Response to the Letter by Zhang “The hidden variable in a ‘neutral’ upgrade: why parameter re-tuning and perceived reliability matter after G6 → G7 in Control-IQ”

Dear Editor,

We thank Dr. Zhang ¹ for the thoughtful and constructive comments regarding our study evaluating the transition from Dexcom G6 to Dexcom G7 in adults with type 1 diabetes using the Tandem t:slim X2 Control-IQ system. ² We appreciate the opportunity to further discuss several methodological aspects of our findings and agree that therapy optimization represents an important component of real-world automated insulin delivery care.

As correctly highlighted, 27 participants (29%) reported modifications in Control-IQ settings during follow-up. ¹ We agree that part of the glycemic improvement observed after transition to Dexcom G7—particularly among participants with baseline TIR_70–180 <70%—may plausibly reflect not only the sensor transition itself, but also adaptive therapy optimization, increased user engagement, and progressive familiarization with the system. ³ In routine clinical practice, these factors are often intrinsically interconnected and difficult to fully disentangle.

To further explore this possibility, we performed additional longitudinal repeated-measures analyses incorporating configuration changes as a between-subject factor across all study visits (−90, −30, −14, +14, +30, and +90 days). Consistent with our original findings, longitudinal TIR_70–180 evolution differed according to baseline glycemic status (time × baseline TIR interaction: F = 5.421, p = 0.022). However, no significant interaction was observed between longitudinal TIR changes and configuration modifications (time × configuration changes interaction: F = 0.422, p = 0.518). Furthermore, the triple interaction among time, baseline TIR subgroup, and configuration changes was not significant (p = 0.818).

These exploratory analyses suggest that the greater improvement observed in participants with baseline TIR_70–180 <70% was unlikely to be explained exclusively by therapy re-tuning alone. At the same time, we agree with Dr. Zhang that therapy optimization likely contributed, at least partially, to the overall clinical evolution observed during follow-up. Rather than representing competing explanations, sensor performance, behavioral adaptation, and individualized parameter optimization probably acted synergistically within the real-world ecosystem of advanced hybrid closed-loop therapy.

Regarding the increase in auto-correction boluses after transition to Dexcom G7, we agree that this finding may represent an interesting mechanistic signal. Although our additional analyses did not support a predominant effect of configuration changes alone, future studies specifically designed to characterize algorithmic behavior during sensor transitions—including objective assessment of automated mode interruptions, connectivity metrics, and parameter modifications—would provide valuable additional insight.

We also thank Dr. Zhang for identifying the reporting inconsistencies in the published manuscript. The correct number of participants with baseline TIR_70–180 <70% was n = 20, as reported in table 1. Likewise, the HbA_1c values in the Results section were inadvertently inverted during manuscript preparation; the correct values are consistent with tables 1 and 3, and the abstract, showing higher baseline HbA_1c in participants with TIR_70–180 <70% (7.3% ± 0.9% vs 6.3% ± 0.5%). We appreciate the opportunity to clarify these points.

Finally, we agree that integrating patient-reported experience measures with objective device-derived metrics represents an important future research direction, particularly for understanding how perceived reliability, workflow burden, and user confidence influence psychosocial outcomes in advanced hybrid closed-loop systems. ⁴

In summary, we share the view that parameter re-tuning, sensor characteristics, behavioral adaptation, and perceived reliability are all relevant contributors to real-world outcomes after transitioning from Dexcom G6 to G7 in Control-IQ users. Our study, designed as a pragmatic multicenter observational evaluation, cannot fully disentangle these interdependent factors. Nevertheless, it provides reassuring safety data, suggests exploratory benefit in individuals with suboptimal baseline glycemic control, and documents favorable patient-reported outcomes following the transition. We appreciate the opportunity to further refine the interpretation of our findings and to highlight directions for more granular, hypothesis-driven research in this evolving area of diabetes technology.

We thank Dr. Zhang again for his interest in our work and for contributing to a constructive scientific discussion around the interpretation of real-world outcomes in diabetes technology.