ABSTRACT
Background
Despite an overall decrease in colorectal cancer (CRC) mortality in recent decades, incidence and mortality of CRC among individuals younger than 50 (early onset CRC; EOCRC) has increased. Individual and population‐level exposures contribute to EOCRC, but it is not clear how rurality and traveling for care impact survival.
Methods
Using the National Cancer Database (NCDB, 2010–2022), we (1) characterized and compared the EOCRC population to the average‐onset (AOCRC) population, and (2) analyzed individual and population‐level sociodemographic and clinical factors associated with 5‐year survival. We used accelerated failure‐time models; hazard ratios (HR) and 95% confidence intervals (CI) are reported.
Results
Among 404,440 individuals with CRC (15.7% EOCRC), more EOCRC patients were non‐white (32.2% vs. 24.6%; p < 0.001), had rectal cancer (43% vs. 34%; p < 0.001), and presented at later stages compared to AOCRC patients. EOCRC patients traveled farther for care overall, and rural EOCRC patients with rectal cancer traveled farthest of any group (median 41.7 miles). In adjusted EOCRC survival models, compared to urban patients traveling for care, survival was worse for urban patients who did not travel (HR 1.09, 95% CI 1.05–1.12) and rural patients regardless of travel (rural, traveled: HR 1.12, 95% CI 1.07–1.18; rural, no travel: HR 1.16, 95% CI 1.09–1.22).
Conclusions
EOCRC individuals living in urban areas experienced improved survival when traveling farther for care, though their rural counterparts did not necessarily gain the same survival benefit by traveling. Using exposures and health behaviors to guide screening beyond age‐based guidelines could improve early detection and survival.
Keywords: disparities, distance to care, early‐onset colorectal cancer, EOCRC, rural‐urban, survival
1. Introduction
Colorectal cancer (CRC) remains a leading cause of cancer‐related mortality in the United States [1]. While overall CRC incidence and mortality have declined over the past several decades, the incidence of early‐onset colorectal cancer (EOCRC), defined as CRC diagnosed before age 50, has risen steadily, contrasting with declining rates among older adults [2]. Further, the age‐adjusted mortality rate of EOCRC increased by 56% between 1999 and 2020 [3]. Compared with average‐onset CRC (AOCRC), EOCRC exhibits unique clinical and sociodemographic characteristics [2, 4]. Patients with EOCRC are more likely to present with advanced‐stage disease, distal tumor locations, and aggressive histopathologic features [2, 5]. Additionally, EOCRC mortality is disproportionately higher among marginalized groups including racial and ethnic minorities [6].
Access to cancer care is multidimensional and often requires financial resources, reliable transportation, time off from work and caregiving duties, and health literacy to navigate healthcare systems [7, 8, 9]. Geographic access may impact outcomes as well, as rural residence has been associated with worse CRC survival, potentially due to delayed diagnosis and limited local oncology services in rural regions [10]. For some patients, traveling longer distances to receive care at higher‐volume or specialized centers may mitigate these disadvantages [11]. Those with EOCRC in particular may benefit from traveling to specialized centers which have multidisciplinary teams, host clinical trials, and can support patients with fertility preservation, financial challenges, psychosocial support, and address parenting concerns [12]. However, it is not known how regional and structural factors such as rurality, distance to care, and access to specialized facilities interact with individual characteristics to influence survival in EOCRC. Understanding these relationships is particularly important for EOCRC, where patients may face barriers related to employment, insurance coverage, and financial hardship.
We sought to examine how geographic and structural dimensions of access influence survival among patients with EOCRC compared with AOCRC. In order to identify targets for intervention to reduce EOCRC disparities, we evaluated differences in sociodemographic characteristics, geographic residence, distance traveled for care, treatment patterns, and surgical approach.
2. Methods
2.1. Data Source and Study Population
This study was deemed exempt from review by the Boston University Institutional Review Board. We used 2010–2022 data from the National Cancer Database (NCDB), a hospital‐based cancer registry produced by the American College of Surgeons and the American Cancer Society. The NCDB captures approximately 70% of newly diagnosed cancer cases in the United States (US) [13].
We included patients diagnosed with primary colon, rectosigmoid, or rectal adenocarcinoma between 2010 and 2022; survival data is not reported yet for those diagnosed in 2023 or later. Individuals with additional primary malignancies and those treated elsewhere from the reporting institution were excluded, as were those missing data for the following variables: clinical stage, surgery and surgical approach, chemotherapy, race, insurance status, rurality, distance to care facility, regional income and/or education, and status as alive or dead. Patients were categorized into two groups based on age at diagnosis: early‐onset colorectal cancer (EOCRC), defined as diagnosis before age 50, and AOCRC colorectal cancer (AOCRC), defined as diagnosis at age 50 or older [4].
Sociodemographic and clinical variables included age, sex, race and ethnicity (non‐Hispanic White, non‐Hispanic Black, Hispanic, other), Charlson‐Deyo comorbidity score (0, 1+), tumor site [colon (C180‐C189)/rectosigmoid (C199), rectum(C209)], stage at diagnosis, treatment modalities (surgery, chemotherapy), surgical approach for those who underwent surgery (open, laparoscopic, robotic), insurance status (uninsured, private, Medicaid, Medicare, other government provider), regional median household income and education levels (quartiles), rurality (urban, rural), distance traveled for care (“crowfly” straight line distance), facility type (Community Cancer Center, Comprehensive Community Center, Academic Comprehensive Cancer Center, Integrated Network Cancer Program), and census region. Education (percent of individuals without a high school degree) and median household income quartiles in NCDB change across diagnosis years are based on corresponding American Community Survey data [13, 14]. Based on prior literature, counties with US Department of Agriculture Rural‐Urban Continuum Codes (RUCC) 1–3 were considered urban, and those with RUCC 4–9 were considered rural [15, 16]. Patients were categorized based on both rurality and median distance traveled for care (7.9 miles for urban, 35.3 miles for rural) to create a combined rurality/distance traveled variable (urban + travel (≥ 7.9 miles) urban + no travel, rural + travel (≥ 35.3 miles), rural + no travel).
2.2. Statistical Analysis
Categorical variables were compared between EOCRC and AOCRC using chi‐square tests and continuous variables were analyzed with Wilcoxon rank‐sum tests. The primary outcome was 5‐year survival, defined as time from diagnosis to death (all‐cause) or last follow‐up in NCDB. Those who survived beyond 5 years were censored as having survived 5 years. We used an accelerated failure time (AFT) model with a Weibull distribution to relax the assumption for baseline proportional hazards.
We stratified AFT models by age‐group and included the following covariates: sex, race (non‐White vs. White (ref)), comorbidities (1+ vs. 0 (ref)), clinical stage at diagnosis [late‐stage (Stage IV) vs. early‐stage (Stage I–III) (ref)], site (rectum vs. colon/rectosigmoid (ref)) insurance status [public (Medicaid, Medicare, other government provider) vs. uninsured vs. private (ref)], rurality + distance traveled for care as a single, combined variable (urban + no travel, rural + travel, rural + no travel vs. urban + traveling for care (ref)), income (quartiles 1–3 vs. quartile 4 (ref)), education (quartiles 1–3 vs. quartile 4 (ref)), chemotherapy (yes vs. no (ref)), surgery (yes vs. no (ref)), and year of diagnosis (continuous). Accelerated failure times were converted to hazard ratios (HR); HR and 95% confidence intervals (CIs) are reported.
2.3. Subgroup Analyses
As surgical resection was associated with significantly lower mortality in the main model, we conducted a post‐hoc analysis of factors associated with survival among surgical patients. The surgical subgroup analysis included the previously described covariates as well as surgical approach (robotic vs. laparoscopic vs. open (ref)). As facility type and region are only available in the NCDB for individuals 40 and older at the time of diagnosis, a sensitivity analysis was performed to examine differences related to region (Midwest vs. South vs. West vs. Northeast (ref)) and facility type (Community Cancer Center vs. Comprehensive Community Center vs. Academic Comprehensive Cancer Center vs. Integrated Network Cancer Program (ref)) [17]; models included the same covariates described previously in addition to region and facility type. Lastly, we used chi‐square tests to study differences among urban EOCRC who traveled vs. those who did not. All subgroup analyses were analyzed by age‐group (EOCRC and AOCRC). All statistical analyses were conducted using SAS version 9.4 (SAS Institute Inc., Cary, NC) and a two‐sided p‐value < 0.05 was considered statistically significant.
3. Results
3.1. Cohort Characteristics
Among 404,440 individuals, 63,335 (15.7%) had EOCRC and 341,105 (84.3%) had AOCRC. The median age among EOCRC patients was 44 (IQR 39–47) and the median age of AOCRC was 66 (58–75). EOCRC patients were overall more non‐White (32.2% vs. 24.6%, p < 0.0001), without any comorbidities (87.7% vs. 71.9%, p < 0.0001), presented at later stages (p < 0.0001) and with more rectosigmoid and rectal cancer (53.2% vs. 42.4%, p < 0.0001), and lived in urban settings (86.6% vs. 84.2%, p < 0.0001). More had private insurance (71.3% vs. 36.3%) but more were also uninsured (7.0% vs. 3.7%), and had higher income (p < 0.0001) and educational attainment (p < 0.0001). A greater proportion of EOCRC patients received chemotherapy (73.7% vs. 52.2%, p < 0.0001). Among those who had surgery (64.5% of both groups), EOCRC patients underwent more robotic surgery (17.1% vs. 13.0%, p < 0.0001; Table 1). EOCRC diagnoses increased over time; of all EOCRC in the NCDB, 7.5% were diagnosed in 2010 vs. 8.5% diagnosed in 2022. In the same timeframe, AOCRC diagnoses decreased over time (8.6% in 2010 vs. 7.0% in 2022). EOCRC patients traveled farther for care, which remained true for both colon and rectal cancer and those living in both urban and rural regions (all p < 0.0001; Table 2).
Table 1.
Sociodemographic and clinical characteristics of study population.
Early‐onset n = 63,335 (15.7%)
Average‐onset n = 341,105 (84.3%)
p value
Male
34,719 (54.8)
184,386 (54.1)
0.0004
White (non‐Hispanic)
42,962 (67.8)
257,242 (75.4)
< 0.0001
Black (non‐Hispanic)
8,940 (14.1)
43,146 (12.7)
Hispanic
7,075 (11.2)
22,813 (6.7)
Other
4,358 (6.9)
17,904 (5.3)
No comorbidities
55,569 (87.7)
245,227 (71.9)
< 0.0001
Stage I
14,528 (22.9)
103,180 (30.3)
< 0.0001
Stage II
8,824 (13.9)
59,282 (17.4)
Stage III
16,050 (25.3)
62,137 (18.2)
Stage IV
23,933 (37.8)
116,506 (34.2)
Colon + rectosigmoid
35,589 (56.2)
222,158 (65.1)
< 0.0001
Rectum
27,746 (43.8)
118,947 (34.9)
Received chemotherapy
46,681 (73.7)
177,877 (52.2)
< 0.0001
Received immunotherapy
8,585 (13.6)
26,743 (7.8)
< 0.0001
Underwent surgery
40,824 (64.5)
220,134 (64.5)
0.7052
Open
18,009 (44.1)
101,728 (46.2)
< 0.0001
Robotic
6,963 (17.1)
28,666 (13.0)
Laparoscopic
15,852 (38.8)
89,740 (40.8)
Rural
8,474 (13.4)
53,989 (15.8)
< 0.0001
Private insurance
45,146 (71.3)
123,844 (36.3)
< 0.0001
Public insurance (Medicare, Medicaid, Other Government)
13,771 (21.7)
204,683 (60.0)
Uninsured
4,418 (7.0)
12,578 (3.7)
Income Q1
10,899 (17.2)
63,400 (18.6)
< 0.0001
Income Q2
13,135 (20.7)
77,558 (22.7)
Income Q3
14,883 (23.5)
81,481 (23.9)
Income Q4
24,418 (38.6)
118,666 (34.8)
Education Q1
14,842 (23.4)
78,755 (23.1)
< 0.0001
Education Q2
17,578 (27.8)
99,619 (29.2)
Education Q3
17,547 (27.7)
96,534 (28.3)
Education Q4
13,368 (21.1)
66,197 (19.4)
Facility type
a
Community
3,052 (6.7)
31,058 (9.1)
< 0.0001
Comprehensive community
16,146 (35.3)
137,409 (40.3)
Academic
18,262 (39.9)
107,176 (31.4)
Integrated Network
8,290 (18.1)
65,438 (19.2)
Region
a
Northeast
8,478 (18.5)
69,692 (20.4)
< 0.0001
Midwest
10,942 (23.9)
86,423 (25.3)
South
17,962 (39.3)
127,455 (37.4)
West
8,371 (18.3)
57,535 (16.9)
Table 2.
Straight‐line distances (miles) traveled for care among study cohort (centroid point of patient’s zip code to the treating facility), median and interquartile ranges (IQR) presented.
Distance traveled for care
Early‐onsetn = 63,335 (15.7%)
Average‐onsetn = 341,105 (84.3%)
p value
Overall (median, IQR)
10.8 (4.9–24.0)
9.0 (4.1–20.6)
< 0.0001
Colona (n = 257,747)
10.1 (4.6–22.3)
8.3 (3.8–18.6)
< 0.0001
Rural
37.4 (19.5–67.2)
31.7 (15.9–55.1)
< 0.0001
Urban
8.9 (4.3–17.6)
7.2 (3.5–13.9)
< 0.0001
Rectum (n = 146,693)
11.8 (5.4–26.4)
10.7 (4.7–24.9)
< 0.0001
Rural
41.7 (22.9–71.9)
38.8 (21.1–64.5)
< 0.0001
Urban
10.0 (4.8–19.7)
8.6 (4.2–17.3)
< 0.0001
3.2. Survival
Unadjusted 5‐year survival was 41.4% among EOCRC (stages I–III: 57.4%, stage IV: 14.9%) and 36.2% among AOCRC (stages I–III: 50.6%, stage IV: 8.6%). Females had lower mortality in both groups, with a stronger protective association in the EOCRC population (HR 0.87, 95% CI 0.84–0.89; Figure 1) compared with the AOCRC population (HR 0.98, 95% CI 0.97–0.99). The effect of race differed between age groups, as race was not significant for EOCRC patients (HR 0.99, 95% CI 0.96–1.02), but non‐white AOCRC patients had lower mortality (HR 0.84, 95% CI 0.83–0.85). Those with comorbidities had higher mortality in both cohorts (EOCRC: HR 1.21, 95% CI 1.16–1.25; AOCRC: HR 1.32, 95% CI 1.31–1.34). Rectal cancer was associated with lower mortality relative to colon in both groups (EOCRC: HR 0.81, 95% CI 0.79–0.84; AOCRC: HR 0.86, 95% CI 0.85–0.87). Late‐stage disease conferred worse mortality in both populations, but to a lesser degree among AOCRC (EOCRC: HR 7.21, 95% CI 7.07–7.34; AOCRC: HR 5.72, 95% CI 5.68–5.76).
Figure 1.
Hazard ratios and 95% confidence intervals (errors bars present) for 5‐year survival in early‐onset colorectal cancer (EOCRC, blue diamonds) and average‐onset colorectal cancer (orange squares).
Among EOCRC, rural patients regardless of travel status (rural, no travel HR 1.16, 95% CI 1.09–1.22; rural, travel HR 1.12, 95% CI 1.07–1.18) and urban patients who did not travel for care had higher mortality (HR 1.09, 95% CI 1.05–1.12) compared to urban patients traveling for care. Similar results were seen in the AOCRC cohort (rural, no travel: HR 1.12, 95% CI 1.10–1.14; rural, travel: HR 1.03, 95% CI 1.01–1.05; urban, no travel HR 1.13, 95% CI 1.11–1.14). Among both age groups, the uninsured (EOCRC: HR 1.46, 95% CI 1.39–1.52; AOCRC: HR 1.46, 95% CI 1.43–1.50) and publicly insured (EOCRC: HR 1.41, 95% CI 1.37–1.45, AOCRC: HR 1.66, 95% CI 1.65–1.68) had higher mortality than those with private insurance. Lower income (EOCRC: HR 1.13, 95% CI 1.09–1.17; AOCRC: HR 1.10, 95% CI 1.09–1.11) and lower educational attainment (EOCRC: HR 1.05, 95% CI 1.01–1.09; AOCRC: HR 1.02, 95% CI 1.00–1.03) were similarly associated with increased mortality in both groups (Figure 1).
Treatment effects differed notably by age group. Compared to those who did not undergo surgery, surgical resection was protective in both populations, (EOCRC: HR 0.47, 95% CI 0.45–0.49; AOCRC: HR 0.41, 95% CI 0.41–0.42). Chemotherapy was associated with slightly decreased mortality among EOCRC patients (HR 0.95, 95% CI 0.91–0.99), but substantially reduced mortality in the AOCRC population (HR 0.49, 95% CI 0.48–0.50). Calendar year was inversely associated with mortality in both groups, indicating improving survival over time (EOCRC: HR 0.97 per year, 95% CI 0.97–0.97; AOCRC: HR 0.98 per year, 95% CI 0.97–0.98; Figure 1).
3.3. Surgery
Among those who underwent surgery, non‐White race was associated with higher mortality among EOCRC patients (HR 1.05, 95% CI 1.00–1.11; Figure 2). Additionally, chemotherapy was associated with higher mortality among EOCRC surgical patients (HR 1.69, 95% CI 1.57–1.83), and lower mortality among AOCRC surgical patients (HR 0.76, 95% CI 0.74–0.77). Geographic residence and travel patterns remained associated with mortality after surgery: relative to urban patients who traveled for care, rural patients—both those who traveled and those who did not—had higher mortality in both cohorts. Urban patients who did not travel also had higher mortality (EOCRC: HR 1.06, 95% CI 1.01–1.12; AOCRC: HR 1.13, 95% CI 1.11–1.15). Surgical approach was significantly associated with survival in both cohorts: compared with open surgery, robotic (EOCRC: HR 0.72, 95% CI 0.67–0.77; AOCRC: HR 0.66, 95% CI 0.64–0.68) and laparoscopic approaches (EOCRC: HR 0.75, 95% CI 0.72–0.78; AOCRC: HR 0.70, 95% CI 0.69–0.71) were associated with lower mortality (Figure 2). Similar trends to the overall cohort were observed related to sex, comorbidities, insurance, income, and education.
Figure 2.
Hazard ratios and 95% confidence intervals for 5‐year survival in early‐onset colorectal cancer (EOCRC, blue diamonds) and average‐onset colorectal cancer (orange squares) among those who underwent surgical resection.
3.4. Sensitivity Analysis: Facility Type and Region
In sensitivity analyses adjusting for facility type and geographic region (only available for individuals 40+ years old at diagnosis), compared with integrated network facilities, treatment at academic centers was associated with lower mortality in both groups (EOCRC: HR 0.93, 95% CI 0.89–0.97; AOCRC: HR 0.89, 95% CI 0.87–0.90). Community and comprehensive community centers were not significantly associated with mortality in EOCRC patients, whereas comprehensive community centers were associated with higher mortality in the AOCRC cohort (HR 1.05, 95% CI 1.04–1.07).
Regional differences in mortality were observed in both groups. Relative to the Northeast, patients treated in the South (EOCRC: HR 1.16, 95% CI 1.10–1.22; AOCRC: HR 1.13, 95% CI 1.12–1.15) and Midwest (EOCRC: HR 1.14, 95% CI 1.08–1.19; AOCRC: HR 1.13, 95% CI 1.11–1.15) experienced higher mortality in both populations. Treatment in the West was only associated with higher mortality in AOCRC patients (HR 1.07, 95% CI 1.05–1.09) (Figure S1). Patterns of association observed in primary analyses remained consistent for all variables in this sensitivity analysis: sex, race, comorbidities, stage, site of cancer, insurance, income, education, rurality + distance traveled, chemotherapy, surgery, and year of diagnosis.
3.5. Subgroup: Urban EOCRC
Among all urban EOCRC, 43.6% traveled for care; individuals who traveled were more often White (69.7% vs. 59.8%, p < 0.001), privately insured (75.2% vs. 68.1%, p < 0.001), and residing in regions with the highest income quartile (47.4% vs. 38.6%, p < 0.001). More urban EOCRC individuals who traveled underwent robotic surgery (19.5% vs. 14.3%, p < 0.001) (Table S1).
4. Discussion
This study is the first to evaluate both individual and regional factors impacting survival in EOCRC. We demonstrate that patients with EOCRC travel farther distances for care compared to AOCRC, and urban patients who traveled farther for care experienced longer survival. However, traveling for care did not confer a survival benefit for rural EOCRC populations. The persistent rural‐urban survival disparity in EOCRC after controlling for clinical and demographic factors suggests that rural patients may be impacted by unmeasured, unaddressed exposures or risk factors prior to diagnosis that are not mitigated by traveling for specialized care. Advocating for rural EOCRC patients to travel farther distances for care may increase travel burden without necessarily improving survival.
Although our survival models controlled for regional income, urban patients (including those living in metropolitan areas) traveling farther distances for care may represent a cohort of people with better access to private transportation, the ability to take more time off from work, and who are more likely to seek out specialized care at high‐volume, high‐performing centers [18, 19, 20]. A recent study using data from the National Household Transportation Survey found that those with higher income tended to travel farther distances for care in less time due to the use of private vehicles. Alternatively, those with lower incomes were more likely to utilize public transportation and traveled shorter distances for care but with a longer travel time [18]. Travel time is longer for EOCRC rural populations, and prior literature suggests that income and rurality intersect to influence survival: younger individuals in low‐income, rural areas face a 50% higher risk of CRC mortality compared to nonrural, nonpoverty areas [21]. Rural patients are less likely to receive chemotherapy and other guideline‐concordant therapies, potentially due to geographic barriers such as distance from radiation and infusion centers and limited local oncology networks. However, our results suggest that traveling for care does not fully mitigate the risk associated with living in a rural region.
Our finding that the rural‐urban survival disparity is not fully mitigated by traveling for care reinforces the notion that EOCRC survival is shaped by a complex interplay of individual and regional socioeconomic, behavioral, and structural factors that extend beyond physical proximity to specialized care. One study of the rural‐urban colon cancer survival disparity using the NCDB (all ages) found no difference in survival between urban and rural patients who traveled to high‐volume centers, with median travel distances of 40 and 108 miles respectively [11]. However, some research suggests that centralization of high‐volume cancer centers worsens access disparities among vulnerable populations, including rural communities and those with travel challenges [22, 23, 24]. For example, non‐White individuals and those with Medicaid have been shown to travel shorter distances and undergo surgery at lower‐volume centers for pancreatectomy, an operation consistently shown to have a strong volume‐outcome relationship [23]. Our results provide further insight into this issue, as we show that traveling farther for care does not necessarily improve survival for rural EOCRC patients.
Consistent with prior literature, we found that lower regional income and educational attainment were associated with worse survival [21, 25, 26]. This finding cannot be attributed solely to stage at presentation, as income and education were associated with survival regardless of stage. Regional income and education may be associated with community‐level health‐related resources and education, as well as risk factors not represented in our data such as exposure to toxic air and water pollutants [27, 28]. Several national studies have identified higher concentrations of carcinogenic environmental exposures in low‐income regions, with a recent study utilizing Environmental Protection Agency data finding a 51% higher burden of carcinogenic air emissions in census tracts with higher poverty and lower educational attainment independent of race and ethnicity [29, 30]. Environmental carcinogens are increasingly associated with CRC tumorigenesis, and some literature suggests a potential link between environmental carcinogens and EOCRC risk [31, 32, 33, 34].
While insurance status, income, and education are often related, we found that insurance status was associated with survival regardless of regional income and educational attainment. EOCRC patients without insurance or with public insurance experienced similarly poor survival compared to those with private insurance. This is a concerning finding, given those with public insurance should theoretically benefit from access to more preventative services than those without insurance. For example, as the United States Preventative Services Task Force (USPSTF) recommends initiating CRC screening at age 45 for average risk individuals, those with insurance, public or private, have access to free CRC screening starting at age 45 [35, 36]. However, structural factors including lower reimbursement rates from public insurance, more limited provider networks, and fewer resources at hospitals serving predominantly publicly insured populations may contribute to the insurance‐based survival disparities we identified [37, 38]. The similarly poor survival outcomes for the publicly insured and uninsured in this study warrants further investigation into drivers of this disparity.
Although the USPSTF recommends CRC screening starting at age 45, our results suggest that individuals with EOCRC often present with advanced disease prior to being screening‐eligible [36]. A growing body of literature demonstrates that CRC rates are increasing among those younger than 45, with the fastest increase among those younger than 40 [2, 3, 39]. In our EOCRC cohort of individuals aged 18–50 years, the median age of diagnosis was 44 with an interquartile range of 38–47. As CRC screening is not covered by insurers for most individuals younger than 45 in the United States, 55% of individuals in this EOCRC cohort would not have had access to routine CRC screening. Prompt recognition of symptoms that may raise concerns for CRC and access to expedited colonoscopy should be prioritized among younger individuals. Further, individualized risk assessments incorporating health behaviors, as well as environmental, familial, and genetic factors, may help to identify those who would benefit from earlier CRC screening [40, 41, 42]. A recent multicenter international study by Archambault et al. combined 16 environmental and lifestyle factors with 141 polygenic risk variants to develop a personal risk score to identify individuals at higher risk for EOCRC [43]. While such individualized risk scores are limited to research at this time, validation and clinical use could improve access to earlier CRC screening for those at high risk of EOCRC.
Finally, treatment modalities impacted survival differently among EOCRC and AOCRC. Chemotherapy was protective among AOCRC but did not confer a survival benefit among EOCRC. The tumor biology and immunologic profile of EOCRC is often distinct from AOCRC, and the response to systemic treatment may be more varied among EOCRC [4, 44, 45, 46]. Several studies have found that despite receiving more and stronger systemic treatment compared to AOCRC, EOCRC patients experienced no to minimal survival benefit stage by stage [47, 48]. There is limited data on EOCRC’s oncotherapeutic sensitivity to specific chemotherapy agents, and age at diagnosis is not currently incorporated into treatment guidelines. For both age groups, receipt of MIS was independently associated with improved survival compared to open surgery. This survival advantage persisted after adjusting for year of diagnosis, disease stage, and site of cancer, suggesting this finding is not merely a reflection of more recent or earlier‐stage interventions. Although patients undergoing open operations may have presented with obstruction or perforation, potentially contributing to their worse survival, prior literature suggests that MIS mediates rural‐urban outcome differences even when controlling for acuity [49]. These findings support the growing body of evidence favoring MIS due to its association with reduced perioperative morbidity and faster recovery, which our results and others suggest may ultimately influence long‐term survival [50, 51].
5. Limitations
The NCDB does not capture CRC screening data, which limits the ability to assess the impact of screening on survival at the patient level. Additionally, given our use of the NDCB, we could not determine why certain individuals may have traveled farther for care, i.e., due to limited resources in their region, whether they were seeking a second opinion, or due to a referral from a local provider to a different hospital for definitive treatment. Lastly, the “crowfly” variable in NCDB is a straight‐line measure of distance between the patient’s home and the reporting facility, and prior studies have suggested that measuring travel time in minutes or based on local road networks may be more accurate [52, 53].
Conclusion
For rural patients with EOCRC, traveling farther for care did not necessarily improve survival. Survival disparities in EOCRC based on geography and sociodemographic factors could be addressed with a multifaceted approach including enhanced early detection strategies, ensuring high‐quality local cancer care for rural EOCRC populations, and expanding access to MIS for EOCRC. Future work should aim to broaden CRC awareness and screening strategies among rural populations and those who do not meet screening guidelines based on age alone.
Conflicts of Interest
The authors declare no conflicts of interest.
Synopsis
Patients with early‐onset colorectal cancer (EOCRC) travel farther for care compared to those with average‐onset colorectal cancer (AOCRC), and rural EOCRC patients travel farthest. However, traveling farther for care did not mitigate the worse survival we identified among rural patients.
Supporting information
Figure S1: Sensitivity analysis including facility type (reference: Integrated Cancer Network) and region (reference: Northeast); data for facility type and region are only available for individuals 40 years and older at the time of diagnosis. Hazard ratios and 95% confidence intervals for 5‐year survival in early‐onset colorectal cancer (EOCRC, blue diamonds) and average‐onset colorectal cancer (orange squares) are represented.
Acknowledgments
Sara Myers is supported by National Research Service Award (NRSA) Institutional Postdoctoral Training Grant T32HP10028. Kelly Kenzik is supported by NIH National Cancer Institute grant R37CA266193. The contents of this manuscript may not represent the views of the HHS, the NIH, or the federal government.
Data Availability Statement
The data are available through the National Cancer Database.
References
-
1.
Siegel R. L., Giaquinto A. N., and Jemal A., “Cancer Statistics, 2024,” CA: A Cancer Journal for Clinicians
74 (2024): 12–49, 10.3322/caac.21820.
[DOI] [PubMed] [Google Scholar] -
2.
Jayakrishnan T. and Ng K., “Early‐Onset Gastrointestinal Cancers: A Review,” Journal of the American Medical Association
334 (2025): 1373–1385, 10.1001/jama.2025.10218.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
3.
Tan J. Y., Yeo Y. H., Ng W. L., Fong Z. V., and Brady J. T., “How Have Us Colorectal Cancer Mortality Trends Changed in the Past 20 Years?,” International Journal of Cancer
155 (2024): 493–500, 10.1002/ijc.34926.
[DOI] [PubMed] [Google Scholar] -
4.
REACCT Collaborative
Characteristics of Early‐Onset vs Late‐Onset Colorectal Cancer: A Review,” JAMA Surgery
156 (2021): 865–874, 10.1001/jamasurg.2021.2380.
[DOI] [PubMed] [Google Scholar] -
5.
Li J., Pan Y., Guo F., et al., “Patterns in Genomic Mutations Among Patients With Early‐Onset Colorectal Cancer: An International, Multicohort, Observational Study,” Lancet Oncology
26 (2025): 1055–1066, 10.1016/S1470-2045(25)00239-6.
[DOI] [PubMed] [Google Scholar] -
6.
Zaki T. A., Liang P. S., May F. P., and Murphy C. C., “Racial and Ethnic Disparities in Early‐Onset Colorectal Cancer Survival,” Clinical Gastroenterology and Hepatology: The Official Clinical Practice Journal of the American Gastroenterological Association
21 (2023): 497–506, 10.1016/j.cgh.2022.05.035.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
7.
Samoil D., Kim J., Fox C., and Papadakos J. K., “The Importance of Health Literacy on Clinical Cancer Outcomes: A Scoping Review,” Annals of Cancer Epidemiology
5 (2021): 3, 10.21037/ace-20-30. [DOI] [Google Scholar] -
8.
Palmer N. R., Weaver K. E., Hauser S. P., et al., “Disparities in Barriers to Follow‐Up Care Between African American and White Breast Cancer Survivors,” Supportive Care in Cancer: Official Journal of the Multinational Association of Supportive Care in Cancer
23 (2015): 3201–3209, 10.1007/s00520-015-2706-9.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
9.
Han L., Sullivan R., Tree A., et al., “The Impact of Transportation Mode, Socioeconomic Deprivation and Rurality on Travel Times to Radiotherapy and Surgical Services for Patients With Prostate Cancer: A National Population‐Based Evaluation,” Radiotherapy and Oncology
192 (2024): 110092, 10.1016/j.radonc.2024.110092.
[DOI] [PubMed] [Google Scholar] -
10.
Sepassi A., Li M., A. Zell J., Chan A., Saunders I. M., and Mukamel D. B., “Rural‐Urban Disparities in Colorectal Cancer Screening, Diagnosis, Treatment, and Survivorship Care: A Systematic Review and Meta‐Analysis,” Oncologist
29 (2024): e431–e446, 10.1093/oncolo/oyad347.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
11.
Raman V., Adam M. A., Turner M. C., Moore H. G., Mantyh C. R., and Migaly J., “Disparity of Colon Cancer Outcomes in Rural America: Making the Case to Travel the Extra Mile,” Journal of Gastrointestinal Surgery
23 (2019): 2285–2293, 10.1007/s11605-019-04270-5.
[DOI] [PubMed] [Google Scholar] -
12.
Fletcher K. M., Revette A., Enzinger A., et al., “Experience and Needs of Patients With Young‐Onset Colorectal Cancer and Their Caregivers: A Qualitative Study,” JCO Oncology Practice
20 (2024): 1604–1611, 10.1200/OP.24.00002.
[DOI] [PubMed] [Google Scholar] -
13.
National Cancer Database
. In: ACS, https://www.facs.org/quality-programs/cancer-programs/national-cancer-database/. Accessed, 10 Sept 2025. -
14.
US Census Bureau American Community Survey (ACS)
. In: Census.gov, https://www.census.gov/programs-surveys/acs.html. Accessed, 16 Feb 2026. -
15.
USDA Economic Research Service. 2020. Rural‐Urban Continuum Codes
. In: US Dep. Agric, https://www.ers.usda.gov/data-products/rural-urban-continuum-codes.aspx. Accessed, 20 May 2022. -
16.
Zahnd W. E., Gomez S. L., Steck S. E., et al., “Rural‐Urban and Racial/Ethnic Trends and Disparities in Early‐Onset and Average‐Onset Colorectal Cancer,” Cancer
127 (2021): 239–248, 10.1002/cncr.33256.
[DOI] [PubMed] [Google Scholar] -
17.
CoC Accreditation
. In: ACS, https://www.facs.org/quality-programs/cancer-programs/commission-on-cancer/coc-accreditation/. Accessed, 20 Jan 2026. -
18.
Labban M., Chen C.‐R., Frego N., et al., “Disparities in Travel‐Related Barriers to Accessing Health Care From the 2017 National Household Travel Survey,” JAMA Network Open
6 (2023): e2325291, 10.1001/jamanetworkopen.2023.25291.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
19.
Stimpson J. P., Billig J., Gurley T., and Liao J. M., “The Impact of Employment Status, Income, and Occupation on the Association Between Workplace Benefits and Health‐Related Work Absences,” American Journal of Industrial Medicine
68 (2025): 598–606, 10.1002/ajim.23733.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
20.
Lueckmann S. L., Hoebel J., Roick J., et al., “Socioeconomic Inequalities in Primary‐Care and Specialist Physician Visits: A Systematic Review,” International Journal for Equity in Health
20 (2021): 58, 10.1186/s12939-020-01375-1.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
21.
Tsai M.‐H., Coughlin S. S., Cortes J., and Thompson C. A., “Intersection of Poverty and Rurality for Early‐Onset Colorectal Cancer Survival,” JAMA Network Open
7 (2024): e2430615, 10.1001/jamanetworkopen.2024.30615.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
22.
Stitzenberg K. B., Sigurdson E. R., Egleston B. L., Starkey R. B., and Meropol N. J., “Centralization of Cancer Surgery: Implications for Patient Access to Optimal Care,” Journal of Clinical Oncology
27 (2009): 4671–4678, 10.1200/JCO.2008.20.1715.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
23.
Fong Z. V., Loehrer A. P., Fernández‐Del castillo C., et al., “Potential Impact of a Volume Pledge on Spatial Access: A Population‐Level Analysis of Patients Undergoing Pancreatectomy,” Surgery
162 (2017): 203–210, 10.1016/j.surg.2017.03.010.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
24.
Diaz A. and Pawlik T. M., “Geographic Disparities in Oncologic Treatment and Outcomes: The Urban–Rural Divide,” Annals of Surgical Oncology
28 (2021): 8011–8013, 10.1245/s10434-021-10653-3.
[DOI] [PubMed] [Google Scholar] -
25.
Salem M. E., Puccini A., Trufan S. J., et al., “Impact of Sociodemographic Disparities and Insurance Status on Survival of Patients With Early‐Onset Colorectal Cancer,” Oncologist
26 (2021): e1730–e1741, 10.1002/onco.13908.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
26.
Aloysius M. M., Goyal H., Shah N. J., et al., “Impact of Race and Socioeconomics Disparities on Survival in Young‐Onset Colorectal Adenocarcinoma—A SEER Registry Analysis,” Cancers
13 (2021): 3262, 10.3390/cancers13133262.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
27.
Mikati I., Benson A. F., Luben T. J., Sacks J. D., and Richmond‐Bryant J., “Disparities in Distribution of Particulate Matter Emission Sources by Race and Poverty Status,” American Journal of Public Health
108 (2018): 480–485, 10.2105/AJPH.2017.304297.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
28.
Weitekamp C. A., Lein M., Strum M., et al., “An Examination of National Cancer Risk Based on Monitored Hazardous Air Pollutants,” Environmental Health Perspectives
129 (2021): 037008, 10.1289/EHP8044.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
29.
Larsen K., Rydz E., and Peters C. E., “Inequalities in Environmental Cancer Risk and Carcinogen Exposures: A Scoping Review,” International Journal of Environmental Research and Public Health
20 (2023): 5718, 10.3390/ijerph20095718.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
30.
Madrigal J. M., Flory A., Fisher J. A., et al., “Sociodemographic Inequities in the Burden of Carcinogenic Industrial Air Emissions in the United States,” JNCI: Journal of the National Cancer Institute
116 (2024): 737–744, 10.1093/jnci/djae001.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
31.
Villanueva C. M., Espinosa A., Gracia‐Lavedan E., et al., “Exposure to Widespread Drinking Water Chemicals, Blood Inflammation Markers, and Colorectal Cancer,” Environment International
157 (2021): 106873, 10.1016/j.envint.2021.106873.
[DOI] [PubMed] [Google Scholar] -
32.
Wen J. and Lin Y., “Invisible Invaders: Unveiling the Carcinogenic Threat of Microplastics and Nanoplastics in Colorectal Cancer—A Systematic Review,” Frontiers in Public Health
13 (2025): 1653245, 10.3389/fpubh.2025.1653245.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
33.
Su L. J., Young S. G., Collins J., Matich E., Hsu P. C., and Chiang T. C., “Geospatial Assessment of Pesticide Concentration in Ambient Air and Colorectal Cancer Incidence in Arkansas, 2013–2017,” International Journal of Environmental Research and Public Health
19, no. 6 (2022): 3258, 10.3390/ijerph19063258.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
34.
Heydari Z., Sarkar G., Helgeson L., et al., “Colon Organoids as Experimental Models to Study the Effect of Micro‐Nanoparticles as a Driver of Early‐Onset Colon Cancer,” Cells
15 (2025): 40, 10.3390/cells15010040.
[DOI] [PMC free article] [PubMed] [Google Scholar] - 35.Affordable Care Act Implementation FAQs—Set 12 | CMS, https://www.cms.gov/cciio/resources/fact-sheets-and-faqs/aca_implementation_faqs12. Accessed, 13 Dec 2024.
-
36.
Davidson K. W., Barry M. J., Mangione C. M., et al., “Screening for Colorectal Cancer: US Preventive Services Task Force Recommendation Statement,” Journal of the American Medical Association
325 (2021): 1965–1977, 10.1001/jama.2021.6238.
[DOI] [PubMed] [Google Scholar] -
37.
Buchheit J. T., Silver C. M., Huang R., et al., “Association Between Racial and Socioeconomic Disparities and Hospital Performance in Treatment and Outcomes for Patients With Colon Cancer,” Annals of Surgical Oncology
31 (2024): 1075–1086, 10.1245/s10434-023-14607-9.
[DOI] [PubMed] [Google Scholar] -
38.
Allar B. G., Abraham L., Eruchalu C. N., et al., “Interaction of Insurance and Neighborhood Income on Operative Colorectal Cancer Outcomes Within a National Database,” Journal of Surgical Research
303 (2024): 95–104, 10.1016/j.jss.2024.08.015.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
39.
Siegel R. L., Miller K. D., Fedewa S. A., et al., “Colorectal Cancer Statistics, 2017,” CA: A Cancer Journal for Clinicians
67 (2017): 177–193, 10.3322/caac.21395.
[DOI] [PubMed] [Google Scholar] -
40.
Tamlander M., Jermy B., Seppälä T. T., et al., “Genome‐Wide Polygenic Risk Scores for Colorectal Cancer Have Implications for Risk‐Based Screening,” British Journal of Cancer
130 (2024): 651–659, 10.1038/s41416-023-02536-z.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
41.
Kastrinos F., Kupfer S. S., and Gupta S., “Colorectal Cancer Risk Assessment and Precision Approaches to Screening: Brave New World or Worlds Apart?,” Gastroenterology
164 (2023): 812–827, 10.1053/j.gastro.2023.02.021.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
42.
Gu J., Li Y., Yu J., et al., “A Risk Scoring System to Predict the Individual Incidence of Early‐Onset Colorectal Cancer,” BMC Cancer
22 (2022): 122, 10.1186/s12885-022-09238-4.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
43.
Archambault A. N., Jeon J., Lin Y., et al., “Risk Stratification for Early‐Onset Colorectal Cancer Using a Combination of Genetic and Environmental Risk Scores: An International Multi‐Center Study,” Journal of the National Cancer Institute
114 (2022): 528–539, 10.1093/jnci/djac003.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
44.
Willauer A. N., Liu Y., Pereira A. A. L., et al., “Clinical and Molecular Characterization of Early‐Onset Colorectal Cancer,” Cancer
125 (2019): 2002–2010, 10.1002/cncr.31994.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
45.
Lieu C. H., Golemis E. A., Serebriiskii I. G., et al., “Comprehensive Genomic Landscapes in Early and Later Onset Colorectal Cancer,” Clinical Cancer Research
25 (2019): 5852–5858, 10.1158/1078-0432.CCR-19-0899.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
46.
Kirzin S., Marisa L., Guimbaud R., et al., “Sporadic Early‐Onset Colorectal Cancer Is a Specific Sub‐Type of Cancer: A Morphological, Molecular and Genetics Study,” PLoS One
9 (2014): e103159, 10.1371/journal.pone.0103159.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
47.
Kolarich A., George T. J., Hughes S. J., et al., “Rectal Cancer Patients Younger Than 50 Years Lack a Survival Benefit from NCCN Guideline–Directed Treatment for Stage II and III Disease,” Cancer
124 (2018): 3510–3519, 10.1002/cncr.31527.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
48.
Kneuertz P. J., Chang G. J., Hu C.‐Y., et al., “Overtreatment of Young Adults With Colon Cancer: More Intense Treatments With Unmatched Survival Gains,” JAMA Surgery
150 (2015): 402–409, 10.1001/jamasurg.2014.3572.
[DOI] [PubMed] [Google Scholar] -
49.
Myers S., Davis E. S., Murillo A., et al., “Surgical Approach as a Mediator of Rural–Urban Disparity After Colon Cancer Resection,” Journal of the American College of Surgeons
242 (2026): 432–443, 10.1097/XCS.0000000000001601.
[DOI] [PubMed] [Google Scholar] -
50.
Kong M., Chen H., Shan K., Sheng H., and Li L., “Comparison of Survival Among Adults With Rectal Cancer Who Have Undergone Laparoscopic vs Open Surgery: A Meta‐Analysis,” JAMA Network Open
5 (2022): e2210861, 10.1001/jamanetworkopen.2022.10861.
[DOI] [PMC free article] [PubMed] [Google Scholar] -
51.
Jadid K. D., Cao Y., Petersson J., Sjövall A., Angenete E., and Matthiessen P., “Long‐Term Oncological Outcomes for Minimally Invasive Surgery Versus Open Surgery for Colon Cancer—A Population‐Based Nationwide Study With a Non‐Inferiority Design,” Colorectal Disease
25 (2023): 954–963, 10.1111/codi.16512.
[DOI] [PubMed] [Google Scholar] -
52.
Boscoe F. P., Henry K. A., and Zdeb M. S., “A Nationwide Comparison of Driving Distance Versus Straight‐Line Distance to Hospitals,” Professional Geographer
64, no. 2 (2011): 188–196, 10.1080/00330124.2011.583586. [DOI] [PMC free article] [PubMed] [Google Scholar] -
53.
Bliss R. L., Katz J. N., Wright E. A., and Losina E., “Estimating Proximity to Care: Are Straight Line and Zipcode Centroid Distances Acceptable Proxy Measures?,” Medical Care
50 (2012): 99–106, 10.1097/MLR.0b013e31822944d1.
[DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Figure S1: Sensitivity analysis including facility type (reference: Integrated Cancer Network) and region (reference: Northeast); data for facility type and region are only available for individuals 40 years and older at the time of diagnosis. Hazard ratios and 95% confidence intervals for 5‐year survival in early‐onset colorectal cancer (EOCRC, blue diamonds) and average‐onset colorectal cancer (orange squares) are represented.
Data Availability Statement
The data are available through the National Cancer Database.
Articles from Journal of Surgical Oncology are provided here courtesy of Wiley
