IJCS | Volume 32, Nº2, March/April 2019

149 Rocha et al. Nutritional assessment in heart failure Int J Cardiovasc Sci. 2019;32(2)143-151 Original Article cachexia and sarcopenia are associated with worse prognosis and higher mortality, 34 it is preferable to use the AMC values even in a potential overestimation of malnutrition due to the impossibility of acquiring and maintaining more sensitive equipment such as EBI. For the APMT, a lower frequency of muscle depletion was found, compared to FFMI, suggesting that this measure has a reduced diagnostic capacity to evaluate compensated individuals followed up in an outpatient setting as in this study. However, its use in an intensive care unit has proven to be a good method of assessing nutritional risk. 35 One of the most striking aspects of our study was the high percentage of individuals who had excessive adiposity/fat. This fact may have a positive impact on the course of the disease, since there is an inverse association between overweight or obesity andmortality in individuals with HFrEF, as found in our population. 36 In our study, there was a uniformity of diagnoses considering the indicators used for this purpose. However, both BMI and TST showed low sensitivity, accuracy and concordance with the percentage of fat evaluated by EBI. Caution is needed in this interpretation, especially when BMI is used for assessing body fat. A prospective study with patients with HF found results similar to ours for the diagnosis of BMI. The authors raise questions that individuals with protein malnutrition may possibly be classified as overweight by BMI, underestimating cases of malnutrition and increasing the risk of mortality. 37 According to a study that investigated the association between body composition assessed by dual energy x-ray absorptiometry (DEXA), an imaging test that promotes the identification of body extracts (gold standard) and mortality of patients with HF, it was observed that the BMI misclassified 41% of patients for body fat, 5 similarly to the one found in our study, where we identified 50% of negative predictive value for BMI. Although our results did not confirm good indicators of accuracy and concordance of TST in the evaluation of excess adiposity, other authors point out the use of TST as a predictor of mortality, further reinforcing the obesity paradox, since excess TST presented an inverse relation with mortality. 7 It is suggested to evaluate subcutaneous fat from the association between multiple skinfolds, used in formulas to evaluate adiposity, as well as was evaluated by Gastelurrutia et al., 37 associated with other methods to identify the nutritional status of individuals with HF, aiming to identify adipose reserve more accurately when there is no evaluation of fat percentage by EBI. Our study is not free of limitations: (1) the low frequency of anthropometric disorders may be biased in the statistical analysis; (2) a small number of participants may have caused lower values ​of sensitivity and concordance, but the results foundwere relevant to signal the need to consider several methods of assessing body composition in this population. The results reveal important information about the use of anthropometric indicators in the evaluation of individuals withHF, reinforcing that, for a more accurate evaluation, EBI assessment may be better indicated. EBI measurements should be standardized to obtain reproducible results and the use of body composition prediction equations must meet the specifications of each population studied. It should be noted that when it is not feasible to use equations already defined in specific studies, it is possible to use data from the EBI provided by the equipment at the time of measurement, such as resistance (R) and reactance (Xc) data, resulting in the formation of data such as phase angle, which is used in the diagnosis of malnutrition and clinical prognosis. Additionally, another potential data for the diagnosis and prognosis of body mass reserves is the analysis of bioelectrical impedance vectors. 38,39 Conclusion In conclusion, it is suggested that although AMC overestimates muscle depletion, it can be used as an indicator of protein malnutrition. In addition, BMI was not sensitive in the evaluation of muscle and body fat components, therefore caution is recommended in the use of this indicator in patients with compensated HF followed up in an outpatient setting. TST did not present good sensitivity for the evaluation of adiposity and body fat, reinforcing the importance of the evaluation considering multiple folds. The anthropometric parameters should be incorporated into the clinical practice for their low cost and practicality, but for greater accuracy in the evaluation of body composition, more accurate methods, such as EBI, should be considered. Author contributions Conception and design of the research: Rocha DO, Diniz RVZ, Lira NRD, Sena-Evangelista KCM.