A commentary on “Extensive study on the associations of 12 composite inflammatory indices with colorectal cancer risk and mortality a cross-sectional analysis of NHANES 2001-2020”

Dear Editor, We read with great interest the recent article by Zhou et al[1], which systematically evaluated the associations of 12 inflammatory indices with colorectal cancer (CRC) risk and mortality. While this work provides valuable insights into the clinical application of inflammatory biomarkers, several methodological considerations merit further discussion. This communication adheres to the TITAN guidelines[2]. First, considerations regarding sample size and model stability. The risk analysis included 126 CRC cases, and the survival analysis included 71 death events, with nine covariates adjusted simultaneously. According to recommendations from methodological literature[3], the events-per-variable in the Cox proportional hazards model was approximately 7.9, which is slightly below the commonly discussed reference threshold. This may, to some extent, affect the stability of coefficient estimates (e.g., the odds ratio for the neutrophil-to-lymphocyte ratio fluctuated from 1.358 to 2.137 in the original manuscript). We acknowledge that the authors candidly discussed the limitation of a small sample size in the supplementary materials, which is certainly commendable. If future studies could further expand the sample size or attempt sensitivity analyses using methods such as Firth’s bias-corrected logistic regression[4] or least absolute shrinkage and selection operator regression, we believe the conclusions would become even more robust. Second, methodological choices for discriminative analysis of survival outcomes. The authors used standard receiver operating characteristic (ROC) curves to evaluate the discriminative ability of inflammatory indices for CRC mortality (Figures 7 and 8 in the original manuscript). We understand that this is an intuitive and commonly used presentation method. However, given that survival data are characterized by time dependence and censoring, standard ROC curves may not fully capture the dynamic predictive performance of these indices. In similar studies, time-dependent area under the curve or Harrell’s concordance index[5] are also worth considering. We suggest that the authors consider using these methods for complementary analyses in the future, as they may provide additional dimensions of information. Third, transparency in reporting missing data. The authors did not provide detailed information on how missing data were handled in the methods section. We understand that missing variables are a common phenomenon in large database analyses. If the authors could supplement information on the proportions of missing data for covariates such as body mass index, smoking, and alcohol consumption, as well as the methods used to address them (e.g., complete-case analysis or multiple imputation), it would help readers better assess the reliability of the results. According to the STROCSS 2025 guidelines[6], clear reporting of missing data is good practice for enhancing research transparency. In summary, we sincerely thank the author team for this systematic work, which provides a comprehensive overview of the associations between 12 inflammatory indices and CRC risk and prognosis, offering valuable data support and theoretical reference for future research. The above observations are by no means intended to negate the academic value of this study but rather to provide some methodological perspectives for discussion, with the hope of collectively promoting the continuous optimization and improvement of research design in this field. Expansion of sample size, appropriate selection of statistical methods, and transparency in data reporting remain three important pillars of quality control in observational studies. We believe that if the author team could further integrate multi-center data, adopt more rigorous statistical strategies, and validate their findings with prospective designs in future work, it would certainly provide higher-level evidence for the clinical application of inflammatory biomarkers in the early screening and prognostic assessment of CRC. We look forward to seeing more high-quality follow-up studies from this author team in this important field.