This is a study based on a large CRC database from West China Hospital of Sichuan University, covering people from all over China, so the sample has good regional representativeness. For the samples obtained after strict inclusion and exclusion, we used an algorithm based on the Cox proportional hazards regression model to group age, and on this basis, we analyzed the difference in postoperative prognosis between young and older patients. To our knowledge, this is the first attempt in this field.

An aggregate of 2374 patients was encompassed within this study, including 1335 in the younger cohort and 1039 in the older cohort. After PSM, 784 patients from each group were enrolled at a 1:1 ratio. No significant differences were observed in the baseline data after PSM implementation. The multivariate analysis results revealed that older age was an independent risk factor for both OS and DSS. Overall OS and DSS were better in the younger group compared to the older group after PSM. In different tumor pathological stages after PSM, both OS and DSS were better in the younger group than in the older group for stages II, III, and IV CRC. The majority of OS and DSS at each time point in the younger group were better than those of the older group when comparing different age groups.

The effect of age on long-term survival in CRC patients remains a contentious issue. Zhao et al. [8] reported that younger age constitutes an independent risk factor for OS and DSS, with older patients demonstrating better OS and DSS outcomes. In addition, several studies reported no significant differences in OS or DSS between younger and older groups [20,21,22]. Nakayama et al. [10] proposed that younger patients exhibit equivalent or improved prognoses compared to their older counterparts. Wang et al. [23] reported that younger patients display better DSS. Liu et al. [24] reported that older age constitutes an independent risk factor for OS, with younger patients exhibiting better OS.

It is crucial to recognize that disparate methods of dividing age groups may impact the results when comparing the prognoses of younger and older patients. Previous studies have employed different cut-off ages, such as 35, 40, 44, 45, and 50 years old, frequently consulting established guidelines [8, 10, 20,21,22,23,24]. Notably, 50 and 45 years old are often selected as the cut-off ages, at which the American Cancer Society (ACS) recommends CRC screening [12]. These are based on United States-centric models, whereas some others have been developed globally [25, 26], rendering it potentially inappropriate to directly adopt cut-off ages from specific previous studies or guidelines. In the present study, utilizing the optimal cut-off age determined by the Cox proportional hazards regression model for age group division, based on the incorporated patient data, might be more suitable. The discovery of improved long-term survival in younger CRC patients under this age group division, which was inconsistent with our clinical observation that young patients with colorectal cancer may have a worse prognosis, may offer novel insights for further elucidating the association between age and prognosis in CRC patients. We may consider whether more intensive comprehensive treatments can bring better long-term prognosis to young patients with colorectal cancer. At the same time, suggestions can also be made for adjusting treatment plans for elderly patients.

For treatment, this may suggest that younger CRC patients may benefit better from CRC resection with preoperative and postoperative treatment in longer treatment cycles and higher intensity. Therefore, young CRC patients should have a more open attitude toward CRC resection. As for the specific age dividing line, perhaps it should be determined in a way that is suitable for local patients. Through an algorithm, we obtained a cut-off age, which might provide partial reference for surgical decision-making for patients of different ages, but its representativeness need to be confirmed by studies with larger samples and in more centers.

What needs to be clarified is that even so, the surgical benefits of older CRC patients should not be completely ignored. Turri et al. [27] thought that the preoperative identification of risk factors for low OS may help the selection of those old patients who may benefit from curative CRC surgery. Willemsen et al. [28] found that a considerable number of octogenarian colorectal cancer patients can still achieve 5-year survival after surgery. Perhaps we can also refer to the postoperative management of gastric cancer found by Qiu et al. [29], that is, to develop personalized follow-up strategies according to age and postoperative time, in order to detect recurrence as early as possible and decide on further treatment, so as to control DSS in old CRC patients.

Previous studies on the relationship between age and prognosis in CRC patients in particular tumor pathological stages have revealed analogous yet not identical results compared to the current study [23, 24]. These findings, on one hand, suggested that younger patients are inclined to exhibit better prognosis than older patients across various tumor pathological stages and, on the other hand, suggested that tumor pathological stage may be a critical independent predictor of prognosis in CRC patients. In fact, the multivariate analysis in this study indicated that tumor pathological stage was the largest independent predictor of OS, barring age, and the primary independent predictor of DSS. Consequently, it would be worthwhile to further explore the association between age and prognosis in CRC patients across different tumor pathological stages.

The multivariate analysis in this study revealed that besides the two most crucial factors—age and tumor pathological stage—tumor differentiation and surgical characteristic also function as independent predictors of both OS and DSS. Tumor stage remains the most pivotal prognostic factor in CRC; more advanced tumor stages and inferior morphological factors, such as lower tumor differentiation, typically signify poorer prognoses [30]. Poorer surgical characteristic entail greater residual tumor presence, resulting in more restricted improvements in tumor morphology, which may contribute to less favorable prognoses.

Regarding the elucidation of the relationship between age and prognosis in CRC patients in this study, the evidence remains circumscribed. In comparison to younger CRC patients, older CRC patients tend to exhibit concomitant diseases, such as chronic conditions, and older patients with concomitant diseases may be associated with inferior OS outcomes [31]. For early-onset CRC, despite enhanced treatment adherence and elevated treatment intensity in younger patients, disease biology is more unfavorable, characterized by advanced tumor stages, lower cell differentiation, and heightened prevalence of signet-ring cell carcinoma [32, 33]. However, younger people make up only a small proportion of CRC patients. Therefore, on a population basis, their prognosis may be more favorable than their older counterparts when controlling for disease, patient, and treatment factors [34]. It is well known that older patients are less tolerant of chemotherapy, which may also not conducive to prognosis. Furthermore, it is imperative to study whether additional age-related factors may influence the prognosis of CRC patients. There may be many reasons why the prognosis of young colorectal cancer patients is better than old ones, including nonspecific and specific reasons and direct and indirect reasons.