Socioeconomic status on survival outcomes in patients with colorectal cancer: a cross-sectional study

Colorectal cancer (CRC) is widely acknowledged as a prevalent malignancy and the second most common cause of cancer-related mortality worldwide. The aim of this study was to examine the independent impact of Median Household Income (MHI) on prognosis and survival outcomes in patients with CRC. Data from 17 cancer registries of the United States Surveillance, Epidemiology, and End Results program, with follow-up extended until November 2022 was analyzed. A Cox proportional hazards regression analysis was conducted to evaluate the influence of different levels of MHI on survival outcomes among patients with CRC. A total of 761,697 CRC patient records were retrieved from the SEER database. The Cox regression analysis results indicated that patients with higher MHI exhibited improved overall survival outcomes when compared to those with lower MHI (MMHI: P < 0.001; HMHI: P < 0.001). Regardless of the specific tumor location, gender, stage of CRC, or treatment method, higher MHI is consistently linked to improved survival outcomes. However, this association was not found to be statistically significant among American Indian/Alaska Native (MMHI: P = 0.017; HMHI: P = 0.081), Asian or Pacific Islander (MMHI: P = 0.223; HMHI: P = 0.002) and unmarried or domestic partner patients (MMHI: P = 0.311; HMHI: P = 0.011). These results emphasize the importance of considering socioeconomic factors, such as income level, in understanding and addressing disparities in survival outcomes of CRC patients.


Introduction
Colorectal cancer (CRC) is widely acknowledged as a highly prevalent malignancy and the second most significant contributor to cancer-related mortality worldwide (Dekker et al. 2019;Sung et al. 2021).Several modifiable risk factors for CRC have been identified, including unhealthy diets and sedentary lifestyles (Carethers 2015).Other factors such as excessive alcohol consumption, and obesity also contribute to the risk of developing CRC (Islami et al. 2018).Furthermore, implementing regular screening, surveillance, and providing high-quality treatment can effectively prevent a significant proportion of CRC cases and reduce the number of CRC-related deaths (Winawer and Zauber 2002).
Numerous studies have consistently demonstrated a strong correlation between socioeconomic status (SES) and survival outcomes (Collaborators 2021;Jansen et al. 2014;Lortet-Tieulent et al. 2020;Shah and Chan 2021;Syriopoulou et al. 2019).For example, a study that encompassed ten population-based registries, representing a population of 32 million individuals, and focused on the 25 most prevalent cancer sites, including CRC demonstrated a significantly lower relative survival rate among patients residing in socioeconomically disadvantaged regions compared to those living in more affluent areas across Germany (Jansen et al. 2014).In addition, a multicenter, international prospective cohort study involving adult patients undergoing surgery has demonstrated that the patients with CRC in low-and middle-income countries (LMICs) experience higher mortality rates following cancer surgery in comparison to patients in high-income countries (Collaborators 2021).
The impact of SES on the utilization and selection of medical treatments remains somewhat uncertain (Carethers 2015(Carethers , 2018;;Carethers and Doubeni 2020).Several studies have indicated that individuals with low SES tend to receive lower rates of surgical interventions, chemotherapy, and radiation therapy (Booth et al. 2016;Harris et al. 2009;Lemmens et al. 2005;McGory et al. 2006).For instance, a study provide evidence that low SES is a predictive factor for inadequate administration of adjuvant chemotherapy and radiation among patients with stage III colon cancer and stage II-III rectal cancer in the United States (McGory et al. 2006).Conversely, other research findings fail to establish a significant correlation in this regard (Campbell et al. 2002;Schrag et al. 2001).
The Median Household Income (MHI) of CRC patients was defined as the estimate of the median household income of the county of residence based on static county attributes (SCAs), which is commonly used as an indicator of the overall economic well-being of a particular geographic area or population.It is often used in socioeconomic research, policy-making, and market analysis to assess the financial status of households and to compare income disparities among different groups or regions.Many studies have reported on the relationship between Median Household Income (MHI) and the prognosis of cancer patients (Bristow et al. 2013;Geng et al. 2023;O'Connor et al. 2018).The findings from these studies suggest that higher MHI is associated with better prognosis for early-stage esophageal adenocarcinoma, ovarian, or other cancer patients.Patients with higher income levels tend to have improved access to healthcare resources, including early detection, timely treatment, better overall quality of care and improved long-term survival rates.
There is currently limited research investigating the specific impact of MHI on outcomes in CRC patients.While studies have explored the association between MHI and cancer outcomes in general, there is a lack of focused research specifically examining the relationship between MHI and CRC outcomes.Thus, we aimed to investigate whether MHI independently influence prognosis and overall survival in patients with CRC.Understanding these associations can inform healthcare policies and interventions aimed at reducing disparities and improving outcomes for individuals across different income levels.

Data source and study population
The study included individuals who were diagnosed with CRC within a 20-year time frame from 2000 to 2020.The collected data was then submitted to the Surveillance, Epidemiology, and End Results (SEER) Program in November 2022, which is an ongoing initiative of the National Cancer Institute.It aims to gather comprehensive data on cancer incidence and survival outcomes, including characteristics of individual patient and tumor, treatment information, and follow-up survival data from cancer registries across the United States.SEER-17 comprises 17 population-based cancer registries, collectively accounting for around 26.5% of the population based on the 2020 census (https:// seer.cancer.gov/).The inclusion criteria for patients encompassed the utilization of the International Classification for Oncology, third edition (ICD-O-3), which facilitated the classification of individuals diagnosed with colon cancer, rectal cancer, or cancer of the rectosigmoid junction (World Health 2013).

Variables of interest
Socioeconomic status is widely recognized as significant causal factors that profoundly impact individuals' lives (Antonoplis 2023).We chose the Median Household Income (MHI) to reflect the socioeconomic status to a certain extent.The MHI were taken from the 2017-2021 American Community Survey (ACS) data, which represents the midpoint of the income distribution among households in a given geographic area covered by the SEER program.The MHI serves as an indicator of the overall economic well-being of a population within a specific SEER region.It provides insight into the income distribution and socioeconomic status of households within that area.The difference in MHI can reflect the gap between different regions of economic growth, income inequality and population financial health.The SEER database categorizes the MHI into four levels: $0-$34,999, $35,000-$54,999, $55,000-$74,999, and more than $75,000.Due to the relatively small population within the first two levels, we consolidated them into a single category of $0-$54,999.Finally, we define it as low MHI (LMHI, $0-$54,999), medium MHI (MMHI, $55,000-$74,999) and high MHI (HMHI, more than $75,000).
Additional demographic and tumor data including sex, age, race, year of diagnosis, tumor site, stage of cancer, tumor grade (Grade I-IV, unknown), radiation therapy (yes, no/unknown), chemotherapy (yes, no/unknown), survival months, vital status, and cause of death were also extracted.

Endpoints
In our study, the primary endpoint focused on assessing survival probability, while the secondary endpoint aimed to evaluate the probability of cancer-specific death in CRC patients.Cancer-specific death, as defined by the SEER classification, referred to deaths attributed specifically to the primary cancer, while deaths caused by other factors were censored.We considered the time from cancer diagnosis until either mortality or loss to follow-up as the measurement of survival time.To capture longterm outcomes, the evaluation of survival time spanned a period of up to 20 years.By using this duration, we were able to evaluate the extended impact of CRC on patient survival and better understand the long-term prognosis associated with the disease.By examining both survival probability and the probability of cancer-specific death, our study aimed to provide a comprehensive understanding of the outcomes and mortality patterns in patients with CRC over an extended follow-up period.

Statistical analysis
The RStudio software (R version 4.2.3,Boston, Massachusetts) was used to clean the data obtained from the SEER database for removing some missing data.The statistical analysis involved the comparison of variables between groups using appropriate tests.For continuous variables that follow a normal distribution, a t-test was conducted.For continuous variables that follow a nonnormal distribution, the Mann-Whitney U test was used.Categorical variables were analyzed using the chi-squared test.All statistical tests conducted in this study were twotailed, adhering to the principle that a P-value of less than 0.05 signifies statistical significance.The Cox proportional hazards regression analysis was conducted using the "survival" package in R.This analysis allowed for the estimation of hazard ratios (HRs) associated with variables related to the cause of death.Primarily, we calculated the survival probability of patients diagnosed with CRC.Then, the probability of cancer-specific death and non-cancer-related death were calculated.In addition, we conducted subgroup analysis according to different population characteristics, including cancer cite, treatment, sex, race, and marital status.Crude HRs were assigned to the univariate regression analysis, while adjusted HRs were assigned to the multivariate regression analysis.The "survminer" package in R was employed to create Kaplan-Meier plots for visualizing and analyzing survival data, and the log-rank (Mantel-Cox) test was utilized to evaluate the clinical significance between survival curves.

Study population
A total of 761,697 CRC patient records were retrieved from the SEER database.After excluding individuals with missing crucial information, 506,042 patients remained in the study.Patients were categorized into three groups according to the MHI of their county of residence.The LMHI group included 82,231 patients.The MMHI group consisted of 214,414 patients.The HMHI group comprised of 209,397 patients.
Table 1 summarized demographic information and tumor characteristics that stratified by MHI.A total of 321,223 patients with colon cancer (63.5%), 47,911 patients with rectosigmoid cancer (9.5%), and 136,908 patients with rectal cancer (27.1%) were included.Males was found to have a higher risk of developing CRC than females, with a higher incidence rate in males across all age groups.Compared to patients with an MMHI or HMHI, the proportion of younger patients with LMHI was higher (43.9% or 44.0%vs 45.0%).Then, patients with higher MHI had lower singles, lower all-cause mortality, and less survival months.

Overall
Figure 1a, b presents a stratified survival distribution plot (Kaplan-Meier) illustrating the overall survival outcomes as well as cancer-specific mortality and alive or mortality from other causes.The plot demonstrates significantly worse survival outcome and higher mortality in the lowest MHI group (log-rank test, p < 0.001).
Table 2 presents the results of our Cox regression analyses, including both univariate and multivariate models, illustrating the associations between several factors such as gender, age, race, and MHI, among others, and the survival outcomes of CRC patients.

Stratified by marital status
Furthermore, we conducted an analysis to investigate the influence of marital status on survival outcomes across different income groups.Our findings revealed that higher MHI

Discussion
In this observational study conducted at the population level, we conducted the Cox regression analysis, which indicated that patients with higher MHI demonstrated better overall survival outcomes compared to those with lower MHI.Regardless of the specific tumor location, gender, or treatment method, higher MHI is consistently linked to improved survival outcomes.However, it is important to note that among certain subgroups, such as Asian or Pacific Islander, American Indian/Alaska Native, and unmarried or domestic partner patients, this association did not reach statistical significance.Socioeconomic measures are closely linked to social determinants of health, which encompass the circumstances individuals encounter throughout their lives and are influenced by the distribution of financial resources, power, and access to various resources (Marmot et al. 2008).Lower socioeconomic status has been consistently linked to reduced access to high-quality treatment for colorectal cancer (CRC) patients (Carethers and Doubeni 2020).A state-based study conducted in 2014 revealed that individuals diagnosed with CRC and belonging to lower socioeconomic status experienced reduced chances of receiving surgical intervention and/or chemotherapy.Consequently, these patients encountered a higher adjusted risk of mortality subsequent to their CRC diagnosis (Thornblade et al. 2020), which aligns with the findings obtained in our study, where we also identified a significant correlation between higher MHI and improved survival outcomes among CRC patients.
HMHI is associated with a range of factors that contribute to improved survival outcomes in CRC patients.These factors include better access to education, higher medical standards, increased availability of medical resources, adoption of healthier lifestyles, and reduced exposure to environmental risk factors.These factors can contribute to a more favorable prognosis and better overall health outcomes for cancer patients.It is vital to highlight that individuals with LMHI are disproportionately affected by lack of insurance coverage, underinsurance, or dependence on Medicaid.This significantly impacts their access to a range of crucial healthcare services, including treatment options, medical providers, hospitals, and screening services.As a result, patients with LMHI may experience barriers to timely and appropriate care, potentially leading to inferior survival outcomes (Daniel et al. 2017;Schlottmann et al. 2020).
Income inequality is intricately associated with race and caste, as individuals belonging to White racial groups are typically overrepresented in the highest income brackets, while individuals from Black racial groups are frequently concentrated in the lowest income brackets (Akee et al. 2019).Research has consistently shown that there is a significant disparity in survival rates between Black and White patients, with Black patients experiencing worse outcomes (Coleman et al. 2008;White et al. 2017).Several factors contribute to this racial disparity in survival rates.One contributing factor to the observed income inequality is that Black patients often present at more advanced Statistically significant differences are in bold stages of the disease compared to White patients.This delayed diagnosis can impact treatment options and overall prognosis.Furthermore, disparities in access to healthcare play a crucial role.Black patients are at a lower likelihood of having health insurance coverage, which can impede their access to essential medical services and treatments (Ward et al. 2008).Additionally, access to clinical care is influenced by various socioeconomic factors, including educational attainment, employment status, income level, and proximity to healthcare providers (Mahmoudi and Jensen 2012).In addition, the combination of increasing income inequality and limited upward mobility for Black individuals has resulted in significant racial disparities over time, which may also have an impact on

Time (months) Strata
Rectum health outcomes.These racial caste differences can have profound implications for health.Studies have shown that individuals from marginalized racial and ethnic groups, such as Black communities, often experience higher rates of chronic illnesses, shorter life expectancies, and poorer overall health compared to their White counterparts (Braveman et al. 2010;Williams and Mohammed 2009).The unequal distribution of wealth and resources, along with systemic racism and discrimination, contribute to these health disparities.
We observe that the association between higher income level and better survival outcomes is not significant among American Indian/Alaska Native and Asian or Pacific Islander patients.According to the American Cancer Society, Asian Americans/Pacific Islanders were most likely to receive an early diagnosis of colorectal cancer when the disease is still localized and treatments are often successful.In contrast, non-Hispanic blacks and American Indians/Native Alaskans were more likely to receive a late-stage diagnosis when the cancer has already spread (ACS 2020).Meanwhile, the stage of cancer at diagnosis has a significant effect on patient survival outcomes.In addition, these population groups face structural barriers to access to health resources and services, such as cultural barriers, language barriers, and low insurance coverage (Lu et al. 2016;McCracken et al. 2007).These factors might diminish or mask the effect of income levels on outcomes.
In regard to unmarried or domestic partner patients, our study did not find a significant difference between the effects of LMHI versus MMHI on survival outcomes among CRC   patients.This lack of statistical significance could be attributed to the relatively small and imbalanced sample sizes in the low income (n = 116) and middle income (n = 448) groups.Given the limitations in sample size, conclusions about income effects should be drawn cautiously from these data.
In addition to its direct impact on CRC patients, LMHI may also contribute to negative survival outcomes in CRC patients through its effect on caregivers as a collective.Cancer caregivers are often required to stay home and care for CRC patients.However, for many cancer caregivers who are employed, taking time off to care for CRC patients can result in a loss of income or even jeopardize their job security (Bradley et al. 2023;Schuster and Chung 2014).According to a retrospective study, approximately one-third of cancer caregivers reported discontinuing their employment (35%)

(b)
and experiencing an increase in household debt (30%).This highlights the significant financial burden faced by caregivers in the context of cancer care (Altice et al. 2017;Ekwueme et al. 2019;Yabroff et al. 2016).This challenge is particularly pronounced for LMHI families, as nearly half of caregivers (46.3%) reported discontinuing their employment (Bradley et al. 2023).In addition, they may already have health issues that hinder their ability to provide the necessary attention required for patients with CRC.Moreover, individuals from LMHI families are more prone to experiencing elevated levels of stress.However, they often face challenges in finding the necessary time and resources to engage in healthy stress relief activities.which can further impact the quality of care they can provide for CRC patients (Pampel et al. 2010).Consistent evidence from multiple studies has consistently linked insurance status to a significant risk factor for the diagnosis of advanced CRC.Being uninsured has specifically been recognized as a crucial risk factor for both overall survival and advanced stage of CRC (Brand et al. 2022;Halpern et al. 2008;Salem et al. 2021).Due to the unavailability of insurance-related data for inclusion in our analysis, we conducted a comprehensive analysis, incorporating single-factor, multi-factor, and subgroup analyses of CRC staging.This approach allows us to elucidate the influence of cancer staging on the survival outcome of CRC, thereby partially substituting the inclusion of insurance status as a variable.

(c) (d)
However, it crucial to acknowledge that the correlation between MHI and cancer prognosis is multifaceted and can be influenced by various factors, including the type and stage of cancer, access to healthcare services, and individual patient characteristics.Additionally, socioeconomic disparities and healthcare inequalities may exist, leading to differential outcomes among patients from different income groups.
This study has several significant limitations that warrant acknowledgement.Firstly, the retrospective nature of the data introduces inherent biases and limitations in terms of data collection and analysis.Additionally, the reliance on registry data may be prone to coding errors and inconsistencies, which could impact the accuracy and reliability of the findings.Furthermore, the study's sample may not be uniformly representative of the entire United States population, potentially limiting the generalizability of the results.
What is more, the insurance coverage plays a crucial role in cancer therapy.However, the insurance-related data was unavailable for inclusion in the analysis, which may cause some bias.Future research incorporating longitudinal data and individual-level analysis could provide a more comprehensive understanding of the temporal dynamics and changes in these populations.These limitations should be duly considered when interpreting the findings and formulating conclusions.

Conclusion
This study uncovered a significant association between lower MHI and poorer survival outcomes in patients with CRC.Notably, our findings demonstrated an increased risk of both cancer-specific mortality and alive or mortality from other causes among individuals with lower MHI.These results emphasize the importance of considering socioeconomic factors, such as income level, in understanding and addressing disparities in survival outcomes of CRC patients.

Fig. 1 a
Fig. 1 a Kaplan-Meier (KM) diagrams for survival outcomes; b Kaplan-Meier (KM) diagrams for cancer-related mortality and alive or death from other causes

Fig. 2
Fig. 2 Kaplan-Meier (KM) diagrams for survival outcomes stratified by tumor location: a colon cancer; b rectum cancer; c rectosigmoid cancer

Fig. 3
Fig.3Forest plot of presenting hazard ratios (HR) of survival outcomes stratified by tumor location

Fig. 4
Fig. 4 Kaplan-Meier (KM) diagrams for survival outcomes stratified by tumor stage: a stage I; b stage II; c stage III; d stage IV

Fig. 5
Fig. 5 Forest plot of presenting hazard ratios (HR) of survival outcomes stratified by tumor stage

Table 2
Cox proportional hazards univariate and multivariate analysis: survival outcomes