Alcohol Overuse After Gastric Bypass Surgery

Alcohol use disorders after gastric bypass procedures are a significant complication of this surgery.  I typically see people who have developed the disorder after the gastric bypass procedure.  The majority of those procedures these days are Roux-en-Y (RYGB).  The surgical mechanics of RYGB are widely available and I am not going to review them here.  This post is about how the procedure can change a person into a problem drinker endangering both their health and in some cases leading to substantial weight gain after initial losses from the bypass procedure.  Although there are other surgical techniques used for gastric bypass that also lead to this complication RYGB is the most common technique and is more likely to be studied.  In the figures above, blood alcohol levels are determined in individuals who had undergone RYGB surgery after ingestion of alcohol.

One of the well known principles in addiction medicine is that overuse of any potentially addictive substance generally follows pharmacological principles especially as tolerance to a drug develops.  Other factors such as economics, whether a drug can be used by a particular route, and biologically determined sensitivities to drugs can modify that basic concept.  To illustrate, most current opioid users with addiction problems started by using prescription painkillers.  They may have modified the original prescription pills by crushing them and either smoking or snorting them.  In some cases, they crush the pills, dissolve them, crudely filter them and inject them.  In this case they are trying to optimize the time to peak drug concentration Tmax and maximum drug concentration Cmax for maximum effect.  In the case of alcohol, it is generally too dangerous to use by the inhaled or intravenous route.  The effects are limited by absorption and elimination by zero order kinetics.

The graphics at the top of this page are from reference 1. It shows the results of blood alcohol concentration of 5 RYGB patients after drinking a solution of 40% vodka and orange juice calculated to contain 0.3 gram per kilogram body weight. The vodka and orange juice mixture was adjusted to contain a 50:50 mix of vodka:orange juice. Mean dose of ethanol given was 26.9 ± 2.3 grams or about 2 standard drinks consumed over a period of 5 minutes. On the graphs alcohol concentration is given as mg/dl. In the US, the current legal driving limit is 80 mg/dl or 0.08%. The researchers decided to look at the time from 0-10 minutes apart from 0-60 minutes because previous work started the measurements at 10 minutes.  The other point to notice in these graphs is that there is no control group.  That is unfortunate given the variable methods (alcohol dose and consumption time) that are used to determine the pharmacokinetics of alcohol in research papers that look at this question.

A comparison with control subjects in the literature can be done, but I could not find any studies that used a dose of 0.3 g/kg of ethanol.  Many studies used a dose of 0.6 g/kg.  The study used in the comparison below used a dose of 0.8 g/kg ingested over 30 minutes.   Comparing Tmax and Cmax for the study on RYGB patients to a study of 12 healthy male controls shows the following differences:

	Male controls (n=12)	RYGB patients (N=5)
Cmax (mg/dl)	106	138.4
Tmax (min)	60	5.4

Experimental differences aside. the Cmax and Tmax for alcohol in the RYGB patients are markedly different.  The time to peak alcohol levels was only 5.4 minutes in the RYGB group and at that time the average Cmax easily exceeded the legal driving limit on what amounts to 2 standard drinks.  Although the study did not correlate the rapid peak of alcohol with effects, they are readily observed by any clinician interested in how intoxicants affect the conscious state of the patient.  People typically tell me that they feel immediate intoxication or in some cases blackout.  The intoxication is a rapid sense of euphoria combined with an immediate cessation of negative mood states.  In the case of blacking out, this is a frequent end point with severe alcohol addiction.  Many people are self professed "black out drunks" and that is their preferred endpoint of acute intoxication and the goal of subsequent drinking sessions.  These states are highly reinforcing for susceptible people and probably explain the greater than expected occurrence of alcohol related disorders following RYGB surgery.  Looking at the prevalence of alcohol use disorder (AUD) before and after RYGB surgery, Mitchell, et al found that 8% had AUD within the 3 year post op period including 48% who had no prior history of AUD.  Including a more permissive screening tool the number increased to 18.4% and 40.6% respectively.  A previous study  showed that the prevalence of AUD symptoms was greater in year 2 post gastric bypass surgery (9.6%) and was associated with a number of preoperative variables including RYGB, male sex, smoking, and regular alcohol consumption before the procedure.

The studies that look at the prevalence of alcohol use are interesting.  From the perspective of an addiction psychiatrist who only sees patients selected for addiction, observing the effect of scale is significant.  As an example, the patients who I see with this problem are predominantly women and practically all of them did not have an alcohol use problem prior to surgery.  In many cases they did not drink alcohol.  Practically all had undergone RYGB but there were also people with other bypass procedures. 

An additional research question has been whether the anatomic changes in RYGB lead to changes in the brain and central nervous system that changes addictive behavior.  The model used for these investigations have been obese rats that have undergone RYGB and intravenous self administration of alcohol and drugs.  That model would eliminate pharmacokinetic considerations from oral administration with rapid increases in plasma concentration in a short while.   Rats with RYGB self administer intravenous alcohol and opioids at a higher rat than rats that have not had gastric bypass surgery suggesting a centrally mediated mechanism more independent of gut absorption (5,6).  One set of authors (5) proposed that in obesity there is blunted dopamine and ghrelin signalling.  They make the case that leptin and ghrelin both affect dopaminergic signalling in the reward system. (see diagram - click to enlarge).  

Despite being well known to eating disorder and addiction specialists, I don't think the phenomenon of alcohol overuse in gastric bypass patients is well known.  It has implications for informed consent for this surgical procedure, some risk factor modification, and follow up.  It also may have implications for the diagnosis and treatment of alcohol use disorders.  As an example, binge drinkers and blackout drinkers are a known subset the population of people with alcohol use disorders.  Does this group have any similarities with the population of RYGB patients who develop alcohol use disorders?  What would pharmacokinetic studies of alcohol in this group show?  If a profile of peptide hormone changes in the RYGB group is defined - should that be studied in binge and blackout drinkers as well?

These might lead to useful endophenotypes for alcohol use disorder that could lead to treatment interventions.


George Dawson, MD, DFAPA


References:

1:  Steffen KJ, Engel SG, Pollert GA, Li C, Mitchell JE. Blood alcohol concentrations rise rapidly and dramatically after Roux-en-Y gastric bypass. Surg Obes Relat Dis. 2013 May-Jun;9(3):470-3. doi: 10.1016/j.soard.2013.02.002. PubMed PMID: 23507629; PubMed Central PMCID: PMC4487806.

2:  Jones AW,  Jönsson KÅ.  Between subject and within subject variation in the pharmacokinetics of ethanol.  Br J Clin Pharmac 1994; 37: 427-431.

.
3:  Mitchell JE, Steffen K, Engel S, King WC, Chen JY, Winters K, Sogg S, Sondag C, Kalarchian M, Elder K. Addictive disorders after Roux-en-Y gastric bypass. Surg Obes Relat Dis. 2015 Jul-Aug;11(4):897-905. doi: 10.1016/j.soard.2014.10.026. PubMed PMID: 25862182; PubMed Central PMCID: PMC4430439.

4:  King WC, Chen JY, Mitchell JE, Kalarchian MA, Steffen KJ, Engel SG, Courcoulas AP, Pories WJ, Yanovski SZ. Prevalence of alcohol use disorders before and after bariatric surgery. JAMA. 2012 Jun 20;307(23):2516-25. doi: 10.1001/jama.2012.6147. PubMed PMID: 22710289; PubMed Central PMCID: PMC3682834.

5:  Polston JE, Pritchett CE, Tomasko JM, Rogers AM, Leggio L, Thanos PK, Volkow ND, Hajnal A. Roux-en-Y gastric bypass increases intravenous ethanol self-administration in dietary obese rats. PLoS One. 2013 Dec 31;8(12):e83741. doi: 10.1371/journal.pone.0083741. PubMed PMID: 24391816; PubMed Central PMCID: PMC3877092.

6:  Biegler JM, Freet CS, Horvath N, Rogers AM, Hajnal A. Increased intravenous morphine self-administration following Roux-en-Y gastric bypass in dietary obese rats. Brain Res Bull. 2016 May;123:47-52. doi: 10.1016/j.brainresbull.2015.08.003. PubMed PMID: 26304761; PubMed Central PMCID: PMC4761525.

Attributions:

1.  Figure at the top of this post is used with permission from copyright holder Copyright © 2013 American Society for Metabolic and Bariatric Surgery. Published by Elsevier Inc. All rights reserved through license 4007221260909 via the Copyright Clearance Center.

2.  Figure at the bottom is from VisiScience and is used per their licensing agreement.

Pancreatitis

There is an outstanding review of pancreatitis in this week's New England Journal of Medicine.  I thought I would add it here as a reference for any addiction or medical psychiatrists who come across this post and may not be regular readers of the NEJM.  I recommend getting the entire article because it has an excellent table and infographic.  The table is on the causes of acute pancreatitis.  The top two - gallstones (40%) and alcohol (30%) have not changed since I was in medical school.  Hypertriglyceridemia (defined as fasting triglycerides >1000 mg/dl) was the third most common cause at about 2-5% or the total.  Some of the causes occur only in a very specific context like endoscopic retrograde cholangiopancreatography (ERCP) and patients undergoing cardiopulmonary bypass.  In both cases, the authors estimated that 5 -10% of patients undergoing these procedures got pancreatitis.  Although the ERCP was expected I was surprised that many cardiopulmonary bypass patients got pancreatitis.  The remaining 5-8% are caused by medications, viral and parasitic infection, and blunt trauma.  Psychiatrists should be aware of valproic acid/valproate correlation with pancreatitis,  especially if they are treating patients with significant alcohol exposure.  Some facilities use a valproate detox protocol and those patients need to be carefully assessed for alcoholic liver diseases and undiagnosed pancreatitis.

The diagnostic features of acute pancreatitis are reviewed and these are important for any acute care psychiatrist who is seeing patient with associated risk factors.  At the clinical level abdominal pain and elevations of amylase and lipase 3 times the upper limit of normal are the initial features that require confirmation by MRI or CT imaging finding consistent with the disease.  From a diagnostic perspective the availability of testing and practicing in a medical facility are limiting factors.  Any abdominal pain with a suspicion of pancreatitis in a non-medical facility makes a trip to the emergency department for rapid assessment and diagnosis most reasonable.

The infographic was on the time course and management of acute pancreatitis.  The time frame used was 6 weeks.  80% of patients with acute pancreatitis have self limited disease and are discharged from the hospital in a few days.  The incidence of acute pancreatitis is rising with a 20% increase in admissions in the past decade.   On the mortality dimension half of the deaths occurred in the first two weeks due to multiple organ system failure.  The other other half of the death occur after two weeks and are due to pancreatic and extrapancreatic infections.  On the pathophysiology dimension, 80-85% of the disease was the interstitial form and 10-15% the necrotizing form.  There was also a therapy dimension outlining critical markers such as aggressive fluid resuscitation in the first 24 hours and enteral nutrition after day 5 if tolerated.   The therapy dimension was linked to the text that reviewed state of the art details on the medical and surgical management of the disorder.  These sections will not apply to psychiatrists, but the authors point out common mistakes in management including the unnecessary use of total parental nutrition and antibiotics for presumed pancreatic infection.   The main lessons for psychiatrists at that stage is that patient management has exceeded the capabilities of psychiatric settings and that the patient must be transferred as soon as possible to an appropriate medical setting.  Once that has occurred, the plan to take the patient back when stable also requires a clear plan with the attending who is discharging the patient.

The other highlights in this article from a psychiatrist's perspective was the estimated dose of alcohol necessary to cause pancreatitis.  The authors give that as 4 - 5 drinks per day for 5 or more years.  This is well below the dose of alcohol required for cirrhosis and probably explains why larger numbers of young patients are seen with pancreatitis.  Apparently, in the 2-5% of heavy drinkers that develop pancreatitis it occurs as a chronic form initially with episodic acute exacerbations superimposed on this chronic form.  That also explains why binge drinking does not precipitate acute pancreatitis.  Diabetes, smoking, and obesity are seen as correlates of acute pancreatitis but not direct causes and these are all significant comorbidities in patients with psychiatric disorders.  Once an episode of pancreatitis has occurred abstinence from alcohol is critical because it decreases the likelihood of recurrence.  The authors reference a structured consistent intervention that they cite as being successful (2).  I don't have access to the full text of this article, but it suggests that an infrequent intervention by a nurse in outpatient clinic is more effective than a single intervention during  hospitalization in preventing relapses.

This was a great overview of acute pancreatitis by some of the top experts in the field.  It is another problem that you never want to miss.  It is a reason to resist simplifying biochemical screening of patients on admission or clinic intake.  Since metabolic syndrome, obesity, and tobacco use is high and elevated triglycerides may be a factors in the pathophysiology of pancreatitis - it seems reasonable to do metabolic screening on patients without recent testing.  If you are seeing patient with risk factors particularly alcohol use, tobacco use, obesity, and diabetes - discussion of lifestyle management, smoking cessation, and abstinence from alcohol is useful in addition to the discussion about their primary psychiatric problem.  Addiction psychiatrists will probably see significant numbers of patients with chronic pancreatitis and a discussion with them on how to prevent recurrences and their understanding of the illness is important.

George Dawson, MD, DFAPA


References:

1. Forsmark CE, Vege SS, Wilcox CM. Acute Pancreatitis. N Engl J Med 2016; 375: 1972-1981.

2.  Nordback I, Pelli H, Lappalainen-Lehto R, Järvinen S, Räty S, Sand J. The recurrence of acute alcohol-associated pancreatitis can be reduced: a randomized controlled trial. Gastroenterology. 2009 Mar;136(3):848-55. doi:10.1053/j.gastro.2008.11.044. PubMed PMID: 19162029.