Abstract
To the Editor:
We read with great interest the recent article by Chang et al, “Association of GLP-1 Receptor Agonist Use With Risk of Lumbar Degenerative Disease and Spine Surgery in Patients With Type 2 Diabetes: A Propensity Score–matched Cohort Study.” 1 The authors should be congratulated for addressing an important and clinically relevant question using a large real-world dataset. In a matched cohort of 196,435 pairs, GLP-1 receptor agonist (GLP-1 RA) use was associated with a lower 3-year and 5-year incidence of lumbar degenerative disease (LDD), and a lower 5-year incidence of lumbar spine surgery. These findings are of considerable interest.
We believe that one methodological issue worth further discussion in this study is that some degree of imbalance in key variables appears to persist after propensity score matching, which may have some implications for the interpretation of the independent protective effect of GLP-1 RA. Given that both HbA1c and BMI are important metabolic factors closely associated with GLP-1 RA exposure and LDD risk, and that the authors themselves discuss weight loss and improved glycemic control as key mechanisms through which GLP-1 RA may exert a potential spinal protective effect, these residual differences may warrant additional clarification. In this context, the observed reduction in LDD risk might partly reflect metabolic differences that remained after matching, rather than being solely attributable to the independent effect of the drug itself. Therefore, we suggest that the current findings may be more appropriately interpreted as an insightful correlation signal, while whether GLP-1 RA has a clear disease-modifying effect remains to be further validated in future research.
We believe that the interpretation of outcomes and underlying mechanisms related to lumbar spine surgery may warrant further caution and refinement. Although the incidence of lumbar spine surgery in the GLP-1 RA group was statistically lower than that in the control group at 5-year follow-up, this result undoubtedly holds certain significance. However, the absolute difference was only approximately 0.1 percentage points, suggesting that the actual effect size may be relatively limited from a clinical perspective. Particularly in large sample sizes approaching 200,000 cases per group, even a small absolute difference could achieve statistical significance.
Meanwhile, the authors proposed that weight loss, anti-inflammatory effects, and improved glycemic control may serve as potential mechanisms, which is biologically plausible and insightful. However, we also observed that there may be a certain degree of overlap between key metabolic factors involved in these mechanisms and baseline characteristics that remained distinct even after matching. Since the study did not provide data on weight changes during follow-up, extent of glycemic improvement, or mediating analyses, these findings should currently be interpreted more as a noteworthy correlation signal. Further research is needed to clarify whether GLP-1 RA agents exert independent and clinically significant protective effects on spinal degenerative progression or surgical requirements.
In summary, these considerations do not diminish the importance of the study. On the contrary, they help refine the interpretation of an otherwise innovative and thought-provoking analysis. We respectfully suggest that the conclusions may be strengthened by emphasizing that the current study supports a promising association between GLP-1 RA use and lower long-term risk of LDD. Additional analyses—such as post-matching adjustment for residual imbalance in HbA1c and BMI, longitudinal assessment of metabolic changes, or mediation-based approaches—could further clarify this relationship in future work. We appreciate the authors’ contribution and believe their work opens an important avenue for future investigation at the intersection of metabolic therapy and spinal degeneration.
