Inappropriate Methods and Flawed Conclusion in: Can Resistance Training Enhance the Rapid Force Development in Unloaded Dynamic Isoinertial Multijoint Movements? A Systematic Review

Request for Clarification:

We have with great interest read the recent article “Can resistance training enhance the rapid force development in unloaded dynamic isoinertial multijoint movements? A systematic review” by van Hooren et al. (¹⁰).

Although the topic is highly relevant to athletes and staffs involved in sports, we have major concerns regarding the methods used to synthesize the available evidence. As the methodological flaws outlined below seem to have led the authors to an erroneous conclusion, we believe that this need to be addressed to avoid misguidance of future clinical practice.

First, and most importantly, the overall conclusion that “Resistance training has a limited transfer to rapid force development in unloaded dynamic isoinertial multijoint movements” (¹⁰) is guided by an inappropriate data synthesis based on the counting of significant and nonsignificant p values, which likely has flawed the conclusion. Importantly, the lack of statistical significance of individual studies neither implies that the totality of evidence supports no or limited training effect nor the opposite (⁴). Furthermore, this way of synthesizing data is not recommended by the PRISMA guidelines (⁷), which the authors state to follow.

Another important methodological consideration is the inclusion of various rates of force development (RFD) metrics in a single analysis. This might be problematic because different RFD metrics describe different physiological mechanisms (⁸). Hence, different metrics may yield different results. In the studies by Kraemer et al. (⁶) and Newton et al. (⁹), trivial-to-small changes were observed in time to peak force; however, small-to-medium improvements were observed in peak force. This would suggest an increased force production during a similar time frame; hence, an increased RFD if calculated as Newton per second using fixed time points.

Finally, we acknowledge that a meta-analysis could not be performed but do not believe that this justifies the use of inappropriate methods as highlighted above. The importance of using appropriate methods is highlighted by the fact that medium-to-large effect sizes are observed for RFD when calculated for fixed time periods and eccentric RFD, whereas trivial-to-small effect sizes are observed for time to peak force and peak RFD, when these are calculated from the available data, using the method described by van Rhee et al. (¹¹). Please see Table 1 for results that support this statement. Therefore, we suggest that the conclusion is revised to underpin that resistance training may in fact improve RFD during unloaded dynamic movements, at least for some RFD metrics.

Table 1.: Combined effect sizes (Cohen's d) based on RFD metric and testing procedure after strength or power training.*†