Impact of bariatric surgery on alcohol‐use disorder and alcohol‐related liver disease: A retrospective study Authors: Thomas J. Williams, Andrew J. Palmer, Jason P. Connor, Gerald Holtmann, Paul J. Clark Published: 2026 ABSTRACT AbstractBackground and aimsIncreased risk of alcohol dependence is recognised following bariatric surgery. Modified gastrointestinal anatomy and neurohormonal profiles lead to altered alcohol metabolism, peak blood alcohol concentrations and enhanced reward circuits and cravings. For patients with metabolic associated fatty liver disease, bariatric surgery has been shown to reduce liver fibrosis; however, the impact of bariatric surgery on alcohol consumption and consequent liver fibrosis has not been well‐characterised using validated measures. This study examined the changes in alcohol consumption patterns, alcohol dependence and validated liver fibrosis measures before and after bariatric surgery.Design, setting and participantsA retrospective cohort study assessing patients seeking treatment for alcohol problems with a history of gastric sleeve (GS) or Roux‐en‐Y Gastric Bypass (RYGB) in a quaternary hospital in Brisbane, Australia.MeasurementsValidated measures of alcohol dependence [Alcohol Use Disorders Identification Test (AUDIT) and brief Michigan Alcohol Screening Test (bMAST)] and non‐invasive liver fibrosis (FIB4 and APRI) were compared pre‐ and post‐surgery.FindingsTwenty patients were identified with prior GS and 10 patients with RYGB. Following bariatric surgery, body mass index was reduced by 14.5 (±9.39) kg/m2. Metabolic co‐morbidities were less frequent post‐surgery. Pre‐surgery, the mean AUDIT score was 10.2 ± 8.9, with 9 patients having scores indicative of alcohol dependence. Mean AUDIT increased statistically significantly post‐surgery (18.1, standard deviation ±11.4; P < 0.001) with 29 patients meeting criteria for alcohol dependence. AUDIT components relating to both consumption (volume, frequency) and harmful consequences of alcohol increased following surgery. Despite surgery lowering metabolic liver fibrosis risk, worsened liver fibrosis was observed [mean FIB4 (0.98 ± 1.8) and APRI (0.41 ± 0.83) increased; P < 0.05].ConclusionWithin a cohort of adults seeking treatment for alcohol problems, increases in alcohol consumption and alcohol‐related harm were observed following bariatric surgery. Despite improvement in risk factors for metabolic‐associated fatty liver disease after surgery, non‐invasive markers for hepatic fibrosis worsened, likely in the setting of harmful alcohol intake and associated liver injury. This supports the role of bariatric surgery pre‐operative assessment and post‐operative follow‐up to identify and manage alcohol‐related problems and associated liver disease risk. INTRODUCTION The obesity epidemic is one of the most pressing issues facing global healthcare. In addition to the resources required for the management of metabolic conditions [1, 2], obesity places substantial financial burden on health economies [3]. The prevalence of obesity and its associated burdens continues to increase world‐wide [1], including in Australia where two‐thirds of the population are classified as overweight or obese [4]. In the setting of the individual and health system pressures, bariatric surgery offers an attractive and effective ‘one‐off’ treatment for the management of obesity, compared to the limited efficacy of dietary interventions and long‐term cost of weight loss medicines such as glucagon‐like peptide‐1 receptor agonists (GLP1‐RA). Gastric sleeve (GS) and Roux‐en‐Y Gastric Bypass (RYGB) are commonly performed and associated with sustained weight loss with improvement in obesity‐related chronic conditions [5]. However, there are also potential adverse effects following bariatric surgery, including growing evidence for possible increased risk of harmful patterns of alcohol consumption [6, 7, 8, 9, 10, 11, 12, 13, 14]. Several mechanisms have been proposed to explain the increased risk of developing alcohol dependence after weight loss surgery. First, the absorption and metabolism of alcohol appears to be altered following the changes in anatomy from both GS and RYGB surgeries, leading to increased and more rapid peaks in blood alcohol concentration, which by increasing the immediacy of reward response to alcohol exposure, may reinforce reward pursuit [15, 16, 17, 18, 19, 20]. First‐pass gastric metabolism of alcohol is reduced with decreased gastric mucosal alcohol dehydrogenase, in addition to accelerated gastric emptying providing a higher concentration of alcohol transferred for absorption in the small bowel [15, 21, 22]. These higher peaks in blood alcohol concentration are expected to increase alcohol concentration and metabolites delivered to the liver via the portal circulation. Second, changes in gut anatomy alters intestinal neuro‐hormonal profiles, including ghrelin and leptin, which may be associated with altered alcohol consumption patterns, alcohol cravings and changes in reward pathway activation [23]. Third, the gut‐brain‐microbiome axis may contribute to alcohol consumption behaviours and cravings, with studies demonstrating distinct microbiome diversity based on alcohol use patterns [24]. Altered gastrointestinal anatomy following bariatric surgery has been associated with changes in microbiome diversity and may contribute to altered alcohol consumption behaviours [25]. Finally, the concept of ‘addiction transfer’ has also been proposed because of imposed restrictions on foods and eating habits, which impairs satisfaction/reward from eating after bariatric surgery. Patients replace the emotionally dysfunctional consumption of food, with alcohol [26, 27]. Studies using insurance databases have suggested that RYGB is associated with an increased hazard of de novo alcohol‐related hepatitis and a two‐fold higher risk of alcoholic cirrhosis in women [28, 29]. Despite existing data being suggestive of an increased risk of both alcohol dependence and possibly alcohol‐related liver disease, metabolic‐associated fatty liver disease (MAFLD) typically improves or completely resolves following bariatric surgeries [30]. An increase in alcohol consumption and risk for alcohol associated liver disease may mute benefits to MAFLD, although there are limited data exploring this hypothesis, using validated measures for alcohol dependence and liver fibrosis assessment. Studies also poorly clarify other comorbid risks for liver disease, and this may lead to attribution bias to surgery as the cause of subsequent liver disease, when pre‐surgical factors may be responsible. We hypothesised that bariatric surgery may contribute to an increased risk of alcohol use disorder (AUD) and that this may worsen liver fibrosis, despite improvements in metabolic risk factors for MAFLD fibrosis. This study aimed to quantify changes in alcohol consumption patterns, alcohol dependence measures and validated liver fibrosis markers before and after bariatric surgery in a treatment‐seeking cohort, using standardised tools in an addiction‐hepatology setting. METHODS AND MATERIALS Study design and setting This retrospective cohort study was conducted at the Princess Alexandra Hospital (PAH), Brisbane. Patients 18 years or older presenting to the Alcohol and Drug Assessment Unit (ADAU) between January 2018 and October 2023, with a history of RYGB or GS were included. Participants Eligible patients had undergone bariatric surgery before ADAU referral and had sufficient data to assess alcohol use and liver fibrosis using validated tools. Patients were excluded if data were incomplete or they were unable to complete structured interviews. Clinical and demographic information was collected from state‐wide electronic medical records and supplemented by structured interviews conducted by trained addiction clinicians. Assessment of alcohol dependence Alcohol consumption was assessed using the Alcohol Use Disorders Identification Test (AUDIT) [31], including total and AUDIT‐C (items 1–3) scores, and the Brief Michigan Alcohol Screening Test (bMAST) [32], a 10‐item tool evaluating lifetime alcohol‐related risk. Note that pre‐operative AUDIT and bMAST scores were calculated based on patient recall, which has been shown to be robust and valid for historic alcohol consumption [33], and particularly for harm‐ and event‐related recall features (e.g. drink‐driving charges, loss of work and marriage breakdown) important in the assessment of dependence. Standard published thresholds were used to define low‐risk use, harmful use and probable dependence. Assessment of liver fibrosis Liver fibrosis was assessed using the Fibrosis‐4 index (FIB4), a validated non‐invasive measure that uses age, alanine aminotransferase (ALT), aspartate aminotransferase (AST) and platelets [34]. Cut‐offs for low‐ and high‐risk fibrosis were less than 1.30 and more than 2.67, respectively [35]. The AST to platelet radio index (APRI) [36] was also calculated. Discordant fibrosis assessments were reviewed by hepatology clinicians. This study and its analysis plan were not pre‐registered on a publicly available platform, therefore, the results should be considered exploratory. Procedures Pre‐operative data were taken from the 3 months before surgery. Post‐operative data were collected at ADAU assessment. Blood tests for fibrosis scores were drawn within 3 months of each time point. Samples with transient hepatic derangement were excluded. Statistical analysis Data normality was assessed using the Shapiro–Wilk test. Normally distributed data were reported as mean (±SD), and skewed data as median [interquartile range (IQR)]. Paired t tests or Wilcoxon signed‐rank tests were used for within‐group comparisons. Bonferroni correction adjusted for multiple comparisons. Cohen's d was used to estimate effect size. Ethics This study was approved by the Metro South Health Human Research Ethics Committee (HREC/2022/QMS/91577). RESULTS Forty‐nine patients were identified who had attended ADAU with a diagnosis AUD with prior GS or RYGB surgery. At the time of the study, 19 individuals were excluded. There were insufficient pre‐operative blood results available for the assessment of FIB4 in 13 patients. Six patients were deceased and unable to complete retrospective assessments. Of the six deceased patients, all had a history of GS, with a mean time between surgery and death of 9.4 years. Of note, five of the deceased patients had an established diagnosis of alcohol‐related cirrhotic liver disease with death occurring as a direct complication of decompensated liver cirrhosis. Records were not available surrounding the death of the sixth patient. Of the 30 patients included within this study, 20 individuals had a history of GS and 10 had previous RYGB. The mean age (±SD) at time of review within ADAU was 49.0 years (±9.2). The median time between surgery and the post‐operative assessment was 1473.4 days (IQR = 1273). 73.3% of patients included within this study were female. Before bariatric surgery, the mean weight for patients was 129.3 kg (±32.9). The mean body mass index (BMI) was 45.9 kg/m2 (±10.9). Two patients had pre‐bariatric surgery diagnosis of type 2 diabetes mellitus (both medicated), 12 were diagnosed with hypertension (8 medicated) and five with hypercholesterolaemia (2 medicated). Bariatric surgery was successful in achieving weight loss and reducing associated metabolic risks for liver disease. At the time of review following bariatric surgery, the mean weight was 88.1 kg (±21.1), with a mean reduction in weight of 41.2 kg (±26.4; P < 0.001). Weight decreased in all patients except for one. The mean post‐operative BMI was 31.4 kg/m2 (±7.2), with a mean reduction in BMI of 14.5 kg/m2 (±9.39; P < 0.001; Cohen's d = 1.59). Post‐operatively, there was a reduction in medications prescribed to treat metabolic complications such as type 2 diabetes mellitus (from 2 to 0 patients), and hypertension (from 12 to 7 patients), with no change in patients prescribed medications for hypercholesterolaemia. Alcohol dependence Before bariatric surgery, the mean AUDIT score was 10.2 (±8.9). A total of 63% of patients (n = 19) recorded low‐risk pre‐operative AUDIT scores (score, ≤7), 11 patients (36.7%) had harmful alcohol use (score, >8) and nine patients (30%) had scores indicative of alcohol dependence (≥13 in females, ≥15 in males). The mean bMAST score before bariatric surgery was 3.7 (±6.07), with a score consistent with problematic or high‐risk alcohol use (score, ≥6) observed in seven patients (23.3%). Following bariatric surgery, the cohort mean comprehensive AUDIT score increased to 28.3 (±9.2) with the mean change in score 18.1 (±11.37; P < 0.001; Cohen's d = 1.57). The mean change in AUDIT score was similar between patients following RYGB (19.7 ± 9.0) and GS surgery (18.3 ± 11.8; P = 0.75). Following surgery, 29 patients reported AUDIT scores suggestive of alcohol dependence, compared with nine pre‐operatively. In those patients considered high‐risk for dependence pre‐operatively (mean pre‐operative AUDIT score 22.5 ± 5.2), the AUDIT risk score increased by a mean of 8.4 points (P < 0.05). Interestingly, patients with low‐risk AUDIT scores pre‐surgery (scores of ≤7), had larger increases in mean AUDIT scores following surgery [mean change in AUDIT score of 26.8 (P < 0.001)]. AUDIT items relating to volume and frequency of alcohol consumption (items 1–3) increased following surgery (Table 1), with the mean AUDIT‐C scores incrementing from 6.07 pre‐surgery to 10.83 following surgery (P < 0.005). Regarding the frequency of alcohol use before bariatric surgery, the most common response was ‘two to four times a month’ (36.7%), followed by ‘two to three times a week’ (30.0%) and ‘monthly or less’ (20.0%). Before surgery, only four patients (13.3%) reported consumption of alcohol ‘four or more times per week’ compared with 24 patients (80%) after surgery [Figure 1(a)]. TABLE 1Measures of metabolic disease, alcohol dependence and liver fibrosis pre‐ and post‐bariatric surgery.Pre‐surgeryPost‐surgeryAge (years) ± SD44.3 ± 8.749.0 ± 9.2Mean weight (kg) ± SD129.3 ± 32.988.1 ± 21.1Mean BMI (kg/m2) ± SD45.9 ± 10.931.4 ± 7.2Frequency of co‐morbidities (prescribed medications)Type 2 diabetes mellitus2 (2)0 (0)Hypertension12 (8)7 (5)Hypercholesterolaemia5 (2)3 (2)Mean AUDIT score10.2 ± 8.928.3 ± 9.2Mean FIB4 score1.04 ± 0.862.02 ± 2.08Mean APRI score0.37 ± 0.420.78 ± 0.97Abbreviations: APRI, aspartate aminotransferase to platelet radio index; AUDIT, Alcohol Use Disorders Identification Test; BMI, body mass index; FIB4, Fibrosis‐4. Measures of metabolic disease, alcohol dependence and liver fibrosis pre‐ and post‐bariatric surgery. Abbreviations: APRI, aspartate aminotransferase to platelet radio index; AUDIT, Alcohol Use Disorders Identification Test; BMI, body mass index; FIB4, Fibrosis‐4. (a) Frequency of alcohol consumption for patients before and following bariatric surgery according to the Alcohol Use Disorders Identification Test (AUDIT) questionnaire. (b) Number of drinks consumed on a typical day when drinking alcohol for patients before and following bariatric surgery according to the AUDIT questionnaire. Using the AUDIT scores, the mean standard drinks per day was calculated pre‐ and post‐surgery (Table 2). Before surgery, the mean standard drinks per day was 1.51 (±1.93), increasing in frequency to 6.63 standard drinks per day (±2.88; P < 0.001) following surgery. TABLE 2Mean comprehensive AUDIT scores for individual questions for patients before and after bariatric surgery.AUDIT itemPre‐surgery scorePost‐surgery scoreDifference P‐valueAlcohol consumption1. How often do you have a drink containing alcohol?2.373.671.30<0.0012. How many drinks containing alcohol do you have on a typical day when drinking?1.833.501.67<0.0013. How often do you have six or more drinks on one occasion?1.873.671.80<0.001Alcohol related problems4. How often during the last year have you found that you were not able to stop drinking once you had started?0.932.631.70<0.0015. How often during the last year have you failed to do what was normally expected from you because of drinking?0.271.501.23<0.0016. How often during the last year have you needed a first drink in the morning to get yourself going after a heavy drinking session?0.101.631.53<0.0017. How often during the last year have you had a feeling of guilt or remorse after drinking?0.732.772.03<0.0018. How often during the last year have you been unable to remember what happened the night before because you had been drinking?0.331.871.53<0.0019. Have you or someone else been injured as a result of your drinking?0.672.732.07<0.00110. Has a relative or friend or a doctor or another health worker been concerned about your drinking or suggested you cut down?1.073.602.53<0.001Abbreviation: AUDIT, Alcohol Use Disorders Identification Test. Mean comprehensive AUDIT scores for individual questions for patients before and after bariatric surgery. Abbreviation: AUDIT, Alcohol Use Disorders Identification Test. An increase in heavy episodic drinking was also observed [Figure 1(b)]. When asked the number of drinks consumed on a typical day when drinking before surgery, the most common response was both ‘one or two’ and ‘five or six’ (23.3%), then ‘three to four’ (20.0%). Following surgery, 66.7% of respondents consumed ‘ten or more’ drinks on a typical day, followed by ‘seven to nine’ (23.3%), ‘five or six’ (6.7%) and ‘one or two’ (3.3%). Items assessing alcohol‐related social and health problems (items 4–10) demonstrated increased risk following bariatric surgery (P < 0.005) (Table 2). Before bariatric surgery, 20.0% of patients reported an injury to themselves or another as a result of drinking, however, this increased to 73.3% following surgery. The frequency of patients with a forensic history of arrest for driving with blood alcohol concentration over the legal limit increased by 10% (13.3%–23.3% following surgery). Liver fibrosis The cohort was low risk generally of liver fibrosis (FIB4 score, <1.30) before surgery, with a mean FIB4 pre‐operative score of 1.04 (±0.86). Higher‐risk score of >2.67 was identified in only two patients. Following bariatric surgery, the mean FIB4 score doubled [2.02 ± 2.08; mean increase in FIB4 score 0.98 (±1.78, P = 0.001; Cohn's d = −0.544)]. The mean increase in FIB4 score was higher following RYGB [1.83 (± 2.4)] than GS [0.559 (±1.2; P = 0.068)]. Following surgery, six patients registered scores suggestive of high risk for liver fibrosis (>2.67). After surgery, less patients had low‐risk fibrosis FIB4 score (FIB4 <1.30; 26 patients pre‐operatively to 18 patients post‐operatively) despite better control of metabolic risks. Similarly, the mean APRI score increased pre‐surgery from 0.37 (±0.42) to 0.78 (±0.97) post‐surgery (P = 0.006; Cohen's d = −0.492), associated with an increase in patients with scores with significant fibrosis (2 pre‐operatively to 7 post‐operatively). No patients were considered cirrhotic before surgery, however, five patients had FIB4 scores consistent with cirrhosis post‐operatively. DISCUSSION Bariatric surgery plays a significant role in the management of obesity, by reducing weight and improving metabolic risk factors. Such beneficial effects reduce the risk of MAFLD causing progressive fibrosis and cirrhosis. In our clinically and psychologically well‐characterised treatment‐seeking cohort, increases in validated measures of alcohol dependence were observed following bariatric surgery. Increases in volume and frequency of alcohol consumption were observed after surgery, but also importantly harmful or consequential drinking. Paradoxically, despite substantial weight loss and a reduction in metabolic risk factors typically associated with improved hepatic fibrosis, we observed significant increases in non‐invasive markers of liver fibrosis. These changes are likely attributable to the escalation in harmful alcohol intake and consequent alcohol‐related liver injury, suggesting that the hepatic improvements in fibrosis associated with bariatric surgery can be negated by increased alcohol intake and/or development of de novo AUD. Emerging literature increasingly supports an association between bariatric surgery and hazardous alcohol consumption, however, the nature of that risk and end‐organ complications such as cirrhosis remain less well defined. Across longitudinal and registry‐based studies, an increase in post‐operative incidence of AUD was observed for both RYGB and GS, with rates up to 18%, typically emerging after 2 to 5 years [6, 7, 8, 9, 10, 11, 12, 13, 14, 37, 38]. Patients who develop alcohol dependence post‐operatively may be more likely to be lost to follow‐up in bariatric outcome studies, which may create a bias that understates these risks. These cohorts were not informed by pre‐ and post‐surgery liver disease assessments. As anticipated, 29 of the 30 patients in this study met criteria for diagnosis of AUD. Importantly, however, AUDIT scores significantly increased post‐surgery relative to pre‐operative estimates. Perhaps of even more concern were the relatively larger score increments and greater increased risk of dependence severity, seen in individuals with low‐risk assessments pre‐surgery. This suggests that screening for AUD should be an active process both before and after surgery, as post‐operative AUD may not be well predicted by validated screens pre‐operatively. Volume and frequency of alcohol consumption increased, as did alcohol‐related problems post‐surgery, which all contributed to the increase in AUDIT scores following surgery. This emphasises that it is not just consuming more alcohol, but greater consequential alcohol‐related harms associated with alcohol dependence are observed post‐surgery. Items associated with alcohol consumption (items 1–3, combining to form AUDIT‐C) demonstrated a significant rise following surgery in our study. Although data are limited surrounding individual AUDIT items, studies have observed an increment in the shorter‐form AUDIT‐C scores following surgery [7, 10, 12]. Additionally, various other measures of alcohol consumption have been used, demonstrating increases in consumption following bariatric surgery [39, 40]. Importantly, our study demonstrated a significant increase in the frequency of heavy episodic drinking (AUDIT item 3) following surgery, which has been described as an independent risk factor for the development of alcohol‐related liver disease [41]. With respect to alcohol associated harm, one retrospective cohort study demonstrated increased alcohol‐related admissions following bariatric surgery [9, 42]. Our study is able to characterise the nature of these consequences, demonstrating significant increases in harms from personal health, social and work‐related function and injury to others. Coluzzi et al. [37] reported a reduction in modified AUDIT assessment 1 year following sleeve gastrectomy, with a reduction in patients meeting criteria for harmful alcohol consumption. Quantity of alcohol being consumed at 1 year was also noted to be lower compared to baseline, however, the study experienced significant loss to follow‐up (172 procedures with 146 patients enrolled in the study), which because of possible selection bias may be important. In our study, all AUDIT assessments were conducted by physicians, nurses and psychologists trained and experienced in addiction medicine, while Coluzzi et al. [37] did not clarify who or how the modified AUDIT was administered and excluded patients with pre‐surgical ‘alcohol abuse’. Using a modified AUDIT‐C tool, Wee et al. [43] also observed a reduction in high‐risk drinking for patients following RYGB and gastric banding at 1 and 2 years following surgery. Again, these findings may be related to the short duration of follow‐up, possible selection bias, different ascertainment strategies and these were not a treatment seeking cohort. In contrast, Mitchell and colleagues followed up a cohort of patients after RYGB [44]. This non‐treatment seeking cohort were screened with either the AUDIT‐C or SCID (Structured Clinical Interview for DSMV Axis), 18.4–40.6% met criteria for alcohol use disorder, of whom 31.6% had no previous diagnosis [44]. Several studies have noted an initial improvement in alcohol consumption following surgery, before subsequent increase in alcohol use [9, 11, 12]. Therefore, outcomes at early follow‐up may not be reflective of long‐term outcomes. In the early post‐operative phase, patients are likely to implement healthy behavioural changes because of both closer medical follow‐up and individual patient motivation, before motivation and follow‐up later declines. This may account for studies with more prolonged follow‐up suggesting increased rates of AUD. Stigma around alcohol is a significant barrier that may lead to under‐reporting in post‐operative screening studies. In contrast, our cohort was composed of individuals with identified alcohol problems seeking treatment. Several factors have been proposed to increase the risk of AUD following bariatric surgery. Male gender, younger age at time of surgery, history of smoking or recreational drug use and reduced degree of social supports have been described as risk factors [13, 38, 45]. Regular alcohol use and a history of AUD before surgery have also been associated with increased risk [11, 13, 38, 45]. Although previous hazardous alcohol consumption appears to increase risk of post‐operative AUD, our study also highlights the risk of de novo AUD diagnosis. Before surgery, 19 patients were categorized as low‐risk alcohol use, however, at time of review 29 patients met criteria for AUD, with a mean change in AUDIT score of 26.8. Kim et al. [28] reported an increased risk of de novo alcohol‐related diagnoses following RYGB, while Wee et al. [43] demonstrated new high‐risk drinking in 7% of patients at 1 to 2 years following bariatric surgery. A proportion of patients without previous hazardous alcohol consumption are at risk of developing new AUD following bariatric surgery. This emphasises the importance of post‐operative monitoring for increased alcohol consumption and harms, even in those with no previous alcohol‐related disorder diagnoses. Despite the association between bariatric surgery and hazardous alcohol consumption, there remains minimal literature exploring the relationship with alcohol‐related liver disease. Kim et al. [28] demonstrated an increased risk of alcohol‐related hepatitis following RYGB, however, this was not reflected for patients following GS. In our cohort of patients, encompassing 20 GS and 10 RYGB, similar changes in alcohol consumption and consequence‐orientated measures of the AUDIT were observed between the two types of surgeries. When analysing specific types of weight loss surgery, Mellinger et al. [29] did not demonstrate an increased risk of alcohol‐related cirrhosis following bariatric surgery. However, combined data of all types of weight loss surgery with a longer follow‐up period, did demonstrate increased risk of alcohol‐cirrhosis [29]. Our study demonstrated an increase in non‐invasive markers of liver fibrosis following bariatric surgery. A reduced proportion of patients were categorised as low‐risk for fibrosis following surgery, with a clinical diagnosis of cirrhosis in five of the patients following surgery. Although regression of fibrosis associated with metabolic risk factors is well described after bariatric surgery, our observed association between worsening of non‐invasive fibrosis markers strengthens the inference that alcohol, rather than residual metabolic risk, is the principal driver of hepatic injury in this subgroup. This finding contrasts with the broader literature demonstrating improvements in both histological and biochemical markers of MAFLD following weight‐loss surgery [46, 47, 48]. Taken together, our data suggest that the hepatic benefits of metabolic improvement can be negated, or even reversed, by emergent hazardous alcohol consumption and/or de novo AUD, contributing to progressive liver fibrosis in a proportion of patients. Our findings should be considered within the context of some limitations. Given the retrospective nature of this study, recall bias is a potential limitation, however our recent scoping review of lifetime drinking‐history instruments found that structured retrospective tools appear to demonstrate moderate‐to‐strong reliability and validity, even when recall spans many years [33]. Information bias may be present in the setting of this retrospective review, however, the use of a state‐wide medical record system is used at our quaternary centre, particularly to access historic pre‐operative liver tests. We also acknowledge that our study population is selected from a treatment‐seeking cohort of patients with no control group present and is enriched for a group of patients presenting and seeking help for alcohol dependence. Our outcome of interest was to assess retrospective alcohol and liver problems in individuals who had presented with alcohol dependence post‐bariatric surgery. Our study question did not incorporate the incidence of AUD among all individuals undergoing surgery, which would require a larger sample size prospectively ascertained via surgical clinics. Importantly, treatment‐seeking typically represents a small selection of patients who develop alcohol dependence (most of whom do not seek treatment), and this group may be indicative of more widespread and unrecognised alcohol problems in this population. Importantly, our study may understate the impacts of bariatric surgery on alcohol‐related liver disease because of survival bias. Six patients who were reviewed post‐surgery had deceased and were excluded from our study because of the inability to assess their pre‐operative addiction measures. Five patients were diagnosed with alcohol‐related cirrhosis, with their death a direct complication of liver disease. Medical records were not available for the sixth patient regarding diagnosis of cirrhosis or cause of death. Although these patients have been excluded from the study, the clinical picture was of alcohol‐related cirrhosis in the setting of worsening alcohol consumption following bariatric surgery. The exclusion of these patients contributes to an underestimation of the risk of increased alcohol consumption and worsened liver‐related harms following surgery (Type I error). Given the large sample size and long follow‐up required to prospectively capture all patients with alcohol dependence and liver fibrosis following surgery, which has not been achieved in other studies, this convenient sample via an addiction service offers a pragmatic sampling strategy. Although bariatric surgery will continue to play an important role in the management of obesity, there has been a suggestion that rates of surgery are declining with the increasing uptake in GLP1‐RA agents [49]. There is growing interest in the impact of GLP1‐RA agents on substance and AUDs [50] with emerging data suggesting these agents may be protective against the development and recurrence of AUD in patients with metabolic disease [51, 52], however, prospective data are lacking. There is limited data currently on the role of these agents in management of AUD following bariatric surgery, although clearly adjunctive use in this setting may be a strategy to maintain weight loss goals and offset the risk of reward‐transfer behaviours that might enable harmful alcohol use [53]. In conclusion, our study suggests an increased risk of alcohol consumption, alcohol‐related harm and worsening of markers of liver fibrosis in some patients following bariatric surgery. Given weight loss surgery remains a key strategy to manage obesity, it is important that clinicians are aware of such risks and have pathways to assess and mitigate potential risks facing patients. Worsening alcohol‐related liver disease may offset benefits for patients trying to reduce their metabolic health risks, particularly MAFLD. Our study suggests that pre‐treatment screening may not identify all patients at risk and supports the role of close follow‐up surveillance for alcohol dependence and the provision of adequate resources to identify and manage alcohol‐related diagnoses and complications, including dependence and liver fibrosis. Prospective studies will be an important additional element to assess both dependence measures and markers of liver fibrosis with longer follow‐up periods to better characterise risk and impacts of these illnesses after bariatric surgery. AUTHOR CONTRIBUTIONS Thomas Williams J: Conceptualization; methodology; data curation; investigation; validation; formal analysis; project administration; visualization; writing—original draft; writing—review and editing. Andrew Palmer J: Writing—review and editing; conceptualization; methodology; validation; formal analysis; investigation; writing—original draft. Jason Connor P: Conceptualization; methodology; writing—review and editing; formal analysis; validation; writing—original draft. Gerald Holtmann: Conceptualization; methodology; writing—review and editing; writing—original draft. Paul Clark J: Writing—original draft; writing—review and editing; conceptualization; methodology; visualization; investigation; formal analysis; validation; supervision. DECLARATION OF INTERESTS None.