Socioeconomic Disparities in the Progression of Gastroesophageal Reflux Disease and Barrett’s Esophagus to Esophageal Adenocarcinoma

Abstract

Objective:

To assess the impact of socioeconomic and other factors on progression from gastroesophageal reflux disease (GERD) and Barrett's esophagus (BE) to esophageal adenocarcinoma (EAC).

Background:

Socioeconomic status (SES) influences outcomes in EAC, but its role in the progression from GERD and BE to EAC is not well-defined. Risk of progression may be influenced by access and engagement with endoscopic screening and surveillance, health care utilization, health literacy, and other competing social determinants of health that vary with SES.

Methods:

Patients diagnosed with GERD between 2015 and 2023 across a regional network of 14 hospitals and >300 clinics were reviewed for BE and EAC. U.S. Census data on wealth, income, education and occupation were used to categorize patients into low, middle, and high SES groups. Univariate and multivariable analyses were conducted to evaluate the influence of demographic, clinical, and SES factors on progression to EAC from GERD and BE.

Results:

Among 162,542 patients with GERD, 7604 (4.70%) developed BE and 331 (0.20%) progressed to EAC. Unemployment, disability, lack of college education, and high school non-graduation were associated with progression to EAC (all P < 0.0001). Independent predictors of disease progression included established demographic and clinical factors (all P < 0.0001), lack of private insurance (odds ratio [OR]: 1.4, P < 0.0001), and low SES (OR: 1.6, P < 0.0001). For BE to EAC progression, predictors were similar, including lack of private insurance (OR: 2.0, P = 0.015) and low SES (OR: 2.0, P = 0.028). Patients with low SES were less frequently endoscopically evaluated (4/1000 vs 10/1000, P < 0.0001) or diagnosed with BE (4/1000 vs 50/1000, P < 0.0001). They were also more likely to be diagnosed with early-onset EAC (13/1000 vs 5/1000, P = 0.0138) and metastatic EAC (77.4% vs 53.1%, P = 0.001). Among patients with BE, those with low SES had a higher rate of progression to EAC (1.77% vs 0.95%, P = 0.026).

Conclusions:

Low SES is associated with increased progression to EAC, with higher rates of early-onset and metastatic disease at diagnosis. Disparities in access to endoscopic screening and surveillance play a significant role, compounded by socioeconomic barriers such as limited health care engagement and competing social priorities. Integrating SES into risk stratification models alongside clinical factors is essential for improving early detection and reducing the burden of advanced disease.

Socioeconomic status (SES) is a significant and well-established determinant of cancer outcomes across multiple malignancies. Low SES contributes to increased incidence, worse survival rates, and inferior overall outcomes.^1–4 High SES has been associated with fewer health care delays, a lower rate of late-stage disease at diagnosis, and a higher likelihood of receiving chemotherapy following surgical resection.⁵ Similarly, in esophageal adenocarcinoma (EAC), socioeconomic and demographic determinants of health have been shown to significantly contribute to disparities in survival rates and access to chemotherapy, surgery, and radiotherapy.^6,7 While numerous studies focus on the repercussions of SES in terms of EAC outcomes, morbidity, and mortality rates, there are limited data on its role in the progression from gastroesophageal reflux disease (GERD) and Barrett esophagus (BE) to EAC.⁸

It is currently understood that some patients with GERD progress through a metaplasia-dysplasia-EAC sequence. The goal of screening and surveillance protocols is to diagnose patients at a point early in this sequence when a disease-modifying intervention can be introduced. Societies recommend surveillance every 3 to 5 years for patients diagnosed with metaplasia and every 6 to 12 months for low-grade dysplasia. High-grade dysplasia or intramucosal adenocarcinoma warrants endoscopic therapy.^9–12 In addition to chronic GERD, several other risk factors for EAC have been previously established to identify patients who need to be screened, including: advanced age, male sex, Caucasian ethnicity, tobacco use, and obesity. Many of these risk factors intersect with socioeconomic factors, such as wealth disparity, education level, occupation, and access to high-quality health care services. Disparity in SES may result in disparity in screenings and surveillance practices crucial for early detection and intervention of EAC. These barriers in health care can precipitate the detrimental advancement of cancers, consequently increasing morbidity and mortality rates.

The progression from GERD to BE and ultimately to EAC mirrors the stepwise progression observed in colorectal cancer, from noncancerous adenomatous polyps to dysplasia and then to malignancy, a progression that has been linked to low SES.^13,14 Given this similarity, it is plausible that SES may also influence the progression from GERD to EAC. In addition to disparities in access and engagement with screening programs, SES disparities in health behaviors—smoking, diet, medication adherence—health care literacy and trust, and other competing social determinants may impact disease progression. However, there is a paucity of data on whether SES has an impact on progression from GERD and BE to EAC. This study aimed to address this gap by examining demographic, clinical, and SES factors for impact on the progression of GERD and BE to EAC within a large regional health network.

METHODS

Study Population and Data Collection

This was a retrospective cohort study including patients diagnosed with GERD between 2015 and 2023 at the Allegheny Health Network hospitals. Our multicenter health network is comprised of 14 hospitals and more than 300 clinics, providing health care services to the population across Western Pennsylvania, East Ohio, North West Virginia, and Southwest New York. Patients were identified from electronic medical records (Epic Systems) by using the International Classification of Disease, 10th version coding system. Patients with a GERD diagnosis were identified by using International Classification of Disease, 10th version-clinical modification K21. These patients were, in addition, queried for the subsequent diagnosis of BE (K22.7) and esophageal carcinoma (C15) in the study time period. Patients who had been diagnosed with esophageal squamous carcinoma, age <18 and age >89 were excluded from the study. This study was evaluated and approved by our Institutional Review Board (IRB 2024-012-WPH).

Patient Characteristics

Patient demographic and clinical characteristics, including age, sex, body mass index, smoking history, race, marital status, employment status, insurance type, number of esophagogastroduodenoscopies (EGD), diagnosis of BE and EAC and stage of the EAC at the time of the diagnosis were collected during the study period. The patient’s race was categorized as Caucasian, African American, Asian, or other. Patients who were active or former smokers were considered to have a smoking history. Insurance status was categorized as private or government (Medicaid or Medicare). Information on occupation was categorized as employed, unemployed/disabled, and retired. The EAC stage was determined based on the 8th edition of the American Joint Committee on Cancer staging manual at the time of the diagnosis.¹⁵ Early-onset EAC was defined as patients who were diagnosed at age <50 years old.

Socioeconomic Status

Patients were linked to their neighborhood of residence in relation to their Zone Improvement Plan (ZIP) code. A summary score was used as an indicator of the SES as previously described.^16–18 The 2020 U.S. Census Bureau data, including wealth and income, education, and occupation, were used to calculate the ZIP code-based SES summary score.

To construct the summary score, mean and SD were calculated for each ZIP code across 6 variables: (1) log of median household income, (2) log of median value of housing unit, (3) percentage of households with interest, dividends, or rental income, (4) percentage of adult residents completing high school, (5) percentage of residents with a college degree, and (6) percentage of residents with management, business, science, or arts occupations. Then, an individual Z score for each ZIP code was calculated by subtracting the mean from an individual measure and dividing by the SD; such that 6 individual Z-scores were calculated for each ZIP code. Finally, a summation of the 6 Z-scores then yielded the summary score. After completion of the summary Z score calculation, SES for all 825 ZIP codes within our health network was determined by applying quantile classification method on the Z score. The 2 lowest quintiles defined the low SES group, the middle quintile defined the middle SES group, and the 2 highest quintiles defined the high SES group.

Statistical Analysis

Baseline demographics, clinical characteristics, and SES levels were compared between patients with and without a subsequent diagnosis of EAC for all GERD subjects and some subgroups, respectively. In addition, SES and other significant/important factors in the univariate analyses were included in a multivariate analysis to assess the impact of SES on the progression from GERD to EAC for all patients with GERD and for patients with BE, respectively. The median number of EGD among the SES groups was calculated per thousand in the population based on 2020 U.S. Census Bureau data.

Among the Census population, the number of EGD per thousand, prevalence of GERD, and progression rates from GERD to BE and from GERD to EAC were compared between the SES groups, respectively. In addition, geographic maps were used to represent distribution of SES, number of EGDs per thousand, and prevalences of BE and EAC across the ZIP codes. Moreover, tumor stages at the time of EAC diagnosis were compared between the SES groups for the patients with GERD who developed EAC.

Values were expressed as either mean with SD or median with interquartile range for continuous variables and frequency and percentage for categorical variables. Comparison analyses among groups were performed using Pearson χ² test for categorical variables, and Mann-Whitney U test or Kruskal-Wallis test for continuous variables when appropriate. A multivariable logistic analysis was conducted to identify predictors of progression to EAC controlling covariates in the model for all patients and patients with BE, respectively. A P value <0.05 was set to be statistical significance. All statistical analyses were performed using the SAS software (version 9.4; SAS Institute).

RESULTS

The final study population consisted of 162,542 patients with GERD, of which 40.1% were males, with a mean (SD) age of 53.4 (17.1) and body mass index of 31.0 (6.9). A total of 7604 (4.70%) patients progressed to BE and 331 (0.20%) to EAC. Of those who progressed to EAC, BE was detected in 82 (24.8%). Demographic, clinical, and socioeconomic factors were compared between individuals who progressed to EAC and those who did not (Table 1). Patients who progressed to EAC were more likely to be male, older, Caucasian, have a smoking history, and have BE. From a socioeconomic standpoint, patients who progressed to EAC were more likely to have a low SES and be below the median household income. They were also more likely to drop out of high school, never receive a college degree, and not have private health insurance. Independent predictors of progression of GERD to EAC using multivariable analysis were found to be male sex [OR: 6.5 (95% CI: 4.8–8.6)], BE [OR: 4.9 (95% CI: 3.7–6.3)], smoking history [OR: 2.5 (95% CI: 2.0–3.3)], age >60 [OR: 1.9 (95% CI: 1.4–2.4)], Caucasian race [OR: 1.5 (95% CI: 1.0–2.5)], low SES [OR: 1.6 (95% CI: 1.2–2.1)], and lack of private insurance [OR: 1.4 (95% CI: 1.1–1.9; Table 2)].

TABLE 1.

Comparison of Demographic and Socioeconomic Factors Between EAC Progression Groups

Characteristics	No progression to EAC (N = 162,211)	Progression to EAC (N = 331)	P
Age, mean (SD)	53 (17)	62 (10)	<0.0001
Sex (M), n (%)	64,890 (40.0)	276 (83.4)	<0.0001
BMI, mean (SD)	31.0 (6.9)	28.0 (6.2)	<0.0001
Race/ethnicity, n (%)
Caucasian	145,956 (90.0)	313 (94.6)	0.0305
African American	10,863 (6.7)	15 (4.5)	—
Asian	1508 (0.9)	1 (0.3)	—
Other	3884 (2.4)	2 (0.6)	—
Marital status, n (%)
Married	88,489 (54.6)	214 (64.7)	0.0016
Single	37,272 (23.0)	49 (14.8)	—
Widowed	17,882 (11.0)	29 (8.8)	—
Divorced	15,152 (9.3)	32 (9.7)	—
Employment, n (%)
Employed	68,371 (43.8)	71 (21.6)	<0.0001
Unemployed/disabled	30,308 (19.4)	53 (16.2)	—
Retired	57,315 (36.7)	204 (62.2)	—
Insurance, n (%)
Medicare	69,005 (43.8)	207 (63.1)	<0.0001
Medicaid	9364 (5.9)	14 (4.3)	—
Private	79,281 (50.3)	107 (32.6)	—
Smoking history, n (%)	74,421 (45.9)	246 (74.4)	<0.0001
BE, n (%)	7522 (4.6)	82 (24.8)	<0.0001
Income ($), median (IQR)	64,151 (55690–77137)	60,633 (54271–72030)	0.0002
Education, mean (SD)
Percent completed high school	94.1 (3.6)	93.7 (3.1)	0.0034
Percent completed college	61.1 (12.8)	58.3 (12.2)	<0.0001
SES, n (%)
Low	18,120 (11.2)	53 (16.0)	0.0008
Middle	19,435 (12.0)	50 (15.1)	—
High	124,160 (76.8)	228 (68.9)	—

IQR indicates interquartile range.

TABLE 2.

Independent Predictors of Progression From GERD to EAC

Variable	Parameter (SE)	Odds ratio (95% CI)	P
Sex (M)	1.864 (0.149)	6.45 (4.8–8.6)	<0.0001
BE	1.588 (0.131)	4.89 (3.7–6.3)	<0.0001
Smoking history	0.931 (0.128)	2.53 (1.9–3.2)	<0.0001
Age >60	0.626 (0.131)	1.87 (1.4–2.4)	<0.0001
Race/ethnicity (Caucasian)	0.434 (0.244)	1.54 (0.9–2.4)	0.0755
Low SES	0.436 (0.153)	1.50 (1.1–2.0)	0.0044
Lack of private insurance	0.350 (0.137)	1.41 (1.0–1.8)	0.0104

Impact of Socioeconomic Status on Demographics, Endoscopic Evaluation, Progression to Barrett Esophagus and Esophageal Adenocarcinoma

Patients in the low SES group were more likely to be African American (11.4% vs 6.2%, P < 0.0001), be single or divorced (39.4% vs 31.0%, P < 0.0001), be unemployed or disabled (26.9% vs 17.1%, P < 0.0001), have a smoking history (52.6% vs 44.4%, P < 0.0001) and lack private insurance (36.3% vs 51.9%, P < 0.0001).

Patients with low SES were significantly less likely to progress from GERD to BE than middle and high SES patients (Fig. 1A); yet there was a stepwise decrease in the rate of progression to EAC as SES increased from low to middle to high (Fig. 1B). In addition, low SES patients with EAC were more likely to be diagnosed at late-stage (stage III or IV) disease than middle and high SES cancer patients (77.4% vs 53.1%, P = 0.001). The rate of early-onset EAC (diagnosis below the age of 50) was significantly higher in the low SES group compared with the rest of population [0.13% vs 0.05%, OR: 2.46 (95% CI: 1.2–5.0) P = 0.0138]. Patients with GERD in the low SES group had a lower number of endoscopic evaluations compared with the middle and high SES groups (4.3 vs 4.5 vs 10.1, P < 0.001). A similar trend was observed in patients with BE (8.4 vs 8.3 vs 14.1, P < 0.001; Fig. 2).

FIGURE 1.

A and B, There was a significant increasing trend in the progression of BE from low to high/highest SES (P < 0.001). However, the progression of EAC showed a significant decrease as SES increased from low to high/highest levels (P = 0.008).

FIGURE 2.

Comparison of median number of EGDs per thousand of population in SES groups for patients with GERD (A) and BE (B). High/highest SES group significantly higher median number of upper GI endoscopies than low and middle SES group (P < 0.001). GI indicates gastrointestinal.

Risk Factor of Progression of Barrett Esophagus to Esophageal Adenocarcinoma

A total of 7604 (4.70%) patients progressed from GERD to BE during the study period. Patients with BE were more likely to be older [57.0 (13) vs 53.2 (17), P < 0.0001], male (50.4% vs 39.4%, P < 0.0001), Caucasian (95.5% vs 89.7%, P < 0.0001), married (63.3% vs 54.1%, P < 0.0001), retired (42.2% vs 35.0%, P < 0.0001) and have smoking history (53.0% vs 45.5%, P < 0.0001) compared with those who did not progress to BE. Patients with high SES were more likely to be diagnosed with BE compared with those with low SES (4.8% vs 4.0%, P < 0.0001). In addition, patients with GERD who were diagnosed with BE had higher median income ($64,278 vs $64,151, P < 0.0001) and completed college (60.9% vs 56.6%, P = 0.0023) than those not diagnosed with BE. Patients in the high SES group underwent significantly more endoscopic evaluations compared with those in the low SES group (14.1 vs 8.4, P < 0.001).

Univariate analysis showed that patients who progressed from BE to EAC were more likely to be older, male, retired, have a smoking history, and less likely to have private insurance, compared with the rest of the population. In addition, patients with BE in the low SES group were more likely to progress to EAC than those in the high SES group (1.77% vs 0.95%, P = 0.026; Table 3). In multivariable analysis, male sex, advanced age, smoking history, lack of private insurance, and low SES were found to be the independent predictors of progression from BE to EAC (Table 4).

TABLE 3.

Comparison of Demographic and Socioeconomic Factors Between EAC Progression Groups in Patients With BE

Characteristics	No progression to EAC (N = 7522)	Progression to EAC (N = 82)	P
Age, mean (SD)	57 (13)	63 (10)	<0.0001
Sex (M), n (%)	3792 (50.4)	70 (85.4)	<0.0001
BMI, mean (SD)	30.6 (6.5)	27.5 (5.7)	<0.0001
Race/ethnicity, n (%)
Caucasian	7184 (95.5)	78 (95.1)	0.794
African American	180 (2.4)	3 (3.7)	—
Asian	38 (0.5)	0	—
Other	120 (1.6)	2 (1.2)	—
Marital status, n (%)
Married	4767 (63.4)	53 (64.6)	0.874
Single	1201(16.0)	15 (18.3)	—
Widowed	747 (9.9)	8 (9.8)	—
Divorced	701 (9.3)	5 (6.1)	—
Employment, n (%)
Employed	2895 (39.7)	13 (16.0)	<0.0001
Retired	3175 (43.6)	51 (63.0)	—
Unemployed/disabled	1212 (16.6)	17 (21.0)	—
Insurance, n (%)
Medicare	3578 (48.4)	57 (71.3)	0.0006
Medicaid	263 (3.6)	3 (3.8)	—
Private	3548 (48.0)	20 (25.0)	—
Smoking history, n (%)	3986 (53.0)	62 (75.6)	<0.0001
Income ($), median (IQR)	64,278 (56,250–79,208)	62,902 (55,000–72,030)	0.0595
Percentage completed college, mean (SD)	60.9 (13.3)	56.6 (13.2)	0.0023
SES, n (%)
Low	717 (9.6)	13 (15.9)	0.075
Middle	913 (12.2)	15 (18.3)	—
High	5874 (78.3)	54 (65.8)	—

IQR indicates interquartile range.

TABLE 4.

Independent Predictors of Progression From BE to EAC

Variable	Parameter (SE)	Odds ratio (95% CI)	P
Sex (M)	1.687 (0.315)	5.40 (2.9–10.0)	<0.0001
Smoking history	0.826 (0.261)	2.28 (1.3–3.8)	0.0015
Lack of private insurance	0.698 (0.289)	2.01 (1.4–3.5)	0.0156
Age >60	0.517 (0.262)	1.67 (1.1–2.8)	0.0484
Low SES	0.693 (0.316)	2.00 (1.0–3.7)	0.0281

DISCUSSION

EAC is a devastating diagnosis with a 5-year survival rate of only 20%.¹⁹ A driving force behind this high mortality is the late-stage diagnosis when treatment with curative intent is no longer possible. The incidence of EAC has more than doubled in recent decades, underscoring the need for appropriate screening and surveillance.²⁰ Multiple society guidelines recommend screening for patients with previously identified risk factors, including chronic GERD, BE, male sex, Caucasian race, advanced age, and history of tobacco use, the same demographic and clinical factors identified in this study.^21–23 However, these studies have not evaluated whether screening guidelines are applied equitably across SES in screening eligible populations. This study identified significant socioeconomic disparities in endoscopic screening and surveillance. Furthermore, among previously identified risk factors, low SES and lack of private insurance were also predictors of progression from GERD to EAC. Disparate screening and surveillance practices, among other SES-related barriers to care, likely contribute to the underdiagnosis of BE in low SES patients and failure to diagnose high-risk patients early in the metaplasia-dysplasia-EAC sequence. Consequently, patients from low SES groups are more likely to present with advanced-stage cancer. Implementing structured screening outreach programs, integrating SES-based risk stratification into screening guidelines, and increasing provider awareness of SES disparities in referrals may improve early detection and reduce advanced-stage disease burden.

BE, the precursor lesion to EAC, was unsurprisingly identified as a key risk factor for disease progression. BE was more commonly diagnosed in high SES patients, consistent with a cohort study of 20,412 patients that found high SES patients were 1.6 times more likely to receive a BE diagnosis than those with low SES.²⁴ Although the study did not specifically evaluate progression to EAC, it suggested that the association between high SES and BE might translate to an increased risk of EAC. However, our findings contradict this assumption: high SES was associated with a lower EAC incidence. Further analysis revealed that high SES patients were also more likely to be scoped, whereas low SES patients were least likely to undergo endoscopy, least likely to receive a BE diagnosis, yet most likely to progress to EAC (Fig. 3). Since the goal of screening is early BE detection when disease-modifying interventions remain possible, these findings suggest that low SES patients are systematically deprived of this opportunity. Instead, BE remains undiagnosed, and patients present only once EAC has developed. A similar conclusion was drawn from a single institution study that examined nonadherence to BE screening recommendations over 10 years. They found that low income was the sole predictor of poor adherence (P = 0.036), with only 36% of patients in the lowest income group following recommendations.²⁵ These findings underscore that disparities in access and adherence to screening recommendations deprive patients of the opportunity for early intervention, increasing the likelihood of disease progression to EAC.

FIGURE 3.

Level of SES (A) and median number of EGD per thousand population (B), prevalence of BE (C), and EAC (D) among GERD population. Maps are colored based on quintile cutoffs.OH indicates Ohio; NY, New York; WV, West Virginia.

Poor adherence to cancer screening among low SES patients extends beyond the esophagus, impacting screening for colorectal, cervical, breast, and lung cancer as well.^26–28 In a multiethnic cross-sectional study of 526 patients, those living in high-poverty neighborhoods were 3.6 times more likely to be non-adherent to colorectal cancer screening (P = 0.01).²⁹ Similarly, in a survey of 438 lung cancer screening eligible patients, those with no intent to be screened were more likely to be low SES (P = 0.026), with 57.7% reporting an annual income <$25,000.³⁰ These studies underscore that the SES disparities observed in the present study align with trends across multiple cancers, contributing to substantially worse outcomes in low SES populations. To improve equity and overall EAC outcomes, expansion of access to screening for low SES patients is critical. Other cancers have addressed these disparities with alternatives to biopsy, such as mailed fecal immunochemical tests for colorectal cancer and primary care-based lung cancer screening. For EAC, a nonendoscopic molecular biomarker test has recently been introduced, demonstrating 87.5% sensitivity and 81.2% specificity compared with EGD with biopsy.³¹ Such advancements may help mitigate SES disparities by reducing the socioeconomic burden of screening.

Once diagnosed with BE, surveillance endoscopy is critical for early cancer detection. The 5-year survival for EAC is only 22.9%, but drops to 5.9% for late-stage cancer.^32,33 Thus, the primary goal of surveillance is to enable diagnosis when disease-modifying intervention remains feasible. A systematic review and meta-analysis demonstrated that endoscopic surveillance increases the likelihood of early-stage EAC detection (relative risk : 2.1, 95% CI: 1.1–4.1), while reducing EAC-related mortality (relative risk: 0.60, 95% CI: 0.5–0.7) and all-cause mortality (hazard ratio: 0.75, 95% CI: 0.6–0.9).³⁴ Similarly, the present study found that high SES was associated with more frequent endoscopic evaluations and fewer late-stage diagnoses. However, the most concerning finding was the reverse: low SES patients underwent fewer endoscopies and were more likely to be diagnosed with EAC, with 77% of these diagnoses being stage 3 or 4.

Education influenced outcomes, with college graduates having lower EAC rates and higher BE rates, while high school dropouts had the highest EAC rates. These trends suggest that poorly educated patients face an elevated EAC risk, a pattern observed in other cancers. Nationally, high school dropouts are 3.0 times more likely to develop lung cancer, 1.5 times more likely to develop colorectal cancer, and 3.2 times more likely to develop cervical cancer.² Institutional interventions designed to address this disparity in breast and cervical cancers have proven effective, increasing adherence among low-education populations by prioritizing organized over opportunistic screening programs and engaging primary care physicians in adherence optimization. Similar strategies for EAC screening may yield comparable benefits.

From 1975 to 2015, early-onset (<50 years old) EAC incidence more than tripled, and the proportion of metastatic disease similarly increased (P < 0.01).³⁵ In a study of 263 patients with EAC, 12.1% were diagnosed before age 50, were more likely to present at late-stage (P = 0.04), and had lower 5-year survival (32.6% vs 45.5%).³⁶ Similarly, a study of 242 patients with EAC found that 12.8% were under 50 and had higher rates of metastases (P < 0.05), highlighting the severity and poor prognosis of early-onset EAC.³⁷ In the present study, low SES patients were 2.5 times more likely to be diagnosed with early-onset EAC. Studies looking at early-onset EAC and SES are limited, but similar trends have been observed in other cancers. Early-onset colorectal cancer now comprises 12% of cases.³⁸ Among low sociodemographic index patients the early-onset incidence increased by 46% (P < 0.001) from 2010 to 2019, compared with 16% in the overall population.³⁹ The rise in early-onset cancers in low SES communities may be due to the western diet and “food swamps,” where unhealthy foods dominate available options.⁴⁰ Further research is necessary to identify and mitigate these risk factors, which are exacerbating disparities in cancer outcomes.

Among SES factors, disability, and unemployment were associated with the highest rate of EAC progression. Beyond financial barriers, patients with disabilities may face functional, interpersonal, and environmental obstacles to care, such as limited mobility, communication difficulties, and lack of transportation.⁴¹ A cross-sectional study of over 250,000 patients found higher rate for multiple cancer types among those with disabilities (P < 0.0001 for all cancers).⁴² Unemployment may impact EAC progression through loss of employer-sponsored insurance, the primary source of coverage for most adults under 65.⁴³ A national database study found unemployed individuals were 4 times more likely to be uninsured.⁴⁴ Insurance status impacts screening adherence—a study of 255,616 individuals showed that uninsured or underinsured adults were significantly less likely to receive colorectal, breast, or cervical cancer screenings.⁴⁵ Similarly, a study of 104 colorectal cancer screening eligible patients found that insured patients were more likely to complete screening (77% vs 33%, P < 0.001).⁴⁶ Insurance type also influences access to health care services.⁴⁷ In a national survey, private insurance holders were more likely than those who had public insurance to meet cancer screening guidelines for breast (80.2% vs 70.9%), cervical (76.9% vs 67.3%) and colorectal (66.6% vs 53.8%) cancer.⁴⁸ Our study suggests a similar trend for BE and EAC screening, where low SES is associated with reduced endoscopic surveillance participation, likely contributing to delayed detection and increased progression rates.

Even among patients with BE, low SES patients were twice as likely to progress to EAC, a disparity partially explained by poor adherence to surveillance. However, health habit differences may also contribute. A survey of 401 mothers found low SES women were less likely to exercise, eat breakfast, or maintain oral hygiene.⁴⁹ Similarly, low SES patients in our study were 1.4 times more likely to smoke, a trend reflected in national data showing persistent smoking rates in low SES populations from 1950 to 2011 despite overall declines.⁵⁰ Studies have demonstrated that smoking rates have fallen over time in high SES groups, but not low SES groups, which has translated to SES disparity in all-cancer mortality rates.^51,52 Therefore, the disparities in progression to EAC extend beyond endoscopic access, reflected the broader impact of low SES on health care access, health habits, medical literacy, and distrust of the health care system.

Male sex has been previously identified as the strongest risk factor for progression from GERD to EAC, a trend reaffirmed in the present study.¹⁹ Despite GERD symptoms being more common in women, men have higher BE and EAC rates.^24,53,54 The biological and behavioral factors underlying this disparity remain incompletely understood, but studies have implicated differences in esophageal mucosal susceptibility, sex hormones, and screening patterns.^55–57 Notably, SES disparities persisted in the present study even after adjusting for sex, suggesting that socioeconomic barriers contribute to delayed diagnosis beyond biological risk factors.

This study is subject to limitations, including its retrospective design and the inherent constraints in the granularity of some of the data. SES was determined using a ZIP code-based summary score, a validated and well-established method that enables standardized, population-level analysis across a large region.^16–18 While ZIP codes do not capture individual-level SES variation, they align closely with health care service areas, patient referral patterns, and health care resource distribution, making them a practical tool for assessing SES disparities in disease progression. In addition, grouping patients into larger geographic units, like ZIP codes, reduces statistical variability, improving the reliability of the socioeconomic comparisons.

Racial and ethnic disparities in cancer have been recognized since the 1970s, but disparities in poverty, education, and health insurance have a greater impact than biological differences.^58,59 In 1991, the National Cancer Institute director Dr Samuel Border⁶⁰ declared, “poverty is a carcinogen.” Nevertheless, intersectionality confounds the impact of race and SES on many cancers. Caucasian race is a known risk factor for EAC, which was demonstrated in the present study. Caucasian patients were also more likely to develop BE. However, despite a higher prevalence of African American patients in the low SES group, both low SES and Caucasian race were independent risk factors of progression to EAC. These findings suggest that racial disparities in EAC progression are not solely due to biological differences but are strongly influenced by systemic health care inequities. Addressing these disparities requires targeted interventions to improve access, reduce financial barriers, and increase screening in underserved communities. Race and SES each have a distinct influence on outcomes—African American Medicaid enrollees utilize fewer primary care services than Caucasian counterparts, reflecting systemic inequities.⁶¹ Clinicians should be mindful of racial and socioeconomic disparities in health care delivery and support programs aimed at achieving equitable access and utilization.

CONCLUSIONS

Innovation in diagnosis and treatment of EAC is necessary to improve outcomes, which remain poor compared with many other malignancies. However, equally important is the equitable application of disease-modifying health care practices. In this study, we assessed the risk of progression from GERD and BE to EAC and its associated demographic, clinical, and socioeconomic risk factors in a cohort of more than 160,000 patients. We found that, in addition to the previously identified factors of male sex, advanced age, Caucasian race, BE, and history of smoking, low SES and lack of private insurance were independent risk factors for EAC. We further demonstrated that patients with low SES were significantly less likely to undergo endoscopic evaluation. This lack of screening and surveillance was further reflected in a reduced rate of BE diagnoses, a higher rate of late-stage cancer diagnoses, and a higher rate of early-onset EAC. Collectively, these findings suggest that low SES patients face disparate barriers to accessing preventative screening and surveillance for BE. As a result, these patients are less likely to be diagnosed at a point when intervention may improve their prognosis. Therefore, epidemiological EAC outcomes may be improved by addressing socioeconomic barriers to health care access and improving equitable screening and surveillance practices.