Long-chain polyunsaturated fatty acid (LC-PUFA) status in severe preeclampsia and preterm birth: a cross sectional study

Long-Chain Polyunsaturated Fatty Acid (LCPUFA) is essential throughout pregnancy, since deficiency of LPUFA may linked to obstetrical complications. This study aimed to investigate LCPUFA status in severe preeclampsia and preterm birth. A cross sectional study was conducted in 104 pregnant women, which divided into normal pregnancy, severe preeclampsia and preterm birth groups. Serum percentage and concentration of total LCPUFA, omega-3, alpha-linolenic acid (ALA), eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA), omega-6, linoleic acid (LA), and arachidonic acid (AA) were measured using gas chromatography/mass spectrometry. Receiver operating characteristic (ROC), bivariate and multivariate analysis were performed. Severe preeclampsia showed the highest concentration of total PUFA and the lowest DHA percentage, with significantly higher Omega-6/Omega-3 ratio (p = 0.004) and lower omega-3 index (p < 0.002) compared to control. Preterm birth showed the least omega-3 concentrations, with significantly low omega-6 derivates (LA (p = 0.014) and AA (p = 0.025)) compared to control. LCPUFA parameters have shown to increase the risk in both conditions, particularly ALA ≤ 53 µmol/L in preeclampsia with OR 5.44, 95%CI 1.16–25.42 and preterm birth with OR 4.68, 95%CI 1.52–14.38. These findings suggest that severe preeclampsia and preterm birth have an imbalance in LCPUFA status.

Sample preparation and fatty acids quantification. Non-fasting maternal blood samples were obtained soon after delivery, and excluded if the delivery had been more than 1 h. Samples were collected using venous puncture into 5-mL tubes (Vacutainer; Becton-Dickinson). The serum was directly separated out from the whole blood, then frozen at − 80 °C until further analysis.
For fatty acids measurement, serum was firstly thawed at room temperature for about 30 min, then extracted using (choloform:methanol, 2:1). 150 uL of Methanol (Merck, Germany) was added to 100 uL serum for protein precipitation and homogenized thoroughly and then added with 300 uL Chloroform (Merck, Germany) contained Nonadecanoic Acid Methyl Ester (Supelco) as internal standard and vortexed. The mixture then centrifuged at 2500 g in 5 min and the chloroform layer which contain fatty acids was transferred to different tubes. The chloroform phase was then dried with gentle nitrogen stream and then reconstituted with n-hexane (Merck, Germany). Tetramethyl Ammonium Hydroxide (TMAH) in methanol (Sigma) was added and incubated 2 h in room temperature for trans-esterification while vortexed, upper phase then transferred to Gas Chromatography (GC) vials [18][19][20][21][22] .
Data were analysed using Statistical Package for Social Sciences (SPSS) version 25.0 (IBM, United States). The numeric data were firstly checked for normal distribution using Kolmogorov Smirnov test, then presented as mean ± SD if normally distributed and as median (IQR) if not normally distributed. One-way ANOVA and Kruskal Wallis test were used to identify significances among groups. Post hoc tests were performed to locate those specific differences in each group, using Bonferroni and Games-Howell test after One-way ANOVA and Mann Whitney test after Kruskal Wallis.
The Receiver Operating Characteristic (ROC) curve analyses were used to analyse corresponding cut-off values, sensitivity and specificity of each LCPUFA concentration variables. The overall performance of each parameter for predicting either preeclampsia or preterm birth were assessed by estimating the area under the curve (AUC). The cut off values for parameters with high overall performance were determined at the point on the curve with the highest value of sensitivity and specificity. Bivariate analysis was performed using Chi-Square test, followed by multivariate association between each LCPUFA serum variables and each pregnancy outcomes in separate logistic regression models after adjusting maternal age, gestational age (only available for preeclampsia), and pre-pregnancy BMI. All results corresponding to p-values < 0.05 (5%) were described as significant and reported.

Results
Characteristics of subjects. All of the samples were included in this study. The characteristics of subjects are shown in Table 1. There was significant variation found in maternal age (p < 0.05), systolic blood pressure (p < 0.05), diastolic blood pressure (p < 0.05), gestational age (p < 0.05), mode of delivery (p < 0.05), BMI (p < 0.05), and birth weight (p < 0.05). Patients with preeclampsia had highest maternal age 32.37 ± 6.41 years old, and highest BMI 26.80 ± 5.76. In contrast, preterm birth had the smallest BMI 21.78 ± 4.43 and lightest birth weight with 1937.5 g (1360.0-2370.0). LCPUFA levels in normal pregnancy, preeclampsia and preterm birth. Details of serum PUFA percentage and concentrations in control, preeclampsia and preterm birth are shown in Table 2. Preeclampsia Table 1. Characteristics of subjects (n = 104). Data presented in Mean ± SD or Median (IQR). *Significances to control (p < 0.05). **Significances to preeclampsia (p < 0.05).  ROC curve analysis. The AUC of ROC curves ranged 0.65-0.93 was taken into account, and all LCPUFA parameters had AUC < 0.65 were not included in further analysis as they are considered poor indicators for severe preeclampsia ( Fig. 1 and Table 3) and preterm birth ( Fig. 2 and Table 4). In predicting the risk of preeclampsia, EPA and AA had the highest AUC with 0.872 and 0.927. The best cut off points used with the sensitivity ranged 60.0-83.3% and specificity ranged 54.5-86.4%. The highest overall performance showed in ≤ 16 µmol/L EPA in having sensitivity 80% and specificity 86.4% in predicting the risk of preeclampsia, followed by AA/EPA with ratio ≥ 35, which had sensitivity 83.3% and specificity 84.1%. In regard of preterm birth, the lowest AUC value which included for further analysis was 0.654 for AA and the highest was 0.709 for ALA. For sensitivity and specificity analysis in preterm birth risk, the best cut off points used with the sensitivity ranged 53.3-66.7% and specificity ranged 63.6-77.3%.

Discussion
This research was dominated by overweight to obese participants. This findings was in conjunction with previous study in Indonesia, where 49.1% of 234 first trimester pregnant women were also found to have BMI ≥ 23 12 . We also found that preterm birth group had the lightest mean of BMI. This data was supported by reviews and meta-analysis, which suggested that low maternal BMI is associated with preterm birth 23 .
In comparing LCPUFA concentrations of those 3 groups, preterm birth group had the lowest total LCPUFA, total omega-3, and omega-6. Even though studies have shown that there is an increasing trend of LCPUFA level in serum, erythrocyte and plasma throughout pregnancy [24][25][26] , studies have also suggested that low total omega-3 had increased the risk of preterm birth < 34 weeks, which can be reduced through LCPUFA supplementation 27,28 . In addition, gestational age seems not affecting total LCPUFA content in preeclampsia, as being the highest level among others. This was thought to influenced by numerous amounts of omega-6 serum concentration.
Preeclampsia and preterm birth group found to have lower amount of omega-3 PUFA than control group, with preterm and control group showed a significant difference (p < 0.05). This is consistent with other findings, where low number of both omega-3 and omega-6 PUFA in preeclampsia and preterm were found 15,29 . Omega-3   30 . Studies have shown that supplementation of omega-3 fatty acids during pregnancy may improve birth weight as well as decrease preterm birth rate 27,31 . Therefore, a good counselling to encourage pregnant women in improving omega-3 dietary intake and supplementation are required. ALA and EPA are essential omega-3 fatty acids which have anti-inflammatory properties and vascular benefit. In our study, we found that both preeclampsia and preterm birth have lower ALA and EPA serum concentration compared to control. These were shown by EPA concentrations ≤ 16 µmol/L was significantly associated with preeclampsia (p < 0.001), followed by ALA maternal serum concentration ≤ 53 µmol/L have shown to increase the risk of both pregnancy complications (preeclampsia by OR 5.44, 95%CI 1. 16-25.42 and preterm birth by OR 4.68, 95%CI 1.52-14.38). These findings were consistent with other studies where low ALA and EPA were associated with higher preeclampsia risk 29,32 . Moreover, a study by Stewart et al. showed lower ALA was found in patients with earlier gestational age as ALA is expected to increase along pregnancy and become the precursor for omega-3 metabolites, which are EPA and DHA 26 . Even though the conversion process is not efficient in human and depend on epigenetic, genetic process, as well as sex 33,34 , another study suggest that the level of maternal ALA is associated with DHA concentration in blood cord 35 . A nutrition study about dairy product with ALA, LA, and DHA found to reduce the rate of early preterm birth and improve the length of gestation 33 . Thus, pregnant women are advised take sufficient amount of ALA and EPA on early terms, using supplementations or omega 3 rich food such as soybean oil, canola oil, green vegetables, and seafood 36,37 .
Furthermore, DHA percentage was found to be significantly low for preeclampsia compared to control (p > 0.05) and preterm birth (p < 0.05). A study have shown that DHA may decrease oxidative stress in deep placentation disorders, including preeclampsia and preterm birth 30 . Another study also showed DHA supplementation during pregnancy significantly reduce the risk of preeclampsia and severe preeclampsia 38 . In addition, preterm subjects showed the least of DHA concentrations result among groups and EPA result was lower compared to control group. Studies have shown that DHA promotes longer gestation period by acting as prostaglandin showed supplementation of 600 mg DHA daily given at early pregnancy resulted in significant longer pregnancy (p = 0.041) and higher birthweight (p = 0.004) 40 . Other study supported this result by showing a decreased preterm birth rates in patients taking Omega-3 PUFA levels up to ~ 600 mg daily 41 . Omega-3 index, which is a combination of EPA and DHA level, is suspected to represent functional PUFA status. Our data also showed that preterm birth had 2.5% omega-3 serum index. This result was close to findings by Olsen et al., which proposed that the risk of preterm labor increase if the value of plasma PUFA below 2.0%, with two measurements of samples in the early pregnancy 15 . Our results also showed that cut off point 2.5% was suitable for predicting the risk of preeclampsia with sensitivity 83% and specificity 54.5%, following the increased risk by OR 12.19, 95%CI 1.65-89.64. Therefore, high omega-3 diet as well as supplementation such as DHA, should be considered for pregnant women.
In addition, we found significantly higher level of total omega-6 PUFA (p < 0.05), LA (p < 0.05), and AA (p < 0.05) in preeclampsia compared to others. These findings were accompanied with lower total omega-3 level along with its derivates (ALA, EPA and DHA) in preeclampsia compared to control group. The result also showed significantly higher ratio (p < 0.05) of omega-6/omega-3 in preeclampsia group compared to control group. This study showed an altered condition of fatty acid level that occur in preeclampsia patient as shown in previous studies before 42,43 .
Our data did not show causality LCPUFA status on preeclampsia, however our result on high total omega-6 level and concordance to high ratio of omega-6/omega-3 may play role in inflammatory reaction in preeclampsia pathogenesis, thus further study is required 6,44 . Higher omega-6 with low omega-3 has been previously associated with promotion of platelet aggregation, vasoconstriction and proinflammatory along with prothrombotic state which contribute to preeclampsia pathophysiology 45,46 . This is in conjunction with our results where both of the pathological conditions had significantly higher value of omega-6/omega-3 ratio compared to control (p < 0.05). Meanwhile, preterm birth had the lowest mean of omega-6/omega-3 ratio. This might be resulted by relatively low total of omega-3 with relatively normal omega-6 level and lower in its derivates such as AA in preterm group.
Our study found that there was a significantly high level of AA/EPA ratio in preeclampsia compared to other groups. Previous studies have suggested that AA/EPA have positive correlation with inflammatory rate of eicosanoids 47,48 , particularly due to AA eicosanoid metabolites 30,49 . On the other hand, AA is also found to generate lipoxins, particularly lipoxin (LXA4), which acts as an anti-inflammatory properties that related to obstetrical complications such as preterm and preeclampsia 50,51 . Here we showed that the value of ≥ 35 AA/EPA has increased the risk for preeclampsia with sensitivity 83.3% and 84.1%, with significant differences compared to control (p < 0.001) and high adjusted odds ratio (OR 24.78). However, the wide 95%CI (4. 25-144.27) found in our study suggesting that this optimal value of AA/EPA should be further elaborated, followed by other metabolites analysis.
Certain limitations need to be considered to interpret our study. We did not asses daily intake of PUFA, as well as other biomarkers which responsible in pathological mechanism related to obstetrical outcomes. We also unable to analyse non-fasting blood samples as preeclampsia and preterm birth maternal condition which required highly nutrient intake. Suggesting that mothers who are in labor, especially with vaginal delivery method, should be energized enough to undergo the process. In addition, triglyceride elimination before trans-esterification was unable to performed. Thus, the calculated fatty acids were both free fatty acids and fatty acids which might still bind to triglycerides and phospholipid. Nevertheless, this initial research throughout this topic could represent an overview of maternal LCPUFA status in pregnant women with preeclampsia and preterm birth, as well as a beginning for further research. Detail evaluation of nutritional intake, LCPUFA supplementation, other metabolites as well as sample processing with wider range of population are required in order to gain better understanding regarding LCPUFA effect on obstetrical outcome.

Conclusion
Our study proposed that both severe preeclampsia and preterm birth have an imbalance in fatty acid metabolism. LCPUFA plays a great role in those pathological conditions, thus dietary balanced intake of omega-3 and omega-6 enriched functional food or supplementation in pregnant women should be highly considered.

Data availability
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.