Abstract
Background and Aims:
Although bariatric surgery is an effective intervention for severe obesity, a subset of patients demonstrates suboptimal weight outcomes. Postoperative loss of control eating (LOCE) and binge eating may influence weight outcomes, though research has not examined differences by surgical procedure, or factors that predict postoperative LOCE. This study aimed to: (1) characterize LOCE and binge eating disorder (BED) over a seven-year period following bariatric surgery; (2) examine concurrent, prospective, and cumulative relationships between LOCE and weight loss; (3) assess whether these associations are moderated by surgery type; and (4) evaluate predictors of LOCE.
Methods:
Participants were 2,156 patients who underwent laparoscopic adjustable gastric banding (LAGB) or Roux-n-Y gastric bypass (RYGB) in the multi-center Longitudinal Assessment of Bariatric Surgery-2 (LABS-2) study. Generalized linear mixed models examined relationships between LOCE and percent weight loss and predictors of LOCE.
Results:
LOCE and BED initially declined then increased after surgery, with a notable number of de-novo cases. LOCE was related to less concurrent but not prospective or cumulative percent weight loss. Self-monitoring of eating, higher daily eating frequency, older age, male gender, and higher self-esteem were associated with a lower likelihood of LOCE.
Conclusions:
Results suggest that LOCE and binge eating are clinically relevant behaviors that may impede weight loss, and findings highlight the importance of ongoing assessment of maladaptive eating following surgery.
Keywords: bariatric surgery, gastric bypass, gastric banding, binge eating disorder, loss of control eating, weight loss
Introduction
Although most bariatric surgery patients maintain substantial weight loss at long-term follow-up, a subset evidences poor weight outcomes.(1–3) Maladaptive eating behaviors, particularly loss of control and binge eating, are factors that may impede successful long-term weight loss.(4) Loss of control eating, LOCE (i.e., feeling one cannot stop eating or control how much one is eating), is a central feature of binge eating (i.e., the consumption of an objectively large amount of food accompanied by a sense of loss of control) and binge eating disorder (BED), which was introduced in the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM) and is characterized by recurrent binge eating in the absence of regular compensatory behaviors.(5)
Both binge eating and BED are more common among individuals with obesity compared to normal weight individuals, and prior research has documented that 43.4% of bariatric surgery candidates report LOCE.(6, 7) While initial decreases in LOCE and binge eating occur following surgery, many individuals continue to endorse or develop these behaviors.(4) Although objective binge eating episodes are often attenuated because the surgery constrains the amount of food patients can consume, over time some appear able to consume larger amounts.(8)
LOCE and binge eating are especially concerning in bariatric populations due to possible attenuation of weight loss.(4, 9) Though the presence of LOCE prior to surgery has not been consistently related to postoperative weight outcomes, converging evidence indicates that LOCE, binge eating, and/or BED following surgery is related to less weight loss or weight regain.(4, 9–12). Some research has also suggested that preoperative BED status is related to attenuated weight loss at 24, but not 12, months after surgery, which demonstrates the importance of examining longer-term outcomes.(13, 14)
There is also limited research examining whether the relationship between LOCE and postoperative weight change varies according to surgical procedure (10), and studies have not examined what factors may predict postoperative LOCE. Previous research found that eating more frequently, taking medication for psychiatric/emotional problems, alcohol use disorder symptoms, lower self-esteem, and depressive symptoms were associated with pre-surgery BED,(7) and thus may be relevant to postoperative LOCE. In addition, successful weight loss has been linked to self-monitoring of weight-related behavior, which may be indicative of self-regulatory capacity.(12, 15) Given that LOCE/binge eating are broadly characterized by self-regulatory difficulties (16–18), a lack of self-monitoring may be related to postoperative LOCE.
Given this evidence, the objectives of this study were to: (1) characterize LOCE and BED in the seven years following bariatric surgery; (2) examine concurrent, prospective, and cumulative relationships between LOCE and weight loss; (3) assess whether these associations differ by surgery type (i.e., laparoscopic adjustable gastric banding [LAGB] or Roux-n-Y gastric bypass [RYGB]); and (4) evaluate predictors of LOCE. It was hypothesized that following initial postoperative declines, LOCE and BED would become more prevalent over time, and that LOCE would be related to less weight loss. It was also expected that less self-monitoring behaviors and greater psychiatric symptoms would be related to greater LOCE.
Materials and Methods
Participants and Procedure
Participants were 2,156 patients (76% female) recruited from 2006–2009 and followed until 2015 from the Longitudinal Assessment of Bariatric Surgery–2 (LABS-2) study who underwent RYGB (n=1,640) or LAGB (n=516). Given the small number of gastric sleeve patients (n=59), these participants were excluded. Patients were assessed prior to bariatric surgery (month 0) and annually to biannually for up to seven years after surgery (months 12, 24, 36, 48, 60, 84). Year 6 (month 72) was excluded due to lack of LOCE and BED assessments. See Courcoulas et al.(1) for a description of the seven-year weight change.
Self-report Measures
LOCE was indicated by a response of “yes” to the following questionnaire items: “During the past 6 months, have you had times when you eat continuously during the day or parts of the day without planning what and how much you would eat?” and “did you experience a loss of control; that is you felt like you could not control your eating?”. BED was evaluated by items assessing the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) criteria analogous to prior LABS-2 studies.(5, 7)
Alcohol use disorder symptoms were assessed by the Alcohol Use Disorders Identification Test (AUDIT)(19); the AUDIT has adequate reliability and validity and has been used to screen for alcohol use disorders in bariatric samples.(20) Depressive symptoms were measured using the Beck Depression Inventory (BDI) version 1 (21), which has demonstrated adequate sensitivity and specificity in bariatric patients.(22) Self-esteem was measured using the Impact of Weight on Quality of Life–Lite (IWQOL-Lite) self-esteem subscale (higher T scores=lower self-esteem); the IWQOL-Lite evidences excellent psychometric properties in patients with obesity and is commonly used in bariatric samples.(23–26) In addition, items assessed whether participants had received psychotherapy or psychiatric treatment since the last assessment, daily frequency of meals/snacks, and whether the participant reported self-weighing (i.e., keeping a graph of weight), recording daily exercise, and recording daily intake over the previous six months.
Statistical Analyses
Analyses used available data without imputation. Patients were grouped into one of four categories for LOCE and BED: Never (not endorsed pre-surgery or at any follow-up assessment), de novo (not endorsed pre-surgery but endorsed at one or more follow-up assessments), recurrent (endorsed pre-surgery and at one or more follow-up assessments), and remitted (endorsed pre-surgery but not at any follow-up).
Concurrent and prospective relationships between LOCE and weight were assessed with generalized linear mixed models. Weight outcome was operationalized as percent of pre-surgery weight loss: weight at the assessment – pre-surgery weight)/pre-surgery weight. The concurrent model included LOCE status over the past 6 months (yes/no), surgery type (RYGB/LAGB), and the interaction between LOCE status and surgery type as predictors of percent weight loss. The prospective model included LOCE status at the prior assessment (i.e. lagged), surgery type, and the lagged LOC status x surgery type interaction as predictors of percent weight loss. Each model included percent weight loss at the previous assessment and pre-surgery LOCE status as covariates. The cumulative effect of LOCE on percent weight loss was also evaluated, in which the total number of LOCE endorsements over the follow-up period for each participant was calculated and entered as a predictor of percent weight loss (measured at the last available assessment) using a generalized linear model.
To examine predictors of LOCE at each time point (1=LOCE; 0=No LOCE), generalized linear mixed models with binomial link functions were used. Predictors of concurrent and subsequent (measured at the next assessment) LOCE were first examined at the univariate level; however, the three self-monitoring variables (i.e., self-weighing, recording eating, recording exercise) were entered simultaneously. Significant predictors were retained and entered into two multivariate models: one assessed predictors of concurrent LOCE, while the second assessed prospective predictors of LOCE. Given that surgery site, age, and gender were found to be related to the number of visits completed, these variables were also added to the aforementioned models as covariates and examined as predictors of LOCE.
Results
LOCE and BED
Table 1 displays sample characteristics, and Table 2 depicts percent weight loss and LOCE and BED at each assessment. LOCE and BED prevalence was highest before surgery (35.0% and 12.7%, respectively), with the greatest decreases observed in the first year. LOCE peaked at 36 months (31.7%) and declined thereafter, though the proportion of patients with BED increased each year to 4.0% by 84 months. With respect to stability and change in LOCE, 39.9% never endorsed LOCE, 25.6% reported de novo LOCE, 25.4% reported recurrent LOCE, and 9.7% reported remitted LOCE. Regarding BED, 82.2% never met criteria, 4.8% reported de novo BED, 3.8% reported recurrent BED, and 9.2% reported remitted BED. However, of the 9.2% remitted BED patients, 46.6% still reported LOCE following surgery.
Table 1.
Pre-surgery sample characteristics and visits completed
| M | SD | Minimum | Maximum | |
|---|---|---|---|---|
| Age | 45.66 | 11.32 | 19 | 78 |
| BMI | 47.06 | 7.36 | 32.85 | 87.04 |
| Weight (lbs) | 293.45 | 56.82 | 165 | 554 |
| AUDIT | 3.49 | 2.11 | 2.00 | 17.00 |
| BDI | 8.96 | 7.26 | 0.00 | 50.00 |
| IWQOL-Lite Self-Esteem | 41.40 | 27.17 | 0.00 | 100.00 |
| Visits completed | 4.90 | 1.95 | 1.00 | 7.00 |
Note. AUDIT=Alcohol Use Identification Test; BDI=Beck Depression Inventory; IWQOL-Lite= Impact of Weight on Quality of Life–Lite (T-scores).
Table 2.
Percent weight loss, LOCE, and BED prior to and following surgery
| Sample | Month | N | Percent weight loss (SD) |
% with LOCE | % with BED | |
|---|---|---|---|---|---|---|
| RYGB | 0 | 1641 | - | - | 33.50 | 12.10 |
| 12 | 1343 | −33.84 | (7.77) | 21.90 | 1.30 | |
| 24 | 1193 | −34.26 | (9.44) | 29.40 | 2.00 | |
| 36 | 1143 | −31.99 | (9.75) | 31.00 | 2.70 | |
| 48 | 1128 | −30.09 | (10.34) | 29.90 | 3.50 | |
| 60 | 1145 | −28.82 | (10.42) | 28.70 | 3.00 | |
| 84 | 812 | −28.13 | (10.86) | 25.60 | 3.30 | |
| LAGB | 0 | 516 | - | - | 39.70 | 14.6 |
| 12 | 431 | −15.86 | (8.28) | 32.90 | 4.50 | |
| 24 | 385 | −17.23 | (9.88) | 32.70 | 4.90 | |
| 36 | 364 | −16.70 | (11.17) | 33.80 | 5.00 | |
| 48 | 352 | −15.35 | (11.10) | 36.60 | 5.10 | |
| 60 | 357 | −14.94 | (11.73) | 36.40 | 5.80 | |
| 84 | 237 | −14.50 | (12.03) | 29.10 | 6.60 | |
| Total sample | 0 | 2157 | - | - | 35.00 | 12.70 |
| 12 | 1774 | −29.45 | (11.04) | 24.60 | 2.10 | |
| 24 | 1578 | −30.03 | (12.06) | 30.20 | 2.70 | |
| 36 | 1507 | −28.20 | (12.08) | 31.70 | 3.30 | |
| 48 | 1480 | −26.58 | (12.25) | 31.50 | 3.90 | |
| 60 | 1502 | −25.62 | (12.22) | 30.60 | 3.70 | |
| 84 | 1049 | −25.26 | (12.42) | 26.40 | 4.00 | |
Note. LOCE=loss of control eating; BED=binge eating disorder; RYGB= Roux-n-Y gastric bypass; LAGB=laparoscopic adjustable gastric banding.
More negative percent weight loss values indicate greater weight loss.
LOCE and Weight Change
Results of the concurrent mixed model indicated a significant association between LOCE and percent weight loss (Table 3), in which participants who reported LOCE over the last six months experienced 1.7% less percent weight loss since the previous assessment compared with participants who did not report LOCE, irrespective of surgery type. There was a significant difference between surgery type in percent weight loss demonstrating greater percent weight loss in the RYGB group, though no significant interaction between LOCE and surgery type. The prospective model also demonstrated significant difference between surgery type in percent weight loss, indicating greater percent weight loss in the RYGB group. There were no significant effects of prior LOCE (measured at the previous assessment) or the interaction between prior LOCE and surgery type on percent weight loss. In both models, age at the time of surgery and gender were also significant predictors, such that older participants and males evidenced greater percent weight loss (ps<.001). There was not a significant cumulative effect of LOCE (i.e., total number of LOCE endorsements over the seven-year follow-up) on percent weight loss (B=0.17, SE=.17, p=.322) after adjusting for surgery site, age, and gender.
Table 3.
Mixed model results assessing LOCE as a predictor of concurrent and prospective percent weight loss
| 95% Confidence Interval |
|||||||
|---|---|---|---|---|---|---|---|
| Concurrent model |
B | SE | df | t | p | Lower | Upper |
| Intercept | −19.47* | 2.22 | 1912.53 | −8.86 | <0.001 | −23.83 | −15.11 |
| Visit | 0.09* | <0.01 | 7058.32 | 21.11 | <0.001 | 0.08 | 0.10 |
| Age | 0.08* | 0.02 | 1844.25 | 4.81 | <0.001 | 0.05 | 0.12 |
| Gender | −2.08* | 0.47 | 1865.92 | −4.39 | <0.001 | −3.01 | −1.15 |
| Surgery site | <0.01 | 0.01 | 1858.00 | 0.68 | 0.497 | −0.01 | 0.02 |
| Surgery type | −15.40* | 0.56 | 2809.17 | −27.37 | <0.001 | −16.51 | −14.30 |
| Pre-surgery LOCE | 0.82* | 0.41 | 1861.61 | 2.01 | 0.045 | 0.02 | 1.62 |
| Lag percent weight loss | 0.05* | <0.01 | 5388.49 | 11.37 | <0.001 | 0.04 | 0.06 |
| LOCE | −1.68* | 0.32 | 5634.87 | −5.29 | <0.001 | −2.31 | −1.06 |
| Surgery type X LOCE | 0.13 | 0.36 | 5670.54 | 0.37 | 0.709 | −0.57 | 0.84 |
| Prospective model | |||||||
| Intercept | −20.32* | 2.24 | 1905.51 | −9.06 | <0.001 | −24.72 | −15.92 |
| Visit | 0.09* | <0.01 | 7050.74 | 20.42 | <0.001 | 0.08 | 0.10 |
| Age | 0.08* | 0.02 | 1849.56 | 4.82 | <0.001 | 0.05 | 0.12 |
| Gender | −2.03* | 0.48 | 1871.19 | −4.22 | <0.001 | −2.97 | −1.08 |
| Surgery site | <0.01 | 0.01 | 1863.49 | 0.66 | 0.508 | −0.01 | 0.02 |
| Surgery type | −15.48* | 0.55 | 2597.31 | −27.94 | <0.001 | −16.57 | −14.39 |
| Pre-surgery LOCE | 0.68 | 0.42 | 1998.70 | 1.62 | 0.105 | −0.14 | 1.51 |
| Lag percent weight loss | 0.04* | <0.01 | 5371.50 | 9.72 | <0.001 | 0.04 | 0.05 |
| Lag LOCE | −0.50 | 0.30 | 5466.50 | −1.65 | 0.098 | −1.09 | 0.09 |
| Surgery type X Lag LOCE | 0.08 | 0.34 | 5605.21 | 0.24 | 0.807 | −0.59 | 0.76 |
Note. Visit=time (months) following surgery. LOCE=loss of control eating; Lag=measured at the prior assessment; Surgery type (RYGB vs. LAGB) was coded such that LAGB was the reference category; LOCE (yes vs. no) was coded such that no LOC was the reference category; gender was coded such that male was the reference category.
More negative percent weight loss values indicate greater weight loss.
p<.05.
Predictors of LOCE
Concurrent predictors.
At the univariate level, higher age, higher BDI scores, higher self-esteem, higher eating frequency, recording eating, taking medications for emotional/psychiatric problems, and engagement in psychiatric treatment were related to a lower likelihood of LOCE, while recording exercise was related to a higher likelihood of LOCE (ps<.001); additionally, LOCE differed significantly across surgery sites (p<.01), and females were significantly more likely to report LOCE than males (p<.001). There were no significant effects of self-weighing or AUDIT scores (Table 4). In the multivariate model including all significant univariate predictors, all of the effects except for psychiatric treatment, exercise self-monitoring, and surgery site remained significant.
Table 4.
Univariate and multivariate predictors of concurrent LOCE
| 95% Confidence Interval |
||||||
|---|---|---|---|---|---|---|
| Univariate models | B | SE | t | p | Lower | Upper |
| Intercept | 1.21 | 1.58 | 0.77 | 0.443 | −1.88 | 4.30 |
| Age | −0.01* | <0.01 | −4.45 | <0.001 | −0.01 | −0.01 |
| Intercept | 0.70 | 1.58 | 0.44 | 0.658 | −2.39 | 3.79 |
| Gender | 0.43* | 0.06 | 7.81 | <0.001 | 0.32 | 0.54 |
| Intercept | 0.82 | 1.56 | 0.53 | 0.600 | −2.25 | 3.89 |
| AUDIT | −0.01* | 0.01 | −1.25 | 0.213 | −0.02 | 0.01 |
| Intercept | 1.18 | 1.55 | 0.76 | 0.446 | −1.86 | 4.22 |
| BDI | −0.06* | <0.01 | −19.30 | <0.001 | −0.06 | −0.05 |
| Intercept | 0.84 | 1.58 | 0.54 | 0.593 | −2.25 | 3.94 |
| Self-weighing | −0.01 | 0.01 | −1.07 | 0.285 | −0.03 | 0.01 |
| Recording eating | −0.17* | 0.03 | −6.37 | <0.001 | −0.23 | −0.12 |
| Recording exercise | 0.08* | 0.02 | 3.43 | 0.001 | 0.04 | 0.13 |
| Intercept | 0.91 | 1.56 | 0.59 | 0.558 | −2.14 | 3.97 |
| Medication | −0.33* | 0.04 | −7.94 | <0.001 | −0.41 | −0.25 |
| Intercept | 0.82 | 1.56 | 0.53 | 0.600 | −2.24 | 3.89 |
| Psychiatric treatment | −0.21* | 0.04 | −5.15 | <0.001 | −0.29 | −0.13 |
| Intercept | −0.43 | 1.53 | −0.28 | 0.78 | −3.43 | 2.57 |
| Self-esteem | 0.02* | <0.01 | 25.93 | <0.001 | 0.02 | 0.02 |
| Intercept | 1.34 | 1.57 | 0.85 | 0.394 | −1.74 | 4.42 |
| Eating frequency | −0.12* | 0.01 | −11.66 | <0.001 | −0.14 | −0.10 |
| Intercept | 1.11 | 1.59 | 0.70 | 0.484 | −2.00 | 4.23 |
| Surgery site | <0.01* | <0.01 | −1.98 | 0.047 | <0.01 | <0.01 |
| Multivariate model | ||||||
| Intercept | 0.49 | 1.50 | 0.33 | 0.742 | −2.44 | 3.43 |
| Psychiatric treatment | −0.06 | 0.04 | −1.39 | 0.165 | −0.15 | 0.03 |
| Eating frequency | −0.11* | 0.01 | −8.10 | <0.001 | −0.13 | −0.08 |
| Self-esteem | 0.02* | <0.01 | 17.34 | <0.001 | 0.02 | 0.02 |
| Medication | −0.13* | 0.04 | −2.97 | 0.003 | 0.02 | 0.02 |
| Recording exercise | 0.04 | 0.05 | 0.83 | 0.407 | −0.06 | 0.14 |
| Recording eating | −0.24* | 0.06 | −4.31 | <0.001 | −0.34 | −0.13 |
| BDI | −0.03* | <0.01 | −6.16 | <0.001 | −0.03 | −0.02 |
| Gender | 0.31* | 0.07 | 4.33 | <0.001 | 0.17 | 0.45 |
| Surgery site | <0.01 | <0.01 | 1.54 | 0.125 | <0.01 | <0.01 |
Note. LOCE=loss of control eating; AUDIT=Alcohol Use Identification Test; BDI=Beck Depression Inventory. Self-esteem was measured by the Impact of Weight on Quality of Life–Lite (higher T-scores indicate lower self-esteem). Gender was coded such that male=reference group.
p<.05.
Prospective predictors.
Regarding predictors of subsequent LOCE, higher age, higher BDI scores, higher self-esteem, higher eating frequency, recording eating, self-weighing, medication use, and psychiatric treatment were related to a lower likelihood of LOCE (ps<.05); and females were significantly more likely to report LOCE than males (p<.001). There were no significant effects of recording exercise, surgery site, or AUDIT scores (Table 5). In the multivariate model, all of the effects except for psychiatric treatment and medication use remained significant.
Table 5.
Univariate and multivariate predictors of subsequent LOCE
| 95% Confidence Interval |
||||||
|---|---|---|---|---|---|---|
| Univariate models | Estimate | SE | t | p | Lower | Upper |
| Intercept | 1.17* | 0.14 | 8.30 | <0.001 | 0.89 | 1.44 |
| Age | −0.01* | <0.01 | −3.37 | 0.001 | −0.01 | <0.01 |
| Intercept | 0.68* | 0.05 | 14.07 | <0.001 | 0.58 | 0.77 |
| Gender | 0.52* | 0.08 | 6.63 | <0.001 | 0.37 | 0.68 |
| Intercept | 0.77* | 0.09 | 8.65 | <0.001 | 0.60 | 0.95 |
| AUDIT | −0.01 | 0.01 | −1.24 | 0.215 | −0.03 | 0.01 |
| Intercept | 1.02* | 0.07 | 15.47 | <0.001 | 0.89 | 1.15 |
| BDI | −0.05* | <0.01 | −10.83 | <0.001 | −0.06 | −0.04 |
| Intercept | 0.93* | 0.08 | 12.38 | <0.001 | 0.78 | 1.08 |
| Self-weighing | −0.03* | 0.01 | −2.11 | 0.035 | −0.06 | <0.01 |
| Recording eating | −0.25* | 0.06 | −4.32 | <0.001 | −0.37 | −0.14 |
| Recording exercise | 0.04 | 0.05 | 0.70 | 0.427 | −0.06 | 0.15 |
| Intercept | 0.84* | 0.05 | 16.94 | <0.001 | 0.74 | 0.94 |
| Medication | −0.21* | 0.05 | −4.37 | <0.001 | −0.30 | −0.11 |
| Intercept | 0.76* | 0.05 | 16.64 | <0.001 | 0.70 | 0.89 |
| Psychiatric treatment | −0.17* | 0.05 | −3.37 | <0.001 | −0.27 | −0.07 |
| Intercept | −0.45* | 0.10 | −4.71 | <0.001 | −0.64 | −0.26 |
| Self-esteem | 0.02* | <0.01 | 14.14 | <0.001 | 0.01 | 0.02 |
| Intercept | 1.32* | 0.09 | 14.89 | <0.001 | 1.15 | 1.49 |
| Eating frequency | −0.11* | 0.02 | −7.53 | <0.001 | −0.14 | −0.08 |
| Intercept | 0.82* | 0.24 | 3.42 | 0.001 | 0.35 | 1.29 |
| Surgery site | <0.01 | <0.01 | −0.21 | 0.833 | <0.01 | <0.01 |
| Multivariate model | ||||||
| Intercept | 1.11* | 0.20 | 5.44 | <0.001 | 0.71 | 1.50 |
| Psychiatric treatment | −0.06 | 0.06 | −1.11 | 0.267 | −0.17 | 0.05 |
| Eating frequency | −0.11* | 0.02 | −6.54 | <0.001 | −0.14 | −0.08 |
| Self-esteem | 0.02* | <0.01 | 10.08 | <0.001 | 0.01 | 0.02 |
| Medication | −0.09 | 0.05 | −1.71 | 0.088 | −0.19 | 0.01 |
| Recording eating | −0.22* | 0.06 | −3.66 | <0.001 | −0.34 | −0.10 |
| BDI | −0.02* | 0.01 | −4.40 | <0.001 | −0.03 | −0.01 |
| Age | −0.02* | <0.01 | −5.50 | <0.001 | −0.02 | −0.01 |
| Gender | 0.44* | 0.09 | 4.93 | <0.001 | 0.26 | 0.61 |
Note. LOCE=loss of control eating; AUDIT=Alcohol Use Identification Test; BDI=Beck Depression Inventory. Self-esteem was measured by the Impact of Weight on Quality of Life–Lite (higher T-scores indicate lower self-esteem). Gender was coded such that male=reference group.
p<.05.
Conclusion
This study extends prior work by examining LOCE and BED seven years after surgery, considering the effect of LOCE on weight across surgical procedures, and assessing predictors of LOCE. Although LOCE and BED initially declined, a substantial proportion of patients reported de novo postoperative LOCE and BED (approximately 25% and 5%, respectively), and almost half of those with remitted BED continued to report LOCE. One possibility is that individuals minimized problematic eating prior to surgery, or it may be that over time individuals have more difficulty adhering to postoperative dietary guidelines, and as a result experience a sense of loss of control. These data suggest it would be beneficial to conduct routine assessments of problematic eating behaviors well beyond the first postoperative year. Results also highlight the importance of assessing a range of behaviors rather than categorical diagnoses, especially given that the objective binge episode BED criterion may be problematic in this population.
The finding that LOCE is indicative of less concurrent weight loss, but not weight loss over subsequent or prolonged time periods (as assessed by the prospective and cumulative models) may suggest that LOCE has a more proximal effect on weight. It could be that over an extended time, additional factors exert more potent influences on weight. The lack of significant interactions also indicates that this effect appears to be independent of surgical procedure.
Importantly, both concurrent and prospective multivariate models suggested older age, self-monitoring of eating, more frequent eating, and higher self-esteem were related to lower likelihoods of current and future LOCE, and that females were more likely to report LOCE than males. Monitoring of dietary intake may be particularly relevant to self-regulation in this population, and may help prevent LOCE. Given that eating smaller portions more frequently is recommended following surgery, more frequent meals/snacks could be an indicator of better adherence to postsurgical guidelines, and perhaps another marker of self-regulatory capacity. Concurrent medication use for emotional/psychiatric problems was also related to lower LOCE likelihood; it could be that such interventions help to reduce negative affect, which is consistently implicated in the onset and maintenance of binge eating.(27, 28) Interestingly, higher depression was related to lower LOCE likelihood, contrary to expectations. It could be that some depressive symptoms (i.e., poor appetite) are related to decreases in eating, and/or BDI scores may be inflated by medical or somatic complaints in this population.(29) Further examination of specific depression symptoms may help clarify this.
It is important to note limitations of the present study, including the operationalization and assessment of LOCE and binge eating following surgery. The “objectively large” food amount criterion becomes problematic because bariatric surgery inherently limits the amount one can consume, and experiences of LOCE may be altered by surgery.(30) The measurement of LOCE, while used in prior bariatric research (7), may overlap with grazing, which is a distinct construct that is also associated with poor weight outcome.(9) LOCE was measured as a dichotomous variable, and thus range restriction may be problematic; given evidence that this construct varies in degree within and between persons, future studies should utilize validated, dimensional measures of LOCE.(31) Additionally, the data with the exception of weight were self-reported, and interview-based assessments may be more precise. The relationship between LOCE and weight assessed in the concurrent model cannot determine directionality, and it may be that individuals who experience less weight loss perceive their eating to be out of control. Finally, given that the large proportion of LAGB in the present sample may not reflect current national trends, the relevance of findings should be interpreted with this in mind.
Summary
Collectively, findings demonstrate the importance of post-surgical LOCE and BED in this population, and notably, these behaviors present de novo in a significant minority of postoperative patients. Future research should also examine other potential trait- and state-level risk factors for LOCE, such as temperament, emotion regulation, and momentary affect. Lastly, results highlight the need for long-term monitoring of LOCE and binge eating in the years following surgery, as well as development of prevention and intervention efforts targeting such behaviors.
Acknowledgments, Funding Information
LABS-2 was funded by a cooperative agreement by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) grants James.Mitchell@SanfordHealth.org (University of Pittsburgh, Data Coordinating Center); U01-DK66667 and UL1-RR024996 (Columbia- Presbyterian in collaboration with Cornell University Medical Center Clinical and Translational Research Center [CTRC]); U01-DK66568 and M01RR-00037 (University of Washington in collaboration with Cornell University Medical Center CTRC); U01-DK66471 (Neuropsychiatric Research Institute); U01-DK66526 (East Carolina University); U01-DK66585 and UL1- RR024153 (University of Pittsburgh Medical Center in collaboration with Cornell University Medical Center CTRC); and U01-DK66555 (Oregon Health & Science University).
Disclosures and Statements of Human Rights
LABS-2 was funded by a cooperative agreement by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) grants U01 DK066557 (University of Pittsburgh, Data Coordinating Center); U01-DK66667 and UL1-RR024996 (Columbia- Presbyterian in collaboration with Cornell University Medical Center Clinical and Translational Research Center [CTRC]); U01-DK66568 and M01RR-00037 (University of Washington in collaboration with Cornell University Medical Center CTRC); U01-DK66471 (Neuropsychiatric Research Institute); U01-DK66526 (East Carolina University); U01-DK66585 and UL1- RR024153 (University of Pittsburgh Medical Center in collaboration with Cornell University Medical Center CTRC); and U01-DK66555 (Oregon Health & Science University). This research was registered at ClinicalTrials.gov (Identifier: NCT00465829)
Footnotes
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed consent was obtained from all individual participants included in the study.
COI statement
All authors have no conflicts of interest.
The authors declare that they have no conflict of interest.
References
- 1.Courcoulas AP, King WC, Belle SH, Berk P, Flum DR, Garcia L, Gourash W, Horlick M, Mitchell JE, Pomp A, Pories WJ, Purnell JQ, Singh A, Spaniolas K, Thirlby R, Wolfe BM, Yanovski SZ. Seven-Year Weight Trajectories and Health Outcomes in the Longitudinal Assessment of Bariatric Surgery (LABS) Study. JAMA Surg. 2018;153(5):427–34. doi: 10.1001/jamasurg.2017.5025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Magro DO, Geloneze B, Delfini R, Pareja BC, Callejas F, Pareja JC. Long-term weight regain after gastric bypass: a 5-year prospective study. Obes Surg. 2008;18(6):648–51. doi: 10.1007/s11695-007-9265-1. [DOI] [PubMed] [Google Scholar]
- 3.Odom J, Zalesin KC, Washington TL, Miller WW, Hakmeh B, Zaremba DL, Altattan M, Balasubramaniam M, Gibbs DS, Krause KR, Chengelis DL, Franklin BA, McCullough PA. Behavioral predictors of weight regain after bariatric surgery. Obes Surg. 2010;20(3):349–56. doi: 10.1007/s11695-009-9895-6. [DOI] [PubMed] [Google Scholar]
- 4.Meany G, Conceicao E, Mitchell JE. Binge eating, binge eating disorder and loss of control eating: effects on weight outcomes after bariatric surgery. Eur Eat Disord Rev. 2014;22(2):87–91. doi: 10.1002/erv.2273. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Association AP. Diagnostic and statistical manual of mental disorders 5ed. Washington, DC: Author; 2013. [Google Scholar]
- 6.Duncan AE, Ziobrowski HN, Nicol G. The Prevalence of Past 12-Month and Lifetime DSM-IV Eating Disorders by BMI Category in US Men and Women. Eur Eat Disord Rev. 2017;25(3):165–71. doi: 10.1002/erv.2503. [DOI] [PubMed] [Google Scholar]
- 7.Mitchell JE, King WC, Courcoulas A, Dakin G, Elder K, Engel S, Flum D, Kalarchian M, Khandelwal S, Pender J, Pories W, Wolfe B. Eating behavior and eating disorders in adults before bariatric surgery. Int J Eat Disord. 2015;48(2):215–22. doi: 10.1002/eat.22275. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Mechanick JI, Youdim A, Jones DB, Garvey WT, Hurley DL, McMahon MM, Heinberg LJ, Kushner R, Adams TD, Shikora S, Dixon JB, Brethauer S, American Association of Clinical E, Obesity S, American Society for M, Bariatric S. Clinical practice guidelines for the perioperative nutritional, metabolic, and nonsurgical support of the bariatric surgery patient−−2013 update: cosponsored by American Association of Clinical Endocrinologists, The Obesity Society, and American Society for Metabolic & Bariatric Surgery. Obesity (Silver Spring). 2013;21 Suppl 1:S1–27. doi: 10.1002/oby.20461. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Colles SL, Dixon JB, O’Brien PE. Grazing and loss of control related to eating: two high-risk factors following bariatric surgery. Obesity (Silver Spring). 2008;16(3):615–22. doi: 10.1038/oby.2007.101. [DOI] [PubMed] [Google Scholar]
- 10.Conceicao EM, Mitchell JE, Pinto-Bastos A, Arrojado F, Brandao I, Machado PPP. Stability of problematic eating behaviors and weight loss trajectories after bariatric surgery: a longitudinal observational study. Surg Obes Relat Dis. 2017;13(6):1063–70. doi: 10.1016/j.soard.2016.12.006. [DOI] [PubMed] [Google Scholar]
- 11.Devlin MJ, King WC, Kalarchian MA, White GE, Marcus MD, Garcia L, Yanovski SZ, Mitchell JE. Eating pathology and experience and weight loss in a prospective study of bariatric surgery patients: 3-year follow-up. Int J Eat Disord. 2016;49(12):1058–67. doi: 10.1002/eat.22578. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Mitchell JE, Christian NJ, Flum DR, Pomp A, Pories WJ, Wolfe BM, Courcoulas AP, Belle SH. Postoperative Behavioral Variables and Weight Change 3 Years After Bariatric Surgery. JAMA Surg. 2016;151(8):752–7. doi: 10.1001/jamasurg.2016.0395. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Chao AM, Wadden TA, Faulconbridge LF, Sarwer DB, Webb VL, Shaw JA, Thomas JG, Hopkins CM, Bakizada ZM, Alamuddin N, Williams NN. Binge-eating disorder and the outcome of bariatric surgery in a prospective, observational study: Two-year results. Obesity (Silver Spring). 2016;24(11):2327–33. doi: 10.1002/oby.21648. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Wadden TA, Faulconbridge LF, Jones-Corneille LR, Sarwer DB, Fabricatore AN, Thomas JG, Wilson GT, Alexander MG, Pulcini ME, Webb VL, Williams NN. Binge eating disorder and the outcome of bariatric surgery at one year: a prospective, observational study. Obesity (Silver Spring). 2011;19(6):1220–8. doi: 10.1038/oby.2010.336. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Burke LE, Wang J, Sevick MA. Self-monitoring in weight loss: a systematic review of the literature. J Am Diet Assoc. 2011;111(1):92–102. doi: 10.1016/j.jada.2010.10.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Loth KA, Goldschmidt AB, Wonderlich SA, Lavender JM, Neumark-Sztainer D, Vohs KD. Could the resource depletion model of self-control help the field to better understand momentary processes that lead to binge eating? Int J Eat Disord. 2016;49(11):998–1001. doi: 10.1002/eat.22641. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Heatherton TF, Wagner DD. Cognitive neuroscience of self-regulation failure. Trends Cogn Sci. 2011;15(3):132–9. doi: 10.1016/j.tics.2010.12.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Haedt-Matt AA, Keel PK. Revisiting the affect regulation model of binge eating: a meta-analysis of studies using ecological momentary assessment. Psychol Bull. 2011;137(4):660–81. doi: 10.1037/a0023660. 21574678; PMCID: PMC3100657. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Saunders JB, Aasland OG, Babor TF, de la Fuente JR, Grant M. Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO Collaborative Project on Early Detection of Persons with Harmful Alcohol Consumption--II. Addiction. 1993;88(6):791–804. [DOI] [PubMed] [Google Scholar]
- 20.Marek RJ, Heinberg LJ, Lavery M, Merrell Rish J, Ashton K. A review of psychological assessment instruments for use in bariatric surgery evaluations. Psychol Assess. 2016;28(9):1142–57. doi: 10.1037/pas0000286. [DOI] [PubMed] [Google Scholar]
- 21.Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry. 1961;4:561–71. [DOI] [PubMed] [Google Scholar]
- 22.Krukowski RA, Friedman KE, Applegate KL. The utility of the Beck Depression Inventory in a bariatric surgery population. Obes Surg. 2010;20(4):426–31. doi: 10.1007/s11695-008-9717-2. [DOI] [PubMed] [Google Scholar]
- 23.Kolotkin RL, Crosby RD, Kosloski KD, Williams GR. Development of a brief measure to assess quality of life in obesity. Obes Res. 2001;9(2):102–11. doi: 10.1038/oby.2001.13. [DOI] [PubMed] [Google Scholar]
- 24.Dymek MP, Le Grange D, Neven K, Alverdy J. Quality of life after gastric bypass surgery: a cross-sectional study. Obes Res. 2002;10(11):1135–42. doi: 10.1038/oby.2002.154. [DOI] [PubMed] [Google Scholar]
- 25.Kolotkin RL, Crosby RD, Gress RE, Hunt SC, Adams TD. Two-year changes in health-related quality of life in gastric bypass patients compared with severely obese controls. Surg Obes Relat Dis. 2009;5(2):250–6. doi: 10.1016/j.soard.2009.01.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Kolotkin RL, Crosby RD, Pendleton R, Strong M, Gress RE, Adams T. Health-related quality of life in patients seeking gastric bypass surgery vs non-treatment-seeking controls. Obes Surg. 2003;13(3):371–7. doi: 10.1381/096089203765887688. [DOI] [PubMed] [Google Scholar]
- 27.Smyth JM, Wonderlich SA, Heron KE, Sliwinski MJ, Crosby RD, Mitchell JE, Engel SG. Daily and momentary mood and stress are associated with binge eating and vomiting in bulimia nervosa patients in the natural environment. J Consult Clin Psychol. 2007;75(4):629–38. doi: 10.1037/0022-006X.75.4.629. [DOI] [PubMed] [Google Scholar]
- 28.Stice E A prospective test of the dual-pathway model of bulimic pathology: mediating effects of dieting and negative affect. J Abnorm Psychol. 2001;110(1):124–35. [DOI] [PubMed] [Google Scholar]
- 29.Munoz DJ, Chen E, Fischer S, Roehrig M, Sanchez-Johnson L, Alverdy J, Dymek-Valentine M, le Grange D. Considerations for the use of the Beck Depression Inventory in the assessment of weight-loss surgery seeking patients. Obes Surg. 2007;17(8):1097–101. [DOI] [PubMed] [Google Scholar]
- 30.Goldschmidt AB, Conceicao EM, Thomas JG, Mitchell JE, Raynor HA, Bond DS. Conceptualizing and studying binge and loss of control eating in bariatric surgery patients-time for a paradigm shift? Surg Obes Relat Dis. 2016;12(8):1622–5. doi: 10.1016/j.soard.2016.09.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Bodell LP, Forney KJ, Chavarria J, Keel PK, Wildes JE. Self-report measures of loss of control over eating: Psychometric properties in clinical and non-clinical samples. Int J Eat Disord. 2018;51(11):1252–60. doi: 10.1002/eat.22957. [DOI] [PMC free article] [PubMed] [Google Scholar]