Racial and ethnic minorities with acute pancreatitis live in neighborhoods with higher social vulnerability scores

Abstract

Objectives:

The primary objective was to determine differences in Social Vulnerability Index (SVI) scores among minorities (African-Americans and Hispanics) with acute pancreatitis (AP) compared to non-Hispanic whites (NHWs) with AP. The secondary objectives were to determine differences in diet, sulfidogenic bacteria gene copy numbers (gcn) and hydrogen sulfide (H₂S) levels between the two groups.

Methods:

AP patients were enrolled during hospitalization (n=54). Patient residential addresses were geocoded and the Centers for Disease Control and Prevention’s SVI scores were appended. Dietary intake and serum H₂S levels were determined. Microbial DNA were isolated from stool and qPCR was used to determine gcn of sulfidogenic bacteria.

Results:

Minorities had higher SVI scores compared to NHWs (p=0.006). They also had lower consumption of beneficial nutrients such as omega-3 fatty acids [stearidonic (p=0.019), and eicosapentaenoic acid (p=0.042)], vitamin D (p=0.025), and protein from seafood (p=0.031). Lastly, minorities had higher gcn of pan-dissimilatory sulfite reductase A (pan-dsrA) (p=0.033) but no statistically significant differences in H₂S levels (p=0.226).

Conclusion:

Minorities with AP have higher SVI compared to NHWs with AP. Higher SVI scores, lower consumption of beneficial nutrients, and increased gcn of pan-dsrA in minorities with AP suggest that neighborhood vulnerability could be contributing to AP inequities.

Introduction

Acute pancreatitis (AP), an inflammatory disease of the pancreas, is a leading cause of gastrointestinal hospitalizations in the USA.^1,2 There has been a gradual increase in the incidence of AP over the past decades, imposing a burden on the health care system.^3,4 Recent studies also showed associations between dietary intake, gut microbiome, and AP risk.^5–7 Furthermore, up to 20% of AP patients develop moderately-severe or severe acute pancreatitis (SAP) leading to increased mortality and approximately 30% of AP patients have recurrent episodes, which can worsen parenchymal injury and increase the risk for subsequent complications, such as exocrine pancreatic insufficiency (EPI) and pancreatogenic diabetes.^8–11

Racial and ethnic minorities are at an increased risk of developing AP and AP-related complications including diabetes and EPI, with African Americans (AA) having a higher incidence of AP followed by Hispanics compared to Caucasian populations (28.6 vs 23.1 vs 19.7 per 1000,000 people per year, respectively).^12,13 African Americans, in particular, also have higher rates of mortality, pain, and disability due to chronic pancreatitis (CP) than other racial-ethnic groups.^12,14 Similarly, Hispanics with AP are shown to experience significant delays in receiving care upon their presentation to emergency department.¹⁵ Although previous studies demonstrate health inequities, the majority of AP studies have been completed in non-Hispanic Whites (NHWs), leaving a significant gap in understanding health inequities in AP and its disproportionate effect on minority groups.

Despite the presence of known health inequities, existing research in AP has focused primarily on proximal risk factors, such as toxic-metabolic, genetic and autoimmune etiologies (Figure 1).¹⁶ While identification of these etiologies is important, focusing primarily on these individual factors does not address the underlying structural factors which might contribute to racial/ethnic inequities in AP. Structural inequities produce uneven societal, economic, and political environments which increases risks and susceptibilities resulting in health inequities.¹⁷ These factors often include inordinate exposure to poverty, crime, violence, and limited access to healthcare and healthy foods, leading to increased morbidity. The Centers for Disease Control and Prevention’s (CDC) Social Vulnerability Index (SVI) provides a quantitative estimate of neighborhood social stress levels and can serve as a proxy for one’s overall exposure to neighborhood contextual risk.¹⁸ It groups 15 social factors, including poverty, housing, transportation, household composition, race, ethnicity, and language into four domains including socioeconomic status, household composition, minority status, and housing type. SVI has been used to examine health care inequities in other infectious and chronic diseases.^19,20 To better conceptualize the relationship between social vulnerability and AP risk, we developed a hypothetical model detailing these relationships, and it can be utilized in future studies focusing on health inequities in AP (Figure 2).

Figure 1. Social determinants of health and health inequities in acute pancreatitis.

Proximal risk factors that are commonly assessed in acute pancreatitis, including toxic metabolic, genetic and autoimmune factors are shown above. However, minimal attention is given to social determinants of health variables during evaluation of AP patients, resulting in significant knowledge gaps regarding their impact on AP severity, recurrence and development of exocrine and endocrine complications.

Figure 2. Conceptual model for understanding the contribution of structural violence and diet to AP disparities.

In our inclusive approach, we will i) utilize a socio-ecological model with a focus on StV, and ii) study biologically and physiologically relevant mechanisms in the context of StV to better understand the root causes of AP disparities.

Structural vulnerabilities such as poverty, poor access to grocery stores, and food insecurity are linked with the consumption of a Western-type diet pattern that is high in animal protein and fat and a pro-inflammatory gut microbiome.^21,22 Minority populations are more likely to consume a Western diet which has been linked to higher abundance of sulfidogenic bacteria that produce hydrogen sulfide (H₂S), a pro-inflammatory metabolite.^23–25 To our knowledge, no previous studies have examined the association between social vulnerability and AP using a validated SVI score. In this study, our primary objective was to examine SVI scores in minorities and NHWs with AP. Secondary objectives included determining differences in i) dietary intake, and ii) sulfidogenic bacteria gene copy numbers (gcn) and their end-product H₂S levels between minorities and NHWs with AP.

Materials and Methods

Patient enrollment and determination of AP diagnosis and severity

A total of 54 AP patients were enrolled during hospitalization. AP was confirmed by the presence of at least two of the following criteria: i) abdominal pain consistent with AP(generally described as acute-onset, persistent, epigastric pain often radiating to the back), ii) serum amylase or lipase ≥ 3 times the upper limit of normal, and iii) features of AP on cross-sectional imaging studies. AP disease severity was determined using the Revised Atlanta Classification.²⁶ Given that the primary focus of this study is health inequities in AP, patients with other pancreatic diseases such as CP and pancreatic cancer were not enrolled. CP was identified if a participant reported history of CP, if there was a documentation of CP in the electronic health record or if there was a cross-sectional imaging study showing findings consistent with CP. Exclusion criteria included: age less than 18 years old, chronic pancreatitis, pancreatic cancer, pancreatic trauma, pain that lasting greater than 7 days before the time of enrollment, acute myocardial infarction, decompensated heart failure, and decompensated liver cirrhosis.

Clinical, and diet data collection

The following clinical data points were collected: age, body mass index (BMI), waist circumference, AP etiology (alcohol, gallstones, hypertriglyceridemia or other), AP disease severity (mild, moderately-severe or severe), in-hospital mortality, and Systematic Inflammatory Response Syndrome (SIRS) on presentation. Dietary intake was collected using a validated food frequency questionnaire (FFQ) (VioCare, VioScreen^™ version4) which queries the past 3 months of dietary intake.

Sociodemographic data and social vulnerability index

Race/ethnicity, sex, education, income, and alcohol and tobacco use data were collected using study questionnaires. To quantify social vulnerability, residential addresses were geocoded using Google Maps^™. We then appended CDC SVI scores using census tract numbers. Given the presence of health inequities in AP in both AAs and Hispanics, and their equal representation in our study, SVI scores of these two groups were combined into a single minority group category.

Determination of sulfidogenic bacteria gcn and hydrogen sulfide levels

Stool samples were stored at −80°C following collection during their hospitalization. The FastDNA SPIN Kit for Feces was used for microbial DNA isolation. Using previously validated primers, pan-dissimilatory sulfite reductase A (pan-dsrA) and B. wadsworthia-specific dsrA (dsrA-Bw) gcn were quantified in triplicate with a CFX Real-Time System (Applied Biosystems, Foster City, California).

Statistical analysis

Distribution of normality for continuous variables were assessed using density, Q-Q plots and Shapiro-Wilk testing. Data was analyzed using t-test for normally distributed parametric data, Mann-Whitney U test for non-parametric data involving 2 groups, Kruskal-Wallis test for non-parametric data involving three or more groups and analysis of variance (ANOVA). Statistical analysis was conducted in R (R Core Team 2023) and tables were generated using the package tableone.^27,28

Results

Comparison of clinical and sociodemographic characteristics by race/ethnicity

A total of 54 patients with AP were enrolled into this study during their hospitalization. The AP group included 22 (40.74%) AAs, 22 (40.74%) Hispanics, and 10 (18.52%) NHWs. There were no significant differences in age by race/ethnicity among the patients enrolled (47.0 ± 14.7 vs. 39.9 ± 12.8 years, p=0.167). The percentage of females were similar across the racial ethnic groups (47.7% vs. 40.0%, p=0.601). In addition, there were no significant differences in BMI (31.1 ± 7.67 vs 27.1 ± 7.3, p=0.133), waist circumference (83.3 ± 25.9 vs 94.9 ± 15.0 cm, p= 0.447), current smoking status (29.5% vs 30.0%, p=1) and daily servings of alcohol (0.086, and 0.359 standard drinks, p=1) between the two groups (Table 1).

Table 1.

Baseline characteristics of Minorities and non-Hispanic Whites with Acute Pancreatitis

Variables	Minorities (n=44)	Non-Hispanic Whites (n=10)	P-value
Age (years)(SD)	47.0 ± 14.7	39.9 ± 12.8	0.167
Sex, Female (%)	47.7	40.0	0.601
BMI (kg/m2)(SD)	31.1 ± 7.67	27.1 ± 7.3	0.133
Waist circumference (cm)(SD)	83.3 ± 25.9	94.9 ± 15.0	0.447
Highest education attained (%)			0.386
Did not complete high school	20.9	0.0
High school	32.6	40.0
Some college education/Associates degree	30.2	30.0
Undergraduate or beyond	16.3	30.0
Daily alcohol use, servings	0.85 ± 0.98	1.11 ± 1.46	0.359
Current smoking status (%)	29.5	30.0	1

Abbreviations: NHW, non-Hispanic Whites; BMI, Body Mass Index

Differences in clinical presentation, pancreatitis etiologies and severity

The most common AP etiology was alcohol (42.6%) followed by gallstones (25.9%), other etiologies including idiopathic AP (22.2%), and hypertriglyceridemia-induced AP (9.3%). Approximately 35% of AP patients had a prior history of AP. There were no significant differences in AP etiology or presence of SIRS on presentation across the racial-ethnic groups (p=0.287 and 0.086, respectively) (Table 2). Additionally, the percentage of those with moderately-severe or severe AP was similar amongst the two groups (20.5% vs. 20.4%, p=1), and there was no significant differences in mortality between the two groups (p=1).

Table 2.

Social Vulnerability Index Scores and Clinical Characteristics Grouped by Minority Status in the Acute Pancreatitis Cohort

Variables	Minorities (n=44)	NHW (n=10)	P-value
Pancreatitis Etiology (%)			0.237
Alcohol	40.9	50.0
Gallstone	29.5	10.0
Hypertriglyceridemia	11.4	0
Other etiologies	18.2	40.0
SIRS on presentation (%)	45.5	10.0	0.086
Moderately-severe/severe AP (%)	20.5	20.0	1
Mortality (%)	4.5	0	1
Living below poverty (%)	27.3 ± 15.3	14.8 ± 5.08	0.014
Social Vulnerability Index, CDC	0.779 ± 0.191	0.58 ± 0.22	0.006

Abbreviations: CDC, Centers for Disease Control and Prevention; NHW, Non-Hispanic Whites

Social vulnerability index scores within the acute pancreatitis group

Minorities with AP had a higher percentage of individuals living in poverty compared to NHWs (27.3% ± 15.3 vs. 14.8% ± 5.08, p=0.014). Furthermore, minorities with AP had significantly higher SVI scores compared to NHWs (0.78 ± 0.19 vs. 0.58 ± 0.22, p=0.006) (Table 2).

Differences in dietary intake, sulfidogenic bacteria gcn and hydrogen sulfide concentrations

Minorities with AP, compared to NHWs, had decreased consumption of omega-3 fatty acids such as stearidonic acid (0.0013 vs 0.0046 g, p=0.019) and eicosapentaenoic acid (0.0094 vs 0.020 g, p=0.042), protein from seafood (0.03 vs 0.105 oz, p=0.031), and Vitamin D3 (2.02 vs 3.5 mcg, p=0.025) (Table 3). No statistically significant differences were observed in total dietary fat, total daily calories, carbohydrates, protein, and whole grain intake amongst the two groups (p=0.586, 0.486, 0.217, 0.946, and 0.182, respectively). Minorities with AP had significantly higher gcn of pan-dsrA compared to NHWs (215000 vs 45000 gcn, p=0.033), but there was no significant difference in gcn of dsrA-Bw amongst the two groups (2670 vs 1270 gcn, p=0.245). Additionally, the mean hydrogen sulfide concentrations were similar between the two groups (3.13 vs. 1.99 μM, p=0.226).

Table 3.

Differences in Habitual Dietary Intake Among Minorities and non-Hispanic Whites

Dietary Variables	Minorities (n=44)	NHW (n=10)	P-Value
Total dietary fat (g)	37.3 (9.93)	38.5 (7.61)	0.586
Total polyunsaturated fatty acids (PUFA) (g)	8.14 (2.73)	6.83 (2.01)	0.183
Stearidonic acid (PUFA 18:4) (g)	0.0013 (0.0023)	0.0046 (0.0075)	0.019
Arachidonic acid (PUFA 20:4) (g)	0.058 (0.039)	0.062 (0.042)	0.823
Eicosapentaenoic acid (PUFA 20:5) (g)	0.0094 (0.0092)	0.020 (0.025)	0.042
Docosapentaenoic acid (PUFA 22:5) (g)	0.0076 (0.0043)	0.012 (0.011)	0.072
Docosahexaenoic acid (PUFA 22:6) (g)	0.021 (0.020)	0.039 (0.048)	0.075
Total monounsaturated fatty acids (MUFA)(g)	14.1 (5.13)	13.3 (5.00)	0.694
Total trans-fatty acids (g)	1.21 (0.683)	1.13 (0.315)	0.734
Total saturated fatty acids (g)	11.6 (3.67)	14.5 (4.78)	0.054
Total daily calories (kcal)	2330 (1270)	2660 (1240)	0.486
Total carbohydrate (g)	120 (28.0)	108 (19.8)	0.217
Total protein (g)	34.8 (8.20)	34.5 (8.61)	0.946
Protein from seafood/fish (oz)	0.032 (0.062)	0.105 (0.165)	0.031
Vitamin D3 (mcg)	2.02 (1.78)	3.54 (1.59)	0.025
Whole grains (oz)	1.88 (2.39)	0.789 (0.540)	0.182
Daily servings of alcohol	0.668 (1.04)	1.11 (1.46)	0.294

Discussion

We have shown that minorities with AP have significantly higher SVI scores compared to NHWs. These results are in accord with our previous retrospective study conducted in the Chicago area, which showed that AAs were more likely to live in underserved neighborhoods and have lower income compared to Whites.²⁹ In addition, we demonstrated that minorities with AP consumed lower levels of beneficial nutrients such as polyunsaturated fatty acids (stearidonic and eicosapentenoic acid), protein from seafood and vitamin D3. Lastly, gcn of pan-dsrA that detects all sulfidogenic bacteria are higher in minorities with AP.

Despite major knowledge gaps in health inequities in AP and data showing worse outcomes in minorities, the majority of population-based studies in AP and RAP have been conducted in predominantly NHW cohorts.^4,30–34 In our diverse clinical cohort, we noted a higher percentage of SIRS at presentation in minorities compared to NHWs but it did not reach statistical significance, likely due to a small sample size. AAs with pancreatic diseases have been shown to have higher rates of organ failure compared to the general population (26.1% vs. 20%), 18% higher mortality, and significantly higher rates of SIRS.^12,35–37 Understanding underlying reasons for these disparities is important for development of targeted strategies to close existing health inequities in AP.

AAs with AP experience a significantly higher rate of hospital admission, and Hispanics are likely to wait twice longer in the ED with frequent delays in care (33 vs. 21.9%, p < 0.001) compared with non-Hispanics.^15,38 There were no significant differences among minorities with AP compared to NHWs with AP in regard to time of symptom onset to hospital presentation (43 vs 28 hours, p=0.374), the length of ED stay (7.33 vs 5.34 hours, p=0.205), and care delivery time upon arrival to ED (2.42 vs 2.43 hours, p=0.995) in our study. Our clinical center’s longstanding experience in serving minorities or relatively small sample size in this study may have resulted in similar care metrics for both groups. While factors such as household income and minority status have been shown to correlate with hospital admission rates and mortality in patients with AP, they do not fully explain existing health inequties. Better understanding of the complex environmental factors and key structural barriers that predispose patients to AP, and affect disease progression and recurrence are needed to close existing health inequities. SVI takes into account several of the structural factors and can be repurposed to identify individuals at increased risk for complications from a disease. Numerous studies in other fields indicate that patients with higher SVI scores have poorer cardiac and surgical outcomes.^39–41 For instance, implementation of such SVI scores in cardiology demonstrated the association between higher SVI scores and poor utilization of cardiac rehabilitation following myocardial infarction, and also with delayed prenatal diagnosis of congenital heart disease.^42,43 This has also been reported in the surgical literature, where it has been consistently observed that patients with higher SVI scores demonstrate poorer surgical outcomes, with increased rates of serious complications and mortality.^20,44,45 The adoption of SVI scores as an epidemiological data point may be useful especially during the post-AP recovery period for decreasing RAP risk in the long-term. RAP can further increase already existing disease burden, especially in minority populations. Utilization of SVI can help us to better understand the epidemiological pathogenesis of AP, and help decrease RAP risk when used in conjunction with known modifiable risk factors for AP, such as diet, smoking, and alcohol use.^7,46,47

Dietary factors have also been implicated in the development of AP, including vitamin D levels, which have been shown to inversely correlate with gallstone-AP, AP severity and ICU admissions.^7,48 Our study is the first to show that vitamin D levels are lower specifically in minority groups with AP compared to NHWs with AP. In addition, higher levels of omega-3 fatty acids, such as stearidonic, eicosapentaenoic and docosahexaenoic acid, have been associated with improved mortality, length of hospital stay, infectious complications and organ-failure in previous meta-analyses.^49,50 These results align with our study, which demonstrated lower levels of stearidonic and eicosapentaenoic acid in minorities with AP. Lastly, the gut microbiome has been implicated as a potential risk factor for AP and sulfidogenic bacteria may play a key role by producing hydrogen sulfide, a pro-inflammatory end-product of sulfidogenic bacteria.^5,51,52 Our results showed significantly higher gcn of pan-dsrA in minorities with AP indicating the potential role of gut microbiome in AP inequities. Taken together, our data suggests that part of the existing inequities in minorities with AP can be explained by increased social vulnerability, suboptimal dietary intake, and differences in gut microbiome composition.

There are some limitations to our study. First, the study includes patients from a single academic center in Chicago and has a relatively modest sample size. Completion of this study in one geographic location and its sample size limits the generalizability of these findings. It would be important to perform this analysis in larger populations and in other geographic regions to see if these findings can be validated. Second, due to overwhelming majority of minorities in our cohort, the findings may not adequately represent the population data at large. Third, this is the first study utilizing SVI as a quantifiable social vulnerability data point in AP and confirmation of its use in other pancreatic disease models is needed. Fourth, patients with chronic pancreatitis were not enrolled. We have not excluded patients with a prior history of AP. In our cohort, 19 AP patients (35.2%) had a prior history of AP, and their inclusion provides a more representative AP group. Studies investigating the contribution of social vulnerability to health inequities in chronic pancreatitis are also needed.

To our knowledge, our study is the first to examine the role of SVI in AP in a clinically diverse cohort. Our results demonstrate the presence of higher SVI scores in minorities compared to NHWs, which reiterates that social vulnerability may have an impact in AP progression and recurrence. Similar analyses need to be conducted on large populations across multiple centers to better understand the implications of social vulnerability on AP and its complications. Additional data from these studies can provide further evidence and help us to develop novel preventive and treatment strategies for AP, especially in minority populations.