Methodological and Statistical Considerations Regarding the Assessment of Psychological Distress and Disordered Eating in Obesity: A Letter to the Editor Authors: Dandan Chi, Yuntao Guan Published: 2026 Dear editor We read with great interest the article by Cyranka et al,1 which examined the prevalence of psychological distress and disordered eating behaviors in adults attending an obesity outpatient clinic. The authors are to be commended for addressing the often-overlooked intersection between mental health and metabolic disease in a real-world clinical setting, and for emphasizing the need for integrated multidisciplinary care. The findings provide valuable insights into the high burden of affective symptoms and maladaptive eating patterns in this population. After careful review of the methodology and statistical presentation, we would like to offer a few observations that we hope may contribute to the ongoing discussion and further strengthen the interpretation of the results. These considerations pertain to sample size justification, consistency in outcome reporting, the appropriate use of screening instruments, confounder adjustment, and the handling of multiple comparisons. Our first observation relates to the sample size. The study included 48 participants, which is reasonable for an exploratory analysis in a real-world clinical setting. However, given that multiple inferential analyses were performed-including three-group comparisons across obesity classes, gender comparisons, and correlation analyses-the study may have been underpowered to detect modest but clinically meaningful effect sizes, particularly in subgroup analyses such as the gender comparison, which involved only 13 males vs 35 females. We would appreciate the authors’ perspective on whether a priori sample size estimation was considered, and whether the exploratory nature of the study influenced the interpretation of nonsignificant findings. As demonstrated by Gorman et al, careful sample size planning is essential for ensuring adequate statistical power in psychological and educational research, and we wonder whether similar considerations were applied in the present study. We also noted a discrepancy between the abstract and the results section regarding the prevalence of depressive symptoms. The abstract reports that 54.2% of participants had moderate to very severe depressive symptoms, whereas Table 3 indicates that 33.3% (n=16) fell into that category, with an additional 20.8% (n=10) classified as having mild symptoms. It would be helpful if the authors could clarify whether different cutoffs or classification schemes were applied between these two presentations, as this would assist readers in accurately interpreting the burden of depressive symptoms in this cohort. Regarding the use of the Eating Attitudes Test-26 (EAT-26), we acknowledge that this is a widely used and well-validated screening tool. However, we wish to note that the EAT-26 is designed to identify risk for disordered eating rather than to establish a formal diagnosis. The authors classified 14.6% of participants as having a “full eating disorder” based solely on EAT-26 scores. While this may have been intended as an approximation, we wonder whether the authors considered using complementary diagnostic interviews or additional clinical assessments to confirm this classification. Recent psychometric evaluations by Papini et al have raised important concerns about the instrument’s performance in certain populations, including differential item functioning based on obesity status, which suggests that caution is warranted when interpreting EAT-26 scores in obesity cohorts.2 Clarifying this point would help readers appreciate the distinction between screening results and formal diagnostic status. We also observed that the correlation analyses between anxiety symptom severity and metabolic parameters-specifically HbA1c and percent body fat-were presented without adjustment for potential confounding variables. Factors such as age, sex, diabetes status, medication use, and comorbidities are known to influence both metabolic control and psychological symptoms. We would be interested to know whether the authors considered multivariable regression or partial correlation analyses to further explore these relationships. Research by Shearrer has demonstrated that the relationship between glycemia and psychological symptoms is complex and may be influenced by neural and metabolic pathways, highlighting the importance of careful confounder adjustment in obesity research.3 Furthermore, evidence from a representative German population study indicates that individuals with obesity demonstrate significantly higher values in depression and anxiety symptoms compared with normal-weight individuals, and that general health status fully mediates the association between BMI and these psychological outcomes, reinforcing the need for comprehensive confounder evaluation in such analyses.4 In addition, we noted that the overall ANOVA for EAT-26 scores across obesity classes was marginally nonsignificant (p=0.054), yet post-hoc testing yielded a statistically significant difference between class I and class III. While the authors report using Tukey’s HSD or Bonferroni correction for post-hoc comparisons, it would be helpful to know whether all pairwise tests were subjected to the same correction and whether these analyses were pre-specified. Recent guidance by Wheelington et al emphasizes the importance of methodological rigor in obesity research, particularly regarding the handling of multiple comparisons and the transparent reporting of analytic decisions, to enhance the reproducibility of findings.5 Finally, we would like to acknowledge the authors’ use of a multidisciplinary approach and their emphasis on the psychological dimensions of obesity care. These are important contributions to the field. We hope that our observations-offered in the spirit of constructive dialogue-may assist the authors in refining their interpretation and in guiding future research in this area. We thank the editors for the opportunity to share these thoughts and look forward to the authors’ response.