Thank you for the opportunity to review this manuscript titled "Efficacy of the Diabeloop Artificial Pancreas for Blood Glucose Control in Patients with Type 1 Diabetes Mellitus Exposed to Challenging Situations in Real Life: A Randomized Controlled Trial".

The authors are investigating the overnight control of the Diabeloop closed-loop system compared to open loop control in the setting of "real life" glycemic challenges that include heavy meals and sustained physical exercise.

Overall, this is a very well written paper addressing the use of a previously validated algorithm in novel scenarios that current artificial pancreas systems will be exposed to during out-patient use.

The main set back to the study's strength is that these scenarios appear to be implemented in a controlled hospital setting rather than in a free-living, home environment that would more fully test the algorithm's robustness.

Despite this, the data gained continues to expand the knowledge of algorithm performance under more extreme conditions, a necessity for ensuring robust glucose control in everyday situations.

Author affiliations and corresponding author information are provided.

The abstract is well written and appropriately concise, but a few points:

• the data for the average glucose and the time-in-range (70-180 mg/dl) over the 3-day period is different from that in the final tables, these should be updated to reflect the data tables, or clarified if this represents different information from the tables

• there is a typo in the last line of the Research Design and Methods section: "... spent in the tight the glucose range..." The second "the" can be removed

The introduction covers the present base of knowledge on blood glucose control in type 1 diabetes and has appropriate citing of articles.

The methods section describes the objective, ethical conduct, patient groups, study materials and procedures, outcome measures, and the statistical plan in appropriate detail. Just a few points here about the methodology:

• Please include the time cutoffs utilized for day and night analysis. I was unable to locate this information throughout the manuscript.

• It is unclear if the extreme conditions are being implemented in the hospital or at home, though this is implied several times. I would ask that the authors clarify this point in the methods to make it clear to the reader that subjects were in hospital during those portions.

• The authors chose overnight glucose control as their primary outcome, and this is an acceptable primary outcome. It might benefit the reader to know why overnight control is the preference for evaluating closed loop control (potentially in the introduction), rather than total average glucose or time in range throughout the entire study.

• For the statistical analysis, the non-inferiority threshold of >0.8 for the 95% lower confidence limit is not a familiar criteria. Turki et al (Br J Clin Pharmacol. 2017 Aug; 83(8): 1636-1642.) provides a good review of assigning the non-inferiority margin for studies, and this should be based on information from prior studies or additional statistical analysis, which have not been discussed in this paper.

Certainly this would be more rigorous for RCTs, but describing the assignment of the method of determining the non-inferiority margin would help improve the overall impact of the manuscript.

The results section is concise and well written as well, and there are only a few points I would recommend to the authors here:

• A recruitment flow chart would be helpful to visualize the final patient sample sizes and help quickly clarify the difference between the n values for the baseline data and the analysis tables for readers.

• Please include results on hardware and software functioning, including active sensor wear time, number of interventions by the study team, communication interruptions, and software failures.

The conclusion is also well written and appropriately references the results. The limitations section is inclusive of the important drawbacks to the study. I would ask the authors to reread this section as there are a few grammatical and syntax errors noted. No other changes are recommended concerning the conclusion.

Funding, author conflicts and contributions, and supplemental material are reviewed and are appropriate for the content.

Aside from the few minor points mentioned, this manuscript was well put together, and it was a pleasure to read and review.

Thank you for this opportunity, and I wish the authors all the best in their future endeavours.

Thank you for your response, all my review points were adequately addressed. I have no further concerns. All the best.