Introduction
Women are more prone to clinically severe obesity (body mass index, BMI, ≥40, or a BMI ≥35 with obesity-related comorbidities) than men [1]. Individuals with classes II (BMI 35–39.9) and III obesity (BMI ≥40) are at higher risk of developing metabolic and cardiovascular diseases when compared with class I obesity (BMI ≥30–34.9) [2], and this places twice the burden on health care systems when compared with healthy weight adults [3]. Health care utilisation for cardiometabolic disorders does not fully account for the true cost of obesity. Increasing obesity and central adiposity lead to greater risks of developing obesity-related disabilities and musculoskeletal injuries [2], which may negatively impact on occupational capacity, social engagements and activities of daily living. Individuals with severe obesity experience a feed-forward cycle of low exercise capacity, physical disability and breathlessness leading to physical inactivity, in turn leading to further weight gain [4] and progressive loss of physical function [5], resulting in frailty in later life [6]. Therefore, strategies to enhance the fitness of adults with clinically severe obesity are important for the prevention of later life physical disability.
Clinical practice guidelines suggest diet and exercise as first-line treatment strategies for overweight and obesity [7]. These interventions produce moderate weight loss in the short term, but the weight is often regained in the long term [8]. Much of the literature concerning energy restriction and physical activity interventions focus on adults who are categorised as having overweight or class I obesity [9]. Whilst exercise can limit the loss of lean body mass and improve physical function when added to energy-restricted diets in adults with class I obesity, exercise plus diet does not produce substantial additional weight loss when compared to diet alone [8, 10]. Whether these findings apply to more severe classes of obesity has not yet been established and so was a focus of this study. There is some evidence that adaptations to exercise could be blunted in obesity [11] due to impaired skeletal muscle protein metabolism, muscle fibre typology [12], sympathetic nervous reactivity [13], brown adipose activation [14] and altered skeletal muscle mitochondrial content and function [15].
An appropriately designed exercise training programme can improve physical fitness and body composition [10, 16]; however, despite the increasing prevalence and the unique challenges that adults with clinically severe obesity face, this population group has not received adequate attention in long-term clinical trials that focus on exercise training. Therefore, the main aim of this study was to investigate whether combined exercise training and an energy-restricted diet improved physical fitness and body composition, compared with an energy-restricted diet alone for women with clinically severe obesity under free-living conditions.
Materials and Methods
Sixty women with clinically severe obesity were recruited for a 12-month, parallel group, randomised clinical trial from Melbourne, Australia, between September 2011 and June 2015 (Fig. 1). All data were available for analysis from November 2016. All participants provided written informed consent and obtained medical clearance prior to enrolling. The full study protocol has been published previously [17]; however, in short, the inclusion criteria were premenopausal women aged 18–50 years, with a BMI ≥40 or a BMI ≥30 with at least 1 obesity-related comorbidity. The exclusion criteria were unstable cardiovascular conditions, type 1 diabetes, active musculoskeletal conditions preventing exercise participation, pregnancy or planning pregnancy, weight loss greater than 5 kg in the past 3 months, using weight loss medications, previous bariatric surgery, medications which significantly influence weight and more than 150 min of moderate intensity exercise per week. Baseline data collection was obtained immediately prior to randomisation. Randomisation was performed by an independent researcher external to the project in a block-randomised (blocks of 6) fashion to either energy restriction alone (ER) or energy restriction plus exercise training (EXER) groups using sealed opaque envelopes (online suppl. Appendix S1; for all online suppl. material, see www.karger.com/doi/10.1159/000506643). Participants and researchers involved in delivering participant interventions were not blinded to the group allocation. All participants randomised to ER were informed that they would have the opportunity to attend supervised exercise training at no cost upon study completion for 3 months in an attempt to improve participant retention.
Energy-Restricted Dietary Intervention
Both groups received identical energy-restricted diet and dietary consultations (Fig. 2) which were designed to resemble that of standard clinical care for free-living adults by dietitians with experience with severe obesity. The consultations and intervention consisted of a very low energy diet protocol (Optifast®, Nestle, Australia) with an energy intake ranging from 450–680 kcal (1,900–2,800 kJ) per day during the intensive phase to 1,200 kcal (5,000 kJ) per day during the stabilisation phase. Participants met every 2 weeks for the first 12 weeks with an accredited practicing dietitian and continued as clinically relevant for the remaining 9 months. All participants started with an intensive phase for at least 12 weeks if tolerated, moving into a transition phase for 6–12 weeks, maintenance phase for the next 6–12 weeks, and then a stabilisation phase for the remaining 6 months. The diet transitioned gradually from full meal replacements to whole food as clinically indicated. There was no attempt to directly control for energy intake beyond the advice and support provided by the study dietitian. The average daily macronutrient content during the intensive phase (3 meal replacement products) for the Optifast® diet was: protein 54 g, carbohydrate 59 g, fat 14 g; plus 2 cups of low starch vegetables and 1 teaspoon of oil of the participant’s choice. Participants randomised to ER were encouraged to be active but were not provided with specific advice on exercise.
Fig. 2.
Overview of the exercise and dietary intervention. VLED, very low energy diet; * Diet prescription was based off the Optifast® clinical treatment protocol. Phase transitions were fluid and determined based on standard clinical practice decision making to facilitate energy deficit or stabilisation of body weight using a client-centred approach.
Exercise Training plus Energy-Restricted Diet Intervention
Participants randomised to EXER completed a supervised exercise training programme (Fig. 2) delivered by accredited exercise physiologists and postgraduate clinical exercise physiology students. Participants underwent a stepped-down approach to supervised exercise training designed to accommodate free-living adults (online suppl. Appendix S1). There was no attempt to control for the degree of energy deficit, energy expenditure or net energy balance between the two interventions. Each supervised exercise training session consisted of 20–30 min aerobic (60–80% heart rate reserve) and 30 min of resistance training (1–3 sets of 8–10 repetitions for 8 different upper and lower body exercises) for a total of 60 min. Exercise training consisted of 3 supervised training sessions per week for the first 6 weeks, 2 sessions per week for weeks 7–12, 1 session per week for the next 3 months and 1 session per fortnight for the final 6 months. The supervised training sessions were supplemented with self-administered sessions for a total of 300 min per week of moderate intensity exercise. The modes of exercise for the self-administered training sessions were developed in consultation with each participant based on their access to facilities and preferences.
Outcome Measures
Outcome measures were collected at baseline, 3, 6 and 12 months. Participants were instructed to fast overnight and to avoid caffeine for 18 h, heavy exercise and alcohol for 24 h prior to testing. The co-primary outcome measures were the difference between groups against time at 12 months for peak aerobic power (VO2 peak) expressed in mL ∙ kg–1 ∙ min–1 and lean body mass in kilograms (dual X-ray absorptiometry).
VO2 peak was assessed using a sign- and symptom-limited incremental treadmill test [17]. The speed remained constant throughout the test, and the gradient was increased by 2% per minute until the participant reached a reported rating of perceived exertion of 17/20 (very hard). Expired respiratory gases were collected and analysed using a breath-by-breath pneumotach system (Innocor Innovision version 6.15, Glamsbjerg, Denmark), and this was calibrated according to the manufacturer guidelines. HbO2 sat% using pulse oximetry was recorded at the end of each minute during the test and recovery.
Three compartment body composition and bone mineral density were assessed using a dual energy X-ray absorptiometry scanner (GE Lunar Prodigy Pro, Madison, WI, USA 53718) with software version 12.1 for total body fat mass, lean body mass, and total body bone mineral content and density. For instances where portions of the trunk remained outside of the scan area, half-body scans were completed. Body weight was measured with participants wearing a hospital gown without shoes. Waist circumference and hip circumference were measured standing using a steel tape measure.
Maximal muscular strength was determined using a 3-repetition maximum for bench press, leg press and seated row using standard resistance training equipment. Full protocol details for strength measures and fasting blood collection and analysis are described in online supplementary Appendix S1 and have been published previously [17].
Statistical Analysis
All randomised participants were analysed following an intention to treat analysis. Descriptive statistics, i.e. mean and standard deviations (mean ± SD), were calculated by treatment group for all the variables at baseline.
The change from baseline to each post-baseline time point (mean ± SEM)was compared between and within the groups using mixed models for repeated measures. This model included fixed effect group, time point (categorical), group and time point interaction, and was adjusted for baseline measurement (centred around mean of each group) by including baseline as a covariate. The model included all individuals who had baseline measurement and at least 1 post-baseline measure and used an unstructured variance-covariance structure to model within-individual errors. There was no formal adjustment for multiple comparisons. Results were reported as mean and standard error unless otherwise specified. All statistical analyses were performed using Stata (StataCorp version 14.2). The sample size for this study was determined based on previous evidence, plus a projected drop-out of 25%, and provided adequate statistical power for the two co-primary outcome measures of VO2 peak (n = 4, α = 0.05, power 0.9) [18] and lean body mass (n = 23, SD = 1.7, α = 0.05, power 0.8) [19]. A p value of less than 0.05 is considered to be statistically significant.
Results
Study Population
A total of 60 participants aged 37 ± 9 years (mean ± SD) and a BMI of 40.4 ± 6.7 were randomised, and 46 participants (76%) completed the study (Fig. 1). The characteristics of the two groups are shown in Table 1. All participants with data collected at baseline and at least 1 follow-up time point (n = 53) were included in the intention to treat analysis. Attendances at dietary consultations were similar between groups, 12 ± 5 (mean ± SD) dietary consultations in ER and 14 ± 4 (mean ± SD) in EXER (mean difference ± SEM; 1.9 ± 1.2; p = 0.139) (online suppl. Appendix S1). A summary of reported adverse events during study participation are presented in the supplementary material (online suppl. Appendix S1).
Peak Aerobic Power
VO2 peak relative to total body mass was significantly greater in both groups when compared to baseline at all time points (Table 2). The group difference at 12 months between EXER and ER for VO2 peak relative to total body mass was not statistically significant (mean diff. ± SEM; 2.3 ± 1.6 mL • kg–1 • min–1, p = 0.15). EXER resulted in significantly greater gains in VO2 peak compared to ER at 3 months (mean diff. ± SEM; 2.5 ± 0.9 mL ∙ kg–1 ∙ min–1, p = 0.006) and 6 months (mean diff. ± SEM; 3.1 ± 1.2 mL ∙ kg–1 ∙ min–1, p = 0.007) (Table 2).
Table 2.
The effects of exercise training when added to energy restriction on physical function outcomes
Absolute VO2 peak (L ∙ min–1) was not significantly different for both EXER and ER over time and there were no significant between-group differences (Table 2). VO2 peak expressed relative to lean mass increased above baseline for EXER at 3 months (4.23 ± 1.51 mL ∙ kg–1 ∙ min–1; p < 0.05) and 6 months (4.84 ± 1.39 mL ∙ kg–1 ∙ min–1, p < 0.05), but not at 12 months. There were no changes for ER at any time point, and there were no between-group effects at any time point (Table 2).
Body Composition
The addition of exercise training to energy restriction facilitated body mass loss at 3 months (EXER vs. ER) but not at 6 or 12 months (Table 3). Lean body mass was similarly reduced at all time points compared to baseline for both groups, and there was no difference between EXER and ER at 12 months for lean body mass (0.3 ± 1.0 kg, p = 0.352; Table 3). Both EXER and ER exhibited significant reductions in total body mass at all time points compared to baseline (Table 3). A 5% body mass loss or greater was achieved by 60% of participants in ER, and 82% in EXER at 12 months (Fig. 3). Fat mass was reduced to a greater extent in EXER at 3 months (p < 0.001), trended to be greater at 6 months (p = 0.062) and was not significantly different at 12 months when compared to ER (Table 3). At 12 months, the ratio of total mass loss to lean body mass loss was 5.6:1 kg in ER, and 9.3:1 kg in EXER. Expressed as percentages, 17.9% of total mass loss in ER was comprised of lean body mass, and 10.8% in EXER at 12 months.
Table 3.
The effects of exercise training when added to energy restriction on anthropometric outcomes
Fig. 3.
Individual percentage weight change between baseline and 12 months for energy restriction and energy restriction plus exercise training. Each bar represents a single participant for their proportion of weight change from baseline to 12 months. The dashed line represents a 5% weight loss at 12 months. The dotted line represents 10% weight loss at 12 months.
Muscle Strength
Total muscle strength (calculated as the sum of weight lifted for all three strength tests) increased significantly in EXER but not in ER at 3 months (mean ± SEM; 15.0 ± 4.0 kg, p < 0.001, –2.88 ± 4.0 kg, respectively; p = 0.476), 6 months (mean ± SEM; 30.5 ± 4.6 kg, p < 0.001, 2.8 ± 4.5 kg, respectively; p = 0.530) and 12 months (mean ± SEM; 36.5 ± 5.6 kg, p < 0.001, 7.1 ± 6.0 kg, respectively; p = 0.242) (Table 2).
Cardiovascular Risk Factors
No differences between groups at 12 months were observed for total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglycerides, systolic or diastolic blood pressure or fasting plasma glucose (Table 4). At 12 months, within-group improvements for high-density lipoprotein cholesterol were observed for both groups. Additionally, ER showed improved fasting glucose while EXER did not. Improvements for fasting glucose occurred in both groups at 3 and 6 months but were maintained in ER (mean –0.18 ± 0.07, p < 0.05) and not EXER (mean –0.06 ± 0.07, p > 0.05) at 12 months. Systolic and diastolic blood pressure improved in both groups at 3 and 6 months but not at 12 months. There were no significant differences between groups at any time point for systolic or diastolic blood pressure.
Table 4.
The effects of exercise training when added to energy restriction on markers of cardiometabolic risk
Discussion
The addition of exercise training to an energy-restricted diet in women with severe obesity did not lead to greater changes in aerobic power or preservation of lean mass at 12 months, compared to energy restriction alone. Both groups experienced significant improvements in aerobic power and a modest, yet statistically significant loss of lean mass at the trial end.
Few prolonged studies such as this 12-month trial have evaluated the effect of exercise training when added to energy restriction in premenopausal women with clinically severe obesity for changes in lean mass. One 12-month study [20] showed that exercise training added to diet limited the loss of lean mass; however, this was in older adults and may not be generalisable to younger cohorts. Preservation of lean mass for shorter duration interventions with younger adults with obesity class I [21], class II [22] and class III [23-25] showed mixed results. The addition of exercise training to surgically induced weight loss in severe obesity may positively influence lean mass [26]; however, randomised controlled trials also showed mixed results. The addition of moderate intensity aerobic exercise training [27], aerobic plus resistance training [28] or resistance training alone [29, 30] over 12 weeks does not lead to differences in lean mass when compared to standard care control groups in participants after Roux-en-Y gastric bypass surgery. There is only one study of 12 months duration which showed that combined aerobic and resistance training after Roux-en-Y gastric bypass protected against lean mass loss when compared to the nonexercise surgery group [31]. However, this was not a randomised clinical trial and the method of group allocation was not described.
While there is evidence that a loss of lean mass occurs concurrently with weight loss in the order of approximately 30% of total weight loss for women with class II obesity [32], we showed that weight loss with and without a structured exercise programme resulted in lower than predicted relative lean mass loss (EXER, 10.8%; ER, 17.9%). The addition of exercise training in our study did not result in a regain of lean mass as reported in a shorter 12-week study of men and women [23]. However, that study implemented a higher-intensity resistance training protocol in the absence of aerobic training, was of shorter duration and differed with respect to participant characteristics. Although various studies have explored the effect of exercise training during weight loss in adults with obesity, there remains insufficient evidence to confirm the ability for protection of lean mass during hypocaloric diets or bariatric surgery in adults with clinically severe obesity. Our data confirm that the addition of a combined aerobic and resistance training programme to energy restriction did not lead to protection of lean body mass when compared to energy restriction alone in free-living women with clinically severe obesity. However, the supervised resistance training intervention in our study also reduced in frequency to twice per month in the last 6 months of the trial. This volume of training is not likely sufficient to observe a gain in lean mass unless participants chose to complete resistance exercise outside of supervised sessions. In the absence of such differences between groups for lean body mass, total-body strength increased in EXER when compared to ER at 12 months. These results are consistent with previous evidence to show that resistance training increases strength and skeletal muscle cross-sectional area in the absence of lean mass gain during a hypocaloric diet [33], which may assist in the protection of future physical disability and early mortality [34].
Peak aerobic fitness is an important determinant of physical function and future physical disability [35], and a strong predictor of cardiovascular events and early mortality [36]. We showed that both ER and EXER enhanced aerobic power relative to body weight at 12 months by 3 and 5.4 mL ∙ kg–1 ∙ min-1, respectively, but there was no significant difference between groups at the study end. This is in contrast to a previous meta-analysis of long duration diet and exercise studies in adults with overweight and obesity, where combined exercise training with an energy-restricted diet led to greater aerobic fitness benefits when compared to energy restriction alone [37], but none of the studies included in that review reported on aerobic fitness in premenopausal women with class II obesity. To our knowledge, no long-term studies have evaluated the effects of combined exercise training and energy restriction in premenopausal women with clinically severe obesity for changes in aerobic fitness. One study of premenopausal women with class I obesity followed over 12 months showed aerobic exercise training during an energy-restricted diet led to greater aerobic power than diet only [38]. This study consisted of a similar aerobic training prescription as our study; however, supervised training continued 3 times per week for 12 months, possibly influencing adherence to the intended training stimulus and therefore the results [39]. This explanation may support our observations at 3 and 6 months where aerobic power increased more in EXER when compared to ER when supervision was greater. Encouragingly, both groups demonstrated improved aerobic power relative to body weight; however, further evidence is required to determine long-term effective exercise prescription approaches which may be suitable for premenopausal women with clinically severe obesity.
Weight loss and specifically fat mass loss are important for the reduction of cardiovascular disease risk and all-cause mortality [40]. Previous evidence from long-term studies shows that exercise when added to diet provides a modest additional loss of body mass and fat mass in adults with overweight and obesity [37]. Our results show that exercise when added to an energy-restricted diet did not reduce fat mass or weight more when compared to the energy-restricted diet-only group at 12 months for women with clinically severe obesity. Similar studies in adults following bariatric surgery confirm a modest additive effect of exercise [41]; however, the only study in women completing aerobic plus resistance training after bariatric surgery showed there was no additional fat mass loss at 24 months when compared to the non-intervention surgery-only group [42]. The lack of additional weight loss and fat mass loss beyond the 3-month follow-up may be associated with a non-linear weight loss trajectory observed in previous weight loss research [8], or the stepped-down approach to exercise supervision. However, the study design does not allow for confirmation of the reason for the lack of difference between groups at 12 months. There continues to be a lack of research in this population group and inconsistent results further support the need for additional research in this area.
Traditional markers of cardiometabolic risk remained relatively unchanged from baseline to the follow-up at 12 months. The non-significant changes in systolic and diastolic blood pressure in both groups likely reflects that participants were normotensive at baseline. Previous evidence suggests that a similar weight loss to that achieved in this study leads to a 7- to 10-mm Hg reduction in systolic blood pressure and a 7-mm Hg reduction in diastolic blood pressure [43]; however, this was not the case in these normotensive women with severe obesity. Similar improvements at 12 months were observed in both groups at 12 months for total cholesterol, low-density lipoprotein cholesterol and triglycerides when compared to baseline. Both groups demonstrated improved high-density lipoprotein cholesterol at 12 months when compared to baseline, while only the ER group demonstrated reduced fasting plasma glucose. It is well accepted that lifestyle interventions which result in 5% or more weight loss using diet, exercise or combined treatments [44] lead to improved biochemical markers of cardiometabolic disease risk in adults with obesity and existing metabolic impairments. The absence of improved cardiometabolic markers in this cohort is likely due to the relatively normal baseline findings.
Based on the results presented here, it appears that premenopausal women with clinically severe obesity do not achieve greater weight loss, fat mass loss or improve aerobic power more when prescribed an aerobic plus resistance training programme with energy restriction over an energy-restricted diet alone under free-living conditions. These findings are important because little is known about how premenopausal women with clinically severe obesity respond to a diet and exercise intervention in free-living conditions for measures of body composition and physical fitness. Although the results presented here showed little additional benefit outside of strength improvement at 12 months with the addition of exercise training, it is reasonable to expect that an exercise intervention that strengthens the support for these women more than in our trial may be effective in enhancing body composition and fitness beyond 6 months in premenopausal women with clinically severe obesity. Based on the superior results at 3 months for the majority of the outcomes in EXER when compared with ER, it is important that further research evaluates the effect of supervision models which retain intensive diet and exercise interventions over a longer period of follow-up.
Strengths and Limitations
Strengths included the prolonged period of follow-up, the method used to evaluate body composition, the recruitment of women with clinically severe obesity and the ecologically valid study design. Limitations included: (i) the reasonably high level of participant attrition that compromised our anticipated statistical power for the primary end points at 12 months; (ii) participants only achieved slightly more than half of the intended exercise training at all time points (see online suppl. Appendix S1). This may bring into question whether 300 min of weekly exercise is feasible for women with clinically severe obesity, despite the recommendations that 200–300 min per week is necessary to promote long-term weight loss [45]; (iii) the groups were not matched for the degree of energy deficit, which makes conclusions about the actual effect of exercise in addition to diet difficult. However, this was not the intention of the study design. This study was developed to compare the effect of the two interventions, in free-living conditions under the guidance of experienced allied health professionals in a community setting. This is an important distinction to make given the real-world applicability the findings from this study may provide for clinical practice.
Conclusion
Combining exercise training with an energy-restricted diet did not lead to greater aerobic power or improvements in lean body mass, fat mass or total body mass at 12 months when compared to energy restriction alone for premenopausal women with clinically severe obesity in free-living situations. Further research is required to determine an effective and feasible exercise and diet approach which can be applied in the community setting for these high-risk individuals.
Acknowledgements
The authors would like to thank the YMCA Victoria, Australia, for its support of this trial, offering in-kind access to its gymnasium facilities at Kew Recreation Centre, RecWest Footscray, Ascot Vale Leisure Centre, Knox Leisureworks, Burwood Fitness Centre and Endeavour Hills Leisure Centre.
The authors would also like to acknowledge the participants who volunteered their time to support this study. The full study protocol has been published previously [17]. Individual participant data will not be made available or shared outside of the research group.
Statement of Ethics
The study was approved by the Human Research Ethics Committees of the Alfred Hospital (HREC reference 2011-154) and Deakin University (HREC reference 330-11) in accordance with the guidelines of the National Health and Medical Research Council of Australia (Australian New Zealand Clinical Trials Registry, ACTRN12611000694910).
Disclosure Statement
John B. Dixon is a board member of Nestlé Health Sciences Australia and has received payment for consultancy from Apollo Endosurgery, Bariatric Advantage, iNova Pharmaceuticals and Novo Nordisk. The laboratories of Professors Dixon and Lambert have recently received research funding from Medtronic. Gavin Lambert has received payment for consultancy from Medtronic. Clint T. Miller, Steve F. Fraser, Steve E. Selig, Toni Rice, Mariee Grima, Daniel J. van den Hoek and Carolina Ika Sari declare no conflicts of interest.
Funding Statement
Nestlé Health Australia provided VLED products at cost in support of the study. John B. Dixon and Gavin W. Lambert were supported by Senior Research Fellowships of the Australian National Health and Medical Research Council. While this study has not received any other external funding, work at the various institutions was supported in part by the Victorian Government’s Operational Infrastructure Support Program.
Author Contributions
The authors declare substantial contribution by all authors to the development of the study design, acquisition of data, analysis and preparation of the manuscript.
Get Permission
PubMed ID
