Impact of nutritional‐behavioral and supervised exercise intervention following bariatric surgery: The BARI‐LIFESTYLE randomized controlled trial

The study's aim was to investigate the impact of a 12‐month adjunctive lifestyle intervention on weight loss and health outcomes after bariatric surgery.


INTRODUCTION
Bariatric surgery is the most effective treatment for people living with severe obesity, resulting in marked, sustained weight loss with improvement/resolution of obesity-linked comorbidities, improved health-related quality of life (HRQoL), and increased life expectancy [1]. At present, the prospective outcome data of bariatric surgery performed in the UK are lacking [2]. The National Bariatric Surgery Registry only collects data on outcomes such as weight loss, resolution of comorbidities, and HRQoL [3], whereas other important parameters such as the changes in body composition, including bone mineral density (BMD), physical activity levels, physical function, and strength are scarce [2]. These crucial data are needed to improve the post-bariatric care provided to patients.
Evidence has shown that weight loss and health benefits of bariatric surgery vary markedly between individuals, with approximately 20% to 30% of patients experiencing suboptimal weight loss [4]. It is also known that weight loss in the early postsurgery period is accompanied by a substantial loss of fat-free mass with a concomitant decline in BMD [5,6]. To optimize weight loss and counteract the unfavorable outcomes, patients are required to follow post-bariatric lifestyle recommendations such as adequate protein intake, lifelong vitamin and mineral supplementation, and engagement in physical activity [7]. However, support from clinical teams in implementing these recommendations is often minimal, which may impact a person's ability to adapt after surgery [8]. In current clinical practice, most of the multidisciplinary input focuses on presurgery preparation [7]. This highlights the need to investigate the integration of effective followup care packages into post-bariatric care services to support patients after surgery [7].
To date, few studies have evaluated the efficacy of an adjunctive lifestyle program starting in the early postoperative period to support patients and maximize health outcomes [9][10][11]. Poor-quality studies, heterogeneity of the lifestyle program, and obsolete bariatric procedures reported in previous randomized controlled trials (RCTs) have meant that a definitive conclusion currently cannot be drawn as to whether a postsurgery lifestyle intervention can elicit further weight loss and health improvement [9][10][11]. Therefore, in the BARI-LIFESTYLE RCT study, we assessed the impact of a combined nutritional-behavioral and supervised exercise intervention, implemented immediately after surgery, on percentage weight loss (%WL) and health outcomes in the first postoperative year, compared with standard care.

Design, setting, and participants
BARI-LIFESTYLE was a two-arm, parallel-group, single-blind, multisite RCT study, embedded within an observational cohort study undertaken at three National Health Service bariatric centers in London, UK. A two-staged randomized consent design was employed [12].
The trial protocol and amendments (available in online Supporting Information) were approved by London-Dulwich Research Ethics Committee (17/LO/0950). This trial was carried out by the Centre for Obesity Research, University College London.
Eligible participants were adults aged 18 to 65 years who were planning to undergo either primary gastric bypass or sleeve gastrectomy (SG) surgery and who fulfilled the National Institute for Health and Care Excellence (NICE) eligibility criteria [7]; were medically safe to participate in an exercise program; were able to read and write in English; were able to comply with the trial protocol; were able to attend a supervised tailored exercise session weekly for 12 weeks; and were willing to wear a Fitbit and ActiGraph (wGT3X-BT, software V.6.13.3).
Patients were deemed ineligible if they had a body weight of 200 kg or more because of the limitation of the dual-energy x-ray absorptiometry (DXA) scan; if they were nonambulatory; if they had a functional limitation; or if they had a medical contraindication for exercise. All trial participants provided the first written informed consent to take part in the BARI-LIFESTYLE observational study that

Study Importance
What is already known?
• About 20% to 30% of patients experience suboptimal weight loss, defined as percentage weight loss less than 20% at 12 months post surgery.
• Prior to bariatric surgery, patients spend more time in sedentary behavior with low levels of physical activity.
Postoperatively, despite weight loss, physical activity levels and sedentary behavior remain similar to preoperative levels.
• Fat-free mass is significantly reduced after surgery, which might contribute to weight regain in the long-term.
What does this study add?
• Provision of an early adjunctive program of nutritionalbehavioral and supervised exercise intervention, implemented immediately after surgery, had no additional favorable impact on weight loss and non-weight-related health outcomes.
How might these results change the direction of research?
• Future studies should investigate whether shifting the start of an adjunctive lifestyle intervention program to a later time (i.e., during weight-loss maintenance phase) would improve weight loss and health benefits.
• Effective strategies are needed to improve physical activity levels, reduce sedentary behavior, and protect fat-free mass following bariatric surgery. aimed to collect outcome data of bariatric surgery. During the consenting process, participants were told there might be the possibility they would be approached again for potential participation in a postsurgery lifestyle program. The study protocol of the observational cohort, including the study rationale, details of recruitment, procedures, outcome measures, and planned data analysis, has been published previously [2]. The details of each surgical technique (Rouxen-Y gastric bypass, one anastomosis gastric bypass, and SG) are outlined in online Supporting Information Methods.

Randomization
Immediately after surgery, all 153 participants in the observational cohort were randomized in a 1:1 ratio to receive either postsurgery standard care or standard care plus a lifestyle program. Randomization was carried out remotely by the investigators using an online randomization facility that generated a sequence with variable blocks of two, four, and six, stratified by the type of surgical procedure and trial site.
All participants randomized to the intervention group were then invited to take part in the BARI-LIFESTYLE intervention study. Those who agreed to take part provided a second written informed consent prior to hospital discharge.
Participants allocated to the control group were not informed of the randomization and intervention study. This was to avoid contamination of the control group, which may have potentially diluted the treatment effects in the intention-to-treat (ITT) analysis. The outcome assessors and the trial statistician were blinded to the group allocation.     [14,15]. Attitude and symptoms of depression were assessed using Beck

Interventions
Depression Inventory-II [16]. Comorbidities resolution (hypertension, hyperlipidemia, and obstructive sleep apnea) was defined as the normalization of the corresponding characteristics without treatment. Complete remission of type 2 diabetes was defined according to the American Diabetes Association criteria of complete remission: no antidiabetic drug, fasting blood glucose <5.6 mmol/L, and glycated hemoglobin in the normal range of at least 1 year's duration. Partial remission was defined as the following: no antidiabetic drug and glycated hemoglobin < 6.5% of at least 1 year's duration [17].

Statistical analysis
The initial sample size was calculated based on the cohort data of patients who underwent bariatric surgery at one of the trial sites with the mean of %WL at 12 months post surgery of 27.8% (SD 8.4%) [4]. We calculated that a sample size of 198 participants (99 participants per study arm) would provide 95% power at a 0.05 significance level to detect a 5% difference in %WL between groups at 12 months, after allowing for up to 25% dropout. However, because of the high attrition rate at 12 months and the COVID-19 pandemic, the primary outcome time point was changed to 6 months post surgery following consultation with the trial steering committee. At this time point, 65 and 69 participants were available in the control and intervention groups. Continuing with the assumed mean difference of 5% and SD of 8.4%, the power to test a difference of 5% at 6 months was 93%.
The longitudinal analysis for the entire cohort has been described previously [2]. The analyses for the primary and secondary outcomes of the RCT were undertaken based on appropriate complete-case groupings for related measures on the ITT principle that no account was paid to the extent to which the treatments were followed. For the primary outcome analysis, the ITT principle was further strengthened by including participants whose weight at 6 months was based on self-measurement. Standard linear regres-   Figure S1 and Table S1). The mean percentage fat mass and fat-free mass loss was also comparable across all bariatric procedures, p = 0.91 and p = 0.52, respectively (Supporting Information Table S2). A detailed exploratory analysis of the primary and secondary outcomes of the observational cohort will be presented separately.

The impact of intervention on weight loss
A total of 49 participants in the intervention group and 55 participants in the control group with available BIA body weight were included for complete-case analysis (Table 2) Table S3).

The impact of intervention on secondary outcomes
For the prespecified secondary outcomes analysis, no between-group differences were observed throughout the first postoperative year in fat mass, fat-free mass, BMD, habitual physical activity, physical function and strength, HRQoL, depressive symptomatology (Table 2), and comorbidities resolution (Supporting Information Table S4).

Adverse events
Twenty-two serious adverse events were reported in 13 participants from the intervention group and 5 participants from the control group who required hospitalization. Of all the serious adverse events, some were expected postsurgery events such as pregnancy, vomiting, cholecystectomy, and surgical-related complications ( Table 3). None of the serious adverse events was related to either the tele-counseling or supervised exercise interventions. All serious adverse events were resolved with appropriate management. Most of the reported adverse events were mild (Supporting Information Table S5). The common adverse events reported in the intervention group were similar to those associated with postbariatric surgery (e.g., constipation, vomiting, abdominal pain). The gastrointestinal and cardiovascular adverse events were more commonly reported by participants in the intervention group than in the control group. These differences were most probably associated with increased reporting by the intervention group compared with the control group, as there were more follow-up visits completed by the former group. T A B L E 2 Main outcomes of the trial at baseline and at 3, 6, and 12 months

DISCUSSION
Aligned with the existing literature [1,3], longitudinal analysis of the BARI-LIFESTYLE cohort showed that bariatric surgery leads to a marked reduction in body weight and fat mass with improvement in physical function, HRQoL, and depressive symptomatology. However, physical activity levels and sedentary behavior did not change after surgery, with an undesirable impact observed on the changes in fatfree mass and BMD, similar to what has been previously reported [5,6,19,20]. Of those participants enrolled in the BARI-LIFESTYLE intervention, their weight loss did not differ compared with those participants receiving post-bariatric standard care. Similarly, the BARI-LIFESTYLE intervention had no favorable impact on nonweight-related outcomes that include fat mass, fat-free mass, BMD, habitual physical activity, physical function and strength, HRQoL, depressive symptomatology, and comorbidities resolution.
The present trial is the first evidence, to our knowledge, to show that combining both nutritional-behavioral intervention and tailored supervised exercise in a single postoperative lifestyle program has no impact on weight loss in the first 12 months after surgery. Our finding further supports a few previous RCTs demonstrating that providing either postsurgery nutritional-behavioral intervention [21][22][23][24][25] or exercise programs [26,27] within the first year after surgery did not lead to additional weight loss.
The exact reasons are not fully understood, although these might be due to the predominant biological effects of weight loss in the early phase of bariatric surgery [28] that may have overridden the potential weight-loss impact of any adjunctive lifestyle interventions. This is consistent with the results of the STEP trials with the glucagon-like peptide-1 receptor agonist semaglutide. The STEP 3 trial that combined semaglutide with intensive behavioral therapy and a low-calorie diet led to no greater weight loss than semaglutide alone in the STEP 1 trial after 68 weeks of treatment [29,30]. Our findings, therefore, highlight the need to identify preoperative predictors to maximize postsurgery weight loss.

No favorable impacts of the BARI-LIFESTYLE intervention on
HRQoL and mental health outcomes were observed, which corroborates findings from previous RCTs [24,31]. This could be explained by the fact that the peak improvement in HRQoL and mental health occurs in the 12 months following surgery [32,33], and so, any adjunctive lifestyle interventions would not offer additional impacts. and better preservation of fat-free mass after surgery [20,39]. Preventing loss of fat-free mass and BMD after bariatric surgery is crucial, as an excessive loss might negatively impact the ability to carry out activities of daily living and increase fracture risk over the long term [5]. This is important as the majority of people undergoing bariatric surgery are above the age of 30 [1,3], a point when muscle mass and strength start to decline involuntarily, with the rate of decline accelerating after the age of 60 [40]. Women especially have an even higher fracture risk once they enter menopause, which represents more than 75% of patients undergoing bariatric surgery [1,3]. Given that the new antiobesity medications that engender weight loss in the bariatric surgery range are likely to be used in a similar group of patients [41,42], studies are needed to examine the impact of these on habitual physical activity and body composition changes.

Strengths and limitations
The main strength of BARI-LIFESTYLE is the use of a high-quality trial design. The adherence rate of the tele-counseling sessions in the present study was also high (90.7%). Both BIA and DXA, the latter being a reference gold standard [43], were used to assess the changes in body composition. The habitual physical activity was measured objectively using an accelerometer, as patients who have undergone bariatric surgery tend to overreport their physical activity levels when assessed using the conventional questionnaires [44]. In addition, the present trial used an obesity-specific HRQoL questionnaire [15] and a validated instrument to assess depressive symptomatology [16], both of which are sensitive to detecting small changes.
We acknowledge several limitations in our study. The prespeci- qualitative substudy that we undertook, the participants reported that the intensity of the remote exercise was similar to the gym exercise they attended before the lockdown. The virtual exercise classes also helped with adherence by removing the barriers when attending the in-person exercise classes before COVID such as traveling time, parking issues, and poor weather conditions [13]. Last, as previously mentioned, approximately one-quarter of participants randomized to the intervention group did not enroll in the exercise program, with reasons similar to a previous RCT study [27].

Implications for practice and considerations for future studies
The high adherence toward the remote dietary tele-counseling suggests that this method of service delivery could be implemented as an alternative to the usual in-person dietetics follow-up. Indeed, a recent survey reported that 81.7% of patients were satisfied with the virtual bariatric follow-up clinics provided during the COVID-19 pandemic mainly because of the time and cost efficiency to attend, with 76% of patients expressing their preference to continue with virtual care [45].
Our trial provides a low-dose exercise program and has no favorable impact compared with previous positive studies [34][35][36][37][38], suggesting that higher frequency, duration, and intensity of resistance training are needed to achieve the beneficial impact on non-weightrelated outcomes.
For future studies, the following recommendations should be considered: 1. We previously showed that patients who lose less than 1 lb a week during the 3-to 6-month postoperative period will not achieve a maximal %WL of more than 20% [4]. Therefore, future studies should target patients who exhibited early poor weightloss response as a main inclusion criteria.
2. The optimal timing to deliver a postsurgery lifestyle intervention program remains unclear, although our data suggest no impact when delivered within the first 12 months after surgery. Therefore, future studies should consider randomizing patients to either receive the lifestyle intervention during the substantial weight-loss phase or during weight-loss maintenance.
3. Studies with longer term follow-up are needed to provide further evidence of whether an early adjunctive lifestyle intervention would have favorable impacts on long-term weight-loss maintenance or the rates of weight regain.
4. Our study highlights that marked weight loss alone, despite support to increase habitual physical activity, does not lead to increased MVPA or reduced sedentary behavior. Given that the use of antiobesity medications that engender weight loss in the bariatric surgery range is likely to become widespread [41,42], future studies should focus on strategies to improve these nonweight-related outcomes in patients who lose significant weight with lifestyle programs, pharmacotherapy, or bariatric surgery.

FUNDING INFORMATION
The BARI-LIFESTYLE trial was supported by the National Institute for