Letter to the Editor: Enhancing the methodological rigor and interpretation of findings in a cross-sectional study on drinking water type and kidney stone risk in U.S. adults

Dear Editor, We read with interest the article entitled “Association between drinking water type and kidney stone risk in U.S. adults: a cross-sectional analysis of NHANES 2009–2016 data”[1], published in the International Journal of Surgery. This study investigates the association between drinking water type and kidney stone prevalence in a nationally representative sample of U.S. adults. The authors report an inverse association between tap water intake and kidney stone risk, whereas bottled water consumption was not significantly associated with the outcome. Despite these findings, several methodological and interpretative issues warrant further consideration to strengthen the robustness and credibility of the conclusions. First, although the authors adjusted for multiple covariates in multivariable models, control of potential confounding may be incomplete. Key socioeconomic indicators, including educational attainment and household income, were not included in the adjustment strategy. These factors are established determinants of health-related behaviors and may influence preferences for tap versus bottled water[2]. Their omission may result in residual confounding and bias the observed associations. Inclusion of socioeconomic variables in future analyses would enhance internal validity and strengthen the assessment of associations. Second, participants with incomplete covariate data were excluded from the analytic sample. In cross-sectional studies, such complete-case analysis, if not adequately justified, may reduce statistical power by decreasing the effective sample size. More importantly, it may introduce selection bias when missingness is related to participant characteristics or outcomes, rather than occurring completely at random. In addition, missing covariate data were not further examined in sensitivity analyses. We suggest that future studies consider advanced methods, such as multiple imputation, to handle missing data more appropriately, preserve sample size, and yield more precise and generalizable estimates[3]. Third, the study applied K-means clustering to identify drinking water consumption patterns. However, the added value of this unsupervised approach was not clearly demonstrated, particularly given that conventional logistic regression analyses had already shown a dose–response relationship[4]. In addition, K-means clustering is sensitive to the prespecification of cluster number and initial centroids, which may limit result stability and interpretability in epidemiological research[5]. We suggest that the authors provide a clearer justification for the use of clustering methods and their advantages over traditional regression approaches, or explicitly discuss their methodological limitations to improve analytical transparency and rigor. Beyond these methodological considerations, further discussion is warranted regarding the potential for reverse causality. In this cross-sectional study, the exposure of interest, namely drinking water type, may not be independent of the outcome. Specifically, a history of kidney stone disease may influence individuals’ subsequent choices regarding drinking water. Individuals diagnosed with kidney stones are commonly advised to increase fluid intake or to modify their drinking water source. As a result, the reported water consumption patterns may reflect behavioral changes after diagnosis rather than habitual exposure prior to stone formation. This lack of clear temporal ordering may bias the observed associations and should be explicitly acknowledged to avoid over interpretation of the findings. In summary, this study provides useful insights into the association between drinking water type and kidney stone risk in the U.S. population. Addressing the methodological and interpretative issues noted above would strengthen the robustness of the analyses, support clearer causal interpretation, and enhance the study’s overall contribution to the field.