arachidonic acid ratio, and arterial stiffness in overweight subjects

Prostaglandins, Leukotrienes and Essential Fatty Acids 130 (2018) 11–18

Contents lists available at ScienceDirect

Prostaglandins, Leukotrienes and Essential Fatty Acids journal homepage: www.elsevier.com/locate/plefa

Associations among FADS1 rs174547, eicosapentaenoic acid/arachidonic acid ratio, and arterial stiffness in overweight subjects M. Kima,1, M. Kima,2, H.J. Yooc, A. Leeb,c, S. Jeongb,c, J.H. Leea,b,c,

T

⁎

a

Research Center for Silver Science, Institute of Symbiotic Life-TECH, Yonsei University, Seoul, Republic of Korea National Leading Research Laboratory of Clinical Nutrigenetics/Nutrigenomics, Department of Food and Nutrition, College of Human Ecology, Yonsei University, Seoul, Republic of Korea c Department of Food and Nutrition, Brain Korea 21 PLUS Project, College of Human Ecology, Yonsei University, Seoul, Republic of Korea b

A R T I C L E I N F O

A B S T R A C T

Keywords: Ba-PWV Fatty acid desaturases Overweight Polymorphism Arterial stiffness

We aimed to evaluate the longitudinal interaction effects between the minor allele of FADS1 rs174547 and overweight on n-3 and n-6 long-chain polyunsaturated fatty acid (PUFA) levels and pulse wave velocity (PWV). Plasma PUFA levels were measured via GC-MS, and arterial stiffness was determined as brachial-ankle PWV (baPWV) at baseline and after a mean follow-up of 3 years. The FADS1 rs174547 T > C genotype was analyzed. At 3-years of follow-up, after adjustment for age, sex, smoking and drinking, there were interaction effects between the FADS1 rs174547 T > C genotype and baseline BMI on the changes (from baseline) in plasma arachidonic acid (AA) levels, in the eicosapentaenoic acid (EPA)/AA ratio, and in ba-PWV (p for interaction = 0.036, 0.022, and 0.001, respectively). There were smaller increases in AA levels from baseline among normal-weight C allele carriers (n = 112) and overweight TT subjects (n = 47) than among normal-weight TT subjects (n = 91). Overweight C allele carriers (n = 37) showed greater reductions in the plasma EPA/AA ratio and greater increases in ba-PWV than the 3 other populations studied. The minor allele of the FADS1 rs174547 polymorphism is associated with age-related decreases in the EPA/AA ratio and increases in ba-PWV among overweight subjects.

1. Introduction Inter-individual variability in human lipid metabolism can be attributed to a combination of lifestyle factors (e.g., diet or smoking) and genetic variation. Fatty acid desaturases (FADSs) are a cluster of enzymes encoded by FADS genes. FADS1, or delta-5-desaturase (D5D), is a rate-limiting enzyme that participates in the endogenous metabolism of n-3 and n-6 long-chain polyunsaturated fatty acids (PUFAs; LCPUFAs) – 20:4 n-6 (arachidonic acid, AA) and 20:5 n-3 (eicosapentaenoic acid, EPA) – by introducing a double bond at the delta-5 position of 20carbon fatty acids – 20:3 n-6 (dihomo-γ-linolenic acid, DGLA) and 20:4 n-3 (eicosatetraenoic acid, ETA), respectively [1,2]. Variation across the FADS gene region appears to be important in modulating long-chain PUFA status. Plasma and tissue LCPUFA concentrations are associated with the risk of several diet-related chronic disease, including cardiovascular disease (CVD) [3,4]. The majority of studies to date suggest that FADS minor alleles, which are related to decreased desaturase activity, are

associated with reduced inflammation, total cholesterol, LDL-cholesterol and coronary artery disease risk [5–7]. However, a few studies have reported contradictory results [2,8,9], which could be due to the ethnicity of the participants or differences in the n-6:n-3 PUFA content of the habitual diet. For example, two studies carried out in ChineseHan population reported the frequency of the rs174556 minor allele to be significantly higher in cases of both coronary artery disease and acute coronary syndrome compared with control groups [8,9]. FADS1 is located on chromosome 11, and the single nucleotide polymorphism (SNP) rs174547, which is located in intron 9 of FADS1, modulates the expression of FADS1 (or D5D) [10,11]. The rs174547 SNP has been reported consistently its association with PUFAs in GWAS [12–14], and is in perfect LD (r2 = 1) with rs174546, which has been associated with triglycerides, LDL-cholesterol and HDL-cholesterol in GWAS [15,16]. Previous studies associated the FADS1 rs174547 genotype with lipids level in Chinese population. For instance, a Chinese study reported that the CC variant of rs174547 was significantly associated with increased TG and decreased HDL-cholesterol [11].

⁎

Corresponding author at: Department of Food and Nutrition, College of Human Ecology, Yonsei University, 50 Yonsei-ro, Seodaemun-gu, Seoul 03722, Republic of Korea. E-mail address: [email protected] (J.H. Lee). Minjoo Kim. 2 Minkyung Kim. 1

https://doi.org/10.1016/j.plefa.2018.02.004 Received 15 March 2017; Received in revised form 29 December 2017; Accepted 14 February 2018 0952-3278/ © 2018 Published by Elsevier Ltd.

Prostaglandins, Leukotrienes and Essential Fatty Acids 130 (2018) 11–18

M. Kim et al.

2. Material and methods

1101, Hico Medical Co., Ltd., Chiba, Japan) with appropriately sized cuffs after a rest period of at least 20 min in the seated position. BP was measured in both arms, and the higher of the two measurements was recorded. Three BP measurements were obtained at each visit, and the differences between the three systolic BP measurements were always < 2 mmHg. The average values of the systolic and diastolic BP measurements were used. Participants were instructed not to smoke or drink alcohol for at least 30 min before each BP measurement. Ba-PWV was measured using an automatic waveform analyzer (model VP-1000; Nippon Colin, Komaki, Japan) in the supine position after 5 min of bed rest. Electrocardiogram electrodes were placed on both wrists, and a microphone for the phonogram was placed on the left edge of the sternum. Pneumonic cuffs were wrapped around both upper arms and ankles and connected to a plethysmographic sensor to determine the volume pulse waveform. Waveforms for the upper arm (brachial artery) and ankle (tibial artery) were stored for 10-s sample times with automatic gain analysis and quality adjustment. An oscillometric pressure sensor was attached to the cuffs to measure blood pressure at the 4 extremities. The baPWVs were recorded using a semiconductor pressure sensor (1200 Hz sample acquisition frequency) and calculated using the equation (La − Lb)/ΔTba. La and Lb were defined as the distance from the aortic valve to the elbow and to the ankle, respectively. The distance from the suprasternal notch to the elbow (La) and that from the suprasternal notch to the ankle (Lb) were expressed by La = [0.2195 × height of subject (in centimeters)] −2.0734 and Lb = [0.8129 × height of subject (in centimeters)] + 12.328. The time interval between arm and ankle distance (ΔTba) was defined as the pulse transit time between brachial and tibial arterial pressure waveforms. The average baPWVs from both the left and right sides were used for the analysis (correlation between the right and left baPWVs: r2 = 0.900, p < 0.001).

2.1. Study subjects

2.4. Clinical and biochemical assessments

Study participants were recruited from a 3-year prospective cohort study that included 800 healthy subjects at the Health Promotion Center of Ilsan Hospital during routine check-up visits between March 2007 and November 2011. Based on the data obtained from the Health Promotion Center, subjects who met the study criteria and agreed to participate were referred to the Department of Family Medicine. The potential subjects’ health was reassessed, and the subjects who met the study criteria were then recommended to participate. A total of 287 participants were ultimately selected according to the study criteria. The clinical and blood tests were performed again at the baseline visit. The exclusion criteria were current and/or history of hypertension, cardiovascular disease, diabetes mellitus, dyslipidemia, liver disease, renal disease, pancreatitis, or cancer; pregnancy or lactation; and regular use of any medication. The aim of the study was carefully explained to all the participants, who provided their written informed consent. The Institutional Review Boards of Yonsei University and Ilsan Hospital approved the study protocol, which complied with the Declaration of Helsinki.

Detailed information about the clinical and biochemical assessments is provided elsewhere [27]. Body weight and height were measured, and BMI was calculated in units of kilograms per square meter (kg/m2). Blood samples were collected following an overnight fast of at least 12 h. Fasting triglycerides; total, HDL, and LDL-cholesterol; glucose; and insulin levels were measured as previously described [27]. Insulin resistance (IR) was determined via homeostasis model assessment (HOMA) using the following equation: HOMA-IR = [fasting insulin level (μIU/mL) × fasting glucose level (mmol/L)]/22.5.

Arterial stiffness, a marker of early vascular aging, can reflect changes in mechanical wall properties that occur early at the onset of vascular disease and in a predisposing factor for the occurrence of CVD [17,18]. Increased arterial stiffness is recognized as an independent risk factor for CVD, and its measurement may become an important routine assessment for patients in daily practice [19]. As a measurement of arterial stiffness, PWV has been shown both in clinically healthy cohorts and in CVD cohorts to be a predictor of future cardiovascular events, independent of the effects of blood pressure [20]. Certain traditional risk factors beyond ageing have been associated with arterial stiffness including insulin resistance/metabolic syndrome [21,22]. Several studies have reported alterations in fuel-metabolism related markers both in patients with clinically significant arterial stiffness [22–24]. A recent study has shown that plasma total n-3 PUFA, EPA and docosahexaenoic acid (DHA) levels are associated with reduced carotidfemoral pulse wave velocity (PWV; cf-PWV) but that plasma total n-6 PUFA and linoleic acid (LA) levels are associated with higher cf-PWV. However, the association of plasma n-3 and n-6 PUFA levels with arterial stiffness could differ according to the genotype of FADS1, such as the SNP rs174547, a dominant SNP in the FADS gene cluster that influences desaturase activity [25]. In addition to this, it is not clear whether the polymorphisms in fatty acid related genes are associated with obesity. A previous study of a Chinese population failed to associate FADS1 rs174547 with BMI [26], but as yet no studies have focused on body composition, especially body fat and its distribution. Therefore, we aimed to evaluate the longitudinal interaction effects between the minor allele of FADS1 rs174547 and body mass index (BMI) on LCPUFA levels and arterial stiffness, as measured by brachialankle PWV (ba-PWV).

2.5. Fatty acid concentrations based on GC-MS (gas chromatography-mass spectrometry)

Genomic DNA was extracted from 5 mL of whole blood using a commercially available DNA isolation kit (WIZARD® Genomic DNA Purification Kit, Promega Corp., Madison, WI) according to the manufacturer's protocol. Genotyping of FADS1 rs174547 was performed via a single-base primer extension assay using the SNaPShot assay kit (Applied Biosystems Inc., Foster City, CA) according to the manufacturer's protocol.

The details of GC-MS have been published previously [27]. Briefly, all analyses were performed using an Agilent Technologies 7890N gas chromatograph coupled to an Agilent Technologies 5977A quadrupole mass selective spectrometer equipped with a triple-axis detector (Agilent, Palo Alto, CA) in the electron ionization mode (70 eV) and the full scan monitoring mode (m/z 50–800). Derivatized samples were separated on a VF-WAX column (Agilent Technologies, Middelburg, Netherlands) using helium as the carrier gas and using a temperature ramp from 50 °C to 230 °C. Metabolites in the samples were identified by comparing their relative retention times and mass spectra with those of authentic reference standards. The relative levels of the metabolites were calculated by comparing their peak areas to the peak area of the corresponding internal standard compound. D5D activity was estimated as the ratio of AA to DGLA, and delta-6-desaturase (D6D) activity was estimated as the ratio of γ-linolenic acid (GLA) to LA.

2.3. Blood pressure (BP) and ba-PWV

2.6. Statistical analysis

2.2. Genotyping of FADS1 rs174547

Statistical analyses were performed using SPSS version 21.0 (IBM/

BP was measured using a random-zero sphygmomanometer (HM12

Prostaglandins, Leukotrienes and Essential Fatty Acids 130 (2018) 11–18

M. Kim et al.

rs174547 T > C and baseline BMI on the changes in the EPA/AA ratio (p for interaction = 0.022) and in ba-PWV (p for interaction = 0.001). Overweight C allele carriers showed greater reductions in the plasma EPA/AA ratio and greater increases in ba-PWV than normal-weight C allele carriers, normal-weight TT subjects, and overweight TT subjects (Fig. 1).

SPSS, Chicago, IL). Hardy-Weinberg equilibrium was assessed using PLINK version 1.07 (http://pngu.mgh.harvard.edu/purcell/plink/). Differences in clinical variables between two groups (normal weight vs. overweight; TT vs. C allele) were tested using independent t-tests. Paired t-tests were performed to determine the differences between the baseline and 3-year follow-up values within each group. The interactions between the FADS1 genotype and baseline BMI were tested using two-way analysis of variance. Multiple linear regression analyses were performed to identify major independent predictors of changes in baPWV. Pearson's correlation coefficients were calculated to examine the relationships between variables. Heat maps were generated to visualize and evaluate the relationships between metabolites and the biochemical parameters in the study population. Logarithmic transformations were performed on skewed variables. The results are expressed as the means ± standard error (SE), and a two-tailed p-value < 0.05 was considered statistically significant.

3.3. Clinical characteristics and biochemical parameters according to the FADS1 rs174547 T > C genotype at baseline and the 3-year follow-up in the normal-weight and overweight groups

We divided the cohort into 2 groups: the normal-weight group (18.5 kg/m2 ≤ BMI < 25 kg/m2, n = 203) and the overweight group (25 kg/m2 ≤ BMI < 30 kg/m2, n = 84). The distribution of the FADS1 rs174547 T > C genotype among the 203 normal-weight subjects was as follows: 91 subjects were homozygous for the T allele (TT), 89 were heterozygous for the C allele (TC), and 23 were homozygous for the C allele (CC). The distribution of the FADS1 rs174547 T > C genotype among the 84 overweight subjects was as follows: 47 were TT, 29 were TC, and 8 were CC. These frequencies did not deviate significantly from Hardy-Weinberg equilibrium (p > 0.05). The minor allele frequency in all subjects was 0.314. We pooled the heterozygotes (TC) and the rare allele homozygotes (CC) to increase the statistical power of the results.

There were no significant differences in baseline age or gender distribution between the normal-weight and overweight groups across the FADS1 rs174547 genotypes (Table 1). Additionally, there were no significant differences in smoking or drinking status between the normal-weight and overweight groups across genotypes at baseline and at the 3-year follow-up (data not shown). After 3 years, overweight C allele carriers showed significant increases in systolic and diastolic BP compared to baseline (Table 1). The systolic and diastolic BPs of overweight TT subjects at both baseline and the 3-year follow-up and those of overweight C allele carriers at the 3year follow-up were greater than those of their respective normalweight counterparts. With respect to serum lipid profiles, after 3 years, normal-weight C allele carriers showed increases in serum triglyceride levels and normalweight TT subjects and C allele carriers showed increases in total and LDL cholesterol levels compared to baseline. At baseline, overweight C allele subjects showed higher total and LDL cholesterol levels than normal-weight C allele carriers. At both baseline and follow-up, normalweight C allele carriers showed lower LDL cholesterol levels than normal-weight TT subjects. The overweight group showed higher HOMA-IR and triglyceride, glucose, and insulin levels as well as lower HDL cholesterol levels than the normal-weight group irrespective of genotype at both baseline and follow-up (Table 1).

3.2. D5D activity, plasma AA levels, the EPA/AA ratio and ba-PWV according to the FADS1 rs174547 T > C genotype at baseline and at the 3year follow-up in the normal-weight and overweight groups

3.4. Plasma fatty acid concentrations according to the FADS1 rs174547 T > C genotype at baseline and at the 3-year follow-up in the normalweight and overweight groups

In the normal-weight group, there was an effect of FADS1 rs174547 T > C on D5D activity at baseline and at the 3-year follow-up; normalweight individuals harboring the C allele had lower D5D activity than those harboring the TT genotype (Fig. 1). A genotype effect on D5D activity was also found in overweight subjects at baseline; overweight individuals with the C allele had lower D5D activity than those with the TT genotype. With regard to AA levels, at baseline, normal-weight and overweight C allele carriers showed lower AA levels than their respective TT counterparts, and overweight TT subjects showed higher AA levels than normal-weight TT subjects. After 3 years, the plasma AA levels were increased from baseline in normal-weight TT subjects and in normalweight and overweight C allele carriers. At the 3-year follow-up, normal-weight C allele carriers showed lower AA levels than normalweight TT subjects. At the 3-year follow-up, after adjustment for age, sex, smoking and drinking, there was a significant interaction effect between FADS1 rs174547 T > C and baseline BMI on the changes (difference from baseline) in plasma AA levels (p for interaction = 0.036). There were smaller increases in AA levels from baseline among normal-weight C allele carriers and overweight TT subjects than among normal-weight TT subjects. After 3 years, there was a decrease in the EPA/AA ratio and an increase in ba-PWV among overweight C allele carriers compared to baseline (Fig. 1). At follow-up, ba-PWV was significantly higher in overweight C allele carriers than in normal-weight C allele carriers. At the 3-year follow-up, after adjustment for age, sex, smoking and drinking, there were significant interaction effects between FADS1

Table 2 shows the plasma n-6 and n-3 fatty acid concentrations, D6D activity and the AA/LA ratio. At baseline, the plasma LA concentrations were higher in normal-weight C allele carriers than in normal-weight TT subjects. At the 3-year follow-up, the plasma LA concentrations were increased in all four groups compared to baseline. The ALA levels at baseline were higher in overweight C allele carriers than in normal-weight C-allele carriers. Furthermore, the plasma ALA concentrations were increased in normal-weight TT subjects as well as in normal-weight and overweight C allele carriers at the 3-year followup compared to baseline. There was a greater increase in ALA levels in normal-weight C allele carriers than in normal-weight TT subjects (p for the interaction effect between FADS1 rs174547 T > C and baseline BMI on changes in ALA levels= 0.048). At follow-up, the ALA level in overweight C allele carriers was higher than that in overweight TT subjects. At both baseline and follow-up, normal-weight and overweight C allele carriers showed lower GLA levels, D6D activity levels and AA/LA ratios than normal-weight and overweight TT subjects, respectively. Overweight C allele carriers showed higher baseline GLA levels, D6D activity levels and AA/LA ratios than normal-weight C allele carriers. After 3 years, compared to baseline, normal-weight C allele carriers as well as normal-weight and overweight TT subjects showed increases in GLA levels, and normal-weight C allele carriers and overweight TT subjects showed increases in D6D activity. At baseline, overweight C allele carriers showed higher EPA, DPA and DHA levels than normal-weight C-allele carriers. After 3 years, normal-weight TT subjects and C allele carriers showed increased EPA,

3. Results 3.1. Frequency of the FADS1 rs174547 T > C polymorphism in normalweight subjects and overweight subjects

13

Prostaglandins, Leukotrienes and Essential Fatty Acids 130 (2018) 11–18

M. Kim et al.

Fig. 1. Effect of the rs174547 genotype on changes in delta-5-desaturase activity, arachidonic acid levels, the EPA/AA ratio, and ba-PWV in the normal-weight and overweight groups at the 3-year follow-up compared with baseline. Mean ± SE. ∮tested by logarithmic transformation. ap < 0.05 comparison between individuals with the TT genotype and C allele in the normal weight group at baseline. bp < 0.05 comparison between individuals with the TT genotype and C allele in the normal weight group at the 3-year follow-up. cp < 0.05 comparison between individuals with the TT genotype and C allele in the overweight group at baseline. dp < 0.05 comparison between individuals with the TT genotype and C allele in the overweight group at the 3-year follow-up. ep < 0.05 comparison of individuals with the TT genotype between the normal weight and overweight groups at baseline. fp < 0.05 comparison of individuals with the TT genotype between the normal weight and overweight groups at the 3-year follow-up. gp < 0.05 comparison of individuals with the C allele between the normal weight and overweight groups at baseline. hp < 0.05 comparison of individuals with the C allele between the normal weight and overweight groups at the 3-year follow-up. *p < 0.05, **p < 0.01, ***p < 0.001 compared with the levels at baseline in each group according to the paired t-test.

normal-weight and overweight TT subjects exhibited greater increases in EPA and DPA levels than overweight C allele carriers.

DPA and DHA levels compared to baseline, and overweight TT subjects also showed increases in DPA and DHA levels compared to baseline. At the 3-year follow-up, overweight TT subjects showed higher DPA levels than normal-weight TT subjects. At the 3-year follow-up, after adjustment for age, sex, smoking and drinking, there were significant interaction effects between FADS1 rs174547 T > C and baseline BMI on the changes in EPA levels (p for interaction= 0.038) and in DPA levels (p for interaction= 0.014). Normal-weight C allele carriers as well as

3.5. Correlation among changes in ba-PWV, clinical characteristics, D5D activity, D6D activity, AA/LA ratio and EPA/AA ratio Fig. 2 shows a correlation matrix for the changes in ba-PWV, clinical characteristics, D5D activity, D6D activity, AA/LA ratio and EPA/AA 14

Prostaglandins, Leukotrienes and Essential Fatty Acids 130 (2018) 11–18

M. Kim et al.

Table 1 Association of rs174547 genotypes with clinical and biochemical characteristics at baseline and 3-year follow-up according to BMI. Normal weight (n = 203) TT (n = 91) Baseline Age (year) Male/Female n, (%) BMI (kg/m2) Systolic BP (mmHg) Diastolic BP (mmHg) Triglyceride (mg/dL)∮ Total-cholesterol (mg/dL)∮ HDL-cholesterol (mg/dL)∮ LDL-cholesterol (mg/dL)∮ Glucose (mg/dL)∮ Insulin (μIU/dL)∮ HOMA-IR∮

Follow-up

44.9 ± 0.98 32 (35.2) / 59 (64.8) 21.9 ± 0.19 22.0 ± 0.20 113.2 ± 1.20 114.5 ± 1.45 69.4 ± 1.03 70.2 ± 1.04 92.5 ± 6.05 95.7 ± 6.83 187.8 ± 3.34 198.4 ± 3.63*** 54.9 ± 1.48 54.6 ± 1.44 114.4 ± 2.83 124.7 ± 2.96*** 90.3 ± 1.03 91.0 ± 1.17 8.36 ± 0.41 7.43 ± 0.27 1.89 ± 0.10 1.69 ± 0.07

Overweight (n= 84) C allele (n = 112)

TT (n = 47)

Baseline

Baseline

Follow-up

44.2 ± 0.82 42 (37.5) / 70 (62.5) 21.9 ± 0.17 22.0 ± 0.20 114.9 ± 1.36 114.9 ± 1.27 70.3 ± 1.03 71.2 ± 0.95 77.4 ± 3.54 92.9 ± 5.59** 179.7 ± 2.79 190.0 ± 4.18** 56.6 ± 1.36 55.1 ± 1.20 107.6 ± 2.82a 116.3 ± 3.86b,** 88.7 ± 0.72 89.3 ± 0.82 7.28 ± 0.22 7.13 ± 0.25 1.59 ± 0.05 1.58 ± 0.06

C allele (n = 37) Follow-up

44.5 ± 1.40 18 (38.3) / 29 (61.7) 26.8 ± 0.22c 26.7 ± 0.25d 122.4 ± 1.84c 124.0 ± 1.67d 75.4 ± 1.36c 75.2 ± 1.37d 131.2 ± 11.3c 152.7 ± 13.8d 197.2 ± 4.10 201.0 ± 4.26 49.6 ± 1.60c 48.4 ± 1.70d 121.4 ± 4.25 122.1 ± 4.26 96.7 ± 1.48c 98.2 ± 2.19d 9.83 ± 0.54c 8.96 ± 0.46d 2.35 ± 0.14c 2.21 ± 0.14d

Baseline

Follow-up

45.1 ± 1.56 15 (40.5) / 22 (59.5) 26.8 ± 0.33e 27.1 ± 0.42f 118.3 ± 2.12 122.4 ± 2.17f,* 73.2 ± 1.69 77.1 ± 1.68f,* 124.5 ± 10.8e 138.6 ± 13.8f 196.2 ± 5.47e 202.0 ± 5.92 50.6 ± 2.81e 50.0 ± 2.17f e 121.3 ± 5.30 124.3 ± 5.64 95.4 ± 1.23e 98.3 ± 1.99f 9.59 ± 0.55e 9.76 ± 0.70f 2.26 ± 0.13e 2.45 ± 0.23f

Mean ± SE. ∮ tested by logarithmic transformation. a p < 0.05 comparison between individuals with the TT genotype and C allele in the normal weight group at baseline. b p < 0.05 comparison between individuals with the TT genotype and C allele in the normal weight group at the 3-year follow-up. c p < 0.05 comparison of individuals with the TT genotype between the normal weight and overweight groups at baseline. d p < 0.05 comparison of individuals with the TT genotype between the normal weight and overweight groups at the 3-year follow-up. e p < 0.05 comparison of individuals with the C allele between the normal weight and overweight groups at baseline. f p < 0.05 comparison of individuals with the C allele between the normal weight and overweight groups at the 3-year follow-up. * p < 0.05. ** p < 0.01. *** p < 0.001 compared with the levels at baseline in each group according to the paired t-test.

4. Discussion

ratio according to the FADS1 rs174547 T > C genotype and baseline BMI. In all 4 groups, changes (△) in systolic BP positively correlated with △diastolic BP; △total-cholesterol positively correlated with △LDL-cholesterol; and △D5D negatively correlated with △D6D. Additionally, in normal-weight TT subjects, △triglyceride positively correlated with △total-cholesterol, △HOMA-IR, and △D6D and negatively correlated with △D5D. Changes in ba-PWV positively correlated with △AA/LA and negatively correlated with △EPA/AA (Fig. 3). Changes in AA/LA positively correlated with △D5D. In normal-weight C-allele subjects, △triglyceride positively correlated with △totalcholesterol, △HOMA-IR, and △D6D and negatively correlated with △D5D. Changes in HDL cholesterol negatively correlated with △LDL cholesterol and positively correlated with △D5D. Changes in ba-PWV positively correlated with △HOMA-IR. In overweight TT subjects, △triglyceride positively correlated with △D6D and negatively correlated with △HDL cholesterol and △D5D. Changes in HDL cholesterol positively correlated with △D5D and △AA/LA. Changes in ba-PWV positively correlated with △AA/LA (Fig. 3). Changes in D5D positively correlated with △AA/LA and △EPA/AA. In overweight C allele carriers, △triglyceride positively correlated with △HOMA-IR and negatively correlated with △D5D and △AA/LA. Changes in HOMA-IR negatively correlated with △AA/LA. Changes in ba-PWV negatively correlated with △EPA/AA (Fig. 3). Changes in D5D positively correlated with △AA/LA (Fig. 2). Because △ba-PWV is a complex parameter that depends on the FADS1 genotype and baseline BMI, a multiple linear regression analysis was performed to determine the independent variables related to △baPWV in all subjects; the following variables were examined: age, gender, FADS1 rs174547 T > C genotype, baseline BMI, △triglyceride, △HDL cholesterol, △D5D, △D6D, △AA/LA, and △EPA/AA. Changes in ba-PWV were significantly independently associated with the FADS1 rs174547 T > C genotype (standardized β = 0.154, p = 0.008), baseline BMI (β = 0.118, p = 0.049), △AA/LA (β = 0.167, p = 0.010), and △EPA/AA (β =−0.197, p = 0.001).

This prospective study evaluated the longitudinal interaction effects between the minor allele of FADS1 rs174547 and overweight on plasma LCPUFA levels and the rate of progression of arterial stiffness. LCPUFA levels and ba-PWV, which reflects arterial stiffness [28,29], were measured at baseline and were then re-examined after a mean followup period of 3 years. The main finding of this study is that the minor allele of FADS1 rs174547 is associated with age-related decreases in the EPA/AA ratio and increases in ba-PWV among overweight subjects. Additionally, multiple regression analysis revealed that the changes in ba-PWV were significantly and independently associated with the FADS1 rs174547 T > C genotype, baseline BMI, and the changes in the AA/LA and EPA/AA ratios. Recently, Ito et al. [30] have found that EPA reduces arterial stiffness in association with an increase in the EPA/AA ratio, independent of other lipid biomarkers. Furthermore, they suggested that the change in the EPA/AA ratio could be the most useful biomarker for predicting the effectiveness of EPA in reducing arterial stiffness. Additionally, there were reports of significantly negative associations between the serum EPA/AA ratio and the incidence of cardiac death or myocardial infarction as well as the coronary plaque score [31,32]. In this study, we also found a negative correlation between changes in the EPA/AA ratio and ba-PWV in normal-weight TT subjects and overweight C allele carriers. This observation could be due to the smaller increases in AA levels from baseline among normal-weight C allele carriers and overweight TT subjects than among normal-weight TT subjects as well as the greater reductions in plasma EPA levels among overweight C allele carriers than in the other three groups studied. Therefore, the present result of a negative association between the EPA/AA ratio and ba-PWV only in specific groups could be attributed to a combined effect of BMI and genetic variations, specifically the FADS1 rs174547 polymorphism, on the EPA/AA ratio. Although ba-PWV may not directly reflect arteriosclerosis, the long-term maintenance of a high EPA/AA ratio is beneficial to the vascular system, especially in overweight C allele carriers, who showed greater age-related increases in ba-PWV than the other populations studied. 15

Prostaglandins, Leukotrienes and Essential Fatty Acids 130 (2018) 11–18

M. Kim et al.

Table 2 Association of rs174547 genotypes with plasma fatty acids levels at baseline and 3-year follow-up according to BMI. Fatty acids (Relative peak area)

Normal weight (n = 203)

Overweight (n = 84)

TT (n = 91) Baseline Linoleic acid (C18:2, n-6) α-Linolenic acid (C18:3, n-3) Change γ-Linolenic acid (C18:3, n-6) Dihomo-γ-linolenic acid (C20:3, n-6) Eicosapentaenoic acid (C20:5, n-3) Change Docosapentaenoic acid (C22:5, n-3) Change Docosahexaenoic acid (C22:5, n-3) Δ6-Desaturase Arachidonic acid/linoleic acid

C allele (n = 112) Follow-up

Baseline

Follow-up

Baseline

Follow-up

1.959 ± 0.082 0.146 ± 0.008g 0.037 0.023 ± 0.002c,g 0.107 ± 0.006

2.283 ± 0.108** 0.183 ± 0.015d,** ± 0.013 0.024 ± 0.002d 0.116 ± 0.007

0.138 ± 0.009 0.171 ± 0.011*** 0.134 ± 0.006

0.158 ± 0.008**

0.172 ± 0.017

0.209 ± 0.022

0.188 ± 0.019g

0.167 ± 0.017

0.034 ± 0.009 0.024 0.058 ± 0.003 0.068 ± 0.003*** 0.055 ± 0.002

± 0.008 0.064 ± 0.002***

0.038 0.065 ± 0.004

± 0.024 −0.020 0.083 ± 0.005f,*** 0.072 ± 0.004g

± 0.018i,j,k 0.073 ± 0.005

0.010 ± 0.002 0.009 0.334 ± 0.013 0.379 ± 0.015*** 0.340 ± 0.010

± 0.002 0.372 ± 0.011***

0.017 0.381 ± 0.022

± 0.004 0.434 ± 0.028*

± 0.004i,j,k 0.414 ± 0.023

0.014 ± 0.001 0.015 ± 0.001 0.271 ± 0.005 0.276 ± 0.005

*

Baseline

2.056 ± 0.084 0.147 ± 0.009 ± 0.009 0.036 ± 0.003** 0.117 ± 0.007

1.993 ± 0.050 0.124 ± 0.005 0.029 0.016 ± 0.001a 0.098 ± 0.003

***

Follow-up

1.916 ± 0.082 0.136 ± 0.008 0.011 0.028 ± 0.002 0.105 ± 0.005

2.089 ± 0.070 0.147 ± 0.005* ± 0.006 0.032 ± 0.002** 0.107 ± 0.004***

a

C allele (n = 37)

2.216 ± 0.062 0.153 ± 0.005*** ± 0.005h 0.023 ± 0.001b,*** 0.110 ± 0.004***

1.821 ± 0.060 0.135 ± 0.005 0.012 0.026 ± 0.002 0.094 ± 0.004

***

TT (n = 47)

0.008 ± 0.000a 0.010 ± 0.000b,*** 0.015 ± 0.001 0.017 ± 0.001* 0.222 ± 0.004a 0.225 ± 0.005b 0.295 ± 0.010e 0.293 ± 0.010

0.001 0.411 ± 0.024g

0.012 ± 0.001c,g 0.011 ± 0.001d 0.242 ± 0.007c,g 0.240 ± 0.009d

Mean ± SE. a p < 0.05 comparison between individuals with the TT genotype and C allele in the normal weight group at baseline. b p < 0.05 comparison between individuals with the TT genotype and C allele in the normal weight group at the 3-year follow-up. c p < 0.05 comparison between individuals with the TT genotype and C allele in the overweight group at baseline. d p < 0.05 comparison between individuals with the TT genotype and C allele in the overweight group at the 3-year follow-up. e p < 0.05 comparison of individuals with the TT genotype between the normal weight and overweight groups at baseline. f p < 0.05 comparison of individuals with the TT genotype between the normal weight and overweight groups at the 3-year follow-up. g p < 0.05 comparison of individuals with the C allele between the normal weight and overweight groups at baseline. h p < 0.05 comparison between individuals with the TT genotype and C allele in the normal weight group at change values. i p < 0.05 comparison between individuals with the TT genotype and C allele in the overweight group at change values. j p < 0.05 comparison between the normal weight and overweight groups among individuals with the C allele at change values. k p < 0.05 comparison between individuals with the TT genotype in the normal weight group and the C allele in the overweight group at change values. * p < 0.05. ** p < 0.01. *** p < 0.001 compared with the levels at baseline in each group according to the paired t-test.

higher serum n-6 PUFA concentrations and markers indicative of elevated arterial stiffness, such as increased ba-PWV [36] and increased cfPWV [37]. The ratio of metabolic precursor to product, such as the AA/ LA ratio, has been used as a surrogate of fatty acid desaturation activity

Inverse associations between n-3 LCPUFA levels and arterial stiffness were previously reported in many studies [33,34], but little is known about the relationship between n-6 PUFA levels and arterial stiffness [35]. A few studies reported a positive association between

Fig. 2. Metabolic pathway analysis of all subjects (n = 68). Correlation matrix for the changes in clinical parameters, ba-PWV, D5D activity, D6D activity, AA/LA ratio, and EPA/AA ratio in the normal-weight and overweight groups according to the rs174547 genotype. Correlations were obtained by calculating Pearson's correlation coefficients. Red indicates a positive correlation, and blue indicates a negative correlation (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.).

16

Prostaglandins, Leukotrienes and Essential Fatty Acids 130 (2018) 11–18

M. Kim et al.

Fig. 3. Correlation between the changes (Δ) in ba-PWV and the EPA/AA ratio as well as between the changes in ba-PWV and the AA/LA ratio in the normal-weight (○) and overweight (*) groups according to the rs174547 genotype. Correlation between the (a) changes in ba-PWV and the EPA/AA ratio for the TT genotype, (b) changes in ba-PWV and the AA/LA ratio for the TT genotype, (c) changes in ba-PWV and the EPA/AA ratio for the TC+CC genotype, (d) changes in ba-PWV and the AA/LA ratio for the TC+CC genotype.

enables the determination of their individual contributions to the agerelated changes in ba-PWV according to the FADS1 rs174547 genotype. These measurements also eliminate the potential biases associated with self-reporting of dietary information, as participants likely under-report their energy intake [39,40], which might lead to an underestimation of true associations. However, the limitations of the present study should be considered. The small sample size of this study was not conducive to the identification of weak associations or those that appear to be less common within the population due to low statistical power. Therefore, the results of this study need to be validated in a separate larger cohort of subjects.

[5,26,38]. In this study, the AA/LA ratio was significantly lower in C allele carriers than in TT subjects at both baseline and follow-up. It was hypothesized that rs174547 C allele carriers might have a lower rate of conversion of LA into AA than rs174547 TT genotype carriers. However, there was no significant difference in the changes in the AA/LA ratio over time across the 4 groups. This effect of the FADS1 rs174547 genotype on n-6 PUFA levels could have produced the positive correlation between the changes in the AA/LA ratio and ba-PWV among normal-weight and overweight TT subjects, although no significant association was observed between the changes in the AA/LA ratio and ba-PWV among overweight C allele carriers. Similar to the observations for the AA/LA ratio, the C allele carriers had lower GLA levels and D6D activity at both baseline and the 3-year follow-up than the TT subjects in this study. GLA is a fatty acid that is generated from LA via desaturation of LA by D6D. However, the GLA level and D6D activity were suggested to be associated with genetic polymorphisms of D5D. The SNP rs174547 has been shown to be linked with other SNPs such as rs174556 and rs174570 [10], which include SNPs related to D6D. Thus, it has been considered that the rs174547 genotype reflects the entire metabolic process of n-6 PUFA conversion from LA to AA, including conversion by D6D. Generally, AA intake is very small at approximately 100 mg/day, representing less than 1% of total energy intake [38]. Therefore, a considerable amount of the AA that is required by the body is supplied via desaturation and elongation of LA. This study has several strengths. First, we measured ba-PWV, which is considered to be a measurement of arterial stiffness. Second, determinations of the plasma n-3 and n-6 PUFA levels at both baseline and the 3-year follow-up have the advantage of identifying and quantifying age-related changes in the levels of individual PUFAs and therefore

5. Conclusions The results of this study show that the minor allele of the FADS1 rs174547 polymorphism is associated with age-related decreases in the EPA/AA ratio and increases in ba-PWV among overweight subjects. The present result of a negative association between the EPA/AA ratio and ba-PWV only among normal-weight TT subjects and overweight C allele carriers could be attributed to a combined effect of BMI and genetic variations, specifically the FADS1 rs174547 genotype, on the EPA/AA ratio. Therefore, long-term maintenance of a high EPA/AA ratio, including EPA supplementation, could be beneficial to the vascular system, especially in overweight C allele carriers, who showed greater age-related increases in ba-PWV than the other populations studied.

Conflict of interest The authors declare no conflict of interest. 17

Prostaglandins, Leukotrienes and Essential Fatty Acids 130 (2018) 11–18

M. Kim et al.

Funding This study was funded by the Bio-Synergy Research Project (NRF2012M3A9C4048762) and the Mid-career Researcher Program (NRF2016R1A2B4011662) of the Ministry of Science, ICT and Future Planning through the National Research Foundation, Republic of Korea.

[20]

[21]

[22]

References

[23] [1] M. Al-Hilal, A. Alsaleh, Z. Maniou, et al., Genetic variation at the FADS1-FADS2 gene locus influences delta-5 desaturase activity and LC-PUFA proportions after fish oil supplement, J. Lipid Res. 54 (2013) 542–551. [2] Y. Lu, A. Vaarhorst, A.H. Merry, et al., Markers of endogenous desaturase activity and risk of coronary heart disease in the CAREMA cohort study, PLoS One 7 (2012) e41681. [3] Q. Wang, X. Liang, L. Wang L, et al., Effect of omega-3 fatty acids supplementation on endothelial function: a meta-analysis of randomized controlled trials, Atherosclerosis 221 (2012) 536–543. [4] P.A. Dacks, D.W. Shineman, H.M. Fillit, Current evidence for the clinical use of long-chain polyunsaturated n-3 fatty acids to prevent age-related cognitive decline and Alzheimer's disease, J. Nutr. Health Aging 17 (2013) 240–251. [5] N. Martinelli, D. Girelli, G. Malerba, et al., FADS genotypes and desaturase activity estimated by the ratio of arachidonic acid to linoleic acid are associated with inflammation and coronary artery disease, Eur. J. Clin. Nutr. 88 (2008) 941–949. [6] Y.S. Aulchenko, S. Ripatti, I. Lindqvist, et al., Loci influencing lipid levels and coronary heart disease risk in 16 European population cohorts, Nat. Genet. 41 (2009) 47–55. [7] J.H. Kwak, J.K. Paik, O.Y. Kim, et al., FADS gene polymorphisms in Koreans: association with omega6 polyunsaturated fatty acids in serum phospholipids, lipid peroxides, and coronary artery disease, Atherosclerosis 214 (2011) 94–100. [8] L. Qin, L. Sun, L. Ye, et al., A case–control study between the gene polymorphisms of polyunsaturated fatty acids metabolic rate-limiting enzymes and coronary artery disease in a Chinese Han population, Prostaglandins Leukot. Essent. Fatty Acids 85 (2011) 329–333. [9] Z. Song, H. Cao, L. Qin, et al., A case–control study between gene polymorphisms of polyunsaturated fatty acid metabolic rate-limiting enzymes and acute coronary syndrome in Chinese Han population, BioMed Res. Int. 2013 (2013) 928178. [10] K. Nakayama, T. Bayasgalan, F. Tazoe, et al., A single nucleotide polymorphism in the FADS1/FADS2 gene is associated with plasma lipid profiles in two genetically similar Asian ethnic groups with distinctive differences in lifestyle, Hum. Genet. 127 (2010) 685–690. [11] S.J. Liu, H. Zhi, P.Z. Chen, et al., Fatty acid desaturase 1 polymorphisms are associated with coronary heart disease in a Chinese population, Chin. Med. J. 125 (2012) 801–806. [12] W. Guan, B.T. Steffen, R.N. Lemaitre, et al., Genome-wide association study of plasma N6 polyunsaturated fatty acids within the cohorts for heart and aging research in genomic epidemiology consortium, Circ. Cardiovasc. Genet. 7 (2014) 321–331. [13] R. Dorajoo, Y. Sun, Y. Han, et al., A genome-wide association study of n-3 and n-6 plasma fatty acids in a Singaporean Chinese population, Genes Nutr. 10 (2015) 53. [14] K. Suhre, S.Y. Shin, A.K. Petersen, et al., Human metabolic individuality in biomedical and pharmaceutical research, Nature 477 (2011) 54–60. [15] T.M. Teslovich, K. Musunuru, A.V. Smith, et al., Biological, clinical and population relevance of 95 loci for blood lipids, Nature 466 (2010) 707–713. [16] C.J. Willer, E.M. Schmidt, S. Sengupta, et al., Discovery and refinement of loci associated with lipid levels, Nat. Genet. 45 (2013) 1274–1283. [17] M. Doumas, V. Papademetriou, V. Athyros, et al., Arterial stiffness and emerging biomarkers: still a long journey to go, Angiology 66 (2015) 901–903. [18] M.F. O’Rourke, J.A. Staessen, C. Vlachopoulos, et al., Clinical applications of arterial stiffness; definitions and reference values, Am. J. Hypertens. 15 (2002) 426–444. [19] A. Dogui, N. Kachenoura, F. Frouin, et al., Consistency of aortic distensibility and

[24] [25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

18

pulse wave velocity estimates with respect to the Bramwell-Hill theoretical model: a cardiovascular magnetic resonance study, J. Cardiovasc. Magn. Reson. 13 (2011) 11. T. Willum-Hansen, J.A. Staessen, C. Torp-Pedersen, et al., Prognostic value of aortic pulse wave velocity as index of arterial stiffness in the general population, Circulation 113 (2006) 664–670. D.R. Webb, K. Khunti, R. Silverman, et al., Impact of metabolic indices on central artery stiffness: independent association of insulin resistance and glucose with aortic pulse wave velocity, Diabetologia 53 (2010) 1190–1198. M.E. Safar, F. Thomas, J. Blacher, et al., Metabolic syndrome and age-related progression of aortic stiffness, J. Am. Coll. Cardiol. 47 (2006) 72–75. M. Zagura, J. Kals, K. Kilk, et al., Metabolomic signature of arterial stiffness in male patients with peripheral arterial disease, Hypertens. Res. 38 (2015) 840–846. A. Mahmud, J. Feely, Adiponectin and arterial stiffness, Am. J. Hypertens. 18 (2005) 1543–1548. D.M. Merino, H. Johnston, S. Clarke, et al., Polymorphisms in FADS1 and FADS2 alter desaturase activity in young Caucasian and Asian adults, Mol. Genet. Metab. 103 (2011) 171–178. J.W. Wang, X. Tang, N. Li, et al., The impact of lipid-metabolizing genetic polymorphisms on body mass index and their interactions with soybean food intake: a study in a Chinese population, Biomed. Environ. Sci. 27 (2014) 176–185. H.Y. Yeo, O.Y. Kim, H.H. Lim, J.Y. Kim, J.H. Lee, Association of serum lycopene and brachial-ankle pulse wave velocity with metabolic syndrome, Metabolism 60 (2011) 537–543. M. Kim, M. Kim, Y.J. Lee, et al., Effects of α-linolenic acid supplementation in perilla oil on collagen-epinephrine closure time, activated partial thromboplastin time and Lp-PLA2 activity in non-diabetic and hypercholesterolaemic subjects, J. Funct. Foods 23 (2016) 95–104. C.S. Hung, J.W. Lin, C.N. Hsu, et al., Using brachial-ankle pulse wave velocity to associate arterial stiffness with cardiovascular risks, Nutr. Metab. Cardiovasc. Dis. 19 (2009) 241–246. H. Tomiyama, Y. Hirayama, H. Hashimoto, et al., The effects of changes in the metabolic syndrome detection status on arterial stiffening: a prospective study, Hypertens. Res. 29 (2006) 673–678. R. Ito, N. Satoh-Asahara, H. Yamakage, et al., An increase in the EPA/AA ratio is associated with improved arterial stiffness in obese patients with dyslipidemia, J. Atheroscler. Thromb. 21 (2014) 248–260. M. Matsuzaki, M. Yokoyama, Y. Saito, et al., Incremental effects of eicosapentaenoic acid on cardiovascular events in statin-treated patients with coronary artery disease, Circ. J. 73 (2009) 1283–1290. M. Ueeda, T. Doumei, Y. Takaya, et al., Serum N-3 polyunsaturated fatty acid levels correlate with the extent of coronary plaques and calcifications in patients with acute myocardial infarction, Circ. J. 72 (2008) 1836–1843. O.Y. Kim, H.H. Lim, M.J. Lee, J.Y. Kim, J.H. Lee, Association of fatty acid composition in serum phospholipids with metabolic syndrome and arterial stiffness, Nutr. Metab. Cardiovasc. Dis. 23 (2013) 366–374. K.D. Monahan, R.P. Feehan, C. Blaha, D.J. McLaughlin, Effect of omega-3 polyunsaturated fatty acid supplementation on central arterial stiffness and arterial wave reflections in young and older healthy adults, Physiol. Rep. 3 (2015) e12438. H. Tomiyama, C. Matsumoto, M. Odaira, et al., Relationships among the serum omega fatty acid levels, serum C-reactive protein levels and arterial stiffness/wave reflection in Japanese men, Atherosclerosis 217 (2011) 433–436. I. Reinders, R.A. Murphy, X. Song, et al., Higher plasma phospholipid n-3 pufas, but lower n-6 pufas, are associated with lower pulse wave velocity among older adults, J. Nutr. 145 (2015) 2317–2324. S. Horiguchi, K. Nakayama, S. Iwamoto, et al., Associations between a fatty acid desaturase gene polymorphism and blood arachidonic acid compositions in Japanese elderly, Prostaglandins Leukot. Essent. Fatty Acids 105 (2016) 9–14. A.F. Subar, V. Kipnis, R.P. Troiano, et al., Using intake biomarkers to evaluate the extent of dietary misreporting in a large sample of adults: the OPEN study, Am. J. Epidemiol. 158 (2003) 1–13. V. Kipnis, A.F. Subar, D. Midthune, et al., Structure of dietary measurement error: results of the OPEN biomarker study, Am. J. Epidemiol. 158 (2003) 14–21 (discussion 22-26).