Sphingolipid metabolites as potential circulating biomarkers for sarcopenia in men

Abstract

Background

Sarcopenia is an age‐related progressive loss of muscle mass and function. Sarcopenia is a multifactorial disorder, including metabolic disturbance; therefore, metabolites may be used as circulating biomarkers for sarcopenia. We aimed to investigate potential biomarkers of sarcopenia using metabolomics.

Methods

After non‐targeted metabolome profiling of plasma from mice of an aging mouse model of sarcopenia, sphingolipid metabolites and muscle cells from the animal model were evaluated using targeted metabolome profiling. The associations between sphingolipid metabolites identified from mouse and cell studies and sarcopenia status were assessed in men in an age‐matched discovery (72 cases and 72 controls) and validation (36 cases and 128 controls) cohort; women with sarcopenia (36 cases and 36 controls) were also included as a discovery cohort.

Results

Both non‐targeted and targeted metabolome profiling in the experimental studies showed an association between sphingolipid metabolites, including ceramides (CERs) and sphingomyelins (SMs), and sarcopenia. Plasma SM (16:0), CER (24:1), and SM (24:1) levels in men with sarcopenia were significantly higher in the discovery cohort than in the controls (all P < 0.05). There were no significant differences in plasma sphingolipid levels for women with or without sarcopenia. In men in the discovery cohort, an area under the receiver‐operating characteristic curve (AUROC) of SM (16:0) for low muscle strength and low muscle mass was 0.600 (95% confidence interval [CI]: 0.501–0.699) and 0.647 (95% CI: 0.557–0.737). The AUROC (95% CI) of CER (24:1) and SM (24:1) for low muscle mass in men was 0.669 (95% CI: 0.581–0.757) and 0.670 (95% CI: 0.582–0.759), respectively. Using a regression equation combining CER (24:1) and SM (16:0) levels, a sphingolipid (SphL) score was calculated; an AUROC of the SphL score for sarcopenia was 0.712 (95% CI: 0.626–0.798). The addition of the SphL score to HGS significantly improved the AUC from 0.646 (95% CI: 0.575–0.717; HGS only) to 0.751 (95% CI: 0.671–0.831, P = 0.002; HGS + SphL) in the discovery cohort. The predictive ability of the SphL score for sarcopenia was confirmed in the validation cohort (AUROC = 0.695, 95% CI: 0.591–0.799).

Conclusions

SM (16:0), reflecting low muscle strength, and CER (24:1) and SM (16:0), reflecting low muscle mass, are potential circulating biomarkers for sarcopenia in men. Further research on sphingolipid metabolites is required to confirm these results and provide additional insights into the metabolomic changes relevant to the pathogenesis and diagnosis of sarcopenia.

Introduction

Sarcopenia is the progressive loss of muscle mass and function with age.^1–3 Sarcopenia is associated with adverse outcomes, including falls, functional decline, frailty, and mortality.^1–3 Therefore, sarcopenia has emerged as a major health concern, causing a huge economic burden.⁴ Currently, there are no specific drugs approved for treating sarcopenia, and the therapeutic effects of exercise and nutritional supplementation are limited after progression³; hence, detecting sarcopenia at an early stage is important for its management.

Diagnosing sarcopenia requires using various modalities to measure muscle mass, strength, and physical performance.^1–3 However, these modalities mainly reflect the skeletal muscle status as a static indicator at a somewhat advanced state. Muscle is a highly active metabolic organ, and changes in energy metabolism have been reported to contribute directly to age-related changes in skeletal muscle in the early stages of sarcopenia.⁵ Because metabolites represent the downstream changes in genome, transcriptome, and proteome expressions, metabolomics is an emerging approach that can reveal inherent omics variation closest to the disease risk/phenotype following recent advances in liquid chromatography–tandem mass spectrometry (LC–MS/MS).⁶ Thus, metabolites may be used as early diagnostic biomarkers of sarcopenia.

Recent studies on metabolomic biomarkers for sarcopenia have reported that some amino acids, lipids, and metabolites related to gut bacterial metabolism are linked with sarcopenia^7–14; however, metabolomic analysis of sarcopenia is in its early stages.⁶ The use of untargeted metabolomics allows for the observation of a wide range of metabolic features, and recent advances in untargeted metabolomics allow better compound identification due to the use of larger libraries of MS/MS spectra; comparatively, targeted metabolomics provides reliable quantitation.¹⁵ Thus, untargeted metabolomics offers better opportunities to discover novel metabolites and generate hypotheses, while targeted metabolomics is useful in hypothesis-driven validation.^6,15

To identify novel biomarkers of sarcopenia, we conducted metabolomic studies designed in three phases: (1) discovering sphingolipids as sarcopenia-related metabolomic features in the plasma of an aging mouse model using untargeted metabolomics, (2) exploring sphingolipid expression levels, using targeted metabolomics, for both the aging mouse model of sarcopenia to demonstrate a change of metabolite profiles and muscle cells to demonstrate a dynamic modulation of metabolites during myogenesis, and (3) analysing sphingolipid expression levels using the plasma of elderly men and women with sarcopenia and age-matched controls in two independent cohorts, to demonstrate the difference in the sphingolipid levels, and their association with sarcopenic parameters. Overall, this study aimed to provide reliable metabolomic biomarkers for sarcopenia.

Sphingolipids, including ceramides (CERs), are bioactive lipids with diverse cellular functions.¹⁶ CERs, central lipids for sphingolipids synthesis, are mainly synthesized via three canonical pathways: (1) de novo synthesis in the endoplasmic reticulum, (2) sphingomyelin (SM) hydrolysis in the cell membrane, and (3) salvaging in the lysosome.¹⁷ Serine deficiency activates serine palmitoyl-transferase, which incorporates other amino acids, including alanine, to form deoxysphingolipids via the non-canonical pathyway.¹⁸ Mammals have six CER synthases (CERS1–6), each exhibiting characteristic substrate specificity toward acyl-CoAs with different acyl chain lengths.¹⁹ Recent studies showed the association of age-related sphingolipid accumulation in skeletal muscle with muscle mass and function^20,21 and plasma sphingolipid levels with lean mass (LM)²² and gait parameters.¹³ Here, we report across three study phases that some sphingolipid metabolites are potential circulating biomarkers for sarcopenia in men.

Methods

Experimental studies using an aging mouse model of sarcopenia and muscle cells

This study used the naturally aging mouse model of sarcopenia.²³ The Institutional Animal Care and Use Committee of the Asan Institute for Life Sciences (No. 2016-12-035) reviewed and approved all animal care and procedures.

Myotube surface area and nuclear fusion index were analysed using MyoCount software (v1.3.1).¹²

Human cohort study

Study participants

Two independent case–control studies were 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 cohort, participants aged ≥65 years 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 were enrolled in the ‘Veterans Sarcopenia Study’.²⁴ 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 validation cohort included patients who visited the AMC (Seoul, Korea) between May 2017 and March 2020 to undergo comprehensive assessment for musculoskeletal disorders. Before the study, all participants completed questionnaires (including medical history), muscle mass measurements, and blood sampling.

Assessment of sarcopenia

For the discovery and validation cohorts, body composition was evaluated using bioelectrical impedance analysis (InBody 570, Biospace Co., Seoul, Korea). Appendicular skeletal muscle mass (ASM) was calculated as the sum of the muscle mass in both arms and legs, and the skeletal muscle mass index (SMI) was calculated by dividing ASM by height squared to ensure an objective comparison of muscle mass between participants. The discovery cohort measured muscle strength as handgrip strength (HGS) using a digital hand dynamometer (T.K.K 5401, Takei, Tokyo, Japan). With the participants standing and forearms fully extended in a sideways position away from the body at thigh level, they were instructed to exert maximum grip strength twice with each hand, and the dominant hand was recorded. For the men in this study, low muscle mass was defined as SMI < 7.0 kg/m², and low muscle strength was defined as HGS < 28 kg, according to the consensus of the Asian Working Group for Sarcopenia (AWGS) 2019.¹ In instances of conflicting results between SMI and HGS, the diagnosis of sarcopenia was determined based on SMI.

Participants with a life expectancy of less than 1 year due to malignancy and those with chronic diseases (heart failure, stroke, Alzheimer's disease, nutrition intake problem, and chronic kidney disease) were excluded. In the discovery cohort, blood samples were collected from 313 eligible participants in the Veterans Sarcopenia Study cohort after excluding ineligible participants.²⁴ For each case, controls were matched in a 1:1 ratio for age within a ± 2-year range. In the validation cohort, controls were matched according to a 2-year age difference for each case.

Metabolome profiling in sample

Metabolites were extracted from samples using liquid–liquid extraction procedures²⁵ and were determined using an LC–MS/MS system. Detailed methods are described in the supporting information.

Statistical analysis

We used the Student's t-test or Mann–Whitney U test for continuous variables and the chi-square test or Fisher's exact test for categorical variables, as needed for comparisons. A P value < 0.05 from the Kolmogorov–Smirnov test for normality assessment indicated a non-normal distribution. The correlation between metabolite levels and relative muscle mass was investigated using Spearman's correlation analysis. Sphingolipid metabolite levels were log-transformed due to their skewness. Unadjusted logistic regression analyses were performed to generate odds ratios (ORs) with a 95% confidence interval (CI). The area under the receiver-operating characteristic (ROC) curve (AUROC) was calculated to evaluate the ability of clinical variables and sphingolipid levels to predict sarcopenia. Associations between metabolite levels and sarcopenia parameters (SMI and HGS) were investigated using unadjusted and adjusted multiple linear regression analyses. All statistical analyses were performed using the R 4.2.2 program (R Foundation, Vienna, Austria), with P < 0.05 indicating statistical significance.

Results

Global metabolome profiling in the plasma of mice

Despite higher body weight, relative muscle mass was significantly lower in the muscles of aged mice than in young mice (P = 0.002), indicating age-related skeletal muscle loss (Table S1).

A total of 699 metabolic features were observed in global metabolome profiling; their confidence levels for compound annotation were either level 2 or 3.²⁶ A total of 146 metabolites showed statistically significant differences between aged and young mice based on the –log₁₀ (P-value) with P < 0.05 and FC > 1.2 (Table S2). Pathway analysis of the 146 significantly changed metabolites showed sphingolipid metabolic pathways potentially related to sarcopenia (Figure 1). Within these pathways, sphingolipid metabolism was selected for further analysis as recent studies showed that the sphingolipid in the skeletal muscle and plasma was associated with age-related decline in the muscle mass and function^20,21 and some parameters of sarcopenia, respectively^13,22.

[IMAGE OMITTED. SEE PDF]

Sphingolipid levels in the muscle and plasma of mice

Absolute muscle weight and relative muscle mass were significantly lower in the muscles of aged mice than in young mice (both P < 0.05; Table S1). Aged mice tended to have higher plasma CER (24:1) than young mice (P = 0.059; Table 1). Plasma CER (24:1) was inversely correlated with relative muscle mass in Spearman's correlation analysis (r = −0.522, P = 0.034; data not shown). Muscle CER (18:1) content was significantly higher in aged mice than in young mice (P = 0.002) and inversely correlated with relative muscle mass (r = −0.553, P = 0.021; data not shown).

Table 1 Sphingolipid levels in mice from the aging mouse model of sarcopenia and young mice

	Plasma concentration (fmol/μL)	Muscle content (fmol/mg)
Young (n = 10)	Old (n = 7)	Pa	Pb	Young (n = 10)	Old (n = 7)	Pa	Pb
CER (14:0)	ND	ND			0.4 [0.4; 0.5]	0.5 [0.4; 0.6]	0.142c	0.484
CER (16:0)	150.7 ± 36.5	175.1 ± 29.9	0.166	0.385	87.5 [67.0; 104.0]	100.0 [88.0; 113.5]	0.379c	0.650
SM (16:0)	26 050.0 [25 400.0; 30 300.0]	29 000.0 [26 800.0; 33 750.0]	0.625c	0.983	4171.5 [3185.0; 5233.0]	4914.0 [4042.5; 5027.5]	>0.999c	>0.999
CER (18:0)	39.7 ± 11.0	38.5 ± 11.3	0.819	0.819	6350.5 ± 938.9	6883.1 ± 1290.4	0.338	0.580
SM (18:0)	1545.0 [1430.0; 1600.0]	1230.0 [1155.0; 1375.0]	0.028c	0.152	4983.7 ± 971.5	5186.6 ± 963.3	0.677	0.738
CER (18:1)	2.6 ± 0.8	2.8 ± 1.1	0.644	0.757	144.4 ± 10.3	181.7 ± 21.0	<0.001	0.002
SM (18:1)	1098.2 ± 270.8	970.7 ± 263.1	0.349	0.640	490.9 ± 90.7	515.1 ± 114.2	0.632	0.738
CER (20:0)	72.9 ± 17.8	56.6 ± 15.4	0.069	0.251	49.2 ± 14.4	48.4 ± 14.4	0.915	0.915
CER (24:0)	1505.0 [1380.0; 2130.0]	1700.0 [1415.0; 1970.0]	0.962c	>0.999	326.9 ± 180.2	225.4 ± 43.7	0.168	0.404
SM (24:0)	110 00.0 [9890.0; 13 600.0]	11 300.0 [10 250.0; 12 850.0]	0.845c	>0.999	2360.0 [1671.0; 3533.0]	2107.0 [1608.0; 2190.0]	0.161c	0.484
CER (24:1)	1458.9 ± 335.8	2054.3 ± 420.0	0.005	0.059	551.9 ± 173.8	690.7 ± 88.8	0.073	0.383
SM (24:1)	29 350.0 [27 800.0; 33 700.0]	34 500.0 [31 950.0; 36 450.0]	0.064c	0.233	3528.0 [3196.0; 4883.0]	4345.0 [3353.5; 4465.0]	0.962c	>0.999

Sphingolipid levels in differentiating muscle cells

Changes in sphingolipid levels in the cell lysate and CM of differentiating muscle cells were also explored. CER (14:0), CER (16:0), CER (24:1), SM (18:0), and SM (24:1) levels in cell lysates were significantly higher in MC than MB but not significantly different between MT and MB. CER (20:0) and SM (24:0) levels were significantly lower in MT than MB (all P < 0.05; Figure 2A). Alternatively, most of the CERs and SMs in CM were significantly lower at the later stages of differentiation (MC vs. MB and MT vs. MB, all P < 0.05; Figure 2B).

[IMAGE OMITTED. SEE PDF]

Characteristics of study participants in the discovery cohort

The main characteristics of the 144 men in the discovery cohort (72 men with sarcopenia and 72 age-matched controls) from the Veterans Sarcopenia Study are presented in Table 2. There were no significant differences in age between the case and control groups (P = 0.989). The weight, height, and BMI of the sarcopenia group were significantly lower than in the control group (all P < 0.001). Muscle mass parameters (lean mass, ASM, and SMI) and HGS in the case group were significantly lower than in the control group (all P < 0.001). Weight was highly correlated with SMI following Spearman's correlation analysis (r = 0.871, P < 2.2E-16; data not shown). There were no significant differences in the chair stand test score, drinking and exercise habits, or prevalence of hypertension and diabetes between the case and control groups (all P > 0.05).

Table 2 Baseline characteristics of men in the discovery cohort (n = 144)

Disease	Control (n = 72)	Sarcopenia (n = 72)	P
Age (years)	74.0 [72.0; 76.0]	74.0 [72.0; 76.0]	0.989
Weight (kg)	72.7 [66.5; 79.5]	56.2 [51.2; 59.1]	<0.001
Height (cm)	167.8 ± 4.8	163.0 ± 5.1	<0.001
BMI (kg/m2)	26.1 [24.1; 27.9]	21.1 [19.4; 22.6]	<0.001
Smoking, N (%)			0.002
Ex-smoker	57 (79.2%)	42 (58.3%)
Non-smoker	12 (16.7%)	13 (18.1%)
Current smoker	3 (4.2%)	17 (23.6%)
Drinking, N (%)			0.174
No alcohol	34 (47.2%)	41 (56.9%)
Alcohol <1/week	20 (27.8%)	19 (26.4%)
Alcohol 1–2/week	12 (16.7%)	4 (5.6%)
Alcohol ≥3/week	6 (8.3%)	8 (11.1%)
Exercise, N (%)			0.397
No exercise	6 (8.3%)	12 (16.7%)
Exercise <1/week	4 (5.6%)	6 (8.3%)
Exercise 1–2/week	12 (16.7%)	10 (13.9%)
Exercise ≥3/week	50 (69.4%)	44 (61.1%)
Hypertension, N (%)	50 (69.4%)	37 (52.1%)	0.051
Diabetes, N (%)	60 (95.8%)	64 (90.1%)	0.314
FM (kg)	21.6 [16.9; 23.9]	14.1 [11.2; 17.8]	<0.001
pFM (%)	28.1 ± 6.2	25.0 ± 6.6	0.003
EQ-VAS	75.0 [60.0; 85.0]	65.0 [50.0; 77.5]	<0.001
SARC-F	0.0 [0.0; 1.0]	1.0 [0.0; 2.0]	0.005
HGS (kg)	34.6 ± 7.0	28.5 ± 5.6	<0.001
Chair stand up test (s)	7.0 [6.0;9.0]	7.0 [5.0;8.0]	0.874
LM (kg)	49.6 [47.7; 52.4]	39.8 [38.0; 41.6]	<0.001
ASM (kg)	22.4 [21.4; 23.9]	16.9 [16.1; 18.2]	<0.001
SMI (kg/m2)	8.0 [7.6; 8.4]	6.5 [6.2; 6.6]	<0.001

The main characteristics of the 72 women in the discovery cohort (36 women with sarcopenia and 36 age-matched controls) from the Veterans Sarcopenia Study are presented in Table S3. Muscle mass parameters (lean mass, ASM, and SMI), but not HGS, were significantly lower in the case group than the control group (all P < 0.001).

Association between sphingolipid metabolites in human plasma and sarcopenia in the discovery cohort

Plasma SM (16:0), CER (24:1), and SM (24:1) levels were significantly higher in men with sarcopenia than in the control group (all P < 0.05; Table 3). Contrastingly, there were no significant differences in the sphingolipid levels in women with or without sarcopenia in the discovery cohort (Table S4). Higher long-chain and very-long-chain dihydroceramide levels were observed in men with sarcopenia via the canonical pathway, although the difference between the two groups was not statistically significant (Table S5). Additionally, there was no significant difference between the two groups of amino acid levels (serine, alanine, and glycine) and deoxyCER levels via the non-canonical pathway.

Table 3 Plasma sphingolipid levels of men in the discovery cohort based on targeted metabolome profiling (n = 144).

	Control (n = 72)	Case (n = 72)	log2(FC)	Pa	Pb
CER (14:0) (μM)	13.6 [11.3;15.2]	13.6 [11.4;16.6]	−0.066	0.740c	0.807
CER (16:0) (μM)	64.1 ± 15.9	69.6 ± 18.2	−0.121	0.052	0.154
SM (16:0) (μM)	40 641.7 ± 6461.3	45 047.3 ± 8671.9	−0.148	0.001	0.003
CER (18:0) (μM)	35.0 [29.1;41.6]	38.3 [27.4;47.1]	−0.039	0.592 c	0.763
SM (18:0) (μM)	23 458.6 [20 875.5; 26 159.6]	24 353.6 [20 977.2; 29 426.1]	−0.098	0.097 c	0.233
CER (18:1) (μM)	7.0 [5.7; 8.5]	7.0 [5.2; 8.8]	−0.007	0.924 c	0.924
SM (18:1) (μM)	5341.0 [4408.0;6126.4]	5312.2 [4152.9;6590.9]	−0.059	0.584 c	0.763
CER (20:0) (μM)	92.8 [77.9;108.5]	92.5 [77.8;117.3]	−0.076	0.636 c	0.763
CER (24:0) (μM)	1682.0 [1445.8;1995.8]	1891.3 [1438.7;2318.8]	−0.149	0.122 c	0.243
SM (24:0) (μM)	13 430.5 [10 920.2; 15 545.4]	14 214.1 [11 524.3; 16 040.8]	−0.046	0.401 c	0.688
CER (24:1) (μM)	573.0 [463.0;653.8]	652.8 [544.0;835.9]	−0.277	0.001 c	0.003
SM (24:1) (μM)	31 586.6 [27 603.5; 35 510.3]	35 924.8 [31 117.2; 40 928.9]	−0.183	<0.001 c	0.003

Univariate analysis showed that increasing SM (16:0) levels (per SD) tended to increase the odds of low muscle strength (OR = 1.43, 95% CI: 0.99–2.08; P = 0.058). The AUROC of SM (16:0) for low muscle strength in men was 0.600 (95% CI: 0.501–0.699). Increasing SM (16:0), CER (24:1), and SM (24:1) levels (per SD) significantly increased the odds of low muscle mass with ORs of 1.76 (95% CI: 1.23–2.52), 2.04 (95% CI: 1.39–3.00), and 1.90 (95% CI: 1.31–2.77), respectively (all P < 0.01; Table 4). The AUROC of SM (16:0), CER (24:1), and SM (24:1) for low muscle mass in men was 0.647 (95% CI: 0.557–0.737), 0.669 (95% CI: 0.581–0.757), and 0.670 (95% CI: 0.582–0.759), respectively.

Table 4 Association of sphingolipid metabolites with low muscle strength and low muscle mass in the discovery cohort male (n = 144)

	OR (95% CI)	P	AUC (95% CI)	P
Low muscle strength
SM (16:0) per 1SD	1.43 (0.99–2.08)	0.058	0.600 (0.501–0.699)	0.049
CER (24:1) per 1SD	1.24 (0.86–1.77)	0.251	0.571 (0.464–0.677)	0.193
SM (24:1) per 1SD	1.23 (0.85–1.76)	0.270	0.563 (0.463–0.663)	0.214
Low muscle mass
SM (16:0) per 1SD	1.76 (1.23–2.52)	0.002	0.647 (0.557–0.737)	0.001
CER (24:1) per 1SD	2.04 (1.39–3.00)	<0.001	0.669 (0.581–0.757)	<0.001
SM (24:1) per 1SD	1.90 (1.31–2.77)	0.001	0.670 (0.582–0.759)	<0.001

SM (16:0), CER (24:1), and SM (24:1) were significantly inversely associated with SMI (all P < 0.05). After adjusting for weight, the statistical significance of the inverse association of CER (24:1) with SMI remained, as shown in Table 5. SM (16:0) tended to be inversely associated with HGS (P = 0.096). Multivariate analysis showed a significant association between CER (24:1) and low muscle mass.

Table 5 Association between sphingolipid metabolite levels and HGS and SMI in the discovery cohort males (n = 144)

	Univariate	Multivariate	Adjustment for weight
βa	SE	βb	P	βa	SE	βb	P	βa	SE	βb	P
SMI
SM (16:0) per 1SD	−0.205	0.075	−0.223	0.007	−0.129	0.079	−0.141	0.103	0.016	0.039	0.017	0.681
CER (24:1) per 1SD	−0.258	0.074	−0.281	0.001	−0.212	0.079	−0.231	0.008	−0.101	0.037	−0.110	0.007
SM (24:1) per 1SD	−0.187	0.075	−0.204	0.014					0.006	0.038	0.006	0.879
HGS
SM (16:0) per 1SD	−0.977	0.583	−0.139	0.096	−0.977	0.583	−0.139	0.096	−0.095	0.544	−0.014	0.862
CER (24:1) per 1SD	−0.444	0.587	−0.063	0.451					0.216	0.533	0.031	0.686
SM (24:1) per 1SD	−0.536	0.587	−0.076	0.363					0.260	0.538	0.037	0.629

Effect of CER treatment on myogenesis in vitro

Treatment with CER (16:0), CER (18:0), and CER (24:1) significantly decreased the MT area and fusion index (Figure S1). Treatment with CER (16:0), CER (18:0), and CER (24:1) also decreased the viability of muscle cells.

Association between sphingolipid metabolites in human plasma and sarcopenia in the validation cohort

The main characteristics of the 164 participants in the validation cohort (36 cases and 128 age-matched controls) from the AMC are presented in Table S6. Muscle mass parameters (lean mass, ASM, and SMI) in the case group were significantly lower than in the control group (all P < 0.001). HGS measurements in 74 of 164 (45.1%) participants in the validation cohort [control group: 50 of 128 (39.1%) and control group: 24 of 36 (66.7%)] revealed that HGS in the case group was significantly lower than in the control group (P = 0.013). The sarcopenia group also had significantly higher SM (16:0), CER (24:1), and SM (24:1) levels than the control group (all P < 0.005). Increasing CER (24:1) and SM (24:1) levels (per SD) significantly increased the odds of low muscle mass with ORs of 1.47 (95% CI: 1.01–2.15) and 2.85 (95% CI: 1.82–4.47), respectively (data not shown).

The predictive ability of the Sphingolipid (SphL) score for sarcopenia in both the discovery and validation cohorts

The SphL score, a diagnostic regression equation for sarcopenia based on multivariate linear regression analysis of CER (24:1) and SM (16:0) with SMI in the discovery cohort, was calculated as follows: SphL score = 7.216–0.212*log (CER (24:1) SD)-0.129*log (SM (16:0) SD). The corresponding ROC curve of the SphL score using a combination of CER (24:1) and SM (16:0) levels had an AUC of 0.712 (95% CI: 0.626–0.798) in the discovery cohort (Figure 3). The addition of the SphL score to low muscle strength (HGS < 28 kg) as a predictor of sarcopenia improved the AUC by 16.3%, from 0.646 (95% CI: 0.575–0.717; HGS only) to 0.751 (95% CI: 0.671–0.831, P = 0.002; HGS + SphL). The AUC of the SphL score was 0.695 (95% CI: 0.591–0.799) in the validation cohort. When analysing the 74 participants with HGS data in the validation cohort, the SphL score was significantly predictive of sarcopenia (AUC: 0.678; 95% CI: 0.541–0.816), although the HGS was not (AUC: 0.553; 95% CI: 0.470–0.636).

[IMAGE OMITTED. SEE PDF]

Discussion

In this study, we aimed to provide reliable metabolomic biomarkers for sarcopenia using mouse and cell models and two cohorts of human participants. Non-targeted metabolome profiling in the plasma of mice from an aging mouse model of sarcopenia showed that sphingolipid metabolism was related to sarcopenia (Figure 1). This result was confirmed by higher plasma CER (C24:1) levels and muscle CER (C18:1) content in aged mice, and both of plasma CER (24:1) levels and muscle CER (18:1) contents were inversely associated with relative muscle mass (Table 1). Moreover, most of CERs and SMs in CM decreased during muscle cell differentiation (Figure 2). Quantitation of sphingolipids from human plasma showed that the SM (16:0), CER (24:1), and SM (24:1) levels were higher in men with sarcopenia than in control participants (Table 3). However, there were no significant differences in sphingolipid metabolites in women with or without sarcopenia. ROC analysis showed the association of the plasma SM (16:0) level in men with lower HGS, which reflects muscle strength (Table 4). ROC analysis also showed the association of plasma SM (16:0), CER (24:1) and SM (24:1) in men with lower SMI, which reflects muscle mass (Table 4). The SphL score, a model for predicting sarcopenia using CER (24:1) and SM (16:0) levels, was significantly predictive for sarcopenia in men in two independent cohorts (Figure 3). The addition of SphL score to HGS in the discovery cohort significantly improved the predictive performance of the latter for sarcopenia, as shown by a 16.3% increase in AUC (Figure 3). To the best of our knowledge, our study is the first to suggest that SM (16:0), reflecting low muscle strength and low muscle mass, and CER (24:1), reflecting low muscle mass, could serve as biomarkers for sarcopenia in men.

The processes underlying sarcopenia pathogenesis are not fully understood and may involve multiple factors, such as a decline in mitochondrial and cellular protein homeostasis.²⁷ Importantly, evidence shows that sphingolipid metabolism perturbations are associated with sarcopenia development and progression of sarcopenia.^{13,17,20,22,28–33} For example, in vitro experimental studies suggest that CERs negatively regulate myogenic differentiation through their involvement in apoptosis, cellular stress, and cell growth arrest.^{17,29–31,33} Additionally, some in vivo experimental studies also demonstrate that sphingolipid metabolism could be involved in the development of sarcopenia and muscle atrophy.^17,22 Furthermore, sphingolipids contribute to muscle aging, and reducing them prevents age-related decline in muscle mass while enhancing strength and exercise capacity^20,21 by affecting mitochondria and protein homeostasis.^18,21 In humans, the muscle CER (16:0) content in a small sample of men (n = 19) was inversely correlated with lean leg mass,²² while plasma CER (16:0) levels in 340 clinically normal participants aged 70 years and older were inversely associated with gait parameters, including gait speed.¹³ Overall, these findings suggest that sphingolipids may play a direct role in sarcopenia and that sphingolipid metabolites may be useful as biomarkers of sarcopenia.

In the present study, plasma levels of three sphingolipid metabolites, SM (16:0), CER (24:1), and SM (24:1), were higher in men with sarcopenia than those without. Although not statistically significant, the higher long- and very-long-chain dihydroCER and serine levels might suggest that the canonical rather than the non-canonical pathway is associated with sarcopenia. The association between SM (16:0) and HGS was similar to that in a previous study, which showed an inverse association between CER (16:0) and gait speed.¹³ Consistent with our findings of an inverse association between SM (16:0), CER (24:1), and SM (24:1) and SMI, another recent study found that increased plasma CER (16:0), CER (24:0), SM (16:0), and SM (24:1) levels were a defining feature of the mouse and human cancer cachexia models; authors suggested that these sphingolipids may contribute to tissue wasting and skeletal muscle atrophy by inhibiting the anabolic signals.³⁴ The disparities between the two groups in humans, especially weight, would be suggested as a discriminator between sarcopenia and age-matched controls. However, (1) statistically persistent significance in the negative association of CER (24:1) level with SMI after adjustment for weight, (2) tendency for higher CER (24:1) levels in the aging mouse model of sarcopenia regardless of higher weight, and (3) no significant difference in the prevalence of comorbidities were observed; thus, differences in sphingolipid levels could be considered as biomarkers of disease status. Our prediction model for sarcopenia, calculated by combining CER (24:1) and SM (16:0) levels, significantly predicted sarcopenia in both cohorts. Adding a model for sarcopenia by combining CER (24:1) and SM (16:0) to HGS improved the discriminatory performance of HGS in the discovery cohort. If confirmed by another large independent cohort, these findings might support using CER (24:1) and SM (16:0) as biomarkers for sarcopenia.

Consistent with our findings, recent studies have shown that reducing sphingolipids, such as CER (24:1) and CER (18:1), accumulating in mouse skeletal muscle with age prevented age-related decline in muscle mass while enhancing strength and exercise capacity.^20,21 In line with findings of CERs as negative regulators of myogenic differentiation,^{17,29–31,33} some CERs and all SMs in CM decreased during muscle cell differentiation. Previous studies showed that inhibiting sphingolipid synthesis can mitigate age-related muscle decline and improve muscle function.^20,21 However, while our study focuses on identifying diagnostic biomarkers, the detrimental cellular effects on myogenesis and cell viability by CER (16:0), CER (18:0), and CER (24:1) observed in the present study and previous studies suggest the role of sphingolipids as a cause of sarcopenia. We recognize the necessity for further research to identify the specific CERs that are not only indicative of sarcopenia but may also serve as targets for treatment. This dual role of CERs as both biomarkers and therapeutic targets warrants deeper exploration in the context of age-related sarcopenia.

Consistent with previous studies showing metabolic differences between sexes,³⁵ we found statistically significant differences in sphingolipid metabolism according to sarcopenia status in men but not women. The target metabolome was selected based on the result generated by untargeted metabolomics using male mice only, which may result in metabolite changes more relevant to men. Untargeted metabolomics for female mice would be necessary to find female-relevant metabolome changes. Additionally, a broad range of target metabolomes must be explored to find common biomarkers for both men and women. Another conflicting finding occurred in the mouse studies, where, compared with young mice, aged mice had higher CER (24:1) but lower SM (18:0) plasma levels. Interestingly, some studies show that supplementation of SM in mice actually increases muscle weight³⁶ and extends swimming time,^36,37 which suggests that there may be SM acyl chain length-specific effects on muscle in mice. Nonetheless, this change was not seen in the human cohort, where the plasma SM (18:0) level tended to be higher in men with sarcopenia than those without, suggesting that sphingolipids may have species–specific effects on muscle in mice and humans.¹⁹ Increased plasma SM (16:0) and CER (24:1) levels in men with sarcopenia reflected increased CER species with fatty acid acyl chain length preferences toward long (C16:0) and very-long chains (C24:0 and C24:1), which CERS6 and CERS2, respectively, produce.²¹ These results align with previous studies that applied transcriptomic analysis of muscle biopsies from aged individuals and patients with muscle disorders.²¹ Another study reported the reduced expression of CERS1 and CERS5 in human skeletal muscles with aging.³⁸ Moreover, the inhibition of CERS1 in mice skeletal muscles and primary human muscle cells increased CER (16:0) and CER (24:1).³⁸

A major strength of our study is that it included a relatively large number of elderly men with matched controls and validation in an independent cohort. Additionally, animal and muscle cell experiments strengthen the robustness of the study's findings and its validity. However, this study has some limitations. First, the causal relationship between identified sphingolipids and sarcopenia cannot be inferred because of the cross-sectional study design. Second, age-matched animal models of sarcopenia would be preferable to the differently aged animals used in this study. We aimed to find the circulating biomarkers for sarcopenia in humans, so we could not measure sphingolipid metabolites in various tissues, including oxidative and glycolytic muscles and other organs. Third, the accuracy of BIA, which was used to diagnose sarcopenia as recommended by the AWGS, had limitations related to equipment accuracy and assessment conditions.³⁹ Despite this concern, the high diagnostic accuracy of BIA, aligning with dual-energy X-ray absorptiometry in a substantial subset of the validation cohort (62 of 164, 37.8%) by 93.5%, with a sensitivity of 100.0% (17 of 17) and specificity of 91.1% (41 of 45), suggested that BIA could be used to diagnose sarcopenia. Fourth, we did not measure HGS in all of the validation cohort participants; thus, the results relating to HGS might not be representative. Accumulating evidence shows the clinical relevance of functional parameters over muscle mass; therefore, recent guidelines emphasize the importance of measuring muscle strength or physical performance over muscle mass.^1,2 However, the association of SM (16:0) with HGS and risk of low muscle strength in the discovery cohort, and that of CER (16:0) with gait speed in another study,¹³ might support that sphingolipids are related to not only muscle mass but also muscle strength or physical performance. We did not define a specific tissue, including muscle types, responsible for the secretion of the sphingolipid metabolites in the plasma, although recent publications related to the specific increases of sphingolipid species in skeletal muscle or reduced/compensatory expression of CERS with aging might support our result.^21,38

Our study demonstrates that SM (16:0), reflecting low muscle strength and low muscle mass, and CER (24:1), reflecting low muscle mass, are potential circulating biomarkers for sarcopenia in men. Further research on sphingolipid metabolites is required to confirm these results and provide additional insights into the metabolomic changes relevant to the pathogenesis and diagnosis of sarcopenia.

Conflict of interest

Je Hyun Seo, Jung-Min Koh, Han Jin Cho, Hanjun Kim, Young-Sun Lee, Su Jung Kim, Pil Whan Yoon, Won Kim, Sung Jin Bae, Hong-Kyu Kim, Hyun Ju Yoo, and Seung Hun Lee declare that they have no conflict of interest.

Funding

This study was supported by the Asan Institute for Life Sciences Grant (grant number: 2023IP0041) and the National Research Foundation of Korea (NRF) grant, funded by the Korean government (Ministry of Science and ICT; grant numbers: 2022R1C1C1002929, 2022R1A2C1007901, and 2022R1A2C1003661). 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

Data availability statement

Data generated/analysed during the study are available from the corresponding author upon reasonable request.

References

Chen LK, Woo J, Assantachai P, Auyeung TW, Chou MY, Iijima K, et al. Asian working group for sarcopenia: 2019 consensus update on sarcopenia diagnosis and treatment. J. Am. Med. Dir. Assoc. 2020;21: [eLocator: e2].

Cruz‐Jentoft AJ, Bahat G, Bauer J, Boirie Y, Bruyere O, Cederholm T, et al. Sarcopenia: revised European consensus on definition and diagnosis. Age Ageing 2019;48:16–31.

Cruz‐Jentoft AJ, Sayer AA. Sarcopenia. Lancet 2019;393:2636–2646.

Norman K, Otten L. Financial impact of sarcopenia or low muscle mass ‐ a short review. Clin. Nutr. 2019;38:1489–1495.

Pugh TD, Conklin MW, Evans TD, Polewski MA, Barbian HJ, Pass R, et al. A shift in energy metabolism anticipates the onset of sarcopenia in rhesus monkeys. Aging Cell 2013;12:672–681.

Liu JC, Dong SS, Shen H, Yang DY, Chen BB, Ma XY, et al. Multi‐omics research in sarcopenia: current progress and future prospects. Ageing Res. Rev. 2022;76: [eLocator: 101576].

Korostishevsky M, Steves CJ, Malkin I, Spector T, Williams FM, Livshits G. Genomics and metabolomics of muscular mass in a community‐based sample of UK females. Eur. J. Hum. Genet. 2016;24:277–283.

Lustgarten MS, Price LL, Chale A, Phillips EM, Fielding RA. Branched chain amino acids are associated with muscle mass in functionally limited older adults. J. Gerontol. A Biol. Sci. Med. Sci. 2014;69:717–724.

Lustgarten MS, Price LL, Phillips EM, Kirn DR, Mills J, Fielding RA. Serum predictors of percent lean mass in young adults. J. Strength Cond. Res. 2016;30:2194–2201.

Miyamoto K, Hirayama A, Sato Y, Ikeda S, Maruyama M, Soga T, et al. A metabolomic profile predictive of new osteoporosis or sarcopenia development. Metabolites 2021;11:278.

Moaddel R, Fabbri E, Khadeer MA, Carlson OD, Gonzalez‐Freire M, et al. Plasma biomarkers of poor muscle quality in older men and women from the Baltimore longitudinal study of aging. J. Gerontol. A Biol. Sci. Med. Sci. 2016;71:1266–1272.

Murphy RA, Moore SC, Playdon M, Meirelles O, Newman AB, Milijkovic I, et al. Metabolites associated with lean mass and adiposity in older black men. J. Gerontol. A Biol. Sci. Med. Sci. 2017;72:1352–1359.

Wennberg AMV, Schafer MJ, LeBrasseur NK, Savica R, Bui HH, Hagen CE, et al. Plasma sphingolipids are associated with gait parameters in the Mayo Clinic study of aging. J. Gerontol. A Biol. Sci. Med. Sci. 2018;73:960–965.

Zhao Q, Shen H, Su KJ, Tian Q, Zhao LJ, Qiu C, et al. A joint analysis of metabolomic profiles associated with muscle mass and strength in Caucasian women. Aging (Albany NY) 2018;10:2624–2635.

Cajka T, Fiehn O. Toward merging untargeted and targeted methods in mass spectrometry‐based metabolomics and lipidomics. Anal. Chem. 2016;88:524–545.

Hannun YA, Obeid LM. Sphingolipids and their metabolism in physiology and disease. Nat. Rev. Mol. Cell Biol. 2018;19:175–191.

Tan‐Chen S, Guitton J, Bourron O, Le Stunff H, Hajduch E. Sphingolipid metabolism and signaling in skeletal muscle: from physiology to physiopathology. Front Endocrinol (Lausanne) 2020;11:491.

Lim EW, Handzlik MK, Trefts E, Gengatharan JM, Pondevida CM, Shaw RJ, et al. Progressive alterations in amino acid and lipid metabolism correlate with peripheral neuropathy in Polg(D257A) mice. Sci. Adv. 2021;7: [eLocator: eabj4077].

Park WJ, Park JW. The effect of altered sphingolipid acyl chain length on various disease models. Biol. Chem. 2015;396:693–705.

Laurila PP, Wohlwend M, Imamura de Lima T, Luan P, Herzig S, Zanou N, et al. Sphingolipids accumulate in aged muscle, and their reduction counteracts sarcopenia. Nat. Aging 2022;2:1159–1175.

Lima TI, Laurila PP, Wohlwend M, Morel JD, Goeminne LJE, Li H, et al. Inhibiting de novo ceramide synthesis restores mitochondrial and protein homeostasis in muscle aging. Sci. Transl. Med. 2023;15:eade6509.

Rivas DA, Morris EP, Haran PH, Pasha EP, Morais Mda S, Dolnikowski GG, et al. Increased ceramide content and NFkappaB signaling may contribute to the attenuation of anabolic signaling after resistance exercise in aged males. J. Appl. Physiol. 1985;2012:1727–1736.

Xie WQ, He M, Yu DJ, Wu YX, Wang XH, Lv S, et al. Mouse models of sarcopenia: classification and evaluation. J. Cachexia. Sarcopenia Muscle 2021;12:538–554.

Kim YA, Lee SH, Koh JM, Kwon SH, Lee Y, Cho HJ, et al. Fatty acid amides as potential circulating biomarkers for sarcopenia. J. Cachexia. Sarcopenia Muscle 2023;14:1558–1568.

Bligh EG, Dyer WJ. A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol. 1959;37:911–917.

Blazenovic I, Kind T, Ji J, Fiehn O. Software tools and approaches for compound identification of LC‐MS/MS data in metabolomics. Metabolites 2018;8:31.

Dao T, Green AE, Kim YA, Bae SJ, Ha KT, Gariani K, et al. Sarcopenia and muscle aging: a brief overview. Endocrinol Metab (Seoul) 2020;35:716–732.

Borsch A, Ham DJ, Mittal N, Tintignac LA, Migliavacca E, Feige JN, et al. Molecular and phenotypic analysis of rodent models reveals conserved and species‐specific modulators of human sarcopenia. Commun Biol. 2021;4:194.

De Larichaudy J, Zufferli A, Serra F, Isidori AM, Naro F, Dessalle K, et al. TNF‐alpha‐ and tumor‐induced skeletal muscle atrophy involves sphingolipid metabolism. Skelet. Muscle 2012;2:2.

Hyde R, Hajduch E, Powell DJ, Taylor PM, Hundal HS. Ceramide down‐regulates system a amino acid transport and protein synthesis in rat skeletal muscle cells. FASEB J. 2005;19:461–463.

Mebarek S, Komati H, Naro F, Zeiller C, Alvisi M, Lagarde M, et al. Inhibition of de novo ceramide synthesis upregulates phospholipase D and enhances myogenic differentiation. J. Cell Sci. 2007;120:407–416.

Sogaard D, Baranowski M, Larsen S, Taulo Lund M, Munk Scheuer C, Vestergaard Abildskov C, et al. Muscle‐saturated bioactive lipids are increased with aging and influenced by high‐intensity interval training. Int. J. Mol. Sci. 2019;20:1240.

Turpin SM, Lancaster GI, Darby I, Febbraio MA, Watt MJ. Apoptosis in skeletal muscle myotubes is induced by ceramides and is positively related to insulin resistance. Am. J. Physiol. Endocrinol. Metab. 2006;291:E1341–E1350.

Morigny P, Zuber J, Haid M, Kaltenecker D, Riols F, Lima JDC, et al. High levels of modified ceramides are a defining feature of murine and human cancer cachexia. J. Cachexia. Sarcopenia Muscle 2020;11:1459–1475.

Costanzo M, Caterino M, Sotgiu G, Ruoppolo M, Franconi F, Campesi I. Sex differences in the human metabolome. Biol. Sex Differ. 2022;13:30.

Ishida Y, Kiyokawa Y, Asai T, Oku N. Ameliorating effects of sphingomyelin‐based liposomes on sarcopenia in senescence‐accelerated mice. Biol. Pharm. Bull. 2016;39:786–793.

Haramizu S, Ota N, Otsuka A, Hashizume K, Sugita S, Hase T, et al. Dietary milk fat globule membrane improves endurance capacity in mice. Am. J. Physiol. Regul. Integr. Comp. Physiol. 2014;307:R1009–R1017.

Knudsen JR, Persson KW, Henriquez‐Olguin C, Li Z, Di Leo N, Hesselager SA, et al. Microtubule‐mediated GLUT4 trafficking is disrupted in insulin‐resistant skeletal muscle. Elife 2023;12: [eLocator: e83338].

Janssen I, Heymsfield SB, Baumgartner RN, Ross R. Estimation of skeletal muscle mass by bioelectrical impedance analysis. J. Appl. Physiol. 1985;2000:465–471.

von Haehling S, Morley JE, Coats AJS, Anker SD. Ethical guidelines for publishing in the Journal of Cachexia, Sarcopenia and Muscle: update 2021. J. Cachexia. Sarcopenia Muscle 2021;12:2259–2261.