Response to the letter to the editor from Pion et al. regarding “Different Carbohydrate Sources in Dog Foods Supported Overall Health and Cardiac Function”

Dear Editor, We thank Pion et al. for their thoughtful engagement with our study, “Different Carbohydrate Sources in Dog Foods Supported Overall Health and Cardiac Function.” We welcome their letter as a valuable opportunity to engage in the important scientific dialogue that moves our field forward. The purpose of this response is to clarify the interpretation of our data and the study’s original conclusions, which were based on a careful and deliberate analysis of the complete clinical picture. A core principle of responsible research is the critical distinction between statistically significant changes and clinically relevant outcomes (Carpenter et al. 2021; Sharma 2021; Willigenburg and Poolman 2023). In keeping with that principle, we reported our findings accurately and without overinterpretation. While Pion et al. correctly identify statistically significant differences in the grain-inclusive with pea protein and fiber (G+PPF) group, cardiac measures for all dogs remained within established normal reference ranges. Thus, our interpretation stops short of claiming “harm” or “myocardial dysfunction” because the data are not sufficient to make such a clinical diagnosis. This principle of cautious interpretation is central to understanding our perspective on both the primary echocardiographic data and the potential influence of confounding factors that introduce bias into the findings like chronic degenerative valvular disease (CDVD) or disproportionate population losses. Pion et al. express concern over the statistically significant changes in the G+PPF group, specifically increased weight-adjusted left ventricular internal diameter at end systole (wLVIDs) and decreased fractional shortening (FS) and ejection fraction (EF) compared to the other dietary treatments, as definitive indicators of reduced systolic function. However, this interpretation overlooks the critical clinical context: all individual and group mean values remained well within the normal reference ranges established from meta-analysis. In our opinion, it would have been an overinterpretation to claim harm based on these findings. The veterinary radiologist who conducted the echocardiograms deemed all results clinically normal with respect to dilated cardiomyopathy (DCM). This reinforces the distinction between a statistically significant observation and a clinically relevant finding. We have no evidence that the changes seen would have continued or progressed into clinical disease in this dietary group or that any other dietary group would have developed changes in heart function if followed further beyond the 18 months of this study. While our study is one of the largest, longest studies to date on this topic, the conclusions made by Pion et al. go beyond the scope of this study. We acknowledge this limitation and encourage these research questions to be the focus of future work. Just as we were cautious in interpreting these primary findings, we were similarly circumspect in discussing potential confounding factors. We acknowledge the authors’ expertise regarding the typical echocardiographic presentation of CDVD. However, our manuscript does not claim the changes in the G+PPF group to be solely attributable to CDVD. Rather, we presented the greater prevalence of CDVD in the G+PPF group as “one possible explanation” for the observed changes, recognizing it as a study limitation and potential confounding factor. Discussing such limitations is a fundamental component of transparent research. This position is supported by the radiologist’s reports, which confirmed CDVD diagnoses in several dogs. Pion et al. state that the purpose of our study was to determine if the diets caused harm; however, the purpose of our study was to evaluate the potential link between grain-free diets and DCM, with the underlying hypothesis that the otherwise nutritionally similar test foods investigated in this study would not negatively impact overall health or heart function indices. Based on these data, where cardiac values for all dogs remained within normal reference ranges, we conclude that, within the limitations of this study, all diets appeared to support both overall health and cardiac function. There was no intent to “ignore or explain away” findings, but rather a deliberate choice to avoid overinterpretation of the data available while acknowledging and discussing the observed statistical differences between dietary treatments. We believe the shared goal of ensuring canine health is best served by rigorous studies and responsible scientific dialogue. We recognize that more research is necessary in this complex field, including evaluating any potential longer term or subclinical effects that were not observed in our study. We welcome further contributions from the scientific community, as continued collaboration between veterinary nutritionists and cardiologists is essential to advancing our collective understanding and to optimizing the health of all dogs. Sincerely, Morris, Stiers, Hancock, and Gross Elizabeth Morris (Writing—original draft, Writing—review & editing), Cheryl A. Stiers (Writing—original draft, Writing—review & editing), Leslie B. Hancock (Writing—original draft, Writing—review & editing), and Kathy L. Gross (Writing—original draft, Writing—review & editing)