Figures
Abstract
Background
Recent studies have determined that branched-chain (BCAAs) and aromatic (AAAs) amino acids are strongly correlated with obesity and atherogenic dyslipidemia and are strong predictors of diabetes. However, it is not clear if these amino acids are capable of identifying subjects with coronary artery disease (CAD), particularly with subclinical atherosclerosis who are at risk of developing CAD.
Methods
Four hundred and seventy two Chinese subjects (272 males and 200 females, 42–97 y of age) undergoing physical exams were recruited at random for participation in the cross-sectional study. Serum BCAAs and AAAs were measured using our previously reported isotope dilution liquid chromatography tandem mass spectrometry method. Bilateral B-mode carotid artery images for carotid intima-media thickness (cIMT) were acquired at end diastole and cIMT values more than 0.9 mm were categorized as increased. Correlations of BCAAs with cIMT and other CAD risk factors were analyzed.
Results
BCAAs and AAAs were significantly and positively associated with risk factors of CAD, e.g., cIMT, BMI, waist circumference, blood pressure, fasting blood glucose, TG, apoB, apoB/apoAI ratio, apoCII, apoCIII and hsCRP, and were significantly and negatively associated with HDL-C and apoAI. Stepwise multiple linear regression analysis revealed that age (β = 0.175, P<0.001), log BCAA (β = 0.147, P<0.001) and systolic blood pressure (β = 0.141, P = 0.012) were positively and independently associated with cIMT. In the logistic regression model, the most and only powerful laboratory factor correlated with increased cIMT was BCAA (the odds ratio of the fourth quartile compared to the first quartile was 2.679; P = 0.009).
Citation: Yang R, Dong J, Zhao H, Li H, Guo H, Wang S, et al. (2014) Association of Branched-Chain Amino Acids with Carotid Intima-Media Thickness and Coronary Artery Disease Risk Factors. PLoS ONE 9(6): e99598. https://doi.org/10.1371/journal.pone.0099598
Editor: Yanqiao Zhang, Northeast Ohio Medical University, United States of America
Received: January 27, 2014; Accepted: May 16, 2014; Published: June 9, 2014
Copyright: © 2014 Yang et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This study was supported by research grants from the National Science & Technology Pillar Program under the Twelfth Five-year Plan of China (2012BAI37B01), the National Natural Science Foundation of China (81171647 and 81301487) and the National Department Public Benefit Research Foundation by Ministry of Health P. R. China (201002011 and 201302008). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Coronary artery disease (CAD) is the leading cause of death in industrialized countries. Given the availability of effective interventions for delaying or preventing the onset of CAD, early identification of at-risk individuals is particularly crucial [1]. Evaluation with developing tools may improve risk stratification and enhance our understanding of the disease process. Recent advances in liquid chromatography tandem mass spectrometry (LC/MS/MS) allow the acquisition of high-throughput profiles of the metabolic status of whole organisms (eg, metabolomics), which may be particularly useful for the risk assessment of human diseases, such as cardiovascular disease [1], [2].
Branched-chain and aromatic amino acids (BCAAs and AAAs) are noteworthy in the context of experimental and clinical data. It has been suggestedthat these amino acids may both be markers and effectors of insulin resistance [3]–[5] and can predict future diabetes [6], [7]. Both sets of findings have been corroborated by more recent studies using LC-MS-based metabolomics platforms. However, these methods for the analysis of such amino acids in biological samples are time consuming. Our previous study developed a simple, fast and reliable isotope dilution LC/MS/MS method for the measurement of serum BCAAs and AAAs [8]. By using this method, we documented strong associations between certain amino acids and lipid profiles, such as a positive correlation with TG and LDLb-C, and a negative correlation with HDL-C and HDL2-C [8]. Because the state of dyslipidemia which often precedes type 2 diabetes is associated with an increased risk of developing atherosclerosis and CAD [9], there is great interest in determining whether these amino acids are capable of identifying subjects with coronary artery disease (CAD), particularly with subclinical atherosclerosis who are at risk of developing CAD.
The aims of the current study was to evaluate the associations of BCAAs and AAAs with increased carotid intima-media thickness (cIMT), a well-established marker of subclinical atherosclerosis that can be measured easily and non-invasively [10], and other risk factors of CAD.
Materials and Methods
Study samples
This is a cross-sectional study conducted on Chinese subjects. From a group of Beijing residents attending an annual physical examination in a period from August to October, 2011, 472 subjects (272 males and 200 females, 42–97 y of age) were recruited randomly. The sera from fasting blood samples obtained from volunteers were isolated and stored at −80°C until being analyzed.
Ethics statement
This study has been reviewed and approved by the Ethics Committee of Beijing Hospital, Ministry of Health. All studied individuals had been made aware in writing of the intended use of their blood sample and provided written consent.
LC/MS/MS measurement for branched-chain and aromatic amino acids
The serum BCAAs [valine (Val), isoleucine (Ile) and leucine (Leu)] and AAAs [tyrosine (Tyr) and phenylalanine (Phe)] were measured using our previously reported isotope dilution LC/MS/MS method [8]. Briefly, 0.05 ml aliquots of calibrators or serum samples were mixed with 0.05 ml of the isotopic labeled internal standard solution and the amino acids were extracted with 0.4 ml of acetonitrile containing 0.1% formic acid, and analyzed using LC/MS/MS. The LC separation was performed on an Agilent 1200 series LC system. Two microliter aliquots of the prepared samples were injected onto a Waters Shield C18 column (3.5 µm, 2.1×150 mm) maintained at 20°C and eluted with a mobile phase of 0.01% formic acid in water-acetonitrile (90∶10) at a flow rate of 0.3 ml/min. An API 4000 triple quadrupole mass spectrometer (Sciex Applied Biosystems) was used for the MS/MS detection. The detection was performed with positive electronic spray ionization (ESI) in multiple reaction monitoring (MRM) mode at a source temperature of 700°C and a voltage of 5500 V. The dwell times were 0.08 s for MRM. Nitrogen was used as the curtain, nebulizer and collision gas at pressures of 50, 60 and 70 psi, respectively. Certain ion transitions for the amino acids and their internal standards were monitored following our previously reported method [8]. The calibration curve was generated using a linear regression of the peak area ratios (y) vs. the amino acid concentrations (x) of the calibrators.
Carotid intima-media thickness assays
Bilateral B-mode carotid artery images for intima-media thickness were acquired at end diastole (defined as the R wave of an electrocardiogram) using a SDU-1200 ultrasound imager (Shimadzu, Japan) with a linear array 5.5/7.5 MHz probe. The posterior wall of the distal common carotid artery, one centimeter below the bifurcation, was assessed by using the automated IMT analysis system. The maximum of the left and right cIMT readings was used for the analyses. The imaging analysis was performed by an experienced technician blinded to the study participants.
Other parameters and laboratory assays
A history of diseases and medications and a family history of diseases were recorded using a list of questions. A subject's smoking status was classified as current, former, or never smoker. At the physical examination center, height, weight, waist circumference (WC) and sitting blood pressure were measured. The serum samples were also tested for fasting blood glucose (FBG), total cholesterol (TC), triglycerides (TG), high density lipoprotein cholesterol (HDL-C), low density lipoprotein cholesterol (LDL-C), apolipoprotein (apo) AI, B, CII, CIII and high sensitive C reactive protein (hs-CRP) using assay kits from Sekisui Medical Technologies (Osaka, Japan) on a Hitachi 7180 chemistry analyzer.
Statistical analyses
Categorical variables are presented as frequencies and percentages, and continuous variables as means and standard deviations, or medians and interquartile ranges (25th to 75th percentile) for variables with skewed distributions. Increased cIMT was defined as a cIMT of 0.9 mm or higher [11]; hypertension as a systolic blood pressure (SBP) of 140 mm Hg or higher, or a diastolic blood pressure (DBP) of 90 mm Hg or higher or receiving antihypertensive treatment; diabetes mellitus as a FBG of 7.0 mmol/l or higher or receiving diabetes treatment; dyslipidemia as a serum TC>6.21 mmol/L, or LDL-C>4.14 mmol/L, or TG>1.70 mmol/L, or HDL-C<1.04 mmol/L; abdominal obesity as a waist circumference of 90 cm or higher for men and 80 cm or higher for women. The one-way analysis of variance (ANOVA) test was used to evaluate the differences in the parametric samples between the normal and increased cIMT groups, while the nonparametric Mann-Whitney test was used to determine significant differences when the data were not normally distributed. The associations of certain AAs with cIMT and other risk factors were assessed with the use of nonparametric Spearman correlation analyses. Stepwise multiple linear regression analysis was used to test the independent relationships of the measured variables with cIMT; meanwhile, collinearity testing was used to avoid including interdependent model variables. The significance levels for entering and removing an explanatory variable were set at 0.05 and 0.10, respectively. A stepwise logistic regression analysis was used to estimate the association between variables and subclinical atherosclerosis status, i.e., the presence or absence of increased cIMT. Odds ratios and 95% confidence intervals (CIs) for increased cIMT versus normal cIMT were calculated. Data were adjusted for age, sex, BMI, smoking status, presence or absence of diabetes, hypertension and dyslipidemias, family history of CAD, and therapies for diabetes, hypertension and dyslipidemias (hypercholesterolemia). All reported P values were two-tailed, with a P value of 0.05 indicating statistical significance. Analyses were performed with the use of SPSS software, version 16.0 (SPSS Inc.).
Results
Characteristics of the cross-sectional study population
The characteristics of the population are shown in Table 1. The most common characteristic was increased cIMT (72% of women, 85% of men). The prevalence of hypertension, diabetes, obesity, increased TG, TC and LDL-C, and decreased HDL-C were 59%, 18%, 12%, 27%, 9%, 7% and15%, respectively. Two percent of the women and 41% of the men were current or former smokers.
Univariate analyses
The concentrations of each AA showed skewed and leptokurtic distributions. As shown in Table 2, the increased cIMT group had significantly higher levels of Val, Ile, Leu, total BCAAs and Phe than the normal cIMT group, P<0.01. However, there were no significant differences for Tyr in the increased cIMT group compared to the normal cIMT group, though increment levels were observed. The normal and increased cIMT individuals differed significantly in age, BMI, WC, SBP and FBG concentration, P<0.05. Individuals with obesity and diabetes had significantly elevated BCAAs (P<0.01) and Phe levels (P<0.05).
Correlation analyses
The relationships between certain amino acids and cIMT, and other parameters are shown in Table 3. In the nonparametric Spearman correlation analyses, Val, Ile, Leu, Phe and the total BCAA and AAA concentrations were significantly and positively correlated with cIMT (P<0.05). There was also a strong correlation between each amino acid and BMI (P<0.001) and waist circumference (P<0.001). Furthermore, all BCAA and AAA concentrations were significantly and positively correlated with TG and the apoB/apoAI ratio (P<0.001) and were inversely associated with HDL-C and apoAI (P<0.001). Ile and Leu were negatively correlated with TC (P<0.05). Positive correlations of BCAA and AAA concentrations with hsCRP, apoCII and apoCIII were also found. Serum BCAAs seemed to be negatively correlated with age, especially Leu (P<0.01). Positive correlations between all AAs and DBP (P<0.05) as well as Val and SBP (P<0.05), were observed. Most of BCAAs and AAAs were correlated with FBG and apoB. In addition, cIMT was significantly correlated with age (r = 0.228, P<0.001), BMI (r = 0.125, P<0.01), WC (r = 0.169, P<0.001), SBP (r = 0.218, P<0.001), FBG (r = 0.105, P<0.05), TG (r = 0.100, P<0.05), HDL-C (r = −0.125, P<0.01), apoAI (r = −0.116, P<0.05), and the apoB/apoAI ratio (r = 0.105, P<0.05). The BCAAs and AAAs were also strongly correlated with each other (P<0.001).
Multiple linear regression model
For multivariate reevaluation of the univariate correlations, all variables given in Table 3 were entered into a stepwise multiple linear regression analysis as independent variables to identify significant contributors to the distribution of cIMT. The stepwise multiple linear regression analysis revealed that age (β = 0.175, P<0.001), log BCAA (β = 0.147, P<0.001) and SBP (β = 0.141, P = 0.012) were independent factors that correlate with cIMT (adjusted R2 = 0.041, P = 0.002) as shown in Table 4.
Logistic regression model
The results of the logistic regression analysis of the differences between the increased and normal cIMT groups are shown in Table 5. Although the univariate models demonstrated significant differences in several variables between increased cIMT and normal cIMT groups (Table 2), many of them did not appear in the model adjusted for age, sex, BMI, smoking, BCAA quartile, AAA quartile, diabetes mellitus, hypertension, dyslipidemias, family history of CAD, and diabetes, hypertension and cholesterol-lowering therapies. In the logistic regression model, BCAA was the most and sole powerful laboratory factor correlated with increased cIMT (the odds ratio of the fourth quartile compared to the first quartile was 2.679; P = 0.009).
Discussion
Recent studies have consistently shown that BCAAs and AAAs are strongly correlated with insulin resistance [3]–[5], diabetes development [6], [7] and intervention outcomes [12]–[14] and are uniquely responsive to therapeutic interventions. Some other investigations have demonstrated that circulating levels of BCAAs and AAAs are not only associated with obesity and impaired glucose tolerance but also with dyslipidemia [15]. One of our previous studies also found that BCAAs and AAAs were related to atherogenic dyslipidemia [8], which is characterized by an increase in serum TG, a decrease in HDL-C and the prevalence of small-dense LDL [16]. The state of dyslipidemia, which often precedes diabetes, is associated with an increased risk of developing atherosclerosis and CAD [9]; thus, it is of interest to explore the associations of BCAAs and AAAs with the progress of atherosclerosis.
There have been a few studies on the association between certain amino acids and CAD. Shah and colleagues [17] used targeted LC-MS to profile 69 metabolites in individuals with and without CAD that had been referred to an academic cardiac catheterization laboratory. BCAAs were positively and independently associated with the presence of CAD (OR, 1.36; P = 0.02). However, they subsequently found that BCAAs were negatively associated with all-cause death (232 cases, HR, 0.86; P = 0.03) and all-cause death plus MI (294 cases, HR, 0.88; P = 0.05) in a cohort study with a median follow-up of 3.1 years [18]. Subjects in the two studies were all patients undergoing cardiac catheterization and the CAD prevalence was high (∼60% with 1–3 diseased vessels). The seemingly discrepant results in the two studies may be related to the fact that older people are more susceptible to all-cause death and BCAAs are likely to be negatively associated with age (discussed below). BCAAs would still be positively, or at least not negatively, associated with atherosclerosis related diseases as the negative correlation became weaker when MI was included (HR 0.88 vs. 0.86). Recently, Magnusson and colleagues [19] reported on a Swedish population based cohort study and found that an amino acid score including branched-chain and aromatic amino acids (Ile, Tyr and Phe) predicted cardiovascular diseases during 12 years of follow-up and was independently correlated with cIMT. This study clearly indicates that certain amino acids are early markers of atherosclerotic CAD development.
In the present study, we explored the association of serum amino acids with atherosclerosis in randomly selected middle-aged and older Chinese subjects. It is well-established that cIMT reflects subclinical atherosclerosis and is an important risk factor for CAD [10], [20]–[22]. The cIMT values of 472 subjects were assessed by B-mode ultrasonography and the subjects were categorized into increased and normal cIMT groups using a cut-point of 0.9 mm [11]. Serum BCAAs and AAAs in the subjects were measured with our previously established LC/MS/MS method. Serum levels of BCAAs and AAAs except for Tyr in the increased cIMT group were significantly higher than those in the normal group. The correlation analyses also showed strong associations of BCAAs and AAAs with cIMT. The relationship between BCAAs and cIMT was still present in the multivariate linear regression analysis, which controlled for traditional CAD risk factors. Total BCAA concentrations, along with age and SBP were independent factors that contributed to explaining the variance in cIMT and individuals with increased cIMT tended to be older and have higher BCAA levels. Among all biochemical markers assessed, BCAAs were the one that were correlated most with increased cIMT. Taken together, the current study demonstrates that a higher level of BCAAs is positively and independently associated with increased cIMT. Our findings are consistent with that of Magnusson et al [19] although our study sample is apparently different in ethnicity, BMI, and probably other culture-related aspects, such as lifestyle and diet.
In addition, we also analyzed the correlations of BCAAs and AAAs with other clinical and laboratory characteristics and the results were consistent with prior studies. Felig and colleagues [23] originally reported that BCAAs and AAAs correlated with obesity and serum insulin. More recent investigations, as well as our previous research, highlighted associations of certain amino acids with obesity, elevated serum glucose, dyslipidemia and high blood pressure [3]–[5], [8], [15]. The present study extends the prior work by demonstrating robust and reproducible associations of circulating levels of BCAAs and AAAs with these previously studied risk factors. Furthermore, the present study also demonstrated significant correlations of BCAAs and AAAs with inflammation related markers, such as hsCRP, and some other apolipoproteins, such as apoCII and apoCIII. It should also be noted that this study showed negative correlations between BCAAs and age. Further studies should be undertaken to determine and evaluate the mechanisms behind these relationships.
Several limitations of this study deserve mentioning. First, we did not analyze the effect of individual BCAAs and AAAs, but used the sum of the BCAAs and AAAs because of their strong intercorrelations which may have influenced the accuracy of explanation of the model. Second, the study population was comprised of predominantly middle-aged to elderly individuals. The generalizability of our findings to younger individuals is not yet known. Third, we did not record sedentary lifestyle, stress/depression and other possible risk factors of atherosclerosis in the subjects. We did not either record the details of the medications. These factors may also affect the observed correlations. Finally, this is a cross-sectional study of small sample size and the findings need to be confirmed in future studies.
In conclusion, using a rapid and precise LC/MS/MS method to measure serum BCAAs and AAAs, we demonstrated obvious associations of BCAAs and AAAs with cIMT, BMI, lipid profiles, blood pressure and inflammation risk factors. These associations would open a new field for the mechanistic understanding of atherosclerosis and the risk assessment of CAD. Further investigation are warranted to test whether plasma amino acid measurements can help predict subclinical atherosclerosis or CAD and to elucidate the biological mechanisms of the identified associations between plasma amino acids levels and various CAD risk factors.
Author Contributions
Conceived and designed the experiments: RY JD WC. Performed the experiments: RY JD HZ HL HG SW CZ SW MW SY. Analyzed the data: RY JD WC. Contributed reagents/materials/analysis tools: RY JD HZ HL HG SW CZ SW MW SY. Wrote the paper: RY JD WC. Reviewed manuscript and made suggestions: HZ HL HG SW CZ SW MW SY.
References
- 1. Gilstrap LG, Wang TJ (2012) Biomarkers and cardiovascular risk assessment for primary prevention: an update. Clin Chem 58: 72–82.
- 2. Nicholson JK, Wilson ID (2003) Opinion: understanding ‘global’ systems biology: metabonomics and the continuum of metabolism. Nat Rev Drug Discov 2: 668–676.
- 3. Huffman KM, Shah SH, Stevens RD, Bain JR, Muehlbauer M, et al. (2009) Relationships between circulating metabolic intermediates and insulin action in overweight to obese, inactive men and women. Diabetes Care 32: 1678–1683.
- 4. Newgard CB, An J, Bain JR, Muehlbauer MJ, Stevens RD, et al. (2009) A branched-chain amino acid-related metabolic signature that differentiates obese and lean humans and contributes to insulin resistance. Cell Metab 9: 311–326.
- 5. Tai ES, Tan ML, Stevens RD, Low YL, Muehlbauer MJ, et al. (2010) Insulin resistance is associated with a metabolic profile of altered protein metabolism in Chinese and Asian-Indian men. Diabetologia 53: 757–767.
- 6. Wang TJ, Larson MG, Vasan RS, Cheng S, Rhee EP, et al. (2011) Metabolite profiles and the risk of developing diabetes. Nat Med 17: 448–453.
- 7. Wurtz P, Soininen P, Kangas AJ, Ronnemaa T, Lehtimaki T, et al. (2013) Branched-chain and aromatic amino acids are predictors of insulin resistance in young adults. Diabetes Care 36: 648–655.
- 8. Yang R, Dong J, Guo H, Li H, Wang S, et al. (2013) Rapid and precise measurement of serum branched-chain and aromatic amino acids by isotope dilution liquid chromatography tandem mass spectrometry. PLoS One 8: e81144.
- 9. Grundy SM (1998) Hypertriglyceridemia, atherogenic dyslipidemia, and the metabolic syndrome. Am J Cardiol 81: 18B–25B.
- 10. Fernandes VR, Polak JF, Edvardsen T, Carvalho B, Gomes A, et al. (2006) Subclinical atherosclerosis and incipient regional myocardial dysfunction in asymptomatic individuals: the Multi-Ethnic Study of Atherosclerosis (MESA). J Am Coll Cardiol 47: 2420–2428.
- 11. Mancia G, De Backer G, Dominiczak A, Cifkova R, Fagard R, et al. (2007) 2007 ESH-ESC Practice Guidelines for the Management of Arterial Hypertension: ESH-ESC Task Force on the Management of Arterial Hypertension. J Hypertens 25: 1751–1762.
- 12. Laferrere B, Reilly D, Arias S, Swerdlow N, Gorroochurn P, et al. (2011) Differential metabolic impact of gastric bypass surgery versus dietary intervention in obese diabetic subjects despite identical weight loss. Sci Transl Med 3: 80re82.
- 13. Hsiao G, Chapman J, Ofrecio JM, Wilkes J, Resnik JL, et al. (2011) Multi-tissue, selective PPARgamma modulation of insulin sensitivity and metabolic pathways in obese rats. Am J Physiol Endocrinol Metab 300: E164–174.
- 14. Shah SH, Crosslin DR, Haynes CS, Nelson S, Turer CB, et al. (2012) Branched-chain amino acid levels are associated with improvement in insulin resistance with weight loss. Diabetologia 55: 321–330.
- 15. Cheng S, Rhee EP, Larson MG, Lewis GD, McCabe EL, et al. (2012) Metabolite profiling identifies pathways associated with metabolic risk in humans. Circulation 125: 2222–2231.
- 16. Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation 106: 3143–3421.
- 17. Shah SH, Bain JR, Muehlbauer MJ, Stevens RD, Crosslin DR, et al. (2010) Association of a peripheral blood metabolic profile with coronary artery disease and risk of subsequent cardiovascular events. Circ Cardiovasc Genet 3: 207–214.
- 18.
Shah SH, Sun JL, Stevens RD, Bain JR, Muehlbauer MJ, et al.. (2012) Baseline metabolomic profiles predict cardiovascular events in patients at risk for coronary artery disease. Am Heart J 163: : 844–850 e841.
- 19. Magnusson M, Lewis GD, Ericson U, Orho-Melander M, Hedblad B, et al. (2013) A diabetes-predictive amino acid score and future cardiovascular disease. Eur Heart J 34: 1982–1989.
- 20. Lorenz MW, Markus HS, Bots ML, Rosvall M, Sitzer M (2007) Prediction of clinical cardiovascular events with carotid intima-media thickness: a systematic review and meta-analysis. Circulation 115: 459–467.
- 21. Polak JF, Pencina MJ, Pencina KM, O'Donnell CJ, Wolf PA, et al. (2011) Carotid-wall intima-media thickness and cardiovascular events. N Engl J Med 365: 213–221.
- 22. Chambless LE, Folsom AR, Clegg LX, Sharrett AR, Shahar E, et al. (2000) Carotid wall thickness is predictive of incident clinical stroke: the Atherosclerosis Risk in Communities (ARIC) study. Am J Epidemiol 151: 478–487.
- 23. Felig P, Marliss E, Cahill GF Jr (1969) Plasma amino acid levels and insulin secretion in obesity. N Engl J Med 281: 811–816.