The paper authored by Dankel et al., published in Archives of Cardiovascular Disease , has been read meticulously. The authors purposed to examine the dose-dependent association of muscle-strengthening activities (MSA) with all-cause mortality. It was concluded that the risk of premature all-cause mortality is decreased remarkably among people with 2−3 MSA sessions per week. Although the findings are useful, some methodological issues must be considered.

First, combined outcomes, including all-cause mortality, are commonly used in studies to improve the sample size and power of statistical tests, but the main assumptions such as homogeneity of strength of associations seems to be neglected in some investigations. In other words, the independent variables such as MSA should be homogenously associated with multiple specific types of mortality, which is difficult to establish. To avoid this problem, some statistical methods, including hierarchical regression modelling, have been newly introduced to examine more efficiently the associations between specific exposures with multiple specific outcomes . Hence, we suggest that the aforementioned regression model is used by the authors in their present and future works.

Second, we wonder why the authors did not examine the adjusted associations between MSA and the combined outcome. Consequently, it was not clarified whether MSA was independently associated with all-cause mortality.

Finally, Dankel et al. used a Cox proportional hazard model to examine the associations between the MSA and all-cause mortality, but the proportional hazard assumption was not evaluated. This important assumption can be checked using different methods and variants of the Cox models, including the Extended Cox regression model, which can be used when this assumption is violated. Otherwise, the findings may be incorrect and misleading .