Fatty acid amides as potential circulating biomarkers for sarcopenia

Abstract

Background

Sarcopenia is characterized by a progressive decrease in skeletal muscle mass and function with age. Given that sarcopenia is associated with various metabolic disorders, effective metabolic biomarkers for its early detection are required. We aimed to investigate the metabolic biomarkers related to sarcopenia in elderly men and perform experimental studies using metabolomics.

Methods

Plasma metabolites from 142 elderly men, comprising a sarcopenia group and an age‐matched control group, were measured using global metabolome profiling. Muscle and plasma samples from an aging mouse model of sarcopenia, as well as cell media and cell lysates during myoblast differentiation, were analysed based on targeted metabolome profiling. Based on these experimental results, fatty acid amides were quantified from human plasma as well as human muscle tissues. The association of fatty acid amide levels with sarcopenia parameters was evaluated.

Results

Global metabolome profiling showed that fatty acid amide levels were significantly different in the plasma of elderly men with sarcopenia (all Ps < 0.01). Consistent with these results in human plasma, targeted metabolome profiling in an aging mouse model of sarcopenia showed decreased levels of fatty acid amides in plasma but not in muscle tissue. In addition, the levels of fatty acid amides increased in cell lysates during muscle cell differentiation. Targeted metabolome profiling in men showed decreased docosahexaenoic acid ethanolamide (DHA EA) levels in the plasma (P = 0.016) but not in the muscle of men with sarcopenia. DHA EA level was positively correlated with sarcopenia parameters such as skeletal muscle mass index (SMI) and handgrip strength (HGS) (P = 0.001, P = 0.001, respectively). The area under the receiver‐operating characteristic curve (AUC) for DHA EA level ≤ 4.60 fmol/μL for sarcopenia was 0.618 (95% confidence interval [CI]: 0.532–0.698). DHA EA level ≤ 4.60 fmol/μL was associated with a significantly greater likelihood of sarcopenia (odds ratio [OR]: 2.11, 95% CI: 1.03–4.30), independent of HGS. The addition of DHA EA level to age and HGS significantly improved the AUC from 0.620 to 0.691 (P = 0.0497).

Conclusions

Our study demonstrated that fatty acid amides are potential circulating biomarkers in elderly men with sarcopenia. DHA EA, in particular, strongly related to muscle mass and strength, can be a key metabolite to become a reliable metabolic biomarker for sarcopenia. Further research on fatty acid amides will provide insights into the metabolomic changes relevant to sarcopenia from an aging perspective.

Introduction

Sarcopenia refers to the loss of skeletal muscle mass and strength as well as a decline in physical performance. ¹ Aging is the most common cause of primary sarcopenia, although various other causes, such as inflammation, sex hormone deficiency, and lack of exercise and nutrition, have also been reported. ² Sarcopenia is regarded as a significant public health concern owing to its association with motility disability, poor quality of life, and increased adverse health outcomes, as well as the frailty and healthcare expenditures. ³ To date, there is no approved treatment for sarcopenia, and the therapeutic effects of strength training and nutritional supplementation are limited after the disease has progressed significantly. ⁴ Therefore, it is important to detect sarcopenia at an early stage.

Measurements of muscle strength and mass using various modalities are key diagnostic features of sarcopenia. ⁵ , ⁶ However, they mainly reflect the skeletal muscle status as a static indicator at a somewhat advanced state. The skeletal muscle is the largest organ contributing to metabolic processes and serves as the principal reservoir for amino acids to maintain protein synthesis. ⁷ Changes in energy metabolism have been reported to directly contribute to skeletal muscle aging in the early stages of sarcopenia. ⁸ These findings suggest that metabolites may be used as early diagnostic biomarkers of sarcopenia. Furthermore, recent advances in liquid chromatography–tandem mass spectrometry (LC–MS/MS) allow direct insight into the metabolic state of the organism and offer important clues on the pathological conditions that cause altered concentrations of specific metabolites. ⁹

Recent studies on metabolomic biomarkers for sarcopenia have shown that plasma ceramides and lysophosphatidylcholine (LysoPC) are associated with gait speed according to the results of the Baltimore Longitudinal Study of Aging. ¹⁰ , ¹¹ In addition, metabolic biomarkers such as lower n‐3 fatty acids or genetic marker‐related to lipid metabolism such as NUDT3 were reported in Asian population studies. ¹² , ¹³ However, the interpretation of the results is limited by the lack of large scale‐replication studies and experimental data which may reveal mechanistic insights and provide valuable information for a better understanding of clinical outcomes.

Towards this goal, we conducted metabolomic studies designed in three phases: (1) discovery of sarcopenia‐related metabolic features in plasma of elderly men with sarcopenia and age‐matched controls using global metabolomics; (2) experimental studies with an aging mouse model of sarcopenia and muscle cells to confirm the target metabolites selected from the results of phase 1; and (3) targeted metabolomics in human plasma (n = 142) and human muscle tissue (n = 10) based on the results of phase 1 and phase 2 studies. Therefore, this study aimed to provide reliable metabolomic biomarkers for elderly men with sarcopenia.

Methods

Human cohort study

Study participants

This prospective case–control study was approved by the Institutional Review Board of Veterans Health Service Medical Center (IRB No. 2020‐02‐015) and Asan medical center (AMC, IRB No. 2017‐0553) and conducted in compliance with the Helsinki Declaration. Written informed consent was obtained from all participants before enrolment. For the discovery phase, participants aged ≥ 65 years were enrolled in the ‘Veterans Sarcopenia Study’. ¹³ They were Koreans who visited the Division of Endocrinology, Department of Internal Medicine, Veterans Health Service Medical Center (Seoul, Korea) to undergo comprehensive geriatric assessment between August 2020 and March 2021. Before the study, all participants completed questionnaires, including medical history, EuroQol Visual Analogue Scale (EQ‐VAS), SARC‐F (strength, assistance in walking, rising from a chair, climbing stairs, falls), muscle mass measurement, muscle strength test, and blood sampling. The prior process was similar but remnant human muscle tissue during orthopaedics surgery was obtained from participants in the AMC cohorts.

Assessment of sarcopenia

Body composition was evaluated using bioelectrical impedance analysis (InBody 570, Biospace Co., Seoul, Korea). The appendicular skeletal muscle mass (ASM) was calculated as the sum of the muscle mass of both arms and legs, and the skeletal muscle mass index (SMI) was defined by adjusting the ASM relative to the height squared to ensure an objective comparison of muscle mass between participants. Muscle strength was measured as handgrip strength (HGS) using a digital hand dynamometer (T.K.K 5401, Takei, Tokyo, Japan). With the participants in a standing position and forearm fully extended in a sideways position away from the body at thigh level, they were instructed to exert maximum grip strength twice each with the left and right hands, and the dominant hand was recorded. In this study, low muscle mass was defined as SMI < 7.0 kg/m² for men, and low muscle strength was defined as HGS < 28 kg for men, according to the consensus of the Asian Working Group for Sarcopenia (AWGS) 2019. ¹⁴ When muscle mass and muscle strength were conflicting, we adopted the SMI criteria for sarcopenia.

Participants with a life expectancy of <1 year due to malignancy and those with chronic diseases (heart failure, stroke, Alzheimer's disease, nutrition intake problem, chronic kidney disease) were excluded. After excluding ineligible participants, blood samples were collected from 313 eligible participants in the Veterans Sarcopenia Study cohort. ¹³ For each case, controls were matched (1:1) according to a 2‐year age difference. Global metabolomic profiling was conducted in two cohorts (batches) and analysed as one based on a previous study's guidelines. ¹⁵

Global metabolome profiling for human plasma using liquid chromatography‐tandem mass spectrometry

Blood samples were collected from the antecubital vein using an EDTA tube in the morning after an overnight fast of at least 8 h. Samples were centrifuged at 3000 rpm for 10 min at 4°C, and the supernatant was then carefully collected to exclude the cellular components. The plasma samples were stored at −80°C prior to the assays. After 100 μL of plasma was combined and mixed well with 375 μL of a chloroform/methanol mixture (1:2), the solution was combined with 125 μL of H₂O and 125 μL of chloroform. After centrifugation at 8000 rpm for 20 min, the upper layer of the aqueous phase and lower layer of the organic phase were collected and dried using a vacuum centrifuge (Speedvac, RVC 2–25 CDplus, Martin Christ, Germany).

The dried sample was reconstituted with 100 μL of the LC mobile solution. After centrifugation of samples at 14 000 rpm for 15 min, the supernatant was carefully moved to the autosampler at 4°C using an autosampler vial. Analyses were performed using a Vanquish/Q Exactive™ Plus Hybrid Quadrupole‐Orbitrap™ mass spectrometer system (Thermo Fisher Scientific, Sunnyvale, CA, USA) along with a reverse‐phase column (Pursuit C18 150 mm × 2.1 mm, 3 μm) and a HILIC column (Zorbax HILIC plus, 100 mm × 2.1 mm, 3.5 μm). Mobile phase A was 0.1% formic acid in water and mobile phase B was 0.1% formic acid in methanol. A scan range of m/z 67–1000 was chosen as the centroid mode, with the resolution for higher‐energy collisional dissociation cell spectra set to 70 000 m/z. The Compound Discoverer software v3.1 (Thermo Fisher Scientific) was used for the extraction of metabolic features and their identification using database searches. Two cohorts were analysed in a separate batch, and the batch effect was adjusted using the ComBat function of the surrogate variable analysis package. Based on the results of the normality test for continuous variables, we decided to use either a parametric or non‐parametric method: a two‐sample t‐test or Mann–Whitney U test was performed. The difference in m/z features was adjusted using the Benjamini–Hochberg (B–H) correction for the false discovery rate (FDR). ¹⁶ Principal component analysis (PCA) was performed to evaluate the experimental reproducibility. Subjects corresponding to outliers, which may affect the validity of the results of PCA, were excluded from further analysis. The differential expression analysis was used to explore the fold change (FC) and the ratio between cases and controls after adjustment for age and the presence of hypertension and diabetes using linear regression analysis. In addition, heat maps for the top 100 metabolites were used to visualize the differential metabolite levels. Based on identified metabolites, metabolic pathway enrichment analysis was conducted using the MetaboAnalyst 5.0 database (https://www.Metaboanalyst.ca/). Statistical analyses were performed using the R 3.6.3 program (R Foundation, Vienna, Austria), and the level of statistical significance was set at P < 0.05.

Experimental studies

Aging mouse model of sarcopenia

Because the natural aging mouse model of sarcopenia can reproduce the aging process to the greatest extent and has been widely used in the study of sarcopenia, ¹⁷ it was used in the present study. All mice were housed at 21–23°C with 12 h light/12 h dark cycles with free access to water and rodent chow under specific pathogen‐free conditions at the Asan Institute for Life Sciences. All animal care and procedures were conducted according to protocols and guidelines approved by the Institutional Animal Care and Use Committee of the Asan Institute for Life Sciences (No. 2016–12‐035). Six‐ and 18‐month‐old male C57BL/6 mice were purchased from the Korea Research Institute of Bioscience and Biotechnology (Daejeon, South Korea). After 4 h of fasting, mice aged 7 and 19 months were euthanized because 7 and 19 months of age in mice correlate with young adult and old age in humans, respectively. ¹⁷ The animals were euthanized by cardiac puncture under anaesthesia induced by an intraperitoneal injection of 50 mg/kg Zoletil 50 (Virbac, Carros, France) and 10 mg/kg Rompun (Bayer Korea, Seoul, South Korea). The quadricep muscles were isolated and weighed. Relative muscle mass (%) was expressed as a percentage of body weight and is shown as the sarcopenic index.

Differentiation of muscle cells

Murine C2C12 myoblasts (MBs) were purchased from American Type Culture Collection (ATCC, Rockville, MD, USA). C2C12 cells were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (Gibco, Grand Island, NY, USA), 100 U/mL penicillin, and 0.1 mg/mL streptomycin in a humidified atmosphere with 5% CO₂. C2C12 MBs were differentiated for 2 and 6 days in DMEM containing 2% horse serum and were considered myocytes (MCs) and myotubes (MTs), respectively. All cultures were incubated in serum‐free DMEM for 24 h. The conditioned medium (CM) was collected, and total cell lysates were prepared in a radioimmunoprecipitation (RIPA) buffer.

Targeted lipidome profiling of fatty acid amides in samples from mice and cell cultures

Muscle (20–30 mg) and plasma (50–100 μL) of mice as well as the CM and cell lysates of MBs, MCs, and MTs were subjected to liquid–liquid extraction after adding 100 μL of 500 nM arachidonoylethanolamide‐d₄ as an internal standard solution. Fatty acid amides were determined using an LC–MS/MS system equipped with a 1290 HPLC system (Agilent, Waldbronn, Germany) and a QTRAP 5500 (AB Sciex, Toronto, Canada). A reverse‐phase column (Pursuit C18, 150 × 2.1 mm) was used with mobile phase A (0.1% formic acid in H₂O) and mobile phase B (0.1% formic acid in methanol). The LC was run at 200 μL/min at 25°C. The isocratic conditions of 90% B were used for 20 min. Multiple reaction monitoring was performed in positive ion mode, and the extracted ion chromatogram corresponding to the specific transition for each analyte was used for quantification. The calibration range for each lipid was 0.1–10 000 nM (r ² ≥ 0.99). Data analysis was performed using Analyst 1.5.2 software. For the CM and cell lysates, lipid amounts were normalized to total protein concentration.

Targeted lipidome profiling of fatty acid amides in human plasma and human muscle tissues

Targeted lipidome analysis was conducted based on the hypothesis that fatty acid amides in blood were candidate biomarkers for sarcopenia. Fifty microliters of human plasma and 25 mg of human muscle were used, and fatty acid amides were quantified as described above.

Statistical analysis

Baseline characteristics of subjects were compared using the Student's t‐test or Wilcoxon rank‐sum test for continuous variables and the chi‐square test was used for categorical variables, as appropriate. If plasma fatty amide levels were not normally distributed based on the Kolmogorov–Smirnov test, natural log‐transformed fatty amide levels were used. Associations of metabolite levels with sarcopenia parameters (ASM, SMI, and HGS) and percent fat mass (pFM) were investigated using multiple linear regression analyses, after adjustment for age and presence of hypertension or diabetes. Receiver‐operating characteristic (ROC) analysis was performed, and the area under the ROC curve (AUC) was calculated to evaluate the ability of fatty metabolites to help predict sarcopenia. The cut‐off for metabolite levels predictive of sarcopenia was calculated using Youden's index. ¹⁸ Unadjusted and adjusted multiple logistic regression analyses were performed to generate odds ratios (ORs) with a 95% confidence interval (CI). All statistical analyses were performed using the SPSS statistical software (SPSS Inc., Chicago, IL), with P < 0.05 indicating statistical significance.

Results

Characteristics of the study participants

The study population initially included 144 participants (72 men with sarcopenia vs. 72 age‐matched men without sarcopenia as controls) from the Veterans Sarcopenia Study. After batch effect correction, one outlier was found in the case group in cohort 1. We removed this outlier in the case group and one matched control for further analysis (Figure S1). Finally, 71 cases of sarcopenia vs. 71 controls were analysed in this study. The main characteristics of the population according to the cases and controls are presented in Table 1. There were no significant age differences between the case and control groups in the entire cohort (P = 0.989). The weight, height, and BMI of the sarcopenia group were significantly lower than those of the control group. Muscle mass parameters (lean mass, ASM, and SMI) and HGS of the case group were significantly lower than those of the control group (Ps < 0.001). The SARC‐F value of the sarcopenia group was significantly higher than those of the control group (P = 0.004), and the EQ‐VAS score of the sarcopenia group was significantly lower than those of the control group (P < 0.001). However, the values obtained following the measurement of 25(OH)₃D levels, chair stand test, drinking evaluations, and prevalence of hypertension and diabetes did not differ between the case and control groups (Ps > 0.05).

Table 1.

Baseline characteristics of the study populations

	Case (n = 71)	Control (n = 71)	P*
Age (years)	74.0 [72.0;76.0]	74.0 [72.0;76.0]	0.989
Weight (kg)	55.7 ± 5.9	73.6 ± 9.5	<0.001
Height (cm)	163.0 ± 5.2	167.8 ± 4.9	<0.001
BMI (kg/m2)	21.1 [19.4;22.6]	26.1 [24.1;27.9]	<0.001
25(OH)3D (ng/mL)	12.7 [8.0;25.1]	14.1 [10.6;21.1]	0.444
Smoking, N (%)			0.002
Current	17 (23.9%)	3 (4.2%)
Ex‐smoker	41 (57.7%)	56 (78.9%)
Non‐smoker	13 (18.3%)	12 (16.9%)
Drinking, N (%)			0.255
≥3/week	19 (26.8%)	20 (28.2%)
1–2/week	4 (5.6%)	11 (15.5%)
<1/week	8 (11.3%)	6 (8.5%)
Non‐drinking	40 (56.3%)	34 (47.9%)
Exercise, N (%)			0.396
≥3/week	12 (16.9%)	6 (8.5%)
1–2/week	10 (14.1%)	11 (15.5%)
<1/week	6 (8.5%)	4 (5.6%)
No exercise	12 (16.9%)	6 (8.5%)
Hypertension, N (%)	37 (52.1%)	49 (69.0%)	0.059
Diabetes, N (%)	64 (90.1%)	68 (95.8%)	0.325
SARC‐F	1.0 [0.0; 2.0]	0.0 [0.0; 1.0]	0.004
EQ‐VAS	65.0 [50.0;77.5]	75.0 [60.0;85.0]	<0.001
HGS (kg)	27.4 ± 5.6	33.1 ± 7.2	<0.001
Chair stand up test (s)	7.0 [5.0;10.0]	7.0 [6.0; 9.0]	0.966
FM (kg)	14.0 [11.2;17.8]	21.8 [16.9;23.9]	<0.001
pFM (%)	24.9 ± 6.6	28.1 ± 6.2	0.003
LM (kg)	39.4 ± 2.7	49.9 ± 4.2	<0.001
ASM (kg)	17.1 ± 1.5	22.6 ± 2.2	<0.001
SMI (kg/m2)	6.5 [6.2; 6.6]	8.0 [7.6; 8.4]	<0.001

Data are presented as mean ± SD, median and IQR, or numbers (percentage) unless otherwise specified. P*: For continuous variables, the Student's t‐test was used when normality was satisfied; the Mann–Whitney U test was used when normality was not satisfied. The chi‐square or Fisher's exact tests were used to analyse categorical variables. Italicized numbers indicate statistically significant values.

ASM, appendicular skeletal muscle mass; BMI, body mass index; EQ‐VAS, EuroQol visual analogue scale; FM, fat mass; HGS, hand grip strength; IQR, interquartile range; LM, lean mass; pFM, percent fat mass; SARC‐F, strength, assistance in walking; rising from a chair, climbing stairs, falls; SD, standard deviation, SMI, skeletal muscle mass index; 25(OH)₃D, 25‐hydroxyvitamin D3.

Global metabolome profiling of plasma obtained from the Veterans Sarcopenia Study cohort

The PCA plot showed clustering of cases and controls and successful correction for batch effects in cohorts 1 and 2 (Figure S1). After adjustment for age and presence of hypertension and diabetes, the volcano plot showed differential metabolites, and statistical significance was determined based on the −log 10 (P‐value) with P < 0.05 and the |log₂ FC| > 1 (Figure 1). Although there was no statistical difference between the ages of the two groups, age was selected as a factor for adjustment owing to its substantial impact on sarcopenia. The heatmaps with differential metabolites between sarcopenia and control groups were created (Figure S2). A total of 312 metabolites showed statistically significant differences between the case and control groups (Table S1). The results of candidate metabolomic biomarkers are summarized after removing overlapping metabolites in Table 2. The metabolite set enrichment analysis showed that unsaturated fatty acid and lipid metabolic pathways were related to sarcopenia (Table 2 and Figure S3). We particularly focused on fatty acid amides because they are increasingly attracting attention as bioactive molecules that play roles in aging diseases. ¹⁹ , ²⁰ In addition, unsaturated fatty acids are included in many fatty acid amide species.

Figure 1.

Volcano plot showing the difference of metabolites after batch effect correction. The volcano plot shows the differential expression of metabolites by fold change (FC) and the ratio between cases and controls after adjustment for age and the presence of hypertension and diabetes following linear regression analysis. Metabolites, which were annotated as ‘UP’, were highly expressed in cases compared with the controls according to the −log 10 (P‐value) with P < 0.05 and the log₂ FC > 1. Metabolites, which were annotated as ‘DOWN’, were highly suppressed in cases than controls according to the −log 10 (P‐value) with P < 0.05 and the log₂ FC < −1. Metabolites, which were annotated as ‘NO’, had no significant differences in the expression between the two groups. Statistical significance was determined based on the −log 10 (P‐value) with P < 0.05 and the log₂ FC < −1 (or >1). LysoPC, lysophosphatidylcholine; PC, phosphatidylcholine.

Table 2.

Classification of significant metabolites related to sarcopenia based on global metabolome profiling

Classification/subclass (ref) a	Metabolites	Formula	Molecular Weight	RT [min]	mzCloud best match	Log2 FC	P‐value*
Fatty acid amides [FA080]/N‐acyl ethanolamines	Oleoyl ethanolamide	C20 H39 N O2	331.31	7.062	88.2	1.511	0.005163
Fatty esters [FA07]/fatty acyl carnitines	Decanoylcarnitine	C17 H33 N O4	315.24	9.477	68.8	1.037	0.024144
Fatty amides [FA08]/N‐acyl ethanolamines [FA0804]	1‐Stearoylglycerol	C21 H42 O4	358.31	23.904	98	0.224	4.01E‐06
Fatty acids and conjugates [FA01]/straight chain fatty acids	Stearic acid	C18 H36 O2	284.27	23.905	98.1	0.211	0.000265
Fatty amides [FA080]/Primary amides	Stearamide	C18 H37 N O	283.29	23.442	93.9	−0.826	3.35E‐28
Fatty amides [FA080]/Primary amides	Oleamide	C18 H35 N O	281.27	18.314	98.9	−0.907	3.60E‐37
Fatty amides [FA080]/Primary amides	Hexadecanamide/palmitamide	C16 H33 N O	255.26	17.278	99	−0.934	4.60E‐27
Fatty amides [FA080]/Primary amides	Docosanamide	C22 H45 N O	339.35	29.318	95.6	−1.429	4.17E‐19
Fatty acid amides [FA080]/N‐acyl ethanolamines	Stearoyl Ethanolamide	C20 H41 N O2	309.30	22.88	69.4	−1.462	1.31E‐16

FC of (case/control) after adjustment for age and the presence of hypertension and diabetes. P*: adjusting for age and presence of hypertension and diabetes with multiple testing corrections (false discovery rate [FDR]). Italicized numbers indicate statistically significant values.

FC, fold change; RT, retention time.

^aReference: LIPIDMAPS (https://www.lipidmaps.org/)

Fatty acid amides in the muscle and plasma of aging mouse model of sarcopenia and based on the differentiation of muscle cells

To determine whether the circulating fatty acid amides in patients with sarcopenia reflect changes in their muscular contents, fatty acid amide levels were examined in the plasma and muscle tissue from the established aging mouse model of sarcopenia. Despite higher body weight, the absolute muscle weight and relative muscle mass were significantly lower in the muscles of aged mice than in those of young mice (all, P < 0.05), indicating that age‐related muscle loss was well induced (Table S2). The levels of several fatty acid amides, like oleamide, linoleamide, palmitamide, palmitoyl EA, arachidonoyl EA, eicosapentaenoyl EA (EPA EA), and docosahexaenoyl EA (DHA EA) were significantly suppressed in the plasma of the aging mouse model of sarcopenia (all Ps < 0.05) compared with those in the plasma of young mice (Table 3). Arachidonoyl EA, EPA EA, and DHA EA levels were positively associated with relative muscle mass (all, P < 0.05, Table 3). By contrast, oleoyl EA levels were significantly high in the muscle tissue of the aging mouse model of sarcopenia and were not associated with relative muscle mass (Table S3). Contrary to the plasma DHA levels, DHA levels in muscle tissue tended to increase in the aging mouse model of sarcopenia and showed an inverse association with relative muscle mass. Changes in fatty acid amides in the CM and cell lysates during muscle cell differentiation were also explored. The levels of most of the fatty acid amides in CM were decreased as differentiation progressed (Figure 2A), although, oleoyl EA, linoleoyl EA, arachidonoyl EA, DHA EA, and EPA EA could not be measured because their levels were below detection limits. On the other hand, fatty acid amides including DHA EA in cell lysates increased as differentiation progressed (Figure 2B).

Table 3.

Primary fatty acid amide and EA levels in the plasma of mice of the aging model of sarcopenia and their association with muscle mass

	Plasma concentration (fmol/μL)			Relative muscle mass (%)
	Young (n = 10)	Old (n = 7)	P*	r	P †
Oleamide	13050.0 [12300.0;13750.0]	7800.0 [6865.0;11900.0]	0.033	0.372	0.142
Linoleamide	4665.0 [4200.0;4735.0]	2630.0 [2220.0;4305.0]	0.033	0.393	0.119
Palmitamide	1100.0 [1020.0;1435.0]	830.0 [707.0;932.0]	0.025	0.304	0.235
Stearamide	2210.0 [1930.0;2650.0]	1670.0 [1375.0;2320.0]	0.193	0.240	0.354
Arachidonoyl amide	7.7 [7.3; 8.4]	6.9 [5.7; 7.8]	0.270	0.299	0.243
Linoleoyl EA	14.1 [13.2;14.6]	10.2 [9.5;14.2]	0.193	0.325	0.203
Oleoyl EA	22.4 [21.3;24.8]	22.3 [18.4;24.2]	0.601	0.307	0.231
Stearoyl EA	18.2 [14.9;25.0]	16.4 [14.3;20.3]	0.601	0.096	0.713
Palmitoyl EA	19.4 [17.6;23.1]	12.8 [11.9;15.3]	0.013	0.454	0.067
Arachidonoyl EA	14.8 [14.2;16.5]	9.3 [8.9;10.3]	0.002	0.581	0.014
EPA EA	0.5 [0.5; 0.6]	0.3 [0.3; 0.4]	0.002	0.643	0.005
DHA EA	8.9 [8.8;10.0]	7.0 [6.8; 7.8]	0.010	0.592	0.012

P*: Statistical analysis was performed using Mann–Whitney U test when normality was not satisfied. P ^† : Pearson's correlation analysis. Italicized numbers indicate statistically significant values.

EA, ethanolamide; EPA, eicosapentaenoyl; DHA, docosahexaenoyl.

Figure 2.

Changes in fatty acid amide and EA levels during myoblast differentiation. Fatty acid amides and EA levels in (A) cell media and (B) cell lysates of myoblasts (MBs), myocytes (MCs), and myotubes (MTs). Oleoyl EA, linoleoyl EA, arachidonoyl EA, DHA EA, and EPA EA in cell media were not measured because their levels were below the detection limits of the analytical method. Significant changes in the levels of each fatty acid amide in MCs and MTs compared with those in MBs are shown as *P < 0.05 and #P < 0.01. The error bar represents the standard deviation. P‐values were calculated using Mann–Whitney U test analysis. EA, ethanolamide; EPA, eicosapentaenoyl; DHA, docosahexaenoyl.

Fatty acid amides in human plasma obtained from veterans sarcopenia study cohorts

Quantitation of fatty acid amides in human plasma revealed that stearamide and stearoyl EA were significantly elevated (P = 0.022 and P = 0.021, respectively), while DHA EA levels reduced (P = 0.016) in the sarcopenia group; the levels of other fatty acid amides did not show statistically meaningful changes (Table S4). Linear regression analysis of fatty acid amides and sarcopenic parameters, such as ASM, SMI, and HGS, revealed a significant positive association with DHA EA levels (P = 0.005, P = 0.001, and P = 0.012, respectively, Table 4) but not with stearamide and stearoyl EA levels. EPA EA showed a positive association with HGS (P = 0.029) but not with ASM and SMI. DHA EA levels and EPA levels showed a positive association and a tendency of positive association with the pFM (P = 0.015 and P = 0.076, respectively). The discriminatory ability of DHA EA levels for sarcopenia diagnosis was assessed by ROC analysis, with an AUC of 0.618 (95% CI: 0.532–0.698) for an optimal cut‐off value of DHA EA ≤4.60 fmol/μL. Univariate (unadjusted) and multivariate (adjusted) logistic regression analyses revealed that OR for sarcopenia was higher in participants with a DHA EA ≤4.60 fmol/μL (OR: 2.11, 95% CI: 1.03–4.30) relative to those with DHA EA >4.60 fmol/μL, even after adjustment for HGS (Table 5). The addition of DHA EA ≤4.60 fmol/μL to age and HGS significantly improved the AUC from 0.620 to 0.691 (P = 0.0497) (Figure 3).

Table 4.

Linear regression analysis of fatty acid amide and EA levels in human plasma with sarcopenia parameters (ASM, SMI, and HGS) and pFM

Metabolites	ASM				SMI				HGS				pFM
	β*	SE	β†	P*	β*	SE	β†	P*	β*	SE	β†	P*	β*	SE	β†	P*
Stearamide	−3.361	2.423	−0.114	0.168	−1.039	0.662	−0.129	0.119	−0.521	4.988	−0.008	0.917	0.220	4.607	0.004	0.962
Stearoyl EA	−2.051	1.933	−0.087	0.291	−.653	0.528	−0.101	0.218	−0.631	3.967	−0.013	0.874	2.604	3.658	0.056	0.478
EPA EA	1.568	1.172	0.118	0.184	1.568	1.172	0.118	0.184	5.112	2.310	0.186	0.029	3.955	2.213	0.149	0.076
DHA EA	4.982	1.758	0.227	0.005	1.646	0.475	0.275	0.001	9.174	3.614	0.200	0.012	8.211	3.342	0.191	0.015

Fatty acid amide and EA levels were log‐transformed because of their skewed distribution. Italicized numbers indicate statistically significant values. β*: Unstandardized coefficient, β†: Standardized coefficient. P*: The enter method was applied to this model after adjustment for age and the presence of hypertension or diabetes.

ASM, appendicular skeletal mass, DHA, docosahexaenoyl; EA, ethanolamide; EPA, eicosapentaenoyl; HGS, hand grip strength; pFM, percent fat mass; SMI: skeletal muscle mass index, SE: standard error.

Table 5.

Logistic regression analysis for DHA EA levels with sarcopenia

Variable	Unadjusted model		Adjusted model*
	OR (95% CI)	P	OR (95% CI)	P
Per unit increase
Per log‐unit increase of DHA EA levels	0.05 (0.01–0.55)	0.013	0.15 (0.01–1.99)	0.148
Per 1 kg increase of HGS	0.87 (0.82–0.92)	<0.001	0.85 (0.79–0.91)	<0.001
According to cut‐off
DHA EA ≤ 4.60 fmol/μL	2.51 (1.28–4.93)	0.008	2.11 (1.03–4.30)	0.041
HGS < 28 kg	3.74 (1.81–7.73)	<0.001	3.33 (1.40–7.97)	0.007

DHA EA, docosahexaenoyl ethanolamide; OR, odds ratio; 95% CI, 95% confidence interval. Italicized numbers indicate statistically significant values.

^*Age and presence of hypertension and diabetes were adjusted.

Figure 3.

The ROC curve for predicting sarcopenia after adding DHA EA level to HGS. Cut‐offs for DHA EA level and HGS were ≤4.60 fmol/μL and <28 kg, respectively. AUC, area under the ROC curve; DHA EA, docosahexaenoyl ethanolamide; HGS, hand grip strength; ROC curve, receiver‐operating characteristic curve; 95% CI, 95% confidence interval.

Fatty acid amides in human muscle tissues obtained from the AMC cohort

Targeted profiling of fatty acid amides in human muscle tissue revealed no significant difference in levels of most primary fatty amides between the case and control groups except for palmitamide which tended to increase in the muscle of cases (Table S5).

Discussion

Global metabolome profiling showed that the levels of fatty acid amides were significantly different in the plasma of elderly men with sarcopenia. These results were confirmed by both decreased levels in the plasma of the aging mouse model of sarcopenia and increased levels in cell lysates during muscle cell differentiation. DHA EA levels in the plasma of the aging mouse model of sarcopenia were positively associated with relative muscle mass. Furthermore, targeted metabolome profiling showed that plasma DHA EA level was decreased in men with sarcopenia and positively correlated with both SMI reflecting muscle mass and HGS reflecting muscle strength. The lower plasma DHA EA levels were significantly associated with sarcopenia independent of HGS. The addition of plasma DHA EA levels to age and HGS significantly improved the discriminatory performance of the latter for sarcopenia, as shown by a 7.1% increase in AUC. These findings suggested that circulating DHA EA levels could serve as a biomarker for sarcopenia.

The processes underlying the pathogenesis of sarcopenia are not fully understood and may involve multiple factors. Importantly, accumulating evidence shows that fatty acids and their lipid intermediates regulate skeletal muscle mass and function. ²¹ Several in vitro ²² , ²³ and in vivo ²⁴ , ²⁵ , ²⁶ studies suggest that saturated fatty acids (SFA) and unsaturated fatty acids (UFA), especially omega‐3 polyunsaturated fatty acids (PUFA) including EPA and DHA, induce promotion or prevention of muscle atrophy and protein degradation, respectively. Furthermore, UFA protects against muscle wasting in response to various pathological conditions. Recently, plasma oleamide ²⁷ and oleyl EA and palmitoyl EA levels ²⁸ were found to decrease significantly in the animal model of sarcopenia. Oleyl EA levels were positively associated with the skeletal muscle function. ²⁸ These findings indicated the potential role of fatty acid amides as biomarkers of sarcopenia.

In the present study, DHA EA levels in the plasma reduced significantly in both the aging mouse model and men with sarcopenia, showing a positive association with muscle mass and strength. Furthermore, lower plasma DHA EA levels were significantly associated with sarcopenia independent of muscle strength. Finally, adding DHA EA levels to muscle strength improved the diagnostic ability for sarcopenia. Although we could not replicate our results in another independent cohort, recently another study using global metabolomics in a small number of older subjects (20 cases and 21 control) showed lower plasma DHA EA levels in sarcopenic subjects, ²⁹ consistent with our results. These findings, if confirmed by another large independent cohort, might support the role of DHA EA as a biomarker for sarcopenia rather than simply reflecting muscle mass or relating to atrophy or cachexia.

The exact function of DHA EA and the mechanism underlying its effect on muscle function in sarcopenia are still unclear. However, existing literature suggests that DHA EA regulates both muscle mass and muscle function through its ability to modulate muscle cell growth, proliferation, and/or differentiation. ²¹ , ³⁰ Consistent with our findings showing a positive association of DHA EA level with muscle mass, DHA EA might maintain skeletal muscle hypertrophy and mass by activating the pathways which are extremely important for protein synthesis in muscle hypertrophy (insulin signalling and Akt‐mTOR‐p70S6K pathways) and the fractional synthesis rate. ²⁴ , ²⁵ , ²⁶ Effect of DHA EA on anti‐inflammatory actions, ³¹ improved mitochondrial function, ³² an enhanced shift from fast glycolytic to slow oxidative fibre types in skeletal muscle, ³³ restoration of autophagy, ²⁷ and increasing the number of activated satellite cells and myogenic progenitor cells for muscle growth and regeneration cells ³⁴ may also contribute to DHA EA's ability to preserve muscle mass and/or function. Recently, DHA EA improved the proliferation and differentiation of MB by mediating the endocannabinoid system which impacts energy homeostasis and cellular activity. ³⁵ Consistent with our findings showing a positive association of DHA EA level with muscle strength in men, ingestion of EPA and DHA can increase muscle strength and/or physical performance by neuromuscular adaptation after exercise. ³⁰ , ³⁶ One possible mechanism underlying this effect is an increase in the incorporation of omega‐3 fatty acids in the cells, particularly in the nerve and muscle, ³⁷ resulting in improved fluidity of the membrane and acetylcholine sensitivity.

Consistent with recent studies reporting decreased levels of oleamide, oleyl EA, and palmitoyl EA in the plasma of animal models of sarcopenia, ²⁷ , ²⁸ the levels of fatty acid amide including EPA EA, and DHA EA in the plasma of aging mouse model reduced significantly. In contrast to the plasma levels, the fatty acid levels including DHA EA in the muscle tissue of the aging mouse model increased significantly. In contrast to the DHA EA levels in the plasma, the muscular content was inversely associated with relative muscle mass in human muscle tissues. A small number of human muscle specimens and differences in species might contribute to differences in fatty acid amides, underlying the changes between animal models and men. This suggests that concentrations of circulating fatty acid amides might not reflect their muscular content but their use or transport into tissues. The results of this cellular experimental study may provide insights into this regard. Fatty acid amides are necessary for the differentiation of muscle cells, and as muscle cells differentiate, their movement into the cells is facilitated. ³⁸ Treatment of DHA EA resulted in improved proliferating and differentiated MB, but the opposite was observed with arachidonoyl EA. ³⁵ In our study, the concentrations of fatty acid amides in the CM decreased as muscle cell differentiation progressed, although DHA EA could not be measured because of the detection limit of the analytical method. By contrast, the cellular contents of most of the fatty acid amides increased with muscle cell differentiation, except arachidonoyl EA. This suggests that fatty acid amides play an important role in MB differentiation, which may cause the accumulation of fatty acid amides in muscle cells in response to the CM, in accordance with the progress of MB differentiation, and result in a reduction of fatty acid amides in CM. These findings suggested a need for different cell culture types and animal models to identify the production and relevance of DHA EA, which might support the role of DHA EA as a biomarker for sarcopenia.

A major strength of our study is the inclusion of a relatively large number of elderly men with matched controls for minimizing gender differences in metabolomic research. In addition, following animal studies and muscle cellular experimental studies have strengthened the study's validation. However, this study has a few limitations. First, an aging mouse model of sarcopenia was used in this study to understand the roles of fatty acid amides in sarcopenia. Age‐matched animal models of sarcopenia would be ideal in light of our findings. Second, we focused on the plasma of only elderly men in a Veterans Hospital, as 90% of the inpatients were men. Although this is an intrinsic limitation of institutional factors, our results need to be confirmed in further studies with independent cohorts for diagnosing and predicting sarcopenia even in women. Future research will provide more insightful biomarkers with clinical implications for functional evaluation through various experimental models, and our study serves as a starting point. Third, global and targeted metabolomic profiling was performed with the same human cohorts; however, for some metabolites, these two strategies did not show the same results. This discrepancy could be explained by intrinsic differences between these two strategies. These two approaches can be characterized by the number of detected metabolites as well as the reliability of the quantification method because untargeted metabolomics focuses on the analysis of a broad range of metabolites while targeted metabolomics emphasizes on reliable quantitation of defined groups of metabolites. ³⁹ Thus, the hypothesis generated from untargeted metabolomics was confirmed by target quantitation.

In conclusion, our study demonstrated that fatty acid amides are potential plasma biomarkers reflecting muscle mass and strength in elderly men with sarcopenia, especially circulating DHA EA levels. Further studies using extensive healthcare data would be needed for validation, and related metabolic pathway exploration with various biological tools is warranted to gain deeper insights.

Conflicts of interest statement

The authors declare that they have no conflict of interest. The authors of this manuscript certify that the study complies with the ethical guidelines for authorship and publishing in the Journal of Cachexia, Sarcopenia and Muscle. ⁴⁰

Supporting information

Table S1. Metabolites identified in untargeted measured with batch‐effect control adjusting for age and presence of hypertension and diabetes

Table S2. Baseline characteristics of aging mice according to the status of sarcopenia

Table S3. Primary fatty acid amide and EA levels in the muscle of aging model of sarcopenia and their association with muscle mass

Table S4. Primary fatty acid amide and EA levels in human plasma from sarcopenia cases based on targeted metabolome profiling.

Table S5. Primary fatty acid amide and EA levels in human muscles in sarcopenia based on targeted metabolome profiling

Fig S1. Principal component analysis of case versus control

Fig S2. Heatmap showing the difference in metabolites between cases and controls after batch correction

Fig S3. Metabolite Set Enrichment Analysis for untargeted metabolites in human plasma using MetaboAnalyst 5.0 (https://www.metaboanalyst.ca/)

Kim Y. A., Lee S. H., Koh J.‐M., Kwon S.‐h., Lee Y., Cho H. J., et al (2023) Fatty acid amides as potential circulating biomarkers for sarcopenia, Journal of Cachexia, Sarcopenia and Muscle, 14, 1558–1568, 10.1002/jcsm.13244