SEARCH

SEARCH BY CITATION

Keywords:

  • Weight loss surgery;
  • substance abuse;
  • addiction;
  • alcohol dependence

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusions
  9. Conflict of Interest Statement
  10. Acknowledgements
  11. References

What is already known about this subject

  • Elevated rates of substance use disorders (SUDs), particularly (but not only) alcohol use, are observed among post-weight loss surgery (WLS) patients.
  • The development of SUDs among post-WLS patients typically occurs 1–2 years post-surgery.
  • Post-WLS patients are developing SUD at a much later time of life than is typical of SUDs in the general population, suggesting they constitute a distinct SUD phenotype that is directly related to having undergone WLS.

What this study adds

  • Overall, findings suggest that post-WLS patients are overrepresented in substance abuse treatment programmes, and the majority of them report no history of SUD before WLS.
  • Relative to non-WLS patients in SUD treatment, post-WLS patients in substance abuse treatment are disproportionally diagnosed with alcohol dependence, including alcohol withdrawal.
  • Post-WLS patients may be at elevated risk for development of New Onset SUD in the absence of a prior SUD history; this group is phenotypically different from those with a history of substance abuse prior to surgery, and such patients may have unique treatment needs.

A comprehensive substance abuse treatment facility began observing increased admissions who reported histories of weight loss surgery (WLS). Emerging evidence suggests that roughly half of post-WLS patients in substance abuse treatment developed their substance use disorder (SUD) after surgery. The present study examined differences between SUD patients who developed New Onset SUD after surgery and those with a reported SUD onset before WLS (SUD Hx+ group). Participants completed a questionnaire and participated in a semi-structured interview. Data were also obtained from participants’ electronic medical records. Of the total treatment sample (n = 4658), 2.8% reported a history of WLS. Post-WLS patients were significantly more likely to be diagnosed with alcohol use disorders (AUDs). Among post-WLS patients who were interviewed (n = 56), 60% were classified as New Onset SUD, while only 40% were SUD Hx+. SUD Hx+ cases reported using significantly more types of substances than New Onset cases and were more likely to report pre-surgical binge eating disorder (BED). Post-WLS patients are overrepresented in substance abuse treatment and are disproportionally diagnosed AUDs. Post-WLS patients may be at elevated risk for development of New Onset SUD at a time in life (middle age) when SUD onset is relatively uncommon.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusions
  9. Conflict of Interest Statement
  10. Acknowledgements
  11. References

Media attention and anecdotal evidence from substance abuse treatment providers suggest elevated rates of alcohol use among post-weight loss surgery (WLS) patients, yet few empirical investigations of this phenomenon have been conducted. While few studies have examined alcohol use among post-WLS patients [1-6], even fewer studies have examined post-surgical abuse of other types of substances. Two recent reports [3, 5] support that substance use disorder (SUD) rates increase after the first post-operative year, particularly after the Roux-en-Y gastric bypass (RYGB) procedure; this increase may not solely include alcohol use disorders (AUDs), but may extend to other substances as well [4-8].

A recent report [4] suggested that post-WLS patients may be overrepresented in in-patient substance abuse treatment, constituting perhaps as many as 6% of treatment admissions, with nearly all such cases having had the RYGB procedure. Previous research from our group has found that at least half of post-WLS patients enrolled in in-patient substance abuse treatment developed SUDs after surgery, a group we refer to as New Onset SUD [6-8]. Specifically, across our previous samples, two-thirds of our SUD cases reported no SUD prior to surgery. This finding emerges across in-patient [6, 7] and community samples [9, 10]. We suggest that this New Onset SUD group may constitute a phenotypically distinct group that likely differs in important ways from those WLS patients who relapse to or continue substance use after WLS. As such, understanding the risk factors for and unique treatment needs of those with New Onset SUD after WLS is of considerable clinical importance.

Some demographic differences have been found between post-WLS SUD cases and those in the general population. Although in general, SUDs are more prevalent among men, post-WLS SUDs appear to be more common in women, although this may simply reflect the fact that 70–80% of WLS patients are female [3]. Interestingly, the age of onset of new onset post-WLS SUDs tends to be between 40 and 50 years of age [8-10], well beyond the period of highest risk in the general population, wherein epidemiologic data indicates the mean age of SUD onset is around age 18 and declines considerably after age 25 [11, 12]. Therefore, post-WLS patients may encounter problematic substance use at a time in life that is not normative for new onset SUD. Post-WLS patients who encounter new onset SUD after surgery may be phenotypically different than the typical in-patient SUD patient and, consequently, may have unique treatment needs and different etiological mechanisms.

In summary, evidence suggests that post-WLS patients may be at risk for SUDs, either of the New Onset variant or by virtue of relapsing to or continuing previous substance use patterns. The present investigation examined differences between New Onset SUD patients and those who admitted to problematic substance use prior to surgery, a group we refer to as SUD History Positive (SUD Hx+).

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusions
  9. Conflict of Interest Statement
  10. Acknowledgements
  11. References

Participants

All participants were voluntarily enrolled at Brighton Hospital, the second oldest comprehensive drug and alcohol treatment programme in the United States, receiving referrals from a broad geographic range primarily encompassing the US Midwestern states.

Electronic medical record (EMR) data were obtained from 4658 patients in Brighton's rehabilitation, detoxification or partial hospitalization programmes from July 2009 to April 2011. This database does not overlap with that used for our previous report [4], which covered April 2006 through May 2009. The new database included psychiatric diagnoses (including alcohol/drug dependence and withdrawal) and history of WLS, which was routinely assessed during intake. Specifically, upon admission to SUD treatment, intake interviewers asked all patients if they had a history of WLS. A sample of 129 post-WLS patients were identified, 56 of whom were recruited to complete the interview portion of this investigation, allowing us to explore differences between New Onset SUD cases (n = 34; 61%) and SUD Hx+ (n = 22; 39%). Of the 56 post-WLS participants, 51 completed the questionnaire for this investigation.

Prior to data collection, this study received approval from both the University and Hospital Institutional Review Boards. EMR data were provided in de-identified fashion to generate estimates of the overall prevalence of WLS history among the full census of an in-patient SUD treatment population. Those who formally enrolled in the questionnaire and/or interview part of the project provided informed consent, which included permission to review their medical records in more detail than what had been obtained as de-identified data from the EMR for the full hospital census.

Procedures

During admission to the hospital, if patients reported a history of WLS, they were asked to voluntarily participate in a research study. If they agreed, a member of the research team obtained informed consent, distributed the study questionnaire and conducted a semi-structured interview.

Measures

The data analysed for this investigation included information obtained from patients’ EMRs, chart reviews, self-report questionnaires and semi-structured interviews. Brighton Hospital provided a de-identified database that included relevant International Classification of Disease 10th edition (ICD-10) diagnoses (e.g. SUDs, psychiatric disorders and medical conditions, including WLS history) for the full hospital census. Chart review, survey and interview data were obtained only from formally enrolled participants, not the full hospital census.

Survey

Enrolled participants (n = 51) were asked to complete a survey assessing demographics, binge eating (Questionnaire on Eating and Weight Patterns-Revised; QEWP-R [13]) and substance use (Alcohol Use Disorders Identification Test Revised, AUDIT-R [14]). The investigators developed additional items to assess substance use trajectories, which included the age of onset of regular substance use, concern about substance use and age when participant first sought treatment for an SUD. The QEWP-R is a 28-item measure designed to identify individuals who may meet criteria for bulimia nervosa and binge eating disorder (BED). The items were modified to assess eating behaviour retrospectively, covering the 6 months prior to WLS. The QEWP-R has adequate internal consistency (alpha = 0.75) and test–retest reliability (kappa = 0.58) [15], although these statistics may not apply to the modified (retrospective) version used in this study. The AUDIT-R is a 10-item measure designed to assess alcohol and other SUDs. The original measure solely assessed AUDs; the AUDIT-R is a modified version that includes other drugs (i.e. cocaine/crack, other stimulants, heroin or other opiates, marijuana, tranquilizers, hallucinogens) to assess the use of a range of drugs. The original measure of the Alcohol Use Disorders Identification Test (AUDIT) has been extensively researched and has demonstrated high internal consistency and good test-retest reliability [16].

Semi-structured interview

Post-WLS participants (n = 56) participated in a brief (i.e. 30 to 60 min) semi-structured interview developed by the investigators. The interview was designed to understand the pre-surgical screening process (e.g. what type of assessments the patient received and if substance use was systematically assessed) and the events that occurred after surgery, with emphasis on the onset of SUDs. More detailed findings from that qualitative piece of this project have been published elsewhere [6].

Participants were classified as ‘SUD Hx+’ if they reported any suggestion of a significant history of illicit drug use or regular alcohol use prior to their WLS, reasoning that the magnitude of their current SUD problems would support that inference. Participants were classified as ‘New Onset’ if they denied all illicit drug use and reported only occasional alcohol use prior to surgery. The classification of New Onset SUD and SUD Hx+ status was obtained from two variables – a question asked during the semi-structured interview which states: ‘Do you feel like your problems with alcohol/drugs began after you had bariatric surgery? If yes, please describe how you began or increased your use of alcohol/drugs, and how you became concerned that it might be a problem for you’, and the substance use trajectory item in the questionnaires which asked for the patient's age at which they first engaged in regular substance use. To be conservative, given that in-patient SUD cases may deny or minimize problems that were present in the past but perhaps less severe than at present, we did not require the patients to acknowledge prior substance use to have been necessarily ‘problematic’ for us to classify them as SUD Hx+.

Chart review data were used to estimate quantity of alcohol consumed at the time of admission. Self-reported alcohol consumption per day was typically reported by patients as a range, e.g. ‘a pint to a fifth of vodka a day’. Therefore, the low and high end of each patient's daily range was converted to minimum and maximum number of standard drinks per drinking day.

Data analysis

For WLS patients identified in the EMR (n = 118), comparisons could be made with those not identified as WLS patients on variables exported and de-identified from the full EMR database.

For the enrolled participants (n = 129), t-tests and chi-square analyses were conducted to compare post-WLS SUD Hx+ and New Onset SUD patients on type and amount of substances used, age of first regular substance use, age of acknowledged concern over use, and age when first sought treatment. They were also compared on rates of current nicotine dependence and pre-surgical BED.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusions
  9. Conflict of Interest Statement
  10. Acknowledgements
  11. References

Prevalence of WLS and related characteristics in the full patient census (n = 4658)

During the 2 years for which EMR data were analyzed, a total of 4658 patients were admitted, 118 of whom were listed as having had WLS in the EMR. This yielded a prevalence rate of 2.5% of all admissions with a recorded diagnosis of having received WLS. This estimate, however, may understate the actual magnitude of the phenomenon, given that even after completing initial chart review and recruitment, an additional 11 cases emerged who had not originally been identified as post-WLS patients. Specifically, 11 additional patients volunteered to be interviewed, whereupon both their self reports and the hospital chart supported that they had WLS (which was not entered into the EMR), yielding a full sample of at least 129 WLS patients, for a more accurate estimate of at least 2.8% of admissions with a WLS history.

Compared with the total non-WLS SUD patient sample (n = 4540), post-WLS SUD patients (n = 118; i.e. the additional 11 WLS cases who volunteered on site could not be identified in the de-identified EMR database) were significantly more likely to be female (72.6% vs. 34.5%) and were significantly older (X = 45.43, SD = 9.73 vs. X = 39.28, SD = 14.06), t (4608) = −4.48, P < 0.001). Post-WLS patients were also significantly more likely to meet criteria for AUDs. Specifically, when compared with non-WLS cases, post-WLS patients were significantly more likely to be diagnosed with alcohol dependence: 68.8% of post-WLS patients were diagnosed with alcohol dependence, while only 54.6% of the non-WLS patients were so diagnosed χ2 (1, n = 4658) = 7.41, P < 0.01. It is also notable that among those with AUDs, 88.9% of WLS patients were treated for alcohol withdrawal, while only 70.1% of non-WLS patients presented with alcohol withdrawal χ2 (1, n = 2808) = 11.96, P < 0.001.

Enrolled post-WLS sample characteristics (n = 56)

Of the 129 WLS patients identified (either from the EMR or by patients simply volunteering when they learned about the study), 56 were recruited for the final study, yielding a 43.4% recruitment rate. The majority of patients at Brighton Hospital receive detoxification treatment for an average stay of 2 days to 2 weeks, and they are very occupied with SUD treatment programming for most of each day, leaving a very brief window for participant recruitment. On a typical day of programming, patients only had 1 h or 2 h of free time to participate in the current study.

Of the post-WLS sample (n = 56), 91.1% identified as White, 71.4% were female, 62.5% were married or living with a partner and 90.6% had undergone the RYGB procedure. Participants had a mean age of 44.80 years (standard deviation, SD = 9.49) and a current mean body mass index (BMI) of 31.18 (SD = 7.11). Consistent with previous findings, 60% of the post-WLS sample met criteria for New Onset SUD and 40% were classified as SUD Hx+ [6].

Comparison of enrolled WLS SUD Hx+ vs. New Onset cases (n = 51)

Additionally, significant differences emerged when comparing SUD Hx+ and New Onset SUD cases (see Table 1). Relative to SUD Hx+ cases, New Onset SUD cases reported a significantly later age of concern about substance use (X= 32.72, SD = 13.23 vs. X= 41.35, SD = 9.94), t (49) = −1.72, P < 0.05. On average, post-WLS patients reported it took them 1.6 years (SD = 1.62) after surgery to become concerned about their drug and/or alcohol problem. Among SUD Hx+ participants, 50% reported first seeking drug/alcohol treatment prior to seeking surgery. Interestingly, among SUD Hx+ participants, the average reported time sober prior to their surgery was 9.18 years (SD = 8.37), with 64.3% of SUD Hx+ participants reporting no substance use 5 years or more prior to surgery.

Table 1. Substance-related and eating behaviour characteristics of SUD Hx+ and New Onset SUD cases
Eating and substance variablesSUD Hx+ (n = 19)New Onset (n = 32)Significance
  1. *P < 0.05.

  2. †Values are expressed as n (%) or M ± SD.

  3. n = 51 except for the following variables: loss of control (n = 28), smoked last month (n = 43), cocaine (n = 48), other stimulants (n = 48), heroin/opiates (n = 50), marijuana (n = 49), tranquilizer (n = 49), hallucinogen (n = 50), AUDIT alcohol (n = 35), AUDIT drug (n = 28).

  4. AUDIT, Alcohol Use Disorders Identification Test; SUD, substance use disorder.

Preoperative binge eating14 (73.7%)13 (40.6%)*
Preoperative binge eating + loss of control over eating12 (85.7%)12 (85.7%)NS
Binge eating disorder11 (57.9%)9 (28.1%)*
Smoke ever15 (78.9%)16 (50%)*
Smoke in the last month10 (58.8%)13 (50%)NS
Alcohol use (current)16 (84.2%)23 (71.9%)NS
Cocaine use (current)1 (5.6%)1 (3.3%)NS
Stimulants (current)2 (12.5%)2 (6.3%)NS
Heroin or opiate use (current)10 (52.6%)12 (38.7%)NS
Marijuana use (current)7 (41.2%)3 (9.4%)*
Tranquilizer use (current)13 (72.2%)13 (41.9%)*
Hallucinogen use (current)0 (0%)0 (0%)NS
AUDIT alcohol score29.58 ± 7.6529.74 ± 9.38NS
AUDIT drug score25.69 ± 13.3925.47 ± 11.22NS

SUD Hx+ cases reported currently using significantly more types of substances than New Onset SUD cases, t (49) = 2.336, P < 0.05 (see Fig. 1). SUD Hx+ cases were significantly more likely to report currently using marijuana, tranquilizers and a history of ever smoking. SUD Hx+ cases were more likely to report a preoperative BED (see Table 1). No significant differences were observed between SUD Hx+ and New Onset cases on the minimum or maximum number of drinks per drinking day (P = 0.117). Both groups, however, were consuming very high quantities of alcohol: SUD Hx+ cases reported a mean of 15.97 (SD = 11.85) minimum drinks per drinking day and 22.51 (SD = 12.13) total maximum drinks per drinking day; New Onset cases reported a mean of 11.81 (SD = 6.11) minimum drinks per drinking day and 16.94 (SD = 10.10) maximum drinks per drinking day.

figure

Figure 1. Total number of substances used for SUD Hx+ vs. New Onset SUD groups.

Note: n = 51 (n = 19, n = 32, for SUD Hx+ and New Onset, respectively), *P < 0.05.

SUD, Substance use disorder.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusions
  9. Conflict of Interest Statement
  10. Acknowledgements
  11. References

Findings indicate that at least 2.8% of the total treatment seeking sample at Brighton Hospital was positive for WLS history, consistent with an earlier report from our research team [4]. As in that earlier report, again, however, it seems that this estimate underestimates the magnitude of this phenomenon. Specifically, WLS history was not entered into the EMR for 11 of the 56 WLS patients who were recruited for the present study; in other words, approximately 20% of identified cases failed to have this surgical history status entered into the EMR. Thus, there may have been additional patients with a WLS history who were not identified during the course of this study.

To put our 2.8% prevalence estimate in context, an estimate of the proportion of the general population that has had WLS would be helpful. Although the total number of WLSs performed to date is unknown, a recent report by Nguyen and colleagues [17] estimates that approximately 54.2 procedures were performed per 100 000 adults in 2008 (i.e. 0.0542%), and although rates of WLS have been increasing, it is difficult to envision how they could come close to accumulating to the 2.8% prevalence we found in this sample of patients entering SUD treatment. Therefore, the present study provides additional support that post-WLS patients – and particularly RYGB patients – are overrepresented in substance abuse treatment. Our estimate that 90.6% of SUD cases had the RYGB procedure can be contrasted with that observed by King and colleagues [3], who tracked rates of AUDs 2 years after WLS, and noted that of their 155 cases, 122 had the RYGB procedure. Excluding their three sleeve gastrectomy patients (because that procedure was rarely offered when our participants would have had their WLS) yields a conservative estimate of about 80% of AUD cases having had the RYGB procedure – lower than our estimate, but perhaps a function of our tracking other SUDs beyond AUDs. Regardless of how estimates might compare, however, the RYGB procedure accounts for the overwhelming majority of SUD cases across studies.

Although the RYGB procedure is the most commonly performed procedure in the United States and Canada, adjustable gastric banding procedures are also very common, as a close second [18], but fewer than 10% of our post-WLS patients had undergone adjustable gastric banding. Therefore, it appears that the RYGB procedure confers unique risk for post-surgical SUD, compared with the adjustable gastric banding procedure.

Post-WLS patients were significantly more likely to be treated for alcohol withdrawal than were their non-WLS counterparts. Contrary to the previous report by Saules and colleagues [4], WLS patients in this study were also more likely to be diagnosed with alcohol dependence. Therefore, results from the current investigation suggest that post-surgical alcohol use may confer high risk for the development of alcohol-related diagnoses. This observation of high rates of alcohol-related diagnoses is buttressed by experimental studies that suggest post-WLS patients have an increased sensitivity to alcohol. For instance, a recent case-crossover study [19] examined RYGB patients pre and post-operatively at 3 and 6 months follow-up and found that after WLS, patients had a breath alcohol content (BAC) well above the legal limit (i.e. above 0.08% BAC, assessed via alcohol metabolism testing with a breathalyzer) after drinking a single glass of wine. These findings are particularly salient given that the WLS patients in this study served as their own controls, affording a within subject examination of changes in WLS patients’ alcohol absorption pre- and post-surgery. Consistent with prior findings [20-22], the researchers found that the post-RYGB patients experienced a higher peak BAC and took longer to return to baseline than prior to surgery. Interestingly, these patients also reported an increased total number of symptoms of intoxication after surgery. In light of the physiological changes in the absorption and response to alcohol for post-bariatric patients, the finding that the number of drinks per day consumed by post-WLS patients was so high is particularly striking, given that post-WLS patients are more, not less, sensitive to alcohol than those who have not undergone WLS. Maluenda and colleagues [22] used a similar design to examine the effect of alcohol among a sample of WLS patients who received the laparoscopic sleeve gastrectomy procedure, and consistent with the aforementioned studies, alcohol absorption was significantly altered after surgery. There were a limited number of sleeve patients in the present study simply because when our participants had surgery, the sleeve procedure was not commonly available. Future research should explore the extent to which sleeve patients may be at risk for the emergence of post-WLS SUDs.

The unique aim of the present investigation, however, was to explore differences between New Onset SUD and SUD Hx+ groups. To our knowledge, this is the first investigation of post-WLS patients to systematically assess New Onset post-WLS substance use in an in-patient substance treatment programme setting. A recent investigation examined post-WLS nurses in a state monitoring programme for addiction and found similar rates of New Onset SUDs [23]. Specifically, they found that the majority (n = 17; 68%) of those who had WLS reported first developing problems post-surgery. Therefore, consistent with our earlier work [6-9] and that of others [23], growing evidence suggests that a majority of post-WLS SUD patients constitute the New Onset phenotype. Findings highlight the importance of better understanding New Onset SUDs after surgery.

Furthermore, our findings suggest that post-WLS, SUD does not have an immediate onset, but rather emerges during the second post-operative year or later. This is consistent with results of a recent large-scale report [3] that noted a significant increase in SUD prevalence between 1 and 2 years post-surgery. Overall, the majority of post-WLS patients in this sample were female, married or living with a partner, and many developed problems with alcohol and other drugs significantly later in life. These demographics are strikingly and significantly inconsistent with those observed for the ‘typical’ SUD in-patient; our post-WLS SUD patients were significantly older and more likely to be female than their non-WLS counterparts. Furthermore, as noted earlier, the majority of post-WLS patients developed problematic substance use and sought substance use treatment relatively soon after WLS. Consequently, in a group therapy setting (which is common for many substance abuse treatment programmes), WLS patients (particularly the New Onset group), many of whom only recently developed problematic use, may have trouble relating to other patients with longer and more pervasive histories of substance use and related lifestyles.

Additionally, among the SUD Hx+ cases, the average time sober prior to surgery was 9.18 years, with 64.3% of SUD Hx+ cases reporting no substance use 5 or more years prior to surgery. This suggests that the majority of our WLS patients with a pre-surgical history of substance use were not actively using substances during the pre-surgical process. However, given that many SUDs do not develop until patients are beyond regular contact with the WLS treatment team, more intensive patient education during the perioperative period is advisable. In addition, ongoing post-surgical assessment of substance use among WLS patients is recommended during routine follow-up visits with the WLS treatment team. This may afford an opportunity to intervene earlier in the progression to AUDs.

Interestingly, SUD Hx+ cases reported using significantly more types of substances than New Onset cases. In particular, SUD Hx+ cases were significantly more likely to engage in marijuana and tranquilizer use, in addition to reporting a greater likelihood of ever smoking cigarettes. Therefore, it seems that New Onset development is associated with less varied drug use and may be a more direct result of the absorption differences in alcohol after surgery. SUD Hx+ cases may have more experience with and exposure to different types of drugs, as they have a longer history of substance use. However, in terms of overall severity of substance abuse, the two groups appear similar, as there were no significant differences in the AUDIT total severity scores.

Of our total WLS sample, 39% met full criteria for BED, using retrospective recall prior to surgery. This is consistent with previous research [24] on retrospective reporting of BED among WLS candidates, which ranges from 37.5% to 49%. Our investigation did not directly examine whether preoperative BED predicts post-surgical substance use (because all of our patients were SUD cases). However, Guisado and Vaz Leal [25] compared morbidly obese WLS candidates with and without BED and found that preoperative binge eaters reported significantly more symptoms of alcohol dependence after surgery.

Interestingly, compared with New Onset cases, SUD Hx+ cases were significantly more likely to have met criteria for BED prior to WLS. Binge eating and ‘food addiction’ are distinct but overlapping constructs [26], and, as such, BED may be an indicator of a tendency towards ‘food addiction’. Given that SUD Hx+ cases, for the most part, were not heavily involved in substance use around the time of their surgery, they may have been instead using food as a coping strategy prior to seeking surgery and subsequently relapsed to (or continued to use) drugs or alcohol after surgery when food was no longer able to be ‘abused’. New Onset cases, however, may develop post-surgical addictions due to the physiological changes after surgery, and more research is necessary to examine what risk factors may relate to New Onset status. Future research should also examine pre-surgical eating behaviour, including both binge eating and ‘food addiction’, to determine if post-WLS patients are ‘transferring’ a food addiction to post-surgical substance use. While prospective studies have examined rates of lifetime and current rates of substance abuse among WLS patients [3], additional longitudinal studies in particular are needed to understand if disordered eating behaviour manifests differently after surgery or is ‘transformed’ into problematic substance use or another form of addiction. Future research is warranted to examine what risk factors may relate to New Onset SUD status. In particular, assessment of ‘food addiction’ and trajectories of substance use may be helpful to understand the unique emergence of post-surgical SUDs after surgery. However, given that all bypass patients have the same altered physiology as a by-product of surgery, it is not clear why only a subset of patients develop post-surgical SUDs. Therefore, further research is warranted to advance our understanding of which WLS patients are most at risk for the development of post-surgical SUDs.

Limitations

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusions
  9. Conflict of Interest Statement
  10. Acknowledgements
  11. References

The current investigation had several limitations. First, the assessments of several of the relevant variables were obtained through retrospective recall. Questions assessing age of SUD onset and other historical variables were asked retroactively through a self-report questionnaire. Ideally, future research should examine post-WLS patients through a prospective, long-term longitudinal study to help elucidate these findings. Additionally, the semi-structured interview, which determined New Onset and SUD Hx+ status, was created by the authors and lacks standardization. Lifetime history of drug or alcohol use could not be systematically evaluated with a well-validated assessment tool due to time constraints imposed by the intensive treatment setting; rather, we relied on the fact of their admission to an in-patient programme and diagnoses in the medical record to support the presence of severe SUDs. The history of those disorders, however, was limited to retrospective recall, but in a setting such as an in-patient detoxification/rehabilitation programme, there is relatively little motivation to minimize past problems in the way that there might be when presenting for WLS.

In addition, the New Onset vs. SUD Hx+ groups had relatively small sample sizes, so the failure to detect group differences on some variables may, in some cases, have been due to low statistical power. The findings that did emerge, however, are perhaps even more notable given the relative lack of statistical power.

Finally, because patients were from only one treatment setting and self-identified for the survey portion of the study, issues of generalizability must be considered. That is, patients who have the resources to obtain in-patient SUD treatment may not represent the full range of those struggling with alcohol and drug problems. Future research should aim to replicate our findings in broader SUD samples.

Conclusions

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusions
  9. Conflict of Interest Statement
  10. Acknowledgements
  11. References

Overall, findings suggest that post-WLS patients may be overrepresented in substance abuse treatment programmes, and the majority did not experience SUDs before WLS. In addition to the recent findings by King and colleagues [3], more longitudinal research is needed to advance our understanding of the trajectory of SUD development among WLS patients and better identify those at highest risk. Findings also highlight the importance of understanding the new onset emergence of SUDs after WLS. It appears that a subset of individuals who have undergone WLS experience the emergence of problematic substance use during a time in life that is uncommon. This New Onset group appears phenotypically distinct and may require unique treatment needs to appropriately address their SUD. Nonetheless, some but not all patients who have undergone WLS appear to develop SUDs. Therefore, future research to understand the risk factors associated with problematic substance use in post-WLS patients is warranted.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusions
  9. Conflict of Interest Statement
  10. Acknowledgements
  11. References

Preparation of this manuscript was supported by the Eastern Michigan University Department of Psychology and the EMU Graduate School. Ms. Wiedemann, Dr. Saules and Dr. Ivezaj conceived the project and assisted in data collection. Ms Wiedemann and Dr Saules analysed data. Ms Wiedemann conducted the literature review and generated the figure and table. All authors were involved in writing the paper and had final approval of the submitted and published version.

We thank Melissa Whelan and Rosa Quezada of Brighton Hospital for their assistance with participant recruitment and access to medical records. We thank the Eastern Michigan University Graduate School for supporting Ms Wiedemann and Dr Ivezaj's graduate studies, and students Summar Reslan, Alisha Serras, Randi Nguyen, Daniel Wood and Lorrianne Kuykendall for their considerable assistance with the data collection and data entry.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusions
  9. Conflict of Interest Statement
  10. Acknowledgements
  11. References
  • 1
    Ertelt TW, Mitchell JE, Lancaster K et al. Alcohol abuse and dependence before and after bariatric surgery: a review of the literature and report of a new data set. Surg Obes Relat Dis 2009; 45: 647650.
  • 2
    Odom J, Zalesin KC, Washington TL et al. Behavioral predictors of weight regain after bariatric surgery. Obes Surg 2010; 20: 349356.
  • 3
    King WC, Chen JY, Mitchell JE et al. Prevalence of alcohol use disorders before and after bariatric surgery. JAMA 2012; 307: 25162525.
  • 4
    Saules KK, Wiedemann AA, Ivezaj V et al. Bariatric surgery history among substance abuse treatment patients: prevalence and associated features. Surg Obes Relat Dis 2010; 6: 615621.
  • 5
    Conason A, Teixeira J, Hsu C et al. Substance use following bariatric weight loss surgery. Arch Surg 2012; 15: 16.
  • 6
    Ivezaj V, Saules KK, Wiedemann AA. “I didn't see it coming” Why are post-bariatric surgery patients in substance abuse treatment? Patients’ perceptions of etiology and future recommendations. Obes Surg 2012; 22: 13081314.
  • 7
    Wiedemann AA, Saules KK, Ivezaj V et al. An examination of post-bariatric patients who develop problematic substance use after surgery: new onset users compared to controls. Ann Behav Med 2011; 41S: s122.
  • 8
    Wiedemann AA, Saules KK, Hopper JA et al. Substance use trajectories of post-bariatric surgery patients enrolled in substance abuse treatment. Ann Behav Med 2010; S39: s38.
  • 9
    Ivezaj V. An examination of psychological risk factors for the development of substance abuse among post-bariatric surgery patients [dissertation]. Ypsilanti (MI): Eastern Michigan University. 2012.
  • 10
    Reslan S. Relationships between food reinforcement and eating behaviors to bariatric surgery weight loss and substance abuse outcomes [dissertation]. Ypsilanti (MI): Michigan: Eastern Michigan University. 2012.
  • 11
    Kessler RC, Berglung P, Demler O et al. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the national comorbidity survey replications. Arch Gen Psychiatry 2005; 62: 593602.
  • 12
    Li T, Hewitt BG, Grant BF. Alcohol use disorders and mood disorders: a National Institute on Alcohol Abuse and Alcoholism perspective. Biol Psychiatry 2004; 56: 718720.
  • 13
    Spitzer RL, Yanovski SZ, Marcus MD. Questionnaire on Eating and Weight Patterns – Revised. Behavioral Measurement and Database Services (Producer) BRS Search (Vendor): McLean, VA, 1994.
  • 14
    Babor TF, de la Fuente JR, Saunders J, Grant M. AUDIT. The Alcohol Use Disorders Identification Test. Guidelines for Use in Primary Health Care. World Health Organization: Geneva, Switzerland, 1992.
  • 15
    Nangle DW, Johnson WG, Carr-Nangle RE, Engler LB. Binge eating disorder and the proposed DSM-IV-TR criteria: psychometric analysis of the questionnaire of eating and weight patterns. Int J Eat Disord 1994; 16: 147157.
  • 16
    Reinert DF, Allen JP. The alcohol use disorders identification test: an update of research findings. Alcohol Clin Exp Res 2007; 31: 185199.
  • 17
    Nguyen NT, Masoomi H, Magno CP et al. Trends in use of bariatric surgery, 2003–2008. J Am Coll Surg 2011; 213: 261266.
  • 18
    Buchwald H, Oien DM. Metabolic/bariatric surgery worldwide 2008. Obes Surg 2009; 19: 16051611.
  • 19
    Woodard GA, Downey J, Hernandez-Boussard T, Morton JM. Impaired alcohol metabolism after gastric bypass surgery: a case-crossover trial. J Am Coll Surg 2010; 212: 209214.
  • 20
    Hagedorn JC, Encarnacion B, Brat GA, Morton JM. Does gastric bypass alter alcohol metabolism? Surg Obes Relat Dis 2007; 3: 543548.
  • 21
    Klockhoff H, Naslund I, Jones AW. Faster absorption of ethanol and higher peak concentration in women after gastric bypass surgery. Br J Clin Pharmacol 2002; 54: 587591.
  • 22
    Maluenda F, Csendes A, De Aretxabala X et al. Alcohol absorption modification after a laparoscopic sleeve gastrectomy due to obesity. Obes Surg 2010; 20: 744748.
  • 23
    Fogger SA, McGuiness TM. The relationships between addictions and bariatric surgery for nurses in recovery. Perspect Psychiatr Care 2011; 48: 1015.
  • 24
    Niego SH, Kofman MD, Weiss JJ, Geliebter A. Binge eating in the bariatric surgery population: a review of the literature. Int J Eat Disord 2007; 40: 349359.
  • 25
    Guisado-Macías JA, Vaz Leal FJ. Psychopathological differences between morbidity obese binge eaters and non-binge eaters after bariatric surgery. Eat Weight Disord 2008; 8: 315318.
  • 26
    Gearhardt AN, Corbin WR, Brownell KD. Preliminary validation of the Yale Food Addiction Scale. Appetite 2009; 52: 430436.