It is time to close the book on device-guided slow breathing

Dear editor,

Barros et al. reported the long-term effects of device-guided slow breathing (DGB) in hypertensive patients comparing DGB using a device called the REPeRATE^® (InterCure Ltd, Tel-Aviv, Israel) to listening to music [1]. The authors observed no significant effects whatsoever on blood pressure, concentrations of catecholamines or muscle sympathetic nerve activity (SNA). We have several remarks related to DBG and this interesting article.

Firstly, we have serious concerns with the interpretation of previous literature. The last paragraph of the discussion section is far too positive concerning the effectiveness of DBG and is not based on an accurate interpretation of the evidence concerning this device [2]. The authors mention that most studies concerning DGB published favorable effect in relation to office blood pressure in patients with hypertension. Of course this is no argument that would favor DGB as such. When considering the methodological quality, the highest quality trials do not show any effect of DGB on blood pressure as was found in the only double-blind (and sham-controlled) trial [3]. Furthermore, the absence of any possible relevant effect has been confirmed in meta-analyses of trials using individual patient data of different trials and aggregated data of more trials without using individual patient data [2,4].

Secondly, we have some questions concerning the methodology. The primary endpoint of this study is the change in SNA and plasma catecholamines. We would like to ask the authors to explain why they have chosen for this endpoint instead of blood pressure, which is a clinically relevant endpoint and has also been used in previous studies? Furthermore, randomization clearly failed, as there is a relevant difference in using antihypertensive drugs between the two groups. It remains unclear whether the authors adjusted for this difference in their analyses. Also, the authors should have performed intention-to-treat analyses, and we would strongly like to ask the authors to present the differences between groups with confidence intervals: this enables the reader to interpret possible relevant or irrelevant differences and is also important for aggregation of data in future meta-analyses.

Thirdly, it should be taken in mind that adverse effects have been reported using DGB. Previously reported side effects included shortness of breath, atypical chest pain with shortness of breath that resulted in an emergency department visit and one patient died of respiratory failure due to underlying heart failure [3]. Therefore, we would like to inquire whether adverse events have taken place during the trial.

In conclusion, we disagree with the authors’ conclusion that ‘DGB still needs more evidence that demonstrates its real effectiveness. At the risk of stating the obvious: the aim of such a study would be to investigate if there is any effect of DBG on blood pressure, not to presume there is an effect. In addition, evidence from meta-analyses of randomized trials and systematic reviews have clearly demonstrated, as this study does (!), that DBG has no beneficial effects on reducing blood pressure. When performing another (identical) study there is risk of a type 1 error, risk for patients and there is no clear scientific need.

Disclosure statement

The authors have no conflicts of interest.