This cross-sectional retrospective study was conducted at Inoue Hospital (Osaka, Japan) (21). Among 334 MHD patients who underwent measurements of muscle mass, muscle strength, and physical performance from January 2019 to December 2019, those who had been receiving therapy for <6 months (n = 3) or with missing data regarding nutritional status (n = 12) were excluded. Furthermore, patients with malignancy, acute illness, infection, or liver cirrhosis (n = 4) were further excluded from the present analysis. As a result, 315 patients undergoing MHD (199 males, 116 females) were retrospectively analyzed in the present study.

Body weight refers to “dry weight” in the present study. Body mass index (BMI) was defined as body weight in kilograms divided by the square of height in meters. Diabetes mellitus was defined based on the American Diabetes Association criteria or treatment for diabetes mellitus (22). Past history of cerebrovascular disease was defined as prior occurrence of intracerebral hemorrhage, transient ischemic attack, subarachnoid hemorrhage, and/or ischemic stroke (23, 24).

Owing to the retrospective nature of the investigation, the Inoue Hospital Ethics Committee waived the need/requirement for obtaining informed consent (approval no. 236). All data subjected to analysis were collected from relevant patient medical charts and records, after approval of the study protocol. The present study was conducted in full accordance with the principles of the Declaration of Helsinki.

Blood was drawn from each patient with or without fasting before the start of the first MHD session of the week. The level of albumin in serum was analyzed using a bromocresol purple method, while that of the serum standard (low-sensitive) C-reactive protein (CRP) was analyzed using a latex agglutination immunoassay (reference range; 0–4.5 mg/L) at the hospital. Other standard laboratory parameters including levels of total cholesterol, creatinine, and hemoglobin were analyzed by use of a routine laboratory method at the hospital (23, 24).

Nutritional status was assessed using NRI score (10), which was calculated based on creatinine, total cholesterol, serum albumin, and BMI (Table 1).

Based on their NRI score, the enrolled patients were divided into the low-risk (score 0–7) and medium-/high-risk (score 8–13) groups (10).

As a part of routine clinical care for MHD patients at Inoue Hospital, muscle mass, muscle strength, and physical performance are generally measured after the end of MHD treatment at the beginning of the week of their birthday (21). The participants were informed that they were scheduled to be evaluated in regard to muscle mass, muscle strength, and physical performance on a day when undergoing treatment prior to the scheduled test date. Muscle strength and physical performance were measured by 10 different examiners, who received training and standardized their methods in advance.

Muscle mass was assessed using a whole body dual-energy X-ray absorptiometry system (Horizon A; Hologic Inc, MA, USA) (25). Appendicular skeletal muscle was considered as an extremity fat-free mass lacking bone mineral content. Height-adjusted muscle mass was calculated as appendicular skeletal muscle in kilograms divided by height in meters squared (kg/m²) (15).

Muscle strength was assessed with a handgrip strength test using a Smedley-type hand dynamometer (T.K.K.5101; Takei, Niigata, Japan), with the examination conducted by an experienced research staff member. The patients were instructed to apply maximum force in an upright position and hold the grip with their arm extended (26, 27). Measurements were repeated two times with both hands, and the maximum value (kg) was recorded.

A five-time chair stand test was conducted to assess physical performance based on the AWGS 2019 criteria (20). Each patient was instructed to stand and sit five times as fast as possible from a sitting position on a chair with a seat height of 40 cm and straight back without armrests, with their hands on their shoulders and arms crossed across the chest, as previously reported in the CHAIR (change hemodialysis patients' activity and impaired functions by chair stand exercise) study (28). They initiated the five-time chair standing test when an experienced research staff member spoke the word “start,” then the time (seconds) to complete the fifth repetition was recorded (29, 30).

Based on the AWGS 2019 criteria, low muscle mass was defined as height-adjusted muscle mass of <7.0 kg/m² for males and <5.4 kg/m² for females, low muscle strength as handgrip strength of <28 kg for males and <18 kg for females, and low physical performance as a five-time chair stand test time of ≥12 seconds for both males and females (20).

For diagnosis of sarcopenia and severe sarcopenia, the AWGS 2019 criteria were used (20). In brief, subjects with low muscle mass, plus either low muscle strength or low physical performance were diagnosed as having sarcopenia, while those with low muscle mass, plus both low muscle strength and low physical performance were diagnosed as having severe sarcopenia.

Values are expressed as the percent or median (interquartile range). A chi-squared test (categorical variables) and Mann–Whitney's U test (continuous variables) were used to compare variables between groups. Spearman's correlation coefficient was used to determine correlations between variables. Serum CRP values below the detection limit (0.1 mg/L) were handled as 0.1 mg/L. Before performing multivariable regression analysis, the serum level of CRP was logarithmically transformed (log) to follow a normal distribution. Multivariable linear regression analysis was used to determine whether the NRI score was independently associated with muscle mass, muscle strength, or physical performance, following adjustments using known risk factors, including age, gender, hemodialysis duration, presence of diabetes mellitus, history of cerebrovascular disease, use of intravenous and/or oral vitamin D, and hemoglobin and CRP levels. Multivariable logistic regression analyses were performed to determine whether NRI score, or a finding of medium- or high-risk assessed by the NRI was independently associated with presence of low muscle mass, low muscle strength, low physical performance, sarcopenia or severe sarcopenia, after adjustments using the known risk factors noted above. All statistical analyses were performed using the Statistical Package for the Social Sciences software package (IBM SPSS Statistics for Windows, 27.0, IBM Corp., Armonk, NY, USA). All reported P values are two-tailed, with a value < 0.05 considered to indicate statistical significance.

Table 2 presents characteristics of the enrolled patients (n = 315). The rates of prevalence of low serum creatinine level, low serum total cholesterol level, high serum total cholesterol level, low serum albumin level, and low BMI based on the NRI score cut-off point were 38.1% (n = 120), 21.0% (n = 66), 7.3% (n = 23), 65.7% (n = 207), and 28.6% (n = 90), respectively. The median NRI score was 5.0 and the prevalence of patients in the medium-/high-risk group (NRI score ≥8) was 31.1% (n = 98). The prevalence ratios for low muscle mass, low muscle strength, and low physical performance were 55.9% (n = 176), 60.6% (n = 191), and 31.4% (n = 99), respectively, while those for sarcopenia and severe sarcopenia were 44.1% (n = 139) and 20.0% (n = 63), respectively. Finally, as compared to female subjects, male subjects had significantly (P < 0.001) higher values for height-adjusted muscle mass and handgrip strength, and significantly (P < 0.05) lower values for five-time chair stand test time (Table 2). However, the frequency of sarcopenia and severe sarcopenia was not significantly different between male and female subjects.

The primary disease related to renal failure in the enrolled patients was diabetic nephropathy in 91 (28.9%), chronic glomerulonephritis in 86 (27.3%), nephrosclerosis in 38 (12.1%), polycystic kidney disease in 20 (6.3%), graft loss in 17 (5.4%), unclassifiable nephritis in 7 (2.2%), congenital anomaly in the kidneys or urinary tract in 4 (1.3%), kidney or urinary tract tumor in 3 (1.0%), pregnancy-induced hypertension in 3 (1.0%), malignant hypertension in 2 (0.6%), autoimmune nephritis in 2 (0.6%), urinary tract obstruction in 2 (0.6%), chronic pyelonephritis in 1 (0.3%), acute kidney injury in 1 (0.3%), other disease in 4 (1.3%), and unknown etiology in 34 (10.8%).

NRI score showed weak to moderate though significant correlation with height-adjusted muscle mass (muscle mass; ρ = −0.359, P < 0.001), handgrip strength (muscle strength; ρ = −0.361, P < 0.001), and five-time chair stand test time (physical performance; ρ = 0.407, P < 0.001). To further examine whether NRI score is independently associated with those factors, multivariable linear regression analyses were performed (Table 3), which showed the score to be significantly associated with muscle mass (β = −0.310, P < 0.001), muscle strength (β = −0.200, P < 0.001), and physical performance (β = 0.195, P = 0.001), while age and gender also showed a significant (P < 0.01) association.

NRI score was significantly higher in patients with low muscle mass [7.0 (4.0–8.0) vs. 4.0 (1.0–7.0), P < 0.001], low muscle strength [7.0 (4.0–8.0) vs. 4.0 (0.3–5.8), P < 0.001], and low physical performance [7.0 (5.0–9.0) vs. 4.0 (3.0–7.0), P < 0.001], as compared to patients without those factors present. Furthermore, the medium-/high-risk group (n = 98) showed a significantly (P < 0.01) higher prevalence for low muscle mass, low muscle strength, and low physical performance as compared to the low-risk group (n = 217; Table 2, Figure 1). To examine whether NRI score and medium-/high-risk were independently associated with those factors in the patient cohort, multivariable logistic regression analyses were performed (Table 4). The results showed that NRI score was significantly associated with low muscle mass [odds ratio (OR) 1.260, 95% confidence interval (CI) 1.157–1.374; P < 0.001], low muscle strength (OR 1.310, 95% CI 1.178–1.457; P < 0.001), and low physical performance (OR 1.216, 95% CI 1.104–1.339; P < 0.001). Furthermore, when NRI score was replaced by presence of medium- or high-risk, medium-/high-risk status retained a significant association with low muscle mass (OR 2.141, 95% CI 1.219–3.760; P = 0.008), low muscle strength (OR 7.665, 95% CI 3.438–17.091; P < 0.001), and low physical performance (OR 2.570, 95% CI 1.401–4.716; P = 0.002). Finally, when use of lipid-lowering medications was added to the multivariable logistic regression model, NRI score and medium-/high-risk status remained significantly associated with low muscle mass, low muscle strength, and low physical performance (Supplementary Table S1).

NRI score was significantly higher in patients with sarcopenia [7.0 (4.0–9.0) vs. 4.0 (2.0–7.0), P < 0.001] as well as those with severe sarcopenia [8.0 (5.0–10.0) vs. 4.0 (3.0–8.0), P < 0.001], as compared to those without. Furthermore, the medium-/high-risk group demonstrated a significantly (P < 0.001) higher prevalence for sarcopenia and severe sarcopenia, as compared to the low-risk group (Table 2, Figure 1). To examine whether NRI score and medium-/high-risk based on NRI score were independently associated with sarcopenia and severe sarcopenia in the cohort, multivariable logistic regression analyses were performed (Table 5). NRI score was shown to be significantly associated with both sarcopenia (OR 1.255, 95% CI 1.143–1.377; P < 0.001) and severe sarcopenia (OR 1.257, 95% CI 1.122–1.407; P < 0.001). Furthermore, when NRI score was replaced by medium-/high-risk, a significant association with sarcopenia (OR 2.960, 95% CI 1.623–5.401; P < 0.001) as well as severe sarcopenia (OR 2.241, 95% CI 1.151–4.362; P = 0.018) was retained. Finally, when use of lipid-lowering medications was added to the multivariable logistic regression model, NRI score and medium-/high-risk status remained significantly associated with sarcopenia and severe sarcopenia (Supplementary Table S2).