Atrophy

Nephrol Dial Transplant (2003) 18: 2074–2081 DOI: 10.1093/ndt/gfg325 Original Article Atrophy of non-locomotor muscle in patients with end-stage renal failure 1 Centre for Biophysical and Clinical Research into Human Movement and 2Exercise Physiology Group, Manchester Metropolitan University, Alsager, Stoke-on-Trent, UK, 3Institute for Fundamental and Clinical Human Movement Sciences, Vrije University, Amsterdam, The Netherlands and 4Directorate of Renal Medicine, North Staffordshire Hospital Trust, Stoke-on-Trent, UK Abstract Background. All previous histological studies of skeletal muscles of patients with renal failure have used locomotor muscle biopsies. It is thus unclear to what degree the observed abnormalities are due to the uraemic state and how much is due to disuse. The present study was undertaken to attempt to investigate this question by examining a non-locomotor muscle (rectus abdominis) in patients with end-stage renal failure. Methods. Biopsies from rectus abdominis were obtained from 22 renal failure patients (RFPs) undergoing surgical Tenchkoff catheter implantation for peritoneal dialysis and 20 control subjects undergoing elective abdominal surgery. Histochemical staining of frozen sections and morphometric analysis was used to estimate the proportion of each fibre type, muscle fibre area and capillary density. Myosin heavy chain composition was examined by SDS–PAGE. Results. There were no differences in fibre type distribution between RFPs and controls. All RFPs showed fibre atrophy [mean cross-sectional area (CSA) 3300 ± 1100 mm2, compared to 4100 ± 1100 mm2 in controls (P < 0.05)]. All fibre types were smaller in mean CSA in RFPs than in controls (15, 26 and 28% for types I, IIa and IIx, respectively). These differences could not be accounted for by differences in age, gender or cardiovascular or diabetic comorbidity. Muscle fibre capillarization, expressed as capillaries per fibre Correspondence and offprint requests to: P. F. Naish, MB FRCP, Renal Medicine, North Staffordshire Hospital Trust, Princes Road, Stoke-on-Trent ST4 7LN, UK. Email: patrick. naish@nstaffsh.wmids.nhs.uk *Present address: University of California, San Francisco, San Francisco General Hospital, Department of Medicine, San Francisco, CA, USA. yPresent address: Department of Biomedical Sciences, University Medical School, Forsterhill, Aberdeen, UK. or capillary contacts per fibre, was significantly less in RFPs. Conclusions. Since a non-locomotor muscle was examined, the effects of disuse as a cause of atrophy have been minimized. It is likely, therefore, that the decreased muscle fibre CSA and capillary density of RFPs compared to controls were due predominantly to uraemia itself. Keywords: atrophy; non-locomotor muscle; renal failure Introduction Muscle wasting and weakness are common in chronic and end-stage renal failure and are associated with physiological impairments and limitations in activities of daily living [1]. An interrelated set of causal factors in these circumstances includes decreased protein/calorie intake, disuse atrophy and disordered muscle protein balance that favours catabolism [2]. In addition, anaemia and neuropathy (uraemic or diabetic) may contribute to functional impairment [3]. All previous studies of skeletal muscle morphology in renal failure have examined upper or lower limb (i.e. locomotor) muscle biopsies [4–10]. It is possible that the abnormalities observed may have been due both to the effects of uraemia and to disuse atrophy. The purpose of the present study, therefore, was to characterize the degree of abnormality found in a non-locomotor muscle of patients with renal failure, thereby largely eliminating the potential influence of disuse atrophy. It was hypothesized that the histochemical and morphometric profile of the renal failure muscle would be abnormal in comparison with age-matched controls free of renal failure. Nephrol Dial Transplant 18(10) ß ERA–EDTA 2003; all rights reserved Downloaded from https://academic.oup.com/ndt/article/18/10/2074/1807592 by guest on 23 February 2023 Giorgos K. Sakkas1,*, Derek Ball1,y, Thomas H. Mercer2, Anthony J. Sargeant1,3, Keith Tolfrey2 and Patrick F. Naish4 Non-locomotor muscle atrophy in renal failure Subjects and methods Patients made using haematoxylin and eosin, Harris’s haematoxylin. To determine the percentage of types I, IIa and IIx fibres, acid-labile myofibrillar ATPase was used [14], with preincubation at pH 4.40 and 4.75. For identifying the muscle fibre capillaries, an -amylase–PAS stain was employed [15]. The morphometric analysis was performed with a camera (E.A.S.Y. 429 K) connected to a light-microscope (Zeiss axioskop) and digitizer (Mitsubishi, Video Copy Processor, Herolab: molecular technique software) connected to a PC. The cross-sectional area (CSA) of the muscle fibres from each sample was determined, the fibre type distribution was determined from sections stained for myofibrillar ATPase and the number of fibres per square millimetre was calculated from the sections stained by the haematoxylin method. A minimum number of 100 fibres was studied from each sample. Morphometric analysis was performed without prior knowledge of the group origin of the biopsy. Atrophic fibres were classified as all those fibres with a CSA of <50% of the control mean CSA, which was specific for each fibre type: 1729 mm2 for type I, 2545 mm2 for type IIa and 1749 mm2 for type IIx. Capillary profile Capillaries were quantified manually from photographs taken from the previously described image analysis system. An area containing at least 150 fibres in each section was selected for capillary counting. Analysis of capillary profiles included the capillary-to-fibre ratio (C:F), i.e. the ratio of the number of capillaries occurring with 100 fibres; capillary density per square millimetre of muscle tissue (CD/mm2); and capillary contact per fibre (CC/F), i.e. the number of capillaries that were adjacent to a single fibre and thus used as a measure of aerobic potential of skeletal muscle. All the capillary measurements were determined from at least 100 fibres, except CD/mm2, which included all the fibres within a specified area (mm2). All transversely cut capillaries were counted; if a capillary was sectioned longitudinally, it was counted as one each time it crossed a junction between three or more muscle fibres. Muscle biopsy A biopsy (200–500 mg wet tissue) of rectus abdominis muscle was obtained from each RFP undergoing Tenchkoff catheter insertion for peritoneal dialysis [13] and from each control subject undergoing abdominal surgery. The biopsy specimen was kept moist on a piece of gauze that was moistened with normal saline. The specimen was cut into two parts at right angles in relation to the direction of muscle fibres; one part was used for histological and histochemical analysis and the remainder for determination of myosin heavy chain (MyHC) composition by gel electrophoresis. Histological and histochemical analysis The biopsy materials were rapidly frozen in isopentane (
155 C) for 10 s and stored in liquid nitrogen (
196 C). Serial sections of the samples were obtained by cutting transverse sections of the biopsy (10 mm thickness in a cryostat Leica CM1800, Germany) at
25 C and collected on glass slides treated with poly-L-lysine. All sections were stained on the same day of cutting, with a time delay of 2 h. A minimum of three serial sections per patient were used for histological examination; the morphological examination was MyHC analysis: one-dimensional gel electrophoresis The proportion of MyHC isoforms present in the whole muscle sample was determined as follows. Muscle specimens (30–50 mg) were homogenized (1 mg of muscle tissue per 10 ml of buffer solution) in a buffer solution composed of 10% glycerol (w/v), 5% 2-mercaptoethanol (w/v), 2.3% SDS (w/v), 62.5 mM Tris and 0.001% bromophenol blue (w/v), pH 6.8 (corrected by HCl), and incubated at 80 C for 10 min. Samples were then centrifuged at 12 000 g at 4 C for 5 min. The supernatant fractions were collected and stored separately (
80 C) and later analysed by SDS–PAGE. Vertical PAGE (0.75 mm thickness) in the presence of SDS was performed according to the method of Talmadge and Roy [16]. MyHCs were analysed in high-glycerol-containing gels (30%). MyHC isoforms from the rectus abdominis muscle were resolved into three separate bands, MyHC I and IIa and IIx, in 8 and 4% polyacrylamide for separating and stacking gels, respectively. The ratio of acrylamide and bisacrylamide in stock was 50:1. The gels were run for 22 h at a constant current of 20 mA and at temperature of 4 C. The relative content of MyHC isoforms was determined by densitometry using gels stained with Coomassie blue (R250). The protein Downloaded from https://academic.oup.com/ndt/article/18/10/2074/1807592 by guest on 23 February 2023 The renal failure patient (RFP) group consisted of 22 predialysis patients (12 women and 10 men), aged 54.7 ± 14.3 (mean ± SD) years. Patients who agreed to take part were recruited sequentially. They had reached end-stage renal failure (serum creatinine 6.4 ± 2.1 mg/dl, blood urea nitrogen 14.3 ± 4.5 mg/dl) and were studied at the time of catheter insertion for peritoneal dialysis, just prior to the institution of dialysis. The primary causes of renal failure were renovascular disease in five cases, reflux nephropathy in four, diabetic nephropathy in three, small kidneys of unknown cause in three, glomerulonephritis in three, and amyloid in one. Twelve patients had no comorbidity, eight had cardiovascular comorbidity (ischaemic heart disease, peripheral vascular disease, left ventricuslar dysfunction) and five were diabetic. The nutritional status of the patients was assessed by the subjective global assessment (SGA) method, using a sevenpoint scale [11]. The SGA involved a careful history and clinical examination. The accuracy, reproducibility and validity of this method have previously been shown in dialysis patients [12]. Blood samples were taken for measurement of creatinine, albumin concentration, haemoglobin and parathyroid hormone. The control (CON) group consisted of 20 subjects (10 women and 10 men), aged 58.5 ± 16.9 (mean ± SD) years, undergoing elective abdominal surgery for various conditions such as hernia repair and elective cholecystectomy, and none had infective, malignant or inflammatory conditions. Blood samples taken as a routine hospital procedure were analysed for creatinine, albumin and haemoglobin. All subjects were given both an oral and written explanation of the purpose and procedures of the study prior to participation and gave written informed consent. All procedures were approved by the Local Research Ethics Committee. 2075 2076 G. K. Sakkas et al. bands were identified according to their apparent molecular masses compared to the migration of high-molecular-weight protein markers (C 3312, Sigma, Poole, UK). The densitometric system used Herolab software as described above, and the coefficient of variation of the gels analysis was 4.87%. Statistical analysis Results Haematological and biochemical evaluation The RFP group showed significantly higher serum creatinine concentration (t40 ¼ 11.8, P < 0.01) and lower albumin (t40 ¼
2.8, P < 0.01) and haemoglobin concentration (t40 ¼
9.4, P < 0.01) than the CON group (Table 1). Muscle fibre type composition: histochemical analysis In the rectus abdominis muscle of the CON group, the dominant fibre types were types I and IIa, which together accounted for 88% of all fibres present, and 12% of the fibres were identified as type IIx (Table 2). A similar pattern was found in the RFP group, in which types I, IIa and IIx fibres accounted for 49, 43 and 8%, respectively. 2 analyses indicated that no significant differences existed between the groups when comparing the fibre type distribution (P > 0.05) or MyHC isoform composition: SDS–PAGE Similar to the histochemical analysis of fibre type composition, the dominant MyHC isoforms were types I and IIa, which accounted for 90% of the total MyHC in both groups. 2 analysis revealed no statistically significant differences between the groups concerning the three different types of MyHC isoforms (P > 0.05) (Table 2). Muscle fibre CSA Type I fibres were revealed to be significantly smaller in CSA than type IIa fibres (F2,39 ¼ 19.654, P < 0.05). This corresponded to a 32 and 22% smaller CSA for the CON and RFP groups, respectively (Figure 1). Type IIx fibres, which accounted for 5–7% of all fibres, also had a smaller mean CSA compared to type IIa in both groups (F2,39 ¼ 19.654, P < 0.05) (Figure 1). All RFPs showed a general muscle fibre atrophy, and the mean CSA was 21% smaller (3300 ± 1100 mm2) than in the CON group (4100 ± 1100 mm2, F1,43 ¼ 7.5, P < 0.05); hence, the number of fibres per square millimetre of muscle was higher (288 ± 54 vs 254 ± 78, F1,43 ¼ 2.069, P ¼ 0.05) in the RFP group than in the CON group. The fibre population most affected in the RFP group was the type IIa fibres, which were 26% smaller than in the CON group (3800 ± 1600 vs 5100 ± 1500 mm2), and the type IIx fibre CSA was 28% smaller than in the CON group (2500 ± 1700 vs 3500 ± 1200 mm2, F1,40 ¼ 7.5, P < 0.05) (Figure 1). Additional analysis of the renal failure data demonstrated that those patients with some form of comorbidity had smaller type I fibres than those with renal failure alone (2518 ± 786 vs 2892 ± 1120 mm2, P ¼ 0.029). There were, however, no differences in the size of the other types of muscle fibres between RFPs with and without some form of comorbidity or diabetes (Table 5). Nutritional status The SGA showed that seven of the 22 RFPs were malnourished (B or Bþ). No significant differences in Table 2. Fibre type distribution (%) and relative myosin heavy chain (MyHC) content (%) in the renal failure patient (RFP, n ¼ 22) and control (CON, n ¼ 20) groups based on histochemical and SDS–PAGE analyses Table 1. Haematological and biochemical variables in the renal failure patient (RFP) and control (CON) groups RFP Creatinine (60–120 mmol/l) Albumin (35–50 g/dl) Haemoglobin (13–18 g/dl) Mean venous TCO2 (24–32 mmol/l) Parathyroid hormone (6–20 pmol/l) CON a 567 ± 187 37 ± 2a 10 ± 1a 22 ± 3 19 ± 13 a Significant differences between groups (P < 0.01). TCO2, total bicarbonate. 92 ± 25 42 ± 6 14 ± 1 – – RFP Fibre type Type I Type IIa Type IIx MyHC content MyHC I MyHC IIa MyHC IIx CON 49 ± 9 43 ± 11 8±8 50 ± 14 38 ± 14 12 ± 12 48.6 ± 11.9 40.6 ± 14.1 10.8 ± 13.2 47.3 ± 14.9 42.6 ± 14.5 10.1 ± 10.9 Values are presented as means ± SD. Downloaded from https://academic.oup.com/ndt/article/18/10/2074/1807592 by guest on 23 February 2023 All analyses were carried out by using the statistical package SPSS 9.0 for Windows (SPSS, Chicago, IL). Standard descriptive statistics, consisting of means ± SDs, were used to characterize the subject population. 2 analyses were used to assess percentage fibre type and MyHC distribution between the two groups. Two-way [group (CON and RFP) by fibre type (I, IIa and IIx)] mixed-model repeated measures ANOVAs were used to examine differences in CSA. Tukey HSD post hoc analysis was performed to indicate differences within analyses. 2 analysis was used to compare the percentage of atrophied types I, IIa and IIx fibres between the RFP and CON groups. Independent t-tests were used to examine differences in mean fibre CSA, CD/mm2, CC/F and C:F between the two groups. An independent Student’s t-test was performed to analyse the differences between malnourished and well-nourished patients as well as the clinical assessment data. An level of P  0.05 was selected to indicate statistical significance. percentage distribution of MyHC isoforms (P > 0.05) (Table 2). Non-locomotor muscle atrophy in renal failure 2077 Table 3. The percentage of atrophic fibres in the renal failure patient (RFP, n ¼ 22) and control (CON, n ¼ 20) groups Total Type I Type IIa Type IIx Type II RFP CON 27.0a 15.5 33.4a 46.8a 36.0a 10.0 6.6 12.3 10.8 12.0 The total percentage value represents the number of atrophic fibres in RFP and CON groups independent of fibre type. The different fibre type percentage values represent the number of atrophic fibres within each specific fibre type population. a Significant differences (P < 0.05). Fig. 2. Photomicrograph of renal failure patient muscle biopsy, showing muscle fibre atrophy and random distribution in fibre size and shape. muscle fibre CSA were observed between malnourished and well-nourished patients (P > 0.05). Atrophy Compared with the CON group, the muscle biopsies of the rectus abdominis muscle from the RFP group showed three times as many atrophied muscle fibres with random distribution in fibre size and shape within the muscle (Figure 2). In the CON group, only 10% of all fibres were classified as atrophic; in comparison, in the RFP group, 27% of all fibres have been calculated as being atrophic (x2 ¼ 42.55, P < 0.05) (Table 3). In the type II fibre populations of the RFP group, 36% of the fibres were atrophic. In the same fibre population of the CON group, the atrophy was nearer to 12% (x2 ¼ 41.33, P < 0.05). Atrophied fibres were found both in isolation and in small groups of 2–5 fibres together; these were surrounded by relatively normal-sized fibres and occasionally adjacent to unusually large hypertrophied fibres (Figure 3). Muscle fibre capillarization The CON group had 20% greater capillary density than the RFP group when expressed as capillaries per fibre (t41 ¼
5.7, P < 0.05) but not when expressed as CD/mm2 of muscle (P > 0.05); as described above, this difference is due to the higher number of fibres per square millimetre of muscle. The CON group also had a greater CC/F than the RFP group (t41 ¼
3.8, P < 0.05) (Table 4). Centronucleation The nuclei of muscle fibres are usually arranged around the periphery of the fibres; nucleii that are located within the fibre are usually indicative of a myopathy or dystrophic process. The proportion of fibres with central located nuclei was 5 ± 4% of the total fibre number in both groups. Downloaded from https://academic.oup.com/ndt/article/18/10/2074/1807592 by guest on 23 February 2023 Fig. 1. Fibre type cross-sectional area (CSA, mean ± SD) for renal failure patients (open bars; n ¼ 22) and controls (closed bars; n ¼ 20). Statistical significant differences in fibre size between groups were found for type IIa (**P < 0.01) and IIx fibres and mean CSA (*P < 0.05). 2078 G. K. Sakkas et al. Discussion Other histological abnormalities The biopsies of RFPs sometimes showed clumping of fibre types and a loss of normal mosaic pattern of fibre type distribution (Figure 4). Other abnormal features, such as large fibres with different shapes and clusters of fibres undergoing phagocytosis, were observed. Table 4. Capillarization of muscle and the number of capillary contacts per muscle fibre and capillaries per muscle fibre in the renal failure patient (RFP) and control (CON) groups Capillary factors RFP CON CD/mm2 CC/F C:F 320 ± 60 3.0 ± 0.4 1.12 ± 0.14 340 ± 90 4.0 ± 0.5a 1.35 ± 0.15a CD/mm2, capillary density per square millimetre of muscle tissue; CC/F, capillary contacts per muscle fibre; C:F, capillary-to-fibre ratio. a Significant differences (P < 0.01). Table 5. Muscle fibre morphometry of the five renal failure patients whose comorbidity was diabetes Type I (mm) Type IIa (mm) Type IIx (mm) Type I (%) Type IIa (%) Type IIx (%) CD/mm2 CC/F C:F 3000 1800 1900 1500 2500 3700 1300 1900 3200 3400 1500 1100 1500 2100 3600 66 35 52 48 59 32 59 46 47 40 2 6 2 5 1 280 400 341 451 330 3.4 2.7 2.9 3.3 3.0 1.10 1.14 0.97 1.32 1.10 CD/mm2, capillary density per square millimetre of muscle tissue; CC/F, capillary contacts per muscle fibre; C:F, capillary-to-fibre ratio. Downloaded from https://academic.oup.com/ndt/article/18/10/2074/1807592 by guest on 23 February 2023 Fig. 3. Photomicrograph of renal failure patient muscle biopsy, showing groups of atrophied fibres surrounded by relatively normalsized fibres. This is the first study of the morphology and morphometry of rectus abdominis muscle biopsies in pre-dialysis patients in end-stage renal failure. The only other study of rectus abdominis in RFPs, by Conjard et al. [17], did not include histological data. Muscle fibre type distribution and MyHC content in RFP biopsies were not different from control subjects and are similar to those previously reported by Haggmark and Thorstensson [18] for abdominal wall muscles of normal control subjects. Patient samples showed significant generalized muscle fibre atrophy, most marked in types IIa and IIx fibres (25–30% smaller than controls), such that the number of fibres per square millimetre of muscle was greater in RFPs. Muscle fibre capillarization, expressed as capillaries per muscle fibre, was significantly reduced compared to controls. Additional abnormal histological features of note were a wide range of fibre sizes, fibre type clumping and some necrotic fibres with associated inflammatory cell infiltrate. In the only published study of pre-dialysis patients, Clyne et al. [1] found, as in the present study, generalized muscle fibre atrophy with a greater degree of atrophy in type II than in type I fibres. In contrast, though, they observed a greater proportion of type I fibres in the patients’ muscle. Since they studied biopsies of vastus lateralis, this discrepancy in fibre type distribution may be due to the confounding affect associated with disuse atrophy, resulting from reduced locomotory activity in the patients they studied. Most other studies of skeletal muscle morphology in renal failure have exclusively examined the locomotor muscle group, the quadriceps, in haemodialysis patients [2,4– 10]. Common findings in these studies, as in the present study, were of fibre atrophy, mostly of type IIa fibres, variation in fibre size and fibre type grouping. Only two other studies have investigated a semi-non-locomotor muscle, the deltoid. Both were conducted in haemodialysis patients, one in adults [9], one in children [10]. The deltoid may be considered to be intermediate between rectus abdominis and quadriceps muscles in terms of involvement in locomotion and therefore subject to the effects of disuse atrophy. Nevertheless, the findings in these two studies were broadly similar to the present study, demonstrating type II fibre atrophy, fibre size variation and fibre type grouping. These Non-locomotor muscle atrophy in renal failure 2079 similarities are, perhaps, all the more remarkable, since the patients of Bautista et al. [9] were reported as showing ‘proximal paresis’. The likely systemic factors that may be associated with fibre atrophy in the RFPs in the present study include metabolic acidosis, peripheral neuropathy, malnutrition, anaemia or diabetic or cardiovascular comorbidity. Since the controls and RFPs were matched for age and gender, these potential influences on muscle fibre CSA cannot have accounted for the differences between the groups. Acidosis has been shown to be associated with increased muscle protein catabolism in patients with chronic renal failure [19]. There are no published muscle biopsy studies of the effects of acidosis correction in renal failure. Although this intervention has been shown to decrease aminoacid catabolism in chronic renal failure [20], neither of the two clinical studies of acidosis correction in dialysis patients was able to document an increase in muscle mass or nutrition [21,22]. Our patients were mildly acidotic, and therefore we cannot rule out a catabolic effect from having a mild metabolic acidosis. Peripheral neuropathy is associated with fibre atrophy and the changes in fibre type grouping [23] similar in pattern to that of other investigators and our observations in RFP muscle. None of our patients exhibited overt signs or symptoms of neuropathy. Our findings may indicate, however, that sub-clinical neuropathy is more widespread than is currently appreciated. Fibre atrophy following denervation is, like that associated with acidosis, mediated by the ubiquitin/proteasome system, but no experimental studies investigating the combined influence of a combination of renal failure and denervation have been published. Malnutrition and starvation have been shown to be associated with muscle fibre wasting in a number of clinical and experimental settings [7]. A decrease in protein/calorie intake is known to occur early in the progression of chronic renal failure, and malnutrition is common in the pre-dialysis population. Fahal et al. [7] found a significant difference in fibre CSA in quadriceps biopsy samples between adequately nourished and malnourished haemodialysis patients. In contrast, we did not find any differences in mean muscle CSA between subjects categorized as adequately nourished or malnourished, as defined by the seven-point SGA scale. They used a combination of plasma concentrations of visceral proteins and the three-point SGA grading as the means of measuring nutritional status. Thus, their categorization of malnutrition was not the same as ours. This, and any possible additional factors contributed by the process of haemodialysis, may explain the differences between the two sets of findings. Since skeletal muscle represents the major store of body protein, the degree of fibre wasting in the present study is potentially very important. It shows that up to about one-fifth loss of skeletal muscle contractile protein may exist in pre-dialysis patients. Thus, the sevenpoint SGA scale, a commonly used clinical nutrition measurement tool, may underestimate the degree of malnutrition in this patient population. It is possible that the lower serum albumin level in the RFPs compared with the controls was in part a reflection of malnutrition. It is possible that anaemia may have had a role in determining fibre atrophy. This proposal is based on the findings of Davenport et al. [8], who found that treatment of anaemia with erythropoeitin was associated with an increase in muscle type I fibre CSA in a group of haemodialysis patients. Since the main abnormality in the RFP biopsies related to type II fibres, it seems unlikely that anaemia was an influence. The mechanism of their observations is not clear, although increased oxygen delivery or increased physical activity may be involved. In line with these observations, it has been reported there is an increase in muscle strength following erythropoeitin treatment [24]. Downloaded from https://academic.oup.com/ndt/article/18/10/2074/1807592 by guest on 23 February 2023 Fig. 4. Photomicrographs of renal failure patient muscle biopsy, showing clusters of fibres undergoing phagocytosis. 2080 muscle bulk may need to be considered as a means of improving not only physical function, but also nutrition, in the renal failure population. Acknowledgements. We express our gratitude to the surgeons Mr Monro, Mr Morgan, Mrs Telfort, Mrs Walsh, Mr Hopkinson and Mr Forrest at Staffordshire Hospital for providing us the muscle biopsies. GKS was supported by the Greek State Scholarship Foundation (IKY) (No 11296). Conflict of interest statement. None declared. References 1. Clyne N, Esbjornsson M, Jansson E, Jogestrand T, Lins LE, Pehrsson SK. Effect of renal failure on skeletal muscle. Nephron 1993; 63: 395–399 2. Kouidi E, Albani M, Natsis K et al. The effects of exercise training on muscle atrophy in haemodialysis patients. Nephrol Dial Transplant 1998; 13: 685–699 3. Ayus J, Frommer P, Young JB. Cardiac and circulatory abnormalities in chronic renal failure. Semin Nephrol 1981; 1: 112–123 4. Bradley JR, Anderson JR, Evans DB, Cowley AJ. Impaired nutritive skeletal muscle blood flow in patients with chronic renal failure. Clin Sci 1990; 79: 239–245 5. Diesel W, Knight B, Noakes TD et al. Morphologic features of the myopathy associated with chronic renal failure. Am J Kidney Dis 1993; 22: 677–684 6. Moore GE, Parsons DB, Stray-Gundersen J, Painter PL, Brinker KR, Mitchell JH. Uremic myopathy limits aerobic capacity in hemodialysis patients. Am J Kidney Dis 1993; 22: 277–287 7. Fahal HI, Bell GM, Bone JM, Edwards RHT. Physiological abnormalities of skeletal muscle in dialysis patients. Nephrol Dial Transplant 1997; 12: 119–127 8. Davenport A, King RFGJ, Ironside JW, Will EJ, Davison AM. The effect of treatment with recombinant erythropoietin on the histological appearance and glycogen content of skeletal muscle in patients with chronic renal failure treated with regular hospital dialysis. Nephron 1993; 64: 89–94 9. Bautista J, Gil-Necija E, Castilla I, Chinchon I, Rafel E. Dialysis myopathy. Acta Neuropathol 1983; 61: 71–75 10. Fernandes Do Prado LB, Fernandes Do Prado G, Oliviera ASB, Schmidt B, De Abreu Carvalhaes JT. Histochemical study of the skeletal muscle in children with chronic renal failure in dialysis treatment. Arq Neuropsiquatr 1998; 56: 381–387 11. Detsky AS, McLaughlin JR, Baker JP et al. What is subjective global assessment of nutritional status? J Parenter Enteral Nutr 1987; 11: 8–13 12. Enia G, Sicuso C, Alati G, Zoccali C. Subjective global assessment of nutrition in dialysis patients. Nephrol Dial Transplant 1993; 8: 1094–1098 13. Gokal R, Nolph KD. Textbook of Peritoneal Dialysis. Kluwer Academic Publishers, Amsterdam, The Netherlands: 1994; 271–314 14. Brooke MH, Kaiser KK. Muscle fibre types: how many and what kind? Arch Neurol 1970; 23: 369–379 15. Andersen P. Capillary density in skeletal muscle of man. Acta Physiol Scand 1975; 95: 203–205 16. Talmadge RJ, Roy RR. Electrophoretic separation of rat skeletal muscle myosin heavy-chain isoforms. J Appl Physiol 1993; 75: 2337–2340 17. Conjard A, Ferrier B, Martin M, Cailette A, Carrier H, Baverel G. Effect of chronic renal failure on enzymes of energy metabolism in individual human muscles fibres. J Am Soc Nephrol 1995; 6: 68–74 Downloaded from https://academic.oup.com/ndt/article/18/10/2074/1807592 by guest on 23 February 2023 Bed-rest and limb immobilization are associated with marked muscle fibre wasting in periods as short as 4 weeks, with type II fibres being particularly affected, together with a decrease in oxidative enzyme content [25]. The appearances associated with bed-rest and those seen in locomotor muscle biopsies in RFPs are, therefore, quite similar. The fact that Kouidi et al. [2] showed that fibre atrophy was decreased following a period of exercise training supports a proposal that disuse contributes to this abnormality in locomotor muscles in RFPs. Rectus abdominis is not directly a locomotor muscle, although it has a role as a trunk stabilizer during movement. It is unlikely, therefore, that disuse atrophy is a significant determinant of the fibre atrophy observed in our patient samples. The finding that mean muscle fibre CSA was not different between those RFPs with and those without cardiovascular or other comorbidity, and that no differences in type II fibre size was found, indicate that these comorbidities are unlikely to have accounted for the fibre size reduction and atrophy in RFPs. Our observation of decreased C:F ratio in RFPs is novel. Disuse atrophy induced by bed-rest or limb immobilization is associated with a decrease in capillary density [26]. Moore et al. [6] observed a low C:F ratio in quadriceps biopsy samples of haemodialysis patients, which was not significantly improved by the exercise intervention they used. Bradley et al. [4] and Kouidi et al. [2] reported muscle capillary abnormalities, seen on electron microscopy, in haemodialysis patients. Kouidi et al. [2] also reported an improvement in capillary density following their exercise intervention trial, but they did not provide any numerical data of capillary density. It is likely, however, that physical activity status explains some of the changes seen in lower limb muscle biopsies in RFPs. As discussed above, this influence is likely to be much less relevant in the abdominal muscle investigated in the present study. It may be of significance that a similar defect in fibre capillarization has been observed in myocardial tissue of dialysis patients [27]. It is possible that these findings and those of Bradley et al. [4], Moore et al. [6] and Kouidi et al. [2], taken with ours, indicate that a systemic defect of angiogenesis may exist in renal failure. The results of studies by previous workers suggest that it is likely that disuse atrophy contributes to the findings of degenerative changes in locomotor muscles in RFPs. There are, however, many features common to those we report in rectus abdominis, a muscle little involved in locomotion. In addition, these features are present in dialysis and in pre-dialysis patients. We conclude that the state of renal failure itself (‘uraemia’) plays an important role in the reduction of the capillary density as well as the muscle size in predialysis patients. These abnormalities in skeletal muscle of patients with end-stage renal disease have at least two important implications: (i) an even greater degree of malnutrition may exist in pre-dialysis patients than is defined by a commonly used clinical measure; and (ii) exercise training interventions aimed at increasing G. K. Sakkas et al. Non-locomotor muscle atrophy in renal failure 18. Haggmark T, Thorstensson A. Fibres types in human abdominal muscles. Acta Physiol Scand 1979; 107: 319–325 19. Mitch WE, Goldberg AL. Mechanisms of muscle wasting. The role of the ubiquitin–proteasome pathway. N Engl J Med 1996; 335: 1897–1905 20. Reaich D, Channon SM, Scrimgeour CM, Daley SE, Wilkinson R, Goodship TH. Correction of acidosis in humans with CRF decreases protein degradation and amino acid oxidation. Am J Physiol 1993; 265: E230–E235 21. Stein A, Moorhouse J, Iles-Smith H et al. Role of improvement in acid-base status and nutrition in CAPD patients. Kidney Int 1997; 52: 1089–1095 22. Brady JP, Hasbargen JA. Correction of metabolic acidosis and its effects on albumin in chronic hemodialysis patients. Am J Kidney Dis 1998; 31: 35–40 2081 23. Brooke MH, Engel WK. The histologic diagnosis of neuromuscular diseases: a review of 79 biopsies. Arch Phys Med Rehab 1966; 47: 99–121 24. Guthrie M, Cardenas D, Eschbach JW, Haley NR, Robertson HT, Evans RW. Effect of erythropoietin on strength and functional status of patients on hemodialysis. Clin Nephrol 1993; 39: 97–102 25. Sargeant AJ, Davies CT, Edwards RH, Maunder C, Young A. Functional and structural changes after disuse of human muscle. Clin Sci Mol Med 1977; 52: 337–342 26. Bloomfield SA. Changes in musculoskeletal structure and function with prolonged bed rest. Med Sci Sports Exerc 1997; 29: 197–206 27. Amann K, Breitbach M, Ritz E, Mall G. Myocyte/capillary mismatch in the heart of uremic patients. Am J Soc Nephrol 1998; 9: 1018–1022 Downloaded from https://academic.oup.com/ndt/article/18/10/2074/1807592 by guest on 23 February 2023 Received for publication: 18.4.02 Accepted in revised form: 2.5.03