We read with interest the article by Fons et al. showing that severe deficits on comprehensive geriatric assessment (CGA) were associated with all-cause and acute hospitalisations in older patients with inflammatory bowel disease (IBD) 1. The prospective multicentre design and focus on older outpatients are notable strengths. However, we believe that several aspects of the study warrant a more cautious interpretation of its clinical implications. The most important unanswered question is not whether CGA is associated with adverse outcomes, but whether it offers meaningful incremental value over simpler frailty screening tools 2-4. This is particularly relevant because the present study appears to be a longitudinal analysis of the same 405-patient cohort previously reported by the authors 2, 5. In their 2022 baseline paper, geriatric deficits were already shown to be prevalent and associated with disease activity and disease burden 5. In their 2024 follow-up study from the same cohort, Geriatric-8 (G8) frailty screening was associated with hospitalisation, decline in quality of life and functional decline 2. Therefore, the present report demonstrates that CGA is also prognostic in this cohort, but it does not show whether CGA performs better than G8 or meaningfully improves risk classification 1, 2, 5. Given that a recent meta-analysis identified substantial heterogeneity in frailty instruments used in IBD and emphasized the need for better risk-assessment tools, a direct head-to-head comparison of discrimination, calibration and reclassification would have been more informative 3, 4. Second, the multivariable adjustment strategy appears too limited to support stronger causal language. The main models were adjusted only for age, sex and biochemical disease activity, whereas prior IBD frailty studies have considered broader clinical context, including comorbidity burden, healthcare utilisation and treatment-related factors 6-8. This matters because frailty in IBD overlaps closely with disease severity and overall clinical complexity. Indeed, the authors' own earlier work showed that geriatric deficits were associated with active disease 5. Thus, the findings may be more appropriately interpreted as showing that CGA captures a subgroup with greater global vulnerability, rather than proving that ‘frailty itself’ has been cleanly separated from disease burden. Third, the observed association seems driven largely by the somatic domain and its components, including comorbidity, polypharmacy and malnutrition. These are not only elements of frailty, but also established predictors of hospitalisation in their own right. Without demonstrating that CGA adds prognostic information beyond these conventional risk factors, the claim of independent added value remains incomplete 6, 7. Finally, the study does not fully address the burden of recurrent hospitalisation. Frailty often manifests through repeated admissions rather than only first events, and prior IBD studies have linked frailty to readmission and cumulative hospitalisation burden 6, 7. Recurrent-event analyses would therefore have provided a more clinically relevant estimate than time-to-first-event modeling alone. Taken together, these findings support the importance of frailty in older IBD, but they do not yet establish the incremental clinical utility of CGA over simpler screening tools or justify routine implementation on this basis alone. Chang Liu: conceptualization, writing – original draft, writing – review and editing, investigation, project administration, resources. Danyang He: conceptualization, investigation, writing – original draft, writing – review and editing, project administration, resources. Ai Wei: investigation, conceptualization, writing – original draft, writing – review and editing, visualization, resources. Liu Xu: conceptualization, project administration, resources, supervision, writing – review and editing, writing – original draft. The authors have nothing to report. This article is linked to Fons et al. paper. To view this article, visit https://doi.org/10.1111/apt.70594. The data that support the findings of this study are available from the corresponding author upon reasonable request.
Liu et al. (Thu,) studied this question.