Abstract
Dear Editor:
We thank the authors for their thoughtful comments and for the careful reading of our work. We appreciate the opportunity to further clarify several aspects of our study.
We agree that some degree of residual imbalance in metabolic variables, particularly hemoglobin A1c and body mass index, remained after propensity score matching. These factors are closely related to both GLP-1 receptor agonist use and the risk of lumbar degenerative disease, and we acknowledge that residual confounding cannot be completely excluded in this type of observational analysis. This is an important point and one that we also considered when interpreting our results.
We also appreciate the authors’ perspective regarding the magnitude of the observed effect on spine surgery. While the difference reached statistical significance, the absolute reduction was small, and we agree that this should be interpreted with caution when considering clinical relevance.
At the same time, we would like to emphasize the intention of our study. Our goal was not to establish a definitive causal relationship, but rather to examine whether a signal exists in a large real-world dataset that might suggest a link between GLP-1 receptor agonist use and spinal outcomes. In this context, we view our findings as hypothesis-generating and as a starting point for further investigation rather than a conclusion in itself.
Notably, our findings are broadly in line with emerging literature suggesting a possible association between GLP-1 receptor agonist use and more favorable spinal or musculoskeletal outcomes, although the evidence remains heterogeneous and largely observational. For example, recent retrospective cohort and registry-based analyses have reported lower rates of spine surgery or postoperative complications among GLP-1 receptor agonist users, whereas randomized data focusing on musculoskeletal endpoints remain limited and sometimes inconclusive.1-3
We fully agree that additional studies are needed to clarify these associations. Future work incorporating more granular adjustment for metabolic variables, mediation analysis, and prospective or randomized designs will be essential to clarify whether any observed benefits are driven by direct disease-modifying effects on the spine or are largely mediated through improvements in metabolic control and weight.
We thank the authors again for their constructive comments, which we believe help place our findings in an appropriate context and highlight key directions for future research.
