Variation in Resting Metabolic Rate Affects Identification of Metabolic Change in Geographically Distinct Cohorts of Patients With ALS.

Background And Objectives: Altered metabolism is observed in amyotrophic lateral sclerosis (ALS). However, without a standardized methodology to define metabolic changes, our understanding of factors contributing to and the clinical significance of altered metabolism in ALS is limited.

Methods: We aimed to determine how geographic variation in metabolic rates influences estimates and accuracy of predicted resting energy expenditure (REE) in patients with ALS and controls, while validating the effectiveness of cohort-specific approaches in predicting altered metabolic rate in ALS. Participants from 3 geographically distinct sites across Australia, China, and the Netherlands underwent REE assessments, and we considered 22 unique equations for estimating REE. Analyses evaluated equation performance and the influence of demographics on metabolic status. Comparisons were made using standardized and local reference values to identify metabolic alterations.

Results: 606 participants were included from Australia (patients with ALS: 140, controls: 154), the Netherlands (patients with ALS: 79, controls: 37) and China (patients with ALS: 67, controls: 129). Measured REE was variable across geographic cohorts, with fat-free mass contributing to this variation across all patients ( = 0.002 to < 0.001). Of the 22 predication equations assessed, the Sabounchi Structure 4 (S4) equation performed relatively well across all control cohorts. Use of prediction thresholds generated using data from Australian controls generally increased the prevalence of hypermetabolism in Chinese (55%, [43%-67%]) and Dutch (44%, [33%-55%]) cases when compared with Australian cases (30%, [22%-38%]). Adjustment of prediction thresholds to consider geographically distinct characteristics from matched control cohorts resulted in a decrease in the proportion of hypermetabolic cases in Chinese and Dutch cohorts (25%-31% vs 55% and 20%-34% vs 43%-44%, respectively), and increased prevalence of hypometabolism in Dutch cases with ALS (1% to 8%-10%).

Discussion: The identification of hypermetabolism in ALS is influenced by the formulae and demographic-specific prediction thresholds used for defining alterations in metabolic rate. A consensus approach is needed for identification of metabolic changes in ALS and will facilitate improved understanding of the cause and clinical significance of this in ALS.