Malnutrition and pectoralis muscle index in medical intensive care unit patients: A matched cohort study

Authors:
Publisher website
Google Scholar
Abstract
Background: Muscle assessment is an important component of nutrition assessment. The Global Leadership Initiative on Malnutrition (GLIM) consortium recently underscored the need for more objective muscle assessment methods in clinical settings. Various assessment techniques are available; however, many have limitations in clinical populations. Computed tomography (CT) scans, obtained for diagnostic reasons, could serve multiple purposes, including muscle measurement for nutrition assessment. Although CT scans of the chest are commonly performed clinically, there is little research surrounding the utility of pectoralis muscle measurements in nutrition assessment. The primary aim was to determine whether CT‐derived measures of pectoralis major cross‐sectional area (PMA) and quality (defined as mean pectoralis major Hounsfield units [PMHU]) could be used to identify malnutrition in patients who are mechanically ventilated in an intensive care unit (ICU). A secondary aim was to evaluate the relationship between these measures and clinical outcomes in this population.Methods: A retrospective analysis was conducted on 33 pairs of age‐ and sex‐matched adult patients who are being mechanically ventilated in the ICU. Patients were grouped by nutrition status. Analyses were performed to determine differences in PMA and mean PMHU between groups. Associations between muscle and clinical outcomes were also investigated.Results: Compared with nonmalnourished controls, malnourished patients had a significantly lower PMA (P = 0.001) and pectoralis major (PM) index (PMA/height in m2; P = 0.001). No associations were drawn between PM measures and clinical outcomes.Conclusion: These findings regarding CT PM measures lay the groundwork for actualizing the GLIM call to action to validate quantitative, objective muscle assessment methods in clinical settings.

This publication has 37 references indexed in Scilit: