Discussion
In the present study, among the South Chinese pregnant cohort, 19.2% were pre-pregnant underweight women according to the WHO criteria for Asian populations, while 19.4% were overweight/obese women. We found that the proportion of underweight women was almost equivalent to the proportion of overweight/obese women. This finding showed that the phenomenon of underweight was actually common in South China. As was expected, a lower BMI cut-off at 25 kg/m2 for defining obesity would be better than BMI cut-off at 28kg/m2 for defining obesity for pregnant women in South China, which had been demonstrated in our previous study12.
We found underweight women had significant higher TC level than normal-weight women and overweight/obese women (Table 3 or Table 4). In line with Kulkarni et al13 , a research in Pune showed that TC concentrations went higher in rural underweight women at 18 and 28 gestational weeks. Kulkarni,et al13 demonstrated in rural, undernourished, normoglycemic Indian pregnant women a significant association between maternal circulating lipids and fetal growth, which was at least as strong as that of glucose. During gestation, maternal lipid alters form anabolic to a catabolic state during, which causes maternal physiological hyperlipidemia(MPH) 14. This phenomenon could be ascribed to an adaptive response to satisfy the increasing fetal demand15, 16. Cholesterol is essential for fetal growth, steroid synthesis and neurodevelopment, at least 1g cholesterol is essential for placenta to synthesis placenta17. Fetus is unable to synthesize cholesterol in the early pregnancy, since their liver and adrenal gland is immature, most of them utilize endogenously cholesterol at term. Facing such greatly demanded, increasing maternal serum TC seems like significant to satisfy fetus rapid growth. Moreover, fetus need to uptake maternal cholesterol to synthesis cell membrane 18 to maintain neurodevelopment. Abnormal cholesterol metabolism appears to related to impair neurological development19 and low birthweight17 or IUGR 20. Insufficiency of LDL-C and TG may associate with FGR21. Besides, a thin-fat phenotype was reported in South Asia22, which was caused by undernourished state, thin-fat women has low body mass with a high concentration of TC. This phenomenon probably indicate that cholesterol is more necessary for underweight pregnant women.
According to the study by Butte NF15, Cholesterol is used by the placenta for steroid synthesis to meet maternal requirement. Estrogen production is 3-8 times higher at term than non-pregnancy, cholesterol as an essential material of steroid is in greatly demanded. Estrogen also strongly stimulates LDL receptor expression and improve uptaking of cholesterol, which results in maternal hypercholesterolemia23, 24. Thus, we presume that lipocatabolic enhance in underweight women than normal weight women and overweight/obese women to meet maternal and fetal demand.
Our results showed that underweight women with high TC level had significantly higher occurrence of LGA, and lower occurrence of SGA, while there was no significant association between serum TC and LGA or SGA in normal-weight women and overweight/obese women. Meanwhile, the increase of LDL-c and TG/HDL-c ratio decreased risk for SGA in underweight women, but no significant relationship was found in overweight/obese women. In line with Adu‐Afarwuah25, TC level, and a lipid‐based nutritional supplement intake could result in increase of birth weight. A research in Pune showed that TC concentrations went higher in rural underweight women at 18 and 28 gestational weeks and positively related to the newborn birth size. Serum cholesterol concentration was directly associated with all newborn measurements except head circumference. A 1-SD-higher maternal TC concentration was associated with a 54-g-higher birth weight. According to the study by Krstevska B26, LGA also attributed to maternal serum LDL-c. High concentrations of HDL-c seemed like a protector for LGA27, 28. However, this conclusion is still controversial. In the contrary, Eslamian L29 and Wang J28 did not find significant association of maternal LDL-C levels and LGA newborn. Our results revealed an association between high TC level and LGA in underweight women, but the potential mechanisms were unknown. We presume that cholesterol plays a more significant and complex role for underweight women in gestation, because it is essential material to maintain maternal pregnancy and fetal development. Moreover, we should pay more attention on this ‘thin-fat’ phenotype, since the proportion of underweight women and overweight/obese women is equivalent. Unfortunately, we had paid much more attention on the diet and weight control of overweight/obese women, while ignored the management of underweight women on their serum lipid profile. Furthermore, underweight women may tend to take more food during gestation, especially in China, and this may aggravate the hypercholesteremia of underweight women, which may lead to adverse outcomes.
In addition, our results are in line with other researchers who reported associations between serum TG levels and adverse maternal and fetal outcomes. Maternal TG was an independent predictor for GDM30 and birth weight31-33. Meanwhile, the prevalence of GDM, PIH, PTL and macrosomia were higher in pregnant women with high TG level34. Thus, the results above remind us the necessity to control serum lipid level.
The main strengthen of our study was that we used a more proper criteria to stratify the underweight women in South of China, and investigate the lipid level and corresponding maternal and fetus outcome. This makes our results more reliable. Additionally, we focus the relationship of underweight women and cholesterol level, which has not been well investigated before. Few researchers explored and interpreted the relationship among underweight women, high level cholesterol and newborn size. At last our study has a huge sample size about 6,000 which seemed like to have compelling results. Thus, this study can contribute new insights into controlling of underweight women serum lipid counseling.
There were some limitations in our study. The height and weight of gravidas were recalled and self-reported by patients, and so it was likely there were more error variation and potential bias than measured. The second limitation was that diet before the blood drawing may affect the serum lipid profile. The amount and composition of dietary fat seem to be associated with serum concentrations of lipids35, 36. In consideration of this, we had given them proper diet guidance before their blood tests.