Branched-Chain Amino Acids and Their Interactions With Lipid Metabolites for Increased Risk of Gestational Diabetes

Abstract

Objective

We aimed to explore associations of branched-chain amino acids (BCAA) in early pregnancy with gestational diabetes mellitus (GDM), and whether high BCAAs and lipidomics markers had interactive effects on the risk of GDM.

Methods

We conducted a 1:1 case-control study (n = 486) nested in a prospective cohort of pregnant women in Tianjin, China. Blood samples were collected at their first antenatal care visit (median 10 gestational weeks). Serum BCAAs, saturated fatty acids (SFA) and lysophosphatidylcholines (LPC) were measured by liquid chromatography–tandem mass spectrometry analysis. Conditional logistic regression was performed to examine associations of BCAAs with the risk of GDM. Interactions between high BCAAs and high SFA16:0 for GDM were examined using additive interaction measures.

Results

High serum valine, leucine, isoleucine, and total BCAAs were associated with markedly increased risk of GDM (OR of top vs bottom tertiles: 1.91 [95% CI, 1.22-3.01]; 1.87 [1.20-2.91]; 2.23 [1.41-3.52]; 1.93 [1.23-3.02], respectively). The presence of high SFA16:0 defined as ≥ 17.1 nmol/mL (ie, median) markedly increased the ORs of high leucine alone and high isoleucine alone up to 4.56 (2.37-8.75) and 4.41 (2.30-8.43) for the risk of GDM, with significant additive interaction. After adjustment for LPCs, the ORs were greatly elevated (6.33, 2.25-17.80 and 6.53, 2.39-17.86) and the additive interactions became more significant.

Conclusion

BCAAs in early pregnancy were positively associated with the risk of GDM, and high levels of leucine and isoleucine enhanced the risk association of high SFA16:0 with GDM, independent of LPCs.

Gestational diabetes mellitus (GDM) is one of the most common complications of pregnancy and its prevalence has been increasing worldwide, affecting approximately 14% of pregnancies worldwide in 2021, according to an estimate of International Diabetes Federation (1). Although it is a transient disorder of pregnancy, GDM is associated with an increased risk of short-term and long-term adverse health outcomes in both the mothers and their offspring (2-5), including the increased maternal risk of diabetes and cardiovascular disease in later life (6, 7) and increased offspring risk of childhood obesity (8). A meta-analysis from our group showed that lifestyle intervention initiated within 15 weeks of gestation could only induce a 20% reduction in GDM risk and lifestyle intervention after the fifteenth gestational week was ineffective in reducing the risk of subsequent GDM (9). Therefore, it is critically important to explore novel biomarkers of GDM in early pregnancy and to better understand the etiology of GDM for accurately identifying women at high risk in early pregnancy and to design better intervention strategies.

Branched-chain amino acids (BCAA), including valine, leucine, and isoleucine, which are essential amino acids for humans, act as substrates and regulators of protein and energy metabolism (10). In recent years, many cross-sectional and cohort studies have consistently reported the associations between increased plasma BCAA levels and risk of obesity, insulin resistance, and diabetes (11-15), and increased BCAA levels have shown to have predictive values for type 2 diabetes (16). There are also a few studies conducted to address the associations between elevated BCAA levels and increased risk of GDM. A small cohort study of Chinese women (65 GDM events out of 431 women enrolled) reported that elevated isoleucine levels in early pregnancy was significantly associated with subsequent occurrence of GDM (17). Studies from our group (18) and other groups (19-21) have reported that abnormal lipidomics was associated with elevated risk of GDM. Animal studies suggested an interaction between abnormal lipid metabolism and BCAA in development of insulin resistance and diabetes (14, 22). Although it is of interest, there are no studies conducted to explore any interactive effects of BCAA levels and lipidomics markers on the risk of GDM. In a related context, we have reported that high saturated fatty acid (SFA) 16:0 played a pivotal role in development of GDM in Chinese pregnant women (23). It is therefore interesting to examine whether high BCAA levels and high SFA16:0 have interactive effects on the risk of GDM.

In the present study, we used a case-control design nested within a population-based prospective cohort of pregnant women in Tianjin, China to test: (1) the association between the BCAA levels in early pregnancy and the risk of GDM; and (2) the interactive effects between high BCAA levels and high levels of SFA16:0 on the risk of GDM.

Methods

Study Design and Participants

The study used an age-matched case-control study nested in a prospective cohort of pregnant women in central Tianjin, China. The study design, methods, and settings of the cohort have been described in detail previously (24, 25). Briefly, between October 2010 and August 2012, 22 302 pregnant women from the 6 central urban districts of Tianjin, China were registered with a participating primary care hospital of the Antenatal Care System of Tianjin. Ethics approval for this study was obtained from the Ethics Committee of Tianjin Women and Children’s Health Center (TWCHC) and all the participants provided written informed consent before data collection.

The Antenatal Care System of Tianjin used a 2-step screening procedure to identify GDM cases. At first, all the participants were offered a 50-g 1-hour glucose challenge test (GCT) in nonfasting status at 24 to 28 weeks of gestation at the primary care hospital. Then, women were referred to TWCHC GDM Clinic to undergo a standard 75-g 2-hour oral glucose tolerance test (OGTT) after at least an 8-hour fasting status for diagnosis of GDM if their GCT level was ≥ 7.8 mmol/L. Based on the International Association of Diabetes and Pregnancy Study Group (IADPSG) criteria, GDM was defined as either fasting plasma glucose ≥ 5.1 mmol/L or 1-hour plasma glucose ≥ 10.0 mmol/L or 2-hour plasma glucose ≥ 8.5 mmol/L (26).

Of the 22 302 enrolled pregnant women, 2991 women provided overnight fasting blood samples at the early stage of the study (at a median of 10 weeks of gestation). Subsequently, 227 pregnant women were excluded due to lack of GCT results or lack of OGTT results if their GCT was 7.8 mmol/L or greater. Among the remaining 2764 pregnant women, 243 women diagnosed with GDM were used as the cases and 243 women without GDM matched on age (±1 year) were selected as the controls. Finally, the 243 pairs of GDM women and their age-matched controls were included in the current analysis.

Data Collection Procedures

The methods of data collection of the cohort have been described in detail previously (24). At the first antenatal care visit to the primary care hospital, the nurses or obstetricians (uniformly trained in a series of workshops) measured maternal height, weight, and systolic/diastolic blood pressure (SBP/DBP). Other information on maternal characteristics were collected by a series of self-administered questionnaires at registration for pregnancy and at the time of the GCT or retrieved from the central database of the Maternal and Child Health Information System, including maternal age, ethnicity, education attainment, parity, family history of diabetes in first-degree relatives, current smoker before or during pregnancy, and alcohol drinker before or during pregnancy. Body mass index (BMI) was calculated as body weight in kilogram divided by the square of body height in meter. Weight gain to GCT was the difference in body weight between the first antenatal care visit and the GCT time.

Measurement of Serum Branched-Chain Amino Acids

The blood samples were stored in a −80 °C freezer and thawed at 4 °C. Each 20 μL sample was taken and mixed with 10 μL L-Norvaline solution (0.1 mmol/L) and 70 μL acetonitrile. After being vortexed for 20 seconds, the mixture was centrifuged at a rotation speed of 15000g for 10 minutes. Then, 8 μL of the supernatant was used for derivatization according to the method of Waters AccQ Tag. After that, it was injected for liquid chromatography–tandem mass spectrometry (LC-MS/MS) analysis.

The BCAA components were identified and quantified using Eksigent Ultra liquid chromatography 100 coupled with a Triple TOF 5600 system (AB SCIEX) and the analyses were separated on a 2.1 × 100 mm XBridge Peptide BEH C18 column (Waters) equipped with a 4 × 2.0 mm C18 guard column (Phenomenex). The temperatures of the column and the autosampler were maintained at 50 °C and 15 °C, respectively. The separation of BCAA was achieved using a controlled gradient of mobile phase A, which consisted of 10 mM ammonium formate, 0.1% formic acid and 99.9% water, and mobile phase B, which was composed of 1.6% formic acid and 98.4% acetonitrile, at a flow rate of 0.4 mL/min. The gradient elution conditions were set as follows: 0.01 minutes, 5% (v/v) B; 0.01 to 15 minutes, 5% to 40% (v/v) B; 15 to 17 minutes, 40% to 100% (v/v) B; 17 to 20 minutes, 100% (v/v) B; 20 to 20.5 minutes, 100% to 5% (v/v) B; 20.5 to 25 minutes, 5% (v/v) B. The injection volume was 5 μL and the total run time was 25 minutes. The ion source was operated in positive mode under an optimal condition: curtain gas, 30 psi; ion source gas 1, 50 psi; ion source gas 2, 50 psi; ion spray voltage floating, 5500 V; and temperature, 550 °C. The Peak View 1.2 was used to identify BCAA in samples and Multi Quant 2.1 was used to quantity BCAA based on the m/z value and sample retention time.

Statistical Analysis

All statistical analyses were performed using the Statistical Analysis System (SAS), release 9.4 (SAS Institute Inc., Cary, NC). All the tests were two-tailed and P values < 0.05 were considered to be statistically significant. Shapiro-Wilk test was used to check normality of the distribution of continuous variables. These variables were expressed as mean ± SD or median (interquartile range, IQR). Differences of these continuous variables between the cases and the controls were compared using paired Student t test or Wilcoxon signed-rank test where appropriate. Categorical variables were presented as number (percentage) and compared using McNemar test or Fisher’s exact test where appropriate.

Conditional binary logistic regression was performed to obtain odds ratios (OR) and their 95% CI of BCAAs for GDM. The linearity of the associations between serum BCAAs concentrations and GDM risk was checked using restricted cubic spline (RCS) analysis. Because the BCAAs were roughly linearly associated with the risk of GDM, we divided BCAAs into 3 groups according to their tertiles. A structured adjustment scheme was used to control for confounders in the conditional binary logistic regression. First, we used univariate analysis to obtain the unadjusted ORs and 95% CI of BCAAs for the risk of GDM. Second, we performed multivariate analysis to adjust for the traditional GDM risk factors, including prepregnancy BMI, SBP, Han nationality, family history of diabetes in first-degree relatives, parity, education attainment, current smoker before or during pregnancy, alcohol drinker before or during pregnancy, and weight gain to GCT.

We used additive interaction to test interactive effects of high BCAAs with high SFA16:0 for GDM. Because we found SFA16:0 was positively associated with the risk of GDM in a linear manner (23), we defined SFA16:0 ≥ 17.1 nmol/mL, that is, the median, as high SFA16:0. Three measures were used to test additive interaction and they are relative excess risk due to the interaction (RERI), attributable proportion due to the interaction (AP) and synergy index (SI). The additive interaction was considered to be significant if any of RERI > 0, AP > 0 or SI > 1 was statistically significant (27).

Results

Characteristics of the Study Participants

The clinical and biochemical characteristics of the study participants are shown in Table 1. At the first antenatal care visit, the mean age of the 486 pregnant women was 29.2 (SD 3.1) years, the mean weight was 60.9 (SD 10.1) kilograms, and the mean gestational age was 10.1 (SD 2.1) weeks. Compared with the women without GDM, the pregnant women with GDM had higher weight, BMI, and SBP/DBP and were more likely to have a family history of diabetes in first-degree relatives. The levels of serum BCAAs, valine, leucine, and isoleucine in the early pregnancy were higher in the GDM group than in the non-GDM group. Other characteristics including height, ethnicity, education attainment, gestational age, weight gain to GCT, and the status of smoking and drinking before and during pregnancy were similar between the GDM group and the non-GDM group.

Associations of Serum BCAAs With GDM

Valine, leucine, and isoleucine were associated with the risk of GDM in roughly linear manners. In univariate analysis, the top tertile of serum valine, leucine, isoleucine and total BCAAs were associated with markedly increased risk of GDM as compared with their bottom tertile (OR: 1.91 [95% CI, 1.22-3.01]; 1.87 [1.20-2.91]; 2.23 [1.41-3.52]; and 1.93 [1.23-3.02], respectively). After adjustment for traditional risk factors, the ORs of top vs bottom tertiles of serum valine, leucine, isoleucine, and total BCAAs for GDM were 1.79 (1.07-2.97), 1.72 (1.04-2.84), 2.05 (1.21-3.48) and 1.84 (1.10-3.09), respectively, with all ORs remaining statistically significant (Table 2).

Additive Interactions Between High Serum BCAAs and High SFA16:0 for the Risk of GDM

Compared with low leucine (ie, <153.3 nmol/mL, the cutoff point for the top tertile) and low SFA16:0 (ie, <17.1 nmol/mL, the median), concurrent presence of both high leucine (≥153.3 nmol/mL) and high SFA16:0 (≥17.1 nmol/mL) markedly increased the ORs of high leucine alone (adjusted OR: 1.45 [95% CI, 0.74-2.84]) and high SFA16:0 alone (adjusted OR: 2.22 [95% CI, 1.33-3.72]) up to OR: 4.56 (95% CI, 2.37-8.75) for the risk of GDM. The additive interaction was significant (AP: 0.41 [95% CI, 0.01 to 0.82]). Similarly, compared with low isoleucine (ie, <36.2 nmol/mL, the cutoff point for the top tertile) and low SFA16:0, concurrent presence of both high isoleucine and high SFA16:0 markedly increased the ORs of high isoleucine alone (adjusted OR: 1.27 [95% CI, 0.66-2.44]) and high SFA16:0 alone (adjusted OR: 2.12 [95% CI, 1.29-3.48]) up to OR: 4.41 (95% CI, 2.30-8.43) for the risk of GDM, with significant additive interaction (AP: 0.46 [95% CI, 0.09 to 0.82]) (Table 3). After additional adjustment for lysophosphatidylcholines (LPC) species, the ORs of copresence of both high leucine and high SFA16:0 and copresence of both high isoleucine and high SFA16:0 were greatly elevated (6.33 [2.25-17.80] and 6.53 [2.39-17.86]) and the additive interactions became more significant (Table 4). However, the additive interaction between high valine and high SFA16:0 for GDM was nonsignificant (AP: 0.25 [95% CI, −0.24 to 0.73]).

Discussion

In this nested case-control study, we found that serum valine, leucine, isoleucine, and total BCAAs in early pregnancy were positively associated with the risk of GDM in Chinese pregnant women. There were significant additive interactions of high leucine and isoleucine with high SFA16:0 for the risk of GDM, independent of LPC species.

Many studies had previously explored associations of BCAAs with insulin resistance and increased risk of diabetes, but their findings were inconsistent. A meta-analysis of cross-sectional and cohort studies evaluating metabolite markers identified for prediabetes and type 2 diabetes showed that BCAA levels were higher in individuals with type 2 diabetes compared with their counterparts who remained free of type 2 diabetes, and that elevated levels of valine, leucine, and isoleucine were associated with a markedly increased risk of type 2 diabetes (28). Furthermore, adding BCAAs to a high-fat diet worsened the ensuing insulin resistance and the development of glucose intolerance in rodents (14), whereas restricting BCAA improved glucose tolerance and insulin sensitivity (29). While there are a number of studies suggesting that BCAAs contribute to the development of insulin resistance and diabetes, there are also some studies reporting negative findings. A few studies in mice reported that leucine supplementation did not have any detectable effects on insulin sensitivity (30-32). In this connection, an animal study of leucine supplement in the drinking water found that doubling dietary leucine reversed many of the metabolite abnormalities and caused a marked improvement in glucose tolerance (31). Although leucine supplementation caused a rise in leucine in the circulation, it also caused a decline in other BCAAs in the circulation in animals (32). Indeed, all BCAAs were elevated in people with insulin resistance and type 2 diabetes (16, 28) and which ones play a causal role in GDM remains inconclusive. In this nested case-control study, we found that high valine, leucine, and isoleucine in early pregnancy had large and significant effects on the risk of GDM.

GDM is characterized by impaired β-cell function and increased insulin resistance. The mechanisms regarding BCAA for the development of GDM via mediating insulin resistance are not fully clear. One possible mechanism is that high levels of BCAAs activate mammalian target of rapamycin complex 1 (mTORC1), resulting in insulin resistance through the phosphorylation of insulin receptor substrate 1 (IRS-1). Leucine is a potent activator of mTORC1 activity; it promotes mTORC1 activation via directly binding Sestrin2, a negative regulator of mTORC1 activity (33). In the absence of leucine, Sestrin2 binds and inhibits GATOR2, a positive regulator of mTORC1 activity. When leucine is available at physiologically relevant concentrations, Sestrin2 releases GATOR2, promoting full mTORC1 activation (34). In addition, an experimental study found that deprivation of individual BCAA in mice and in cell lines could improve insulin sensitivity, which was possibly mediated by decreasing mTOR pathway but increasing AMP-activated protein kinase (AMPK) signaling pathway (35). Another possible mechanism is that accumulation of potentially toxic BCAA metabolites (caused by abnormal BCAA metabolism) may lead to induction of oxidative stress, mitochondrial dysfunction, impaired insulin action, and ultimately to perturbation of glucose homeostasis. Branched-chain alpha-keto acids (BCKA), that is the metabolites of BCAA, significantly compromised the nonenzymatic and the enzymatic antioxidant defenses and increased the levels of nitrogen reactive species in C6 cells and rats (36, 37). As a result, there was an unbalance between pro-oxidants and antioxidants, a situation defined as oxidative stress, which can affect insulin signal transduction. The molecule 3-hydroxyisobutyrate (3-HIB), also a catabolic intermediate of the BCAA, could alter mitochondrial function and transendothelial fatty acid transport, stimulate muscle fatty acid uptake in vivo, and promote lipid accumulation in muscle, leading to insulin resistance (38).

It is biologically plausible that there were significant additive interactions between individual BCAAs with SFA16:0 for the risk of GDM. High levels of BCAAs can persistently activate mTORC1, followed by activation of ribosomal protein S6 kinase 1 (S6K1), resulting in insulin resistance through the phosphorylation of IRS-1 (14, 39). SFA16:0 (ie, palmitic acid) was also shown to promote mTOR signaling in rat hepatocytes and played a key role in insulin regulation and β-cell function (40-42). It is conceivable that a synergistic effect of BCAA and SFA16:0 on mTORC1 activity and inducing insulin signaling disorders. Nevertheless, high LPCs increased the risk of GDM not by activating mTORC1 to induce insulin resistant, but by increased inflammation and oxidative stress (43). High LPCs not only could increase the release of inflammatory cytokines, such as interleukin (IL)-6 and IL-8 (44, 45), but also increase oxidative stress via inducing the overproduction of nitric oxide (46). High BCAAs, SFA16:0, and LPCs promoted the development of GDM by a different mechanism. In this regard, we found that high leucine and isoleucine had significant additive interactions with high SFA16:0 for the risk of GDM, independent of LPCs.

Our study has some strengths. First, the study collected the participants’ blood sample in early pregnancy (at a median of 10 gestational weeks), thus a reverse causation was unlikely. Second, traditional risk factors for GDM were carefully collected in our cohort and were available for the current analysis. Consequently, a large proportion of the confounding effect of traditional factors could be removed. Our study also has some limitations. First, the BCAA levels and SFA16:0 were influenced by dietary intake, but we didn’t collect the information on dietary habits from the participants. However, diet is associated with many demographic and clinical factors. Therefore, diet habit was more likely to be a cause but not a consequence of altered BCAA levels. Second, we did not collect history of previous GDM in the cohort at the enrollment. All pregnant women of this study were recruited in Tianjin, China, between October 2010 and August 2012. During this period, China still implemented its one-child policy and more than 95% of the pregnant women in the cohort were nulliparous (24). Given that the prevalence of GDM in the cohort was not too high (ie, 9.3%), the number of pregnant women with history of previous GDM in this case-control study would be very small. Third, our findings were generated by a single pregnant women population from urban Tianjin. These findings need to be replicated in other Chinese and non-Chinese populations of pregnant women.

In conclusion, we found that high serum levels of valine, leucine, and isoleucine in early pregnancy were associated with markedly increased risk of GDM. The high levels of leucine and isoleucine had significant additive interactions with high SFA16:0 for the risk of GDM, independent of LPC species. It is warranted to explore biological mechanisms underlying these interesting observations and to test their predictive values for GDM in early pregnancy in Chinese and non-Chinese women.

Abbreviations

Abbreviations
 
• AP

  attributable proportion due to the interaction

 
• BCAA

  branched-chain amino acid

 
• BMI

  body mass index

 
• DBP

  diastolic blood pressure

 
• GCT

  glucose challenge test

 
• GDM

  gestational diabetes mellitus

 
• IRS-1

  insulin receptor substrate 1

 
• LPC

  lysophosphatidylcholine

 
• mTORC1

  mammalian target of rapamycin complex 1

 
• OGTT

  oral glucose tolerance test

 
• OR

  odds ratio

 
• RERI

  relative excess risk due to the interaction

 
• SBP

  systolic blood pressure

 
• SFA

  saturated fatty acid

 
• SI

  synergy index

 
• TWCHC

  Tianjin Women and Children’s Health Center

Acknowledgments

The authors thank all doctors, nurses, and research staff at the primary care hospitals, 6 district-level women and children’s health centers (WCHCs) and Tianjin Women and Children’s Health Center (TWCHC) for their participation in this study.

Funding

This research was supported by the National Natural Science Foundation of China (Grant Nos: 81870549; 81900724) and National Key Research and Development Program of China (Grant Nos: 2021YFA1301200; 2021YFA1301202).

Author Contributions

X.Y. and Z.F. conceived the idea and designed the study; J.Leng and W.L. collected the data; K.Y & Z.F. conducted the measurement of serum amino acids and lipidomics; N.L. analyzed the data; N.L and J.Li wrote the first draft. All others gave critical comments and edited the manuscript. All authors gave comments and contributed to the writing of the manuscript and agreed to submit and publish the manuscript. X.Y. and N.L. took full responsibility for the work as a whole, including the study design, access to the data, and decision to submit.

Conflict of Interest Statement

The authors declared no conflict of interest.

Data Availability

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

References

Author notes