Abstract
Individual-level social determinants of health are associated with pancreatic ductal adenocarcinoma; however, it is currently unknown whether neighborhood-level socioeconomic disadvantage is related to the risk of pancreatic ductal adenocarcinoma diagnosis. Area deprivation index is a validated tool to measure neighborhood-level disadvantage. We conducted a retrospective cohort study of 5 069 429 patients in the Veterans Health Administration between October 1, 2001, and December 31, 2021. Area deprivation index percentiles were grouped using national area deprivation index decile cutoffs. In multivariable analysis, the lowest area deprivation index group, representing the highest neighborhood-level socioeconomic status, was associated with increased hazards for pancreatic ductal adenocarcinoma (adjusted hazard ratio [HR] = 1.13, 95% confidence interval [CI] = 1.06 to 1.21) compared with those with median area deprivation index percentiles. Differences in pancreatic ductal adenocarcinoma hazards were not seen in the other area deprivation index percentiles. These results suggest that within the Veterans Health Administration, a relatively equal access health-care system, there is limited contribution of neighborhood-level socioeconomic deprivation to pancreatic ductal adenocarcinoma, except for patients with the highest neighborhood-level socioeconomic status (lowest area deprivation index).
Pancreatic ductal adenocarcinoma is the third-leading cause of cancer death in the United States.1 Individual-level socioeconomic factors such as Black race have been associated with increased risk for pancreatic ductal adenocarcinoma,2-6 but findings for factors such as education level are mixed.7,8 Additionally, the contribution of neighborhood-level socioeconomic disadvantage to pancreatic ductal adenocarcinoma risk remains poorly understood. Understanding neighborhood context is important because it captures structural and environmental factors, such as health-care access, education quality, and housing stability, that individual-level socioeconomic status (SES) alone may not reflect.9 People of color, who have higher pancreatic ductal adenocarcinoma incidence and mortality, are also more likely to live in high-poverty neighborhoods.10
Area deprivation index is a validated composite measure of neighborhood-level deprivation incorporating variables related to income, education, employment, and housing quality.11 High area deprivation index, indicating more social disadvantage, has been associated with increased risk for common cancers including lung and colorectal.12,13 Neighborhood-level deprivation has not been investigated with respect to diagnosis of pancreatic cancer.
We aimed to investigate whether neighborhood-level socioeconomic disadvantage, as measured by area deprivation index, is independently associated with pancreatic cancer among veterans. The Veterans Health Administration (VA) is the largest nationwide, integrated health-care system in the United States.14 The VA is a relatively equal-access health-care system, providing comprehensive health-care services based on income-based tiers,15 although age, sex, and racial and ethnic disparities have been previously reported.16,17
In this retrospective cohort study, we included veterans aged at least 20 years in the VA with at least 1 inpatient or outpatient encounter and an index date between October 1, 2001, and December 31, 2020. To account for random individualized pancreatic ductal adenocarcinoma risk, we defined the baseline index date as a randomly selected visit date at least 18 months after a patient’s first VA encounter. Patients with pancreatic ductal adenocarcinoma diagnosed prior to their index date were excluded. Incident pancreatic ductal adenocarcinoma was determined by 1 inpatient or 2 outpatient VA or VA-linked Medicare encounters (Table S1). We adjusted for several demographic variables and medical conditions measured at the index date, including age (younger than 50 years, then 5-year increments until age 90 years), sex, race and ethnicity, body mass index categories (Table 1), Medicaid eligibility, alcohol use disorder, smoking status (current, former, never, uninterpretable), acute and chronic pancreatitis, and pancreatic cysts. We also included a validated measure of all-cause mortality known as the Veterans Aging Cohort Study (VACS)–Charlson Comorbidity Index, along with human immunodeficiency virus, diabetes, and liver disease defined per Charlson Comorbidity Index18 (Table S1).
Table 1.
Baseline characteristics
| ADI rank | ADI 1-10 (n = 272 363) | ADI 11-20 (n = 435 413) | ADI 21-30 (n = 590 647) | ADI 31-40 (n = 731 552) | ADI 41-50 (n = 805 147) | ADI 51-60 (n = 823 115) | ADI 61-70 (n = 802 147) | ADI 71-80 (n = 754 946) | ADI 81-90 (n = 709 259) | ADI 91-100 (n = 586 848) |
|---|---|---|---|---|---|---|---|---|---|---|
| Individuals with pancreatic ductal adenocarcinoma, No. (%) | 1704 (0.63) | 2109 (0.48) | 2794 (0.47) | 3256 (0.45) | 3660 (0.45) | 3855 (0.47) | 3678 (0.46) | 3549 (0.47) | 3523 (0.50) | 3114 (0.53) |
| Age at baseline, median (IQR), y | 66 (24) | 64 (24) | 63 (24) | 63 (24) | 63 (24) | 63 (23) | 63 (22) | 63 (21) | 63 (20) | 62 (18) |
| Younger than 50 | 60 716 (22.3) | 113 387 (26.0) | 158 883 (26.9) | 198 679 (27.2) | 216 314 (26.9) | 212 869 (25.9) | 195 928 (24.4) | 171 731 (22.7) | 147 126 (20.7) | 111 336 (19.0) |
| 50-54 | 16 488 (6.1) | 30 655 (7.0) | 43 129 (7.3) | 53 815 (7.4) | 59 724 (7.4) | 62 205 (7.6) | 61 467 (7.7) | 58 762 (7.8) | 57 776 (8.1) | 54 371 (9.3) |
| 55-59 | 21 251 (7.8) | 36 880 (8.5) | 50 892 (8.6) | 64 661 (8.8) | 74 168 (9.2) | 78 689 (9.6) | 80 758 (10.1) | 80 174 (10.6) | 80 656 (11.4) | 77 212 (13.2) |
| 60-64 | 27 582 (10.1) | 46 751 (10.7) | 64 525 (10.9) | 81 149 (11.1) | 93 617 (11.6) | 99 810 (12.1) | 102 169 (12.7) | 101 435 (13.4) | 101 023 (14.2) | 90 352 (15.4) |
| 65-69 | 35 037 (12.9) | 57 744 (13.3) | 79 974 (13.5) | 100 990 (13.8) | 112 859 (14.0) | 117 095 (14.2) | 116 009 (14.5) | 111 662 (14.8) | 106 550 (15.0) | 85 886 (14.6) |
| 70-74 | 34 405 (12.6) | 52 589 (12.1) | 71 333 (12.1) | 87 630 (12.0) | 94 101 (11.7) | 95 880 (11.6) | 92 959 (11.6) | 88 536 (11.7) | 82 423 (11.6) | 63 237 (10.8) |
| 75-79 | 26 566 (9.8) | 36 652 (8.4) | 47 441 (8.0) | 57 365 (7.8) | 62 119 (7.7) | 63 250 (7.7) | 61 898 (7.7) | 58 614 (7.8) | 55 726 (7.9) | 43 381 (7.4) |
| 80-84 | 24 647 (9.0) | 30 973 (7.1) | 38 227 (6.5) | 45 874 (6.3) | 48 863 (6.1) | 49 909 (6.1) | 48 710 (6.1) | 45 353 (6.0) | 42 579 (6.0) | 33 350 (5.7) |
| 85-90 | 19 344 (7.1) | 22 984 (5.3) | 27 922 (4.7) | 32 182 (4.4) | 33 875 (4.2) | 33 981 (4.1) | 33 180 (4.1) | 30 406 (4.0) | 27 979 (3.9) | 21 859 (3.7) |
| Unknown | 6327 (2.3) | 6798 (1.6) | 8321 (1.4) | 9207 (1.3) | 9507 (1.2) | 9427 (1.1%) | 9069 (1.1) | 8273 (1.1) | 7421 (1.0) | 5864 (1.0) |
| Male, No. (%) | 254 876 (93.6) | 402 530 (92.4) | 542 555 (91.9) | 670 716 (91.7) | 737 778 (91.6) | 755 888 (91.8) | 740 000 (92.3) | 700 207 (92.7) | 662 663 (93.4) | 549 950 (93.7) |
| Race, No. (%) | ||||||||||
| White | 195 665 (71.8) | 314 417 (72.2) | 445 147 (75.4) | 566 418 (77.4) | 623 248 (77.4) | 634 130 (77.0) | 611 517 (76.2) | 564 475 (74.8) | 498 683 (70.3) | 334 791 (57.0) |
| Black | 30 617 (11.2) | 60 541 (13.9) | 81 374 (13.8) | 98 978 (13.5) | 113 518 (14.1) | 121 566 (14.8) | 123 875 (15.4) | 126 715 (16.8) | 139 554 (19.7) | 179 530 (30.6) |
| Hispanic | 15 498 (5.7) | 25 882 (5.9) | 30 107 (5.1) | 31 540 (4.3) | 32 763 (4.1) | 32 262 (3.9) | 32 398 (4.0) | 31 486 (4.2) | 39 599 (5.6) | 42 897 (7.3) |
| Other | 30 302 (11.1) | 34 077 (7.8) | 33 432 (5.7) | 33 908 (4.6) | 34 965 (4.3) | 34 589 (4.2) | 33 793 (4.2) | 31 791 (4.2) | 31 016 (4.4) | 29 297 (5.0) |
| Unknown | 281 (0.1) | 496 (0.1) | 587 (0.1) | 708 (0.1) | 653 (0.1) | 568 (0.1) | 564 (0.1) | 479 (0.1) | 407 (0.1) | 333 (0.1) |
| Smoking, No. (%) | ||||||||||
| Never | 90 664 (33.3) | 140 926 (32.4) | 188 152 (31.9) | 226 397 (30.9) | 240 518 (29.9) | 238 146 (28.9) | 224 195 (27.9) | 202 696 (26.8) | 189 118 (26.7) | 158 249 (27.0) |
| Former | 72 403 (26.6) | 117 640 (27.0) | 164 394 (27.8) | 207 696 (28.4) | 228 335 (28.4) | 232 146 (28.2) | 222 319 (27.7) | 205 891 (27.3) | 183 508 (25.9) | 141 976 (24.2) |
| Current | 40 791 (15.0) | 75 081 (17.2) | 111 658 (18.9) | 152 167 (20.8) | 185 167 (23.0) | 205 729 (25.0) | 217 101 (27.1) | 219 590 (29.1) | 219 157 (30.9) | 196 530 (33.5) |
| Uninterpretablea | 20 769 (7.6) | 34 648 (8.0) | 50 194 (8.5) | 63 399 (8.7) | 67 539 (8.4) | 67 509 (8.2) | 64 767 (8.1) | 60 583 (8.0) | 60 352 (8.5) | 48 377 (8.2) |
| Unknown | 47 736 (17.5) | 67 118 (15.4) | 76 249 (12.9) | 81 893 (11.2) | 83 588 (10.4) | 79 585 (9.7) | 73 765 (9.2%) | 66 186 (8.8) | 57 123 (8.1) | 41 716 (7.1) |
| Alcohol use disorder, No. (%) | 31 125 (11.4) | 51 407 (11.8) | 69 909 (11.8) | 88 771 (12.1) | 101 195 (12.6) | 107 920 (13.1) | 111 924 (14.0) | 111 206 (14.7) | 112 421 (15.9) | 110 685 (18.9) |
| Body mass index, No. (%), kg/m2 | ||||||||||
| Mean (SD) | 28.5 (5.2) | 29.2 (5.4) | 29.5 (5.6) | 29.7 (5.7) | 29.8 (5.8) | 29.9 (5.9) | 29.8 (6.0) | 29.8 (6.1) | 29.7 (6.1) | 29.4 (6.3) |
| Underweight, <18.5 | 2015 (0.7) | 2865 (0.7) | 3914 (0.7) | 4775 (0.7) | 5603 (0.7) | 6570 (0.8) | 7243 (0.9) | 7392 (1.0) | 8174 (1.2) | 8442 (1.4) |
| Normal weight, 18.5-25 | 64 724 (23.8) | 88 426 (20.3) | 112 310 (19.0) | 133 798 (18.3) | 145 724 (18.1) | 150 363 (18.3) | 152 045 (19.0) | 145 800 (19.3) | 143 867 (20.3) | 129 658 (22.1) |
| Overweight, 25-30 | 112 782 (41.4) | 172 940 (39.7) | 228 071 (38.6) | 275 658 (37.7) | 297 187 (36.9) | 299 203 (36.4) | 286 699 (35.7) | 267 143 (35.4) | 249 307 (35.2) | 203 496 (34.7) |
| Obese, >30 | 86 903 (31.9) | 162 535 (37.3) | 236 220 (40.0) | 305 602 (41.8) | 344 352 (42.8) | 354 498 (43.1) | 343 928 (42.9) | 323 231 (42.8) | 297 151 (41.9) | 236 180 (40.2) |
| Unknown | 5939 (2.2) | 8647 (2.0) | 10 132 (1.7) | 11 719 (1.6) | 12 281 (1.5) | 12 481 (1.5) | 12 232 (1.5) | 11 380 (1.5) | 10 760 (1.5) | 9072 (1.5) |
| Medicaid eligibility, No. (%) | 3690 (1.4) | 4552 (1.0) | 5226 (0.9) | 5707 (0.8) | 6007 (0.7) | 6489 (0.8) | 6780 (0.8) | 6585 (0.9) | 6957 (1.0) | 7546 (1.3) |
| Pancreatitis, No. (%) | 3646 (1.3) | 5663 (1.3) | 7784 (1.3) | 9811 (1.3) | 11 679 (1.5) | 12 477 (1.5) | 12 884 (1.6) | 12 973 (1.7) | 13 189 (1.9) | 11 923 (2.0) |
| Pancreatic cyst, No. (%) | 921 (0.3) | 1371 (0.3) | 1621 (0.3) | 1902 (0.3) | 2126 (0.3) | 2112 (0.3) | 2054 (0.3) | 1922 (0.3) | 1864 (0.3) | 1639 (0.3) |
| Human immunodeficiency virus, No. (%) | 1819 (0.7) | 2177 (0.5) | 2426 (0.4) | 2630 (0.4) | 2721 (0.3) | 2867 (0.3) | 2979 (0.4) | 3024 (0.4) | 3216 (0.5) | 3475 (0.6) |
| Diabetes, No. (%) | 54 912 (20.2) | 91 034 (20.9) | 126 035 (21.3) | 161 361 (22.1) | 185 290 (23.0) | 197 054 (23.9) | 200 889 (25.0) | 196 314 (26.0) | 193 742 (27.3) | 165 232 (28.2) |
| Liver disease, No. (%) | 10 532 (3.9) | 15 828 (3.6) | 20 555 (3.5) | 25 087 (3.4) | 27 975 (3.5) | 29 360 (3.6) | 29 988 (3.7) | 29 562 (3.9) | 29 770 (4.2) | 28 630 (4.9) |
| Veterans Aging Cohort Study-Charlson Comorbidity Index 1, mean (SD) | 35.2 (14.2) | 33.5 (14.1) | 33.0 (14.0) | 32.9 (14.0) | 33.1 (13.9) | 33.5 (13.9) | 34.0 (13.8) | 34.6 (13.6) | 35.2 (13.3) | 35.7 (13.0) |
Abbreviations: ADI = area deprivation index; IQR = interquartile range.
Smoking status was categorized as uninterpretable if unclear or contradictory to prior smoking status data points.
Area deprivation index data are available in the VA between 2012 - 2021. We obtained the area deprivation index closest to the index date, sourced from the Neighborhood Atlas for each census block group and individually linked using geocoded residence data.19 We categorized area deprivation index into 10 groups using national area deprivation index decile cutoffs. Area deprivation index was inversely related to neighborhood-level SES (eg, area deprivation index percentile 1-10 corresponds to the highest advantaged neighborhoods).
Using STATA v18 (College Station, TX, USA), we used complete case analysis and performed multivariable Cox regression to assess associations with time to pancreatic ductal adenocarcinoma, censoring at death, 2 years after the last VA or Medicare encounter to address loss to follow-up, or the end of the study period (December 31, 2021). The proportional hazards assumption was verified by plotting log-log survival curves. We performed subgroup analyses stratifying for race, smoking, and diabetes.
We gathered components of area deprivation index and census variables using the Sociome package in R.20 We evaluated the most meaningful components of the composite area deprivation index score through linear regression (components with absolute T values greater than 100) and a random forest model (components that contributed most to node purity on a variable importance plot) and identified the overlap in both models.
There were 6 511 437 eligible Veterans with available area deprivation index data. Of these individuals, 92.4% (n = 6 017 163) were male, 73.5% (n = 4 788 491) were White, 16.5% (n = 1 076 268) were Black, 4.8% (n = 314 432) were Hispanic, and 5.0% (n = 327 170) were of other race and ethnicity (American Indian or Alaska Native, Asian, Native Hawaiian or other Pacific Islander, and multiracial). White patients had lower mean area deprivation index percentiles (53.9) compared both Black patients (mean difference = 6.9, P < .001) and Hispanic patients (mean difference = 2.2, P < .001). In total, 5 069 429 (77.9%) qualified for our complete case analysis.
Overall, 31 242 (0.62%) pancreatic ductal adenocarcinoma patients were identified, with a median baseline age of 63 years (interquartile range [IQR] = 50-72 years) (Table 1). The median time to follow-up was 4.8 years. Compared with other area deprivation index percentiles, patients in the most advantaged neighborhoods had an older age at pancreatic ductal adenocarcinoma onset (66 years vs 62-64 years; χ2 test, P < .001) and a higher rate of incident pancreatic ductal adenocarcinoma (91.5 vs 69.2-73.6 cases per 100 000 person-years) (Figure S1).
Compared with veterans in the median area deprivation index (51-60), those in the most advantaged neighborhoods were associated with an increased hazard for incident pancreatic ductal adenocarcinoma (Table 2; unadjusted hazard ratio [HR] = 1.22, 95% confidence interval [CI] = 1.14 to 1.30), which persisted after adjustment for race and lifestyle factors, clinical factors, and VACS Charlson Comorbidity Index (adjusted HR = 1.13, 95% CI = 1.06 to 1.21). There were no clinically significant differences in pancreatic ductal adenocarcinoma hazard seen among the other area deprivation index percentiles.
Table 2.
Hazard of pancreatic ductal adenocarcinoma by area deprivation index percentile
| Area deprivation index percentile | Pancreatic ductal adenocarcinoma events, No. (%) | Unadjusted, HR (95% CI) (n = 5 069 429) | Model adjusted for demographic,a lifestyle,b and clinical factors,c HR (95% CI) |
Model additionally adjusted for Veterans Aging Cohort Study-Charlson Comorbidity Index, HR (95% CI) |
|---|---|---|---|---|
| 1-10 | 1704 (0.63) | 1.22 (1.14 to 1.30) | 1.15 (1.07 to 1.22) | 1.13 (1.06 to 1.21) |
| 11-20 | 2109 (0.48) | 1.00 (0.94 to 1.06) | 1.00 (0.94 to 1.06) | 1.00 (0.94 to 1.06) |
| 21-30 | 2794 (0.47) | 1.00 (0.95 to 1.06) | 1.02 (0.97 to 1.08) | 1.02 (0.97 to 1.08) |
| 31-40 | 3256 (0.45) | 0.95 (0.90 to 1.00) | 0.97 (0.92 to 1.02) | 0.97 (0.92 to 1.02) |
| 41-50 | 3660 (0.45) | 0.97 (0.93 to 1.02) | 0.99 (0.94 to 1.04) | 0.99 (0.94 to 1.04) |
| 51-60 | 3855 (0.47) | 1 (Referent) | 1 (Referent) | 1 (Referent) |
| 61-70 | 3678 (0.46) | 0.96 (0.91 to 1.01) | 0.94 (0.90 to 0.99) | 0.94 (0.89 to 0.99) |
| 71-80 | 3549 (0.47) | 0.98 (0.93 to 1.03) | 0.94 (0.90 to 0.99) | 0.94 (0.89 to 0.99) |
| 81-90 | 3523 (0.50) | 1.01 (0.96 to 1.07) | 0.95 (0.90 to 1.00) | 0.94 (0.90 to 0.99) |
| 91-100 | 3114 (0.53) | 1.04 (0.99 to 1.10) | 0.95 (0.90 to 1.00) | 0.94 (0.90 to 0.99) |
Abbreviations: CI = confidence interval; HR = hazard ratio.
Demographic factors include age, sex, and race and ethnicity.
Lifestyle factors include body mass index, smoking, alcohol use disorder, and Medicaid eligibility.
Clinical factors include pancreatitis, pancreatic cysts, diabetes, human immunodeficiency virus, and liver disease.
The lowest area deprivation index group remained associated with slightly increased hazard of pancreatic ductal adenocarcinoma after separate stratification by race (White vs Black), smoking status (former, never, current), and diabetes (diabetics vs nondiabetics). However, the hazard values were no longer statistically significant among Black race and current smokers, likely because of sample size. Additional details are in Table S2.
We analyzed the main contributing components to composite area deprivation index percentile. Median house value, financial strength, median mortgage, and median rent appeared to have the most impact on area deprivation index with agreement in variable importance between the linear model and the random forest model (Figure S2, Table S3).
Among veterans, we found that the highest neighborhood-level SES was associated with an elevated hazard for pancreatic ductal adenocarcinoma, even after adjusting for demographic, lifestyle, and clinical factors. None of the other area deprivation index percentiles showed clinically significant associations with pancreatic ductal adenocarcinoma.
There may be small lifestyle and medical utilization differences not measured in the electronic health record, such as increased medical testing, surveillance within and outside the VA, or military exposures, that may have unexpectedly placed those from more affluent neighborhoods at higher hazard of pancreatic ductal adenocarcinoma diagnosis.21 It is also possible that those with the highest neighborhood-level SES are living longer and therefore more likely to be diagnosed with pancreatic ductal adenocarcinoma, as this group was found to have older ages at baseline and at pancreatic ductal adenocarcinoma diagnosis. However, after adjusting for mortality indicators using the VACS Charlson Comorbidity Index, there remained an increased hazard among those living in the most advantaged neighborhoods. Notably, the association between highest neighborhood-level SES and pancreatic ductal adenocarcinoma was relatively small and would need to be further investigated in other populations, such as those with less access to health-care resources. Although underinsurance, low income, and less educational attainment are noted to be barriers to cancer diagnosis, the VA emphasis on primary prevention and relatively equal health-care access likely overcomes some of the systems level barriers to care, leading to less differentiated pancreatic ductal adenocarcinoma hazard among other area deprivation index groups.22,23
This is the first study that addresses the overall relationship of pancreatic ductal adenocarcinoma with neighborhood-level socioeconomic deprivation. Conducting this study in a large, national population with longitudinal data improves the generalizability and strength of our findings. With the extensive VA and VA-linked Medicare data repository, we accounted for a multitude of clinical conditions, including liver disease and pancreatic diseases, along with lifestyle factors, such as smoking and alcohol use.
Previously, the 9 studies investigating area deprivation index and pancreatic ductal adenocarcinoma have solely focused on pancreatic ductal adenocarcinoma outcomes.24-32 Of these studies, 7 found an association with high area deprivation index and decreased receipt of adjuvant therapy, increased complications, or decreased overall survival.24,26-31 Del Valle et al.33 used a similar tool to the area deprivation index, known as the Social Deprivation Index, to investigate pancreatic ductal adenocarcinoma outcomes of resection and survival within the VA population and similarly observed no differences among levels of neighborhood social deprivation, suggesting that the VA’s relatively equal access system may overcome neighborhood-level inequalities.
Previous studies of breast cancer,34,35 thyroid cancer,36 and melanoma37 notably found a higher cancer incidence or risk among advantaged neighborhoods, hypothesized to be associated with lifestyle behaviors such as fewer social connections, increased health-care utilization, and environmental exposures that conferred greater risk.
Our study has a few limitations. Our cohort was mostly male, limiting generalizability; however, pancreatic ductal adenocarcinoma incidence does not differ by sex.1 Other social determinants of health previously associated with worse pancreatic ductal adenocarcinoma outcomes, such as rurality and health literacy, were not included.38,39 However, area deprivation index does incorporate education level, which correlates with health literacy.40 Death was treated as a censoring event rather than a competing risk; thus, we adjusted for baseline comorbidity using the VACS Charlson Comorbidity Index, which partially captures mortality risk.
Overall, there was a mildly elevated pancreatic ductal adenocarcinoma hazard in the highest neighborhood-level SES. Future studies should investigate this relationship in settings with greater health-care access disparities, as neighborhood-level SES may be less important in relatively equal-access health-care systems. Factors related to income and housing quality were most impactful to the composite area deprivation index score among veterans, and their influence on pancreatic ductal adenocarcinoma in future studies should be explored.
Supplementary Material
Acknowledgments
The funder had no role in the design of the study; the collection, analysis, or interpretation of the data; or the writing of the manuscript and decision to submit it for publication.
Contributor Information
Rachel N Levinson, Section of Digestive Diseases, Department of Internal Medicine, Yale School of Medicine, New Haven, CT, United States; Department of Medicine, VA Connecticut Healthcare System, West Haven, CT, United States.
Ryan Bushman, Section of Digestive Diseases, Department of Internal Medicine, Yale School of Medicine, New Haven, CT, United States; Department of Medicine, VA Connecticut Healthcare System, West Haven, CT, United States.
Catherine Mezzacappa, Section of Digestive Diseases, Department of Internal Medicine, Yale School of Medicine, New Haven, CT, United States.
Janet P Tate, Department of Medicine, VA Connecticut Healthcare System, West Haven, CT, United States; Section of General Medicine, Department of Internal Medicine, Yale School of Medicine, New Haven, CT, United States.
Amy C Justice, Department of Medicine, VA Connecticut Healthcare System, West Haven, CT, United States; Section of General Medicine, Department of Internal Medicine, Yale School of Medicine, New Haven, CT, United States; School of Public Health, Yale University, New Haven, CT, United States.
Louise L Wang, Section of Digestive Diseases, Department of Internal Medicine, Yale School of Medicine, New Haven, CT, United States; Department of Medicine, VA Connecticut Healthcare System, West Haven, CT, United States; Division of Gastroenterology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, United States.
Author contributions
Rachel Levinson (Conceptualization, Formal analysis, Investigation, Methodology, Software, Validation, Visualization, Writing—original draft, Writing—review & editing), Ryan Bushman (Formal analysis, Investigation, Writing—original draft), Catherine Mezzacappa (Methodology, Writing—review & editing), Janet Tate (Data curation, Resources, Writing—review & editing), Amy Justice (Project administration, Supervision, Writing—review & editing), and Louise Wang (Conceptualization, Funding acquisition, Methodology, Project administration, Supervision, Visualization, Writing—review & editing)
Supplementary material
Supplementary material is available at JNCI Cancer Spectrum online.
Funding
Research reported in this publication was supported by the VA Career Development Award-2 (IK2 Act BX005891) and the Yale School of Medicine Leon Rosenberg, M.D. Medical Student Research Fund in Genetics.
Conflicts of interest
There are no conflicts of interest to report from any of the authors.
Data availability
The data and analytic methods will be shared upon reasonable request to the corresponding author. Data access is restricted to qualified researchers with a VA appointment and can be made available only after obtaining the necessary approvals. Access is granted through the VA Informatics and Computing Infrastructure. Data cannot be shared outside the VA system.
References
- 1. Surveillance Research Program, National Cancer Institute. Cancer Stat Facts: Common Cancer Sites. SEER. 2023. Accessed November 15, 2023. https://seer.cancer.gov/statfacts/html/common.html
- 2. Moshayedi N, Escobedo AL, Thomassian S, Osipov A, Hendifar AE. Race, sex, age, and geographic disparities in pancreatic cancer incidence. J Clin Oncol. 2022;40:520-520. 10.1200/JCO.2022.40.4_suppl.52034878806 [DOI] [Google Scholar]
- 3. Tavakkoli A, Singal AG, Waljee AK, et al. Racial disparities and trends in pancreatic cancer incidence and mortality in the United States. Clin Gastroenterol Hepatol. 2020;18:171-178.e10. 10.1016/j.cgh.2019.05.059 [DOI] [PubMed] [Google Scholar]
- 4. SEER*Explorer Application. All Cancer Sites Combined Recent Trends in SEER Age-Adjusted Incidence Rates, 2000-2022. Accessed October 9, 2025. https://seer.cancer.gov/statistics-network/explorer/application.html? site=40&data_type=1&graph_type=2&compareBy=race&chk_race_6=6&chk_race_5=5&chk_race_4=4&chk_race_9=9&chk_race_8=8&rate_type=2&sex=1&age_range=1&stage=101&advopt_precision=2&advopt_show_ci=on&advopt_show_count=on&hdn_view=0&advopt_show_apc=on&advopt_display=2#resultsRegion0
- 5. Gordon-Dseagu VL, Devesa SS, Goggins M, Stolzenberg-Solomon R. Pancreatic cancer incidence trends: evidence from the Surveillance, Epidemiology and End Results (SEER) population-based data. Int J Epidemiol. 2018;47:427-439. 10.1093/ije/dyx232 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Chang KJ, Parasher G, Christie C, Largent J, Anton-Culver H. Risk of pancreatic adenocarcinoma. Cancer. 2005;103:349-357. 10.1002/cncr.20771 [DOI] [PubMed] [Google Scholar]
- 7. Mouw T, Koster A, Wright ME, et al. Education and risk of cancer in a large cohort of men and women in the United States. PLoS One. 2008;3:e3639. 10.1371/journal.pone.0003639 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Cirera L, Huerta JM, Chirlaque MD, et al. Socioeconomic effect of education on pancreatic cancer risk in western Europe: an update on the EPIC Cohorts Study. Cancer Epidemiol Biomarkers Prev. 2019;28:1089-1092. 10.1158/1055-9965.EPI-18-1153 [DOI] [PubMed] [Google Scholar]
- 9. Cheng E, Soulos PR, Irwin ML, et al. Neighborhood and individual socioeconomic disadvantage and survival among patients with nonmetastatic common cancers. JAMA Netw Open. 2021;4:e2139593. 10.1001/jamanetworkopen.2021.39593 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Neighborhood Poverty: All neighborhoods should be communities of opportunity. National Equity Atlas. Accessed July 31, 2024. https://nationalequityatlas.org/indicators/Neighborhood_poverty
- 11. Kind AJH, Buckingham WR. Making neighborhood-disadvantage metrics accessible–the neighborhood atlas. N Engl J Med. 2018;378:2456-2458. 10.1056/NEJMp1802313 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Fairfield KM, Black AW, Ziller EC, et al. Area deprivation index and rurality in relation to lung cancer prevalence and mortality in a rural state. JNCI Cancer Spectrum. 2020;4:pkaa011. 10.1093/jncics/pkaa011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Purrington KS, Hastert TA, Madhav KC, et al. The role of area-level socioeconomic disadvantage in racial disparities in cancer incidence in metropolitan Detroit. Cancer Med. 2023;12:14623-14635. 10.1002/cam4.6065 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Brown SH, Lincoln MJ, Groen PJ, Kolodner RM. VistA—U.S. Department of Veterans Affairs national-scale HIS. Int J Med Inform. 2003;69:135-156. 10.1016/S1386-5056(02)00131-4 [DOI] [PubMed] [Google Scholar]
- 15. Silva A, Tarlov E, French DD, Huo Z, Martinez RN, Stroupe KT. Veterans affairs health system enrollment and health care utilization after the Affordable Care Act: initial insights. Military Med. 2016;181:469-475. 10.7205/MILMED-D-15-00094 [DOI] [PubMed] [Google Scholar]
- 16. Saha S, Freeman M, Toure J, Tippens KM, Weeks C, Ibrahim S. Racial and ethnic disparities in the VA health care system: a systematic review. J Gen Intern Med. 2008;23:654-671. 10.1007/s11606-008-0521-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Rockville (MD): Agency for Healthcare Research and Quality (US). Contrast 2 findings: disparities within the veteran population. National Healthcare Quality and Disparities Report: chartbook on Healthcare for Veterans. 2020. Accessed August 21, 2024. https://www.ncbi.nlm.nih.gov/books/NBK578547/ [Google Scholar]
- 18. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373-383. 10.1016/0021-9681(87)90171-8 [DOI] [PubMed] [Google Scholar]
- 19. Hahn Z, Hotchkiss J, Atwood C, et al. Travel burden as a measure of healthcare access and the impact of telehealth within the Veterans Health Administration. J Gen Intern Med. 2023;38:805-813. 10.1007/s11606-023-08125-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Krieger N, Dalton J, Wang C, Perzynski A. sociome: Operationalizing Social Determinants of Health Data for Researchers version 2.2.5 from CRAN. Accessed September 11, 2024. https://rdrr.io/cran/sociome/
- 21. Welch HG, Fisher ES. Income and cancer overdiagnosis–when too much care is harmful. New Eng J Med. 2017;376:2208-2209. 10.1056/NEJMp1615069 [DOI] [PubMed] [Google Scholar]
- 22. Kaboli PJ, Shimada SL. A decade of focus on and improvement in access to care in the Veterans Health Administration. J Gen Intern Med. 2023;38:801-804. 10.1007/s11606-023-08208-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Fabregas JC, Riley KE, Brant JM, George TJ, Orav EJ, Lam MB. Association of social determinants of health with late diagnosis and survival of patients with pancreatic cancer. J Gastrointest Oncol. 2022;13:1204-1214. 10.21037/jgo-21-788 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Bonner SN, Ibrahim AM, Kunnath N, Dimick JB, Nathan H. Neighborhood deprivation, hospital quality, and mortality after cancer surgery. Ann Surg. 2023;277:73-78. 10.1097/SLA.0000000000005712 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Brooks GA, Tomaino MR, Ramkumar N, et al. Association of rurality, socioeconomic status, and race with pancreatic cancer surgical treatment and survival. JNCI: J Natl Cancer Inst. 2023;115:1171-1178. 10.1093/jnci/djad102 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Khalid A, Pasha SA, Demyan L, et al. Evaluating the Association of Area Deprivation Index (ADI) on postoperative outcomes in pancreatic adenocarcinoma. J Surg Oncol. 2025;131:637-645. 10.1002/jso.27996 [DOI] [PubMed] [Google Scholar]
- 27. Lim JS, Lozano V, Heard J, et al. Redefining at-risk patients undergoing pancreaticoduodenectomy: impact of socioeconomic factors including Area Deprivation Index and distance traveled. Surgery. 2025;179:108804. 10.1016/j.surg.2024.07.059 [DOI] [PubMed] [Google Scholar]
- 28. Marcinak CT, Praska CE, Vidri RJ, et al. Association of neighborhood disadvantage with short- and long-term outcomes after pancreatectomy for pancreatic ductal adenocarcinoma. Ann Surg Oncol. 2024;31:488-498. 10.1245/s10434-023-14347-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Mora J, Krepline AN, Aldakkak M, et al. Adjuvant therapy rates and overall survival in patients with localized pancreatic cancer from high Area Deprivation Index neighborhoods. Am J Surg. 2021;222:10-17. 10.1016/j.amjsurg.2020.12.001 [DOI] [PubMed] [Google Scholar]
- 30. Murimwa GZ, Meier J, Nehrubabu M, Zeh HJ, Yopp AC, Polanco PM. Implications of the interaction between travel burden and area deprivation for patients with pancreatic cancer. Am J Surg. 2023;226:515-522. 10.1016/j.amjsurg.2023.06.013 [DOI] [PubMed] [Google Scholar]
- 31. Pothuri V, Zárate Rodriguez JG, Kasting C, et al. Area deprivation and rurality impact overall survival and adjuvant therapy administration in patients with pancreatic ductal adenocarcinoma (PDAC). HPB. 2023;25:1545-1554. 10.1016/j.hpb.2023.08.005 [DOI] [PubMed] [Google Scholar]
- 32. Powers BD, Fulp W, Dhahri A, et al. The impact of socioeconomic deprivation on clinical outcomes for pancreatic adenocarcinoma at a high-volume cancer center: a retrospective cohort analysis. Ann Surg. 2021;274:e564-e573. 10.1097/SLA.0000000000003706 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Del Valle JP, Fillmore NR, Molina G, et al. Socioeconomic disparities in pancreas cancer resection and survival in the Veterans Health Administration. Ann Surg Oncol. 2022;29:3194-3202. 10.1245/s10434-021-11250-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Conroy S, Clarke CA, Yang J, et al. Contextual impact of neighborhood obesogenic factors on postmenopausal breast cancer: the Multiethnic Cohort. Cancer Epidemiol Biomarkers Prev. 2017;26:480-489. 10.1158/1055-9965.EPI-16-0941 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Keegan THM, John EM, Fish KM, Alfaro-Velcamp T, Clarke CA, Gomez SL. Breast cancer incidence patterns among California Hispanic women: differences by nativity and residence in an enclave. Cancer Epidemiol Biomarkers Prev. 2010;19:1208-1218. 10.1158/1055-9965.EPI-10-0021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Li N, Du XL, Reitzel LR, Xu L, Sturgis EM. Impact of enhanced detection on the increase in thyroid cancer incidence in the United States: review of incidence trends by socioeconomic status within the Surveillance, Epidemiology, and End Results registry, 1980–2008. Thyroid. 2013;23:103-110. 10.1089/thy.2012.0392 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Clarke CA, McKinley M, Hurley S, et al. Continued increase in melanoma incidence across all socioeconomic status groups in California, 1998–2012. Journal of Investigative Dermatology. 2017;137:2282-2290. 10.1016/j.jid.2017.06.024 [DOI] [PubMed] [Google Scholar]
- 38. Chu QD, Hsieh MC, Gibbs JF, Wu XC. Social determinants of health associated with poor outcome for rural patients following resected pancreatic cancer. J Gastrointest Oncol. 2021;12:2567-2578. 10.21037/jgo-20-583 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Wall NR, Fuller RN, Morcos A, De Leon M. Pancreatic cancer health disparity: pharmacologic anthropology. Cancers (Basel). 2023;15:5070. 10.3390/cancers15205070 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.U.S. Department of Health and Human Services. America’s Health Literacy: Why We Need Accessible Health Information. Published online 2008. https://www.ahrq.gov/sites/default/files/wysiwyg/health-literacy/dhhs-2008-issue-brief.pdf
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data and analytic methods will be shared upon reasonable request to the corresponding author. Data access is restricted to qualified researchers with a VA appointment and can be made available only after obtaining the necessary approvals. Access is granted through the VA Informatics and Computing Infrastructure. Data cannot be shared outside the VA system.