Dear Editor,

We read with great interest the article by Hein et al. [1], which described the meta-analysis study on the impact of disease-modifying anti-rheumatic drugs (DMARDs) therapy on skeletal muscle mass in rheumatoid arthritis (RA) patients. While the data presented are impressive, we would like to add some remarks about methodological issues that should be considered.

First, this meta-analysis does not include several necessary studies that have provided data on the relationship between anti-tumor necrosis factor (anti-TNF) therapy and body composition. For instance, Serelis et al. showed the information on body composition before and one year following anti-TNF therapy [2]. Additionally, Eric Toussirot et al. also presented data on body composition before, 1 year after, and 2 years after anti-TNF medication [3]. These studies appear to meet the criteria proposed by the authors; thus, it could be necessary to include them to make the meta-analysis more comprehensive.

Second, this study did not employ a representative measure of skeletal muscle mass that was adjusted for body size, such as skeletal muscle mass index (SMI). It is well recognized that skeletal muscle mass varies with body size, particularly height and body mass index (BMI). This raises the possibility that even when the same proportion of skeletal muscle mass changes, the absolute degree of that change may vary between populations with larger and smaller body sizes. For this reason, several recommendations, including those for sarcopenia, recommend using skeletal muscle mass that has been adjusted for height or BMI when determining and contrasting the quantity of skeletal muscle mass [4, 5]. Given the heterogeneity background of the studies included in this meta-analysis, it may be worthwhile to conduct an additional analysis regarding the associations between DMARDs and the adjusted indicator of skeletal muscle mass such as SMI, which indicators are included in several studies in this meta-analysis.

Third, when determining body composition, there are variances between bioelectrical impedance analysis (BIA) and dual-energy X-ray absorptiometry (DEXA). In comparison to DEXA, BIA is known to overestimate skeletal muscle mass in the general population [6]. Similar to this, BIA has overestimated the amount of fat-free mass in RA patients [7]. In this regard, it may not be appropriate to simultaneously perform a meta-analysis of skeletal muscle mass determined by DEXA and BIA.

We conclude by recommending more investigations to strengthen the arguments presented by this informative meta-analysis due to the significant methodological issues we raised.