Introduction

Bariatric surgery (BS) has been considered superior to both medical therapy and intensive lifestyle interventions to achieve significant and long-lasting weight loss in obese individuals [1]. BS has also been shown to restore the ability to perform activities of daily living in obese individuals [2]. Exercise training following bariatric surgery can provide additional benefits in the obese population (i.e., increase maximum oxygen consumption (VO2max) and greater improvements in skeletal muscle metabolism) and has also been shown to be an efficient additional intervention to increase cardiorespiratory fitness in post-BS patients [3].

Although, a few studies show VO2max increase after aerobic and/or resistance training in post-BS patients [4, 5], other studies failed to show an additive effect of exercise training on exercise capacity compared to BS alone [6, 7]. This inconsistency could be due to the lack of standardization among exercise training protocols, cardiorespiratory fitness testing procedures, time to start intervention post-BS, intervention duration, lack of statistical power, or other methodological differences such as assessment of VO2max in absolute or relative values. Therefore, our goal was to perform a meta-analysis in order to determine the effect size of exercise training (aerobic, resistance, or both) on VO2max in adults following bariatric surgery weight loss.

Methods

Data Sources and Searches

The search was conducted on studies published prior to August 21, 2018 in the MEDLINE database (through PubMed), using the following terms: “Bariatric Surgery” OR “Metabolic Surgery” OR “Bariatric Surgeries” AND “exercise” and included clinical studies, clinical trial, comparative studies, multicenter studies, observational studies, and randomized controlled trials.

Study Selection

For the meta-analysis, we considered all eligible prospective cohort models, human subjects (adults over 19 years old), written in English language, studies investigating the association between CRF and measured cardiorespiratory variables following BS by cardiopulmonary exercise testing, and studies which included a description of the exercise training protocol. Exclusion criteria: non-original publications (i.e., errata, letters to the editor, reviews), retracted papers, guidelines, qualitative studies, lack of an exercise training intervention (aerobic, resistance, or combined), and studies which did not investigate the effects of exercise training on VO2max. Two reviewers (AG and AS) independently screened all titles and abstracts for eligibility. Any disagreements about inclusion of studies were resolved by discussion.

Data Extraction and Statistical Analysis

The average, standard deviation, and sample size of VO2max following BS at baseline and post-intervention in exercise training groups and control groups (CG) were extracted from each study. The meta-analysis was performed using comprehensive meta-analysis software version 3.3.070. The effect size was calculated based on the standardized mean difference (SMD) of the change in VO2max (baseline to post-intervention) between exercise training groups and CG. The quality of studies was assessed using the PEDro scale (0–10). Heterogeneity of studies was assessed using I2 (p = 0.05). The risk of publication bias was assessed by Egger’s test (p = 0.05).

Results

Description of Included and Excluded Studies

From 306 studies found in the first search, only 7 randomized control trials (RCTs) were met the criteria for inclusion. Descriptions of included study characteristics are described in Table 1. The studies that did not met the criteria included an erratum (n = 1); a retracted paper (n = 1); review papers (n = 106); letters to the editor (n = 9); other non-original studies (n = 13); guidelines (n = 7); qualitative studies (n = 5); studies with children (n = 2); no exercise effects on VO2max (n = 151); no aerobic, resistance, or combined training intervention (n = 3); and study without extracted data (n = 1).

Table 1 Characteristics of included studies in the meta-analysis
Full size table

Quality of the Studies and Publication Bias

Among the seven studies selected, three studies scored 4 [6, 8, 10], one study scored a 6 [4], and other three studies scored 7 [3, 5, 7] on the PEDro scale. Two questions regarding blinding patient and care provides were nulled as it is not possible in exercise intervention RCTs. The p value for Egger’s test was 0.25, suggesting no risk of publication bias. Since there was no statistical significance for heterogeneity (I2 = 0.0%, p = 0.76), fixed effect models were selected for all analysis.

Outcomes

The pooled results of these studies demonstrate that exercise training leads to a moderate and significant increase of VO2max in post-BS patients (SMD = 0.430, 95% CI 0.157; 0.704, p = 0.002) (Fig. 1).

Fig. 1
figure 1

Forest plot for differences between exercise training (ET) and control group (CG) (Reviewer #2/Comment #2) on reduction or increase VO2max effect size. SMD, standardized mean difference; LL, lower limit of 95% CI; UL, upper limit of 95% CI; CI, confidence interval

Full size image

Discussion

The results of this meta-analysis demonstrate that exercise training following BS results in a moderate increase in VO2max following BS. This suggests an important risk reduction effect of exercise in this population when considering that CRF is inversely correlated with mortality. Increasing CRF by 1 MET (3.5 ml/kg/min) is associated with a 10 to 12% of reduction in mortality rate independent of disease status [9, 11]. An isolated meta-analysis analysis with a subgroup of 5 studies measuring relative VO2max showed that exercise training following BS results in a raw mean increase of 0.73 ml/kg/min (95% CI 0.0; 1.47) [3, 5,6,7, 10].

Following BS, patients undergo substantial and immediate weight loss and tend to become more physically active which may increase VO2max [10, 12]. The increase in relative VO2max could be due to reduction in body weight, as confirmed by the VO2max increase in the CG (Table 1). In this context, it is very important to isolate the effect of the exercise on VO2max from the weight loss effects on exercise capacity [10]. In this study, we were not able to completely eliminate this confounding factor (the majority of the studies included did not provide absolute VO2max values); however, reducing body weight alone was not enough to increase the VO2max post-BS in two studies [3, 6] and the comparison with a CG in the present meta-analysis ensure the exclusive training intervention influence on the increase in VO2max post-BS.

Although the reduction in body weight following BS could partially contribute to relative VO2max increase, CRF could be negatively affected by the cardiorespiratory and metabolic function changes post-BS such as elevated cardiac stress, loss of muscle mass and strength, and the deterioration of oxidative muscle metabolism [4, 8, 10, 12]. Given these changes and the role of skeletal muscle metabolism in exercise capacity, the restoration or prevention of muscle mass [7, 10] must be considered in exercise training protocol design [5] for patients undergoing BS. This demonstrates the importance of including resistance/aerobic or combined training in exercise training protocols for patients undergoing BS in order to improve muscle mass and CRF. Interestingly, our exploratory analysis also demonstrated no significant differences between increases in RCTs using absolute [4, 8] or relative VO2max [3, 5,6,7, 10] (SMD 0.60 (95% CI 0.12; 1, 07) and SMD 0.34 (95% CI: 0.01; 0.68), respectively), in RCTs using programs with training durations of ≤ 3 months [3, 6, 7, 10] and > 3 months [4, 5, 8] (SMD 0.36 (CI 0.004; 0.71) and SMD 0.53 (CI: 0. 10; 0.95), respectively). The results from these analyses indicate that exercise training can significantly improve VO2max, in both shorter and longer interventions. The results were similar regardless of the use of relative or absolute VO2max.