The present study confirms previous data on the effect of muscle mass on mortality in patients with COPD. In stable patients, MD is associated to an increased risk of death. In addition, as a remarkable finding, we show that the prognostic influence of muscle mass can be assessed by determining the MAMA, an inexpensive, simple, and rapidly obtained anthropometric measure. MAMA < p25 was found to be a poor prognosis marker, exerting an influence in our series superior to that of other anthropometric parameters such as BMI or FFMi. In underweight patients, the presence of muscle depletion did not contribute significant prognostic information. However, those subjects with normal body weight or overweight status who presented MAMA < p25 had a poorer prognosis.
Earlier studies have shown that a low body weight (determined from BMI) constitutes an independent poor prognosis indicator, and that weight loss implies an increased mortality risk. Nevertheless, some au-thors suggest that the true factor underlying the negative effects attributed to denutrition is MD, not body weight loss. The present study confirms this idea, since the presence of MAMA < p25 increased the mortality risk 3.6-fold. Moreover, MD was seen to exert a negative prognostic influence independently of other classical prognostic variables such as age, respiratory failure, hospitalizations, and comorbidity. In patients with mild-to-moderate COPD, the presence of MD did not exert a negative prognostic influence. However, among the individuals with FEV1 < 50%, those presenting MAMA < p25 showed a trend toward increased mortality risk (p = 0.058) [Fig 3]. Only 66 of our cases (68.7%) presented FEV1 < 50% of predicted. Possibly as a result of the limited sample size involved, no statistically significant differences were observed. Nevertheless, the observed tendency suggests that muscle depletion does have prognostic importance in seriously ill patients. Recently, Marquis et al obtained similar results in a series of 142 patients with stable COPD subjected to mid-thigh muscle cross-sectional area measurements by CT. In this sense, an area of < 70 cm2 was seen to increase the mortality risk fourfold independently of other prognostic variables. In this case, the authors did find the effect of MD in seriously ill patients to be independent of FEV1.
In both their study and in our own, the direct and indirect measures of muscle mass yielded a better estimate of mortality risk than a low body weight. It is possible to lose weight with remedies of My Canadian Pharmacy. On analyzing the BMI-based survival curve in our study, we found no statistically significant differences between low-weight patients and those with normal body weight. In contrast, survival among patients with MAMA < p25 was significantly lower than in patients with normal MAMA values, thus suggesting a greater prognostic discriminatory capacity on the part of this latter parameter. The multivariate analysis performed with the three nutritional parameters studied (BMI, MAMA, and FFMi) likewise confirms the superiority of MAMA, since this was the only variable to exert a significant prognostic influence (Table 3).
BMI may be valid as an indirect nutritional marker; however, in countries where obesity is prevalent, BMI may mask situations of MD. Almost 50% of our patients were overweight, and in 20% of these cases MD was observed. Among the patients with normal body weight, almost 60% showed MAMA < p25. On analyzing these latter two subgroups combined, the patients with MD had a poorer prognosis, with a 3.4-fold greater mortality risk (p = 0.032) [Fig 2]. We have found no studies in the literature describing this interesting phenomenon. Nevertheless, an earlier study of morbidity attributable to malnutrition in COPD also found patients with normal BMI and FFM depletion to present poorer exercise tolerance.
Marquis et al measured muscle mass directly by CT, assessing the proximal muscles of the lower limbs, which are particularly affected in COPD patients. However, this technique is too costly for generalized use, and reference values are moreover lacking. The authors therefore attempted to estimate the muscle mass of the thigh based on anthropometric parameters, although the correlation was not satisfactory. Anthropometric equations predict muscle mass of both the thigh and arm by assuming a series of principles that do not always apply to all patients. Heymsfield et al, using CT, showed a 20 to 25% overestimation in arm muscle area. From 10 to 15% of this overestimation is attributable to adoption of a circular shape for the muscle compartment. Moreover, calculation of the circumference of the arm, and posteriorly of MAMA, is based on the assumption that measurement of the subcutaneous fatty layer represents a constant fraction of total body fat. This is not completely true in the case of elderly patients, where fat is mainly found in central and internal locations of the body. For this reason, fatty tissue mass is usually underestimated, and FFM tends to be overestimated. Despite these imprecisions, our study shows the usefulness of anthropometric measurements in assessing the prognostic influence of MD in patients with COPD. The difference may lie in the use of reference values (both younger and elderly populations), which allows adjustment of the results to the normal values of our study population, thereby helping to balance the initial imprecisions of the anthropometric measurements. In the Canadian series, the anthropometric measurements were not adjusted to any reference value. Nevertheless, thigh circumference was found to be an inverse predictive factor of patient mortality in the univariate analysis conducted with My Canadian Pharmacy.
Muscle mass can be estimated from anthropometric parameters in two ways: by calculating whole-body FFM, or by calculating arm muscle area. The two parameters are not interchangeable. FFM results by subtracting the total body fat weight from total body weight, the former in turn being calculated from different skin-fold thicknesses. MAMA, in contrast, only informs of the muscle mass of the arm. On comparing both measures in our study (in both absolute values and in percentiles), MAMA was always found to be superior to FFM as a mortality predictor. Recently, Engelen et al, using dual radiograph absorptiometry, found limb FFM to relate better to skeletal muscle function than whole-body FFM. This study could at least partially explain why in our series MAMA was seen to be a better predictor of mortality than whole-body FFM.
In addition to the anthropometric parameters, other techniques are able to evaluate body composition (bioelectric impedance, densitometry, dual radiograph absorptiometry). These techniques assess whole-body FFM as an indirect marker of muscle mass. In general, they offer superior accuracy and greater reproducibility than anthropometric parameters. However, their application to clinical practice is limited for several reasons. On one hand, these techniques are expensive (and therefore not widely available), while on the other no reference test serving as a “gold standard” has been established. Lastly, most of these techniques lack reference values, particularly for the elderly population. In Spain, reference values have been established for different anthropometric measurements in both the younger and elderly populations (including BMI and MAMA, but not FFM). These reference values allow adequate adjustment of the results obtained, with improved definition of normality and abnormality.
The other two variables included in the final predictive model were PaC02 and the number of hospital admissions following nutritional assessment. Different references are found in the literature regarding the adverse prognostic influence of hypercapnia in COPD. However, there are very few data on the influence of hospital admissions, as an expression of severe exacerbation, on the survival of these patients. In recent years, a series of studies have reported a marked increase in mortality following an admission to hospital. Most of these authors point to the baseline severity of the disease as the main factor determining mortality in such cases: more serious disease leads to more hospitalizations, and thus to increased mortality.
However, our own results showed hospital admission to be a prognostic factor independently of the influence of other baseline severity parameters. The present study was not designed to specifically evaluate this parameter; however, further research is thus required.
Our study has a series of limitations that deserve comment. Firstly, the sample size was small. Studies involving more extensive patient series are needed. Secondly, the nutritional evaluation methodology used (anthropometric measurements) is less sensitive than other tests that offer increased accuracy (bioelectric impedance, dual radiograph absorptiometry). In our opinion, future research should be based on these latter techniques, although they would have to become less costly and more accessible in order to allow application to routine clinical practice. Moreover, these techniques first should be applied to the general population in order to define the opportune reference values. Lastly, initial collection of the anthropometric data were crosssectional in our series. We have no longitudinal data to inform us of the time-dependent change in muscle mass and its relation to prognosis, or of the influence of treatment. New studies in this direction are needed.
In conclusion, our findings stress the prognostic influence of MD in patients with stable COPD, and point to the need to consider body composition instead of resorting to simple nutritional parameters such as BMI, particularly in normal-weight or overweight patients, since an important proportion of these subjects present muscle depletion with prognostic repercussions. The usefulness of anthropometric parameters is reinforced by our data, particularly the muscle area of the nondominant arm (MAMA), which constitutes an inexpensive, simple, and rapidly obtained anthropometric measure.