This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) statement (Moher et al., 2009). The Population, Intervention, Comparator, Outcomes, and Study design (PICOS) approach was used to identify the inclusion criteria (Table 1). Studies that have examined the effects of implementing a structured exercise or a physical activity intervention before BS on postoperative outcomes “body composition, weight loss, physical activity level, performance, and metabolic parameters” were eligible for inclusion. The studies were included in the current systematic review if they met the following criteria: 1) published in peer-reviewed journals, 2) included adult participants, and 3) compared BS outcomes pre- and/or postoperatively. The studies were excluded if they 1) reported only subjective measures, 2) were not written in English, or 3) were retrospective. Moreover, review articles were not included in the current systematic review.

Literature searches were conducted in four electronic databases, including PubMed, Institute for Scientific Information (ISI) Web of Knowledge, Web of Science, and SPORTDiscus, to identify studies of preoperative exercise intervention or preoperative PA practices using the search terms “bariatric surgery” or “weight loss surgery” or “obesity surgery” or “weight reduction surgery” or “biliopancreatic diversion” or “laparoscopic band” or “lap and” or “gastric band” or “gastric bypass” or “gastroplasty” or “gastric sleeve” or “sleeve gastrectomy” and “preoperative exercise intervention” or “preoperative physical activity” or “preoperative lifestyle modification.”

The search was completed with a manual search of reference lists of key articles. Since the scope of this review is large in terms of outcome measures, a systematic review and not a meta-analysis was performed.

The final screening was performed by the principal investigator (GJ) based on the relevance of the inclusion and exclusion criteria and the identified items for assessing the effects of preoperative exercise intervention on anthropometric characteristics and body composition (e.g., body mass, body fat, and BMI), physical performances (e.g., muscular strength and physical capacity), cardiorespiratory fitness and function (e.g., oxygen uptake and heart rate), energy expenditure and metabolism parameters (e.g., resting metabolic rate, insulin resistance, and lipid profile), and hospital stay in obese adults of both genders undergoing BS using PICOS criteria. If the citation showed any potential relevance, the abstract was screened. When abstracts indicated potential inclusion, full-text articles were reviewed.

Our primary research identified 999 records, including 802 duplicates (Figure 1). After screening titles, abstracts, and full texts, 21 studies were included in our final analysis, and the characteristics of these studies are displayed in Table 2.

Out of 21 studies, 19 were prospective cohorts (Hickey et al., 1999; Still et al., 2007; Alger-Mayer et al., 2008; Colles et al., 2008; Funderburk and Callis, 2010; Marcon et al., 2011; King et al., 2012; Baillot et al., 2013; Browning et al., 2014; Bond et al., 2015a; Bond et al., 2015b; Baillot et al., 2016; Baillot et al., 2017; Bond et al., 2017; Marcon et al., 2017; Baillot et al., 2018; Daniels et al., 2018; Gilbertson et al., 2020; García-Delgado et al., 2021) and compared pre- to post-BS or pre- to postintervention outcomes in adult patients (Table 2). In total, 14 studies introduced an intervention pre-BS, of which 10 were randomized controlled trials that used a structured exercise program (Funderburk and Callis, 2010; Marcon et al., 2011; Baillot et al., 2013; Baillot et al., 2016; Baillot et al., 2017; Marcon et al., 2017; Baillot et al., 2018; Daniels et al., 2018; Gilbertson et al., 2020; García-Delgado et al., 2021), three studies used a physical activity program (Bond et al., 2015a; Bond et al., 2015b; Bond et al., 2017), and one study used a lifestyle modification program (Still et al., 2007) (Table 2). Among the 14 studies with an intervention, 11 studies performed a pre- to postintervention comparison, nine with exercise (Funderburk and Callis, 2010; Marcon et al., 2011; Baillot et al., 2013; Baillot et al., 2016; Baillot et al., 2017; Marcon et al., 2017; Daniels et al., 2018; Gilbertson et al., 2020), and two with PA intervention (Bond et al., 2015a; Bond et al., 2015b), while the remaining three studies performed a pre- to post-BS comparison (Still et al., 2007; Bond et al., 2017; Baillot et al., 2018) (Table 2).

Nine studies reported the effect of exercise training on anthropometric variables, among which five studies found no changes induced by the intervention (Table 3). Baillot et al. (2013) found a reduction in body mass, BMI, and fat mass after 12 weeks of supervised combined endurance and strength training (PreSET). In another study, Baillot et al. (2018) compared BMI between usual care and PreSET groups on seven different occasions (preintervention, 12 weeks postintervention, 2 weeks pre-BS, and 3, 6, 9, and 12 months post-BS). They found that the PreSET group experienced a greater decrease in BMI than the usual care group at 9 and 12 months post-BS. Furthermore, (Marcon et al. (2011 and 2017) reported larger decreases in body mass and BMI in the experimental groups than in the control group in two studies (Table 3). Funderburk and Callis (2010) reported a reduction in body mass after 12 weeks of supervised aquatic exercises, without a difference between the aquatic exercise and control groups. The rest of the studies did not find any anthropometric differences when comparing pre- to postintervention states or when comparing experimental (with exercise intervention) to control groups (without exercise intervention) (Hickey et al., 1999; Baillot et al., 2016; Baillot et al., 2017; Daniels et al., 2018; Gilbertson et al., 2020) (Table 3).

Nine studies reported the effect of exercise intervention on physical fitness parameters (Funderburk and Callis, 2010; Marcon et al., 2011; Baillot et al., 2013; Baillot et al., 2016; Baillot et al., 2017; Marcon et al., 2017; Baillot et al., 2018; Daniels et al., 2018; Gilbertson et al., 2020), of which two found no changes in measured parameters induced by the intervention (Marcon et al., 2017; Gilbertson et al., 2020) (Table 3). All four studies that used concurrent training (endurance and strength training) reported improvement in cardiovascular and/or muscular fitness parameters (Baillot et al., 2013; Baillot et al., 2016; Baillot et al., 2017; Baillot et al., 2018). Baillot et al. (2013 and 2016) compared patients’ baseline measures to 12 weeks post-training. They found an improvement in the 6-min walk test (6MWT) distance, percentage of theoretical 6MWT distance reached, 6MWT heart cost, half-squat test, and arm curl test. No differences were found in the sit-to-stand test or maximum aerobic capacity. In another study, Baillot et al. (2017) performed within (pre- and 12 weeks post-training) and between groups (training vs. conventional care group) comparisons. Compared to baseline measures, 6MWT distance, sit-to-stand repetitions, arm curl repetitions, and maximal aerobic capacity improved after 12 weeks of training. However, only the 6MWT distance, arm curl repetitions, and 6MWT heart cost improved in the training group compared to the conventional care group (Table 3). Another study between-group comparison revealed an improvement in 6MWT heart cost and the half-squat test for the training compared to the conventional care group (Baillot et al., 2018). Notably, the 6MWT distance was found to improve in two studies after aerobic training programs (Marcon et al., 2011) and aquatic exercise programs (Funderburk and Callis, 2010). Finally, only one study evaluated the effect of a 12-week resistance training program and found improvements in leg press strength, leg extension strength, and leg press muscle quality (Daniels et al., 2018) (Table 3).

Fourteen studies introduced an intervention, among which only four studies reported health-related parameters (Marcon et al., 2011; Baillot et al., 2016; Marcon et al., 2017; Gilbertson et al., 2020) (Table 3). Baillot et al. (2016) reported that BS candidates who were committed to a 12-week exercise intervention were protected from worsening of musculoskeletal pain. Gilbertson et al. (2020) tested the effect of a pre-BS aerobic exercise intervention on insulin sensitivity, metabolic flexibility, adipokines, and length of hospital stay. Marcon et al. (2011) evaluated the effect of an aerobic exercise program on the cardio-metabolic risk of BS candidates. A significant decrease in systolic and diastolic pressure and the Framingham risk score was found after 6 months of the supervised aerobic exercise program (Table 3).

Six studies examined the effect of PA and/or exercise intervention on body composition parameters pre- vs. postoperatively (Brandenburg and Kotlowski, 2005; Still et al., 2007; Alger-Mayer et al., 2008; Colles et al., 2008; Browning et al., 2014; Baillot et al., 2018), and five studies reported relevant data regarding body composition parameters before and after PA/exercise intervention in preoperative BS candidates (Baillot et al., 2013; Baillot et al., 2016; Baillot et al., 2017; Marcon et al., 2017; Baillot et al., 2018) (Table 4). Some studies showed a significant decrease in preoperative body mass (Still et al., 2007; Alger-Mayer et al., 2008; Bond et al., 2015a) or BMI (King et al., 2008) after an intervention. For these studies, the positive impact procured by a PA intervention prior to BS may be explained primarily by the improvement in PA and physical fitness levels among BS candidates, which is an important step toward improving their overall health parameters. Despite these promising results, it remains difficult to attribute all of these improvements solely to PA considering that many limitations have not been addressed, such as the lack of any control of PA (in the majority of cases, patients were only advised to practice PA) without excluding the interference of BS candidates’ existing conditions (such as BMI and comorbidities) as well as their diet and lifestyle prior to surgery.

Nonetheless, the effect of PA and exercise interventions on postoperative anthropometric parameters has been considered an interesting topic. Studies using interdisciplinary individual lifestyle counseling and helping BS patients adopt behavioral changes (e.g., to be active) (Still et al., 2007; Colles et al., 2008; Bond et al., 2015b) reported significant decreases in body mass during the post-BS period (Table 4). However, the lack of a control group, the small sample size, and the specificity of the sample may limit the generalizability of these results.

Other studies (Funderburk and Callis, 2010; Baillot et al., 2016; Baillot et al., 2017; Daniels et al., 2018; Gilbertson et al., 2020) applying supervised exercise training during the preoperative period found no significant pre- to postoperative changes in any body composition parameters between the intervention and usual care groups. It seems that BS induces a strong influence on weight loss and therefore can mask any eventual effect of preoperative intervention. In a study by Baillot et al. (2016), participants received lifestyle counseling for an average of 10.4 ± 4.0 months (5.8 ± 1.8 dietician and 5.6 ± 1.8 PA specialist visits) before inclusion in the study. Thus, significant changes might have occurred before inclusion in the surgical treatment option. In contrast, Baillot et al. (2018) reported a larger BMI decrease in the BS group undergoing preoperative exercise intervention (PreSET) compared to that in the usual care group and attributed this improvement to the higher loss of fat-free mass (FFM) in the PreSET group. Moreover, Gilbertson et al. (2020) reported a significant decrease in FFM in participants undergoing preoperative home-based walking for 30 min per day. In addition, Marcon et al. (2017) reported similar results after 4 months of aerobic and stretching exercises. These discrepancies among results might be primarily attributed to the duration and form of intervention. Moreover, the characteristics of patients prior to the PA and exercise intervention must be considered. In fact, many BS candidates encounter remarkable difficulties (e.g., musculoskeletal problems, preoperative fitness level) that might affect their exercise tolerance and adherence, consequently limiting or reducing the PA/exercise intervention benefits. Therefore, more support in selecting an appropriate activity along with a feasible monitoring technique is highly required in such a context.

Reduced physical fitness, reported mostly in BS candidates, may affect the BS results. Current evidence supporting the importance of increasing the physical fitness level in BS candidates is not abundant but is promising. In fact, several studies found that a preoperative intervention based on exercise or Pa that aimed at improving physical fitness and performance indicators (e.g., strength, 6MWT distance, and maximal aerobic capacity) among individuals awaiting BS may be an effective strategy to improve the BS candidates’ overall health parameters and their BS outcomes (Funderburk and Callis, 2010; Baillot et al., 2013; Baillot et al., 2016; Baillot et al., 2017; Marcon et al., 2017; Baillot et al., 2018; Daniels et al., 2018). In this regard, Baillot et al. (2013 and 2016), in their randomized controlled trial, reported a significant increase in some physical function parameters assessed with a test battery (6MWT, sit-to-stand, half-squat, and arm curl test) after 12 weeks of supervised exercise training either with or without an individual lifestyle counseling intervention. One study by Gilbertson et al. (2020) reported significant increases in the VO₂ peak among BS candidates after adding preoperative aerobic exercise to standard medical care. To the best of our knowledge, this is the only study to investigate the effect of a preoperative PA intervention on aerobic performance. The increase in the VO₂ peak may be associated with a shorter operation time and length of hospital stay and with prevention of muscle loss along with a concomitant increase in the PA level among BS candidates. However, the mechanism underlying these improvements remains to be studied.

The PA level is an interesting parameter that has been evaluated. In a long-term study (1 year after surgery), Baillot et al. (2018) reported that the addition of preoperative supervised exercise training to individual lifestyle counseling improved PA levels and submaximal physical fitness 1 year post-BS. Similarly, King et al. (2012) reported a gain in the PA level 1 year after surgery among patients with higher preoperative PA levels. Other short-term studies (Bond et al., 2015a; Baillot et al., 2016; Bond et al., 2017) reported similar results regarding PA levels in response to preoperative intervention. A small amount of evidence suggests that improvements in PA in response to a preoperative PA/exercise intervention may mainly be attributed to improvements in PA barriers, social interactions, and feelings of embarrassment (Bond et al., 2015a; Baillot et al., 2016). In addition, the impact of the overall improvement in the fitness level following a PA/exercise intervention aimed at improving the physical fitness of preoperative BS candidates cannot be overlooked. However, the small sample size and exclusion criteria applied in the aforementioned studies prevent their generalization to all subjects awaiting BS. Moreover, the study recruitment process was limited to volunteers who were able to frequently visit the facility and were without major functional limitations.

Only four studies have explored a limited number of health parameters (Marcon et al., 2011; Baillot et al., 2016; Marcon et al., 2017; Gilbertson et al., 2020). Baillot et al. (2016) looked at the effect of a 12-week exercise program on musculoskeletal pain and found that BS candidates can be protected from worsening of pain associated with daily life activities. Additional studies are required to confirm the impact of PA on musculoskeletal pain before and after BS. In a study by Gilbertson et al. (2020), patients undergoing preoperative EX + SC prior to bariatric surgery had a shorter length of hospital stay than patients undergoing preoperative SC. Two studies by Marcon et al. (2011 and 2017) reported a significant decrease in systolic and diastolic blood pressure and the Framingham risk score after 6 months of a supervised aerobic exercise program.

To date, the mechanisms responsible for such improvement have not been explored, although current evidence clearly favors a preoperative PA/exercise intervention for facilitating better postoperative outcomes. Few studies have investigated the role of preoperative interventions on BS outcomes. The differences in study design and the lack of randomized controlled trials decrease the evidence level of the results. Moreover, the heterogeneity of activities performed (with or without supervision), the inclusion criteria, and the interference of covariates (e.g., participant characteristics) were not well controlled. Finally, it is important to mention that the included studies were very small and had a short follow-up time, thus making the results less convincing. The data provided by this review did not consider BS procedures (i.e., type of surgery), making the results difficult to interpret.

In conclusion, this review summarizes the benefits of a preoperative PA/exercise intervention among BS candidates and highlights the importance of such strategies as a component of medical therapy. A good understanding of the beneficial effect of improving the preoperative physical condition on postoperative outcomes is highly recommended for future interventional studies to potentiate the beneficial effect of BS among obese candidates. Based on significant evidence, including optimized weight loss post BS, reduced cardiovascular risk, and increases in patients’ regular PA and fitness levels, there is a need to implement a PA/exercise program before BS to promote and optimize BS outcomes. Nevertheless, there is a need for future investigations in this field to determine the most appropriate form(s) of PA/exercise intervention according to the patient profile.