ABSTRACT
-
Background/Aims
Alcohol-associated liver disease (ALD) is a public health concern. ALD patients often have psychiatric comorbidities, but the effects of psychiatric interventions on ALD are not well-established. This study explores the prognostic impact of psychiatric intervention on ALD within UK Biobank cohort.
-
Methods
This population-based study included 2,417 ALD patients from the UK Biobank cohort. Psychiatric intervention was defined by a consultation with psychiatrists during hospitalization or a history of medication related to alcohol use disorder and psychiatric comorbidities. Survival analysis was conducted, incorporating propensity score matching (PSM), to precisely assess the impact of psychiatric intervention.
-
Results
Among 2,417 ALD patients, those with F10 (mental disorders due to alcohol) codes had poorer survival outcomes. Psychiatric intervention significantly improved the outcomes of both all-cause and liver-related mortality and reduced the incidence of liver cirrhosis. In subgroup or 2-year landmark analyses, psychiatric intervention consistently showed a survival benefit in ALD patients. In the multivariate analysis, psychiatric intervention was identified as a favorable prognostic factor (hazard ratio, 0.780; P=0.002 after PSM).
-
Conclusions
This study demonstrates the favorable effect of psychiatric intervention in ALD patients with psychiatric comorbidities. These findings emphasize the importance of integrated management for ALD patients to address both their medical and psychiatric aspects. Therefore, we suggest the potential benefits of early psychiatric interventions in improving survival outcomes in ALD.
-
Keywords: Alcohol-associated liver disease; Psychiatric comorbidities; Psychiatric medication; Psychiatric consultation
Study Highlights
• Psychiatric intervention improved all-cause and liver-related mortality in ALD patients.
• Reduced mortality is associated with a decreased incidence of liver cirrhosis, attributable to psychiatric intervention.
• The study emphasizes the importance of integrated management for ALD patients, addressing both their medical and psychiatric aspects.
Graphical Abstract
INTRODUCTION
Alcohol-associated liver disease (ALD) is a complicated health concern with high morbidity and mortality rates [
1,
2]. The disease appears in various forms, ranging from alcohol-associated fatty liver to the more severe phenomenon of alcoholic hepatitis and liver cirrhosis (LC) [
2-
4]. The treatment of ALD should involve a comprehensive approach targeting the reduction of alcohol consumption, managing complications of liver disease, and addressing the risk factors associated with the disease progression [
5,
6]. In particular, a reduction in alcohol consumption is associated with significant improvements among patients with alcohol use disorder (AUD) [
7,
8]. Therefore, the recent practice guidelines for ALD emphasize that abstinence is the most crucial factor in determining the prognosis for patients [
5,
6].
Various psychosocial, behavioral, and pharmacological interventions have been explored as strategies in the comprehensive management of AUD [
9]. Disulfiram (an aldehyde dehydrogenase inhibitor), acamprosate (a glutamate system modulator), and naltrexone (an opioid receptor antagonist) are drugs approved by the U.S. Food and Drug Administration to reduce cravings for alcohol [
10-
12]. Recently, other off-label medications, such as topiramate and gabapentin, have been trialed in AUD patients to regulate the release of neurotransmitters in the central nervous system [
13,
14]. However, due to the side effects and limited efficacy, it is recommended that ALD patients receive individual psychosocial interventions, such as motivational interviewing or psychiatric consultations, alongside psychiatric medications [
5,
11,
15]. Unfortunately, for ALD patients who primarily require management for physical symptoms, it is difficult for patients to be exposed early to psychiatric intervention in real-world clinical settings.
The prevalence of psychiatric comorbidities in AUD patients— which include mood disorders (30%), anxiety disorders (19%), substance use disorders (47%), attention-deficit hyperactivity disorder (24%), and psychotic disorder (4%) that are often exacerbated by the medical stress associated with ALD [
16]—is a significant concern. Therefore, when ALD is initially diagnosed, it is essential not only to screen patients for psychiatric comorbidities but also to ensure that their psychosocial status is clearly assessed through early psychiatric consultation [
6,
17]. In fact, the presence of any psychiatric illness in patients with LC exacerbates the all-cause mortality rate [
18]. Otherwise, major depressive disorder is the most common co-occurring mental disorder in patients with AUD, though the effectiveness of antidepressants is still uncertain in terms of patients’ prognosis [
19]. Based on the information above, we propose a hypothesis that the implementation of psychiatric interventions could improve the survival outcomes for patients with ALD.
Here, we first analyzed the impact of psychiatric intervention on the survival prognosis of patients diagnosed with ALD in the UK Biobank cohort. Furthermore, we investigated the influence of psychiatric intervention on the incidence of LC among patients with ALD and aimed to minimize statistical bias by performing various subgroup analyses. We also used methodological techniques like propensity score matching (PSM) to precisely estimate the effects of psychiatric intervention.
MATERIALS AND METHODS
Study design and participants
The UK Biobank is a population-based, prospective cohort study that has enrolled 502,370 participants to date (with a response rate of 5.5%) all aged 40–69 years and residing in the United Kingdom [
20]. This study has been ongoing since 2006 and operates from 22 assessment centers across Scotland, England, and Wales (
http://www.ukbiobank.ac.uk). During the baseline assessment from 2006–2010, all participants completed a questionnaire to provide detailed information about their lifestyles using an electronic device with a touchscreen. Additionally, they regularly provided biological samples, such as blood, urine, and saliva [
20]. Among these, 2,417 inpatients diagnosed with ALD were included in this study (
Fig. 1). ALD was defined as a condition in participants who had the International Classification of Diseases, 10th Revision (ICD-10), code K70 (for alcoholic liver disease), as acquired from the dataset. The development of LC was defined based on the first report of the ICD-10 code K74, which indicates cirrhosis of the liver. We excluded patients with extrahepatic malignancies and those with a history of LC prior to the diagnosis of ALD. The presence of ICD-10 code F10 (for mental and behavioral disorders due to the use of alcohol) for AUD was also extracted.
The UK Biobank study received approval from the North West Multi-Centre Research Ethics Committee, and all participants provided signed written informed consent forms (
http://www.ukbiobank.ac.uk). This research was performed using the UK Biobank dataset under application number 85156.
Psychiatric intervention was defined as either consulting with a psychiatrist during hospitalization or having a history of medication use for AUD and associated psychiatric comorbidities.
Supplementary Table 1 includes detailed information on psychiatric medications based on the recent AUD guidelines [
11]. Among the 497 patients who underwent psychiatric intervention, 345 (69.4%) received psychiatric medication, and the medications were categorized into anti-craving drugs (10.1%), antidepressants (54.5%), and anxiolytics (10.5%). Among the anti-craving drugs, naltrexone was not listed in the UK Biobank medication codes and was therefore excluded from the analysis. The codes for psychiatric consultations in the UK Biobank used in this study are summarized in
Supplementary Table 2.
The significance of continuous factors (mean±standard deviation values) was analyzed using Wilcoxon rank-sum tests, while Pearson’s chi-squared or Fisher’s exact tests were applied for categorical variables (number, %) to evaluate their significance. Overall survival and liver-related survival were analyzed from the time of ALD diagnosis to allcause death and liver-related death, respectively. The logrank test was used in survival analyses to identify significant differences in outcomes between various patient groups. Cox regression analyses were used for univariate and multivariate assessments to study the effects of various factors on patient survival, accounting for potential confounders. PSM was conducted using the nearestneighbor method with a 1:2 matching ratio ensuring a balanced comparison between the groups. Time zero for the survival and landmark analyses was defined as the date of diagnosis of ALD, identified using the ICD-10 code K70. To address significant differences in baseline demographic and clinical characteristics, we included variables such as age, sex, presence of F10, aspartate aminotransferase (AST), gamma-glutamyl transferase (GGT), creatinine, alcohol intake frequency, and physical activity frequency in the propensity score estimation. R (version 4.2.3; R Foundation for Statistical Computing, Vienna, Austria) was the primary tool used for statistical analyses.
RESULTS
Baseline demographic and clinical characteristics of patients
Table 1 outlines the baseline demographic and clinical characteristics of ALD patients before PSM. Our cohort consisted of 2,417 patients diagnosed with ALD, divided into two groups of 1,920 patients who did not receive psychiatric intervention and 497 who did, respectively. The mean age was 56.78±7.51 years, with those receiving psychiatric care being slightly younger (
P<0.001). Out of 2,417 patients, 1,630 (67.4%) had ICD-10 code F10, with a significantly higher prevalence (82.9%) of this code in those receiving psychiatric intervention (
P<0.001). In terms of liver function test results, the intervention group had lower mean AST and GGT levels. Alcohol consumption was a daily habit for 48.2% of the total cohort, with significant variations between the two groups (
P<0.001). Regular physical activity was reported by 45.6% of patients, with a statistically significant difference in frequency observed between the groups (
P=0.028). In the psychiatric intervention group, 69.4% of patients were administered psychiatric medications and 44.5% had undergone consultations.
Impact of psychiatric intervention on ALD
The presence of alcohol-related psychiatric illness is verified by identifying the F10 code in their medical records. Therefore, we first analyzed the impact of this code on ALD patients’ survival, as its presence is associated with longer hospital stays and increased mortality rates [
21]. As expected, ALD patients with the F10 code experienced a remarkably higher rate of all-cause and liver-related mortality (
Supplementary Fig. 1A and
1B). Interestingly, there has been a noticeable increase in the incidence of LC among patients with the F10 disease code, and it can be inferred that this is associated with a worsening mortality rate (
Supplementary Fig. 1C). The finding that alcohol-related psychiatric problems have a decisive impact on the prognosis of ALD patients has led us to propose the hypothesis that psychiatric intervention could improve the survival of these patients. In Figure 2A–C, patients who received psychiatric intervention experienced a notable reduction in all-cause mortality (
P<0.0001) (
Fig. 2A), liver-related mortality (
P=0.0003) (
Fig. 2B), and the incidence of LC (
P=0.0005) (
Fig. 2C). In the univariate analysis, psychiatric intervention was associated with improved survival (HR 0.700; 95% CI 0.607–0.808;
P<0.001). In the multivariate analysis, psychiatric intervention remained a significant factor, with an HR of 0.727 (95% CI 0.618–0.856;
P<0.001), indicating a 27% reduction in the risk of death for patients receiving psychiatric care (
Supplementary Table 3).
PSM was utilized due to notable imbalances in baseline characteristics between two groups, aiming to ensure a balanced comparison and enhance the study’s accuracy (
Table 2). All clinical parameters were not significantly different, indicating that PSM effectively balanced these characteristics. After PSM, the presence of the ICD-10 code F10 significantly impacted the prognosis negatively (
Supplementary Fig. 2). The Kaplan–Meier curves consistently demonstrate the existence of significant improvements in all-cause mortality (
P=0.0008), liver-related mortality (
P=0.0008), and the incidence of LC (
P=0.0007) for patients who received psychiatric intervention (
Fig. 2D–
F). Age, presence of F10, AST, GGT, albumin levels, reduced alcohol consumption, and psychiatric intervention were the significant factors for overall survival in the multivariate analysis (
Table 3). The median survival time of the psychiatric intervention group in
Table 4 remained significant after PSM except for consultation only group: consultation only (12.60 years for the control and 13.95 years for the intervention group;
P=0.193), medication only (12.02 years for the control and 17.26 years for the intervention group;
P<0.001) or both (11.55 years for the control and 15.00 years for the intervention group;
P=0.001).
Next, we conducted a subgroup analysis of ALD patients with the F10 code in their medical records. The survival analysis conducted on patients with F10 code revealed a marked reduction in all-cause mortality, liver-related mortality, and the incidence of LC for those who underwent psychiatric intervention (
Supplementary Fig. 3). This finding was corroborated by the univariate and multivariate analysis, where psychiatric intervention remained a strong predictor of enhanced survival, with an HR of 0.709 (
P<0.001) (
Supplementary Table 4). To reduce the lead time bias where psychiatric illness is diagnosed before ALD, leading to prior exposure to psychiatric medication or consultation, we re-analyzed the patients who had the F10 code entered in their medical records after a diagnosis of ALD. As a result, we observed a remarkable decrease in mortality among patients receiving psychiatric intervention (
Fig. 3A–
C). In the multivariate analysis, psychiatric intervention, along with reduced alcohol consumption and higher albumin levels, were still identified as significant factors (
Supplementary Table 5). To make the evidence more concrete, we conducted the same analysis after performing PSM on the cohort used in Figure 3A–C, and psychiatric intervention showed a similar effect on survival outcomes (
Fig. 3D–
F and
Supplementary Table 6). In the median survival time analysis (
Table 4), all ALD patients with the F10 code showed significantly longer survival time when treated with consultation only, medication only, or both (all
P<0.001). Additionally, patients diagnosed with ALD and subsequently diagnosed with F10 had significantly longer survival times both before and after PSM only when treated with a combined psychiatric intervention of consultation and medication.
Next, a 2-year landmark analysis was conducted to account for the dynamic nature of patient characteristics and other covariates over time, similar to what was done in other prospective research of UK Biobank cohorts [
22,
23]. This approach allowed for a focused examination of patient outcomes at a specific time point, mitigating the potential confounding effects of time-varying factors. As such, this approach ensured an unbiased estimation of the psychiatric intervention’s impact and minimized the effects of reverse causality on ALD patients. The results were consistent with previous subgroup analyses, showing that psychiatric treatment significantly reduced mortality and the incidence of LC, while maintaining a significant protective effect in multivariate analysis (
Supplementary Fig. 4,
Supplementary Table 7, and
Table 4). The prognostic effect of psychiatric intervention on liver-related mortality and the incidence of LC in the cohorts used for all the above analyses is summarized in
Supplementary Tables 8 and
9.
We conducted additional subgroup analyses focusing on the impact of different psychiatric medications, as well as the influence of gender on prognosis. The analysis indicated that the use of antidepressants effectively improved overall survival in the total cohort both before and after PSM (
Supplementary Table 10). Anxiolytic drugs did not show a significant impact on survival due to the small sample size, showing only a trend towards significance. Gender-specific analysis revealed that psychiatric intervention was associated with a significant reduction in mortality in males before and after PSM (
Supplementary Table 11). In females, the intervention improved survival significantly before PSM, but did not show significance after PSM. The effect of psychiatric intervention was relatively less in females compared to males and was only significant in some cohorts. These analyses confirm that the prognostic benefits of psychiatric interventions vary by gender, highlighting the importance of gender-specific approaches in treatment.
DISCUSSION
Despite ongoing multifaceted efforts to reduce the social and economic burdens caused by ALD, the mortality rate from ALD has continued to rise recently, highlighting an urgent need for therapeutic interventions using novel approaches [
24]. In the present study, we unveiled a significant impact of psychiatric intervention that led to improved survival outcomes of patients with ALD. Our findings, derived from a comprehensive investigation of a UK Biobank cohort, underscore a marked reduction in all-cause and liverrelated mortality as well as a decreased incidence of LC among ALD patients receiving psychiatric care. The statistical rigor of our research is anchored in PSM, DR estimation, various subgroup analyses, and a 2-year landmark analysis, ensuring a balanced and precise evaluation of psychiatric intervention effects. To the best of our knowledge, this is the first study to explore the impact of psychiatric management on ALD patients using UK Biobank cohort data, unveiling new insights into ALD management.
Abstinence from alcohol emerges as the cornerstone of therapy, playing a pivotal role in mitigating disease progression and reducing mortality [
25,
26]. Even modest alcohol intake is associated with increased mortality among individuals with elevated alanine transaminase levels, underscoring the imperative of complete abstinence [
27]. Consistent with this finding, a significant benefit of alcohol abstinence is a reduced risk of hepatocellular carcinoma among patients with alcohol-related LC [
28]. These findings proved that abstinence not only prevents liver damage but also reduces the risk of life-threatening complications such as cancer. A German cohort study demonstrated that healthy alcohol abstainers without risk factors do not exhibit greater mortality, cautioning against generalized recommendations for alcohol consumption [
29]. Therefore, psychiatric interventions, customized to the individual’s unique psychological and behavioral profile, can target the underlying aspects of alcohol dependence, creating a comprehensive and enduring pathway to abstinence.
Several previous studies have investigated the role and impact of AUD medication on prognosis among ALD patients. Patients with both AUD and LC exhibit a significant reduction in hepatic decompensation when undergoing behavioral and pharmacological treatments, highlighting the essential role of comprehensive AUD therapy in enhancing clinical outcomes [
30]. A recent study has identified that medical addiction therapy for AUD is significantly associated with a lower risk of developing ALD and a reduced incidence of hepatic decompensation [
31]. On the other hand, maintaining alcohol abstinence showed remarkable benefits for patients with alcohol-related LC at every stage of portal hypertension, thereby improving their overall prognosis [
32]. Compared to the above reports, the present study incorporates a more comprehensive approach by integrating the concept of psychiatric consultation during hospitalization and assessing the detailed effect of other psychiatric medications in ALD patients. The inclusion of antidepressants and anxiolytics, alongside conventional AUD medications, allows us to explore the broader effects of psychiatric intervention on prognostic outcomes. Moreover, the longitudinal design of our study enables a thorough analysis of the long-term effects of combined psychiatric and pharmacological interventions on the incidence of LC. These findings contribute to a more nuanced understanding of the complex characteristics of psychiatric interventions, proposing an integrated approach for enhanced patient outcomes.
Considering the nature of AUD, the presence of various psychiatric comorbidities, including mood and anxiety disorders, is inevitable. Particularly, mood disorders are the most common psychiatric disorders among people with AUD, and their co-occurrence is associated with greater severity and worse prognosis for both disorders [
16]. In fact, more than 30% of people in treatment for AUD met criteria for major depressive disorder, and depression interacts significantly with heavy alcohol consumption, leading to an increased risk of all-cause mortality [
16,
33]. Furthermore, the manifestation of alcohol-induced psychotic disorder and delirium has been associated with poor outcomes in patients, serving as another compelling piece of evidence supporting the need for psychiatric intervention in ALD patients [
34]. The robust results from our study consistently identified the crucial role of medications in managing psychiatric comorbidities, demonstrating a significant improvement in both all-cause and liver-related mortality rates as well as the incidence of LC. This evidence emphasizes the need to prioritize psychiatric care during the early stages of ALD management, aiming to improve patient outcomes and their quality of life.
Generally, medications for AUD or psychiatric comorbidities are often prescribed by psychiatrists rather than primary care physicians, so we included the psychiatric consultation during hospitalization as a variable for survival analyses. In a recent study, inpatient psychiatric consultation for addiction played a pivotal role during the hospitalization period and was associated with a reduction in 90-day all-cause mortality among patients with ALD [
35]. Moreover, early psychiatric intervention through integrated management is crucial for ALD patients, as it helps stabilize their mental health, manage alcohol use, and support long-term recovery [
36]. Integrated care of ALD involves a combination of laboratory assessments, pharmacological and behavioral therapies, and follow-up care by specialists [
37]. In our study, patients who received at least one psychiatric consultation, medication related to AUD, or medication associated with psychiatric comorbidities during the observation period were included in the psychiatric intervention group. Therefore, there is a high likelihood that these individuals received care from psychiatric specialists, leading to improved survival outcomes.
However, it is important to note that in clinical settings, clinicians sometimes hesitate to prescribe AUD-related medications to ALD patients due to concerns about potential side effects. These side effects can include hepatotoxicity, gastrointestinal disturbances, and neurological symptoms, which may complicate the management of ALD patients already at risk of liver dysfunction [
10,
11]. This hesitancy is particularly pronounced when dealing with medications like disulfiram, which can cause severe adverse reactions if alcohol is consumed concurrently [
10]. The potential for these adverse effects underscores the necessity for careful patient selection and monitoring when implementing pharmacotherapy for AUD in ALD patients . This reality highlights the critical role of psychiatric consultations and individualized treatment plans that weigh the benefits of psychiatric interventions against potential risks. It also underscores the importance of further research to develop safer and more effective therapeutic options for this vulnerable population .
One critical challenge of our study is the potential for immortal time bias, which occurs when patients who die before receiving psychiatric intervention are classified into the no intervention group, potentially overestimating the effect size. To address this, we re-analyzed patients with the F10 code (mental and behavioral disorders due to alcohol use) entered after their ALD diagnosis, ensuring psychiatric intervention occurred post-ALD diagnosis. This approach demonstrated improved survival outcomes both before and after PSM (
Fig. 3). Additionally, we employed a two-year landmark analysis, defining time zero as the ALD diagnosis date (ICD-10 code K70), to capture the majority of psychiatric interventions and robustly assess their impact on survival. This method aligns with other similar UK Biobank studies and mitigates confounding effects of time-varying factors [
22,
23]. Our results remained consistent across different landmark points (1, 2, and 3 years; data not shown), confirming the robustness of our findings and demonstrating that psychiatric treatment significantly reduced mortality and the incidence of LC. Future research should aim to collect more precise data on the timing and duration of psychiatric interventions to further minimize time bias.
There are several limitations to discuss in this study. First, the psychiatric intervention group included patients exposed to at least one consultation or medication during the observational period, but specific dosages, start dates, and durations of the medications remain unclear. Additionally, the specific psychiatric assessments conducted by specialists during the consultations of ALD patients remain unidentified. The inclusion of antidepressants and anxiolytic drugs in the psychiatric medication category raises another concern, as it is uncertain which specific psychiatric comorbidities were being treated with these medications. Our results suggest a potential reduction in alcohol consumption among ALD patients receiving psychiatric intervention, inferred from the decreased incidence of LC and mortality rates. However, there is a lack of concrete data on the actual changes in alcohol-consumption patterns following the intervention. Therefore, we have emphasized the importance of future research using more detailed longitudinal data to capture the frequency, duration, and adherence to psychiatric interventions. Such studies would provide a more comprehensive understanding of how these factors influence the prognosis of ALD and help validate our findings.
Incorporating the insights from this study, we conclude with an emphasis on the tangible benefits of psychiatric intervention in the management of ALD. The reduction in both all-cause and liver-related mortality underscores the pivotal role of psychiatric care in enhancing survival outcomes. The individualized approach, adapting to each patient’s distinct mental health status, is not just a recommendation but a necessity, reflecting the complexity and multifaceted nature of ALD treatment. In conclusion, our findings support a shift in ALD management, emphasizing the early integration of psychiatric care as a core component to address both physical and mental health in patients with ALD.
ACKNOWLEDGMENTS
This study was carried out using resources provided by the UK Biobank. We extend our thanks to the UK Biobank participants for their invaluable contribution. This work was supported by the National Research Foundation of Korea (NRF) grant (RS-2023-00208767 to S.H.B.), the Medical Scientist Training Program (N11220102 to S. K.) funded by the Korea government (Ministry of Science, ICT, and Future Planning), the Brain Research Program through the NRF funded by the Ministry of Science & ICT (Grant no. NRF-2022M3E5E8081200 to B.J.), and the Basic Medical Science Facilitation Program through the Catholic Medical Center of the Catholic University of Korea funded by the Catholic Education Foundation (to H.K.L).
FOOTNOTES
-
Authors’ contribution
S.H.B., H.K.L., and B.J.: general conception, design of the study, and drafting of the manuscript. K.Y. and S.K.: detailed design of the study, writing of the manuscript, statistical analysis, and interpretation of data. H.Y. and S.W.: data collection and supporting statistical analysis. All authors were responsible for the critical revision of the manuscript and approved the manuscript to be published.
-
Conflicts of Interest
The authors have no conflicts to disclose.
SUPPLEMENTAL MATERIAL
Supplementary material is available at Clinical and Molecular Hepatology website (
http://www.e-cmh.org).
Supplementary Figure 1.
Kaplan–Meier curves present the overall survival rates and incidence of liver cirrhosis in the total cohort (n=2,417) before propensity score matching. (A) The impact of mental and behavioral disorders due to use of alcohol (F10) on all-cause mortality (P<0.0001), (B) liver-related mortality (P<0.0001), and (C) incidence of liver cirrhosis (P=0.0015). OS, overall survival; LC, liver cirrhosis; PSM, propensity score matching.
Supplementary Figure 2.
Kaplan–Meier curves show the overall survivals and occurrence of liver cirrhosis among 1,491 patients with alcohol-associated liver disease after propensity score matching. (A) The effect of mental and behavioral disorders due to use of alcohol (F10) on all-cause mortality (P<0.0001), (B) liver-related mortality (P=0.0017), and (C) incidence of liver cirrhosis (P=0.047). OS, overall survival; LC, liver cirrhosis; PSM, propensity score matching.
Supplementary Figure 3.
Survival analysis in patients with mental and behavioral disorders due to use of alcohol (F10) (n=1,630). (A) All-cause mortality (P<0.0001), (B) liver-related mortality (P<0.0001), and (C) incidence of liver cirrhosis (P=0.0002) according to the psychiatric intervention. OS, overall survival; LC, liver cirrhosis.
Supplementary Figure 4.
2-year landmark analyses investigate the overall survivals and incidence of liver cirrhosis using Kaplan–Meier curves. (A) The impact of psychiatric intervention on all-cause mortality (P=0.0004), (B) liver-related mortality (P=0.035), and (C) incidence of liver cirrhosis (P=0.0011). OS, overall survival; LC, liver cirrhosis.
Figure 1.A research flow for statistical analysis. ICD-10, International Classification of Diseases, 10th Revision.
Figure 2.Kaplan–Meier curves present the overall survival rates and the incidence rate of liver cirrhosis in the total cohort. Before PSM: (A) The impact of psychiatric intervention on all-cause mortality (P<0.0001), (B) liver-related mortality (P=0.0003), and (C) the incidence of liver cirrhosis (P=0.0005). After PSM: (D) The effect of psychiatric intervention on all-cause mortality (P=0.0008), (E) liver-related mortality (P=0.0008), and (F) incidence of liver cirrhosis (P=0.0007). LC, liver cirrhosis; PSM, propensity score matching.
Figure 3.Survival analysis in patients with mental and behavioral disorders due to the use of alcohol (F10) following the diagnosis of alcohol-associated liver disease. Before PSM: (A) All-cause mortality (P=0.0014), (B) liver-related mortality (P=0.014), and (C) the incidence of liver cirrhosis (P=0.017). After PSM: (D) All-cause mortality (P=0.047), (E) liver-related mortality (P=0.027), and (F) incidence of liver cirrhosis (P=0.037) according to the psychiatric intervention. LC, liver cirrhosis; PSM, propensity score matching; ALD, alcohol-associated liver disease.
Table 1.Baseline demographic and clinical characteristics before propensity score matching
Table 1.
|
Characteristic |
Overall, (n=2,417) |
Psychiatric intervention
|
P-value*
|
SMD |
|
No (n=1,920) |
Yes (n=497) |
|
Age at recruitment (year) |
56.78±7.51 |
57.41±7.43 |
54.35±7.33 |
<0.001 |
0.41 |
|
Sex (male) |
1,859 (76.9) |
1,497 (78.0) |
362 (72.8) |
0.016 |
0.12 |
|
Presence of F10 |
1,630 (67.4) |
1,218 (63.4) |
412 (82.9) |
<0.001 |
–0.45 |
|
Body mass index (kg/m2) |
28.73±5.38 |
28.83±5.32 |
28.36±5.57 |
0.057 |
0.09 |
|
Protein (g/L) |
73.89±4.96 |
74.00±4.93 |
73.49±4.70 |
0.034 |
0.11 |
|
Albumin (g/L) |
44.41±3.69 |
44.37±3.74 |
44.60±3.28 |
0.508 |
–0.06 |
|
AST (U/L) |
49.73±40.67 |
50.73±40.68 |
45.36±39.28 |
<0.001 |
0.12 |
|
GGT (U/L) |
157.58±187.49 |
160.66±188.05 |
141.11±178.39 |
<0.001 |
0.08 |
|
Cr (µmol/L) |
72.42±24.95 |
73.40±26.69 |
68.78±15.95 |
<0.001 |
0.21 |
|
HbA1c (mmol/mol) |
37.95±10.32 |
37.84±10.21 |
38.35±10.74 |
0.375 |
–0.05 |
|
LDL (mmol/L) |
3.32±0.98 |
3.32±0.97 |
3.31±0.99 |
0.902 |
0.01 |
|
Coffee intake |
|
|
|
0.209 |
0.06 |
|
≥3 cups/day |
713 (29.5) |
555 (28.9) |
158 (31.8) |
|
|
|
<3 cups/day |
1,704 (70.5) |
1,365 (71.1) |
339 (68.2) |
|
|
|
Alcohol intake frequency |
|
|
|
<0.001 |
–0.27 |
|
Daily or almost daily |
1,164 (48.2) |
978 (50.9) |
186 (37.4) |
|
|
|
Less than 3 times a week |
1,253 (51.8) |
942 (49.1) |
311 (62.6) |
|
|
|
Physical activity frequency |
|
|
|
0.028 |
–0.11 |
|
≥4 times a week |
1,103 (45.6) |
898 (46.8) |
205 (41.2) |
|
|
|
<4 times a week |
1,314 (54.4) |
1,022 (53.2) |
292 (58.8) |
|
|
|
Psychiatric medication |
345 (14.3) |
0 (0.0) |
345 (69.4) |
<0.001 |
–2.10 |
|
Psychiatric consultation |
221 (9.1) |
0 (0.0) |
221 (44.5) |
<0.001 |
–1.31 |
Table 2.Baseline demographic and clinical characteristics after propensity score matching
Table 2.
|
Characteristic |
Overall, (n=1,491) |
Psychiatric intervention
|
P-value*
|
SMD |
|
No (n=994) |
Yes (n=497) |
|
Age at recruitment (year) |
54.74±7.41 |
54.93±7.45 |
54.35±7.33 |
0.125 |
0.08 |
|
Sex (male) |
1,095 (73.4) |
733 (73.7) |
362 (72.8) |
0.709 |
–0.09 |
|
Presence of F10 |
1,299 (82.4) |
817 (82.2) |
412 (82.9) |
0.736 |
–0.02 |
|
Body mass index (kg/m2) |
28.37±5.52 |
28.38±5.49 |
28.36±5.57 |
0.945 |
<0.01 |
|
Protein (g/L) |
73.70±4.64 |
73.81±4.60 |
73.49±4.70 |
0.240 |
0.07 |
|
Albumin (g/L) |
44.44±3.40 |
44.36±3.45 |
44.60±3.28 |
0.334 |
–0.07 |
|
AST (U/L) |
45.87±38.65 |
46.47±38.34 |
45.36±39.28 |
0.250 |
0.05 |
|
GGT (U/L) |
142.19±175.11 |
142.74±173.54 |
141.11±178.39 |
0.204 |
0.01 |
|
Cr (µmol/L) |
68.97±16.40 |
69.07±16.62 |
68.78±15.95 |
0.609 |
0.02 |
|
HbA1c (mmol/mol) |
37.64±10.15 |
37.29±9.83 |
38.35±10.74 |
0.096 |
–0.10 |
|
LDL (mmol/L) |
3.34±1.00 |
3.35±1.00 |
3.31±0.99 |
0.544 |
0.04 |
|
Coffee intake |
|
|
|
0.525 |
0.03 |
|
≥3 cups/day |
458 (30.7) |
300 (30.2) |
158 (31.8) |
|
|
|
<3 cups/day |
1,033 (69.3) |
694 (69.8) |
339 (68.2) |
|
|
|
Alcohol intake frequency |
|
|
|
0.217 |
–0.06 |
|
Daily or almost daily |
591 (39.6) |
405 (40.7) |
186 (37.4) |
|
|
|
Less than 3 times a week |
900 (60.4) |
589 (59.3) |
311 (62.6) |
|
|
|
Physical activity frequency |
|
|
|
0.095 |
–0.09 |
|
≥4 times a week |
667 (44.7) |
462 (46.5) |
205 (41.2) |
|
|
|
<4 times a week |
824 (55.3) |
532 (53.5) |
292 (58.8) |
|
|
|
Psychiatric medication |
345 (23.1) |
0 (0.0) |
345 (69.4) |
<0.001 |
–2.10 |
|
Psychiatric consultation |
221 (14.8) |
0 (0.0) |
221 (44.5) |
<0.001 |
–1.31 |
Table 3.Univariate and multivariate analysis for overall survival after propensity score matching
Table 3.
|
Characteristic |
Univariate
|
Multivariate
|
|
HR |
95% CI |
P-value |
HR |
95% CI |
P-value |
|
Age at recruitment |
1.022 |
1.012, 1.033 |
<0.001 |
1.025 |
1.015, 1.036 |
<0.001 |
|
Sex (male) |
1.221 |
1.023, 1.456 |
0.027 |
1.097 |
0.917, 1.312 |
0.312 |
|
Presence of F10 |
1.612 |
1.294, 2.007 |
<0.001 |
1.573 |
1.258, 1.966 |
<0.001 |
|
Body mass index (kg/m2) |
0.998 |
0.983, 1.012 |
0.754 |
|
|
|
|
Protein (g/L) |
1.017 |
1.001, 1.034 |
0.034 |
1.011 |
0.994, 1.028 |
0.220 |
|
Albumin (g/L) |
0.948 |
0.929, 0.967 |
<0.001 |
0.960 |
0.940, 0.981 |
<0.001 |
|
AST (U/L) |
1.005 |
1.004, 1.007 |
<0.001 |
1.003 |
1.001, 1.005 |
0.001 |
|
GGT (U/L) |
1.001 |
1.001, 1.002 |
<0.001 |
1.001 |
1.000, 1.001 |
0.003 |
|
Cr (µmol/L) |
0.998 |
0.993, 1.002 |
0.331 |
|
|
|
|
HbA1c (mmol/mol) |
1.005 |
0.998, 1.001 |
0.171 |
|
|
|
|
LDL (mmol/L) |
0.927 |
0.856, 1.003 |
0.060 |
|
|
|
|
Coffee intake (<3 cups/day) |
1.004 |
0.930, 1.083 |
0.918 |
|
|
|
|
Alcohol intake frequency (less than 3 times a week) |
0.778 |
0.719, 0.842 |
<0.001 |
0.806 |
0.688, 0.944 |
0.007 |
|
Physical activity frequency (<4 times a week) |
1.098 |
0.982, 1.228 |
0.102 |
|
|
|
|
Psychiatric medication |
0.723 |
0.604, 0.865 |
<0.001 |
|
|
|
|
Psychiatric consultation |
0.875 |
0.717, 1.069 |
0.193 |
|
|
|
|
Psychiatric intervention (medication or consultation) |
0.765 |
0.654, 0.895 |
0.001 |
0.780 |
0.665, 0.913 |
0.002 |
Table 4.Median survival time of psychiatric consultation, medication, and both interventions for each group
Table 4.
|
Median survival time, years (95% CI)
|
P-value |
|
Control group |
Intervention group |
|
Total cohort before propensity score matching |
|
|
|
|
Consultation |
11.06 (9.86, 12.02) |
13.95 (12.19, 17.56) |
0.014 |
|
Medication |
10.52 (9.63, 11.51) |
17.26 (13.70, 20.90) |
<0.001 |
|
Consultation+Medication |
10.08 (9.29, 11.29) |
15.00 (13.29, 17.56) |
<0.001 |
|
Total cohort after propensity score matching |
|
|
|
|
Consultation |
12.60 (11.47, 13.84) |
13.95 (12.19, 17.56) |
0.193 |
|
Medication |
12.02 (10.69, 13.04) |
17.26 (13.70, 20.90) |
<0.001 |
|
Consultation+Medication |
11.55 (10.19, 12.96) |
15.00 (13.29, 17.56) |
0.001 |
|
ALD patients with F10 |
|
|
|
|
Consultation |
9.12 (8.46, 10.10) |
13.53 (12.05, 17.26) |
<0.001 |
|
Medication |
9.19 (8.33, 10.10) |
14.84 (12.27, 18.52) |
<0.001 |
|
Consultation+Medication |
8.57 (7.78, 9.59) |
13.75 (12.31, 17.26) |
<0.001 |
|
F10 patients after diagnosis of ALD before propensity score matching |
|
Consultation |
11.74 (10.43, 12.89) |
16.87 (12.70, NA*) |
0.077 |
|
Medication |
11.41 (9.86, 12.69) |
17.42 (13.70, NA*) |
0.003 |
|
Consultation+Medication |
11.20 (9.53, 12.26) |
17.26 (13.53, 25.62) |
0.001 |
|
F10 patients after diagnosis of ALD after propensity score matching |
|
Consultation |
14.31 (12.26, 16.45) |
16.87 (12.70, NA*) |
0.359 |
|
Medication |
13.04 (11.74, 15.84) |
17.42 (13.70, NA*) |
0.055 |
|
Consultation+Medication |
12.84 (11.41, 15.84) |
17.26 (13.53, 25.62) |
0.048 |
|
2-year landmark analyses |
|
|
|
|
Consultation |
15.71 (14.61, 17.42) |
17.26 (14.00, 20.90) |
0.274 |
|
Medication |
14.97 (13.72, 16.04) |
20.90 (17.53, NA*) |
<0.001 |
|
Consultation+Medication |
14.84 (13.59, 16.04) |
18.52 (16.87, 24.41) |
<0.001 |
Abbreviations
alcohol-associated liver disease
aspartate aminotransferase
gamma-glutamyl transferase
propensity score matching
REFERENCES
- 1. Seitz HK, Bataller R, Cortez-Pinto H, Gao B, Gual A, Lackner C, et al. Alcoholic liver disease. Nat Rev Dis Primers 2018;4:16.
- 2. Singal AK, Mathurin P. Diagnosis and treatment of alcoholassociated liver disease: a review. JAMA 2021;326:165-176.
- 3. Wu X, Fan X, Miyata T, Kim A, Cajigas-Du Ross CK, Ray S, et al. Recent advances in understanding of pathogenesis of alcohol-associated liver disease. Annu Rev Pathol 2023;18:411-438.
- 4. Gómez-Medina C, Melo L, Martí-Aguado D, Bataller R. Subclinical versus advanced forms of alcohol-related liver disease: need for early detection. Clin Mol Hepatol 2023;29:1-15.
- 5. Crabb DW, Im GY, Szabo G, Mellinger JL, Lucey MR. Diagnosis and treatment of alcohol-associated liver diseases: 2019 practice guidance from the American Association for the Study of Liver Diseases. Hepatology 2020;71:306-333.
- 6. European Association for the Study of the Liver. EASL clinical practice guidelines: management of alcohol-related liver disease. J Hepatol 2018;69:154-181.
- 7. Roerecke M, Gual A, Rehm J. Reduction of alcohol consumption and subsequent mortality in alcohol use disorders: systematic review and meta-analyses. J Clin Psychiatry 2013;74:e1181-e1189.
- 8. Laramée P, Leonard S, Buchanan-Hughes A, Warnakula S, Daeppen JB, Rehm J. Risk of all-cause mortality in alcoholdependent individuals: a systematic literature review and meta-analysis. EBioMedicine 2015;2:1394-1404.
- 9. Ramkissoon R, Shah VH. Alcohol use disorder and alcoholassociated liver disease. Alcohol Res 2022;42:13.
- 10. Kranzler HR, Soyka M. Diagnosis and pharmacotherapy of alcohol use disorder: a review. JAMA 2018;320:815-824.
- 11. Reus VI, Fochtmann LJ, Bukstein O, Eyler AE, Hilty DM, Horvitz-Lennon M, et al. The American Psychiatric Association practice guideline for the pharmacological treatment of patients with alcohol use disorder. Am J Psychiatry 2018;175:86-90.
- 12. Carvalho AF, Heilig M, Perez A, Probst C, Rehm J. Alcohol use disorders. Lancet 2019;394:781-792.
- 13. Blodgett JC, Del Re AC, Maisel NC, Finney JW. A metaanalysis of topiramate’s effects for individuals with alcohol use disorders. Alcohol Clin Exp Res 2014;38:1481-1488.
- 14. Mason BJ, Quello S, Goodell V, Shadan F, Kyle M, Begovic A. Gabapentin treatment for alcohol dependence: a randomized clinical trial. JAMA Intern Med 2014;174:70-77.
- 15. Khan A, Tansel A, White DL, Kayani WT, Bano S, Lindsay J, et al. Efficacy of psychosocial interventions in inducing and maintaining alcohol abstinence in patients with chronic liver disease: a systematic review. Clin Gastroenterol Hepatol 2016;14:191-202.e1-e4 quiz e20.
- 16. Castillo-Carniglia A, Keyes KM, Hasin DS, Cerdá M. Psychiatric comorbidities in alcohol use disorder. Lancet Psychiatry 2019;6:1068-1080.
- 17. Winder GS, Fernandez AC, Mellinger JL. Integrated care of alcohol-related liver disease. J Clin Exp Hepatol 2022;12:1069-1082.
- 18. Shaffer LR, Kaplan DE, Taddei TH, Mahmud N. The association between mental illness and all-cause mortality in patients with cirrhosis: a Veterans Affairs retrospective cohort study. Hepatol Commun 2023;7:e0129.
- 19. McHugh RK, Weiss RD. Alcohol use disorder and depressive disorders. Alcohol Res 2019;40:arcr.v40.1.01.
- 20. Collins R. What makes UK Biobank special? Lancet 2012;379:1173-1174.
- 21. Lee DU, Chou H, Wang E, Fan GH, Han J, Chang K, et al. The clinical implication of psychiatric illnesses in patients with alcoholic liver disease: an analysis of US hospitals. Expert Rev Gastroenterol Hepatol 2022;16:689-697.
- 22. Petermann-Rocha F, Gray SR, Forrest E, Welsh P, Sattar N, Celis-Morales C, et al. Associations of muscle mass and grip strength with severe NAFLD: a prospective study of 333,295 UK Biobank participants. J Hepatol 2022;76:1021-1029.
- 23. Petermann-Rocha F, Wirth MD, Boonpor J, Parra-Soto S, Zhou Z, Mathers JC, et al. Associations between an inflammatory diet index and severe non-alcoholic fatty liver disease: a prospective study of 171,544 UK Biobank participants. BMC Med 2023;21:123.
- 24. Moon AM, Yang JY, Barritt AS 4th, Bataller R, Peery AF. Rising mortality from alcohol-associated liver disease in the United States in the 21st century. Am J Gastroenterol 2020;115:79-87.
- 25. Pimpin L, Cortez-Pinto H, Negro F, Corbould E, Lazarus JV, Webber L, et al. Burden of liver disease in Europe: epidemiology and analysis of risk factors to identify prevention policies. J Hepatol 2018;69:718-735.
- 26. Asrani SK, Mellinger J, Arab JP, Shah VH. Reducing the global burden of alcohol-associated liver disease: a blueprint for action. Hepatology 2021;73:2039-2050.
- 27. Sinn DH, Kang D, Guallar E, Hong YS, Cho J, Gwak GY. Modest alcohol intake and mortality in individuals with elevated alanine aminotransferase levels: a nationwide cohort study. BMC Med 2022;20:18.
- 28. Rodríguez M, González-Diéguez ML, Varela M, Cadahía V, Andrés-Vizán SM, Mesa A, et al. Impact of alcohol abstinence on the risk of hepatocellular carcinoma in patients with alcohol-related liver cirrhosis. Am J Gastroenterol 2021;116:2390-2398.
- 29. John U, Rumpf HJ, Hanke M, Meyer C. Alcohol abstinence and mortality in a general population sample of adults in Germany: a cohort study. PLoS Med 2021;18:e1003819.
- 30. Rogal S, Youk A, Zhang H, Gellad WF, Fine MJ, Good CB, et al. Impact of alcohol use disorder treatment on clinical outcomes among patients with cirrhosis. Hepatology 2020;71:2080-2092.
- 31. Vannier AGL, Shay JES, Fomin V, Patel SJ, Schaefer E, Goodman RP, et al. Incidence and progression of alcohol-associated liver disease after medical therapy for alcohol use disorder. JAMA Netw Open 2022;5:e2213014.
- 32. Hofer BS, Simbrunner B, Hartl L, Jachs M, Bauer DJM, Balcar L, et al. Alcohol abstinence improves prognosis across all stages of portal hypertension in alcohol-related cirrhosis. Clin Gastroenterol Hepatol 2023;21:2308-2317.e7.
- 33. Greenfield TK, Rehm J, Rogers JD. Effects of depression and social integration on the relationship between alcohol consumption and all-cause mortality. Addiction 2002;97:29-38.
- 34. Perälä J, Kuoppasalmi K, Pirkola S, Härkänen T, Saarni S, Tuulio-Henriksson A, et al. Alcohol-induced psychotic disorder and delirium in the general population. Br J Psychiatry 2010;197:200-206.
- 35. Wilson JD, Altieri Dunn SC, Roy P, Joseph E, Klipp S, Liebschutz J. Inpatient addiction medicine consultation service impact on post-discharge patient mortality: a propensitymatched analysis. J Gen Intern Med 2022;37:2521-2525.
- 36. Yule AM, Kelly JF. Integrating treatment for co-occurring mental health conditions. Alcohol Res 2019;40:arcr.v40.1.07.
- 37. Holbeck M, DeVries HS, Singal AK. Integrated multidisciplinary management of alcohol-associated liver disease. J Clin Transl Hepatol 2023;11:1404-1412.
, Hyun Kook Lim2,4
