Homocysteine during the third trimester is a risk factor for preeclampsia: A prospective study

Background: The purpose of this study was to investigate the risk factors for elevating homocysteine during pregnancy and the relative effects on preeclampsia, so as to further understand whether Hcy had predictive value for PE. Method: This is a prospective study that only covers pregnant women with singleton who received regular prenatal care from July to September 2018 exclusively at IPMCH (N=1142). Homocysteine, folic acid and vitamin B12 were tested in the 1 st trimester (10-14 weeks), 2 nd trimester (24-28 weeks), and 3 rd trimester (30-34 weeks), respectively, and MTHFR genes (rs1801133, rs1801131, rs17367504) were detected. Therefore, the analysis of this case includes the variation in Hcy levels during pregnancy, risk factors for elevating homocysteine and the risk factors on preeclampsia. Results: (1) Homocysteine was lowest in the 1 st trimester. (2) Homocysteine was negatively correlated with folic acid (r=-0.17, p<0.001) and vitamin B12 (r=-0.15, p<0.001) in the same trimester. (3) Both of heterozygous CT (p=0.025, 95% CI 0.018, 0.275) and homozygous TT (p<0.001, 95% CI 0.185, 0.501) in MTHFR rs1801133 might be risk factors that caused an increase in Hcy. G-spot mutations in MTHFR rs17367504 might be a risk factor that caused a decline in homocysteine. (4) Homocysteine in the 3 rd trimester might be signicantly correlated with increasing risk of preeclampsia (OR = 1.2, 95% CI 1.01,1.42), particularly early-onset preeclampsia (OR = 3.63, 95% CI 1.71,7.71) and severe preeclampsia (OR = 3.63, 95% CI 1.71,7.71). Conclusions: The variation in homocysteine level in trimester (30-34 weeks) regular blood during The follow-up was completed until 6 weeks postpartum or withdraw. General data, blood biochemical results and pregnancy outcomes of the subjects were collected. The diagnostic criteria for PE were based on American College of Obstetricians and Gynecologists (ACOG) for gestational hypertension and PE.[4] All participants provided written informed consent.


Background
Preeclampsia (PE) is a pregnancy-speci c disease that occurs after 20 weeks of gestation and manifests as hypertension and proteinuria with or without visceral damage. The incidence of PE is 3-5% [1], which reasonably accounts for the increased maternal and prenatal morbidity and mortality.
Previous studies have suggested that PE may be associated with elevated homocysteine (Hcy) during pregnancy. Hcy is a sulfur-containing amino acid intermediate in the metabolism of methionine and cysteine. Cotter [2] has found that Hcy is signi cantly higher in PE than in normal pregnancies, especially in the 1 st trimester, which in turn increases the risk of severe PE. However, Hietala [3] found no signi cant differences in gestational Hcy between the PE group and the normal pregnancy group. The role of Hcy remains controversial. Therefore, this study identi es the changes in Hcy during pregnancy, risk factors for hyper homocysteine (H-Hcy) and effects of gestational Hcy on PE.

Participant
Single pregnant women who received regular prenatal care at the International Peace Maternal and Child Health Hospital (IPMCH) a liated with Shanghai Jiao Tong University School of Medicine from July 2018 to September 2018 were selected and they all taken folic acid or multivitamins (containing folic acid) during pregnancy. The exclusion criteria were as follows: (1) twin or multiple pregnancy; (2) pre-pregnancy with hypertension, diabetes, kidney disease, severe anemia and other internal and surgical diseases; (3) use of aspirin or other drugs during pregnancy; and (4) non-Han pregnancy. A total of 1142 participants were enrolled in this study. After the pregnant women were enrolled in the 1 st trimester (10-14 weeks), they were followed up when the need antenatal care and blood samples were collected in the 1 st trimester (10-14 weeks), the 2 nd trimester (24-28 weeks), and the 3 rd trimester (30-34 weeks) when they have to do regular blood test during pregnancy. The follow-up was completed until 6 weeks postpartum or withdraw. General data, blood biochemical results and pregnancy outcomes of the subjects were collected. The diagnostic criteria for PE were based on American College of Obstetricians and Gynecologists (ACOG) guidelines (2019) for gestational hypertension and PE. [4] All participants provided written informed consent.

Method
According to the above inclusion and exclusion criteria, a total of 1142 pregnant women were included in this study. Levels of Hcy, FA and vitamin B12 were detected in the 1 st trimester (10-14 weeks), 2 nd trimester (24-28 weeks), and 3 rd trimester (30-34 weeks), and MTHFR genes (rs1801133, rs1801131, rs17367504) were also detected. At the same time, clinical information of the enrolled pregnant women was collected, including general information, test results and pregnant outcomes.

Blood sampling
Samples of venous blood (4 ml) were collected in the rst trimester (10-14 weeks), second trimester (24 to 28 weeks), and third trimester (30-34 weeks). Two milliliters of blood were placed in the tube containing inert separation of coagulant, and the upper serum was taken for Hcy, folic acid and vitamin B12 detection after centrifugation. The other 2 ml of blood was placed in a tube containing EDTA anticoagulant, and the lower blood cells were collected for DNA extraction after centrifugation to detect the MTHFR gene.

Biochemical analyses
In this study, the cyclic enzymatic method was used to detect Hcy (Wuhan Hi-tech Medical Devices Park, Wuhan, China) with Roche reagent (Cabas8000 automatic biochemical analyzer, Roche Ltd., Shanghai, China). Folic acid (Daoyuan Biological Technology Ltd., Shanghai, China) and vitamin B12 (Daoyuan Biological Technology Ltd., Shanghai, China) were detected by chemiluminescence (ARCHITECT i2000, Abbott, Shanghai, China), and the Abbott ARCHITECT i2000 biochemical analysis system was used.
Statistical analyses R 3.5.3 software (Mathsoft, USA) was used for statistical analysis. For demographic description, continuous variables were described by the mean±SD and quartile spacing, and classi cation variables were described by the proportion of each subgroup.
For continuous variables, rst, according to the single factor linear model, the initial screening was used to identify the risk factors related to Hcy, folic acid and vitamin B12 (p≤0.1), the optimal multiple linear model (AIC minimum) was determined based on the backward elimination method (AIC minimum), and then VIF < 2 was taken as the standard to determine whether collinearity existed between variables. Background factors related to Hcy, folic acid and vitamin B12 were analyzed according to the determined multiple linear model (P<0.05). The Mauchly sphericity test determines whether the variance in the difference between different measurements is equal when repeated measurements are made. According to the random intercept model, the changes in Hcy, folic acid and vitamin B12 levels in different periods of pregnancy were analyzed, and the curves of Hcy, folic acid and vitamin B12 levels in different periods of pregnancy were drawn in the form of the mean ± standard deviation.
According to pregnancy outcome, for the classi ed variable, the optimal multiple linear model (AIC minimum) was determined based on the backward elimination method based on the initial screening of in uencing factors related to outcome (P≤0.1) in the univariate logistic regression model, and collinearity among variables was judged based on VIF<2. The regression model was used to analyze the in uencing factors related to pregnancy outcome (p < 0.05). The statistical tests were all double-sided, and the con dence interval was 95% reliable.

Baseline characteristics
Demographic description was compared using mean±SD and the proportion of each subgroup (Table1). A total of 1142 participants, with an average age of 31.29±3.96 years and a pre-gestational BMI of 21.16±2.75 kg/m 2 was selected for this study. Three MTHFR genes were detected as follows: (1) MTHFR rs1801133: 366 cases with CC type (32.05%), 529 cases with CT type (46.32%), and 247 cases with TT type (21.63%). The T mutation rate was 67.95%. (2) MTHFR rs1801131: 752 cases with AA type (65.85%), 356 cases with AC type (31.17%), and 34 cases with CC type (2.98%). The C mutation rate was 18.56%.

Changes in Hcy during pregnancy
Hcy levels in different trimester were compared using paired t test. Hcy was the lowest in the rst trimester compared with the second and third trimesters (p<0.001), but the participants appeared to have no signi cant differences between the second and third trimesters ( Figure 2 and Table 2).

Risk factors for Hcy
Pearson correlation coe cient r was used to determine the correlation between Hcy and FA, vitamin B12 in the same period. Hcy during pregnancy was negatively correlated with folic acid (r=-0.17, p<0.001) and vitamin B12 (r=-0.15, p<0.001) in the same period (Table 3), and Hcy increased with the decrease in folic acid and vitamin B12. We used single factor linear model and multiple linear model to analyze the risk factors of Hcy. Along with the gradual decrease in folic acid and vitamin B12 during pregnancy, Hcy increased as gestational weeks increased ( Figure 2). MTHFR rs1801133 had a signi cant effect on Hcy during pregnancy, and heterozygous CT (p=0.025, 95% CI 0.018,0.275) and homozygous TT (p<0.001, 95% CI 0.185,0.501) both caused an increase in Hcy. However, the impact of homozygous TT was stronger than that of heterozygous CT, and G-spot mutations in MTHFR rs17367504 have protective effects on Hcy, which may cause Hcy to decline during pregnancy (Table 4).

Impact of Hcy on PE
The regression model was used to analyze the in uencing factors related to PE (Table5). The rise in Hcy levels in the third trimester signi cantly increased the risk of PE (OR = 1.2, 95% CI 1.01,1.42), particularly early-onset PE (OR = 3.63, 95% CI 1.71,7.71) and severe PE (OR = 3.63, 95% CI 1.71,7.71), but there is no correlation between elevating Hcy and late-onset PE (p>0.05). Hcy levels in the rst trimester and PE had no signi cant correlation. We also found that the three MTHFR genes referred to in this study had no signi cant correlation with the risk of PE.

Discussion
We conducted a prospective study that focused on the changes in Hcy during pregnancy, risk factors for Hcy and the impacts of Hcy on PE because there were controversies about Hcy, MTHFR genes and PE.

Changes in Hcy
The average value for Hcy in the subjects in this study was 4.36±1.05 μmol/L in the rst trimester (10-14 weeks), 5.95±1.46 μmol/L in the second trimester (24-28 weeks), and 5.95±1.74 μmol/L in the third trimester (30-34 weeks). Hcy in the second and third trimesters was signi cantly higher than that in the rst trimester, similar to the results of Rolf [5] and MW Wallace [6], which may be due to physiological changes and endocrine hormone changes during pregnancy such as HCG. Boxmeer [7] found that after HCG administration, Hcy level had been signi cantly lower than before during the menstrual cycle. This research found that pregnancy folic acid and vitamin B12 gradually decrease with increasing gestational age, which could explain the changes in Hcy during pregnancy.
Mark Walker [8] found that Hcy drops in the second trimester and then rises in the third trimester, which is different from the results of this study, and the possible reasons are as follows: (1)  In this study, MTHFR was found that it might be another factor causing the increase in Hcy, and the mutation rate of the MTHFR gene involved in this study was similar to the results of Moll [9] and Wu. X [10]. Mutations in the MTHFR gene can signi cantly weaken the function of tetrahydrofolate reductase, and the activity of tetrahydrofolate reductase of heterozygous MTHFR rs1801133 can be reduced to 65% of that of normal people, while the activity of homozygous individuals is only 30% [9]. The enzyme activity of homozygous MTHFR rs1801131 was 60% of normal [9]. Therefore, mutations in the MTHFR gene might be one of the risk factors affecting Hcy in pregnancy [11], which was similar to result in this study. At the same time, we also found that the G mutation of MTHFR rs17367504 had had a protective effect on Hcy during pregnancy, and the G mutation might cause Hcy decline. This might explain Thomsen's ndings [12]. Thomsen [12] rst found that the G mutation at MTHFR rs17367504 was a protective factor for PE (OR=0.65, 95% CI 0.53-0.80). In addition to the MTHFR gene, creatinine and protein levels in pregnancy were also possible factors that in uence Hcy in pregnancy. The reason might be that the increase in creatinine affects the metabolism of Hcy and reduces urine excretion, which further causes the increase in Hcy in blood [11].

Impact of Hcy on PE
At present, there have been many studies on the correlation among Hcy, MTHFR genes and PE, and the ndings are controversial. One study found that the Hcy of the T mutants at MTHFR rs1801133 was signi cantly higher than that of the non-mutants [13], which further increases the risk of the occurrence of PE [14]. Sunkara [15] found that in the PE group, the T mutation rate of MTHFR rs1801133 was signi cantly higher than that in the normal group. Furthermore, Dickerson [16] found that the T mutation of MTHFR rs1801133 was not signi cantly correlated with PE. There was no signi cant correlation among the three MTHFR genes referred to in this study and the risk of PE (including early-onset PE and severe PE) (p>0.05). Therefore, this study indicates that there is no direct correlation between the MTHFR genes and PE, and mutations in the MTHFR gene are risk factors for elevating Hcy during pregnancy which might directly or indirectly lead to vascular endothelial dysfunction, oxidative stress response, stimulate vascular smooth muscle cell proliferation, disrupt the body's coagulation and brinolytic system, and ultimately increase the risk of PE [17].
This study found that elevating Hcy might be correlated with the risk of PE in the third trimester (OR=1.2, 95% CI 1.01,1.42), especially in early-onset PE (OR=3.63, 95% CI 1.71, 7.71, p=0.001) and severe PE (OR=1.46, 95% CI 1.21, 1,76, p<0.001), but not late-onset PE (p>0.05). There was no correlation between Hcy in the rst trimester and PE (p>0.05). However, most researchers believe that Hcy in the PE group was signi cantly higher than that in normal pregnancies, and the increasing Hcy level, especially in the rst trimester, signi cantly increases the risk of PE, which is different from the results of this study [2,[18][19][20]. In the Finnish study, Hietala [3] found no signi cant differences in Hcy between the PE group and normal group. Such differences may be due to different gestational weeks detecting Hcy. Kahn [21] has found that Hcy detection at 22-26 weeks of gestation showed no signi cant difference between the PE group and the normal group, while Hogg [22] found that the same group, with no difference at 26 weeks, had a signi cant difference in Hcy detection at 37 weeks. This is similar to the results of this study, and the possible reasons for such differences are as follows: (1) The gestational weeks detected in this study were 10-14 weeks, 24-28 weeks, and 30-34 weeks, and it can be inferred that Hcy before 28 gestational weeks had no signi cant correlation with PE. The detection of Hcy at 30-34 gestational weeks was signi cantly associated with PE, especially early-onset PE and severe PE. (2) The sample size of this study is limited, and the number of cases of PE is small, which is not enough to re ect the predictive value of Hcy in the rst trimester for PE. (3) The detection results of Hcy are greatly affected by blood lipids [23], which are prone to increase during pregnancy, especially for patients with PE who usually have abnormal lipid metabolism [24,25], and it may have a certain in uence on the detection results. (4) This study nd that Hcy in the third trimester was more signi cantly correlated with early-onset PE, possibly because its detection of gestational weeks were very similar to the onset of the disease, or the test samples were collected at the onset of early-onset preeclampsia. (5) This study nd that Hcy in the third trimester was correlated with early-onset PE, other than late-onset PE, possibly due to the different pathogenesis between early-onset PE and late-onset PE. Early-onset PE arises owing to defective placentation which is similar to the pathogenesis mechanism of adverse pregnancy outcomes caused by elevated Hcy , whilst late-onset PE may center around interactions between normal senescence of the placenta and a maternal genetic predisposition to cardiovascular and metabolic disease. [26] Therefore, although the correlation between Hcy in the third trimester and PE is statistically signi cant, further studies are necessary to con rm whether Hcy can be used as an early indicator of clinical disease.

Strengths and limitation
One of the main strengths of this study is that this was a prospective study by testing the same patient at different gestational ages. We analyzed the correlation between Hcy levels in the rst and third trimester of the same patient who nally progressed to PE and took MTHFR genes variation into account.
The limitation of this study regards assessments of folic acid supplementation dose, gestational age, and dietary intake of folate and vitamin B12. Because the dietary structure and living habits of the included subjects are quite different and the use of multivitamins during pregnancy is not the same throughout the gestational period, it is di cult to conduct a quantitative study on the intake of folic acid and vitamin B12 during pregnancy, but a maternal blood test was taken to measure folic acid and vitamin B12 levels instead. On the other hand, the sample size of this study is limited, and the number of cases of PE is small, which is not enough to re ect the predictive value of Hcy in the rst trimester for PE.

Conclusions
In this study, the correlation between Hcy and PE is discussed in detail, and Hcy in pregnancy is lowest in the rst trimester and increases with increasing gestational age. MTHFR, folic acid and vitamin B12 might be the most critical factors for changing Hcy in pregnancy. Hcy level in the third trimester might be associated with PE, especially early-onset PE and severe PE, rather than that in the rst trimester.

Declarations
Ethics approval and consent to participate: The study was approved by the Ethics Committee of the International Peace Maternity and Child Health Hospital, Shanghai Jiaotong University, Shanghai, China. (GKLW2017-102) All participants has given written informed consent.

Consent for publication
Not applicable

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests:
There is no con ict of interest in relation to this article.