The research question was supported by inclusion criteria for study design, outcomes and intervention but not for participants, which may have led to subjective decisions when selecting studies. Only one database was searched and only published English-language studies were included, which may have increased the possibility of publication and language biases. Data were extracted by two reviewers, reducing the possibility of reviewer bias and error. It was not reported whether similar steps were taken for study selection. The authors stated that the primary studies were too heterogeneous to pool in a meta-analysis, but pooled effect sizes were reported, which may not have been appropriate. The pooled effect sizes did not include an assessment of statistical heterogeneity. As the validity of these studies was not assessed, it was unknown whether their results and, therefore, the synthesis, was reliable. The number of participants was also unclear as the total numbers in the text did not match those in the table.
Although the authors' cautious conclusions adequately reflected the limitations of the evidence, a cautious interpretation is advised given concerns about the review methods and analyses.