Racial and Ethnic Differences in Medical Mistrust Among Patients with Non-Decompensated ALD and MetALD

Introduction

Alcohol-associated liver disease (ALD) is a leading cause of liver-related mortality and represents a growing healthcare and economic burden.¹ Early stages of ALD represent an important window for intervention, as alcohol cessation is a major determinant of long-term outcomes. However, patients with ALD often present at an advanced stage, after liver-related complications have developed.¹ Earlier engagement in care may therefore help prevent disease progression and reduce the burden of advanced ALD on healthcare systems.

Medical mistrust, defined as a lack of trust in healthcare systems and providers, disproportionately affects racial and ethnic minority populations and represents an important barrier to engagement in liver care.² This may be especially relevant in ALD, where early engagement in both liver and alcohol use disorder (AUD) care is central to improving outcomes. However, medical mistrust remains poorly characterized in the early stages of ALD, with prior liver-specific work largely focusing on cohorts with cirrhosis of mixed etiologies.

Here, we report interim findings from a well-characterized, prospectively enrolled cohort of patients with non-decompensated ALD and metabolic dysfunction and alcohol-associated liver disease (MetALD). We aimed to characterize medical mistrust in this understudied population and to identify factors associated with it. These findings may help inform strategies to improve engagement in care and guide healthcare efforts aimed at addressing barriers to care in ALD.

Materials and Methods

The Dallas Dionysus Study is a prospective observational cohort of patients with non-decompensated ALD and MetALD enrolled at UT Southwestern Medical Center and Parkland Health in Dallas, Texas. Eligible participants are adults (≥18 years) with (1) hepatic steatosis on imaging, (2) alanine aminotransferase (ALT) above sex-specific thresholds (>33 U/L in men or >25 U/L in women), (3) alcohol intake ≥210 g/week for men or ≥140 g/week for women within the prior 2 years. Patients with a history of hepatic decompensation (jaundice, ascites, variceal bleeding, or hepatic encephalopathy), hepatocellular carcinoma, or a liver disease etiology not related to excessive alcohol consumption were excluded. Liver stiffness was assessed by vibration-controlled transient elastography (VCTE) with a cutoff of ≥12.1 kPa defining advanced fibrosis/cirrhosis.¹ The study protocol (STU-2023-0685) was approved by the UT Southwestern Institutional Review Board. All participants provided written informed consent and the study was conducted in accordance with the Declaration of Helsinki. The present study is a cross-sectional analysis of participants enrolled between September 2024 and November 2025.

At enrollment, participants completed an in-person interviewer-administered survey capturing sociodemographic data, medical and psychiatric comorbidities, and barriers to AUD care (financial/insurance, transportation, stigma, social anxiety or other barriers). Race and ethnicity were self-reported, and patients were classified as non-Hispanic White (White), non-Hispanic Black (Black), Hispanic, or “other” race. Severity of AUD was assessed by the Structured Clinical Interview for DSM-5 (SCID-5).

Medical mistrust was assessed using the validated Group-Based Medical Mistrust Scale (GBMMS), a 12-item Likert-based instrument assessing perceived group-based disparities, suspicion toward healthcare systems and providers, and expectations of mistreatment.³ Items are scored from 1 (strongly disagree) to 5 (strongly agree), with higher scores indicating greater mistrust. A mean score >3 defined high mistrust.

Continuous variables are presented as medians with interquartile ranges (IQRs), and categorical variables as frequencies and percentages. Group differences between participants with high mistrust and those without high mistrust were assessed using Mann–Whitney U-tests for continuous variables and Fisher’s exact tests for categorical variables. Univariable and multivariable linear regression were used to evaluate associations with medical mistrust, using mean GBMMS score as a continuous outcome. Variables associated with GBMMS score at p<0.05 in univariable analysis were entered into the multivariable linear regression model. Adjusted beta coefficients with 95% confidence intervals (CIs) are reported. Multicollinearity was assessed using variance inflation factors, with all values below 5. Statistical significance was defined as p<0.05. Analyses were performed using R (v4.5.1).

Results

As of November 2025, 105 participants were enrolled, of whom 102 (97.1%) completed the GBMMS. The cohort included 58 (55.2%) White, 27 (25.7%) Hispanic, 14 (13.3%) Black, and 6 (5.7%) participants of other race/ethnicity. Median age was 48 years (IQR, 37–60), 83.8% were men, and 51.4% were enrolled through outpatient hepatology clinics. Median liver stiffness was 7.8 kPa (IQR, 5.6–13.7), with 29.5% having advanced fibrosis/cirrhosis. Severe AUD was present in 57.1%, and 40.6% reported at least one barrier to AUD care (Supplementary Table 1).

Overall, 13.7% (14/102) of participants reported high medical mistrust (GBMMS >3). High mistrust was more frequent among Black (35.7%) and Hispanic (23.1%) participants than White participants (5.3%; p=0.006 and p=0.024, respectively). Participants with high mistrust were also younger (median age, 38 vs 50 years; p=0.003), less frequently married (7.1% vs 41.4%; p=0.016), and less frequently enrolled through outpatient hepatology clinics (21.4% vs 55.7%; p=0.022) (Supplementary Table 2).

Median GBMMS scores differed significantly by race/ethnicity, at 1.2 (IQR, 1.0–1.5) among White, 1.7 (IQR, 1.1–2.7) among Hispanic, and 2.1 (IQR, 1.2–3.7) among Black participants (p=0.004) (Figure 1). In univariable analysis, Black race, Hispanic ethnicity, younger age, unmarried status, and lack of hepatology clinic enrollment were significantly associated with higher mistrust scores (all p<0.05). In multivariable analysis, Black participants were more likely to have higher mistrust scores than White participants (adjusted β=0.938; 95% CI, 0.417–1.459; p<0.001), while participants enrolled in outpatient hepatology clinic had significantly lower mistrust scores than those who were not (adjusted β=−0.479; 95% CI, −0.906 to −0.051; p=0.029) (Table 1).

Figure 1.

Mean medical mistrust scores stratified by race and ethnicity. Box-and-whisker plots showing the distribution of medical mistrust scores assessed by the GBMMS, stratified by race and ethnicity. GBMMS scores differed across groups (P=0.004, Kruskal–Wallis test).

Abbreviation: GBMMS, Group-Based Medical Mistrust Scale.

Table 1.

Factors Associated with Medical Mistrust Score

Variable	Univariable β (95% CI)	P value	Adjusted β (95% CI)	P value
Hispanic (reference: White)	0.632 (0.204 to 1.060)	0.004	0.367 (−0.070 to 0.804)	0.099
Black (reference: White)	1.073 (0.534 to 1.613)	<0.001	0.938 (0.417 to 1.459)	<0.001
Age, per year	−0.025 (−0.038 to −0.013)	<0.001	−0.012 (−0.027 to 0.003)	0.106
Married	−0.492 (−0.884 to −0.100)	0.014	0.073 (−0.362 to 0.508)	0.739
Immigrant status	0.362 (−0.048 to 0.772)	0.083	—	—
High school education or less	0.285 (−0.104 to 0.673)	0.149	—	—
Annual household income <$50,000	0.298 (−0.098 to 0.694)	0.138	—	—
Insured	−0.122 (−0.787 to 0.542)	0.715	—	—
Enrolled through hepatology clinic	−0.771 (−1.128 to −0.413)	<0.001	−0.479 (−0.906 to −0.051)	0.029
Advanced fibrosis (liver stiffness ≥12.1 kPa)	−0.211 (−0.632 to 0.209)	0.322	—	—
Any barrier to AUD care	0.390 (−0.021 to 0.801)	0.063	—	—
Low health literacy	0.162 (−0.490 to 0.815)	0.623	—	—

Note: Linear regression was used to evaluate associations with mean GBMMS score as a continuous outcome (range, 1–5). Variables associated with GBMMS score at P<0.05 in univariable analysis were included in the multivariable model: race/ethnicity, age, marital status, and hepatology clinic enrollment (n=97). The Other race/ethnicity group (n=6) was excluded from race-based regression analyses because of its small size and heterogeneous composition. Boldface indicates P<0.05.

Abbreviations: AUD, alcohol use disorder; CI, confidence interval; GBMMS, Group-Based Medical Mistrust Scale.

Item-level responses further characterized racial and ethnic differences in perceived healthcare experiences. Black participants more frequently reported having been treated poorly or unfairly because of their ethnicity (42.9% vs 22.2% of Hispanic and 1.7% of White participants; p<0.001), that healthcare professionals did not take complaints from members of their ethnic group seriously (42.9% vs 14.8% and 3.4%, respectively; p<0.001), and concerns about being treated as “guinea pigs” (21.4% vs 7.4% and 1.7%, respectively; p=0.020) (Supplementary Table 3).

Discussion

Medical mistrust is prevalent in our cohort of patients with non-decompensated ALD and MetALD, indicating that mistrust is already present at the early stages of ALD, when timely intervention may still alter disease trajectories. Consistent with prior work, higher rates of mistrust were observed among Black and Hispanic patients than among White patients.² Racial and ethnic minority populations have been shown to experience discrimination and mistrust at multiple levels, including societal, institutional, and interpersonal, rooted in both historical injustices and personal experiences within the healthcare system.² In contrast, mistrust was lower among patients engaged in outpatient hepatology care, suggesting a potential avenue to mitigate mistrust through patient-centered communication and shared decision-making.

Patient trust is shaped by modifiable healthcare-system and provider factors, including communication, service accessibility, and the healthcare environment.⁴ Mistrust may further compound the barriers that impede co-management of ALD and AUD. Strategies to address these barriers such as co-located AUD treatment, embedded behavioral health clinicians, stigma-reducing clinician training, and telemedicine have been proposed to help sustain engagement.⁵ Culturally tailored approaches may be particularly relevant, as alcohol consumption patterns and motivations for drinking vary by race and ethnicity.⁶

These findings carry implications for healthcare policy, administration, and risk management. Trust-building has been proposed as a people-centered strategy for engaging vulnerable and minority populations and fostering resilient, health-seeking communities.⁷ The high prevalence of mistrust in socially vulnerable populations suggests that unaddressed medical mistrust may represent an actionable target. Identifying populations with greater mistrust may allow healthcare systems and payers to recognize patients at risk of disengagement and target interventions earlier in the ALD care pathway, before progression to more morbid and costly disease. Addressing mistrust may require system- and policy-level action rather than clinician effort alone, including alternative payment models and insurance and national-policy advocacy to expand access to integrated ALD and AUD care.⁵ Incorporating measures of patient trust and experiences of discrimination into quality-improvement and risk-mitigation initiatives may help guide allocation of supportive services and inform administrative policies aimed at reducing the burden of ALD.

Our results should be interpreted within the limitations of our study design. This is an interim analysis of an ongoing, prospectively enrolled cohort, and the sample size limited statistical power for some subgroup comparisons. The cross-sectional nature of the analysis precludes causal inference. Although the cohort was racially and ethnically diverse, recruitment from two sites within a single metropolitan area may limit generalizability.

Conclusion

Medical mistrust is common among patients with non-decompensated ALD, particularly among Black patients and those not engaged in outpatient hepatology care. Recognizing and addressing mistrust early in the disease course is an actionable target that may improve engagement and reduce disease burden both on patients and healthcare systems. These findings support healthcare strategies and policies that promote patient-centered, culturally responsive, and non-stigmatizing care, particularly among populations at greater risk of mistrust. Ongoing longitudinal follow-up in the Dallas Dionysus Study will assess whether mistrust affects adherence to AUD and liver care, longitudinal alcohol use patterns, and long-term liver-related outcomes.

Funding Statement

This work is supported by the American Association for the Study of Liver Diseases (AASLD) Clinical, Translational and Outcomes Research Award (CTORA) and National Institute on Alcohol Abuse and Alcoholism (NIAAA) K23AA031310 grant. The views expressed do not necessarily represent the views of the NIAAA.

Abbreviations

ALD, Alcohol-associated liver disease; ALT, Alanine aminotransferase; AUD, Alcohol use disorder; CI, Confidence interval; DSM-5, Diagnostic and Statistical Manual of Mental Disorders, 5th Edition; GBMMS, Group-Based Medical Mistrust Scale; IQR, Interquartile range; kPa, Kilopascal; MetALD, Metabolic dysfunction and alcohol-associated liver disease; SCID-5, Structured Clinical Interview for DSM-5; U/L, Units per liter; VCTE, Vibration-controlled transient elastography.

Data Sharing Statement

Data supporting the findings of this study are available within the article. Additional study materials and data are available from the corresponding author upon reasonable request.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

Thomas G. Cotter has served as a consultant for Madrigal Pharmaceuticals and Novo Nordisk. Dr. Cotter’s research is supported by (NIAAA) K23AA031310. Amit G. Singal has served as a consultant or on advisory boards for Genentech, AstraZeneca, Eisai, Bayer, Exelixis, Merck, Elevar, Boston Scientific, Sirtex, FujiFilm Medical Sciences, Exact Sciences, Helio Genomics, Curve Biosciences, Glycotest, Abbott, DELFI, Mursla Bio, ImCare, and Universal Dx. Dr. Singal’s research is supported by R01 CA256977. Thomas A. Kerr has served as a consultant for Alexion, M42, and Madrigal Pharmaceuticals and receives authorship royalties from UpToDate. Dr. Kerr receives research support from Alexion and GSK. Dr Nicole Rich reports personal fees from AstraZeneca, personal fees from Eisai, personal fees from Exelixis, personal fees from Genentech, personal fees from Madrigal, outside the submitted work. Dr Mack Mitchell reports personal fees from GlaxoSmithKline, personal fees from Amygdala Neurosciences, grants from Altimmune, outside the submitted work. The authors report no other conflicts of interest in this work.