carbon length and degree of saturation [29,30]. They can influence foetal growth through different mechanisms: (1) by altering their own specific transfer from the mother to the foetus [31]; (2) by regulating trophoblast amino acid transport through the modulation of the mammalian target of rapamycin (mTOR) and insulin-like growth factor (IGF) pathways [29,32]; and (3) by initiating innate immune responses via the toll/like receptor-4 [30]. These mechanisms has not been extensively explained in SFAs and MUFAs, for which research mainly focused on palmitic, stearic, and oleic acids [29,30].
In particular, it has been hypothesised that the high long chain saturated fat diet, known to cause insulin resistance, could be an inciting factor in the decreased expression of the embryonic IGF-1 receptor which manifests later through differences in offspring size, growth patterns, and metabolic response [24]. Alternatively, this diet may affect Igf1r (IGF-I receptor gene) expression in the blastocysts, resulting in subsequent insulin resistance [24]. Interestingly, Yang et al. showed how the saturated FAs palmitic acid and stearic acid play a dynamic role in the placenta inflammation status [30]. Moreover, Lager et al. found DHA and oleic acid to be associated to a decrease and increase of amino acids transfers, respectively. Palmitic acid (C16:0) was not found to affect trophoblast amino acid transport [29].
Our results on SFAs suggest a possible enhanced transfer of these fatty acids from mothers to the foetus in SGA infants, perhaps in relation their increased representation in the former, likely contributing to an altered placental metabolism as well as an impaired placental nutrient transport capacity, with consequent reduced foetal growth. Our findings mainly refer to myristic, pentadecanoic, heptadecanoic, and erucic acids, whose roles have not been investigated yet.
The present study has strengths as well as limitations. The main strengths include the measurement of erythrocyte FAs, the comparison between the three classes of infants (SGA, AGA, and LGA groups) and the large sample size. However, we were not able to measure other factors or residual confounds, such as genetic variability, which may have affected erythrocyte lipid profiles; we were also unable to rule out whether the influence of maternal fatty acid profiles on the new-borns' body size reflects dietary differences. In our series, the use of FAs concentrations in the analyses instead of percentages did not affect results as far as FAs classes are concerned. Therefore, we used percentages in order to allow comparisons to previous studies. Nevertheless, this may be considered a limitation of the study.

Conclusions
As far as we know, the "Feeding Low-Grade Inflammation and Insulin Resistance of the Foetus" study is one of the largest cohort studies investigating the association between maternal/foetal erythrocyte fatty acid profiles and birth weight. It is worth mentioning that this was also the first study comparing maternal and foetal erythrocyte FAs in the three classes of AGA, SGA, and LGA infants.
Our study results has shown that both maternal and foetal FAs profiles may affect foetal growth during pregnancy, likely by modulating placenta metabolism. Excess of SFAs, trans, and erucic acid may play a role to shape the risk of low intrauterine growth.