ABSTRACT
Background
Sarcopenia is a common and serious complication of dialysis patients. In addition to the reduction in total muscle mass, fat infiltration of skeletal muscle as represented by the percentage of intermuscular adipose tissue (IMAT%) affects muscle quality and may have prognostic implications.
Methods
We investigated 86 incident peritoneal dialysis (PD) patients. Their total muscle area, skeletal muscle index and IMAT% at the level of third lumbar vertebra (L3) was measured by computer tomography. The relation with functional assessment, bioimpedance spectroscopy (BIS) parameters, survival and peritonitis rate were analyzed.
Results
L3 total muscle area and IMAT% had opposite but both significant correlation with functional scores (Morse Fall Scale and Norton Scale) and lean tissue mass measured by BIS, but IMAT% did not have correlation with adipose tissue mass by BIS. IMAT%, but not total muscle area, was independently associated with peritonitis-free survival (adjusted hazard ratio 1.084, 95% confidence interval (CI) 1.027 to 1.144, P = .004) and peritonitis rate (adjusted β = 1.091, 95% CI 1.032 to 1.155, P = .002). Neither total muscle area nor IMAT% was associated with patient survival or hospitalization rate.
Conclusion
IMAT% is an independent predictor of peritonitis-free survival and peritonitis rate in incident PD patients. Further studies are needed to validate our results and to develop convenient non-invasive methods for the assessment of IMAT%.
Keywords: frailty, malnutrition, renal failure, sarcopenia
KEY LEARNING POINTS.
What was known:
Sarcopenia is a prevalent and serious complication among dialysis patients and is linked to poor functional outcomes and higher morbidity.
Fat infiltration within skeletal muscle, such as increased intermuscular adipose tissue (IMAT%), may impair muscle quality, but its prognostic significance in peritoneal dialysis (PD) patients was not well established.
This study adds:
IMAT% is associated with worse functional performance and muscle quality in incident PD patients, independent of overall muscle area or total adipose tissue.
IMAT% is an independent predictor for both peritonitis-free survival and peritonitis rates.
Potential impact:
Identifying IMAT% as a predictor enables early risk stratification for peritonitis in PD patients.
Incorporation of IMAT% measurement into routine assessments could facilitate targeted interventions, reducing peritonitis-related complications and healthcare burdens in this vulnerable population.
INTRODUCTION
Peritoneal dialysis (PD) is a life-saving treatment for patients with end-stage kidney disease. As compared with hemodialysis, PD offers several advantages, including greater flexibility, improved quality of life and better preservation of residual kidney function [1, 2]. However, patients undergoing PD face an increased risk of peritonitis and sarcopenia [3, 4], both of which are major causes of morbidity and mortality.
Sarcopenia, characterized by low lean tissue mass (LTM) and reduced body mass index after 6 months of dialysis, is strongly associated with adverse clinical outcomes in dialysis patients, including a higher peritonitis incidence and increased all-cause mortality [5, 6]. Furthermore, PD patients who developed peritonitis tended to have a lower sarcopenic index after 1 year of dialysis, but not at baseline [5]. This suggests that a reduction in muscle mass following dialysis may predict peritonitis—either because frail, wasted PD patients are more susceptible to technical errors during dialysis exchanges, or because sarcopenia serves as a surrogate marker of immune system dysfunction. In addition to its physical function, it is now recognized that skeletal muscle is an active endocrine organ that secretes signaling molecules called myokines, which affect energy metabolism and inflammation, both locally and to distal organs [7, 8]. In CKD patients, the myokine profile is altered [8]. Specifically, growth differentiation factor-15 (GDF-15) and activin A are linked to muscle wasting [9–11]. GDF-15 triggers anorexia and nausea upon binding to its receptors in the brainstem [12], which limits the protein intake, while myostatin and activin A bind to the ActRII receptors on muscle cells to inhibit protein synthesis and suppress proliferation [13, 14]. Myoglobin is a marker of muscle breakdown and may serve as an indicator of muscle loss [15]. Irisin and follistatin are myogenic and may correlate with better muscle health [16, 17].
Although sarcopenia means the reduction in skeletal muscle bulk literally, recent studies have revealed that replacement of skeletal muscle fiber by adipose tissue, which may not result in the reduction in overall muscle size, has considerable clinical relevance [18]. The accumulation of adipose tissue within skeletal muscle, known as intermuscular and intramuscular adipose tissue (IMAT), can be quantified by computed tomography (CT) scans at the third lumbar vertebra (L3) [19]. In the general population, IMAT has been reported to be associated with various clinical conditions, including muscle dysfunction, systemic inflammation, insulin resistance and various infections [20–24]. In patients undergoing hemodialysis, IMAT levels are markedly elevated and have been linked to increased inflammation and reduced muscle quality [25]. On the other hand, the clinical significance of IMAT in PD patients remains poorly understood. In this study, we investigated the prognostic value of IMAT in PD patients.
MATERIALS AND METHODS
The study received approval from the Joint Chinese University of Hong Kong-New Territories East Cluster Clinical Research Ethics Committee (approval number CRE-2015.250 and CRE-2023.363). All procedures adhered to the guidelines outlined in the Declaration of Helsinki.
Patient selection
This prospective observational cohort study involves new adult PD patients from a single center between 2020 and 2022. Patients were excluded if they were using pacemakers or metallic prostheses, or had planned for living donor kidney transplants or transfers to other hospitals within 6 months. After obtaining the written informed consent, baseline assessments, including multi-frequency bioimpedance spectroscopy, dialysis adequacy and a range of laboratory tests, were performed about 1 month after patients became stabilized on PD.
Functional assessment
Functional assessment of patients, including the Morse Fall Scale and Norton Scale as previously described [26], was determined at baseline. In brief, the Morse Fall Scale is used to assess the patient’s risk of falling. It is composed of six categories, namely the history of falling, secondary diagnosis, usage of ambulatory aids, intravenous therapy, gait and mental status, with the total score ranging from 0 to 120 (higher scores indicate a higher risk of fall) [27]. The Norton Scale incorporates five elements, namely physical condition, mental condition, activity, mobility and incontinence, with a total score ranging from 5 to 10, with a lower score indicating a higher risk of pressure injury [28].
CT image analysis
Non-enhanced and contrast-enhanced abdominal CT scan were performed within 6 months after the initiation of dialysis, in the presence of dialysis solution dwell. All the scans were performed by the 64-slice GE Healthcare Light Speed VCT (GE Medical Systems, Uppsala, Sweden), with a 0.35-s gantry rotation speedand a slice thickness of 0.625 mm. Our abdominal CT protocol aimed for a 120 kVp setting, with specific Noise Index values ranging from 15 to 22 Hounsfield units (HU) depending on the patient size. The system was calibrated to define water as 0 HU and air as –1000 HU. A single slice at the L3 level of each patient was analyzed by the ImageJ software version 1.54 p (National Institutes of Health, Bethesda, MD, USA). The cross-sectional skeletal muscle area and the IMAT at L3 were obtained via manual labeling and threshold adjustment. Skeletal muscle density was identified by values ranging from –30 HU to +150 HU; adipose tissue, including IMAT, was identified in the range of –190 HU to –30 HU. IMAT was then converted into the percentage of the total cross-sectional area of muscle at L3 for subsequent analysis (IMAT%). The muscle segment contains the rectus abdominis, abdominal wall, psoas and paraspinal muscle groups. The skeletal muscle index was calculated by dividing the total muscle area by the square of body height.
Serum myokine profile
Serum levels of a myokine panel were measured by enzyme-linked immunosorbent assay (ELISA) kits and performed as per the manufacturers’ instructions. In this study, we measured the serum concentrations of GDF-15, activin A, irisin, myostatin, follistatin and myoglobin as recommended by previous studies [29]. The details of the ELISA kits used in this study are summarized in Supplementary data, Table S1.
Small solute clearance and other laboratory indices
Evaluation of dialysis small solute clearance was conducted via a 24-h collection of dialysate and urine, as previously described [30], and the total weekly Kt/V was computed using the standard formula. The residual glomerular filtration rate (GFR) was calculated by averaging the 24-h urinary urea and creatinine clearances [31]. The bromocresol purple method was used to analyze the serum albumin levels [32]. Additional laboratory tests, such as plasma biochemistry and hemoglobin levels, were conducted as part of routine clinical care.
Multi-frequency bioimpedance spectroscopy
In this study, we used bioimpedance spectroscopy (BIS) study to assess the LTM and adipose tissue mass (ATM) in PD patients as previously described [33]. Briefly, electrodes were positioned on the patient’s right hand and foot in a supine position. Measurements were performed and recorded using the Body Composition Monitor (BCM, Fresenius Medical Care, Germany). Moreover, assessment was also made of overhydration volume, total body water, extracellular water and intracellular water. The extracellular-to-intracellular volume ratio was also calculated. All body composition measurements were carried out with PD fluid dwell, since our earlier research showed that peritoneal dialysate had inconsequential effects on the measurements [34].
Follow-up and outcome parameters
This study cohort was followed until 1 July 2024. During the follow-up period, the treatment decisions were determined by the individual clinician in charge of patient care, without interference from the study. The primary outcome measures included peritonitis-free survival and the peritonitis rate. Secondary outcome measures included patient survival, technique survival, number of hospital admissions and total duration of hospitalizations. In this study, technique failure was defined as a transfer to hemodialysis, receiving kidney transplant or mortality. Transfers to other medical centers or regaining renal function were deemed censoring events. The peritonitis rate was presented as the number of peritonitis episodes per patient-year [35].
Statistical analysis
Statistical analysis were performed through SPSS (version 28.0, IBM Corporation, Armonk, NY, USA) and GraphPad Prism (version 10.1.1, GraphPad Software, San Diego, CA, USA). Data are shown as mean ± standard deviation or median (interquartile range), unless stated otherwise. Pearson’s and Spearman’s correlation coefficients were utilized to assess the relationship between each parameter. The Kaplan–Meier method was employed to present data regarding the peritonitis-free survival rates. Multivariate Cox regression models were constructed to adjust for potential clinical confounders, which, in addition to all six myokines, included Charlson comorbidity score, serum albumin and residual GFR because of their well reported relation with clinical outcome [36–38]. Poisson regression analysis was used to analyze the peritonitis rate, while the log-linear regression was used to analyze the correlation with the hospitalization rate and the duration of hospitalization because the assumption of Poisson’s distribution that mean equals variance was not fulfilled. A P-value of <.05 was considered statistically significant. All probabilities were two-tailed.
RESULTS
We identified 205 new PD patients during study period. After excluding patients for clinical factors, missing data and samples, 86 patients were analyzed (Supplementary data, Fig. S1). Their baseline demographic, clinical and biochemical characteristics, as well as the serum myokine profile of the study population, are summarized in Tables 1 and 2. The baseline L3 total muscle area was 122.3 ± 32.2 cm2 in males and 91.8 ± 19.0 cm2 in females (P < .0001); the L3 skeletal muscle index was 43.9 ± 11.8 cm2/m2 in males and 38.6 ± 7.7 cm2/m2 in females (P < .0001). The overall baseline IMAT% was 22.6 ± 6.1%, and IMAT% was significantly lower in male than female patients (21.1 ± 5.4% vs 25.0 ± 6.4%, P = .003). IMAT% has a significant but only modest correlation with L3 total muscle area (r = –0.521, P < .0001) and L3 skeletal muscle index (r = –0.414, P < .0001). The internal correlation between serum myokine levels is summarized in Supplementary data, Table S2.
Table 1:
Baseline demographic and clinical characteristics.
| No. of patients | 86 |
|---|---|
| Sex (male:female) | 54:32 |
| Age (years) | 62.3 ± 11.7 |
| Body build | |
| Weight (kg) | 65.0 ± 14.5 |
| Height (cm) | 162.3 ± 8.9 |
| Body mass index (kg/m2) | 24.5 ± 4.2 |
| Blood pressure (mmHg) | |
| Systolic | 139.9 ± 20.5 |
| Diastolic | 77.1 ± 13.4 |
| Renal diagnosis, no. of case (%) | |
| Diabetic nephropathy | 42 (48.8) |
| Glomerulonephritis | 21 (24.4) |
| Hypertension | 9 (10.5) |
| Urological problem | 5 (5.8) |
| Polycystic kidney disease | 2 (2.3) |
| Others or unknown | 7 (8.1) |
| Major comorbidities, no. of case (%) | |
| Diabetes mellitus | 50 (58.1) |
| Coronary artery disease | 15 (17.4) |
| Cerebrovascular accident | 8 (9.3) |
| Peripheral vascular disease | 5 (5.8) |
| Charlson’s comorbidity index | 6.01 ± 2.44 |
| Morse Fall Scalea | 45 (35 to 60) |
| Norton Scalea | 19 (17 to 19) |
| Type of PD, no. of cases (%) | |
| Machine-assisted PD | 19 (22.1) |
| Low GDP solution | 16 (18.6) |
| Glucose polymer solution | 30 (34.9) |
Date expressed as mean ± standard deviation or amedian (interquartile range). GDP, glucose degradation product.
Table 2:
Baseline biochemical, bioimpedance and nutritional characteristics.
| Myokine profile | |
|---|---|
| GDF-15 (ng/mL) | 5.34 (3.47 to 7.99) |
| Activin A (pg/mL) | 149.5 (57.4 to 676.1) |
| Irisin (ng/mL) | 113.8 (54.0 to 190.5) |
| Myoglobin (ng/mL) | 334.8 (240.3 to 502.7) |
| Myostatin (ng/mL) | 2.28 (1.26 to 3.26) |
| Follistatin (ng/mL) | 1.44 (1.16 to 1.85) |
| Bioimpedance spectroscopy | |
| LTM (kg) | 41.4 (31.6 to 48.8) |
| ATM (kg) | 18.7 (9.2 to 25.7) |
| Overhydration (L) | 3.3 (1.8 to 5.4) |
| E/I ratio | 1.0 (0.9 to 1.1) |
| Fasting plasma glucose (mmol/L) | 7.9 (6.1 to 11.8) |
| Insulin (µU/mL) | 3.6 (0.9 to 10.7) |
| HOMA-IR | 1.7 (0.3 to 4.7) |
| C-reactive protein (mg/dL) | 1.0 (0.5 to 2.3) |
| Hemoglobin (g/dL) | 10.6 (9.5 to 11.5) |
| Serum albumin (g/L) | 28.1 (25.5 to 31.8) |
| Total weekly Kt/V | 2.1 (1.7 to 2.4) |
| Residual GFR (mL/min/1.73 m2) | 4.5 (2.6 to 6.4) |
| NPNA (g/kg/day) | 1.1 (0.9 to 1.3) |
| FEBM (kg) | 21.1 (18.3 to 24.9) |
| Peritoneal transport characteristics | |
| D/P4 | 0.62 (0.53 to 0.72) |
| MTAC creatinine (mL/min/1.73 m2) | 8.14 (6.00 to 12.19) |
| Arterial pulse wave velocity (cm/s) | |
| Carotid-radial | 10.7 (9.6 to 11.8) |
| Carotid-femoral | 11.3 (9.8 o 12.5) |
Data are presented as median (interquartile range). IMAT%, inter- and intramuscular fat percentage; D/P4, dialysate-to-plasma creatinine at 4 h; E/I ratio, extracellular to intracellular volume ratio; HOMA-IR, homeostatic model assessment of insulin resistance index; Kt/V, dialysis adequacy; MTAC, mass transfer area coefficient; NPNA, normalized protein nitrogen appearance; FEBM, fat-free-edema-free body mass; PWV, pulse-wave velocity.
Relation with clinical factors
The relation between the baseline L3 muscular parameters and serum myokine levels is summarized in Fig. 1, and the correlation with other clinical factors is summarized in Table 3. In essence, the L3 total muscle area significantly correlated with Norton Scale, LTM and myostatin level, and inversely with the Morse Fall Scale, after adjusting for multiple comparisons, while the L3 muscle index did not have a significant correlation with any clinical parameters. IMAT% did not have a significant correlation with the ATM (r = 0.068, P = .543), but it correlated significantly with age, body weight, height and Morse Fall Scale, and inversely with the LTM, serum albumin, myoglobin and myostatin levels. IMAT% was similar between patients with and without diabetes (22.3 ± 6.2% vs 22.8 ± 6.0%, respectively, P = .707), as well as patients with and without CVD (23.4 ± 8.8% vs 22.4 ± 5.4%, respectively, P = .687). The correlation of serum myokine levels and other clinical factors is summarized in Supplementary data, Table S3. In short, LTM had a significant correlation with myoglobin (r = 0.411, P = .003) and myostatin levels (r = 0.457, P = .0003).
Figure 1:
Correlation of the baseline L3 total muscle area, muscle index and the IMAT% with the serum level of myostatin, GDF-15, activin A, irisin, myoglobin and follistatin. Correlation was analyzed by Spearman’s rank correlation coefficient.
Table 3:
Correlation between the L3 muscular parameters with clinical characteristics.
| L3 total muscle area | L3 muscle index | IMAT% | ||||
|---|---|---|---|---|---|---|
| Spearman’s r | P-values | Spearman’s r | P-values | Spearman’s r | P-values | |
| Age | –0.297 | .056 | –0.215 | .311 | 0.357 | .014 |
| Weight | 0.402 | .001 | 0.155 | .339 | –0.328 | .014 |
| Height | 0.425 | .001 | 0.039 | .844 | –0.394 | .002 |
| Body mass index | 0.175 | .209 | 0.123 | .431 | –0.102 | .591 |
| Systolic blood pressure | 0.090 | .505 | 0.134 | .412 | –0.014 | .945 |
| Diastolic blood pressure | 0.216 | .136 | 0.171 | .311 | –0.158 | .327 |
| Charlson’s comorbidity index | –0.086 | .507 | –0.052 | .815 | 0.172 | .280 |
| Morse Fall Scale | –0.316 | .018 | –0.189 | .311 | 0.301 | .029 |
| Norton Scale | 0.355 | .008 | 0.264 | .232 | –0.274 | .052 |
| LTM | 0.523 | <.0001 | 0.270 | .232 | –0.460 | .0002 |
| ATM | –0.050 | .680 | –0.084 | .604 | 0.068 | .771 |
| Overhydration | 0.080 | .520 | –0.030 | .844 | –0.071 | .771 |
| E/I ratio | –0.258 | .056 | –0.228 | .311 | 0.235 | .100 |
| Fasting plasma glucose | 0.050 | .680 | –0.031 | .844 | 0.084 | .708 |
| Insulin | –0.166 | .247 | –0.201 | .311 | –0.003 | .975 |
| HOMA-IR | –0.126 | .417 | –0.189 | .311 | 0.025 | .945 |
| C-reactive protein | –0.192 | .192 | –0.179 | .311 | 0.208 | .168 |
| Hemoglobin | –0.103 | .483 | –0.123 | .431 | 0.106 | .591 |
| Serum albumin | 0.145 | .329 | 0.085 | .604 | –0.311 | .024 |
| Weekly Kt/V | –0.097 | .494 | 0.006 | .955 | 0.151 | .344 |
| Residual GFR | 0.201 | .171 | 0.134 | .412 | –0.047 | .830 |
| NPNA | 0.203 | .171 | 0.186 | .311 | –0.013 | .945 |
| FEBM | 0.337 | .011 | 0.155 | .339 | –0.228 | .119 |
| D/P4 | –0.013 | .909 | –0.014 | .936 | 0.050 | .830 |
| MTAC | –0.118 | .417 | –0.046 | .828 | 0.147 | .344 |
| Carotid-radial PWV | 0.178 | .209 | 0.167 | .322 | –0.018 | .945 |
| Carotid-femoral PWV | 0.116 | .417 | 0.087 | .604 | 0.063 | .771 |
E/I, extracellular to intracellular volume ratio; HOMA-IR, homeostatic model assessment of insulin resistance; GFR, glomerular filtration rate; NPNA, normalized protein nitrogen appearance; FEBM, fat-free-edema-free body mass; D/P4, dialysate to plasma ratio at 4 h; MTAC, mass transfer area coefficient; PWV, pulse wave velocity. P-values were adjusted by the Benjamini–Hochberg method.
Peritonitis
During the study period, there were 67 peritonitis episodes in 34 patients; 52 patients (60.5%) were peritonitis-free. The overall peritonitis rate was 0.61 episodes per patient-year follow-up. The peritonitis-free survival rate was significantly associated with IMAT%, but not L3 total muscle area or L3 skeletal index, by univariate Cox analysis (Table 4). The Kaplan–Meier estimate of 2-year peritonitis-free survival rates for IMAT% quartile I to IV (with quartile IV having the highest values) were 89.1%, 72.2%, 71.8% and 51.3%, respectively (P = .0002) (Fig. 2). The association between IMAT% and peritonitis-free survival remained significant after adjusting for confounding factors by multivariable Cox regression analysis (adjusted hazard ratio 1.084, 95% CI 1.027 to 1.144, P = .004).
Table 4:
Cox regression model for peritonitis-free survival.
| Univariate | Multivariate | ||||
|---|---|---|---|---|---|
| Unadjusted HR | P-values | Adjusted HR | 95% CI | P-values | |
| IMAT% | 1.094 | .0004 | 1.100 | 1.026 to 1.180 | .007 |
| GDF-15a | 1.203 | .502 | 0.799 | 0.446 to 1.432 | .451 |
| Activin Aa | 1.050 | .509 | 1.071 | 0.915 to 1.253 | .393 |
| Irisina | 2.245 | .002 | 2.455 | 1.388 to 4.341 | .002 |
| Myoglobina | 1.789 | .107 | 2.596 | 1.141 to 5.906 | .023 |
| Myostatina | 0.506 | .030 | 1.125 | 0.483 to 2.616 | .785 |
| Follistatina | 1.474 | .401 | 1.211 | 0.444 to 3.300 | .709 |
| Charlson’s index | 1.095 | .224 | 0.987 | 0.829 to 1.175 | .884 |
| Serum albumin | 0.922 | .066 | 0.952 | 0.848 to 1.069 | .403 |
| Residual GFR | 0.944 | .409 | 0.972 | 0.833 to 1.133 | .713 |
aData were log transformed for the analysis. HR, hazard ratio.
Figure 2:
Kaplan–Meier plots of the peritonitis-free survival in PD patients. Patients were divided into quartiles according to their baseline IMAT%, with quartile 1 having the lowest value. Data were compared by the Cox analysis with the IMAT% quartile as the independent variable.
Similarly, the peritonitis rate significantly correlated with IMAT% but not L3 total muscle area or L3 skeletal index by Poisson regression (Table 5). After adjusting for confounding clinical factors, IMAT% remained significantly associated with the peritonitis rate (adjusted β = 1.091, 95% CI 1.032 to 1.155, P = .002) (Table 5). The peritonitis rates for IMAT% quartile I, II, III and IV were 0.16 ± 0.11, 0.19 ± 0.08, 0.42 ± 0.14 and 1.71 ± 0.99 episodes per patient-year follow-up, respectively (P = .003) (Fig. 3). In this analysis, none of the myokines had an independent association with peritonitis rate.
Table 5:
Multivariate Poisson regression analysis of the peritonitis rate.
| Unadjusted β | P-values | Adjusted β | 95% CI | P-values | |
|---|---|---|---|---|---|
| IMAT% | 1.066 | .0009 | 1.069 | 1.012 to 1.130 | .016 |
| GDF-15a | 0.890 | .555 | 0.580 | 0.364 to 0.924 | .022 |
| Activin Aa | 1.060 | .277 | 1.070 | 0.966 to 1.185 | .194 |
| Irisina | 1.574 | .004 | 1.388 | 1.014 to 1.898 | .040 |
| Myoglobina | 1.089 | .023 | 1.986 | 1.172 to 3.365 | .011 |
| Myostatina | 0.685 | .077 | 0.905 | 0.523 to 1.567 | .722 |
| Follistatina | 1.632 | .116 | 1.398 | 0.697 to 2.803 | .346 |
| Charlson’s index | 1.079 | .106 | 1.028 | 0.920 to 1.149 | .626 |
| Serum Albumin | 0.985 | .639 | 1.033 | 0.959 to 1.114 | .390 |
| Residual GFR | 0.933 | .157 | 0.959 | 0.861 to 1.069 | .452 |
aData were log-transformed for the analysis.
Figure 3:
Peritonitis rate of patients from different IMAT% quartiles. Data was compared by the Kendall’s tau test. Error bars represent the standard error of the mean (SEM).
Patient and technique survival
During a median follow-up of 30.0 (interquartile range 20.2 to 35.6) months, 23 patients died. The causes of death were peritonitis (5 cases), non-peritonitis infections (9 cases), ischemic heart disease (5 cases), cerebrovascular accident (2 cases), termination of dialysis (1 case) and malignancy (1 case). During the same period, 11 patients were converted to hemodialysis, 5 patients received kidney transplantation and 1 patient recovered. None of the L3 muscle parameters was associated with patient or technique survival (Supplementary data, Table S4). Among the six myokines, serum myostatin level was associated with patient survival rate by univariate analysis (Supplementary data, Table S4), but the association was not significant after adjusting for other confounding factors by multivariable Cox regression analysis.
Hospitalization
During the study period, there were 323 hospital admissions for a total of 3051 days. The hospitalization rate did not have a significant correlation with the L3 total muscle area (P = .524), L3 skeletal muscle index (P = .948) or the IMAT% (P = .520). Similarly, the duration of hospital stay, adjusted for the duration of follow-up, did not have a significant correlation with the L3 total muscle area (P = .901), L3 skeletal muscle index (P = .727) or the IMAT% (P = .302).
DISCUSSION
In this study, we evaluated 86 incident PD patients and examined the clinical significance of CT-derived parameters of muscle mass at L3, including total muscle area, skeletal muscle index and IMAT%. Both total muscle area and IMAT% showed significant associations with functional assessments such as the Morse Fall Scale and Norton Scale, as well as with lean tissue mass measured by bioimpedance spectroscopy. Notably, IMAT% did not correlate with adipose tissue mass, although both are markers of body fat content. Our findings highlight the limitations of non-invasive assessment of body fat content, as these methods do not differentiate intermuscular adipose tissue from other types of fat.
The relationships between CT-derived parameters and myokines are complex. In summary, both the L3 total muscle area and the muscle index showed positive correlations with myostatin and myoglobin levels, while IMAT% was inversely related to myostatin and myoglobin but not the level of other myokines. Since myoglobin is a marker of muscle wasting [15] and myostatin inhibits muscle growth [39], these findings may seem counterintuitive at first. However, they could be reasonable in the context of PD. Because muscle wasting is common in PD patients [40], a larger muscle mass may release more myostatin, which could account for the observed positive correlation with the L3 total muscle area and the muscle index. Since the IMAT% is inversely related to the overall muscle mass (r = –0.460, P = .0002), it is reasonable to reveal a negative correlation between IMAT% and myostatin. In addition, a previous study in PD patients has found that, in contrast to findings in the general population, myostatin is positively associated with muscle mass [39], further supporting our observations. For the same reason, since myoglobin is primarily released from muscle cells, the baseline serum myoglobin level is expected to correlate positively with the L3 muscle area and the muscle index but negatively with IMAT%. A previous study in PD patients has found that, in contrast to findings in the general population, myostatin is positively associated with muscle mass [41]. Our observation suggests that the regulatory roles of these myokines in PD patients may be distinct from patients without kidney disease, underscoring the need for further research.
While numerous previous studies have identified the L3 muscle cross-sectional area/index and IMAT% as predictors of all-cause mortality in various patient populations [42–44], our study found no significant association with either patient survival or technique survival in incident PD patients. This discrepancy may be partly due to our smaller sample size, which limits statistical power. However, we believe a more plausible explanation is that PD patients have distinct risk profiles and underlying pathophysiology, so traditional risk factors may not hold the same prognostic value in this group.
Our study found that IMAT% was independently associated with peritonitis-free survival, whereas neither absolute LTM nor L3 muscle area showed any association, despite IMAT% being significantly related to these measures. This contrasts with previous literature that emphasized muscle mass as protective against peritonitis [5]. Our results suggest that IMAT% may be a superior predictor of peritonitis risk, potentially because it reflects the detrimental effects of pro-inflammatory adipose tissue on the immune system [18, 22]. In addition, a high IMAT% may also indicate a poor muscle quality and a compromised general health, which may also contribute to the peritonitis risk [45–47]. In this study, IMAT% was also identified as an independent risk factor for peritonitis rate, while none of the myokines showed an independent association. Taken together, these findings underscore the importance of measuring IMAT% for peritonitis risk stratification.
While our study demonstrated the prognostic value of IMAT%, several limitations should be noted. Most importantly, there are numerous parameters and correlations included in this analysis, which raises concerns about overfitting and model stability. With a limited number of patients and events, the analysis included too many parameters relative to the sample size, undermining the reliability of multi-parameter estimates and increasing the risk of spurious associations. In addition, IMAT% was measured using manual segmentation of CT images with ImageJ, a process that is time-consuming, expensive and subject to inter-observer variability. As CT imaging is expensive and requires exposure to radiation, it may not be suitable for serial measurements [5]. Future studies should focus on developing automated methods for IMAT% measurement to replace manual segmentation and exploration of alternative imaging techniques to reduce cost and facilitate serial measurement. In this regard, ultrasonography (USG) may serve as a potential alternative approach as it is non-invasive, less expensive and does not have radiation. A previous study in critically ill patients showed that IMAT% could be measured by USG [48]. Nonetheless, the results of USG-measured IMAT% have not been validated with the gold standard CT or MRI measurement, and the role of USG-measured IMAT% has not been explored in dialysis patients.
In conclusion, our study is the first to evaluate the clinical significance of IMAT% in PD patients. We showed that IMAT% had an independent association with peritonitis-free survival and the overall peritonitis rate. Our findings contribute to the development of risk-stratification models for PD patients.
Supplementary Material
Contributor Information
Lixing Xu, Carol & Richard Yu Peritoneal Dialysis Research Centre, Department of Medicine & Therapeutics, Prince of Wales Hospital, NT, Hong Kong, China; Li Ka Shing Institute of Health Sciences (LiHS), Faculty of Medicine, The Chinese University of Hong Kong, Shatin, NT, Hong Kong, China.
Jack Kit-Chung Ng, Carol & Richard Yu Peritoneal Dialysis Research Centre, Department of Medicine & Therapeutics, Prince of Wales Hospital, NT, Hong Kong, China; Li Ka Shing Institute of Health Sciences (LiHS), Faculty of Medicine, The Chinese University of Hong Kong, Shatin, NT, Hong Kong, China.
Winston Wing-Shing Fung, Carol & Richard Yu Peritoneal Dialysis Research Centre, Department of Medicine & Therapeutics, Prince of Wales Hospital, NT, Hong Kong, China.
Gordon Chun-Kau Chan, Carol & Richard Yu Peritoneal Dialysis Research Centre, Department of Medicine & Therapeutics, Prince of Wales Hospital, NT, Hong Kong, China.
Wing-Fai Pang, Carol & Richard Yu Peritoneal Dialysis Research Centre, Department of Medicine & Therapeutics, Prince of Wales Hospital, NT, Hong Kong, China.
Kai-Ming Chow, Carol & Richard Yu Peritoneal Dialysis Research Centre, Department of Medicine & Therapeutics, Prince of Wales Hospital, NT, Hong Kong, China.
Cheuk-Chun Szeto, Carol & Richard Yu Peritoneal Dialysis Research Centre, Department of Medicine & Therapeutics, Prince of Wales Hospital, NT, Hong Kong, China; Li Ka Shing Institute of Health Sciences (LiHS), Faculty of Medicine, The Chinese University of Hong Kong, Shatin, NT, Hong Kong, China.
AUTHORS’ CONTRIBUTIONS
Research idea and study design: L.X.X., J.K.-C.N., C.-C.S.; data acquisition: J.K.-C.N., W.W.-S.F., G.C.-K.C.; data analysis/interpretation: L.X.X., J.K.-C.N., C.-C.S.; statistical analysis: L.X.X., J.K.-C.N., C.C.S.; supervision or mentorship: K.-M.C., W.-F.P.; manuscript preparation: L.X.X., C.-C.S. Each author contributed important intellectual content during manuscript drafting or revision and accepts accountability for the overall work by ensuring that questions pertaining to the accuracy or integrity of any portion of the work are appropriately investigated and resolved.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflict of interest.
FUNDING
This study was supported by the Richard Yu Chinese University of Hong Kong (CUHK) PD Research Fund, and CUHK research accounts 6905134, 6906662 and 8601286. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
DATA AVAILABILITY STATEMENT
The data underlying this article are publicly available at the following site: https://github.com/szetocc/IMAT_p1.
References
- 1. Auguste BL, Bargman JM. Peritoneal dialysis prescription and adequacy in clinical practice: core curriculum 2023. Am J Kidney Dis. 2023;81:100–9. 10.1053/j.ajkd.2022.07.004 [DOI] [PubMed] [Google Scholar]
- 2. Jung HY, Jeon Y, Park Y et al. Better quality of life of peritoneal dialysis compared to hemodialysis over a two-year period after dialysis initiation. Sci Rep. 2019;9:10266. 10.1038/s41598-019-46744-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Kang SH, Kim AY, Do JY. Association of sarcopenia and its components with clinical outcomes in patients undergoing peritoneal dialysis. Kidney Res Clin Pract. 2022;41:741–52. 10.23876/j.krcp.21.278 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Dzekova-Vidimliski P, Nikolov IG, Gjorgjievski N et al. Peritoneal dialysis-related peritonitis: rate, clinical outcomes and patient survival. Pril (Makedon Akad Nauk Umet Odd Med Nauki). 2021;42:47–55. 10.2478/prilozi-2021-0034. [DOI] [PubMed] [Google Scholar]
- 5. Do JY, Kang SH. Association between peritonitis and low muscle mass in peritoneal dialysis patients. J Ren Nutr. 2020;30:341–6. 10.1053/j.jrn.2019.09.003 [DOI] [PubMed] [Google Scholar]
- 6. Kojima D, Washida N, Uchiyama K et al. The body mass index change is associated with death or hemodialysis transfer in Japanese patients initiating peritoneal dialysis. Ren Fail. 2023;45:2163904. 10.1080/0886022X.2022.2163904 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Picciotto D, Maccio L, Verzola D et al. Pathophysiology of physical exercise in kidney patients: unveiling new players—the role of myokines. Kidney Blood Press Res. 2024;49:457–71. 10.1159/000539489 [DOI] [PubMed] [Google Scholar]
- 8. Quiroga B, Diez J. The kidney-skeletal muscle-heart axis in chronic kidney disease: implications for myokines. Nephrol Dial Transplant. 2025;40:255–63. 10.1093/ndt/gfae193 [DOI] [PubMed] [Google Scholar]
- 9. Elkina Y, von Haehling S, Anker SD et al. The role of myostatin in muscle wasting: an overview. J Cachexia Sarcopenia Muscle. 2011;2:143–51. 10.1007/s13539-011-0035-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Ling T, Zhang J, Ding F et al. Role of growth differentiation factor 15 in cancer cachexia (Review). Oncol Lett. 2023;26:462. 10.3892/ol.2023.14049 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Kawaguchi Y, Kataoka Y, Okamoto K et al. Activin A is deeply involved in the progression of sarcopenia and leads to poor prognosis in lung cancer patients. Sci Rep. 2025;15:13641. 10.1038/s41598-025-97907-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Li J, Hu X, Xie Z et al. Overview of growth differentiation factor 15 (GDF15) in metabolic diseases. Biomed Pharmacother. 2024;176:116809. 10.1016/j.biopha.2024.116809 [DOI] [PubMed] [Google Scholar]
- 13. Bataille S, Chauveau P, Fouque D et al. Myostatin and muscle atrophy during chronic kidney disease. Nephrol Dial Transplant. 2021;36:1986–93. 10.1093/ndt/gfaa129 [DOI] [PubMed] [Google Scholar]
- 14. Rodgers BD, Ward CW. Myostatin/activin receptor ligands in muscle and the development status of attenuating drugs. Endocr Rev. 2022;43:329–65. 10.1210/endrev/bnab030 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Weber MA, Kinscherf R, Krakowski-Roosen H et al. Myoglobin plasma level related to muscle mass and fiber composition: a clinical marker of muscle wasting?. J Mol Med. 2007;85:887–96. 10.1007/s00109-007-0220-3 [DOI] [PubMed] [Google Scholar]
- 16. Zhao C, Wu Y, Zhu S et al. Irisin protects musculoskeletal homeostasis via a mitochondrial quality control mechanism. Int J Mol Sci. 2024;25:10116. 10.3390/ijms251810116 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Zhu J, Li Y, Lu A et al. Follistatin improves skeletal muscle healing after injury and disease through an interaction with muscle regeneration, angiogenesis, and fibrosis. Am J Pathol. 2011;179:915–30. 10.1016/j.ajpath.2011.04.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Addison O, Marcus RL, Lastayo PC et al. Intermuscular fat: a review of the consequences and causes. Int J Endocrinol. 2014;2014:1. 10.1155/2014/309570 [DOI] [Google Scholar]
- 19. Guo X, Cao N, Deng X et al. Intermuscular adipose tissue affected muscle density more than intramuscular adipose tissue content with opportunistic screening at abdominal CT. Sci Rep. 2025;15:8172. PMID: 40059242; PMCID: PMC11891326. 10.1038/s41598-025-85946-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Jones HG, Kopinke D, Meyer GA. Intramuscular adipose tissue: from progenitor to pathology. Am J Physiol Cell Physiol. 2025;329:C1268–82. 10.1152/ajpcell.00613.2025 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Zoico E, Rossi A, Di Francesco V et al. Adipose tissue infiltration in skeletal muscle of healthy elderly men: relationships with body composition, insulin resistance, and inflammation at the systemic and tissue level. J Gerontol Ser A Biol Sci Med Sci. 2010;65A:295–9. 10.1093/gerona/glp155 [DOI] [Google Scholar]
- 22. Lim JP, Chong MS, Tay L et al. Inter-muscular adipose tissue is associated with adipose tissue inflammation and poorer functional performance in central adiposity. Arch Gerontol Geriatr. 2019;81:1–7. 10.1016/j.archger.2018.11.006 [DOI] [PubMed] [Google Scholar]
- 23. Kamo N, Kaido T, Miyachi Y et al. Preoperative abnormal body composition is closely related to bacteremia after living donor liver transplantation. Nutrition. 2020;77:110798. 10.1016/j.nut.2020.110798 [DOI] [PubMed] [Google Scholar]
- 24. Akazawa N, Kishi M, Hino T et al. The degree of recovery in swallowing ability in older inpatients with aspiration pneumonia is related to intramuscular adipose tissue of the quadriceps than to muscle mass. PLoS One. 2022;17:e0275810. P. 10.1371/journal.pone.0275810 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Dilaver RG, Demirci M, Crescenzi R et al. Intermuscular adipose tissue and muscle function in patients on maintenance hemodialysis. Physiol Rep. 2025;13:e70363. 10.14814/phy2.70363 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Cooper L, Siam B, Sagee A et al. Some nursing screening tools can be used to assess high-risk older adults who undergo colorectal surgery for cancer. Clin Interv Aging. 2020;15:1505–11. 10.2147/CIA.S258992 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Kim E, Seol EM, Lee HJ. The association of body mass index on falls risk and mortality in hospitalized patients of different old-age categories requiring nutritional support. Clin Nutr Res. 2024;13:96–107. 10.7762/cnr.2024.13.2.96 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Natanzon SS, Maor E, Klempfner R et al. Norton score and clinical outcomes following acute decompensated heart failure hospitalization. J Cardiol. 2020;76:335–41. 10.1016/j.jjcc.2020.05.016 [DOI] [PubMed] [Google Scholar]
- 29. Ladang A, Beaudart C, Reginster J-Y et al. Biochemical markers of musculoskeletal health and aging to be assessed in clinical trials of drugs aiming at the treatment of sarcopenia: consensus paper from an expert group meeting organized by the European Society for Clinical and Economic Aspects of Osteoporosis, Osteoarthritis and Musculoskeletal Diseases (ESCEO) and the Centre Académique de Recherche et d’Expérimentation en Santé (CARES SPRL), under the auspices of the World Health Organization Collaborating Center for the Epidemiology of Musculoskeletal Conditions and Aging. Calcif Tissue Int. 2023;112:197–217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Szeto CC, Wong TY, Chow KM et al. Oral sodium bicarbonate for the treatment of metabolic acidosis in peritoneal dialysis patients: a randomized placebo-control trial. J Am Soc Nephrol. 2003;14:2119–26. 10.1097/01.ASN.0000080316.37254.7A [DOI] [PubMed] [Google Scholar]
- 31. van Olden RW, Krediet RT, Struijk DG et al. Measurement of residual renal function in patients treated with continuous ambulatory peritoneal dialysis. J Am Soc Nephrol. 1996;7:745–50. 10.1681/ASN.V75745 [DOI] [PubMed] [Google Scholar]
- 32. Delanghe S, Biesen WV, Velde NV et al. Binding of bromocresol green and bromocresol purple to albumin in hemodialysis patients. Clin Chem Lab Med. 2018;56:436–40. 10.1515/cclm-2017-0444 [DOI] [PubMed] [Google Scholar]
- 33. Moissl UM, Wabel P, Chamney PW et al. Body fluid volume determination via body composition spectroscopy in health and disease. Physiol Meas. 2006;27:921–33. 10.1088/0967-3334/27/9/012 [DOI] [PubMed] [Google Scholar]
- 34. Parmentier SP, Schirutschke H, Schmitt B et al. Influence of peritoneal dialysis solution on measurements of fluid status by bioimpedance spectroscopy. Int Urol Nephrol. 2013;45:229–32. 10.1007/s11255-012-0216-y [DOI] [PubMed] [Google Scholar]
- 35. Chan GC, Ng JK, Chow KM et al. Depression does not predict clinical outcome of Chinese peritoneal dialysis patients after adjusting for the degree of frailty. BMC Nephrol. 2020;21:329. 10.1186/s12882-020-01994-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Fried L, Bernardini J, Piraino B. Charlson comorbidity index as a predictor of outcomes in incident peritoneal dialysis patients. Am J Kidney Dis. 2001;37:337–42. 10.1053/ajkd.2001.21300 [DOI] [PubMed] [Google Scholar]
- 37. Mehrotra R, Duong U, Jiwakanon S et al. Serum albumin as a predictor of mortality in peritoneal dialysis: comparisons with hemodialysis. Am J Kidney Dis. 2011;58:418–28. 10.1053/j.ajkd.2011.03.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Perl J, Bargman JM. The importance of residual kidney function for patients on dialysis: a critical review. Am J Kidney Dis. 2009;53:1068–81. 10.1053/j.ajkd.2009.02.012 [DOI] [PubMed] [Google Scholar]
- 39. Esposito P, Picciotto D, Battaglia Y et al. Basic biology to clinical application. Adv Clin Chem. 2022;106:181–234. PMID: 35152972. [DOI] [PubMed] [Google Scholar]
- 40. Xu L, Ng JK, Chan GC et al. Early change in lean tissue mass after peritoneal dialysis: prevalence, risk factors, and clinical implications. BMC Nephrol. 2025;26:411. 10.1186/s12882-025-04350-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Silva MZC, Vogt BP, Reis NSC et al. Serum myostatin levels are associated with physical function and hospitalization in peritoneal dialysis patients. Eur J Clin Nutr. 2023;77:292–4. 10.1038/s41430-022-01227-x [DOI] [PubMed] [Google Scholar]
- 42. Weijs PJ, Looijaard WG, Dekker IM et al. Low skeletal muscle area is a risk factor for mortality in mechanically ventilated critically ill patients. Crit Care. 2014;18:R12. 10.1186/cc13189 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Portal D, Hofstetter L, Eshed I et al. L3 skeletal muscle index (L3SMI) is a surrogate marker of sarcopenia and frailty in non-small cell lung cancer patients. Cancer Manag Res. 2019;11:2579–88. 10.2147/CMAR.S195869 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Miljkovic I, Kuipers AL, Cauley JA et al. Greater skeletal muscle fat infiltration is associated with higher all-cause and cardiovascular mortality in older men. J Gerontol A Biol Sci Med Sci. 2015;70:1133–40. 10.1093/gerona/glv027 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Marcus RL, Addison O, LaStayo PC. Intramuscular adipose tissue attenuates gains in muscle quality in older adults at high risk for falling. A brief report. J Nutr Health Aging. 2013;17:215–8. 10.1007/s12603-012-0377-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Waters DL. Intermuscular adipose tissue: a brief review of etiology, association with physical function and weight loss in older adults. Ann Geriatr Med Res. 2019;23:3–8. 10.4235/agmr.19.0001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Xi F, Zheng N, Xiong B et al. Skeletal muscle density as a new predictor of abdominal infection in abdominal trauma patients. Clin Nutr. 2024;43:209–17. 10.1016/j.clnu.2024.10.035 [DOI] [PubMed] [Google Scholar]
- 48. Molinger J, Whittle J, Endeman H et al. A novel ultrasound approach in assessing IMAT in critically ill patients. Front Anesthesiol. 2024;3:1458633. 10.3389/fanes.2024.1458633. [DOI] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data underlying this article are publicly available at the following site: https://github.com/szetocc/IMAT_p1.