Wasting, but not malnutrition, predicts cardiovascular mortality in end-stage renal disease.
P. Stenvinkel,O. Heimbürger,B. Lindholm
DOI: https://doi.org/10.1093/NDT/GFH296
2004-09-01
Abstract:Beddhu et al. [1] recently have reported that there were no positive associations of malnutrition with documented acute coronary syndromes requiring hospitalization in a large incident Medicare dialysis population. The authors state that their findings do not support our suggestion [2] of an association between malnutrition, inflammation and atherosclerosis (MIA hypothesis) in patients with end-stage renal disease (ESRD). No doubt the study by Beddhu et al. [1] is a well performed study conducted in a large group of dialysis patients. However, we feel that some fundamental issues regarding the nomenclature and definition of ‘malnutrition’ need to be discussed. First, as discussed by Mitch [3], the use of the word ‘malnutrition’ has often been used incorrectly in the renal literature, and, according toMitch, ‘malnutrition’ should be used to describe a deficient nutritional status caused by insufficient nutritient intake. However, literally, the word ‘malnutrition’ (derived from the latin word ‘malus’) means, ‘not correctly nourished’. Thus, both underand over-nourished (obese) patients could be considered to be ‘malnourished’. Secondly, whereas malnutrition is usually defined as a consequence of insufficient food intake and low serum protein levels, the loss of muscle mass (i.e. cachexia or wasting) in the ESRD patients is usually the consequence of a number of catabolic mechanisms stimulated by renal insufficiency. Indeed, the aetiology of loss of lean body mass in ESRD is very complex and may include numerous factors apart from poor food intake (i.e. true malnutrition), such as delayed gastric emptying, hormonal derangements, inadequate control of acidosis, co-morbidity and inflammation [3]. Whereas Beddhu et al. [1] used a low body mass index (BMI) and low creatinine excretion in urine as a marker of malnutrition, we used subjective global assessment (SGA). It is obvious that SGA and BMI are crude surrogate markers of nutritional status that cannot be readily compared. In our opinion, BMI is not a very precise parameter of nutritional status, especially in patients in whom gross imbalances in fluid homeostasis are commonly observed, such as in patients with ESRD, congestive heart failure and liver disease. Also, protein malnutrition with loss of muscle mass (sarcopoenia) is often associated with a relatively well-preserved fat mass in dialysis patients, resulting in small changes in BMI that may be obscured further by imbalances in fluid homeostasis. In fact, as BMI alone does not always capture adequately the joint relationship of body composition and body size to outcome [4], the waist circumference may have more potential in detecting overweight than BMI [5]. Indeed, although there may be a marked statistical difference in BMI levels comparing ESRD patients classified as well nourished (SGA1⁄4 1) or malnourished (SGA 2), there is a considerable overlap between the two groups (Figure 1). Thus, patients considered to be malnourished according to SGA may have a low as well as a high BMI (Table 1). Clearly, high-BMI patients could also have an ongoing catabolic wasting process, and low-BMI patients may show no signs of catabolism. To be able to compare the results presented by Beddhu et al. [1], we have analysed the independent predictive power of both BMI and SGA in 271 incident ESRD patients (173 males; mean 53±1 years) starting renal replacement therapy (Table 1). During the observation period (3.7±0.2 years; range 0.1–9.2 years), 94 patients died (66 of cardiovascular disease; CVD). Like Beddhu et al. [1], we divided the patients into three groups according to their BMI. Whereas no differences in the prevalence of diabetes mellitus, clinical overt CVD or inflammation [C-reactive protein (CRP) 10mg/l] were noted between the different BMI groups, the prevalence of malnutrition (according to SGA) differed significantly (Table 1). However, it should be noted that 38% of all patients in the low-BMI group were classified as having a normal nutritional status according to SGA, whereas 45% of the patients in the normal-BMI group and 17% in the high-BMI Correspondence and offprint requests to: Peter Stenvinkel, MD, PhD, Division of Renal Medicine K56 St, Karolinska University Hospital, Huddinge, S-141 86 Stockholm, Sweden. Email: peter.stenvinkel@klinyet.ki.se