Modifiable factors associated with weight regain after bariatric surgery: a scoping review

Search strategy

A systematic search of the literature was conducted in Medline, Google Scholar, Cochrane, National Collaborating Centre for Methods and Tools (NCCMT), and Practice-based Evidence in Nutrition (PEN). The most recent search was conducted on February 2, 2017. We included studies published in English between January 1990 and January 2017, using Boolean search terms—such as, “weight loss maintenance” OR “weight regain” AND “behavior” AND “English” AND “adult” (see Extended data) - that were identified by the research team¹¹. An example of the search strategy used in Medline includes the following search terms: ("weight loss maintenance"[All Fields] OR "weight regain"[All Fields]) AND "behavior"[All Fields] AND (("1990/01/01"[PDAT] : "2017/02/02"[PDAT]) AND "humans"[MeSH Terms] AND English[lang] AND medline[sb] AND "adult"[MeSH Terms]). Manual searches of cited references were also conducted to identify additional articles, as described in the eligibility criteria explained below. In total, 276 studies were identified using this search strategy once duplicates were removed.

Inclusion and exclusion criteria

Our inclusion criteria were any published studies, reviews, practice guidelines and expert opinions involving adults (18+ years) in English between January 1, 1990 and January 31, 2017. All studies and reviews published after this date were excluded. To further narrow the number of papers for full-text review, only titles and/or abstracts containing the terms “bariatric surgery” or “weight loss surgery”, or an association of these terms, were included (n=32). All bariatric surgery types were considered, even those no longer being commonly performed (e.g. vertical banded gastroplasty, VBG). Given that the research community has not yet reached consensus on how weight regain is defined or compared, all studies addressing weight regain either as a percentage of regain or in weight gained were included regardless of the extent of regain. We retained studies that examined modifiable factors (such as diet, behavior/psychology, support, and physical activity) either exclusively or in conjunction with other influences. Studies that focused solely on non-modifiable factors including, but not limited to, gastric pouch size, age, sex, and pre-operative body mass index (BMI), were excluded. The included studies must have identified or referenced weight regain, or explored a period of possible weight regain or weight maintenance; however, no limit in time-frame was applied. Studies that only addressed insufficient weight loss without addressing weight regain were excluded. A total of 32 full-text reviews were performed and 22 articles were included for thematic analysis, after being assessed for relevancy and removing articles mentioned in the systematic reviews that were included (see Figure 1).

Figure 1. PRISMA flowchart.

This PRISMA flowchart depicts the number of records identified through database searching and other sources, and the final number of articles – 22 articles – included in this scoping review.

Data was charted into an excel spreadsheet. This data was gathered, charted, and inputted independently, and reviewed by two other authors (AH, MJ). All authors discussed the themes that emerged in order to define the variables. Emerging themes that were factors in weight regain were documented and grouped into variables that summarized the themes, namely into behavioral, environmental, support, and exercise. All articles were assigned and coded to one, or more, of the stated variables.

Poor dietary adherence

Of the 22 studies included in our thematic analysis, eight (36%)^8,12–18 suggested that diet was significantly associated with weight regain (p = 0.001-0.05) after bariatric surgery. Poor adherence to dietary guidelines - represented by higher carbohydrate intake¹⁶, higher alcohol intake^8,15,16 and lower dietary quality⁸ - were key contributors.

Two studies reported that a higher energy intake (2000 vs. 1500 Kcal/day) was associated with a weight regain of 11 kg from nadir weight as of two years after surgery^8,12 Similarly, Himes et al.¹⁷ reported a 10% weight loss of the total weight regained in as little as six weeks by reducing the frequency of eating episodes from 6.7 to 5.5 episodes, and 86% of participants in Faria et al.¹⁴ lost 54% of their weight regain in three months following a 1400 Kcal/day prescription. This suggests that a higher frequency of eating episodes and higher energy intake over time may have contributed to weight regain prior to the interventions.

In terms of dietary quality and alcohol intake, Reid et al.¹⁶ observed a 26% difference in weight outcomes (p ≤ 0.01), at 12 years post-operation, between people who maintained their weight and people who regained some weight. In this study, people who regained some weight consumed more carbohydrates than people who maintained their weight (222 vs. 162 g/day, p < 0.05); however, there was no difference in the percentage of energy intake from carbohydrates in both groups (43% vs. 42%, respectively)¹⁶ While people who regained some weight consumed more alcohol than people who maintained their weight (1.32 vs. 0.19 standard drinks/day; p < 0.05) in Reid et al.’s study¹⁶ reported consumption was still within the suggested limits for the general population. Other studies, however, have determined that alcohol misuse or abuse was associated with weight regain^8,15. A higher median fat intake was observed in people who regained some weight (88.6 vs. 64.3 g/d, p < 0.05), although the percentage of energy intake from fats was similar in both groups (41.7% vs. 37.4%, respectively)¹². Similarly, Karmali et al. refer to studies that reported a higher energy intake¹⁹ and poor dietary quality³, including higher sugar, sweets and fatty foods, were attributable to weight regain as of two years after RYGB, VBG, and adjustable gastric band (AGB).

Finally, Bastos et al. examined the influence of having a food-related occupation on weight regain. They determined that working in food production - whether as a baker or working in a cafeteria, snack bar, restaurant, or grocery store - was correlated with significant weight regain (p = 0.003)¹³.

Of the eight studies that observed diet as a factor associated with weight regain after bariatric surgery, seven of them observed calorie intake as a contributing factor. A higher calorie intake - whether from carbohydrates, alcohol, low nutritive value sweets, fatty foods, or as a result of a higher frequency of eating episodes - are associated with weight regain^8,12–18. Only one study observed the association of working in the food industry with weight regain. Contrary to other studies, however, they did not observe alcohol intake as a contributing factor.

Behavioral/psychological issues

Thirteen studies (59%) identified post-operative diet-related behaviors, or eating habits, and psychological factors were associated with long-term post-operative weight regain.

Maladaptive eating behaviors. Of the 12 studies that examined diet-related behaviors, nine studies (75%) found a significant association with post-operative weight regain^8,18,20–27. Variations of these habits, including binge eating, disinhibited eating, picking and nibbling or grazing, and loss of control eating behaviors, have contributed to a weight regain ranging from 10 kg to 17 kg^8,17, nearly 11% gain from nadir weight at two years after surgery²². At five years post-operative, weight regain was found to be as high as 44% from nadir weight, as a result of disinhibited eating, or the tendency to overeat, and unsuitable eating behaviors²¹. While most studies have demonstrated the association of maladaptive dietary behaviors with weight regain^8,18,20–27, Himes et al.¹⁷ demonstrated that these behaviors improved with interventions; for example, a group behavioral cognitive behavioral therapy (CBT) intervention led to a 1.6 kg weight loss over six weeks in those who had been on a weight regain trend post bariatric surgery. Meanwhile, Mitchell et al.²⁰ determined that grazing and binging habits, in addition to a lack of self-monitoring, accounts for 16% variability in weight outcomes. Pekkarinen et al.²⁴ showed that although people who binged and people who did not binge had similar outcomes at one year post-VBG (55% vs. 57%, respectively), people who binged regained significantly more weight than people who did not binge two years after their operation (24% vs. 50%, respectively, p = 0.04). Finally, Hsu et al.²⁷ showed that those who suffered from binge-eating disorder (BED) prior to surgery continued to struggle with this after bariatric surgery. This finding was further confirmed by Colles et al.,²⁸ who found that grazing habits increased post bariatric surgery (26% pre-operation to 38% by one year post-operative), while Nicolau et al. determined that 72% of people who grazed gained weight vs. people who did not graze (p < 0.0001)²³. However, Alvarez et al. did not find BED to be associated with post-operative weight regain.

Of all 22 studies reviewed, only one had identified nocturnal eating as an important determinant of weight regain (p = 0.01)¹⁵.

Psychological factors. Three articles determined that depression^8,15, anxiety¹², and alcohol and/or substance abuse^8,12,15 were associated with post-operative weight regain.

Lack of support

Five articles (23%) suggested poor post-operative support was associated with long-term weight regain^8,29–32. Four of these studies examined the impact of follow-ups with a bariatric team; they determined that little to no post-operative follow-up can lead to poor long-term outcomes^8,29–31. Those who maintained regular follow-up, for up to three years post-operation, had better long-term weight outcomes (74% excess weight loss, EWL)^30,31 than those who were lost to follow-up within the first year of their surgery (56% EWL, p < 0.05)³⁰. Among the 75% of patients who no longer received follow-up as of three years post-operation, they were observed to regain up to 14% more weight; in comparison, only 25% of patients who received an annual follow-up up to five to six years post-operation were found to experience weight regain²⁹. In another study, regular follow-up represented 47% difference in %EWL at two years post-operation³⁰.

Post-operative support from health care professionals was found to be an important component in long-term success. Karmali et al.⁸ cited that, among those who failed surgery, 60% had never seen a dietitian and 80% had never seen a psychologist. Comparatively, Gould et al.³¹ reported greater long-term outcomes in patients who attended all post-operative follow-up visits with a multidisciplinary team. A post-operative bariatric surgery patient who received individual or group CBT sessions had a higher %EWL (90% vs. 43% EWL) at two years post-surgery and better weight loss outcomes than the controls who did not receive support³⁰. Yet, despite the known benefits of post-operative support, there was little evidence in the literature in support group attendance and its influence on weight regain or weight maintenance. Liebl’s qualitative study³² described the experiences of post-operative bariatric patients who were successful at maintaining weight loss (average of 8% of EWL regained) at an average of 69 months post-surgery. The patients surveyed in the study reported that support from their family, peers and professionals from their bariatric surgery clinic had been necessary to achieve positive outcomes.

Physical inactivity

Seven studies (32%) addressed the relationship between physical activity and weight regain. Questionnaires, which primarily captured moderate to vigorous intensity activity, were used to determine activity levels in three of the studies^15,18,21. Four systematic reviews reinforced that lower physical activity levels were associated with poorer weight loss outcomes and higher weight regain, despite undergoing bariatric surgery^{8,18,29,33,34}. Furthermore, Amundsen et al.²¹ observed that people who significantly regained weight (total weight regain ≥ 15% from nadir weight), were less active than people who regained a normal amount of weight (p=0.06). People who maintained their weight spent remarkably more time being moderately active (567 min per week) compared to people who regained weight (287 min per week).

Poor dietary adherence

Studies related to dietary adherence suggested that poor observance of the dietary guidelines -represented by higher carbohydrate intake, higher alcohol intake and lower dietary quality - were key contributors to weight regain in the long-term recovery from bariatric surgery.

Changes in dietary adherence over the course of the post-operative phase may be associated with weight regain in patients after bariatric surgery^8,12,16,18. Higher carbohydrate consumption appears to be the most evident dietary cause associated with weight regain¹⁶. Although the source of carbohydrates was not clearly defined in all studies, some have demonstrated that an increased consumption of liquid calories and sugar intake from non-nutritive sources were attributable to weight regain²⁶. Thus, increases in patient’s non-nutritive, free- and added-sugar intake potentially explain some of the weight regain following bariatric surgery.

Alcohol is a source of empty liquid calories which contributes significantly to one’s caloric intake, and some studies have found a positive association between weight regain and alcohol abuse or misuse^8,15,16. This is particularly concerning because alcohol abuse has been shown to be higher among people who have had bariatric surgery, in comparison to the general population³⁵. While there does not appear to be a consensus on post-operative weight regain and alcohol intake, one of the studies reported an association between intake levels at or beneath the suggested alcohol intake limits^16,36. The discrepancies seen in the literature may be due to the nature of the data collection. Alcohol intake is known to be underreported by up to 50% when self-reported³⁷. This is especially true among middle-aged women, which coincides with the demographic of the bariatric population³⁷. Patients need to be educated on the effects of alcohol consumption after malabsorptive bariatric procedures due to changes in alcohol metabolism, particularly after a gastric bypass. Studies have demonstrated an acceleration in alcohol absorption after a gastric bypass such that it takes a shorter time to reach a maximum concentration. In addition, there is a higher maximum alcohol concentration achieved, as well as it taking longer to fully metabolize and eliminate alcohol³⁸. As such, despite the inconclusive results, results from this review suggest alcohol intake should be more closely assessed and monitored in the people who have had bariatric surgery, even if their consumption is within the suggested limits according to the Centers for Disease Control and Prevention³⁶.

Given the gastric restriction of bariatric surgery, increased calorie intake occurs in the form of more frequent eating episodes and/or in the consumption of more calorie-dense foods or liquid calories. This demonstrates that dietary patterns after bariatric surgery do not remain consistent; rather, there is a gradual onset of undesirable dietary habits that develops, and some patients may not be cognizant of the effects this will have on their future outcome. Therefore, bariatric surgery alone is not protective in the long term; patients will likely require the ongoing support and monitoring of dietitians and their bariatric team to find solutions and alternatives to the challenges to maintain adherence to dietary recommendations.

Behavioral/psychological issues

Behavioral and psychological factors may impede one’s ability to comply with post-operative lifestyle recommendations. While restrictive and malabsorptive procedures may limit the amount of food consumed in a given sitting, it does not generally limit the ability to eat significant volumes over the course of a day. Grazing and binging were the most commonly identified eating behaviors associated with weight regain. All, except for one study, clearly observed this relationship. Although maladaptive eating habits do not negatively affect one’s weight outcomes at one year post-operation, people who continue to binge have a higher risk of regaining weight by the second year following surgery²⁴. Furthermore, Himes et al.¹⁷ suggests that therapy aimed at reducing binge eating behaviors can lower the number of daily eating episodes and encourage weight loss following weight regain. Therefore, targeted therapy towards maladaptive eating behaviors provided early on in the patient’s recovery process may help to prevent occurrence of weight regain.

Given that maladaptive eating behaviors that exist prior to the surgery appear to remain an issue in the post-operative phase²⁷, dietary behaviors are an important target point for discussion even before the weight loss surgery is performed. Health professionals should educate patients on the evolution of maladaptive eating behaviors after surgery to encourage therapy and treatment prior to the operation and on an ongoing basis. Professionals should continue to monitor and probe for various types of maladaptive eating behaviors long-term after surgery.

Most of the articles that studied behaviors looked at maladaptive eating behaviors also found lower levels of physical activity as contributing factor to weight regain. This suggests that there may be a relationship―and, potentially, interactive or cumulative impacts―between eating behaviors, dietary intake and physical activity.

Lack of support

The American Society for Metabolic and Bariatric Surgery (ASMBS) has produced guidelines on post-operative follow-up care with a bariatric team, which emphasizes the importance of follow-up care on a patient’s progress and outcomes. Follow-up care helps to track the patient’s progress, ensure their adherence to recommendations made by their health care team and offers solutions for managing weight and lifestyle. With 50% of people who have had bariatric surgery regaining weight within two years of their surgery, it is not surprising that 60% of these patients had not received regular nutrition follow-up and 80% did not receive psychological follow-up³⁹; in contrast, patients who attended all of their post-operative follow-up were more successful in the long-term³¹. Follow-up maintained for up to three years post-operation resulted in 18% higher EWL compared to a patient who was lost to follow-up within the first year of surgery³¹, which may infer a positive correlation between longer follow-up periods and improved health outcomes. The impact of support has been well studied post bariatric surgery, as evidenced by several systematic reviews on the subject. Unfortunately, studies included in this scoping review were only able to highlight the benefit of post-operative support with a bariatric team; it was not clear from the literature if there is a single member of the bariatric team that is pivotal in providing the follow-up, even though several studies observed follow-up with either a dietitian or a psychologist.

From the patient’s perspective, positive family, peer, and professional bariatric support were identified as being vital in achieving long-term post-operative success. This offers important insight into the patients’ perspectives and the three prongs of support considered necessary to optimize their long-term outcomes. Based on these findings, a patient’s health care provider and social network - which may include family members and friends - should encourage the patient to present to follow-up appointments and, ideally, accompanied by their family, partners or friends. Given that patients usually forget the information they have been taught prior to surgery, which further worsens at one year post-operation⁴⁰, inviting the patients’ support systems to follow-up appointments may encourage them to maintain regular contact with their bariatric team and also serve as an aid to remind patients of the positive interventions and behaviors discussed during consultations.

Physical inactivity

Bariatric surgery could greatly improve mobility and reduces the burden of osteoarthritis. As previously supported in the literature, participants with obesity tend to over-report their physical activity levels⁴¹. While 89% of patients self-reported to be regularly active at two years post-operation³³, other studies observed that only about half of patients were engaged in moderate to vigorous activity for more than one session per week⁴², which is lower than the activity of adults in the general population (48% vs 53%, respectively)⁴³. Furthermore, only 11%¹⁶ of people who have had bariatric surgery, as compared to 35%⁴³ of the general population, actually achieve 10,000 steps per day when measured using accelerometers.

Although there were several systematic reviews on the impact of physical activity on post-operative bariatric surgery outcomes, there is little information available on the subject of physical activity and weight regain; most articles cite that ‘low energy expenditure’ was associated with weight regain. It has also been suggested that current physical activity guidelines may be too low to prevent long-term weight regain after bariatric surgery²¹. However, it is observed that despite having similar moderate to vigorous physical activity habits to the general population, people who have had bariatric surgery are less active on a daily basis.

Limitations in the literature

There were several identified gaps in the literature, which could limit the generalizability of the conclusions presented here. Firstly, there is not a consistent definition and measure for “weight regain”, or a clear indication of what is considered to be normal weight regain. This makes it difficult to compare the effect size across studies and modifiable factors in relation to weight regain. The ASMBS has suggested standardized outcomes reporting; however, the association has not offered guidelines on reporting magnitude of weight regain. In addition, although participants were stratified into groups, the cut-off points used in some studies to describe percentage of weight regain may not be sensitive enough (i.e. >50% EWL vs ≤50% EWL) to observe clinically significant changes. As a result, there may have been poor differentiation between outliers and participants who narrowly fell within the cut-off range¹².

Secondly, the surveys and questionnaires used were unique to each study, resulting in inconsistent metrics that are often not directly comparable. This is important since different surveys for assessing lifestyle habits may lead to varying results and conclusions. The use of different tools and questionnaires that have not been validated for the bariatric surgery population may also explain the discrepancies observed for the expected similar associations.

Thirdly, the majority of the studies relied on self-reported measures of modifiable behaviors. Dietary misreporting in the population affected by obesity is particularly concerning and has been well-documented in a previous study, such that people who under-reported their energy intake were more likely to be inflicted with obesity⁴⁴. Similar results have been observed for self-reported physical activity, which is generally over-reported. All but one study included in this scoping review relied on questionnaires and self-reported data; hence, making it difficult to reach a robust conclusion on physical activity levels post-surgery. Therefore, self-reported data inevitably limits the external validity of the findings.

Another source of variability to consider is that this scoping review included all types of bariatric surgery, even though the VBG is no longer performed and the AGB is being phased out as a routine procedure. However, even though each procedure is different in the weight loss achieved and outcome experiences (with regards to tolerance of certain foods), the modifiable behaviors would be similar across surgery types.

Lastly, most of the studies included in this scoping review were observational. However, even among the interventional studies, the sample sizes were very small (n<30), making it difficult to determine the true cause and effect.

Underlying data

All data underlying the results are available as part of the article and no additional source data are required.

Extended data

Open Science Framework: Modifiable Factors Associated with Weight Regain After Bariatric Surgery: A Scoping Review. https://doi.org/10.17605/OSF.IO/9Q78A¹¹.

This project contains the following extended data:

Data are available under the terms of the Creative Commons Zero "No rights reserved" data waiver (CC0 1.0 Public domain dedication).