Colonoscopy-detected high-risk adenomas and their association with alcohol consumption among first-degree relatives:...

First-degree relatives (FDRs) of individuals with colorectal neoplasia are at increased risk of advanced colorectal lesions. However, risk stratification within this high-risk population remains suboptimal. This study aimed to identify factors associated with high-risk adenomas (HRAs) among FDRs and to evaluate the predictive performance and clinical utility of a multivariable model integrating lifestyle, clinical, and laboratory variables.

In this observational case-control study, 210 FDRs undergoing colonoscopy were enrolled, including 105 FDRs of probands with HRAs and 105 FDRs of probands with normal colonoscopy findings. Baseline characteristics, lifestyle factors, fecal immunochemical test (FIT) results, and serum biomarkers were collected. Univariable and multivariable logistic regression analyses were performed to identify factors associated with HRAs. Model performance was assessed using ROC curves, calibration analysis, and decision curve analysis (DCA), with internal validation using bootstrap resampling; however, no external validation was performed.

HRAs were detected in 34.8% of FDRs in this study sample. Alcohol consumption was significantly associated with higher odds of HRAs in both univariable (OR = 4.30, 95% CI: 1.54-11.98, P = 0.005) and fully adjusted models (OR = 5.75, 95% CI: 1.49-22.20, P = 0.011). In contrast, proband HRA status was not significantly associated with HRAs among FDRs. The prediction model demonstrated moderate discrimination (area under the curve = 0.707) and acceptable calibration, with potential clinical utility primarily at low-to-intermediate threshold probabilities, although net benefit became less stable at higher thresholds. The observed difference in the proportion of FDRs with HRAs between groups was smaller than expected.

A substantial proportion of FDRs had HRAs detected, underscoring the importance of targeted screening in this population. Alcohol consumption appears to be an important modifiable factor associated with HRAs, whereas proband HRA status provides limited additional value for risk stratification. The proposed model may aid in risk stratification, although further studies incorporating genetic and environmental factors and external validation are needed to improve predictive accuracy. These findings support the need for colonoscopy-based screening and targeted lifestyle interventions among FDRs.