Responses of different functional tests in candidates for bariatric surgery and the association with body composition, metabolic and lipid profile

Individuals with obesity can have metabolic disorders and may develop impairments that affect the ability to exercise. The maximal incremental cardiopulmonary exercise test is widely used to assess functional capacity. However, submaximal tests such as the two-minute step test (2MST) and the six-minute walk test (6MWT) also allow this assessment. We propose to analyze whether body composition, metabolic and lipid profile influence the maximal and submaximal performance, and investigate these variables in response to different functional tests. Forty-four individuals with obesity, aged 18–50 years, underwent analysis of body composition, metabolic and lipid profile, incremental treadmill test (ITMT), 6MWT, and 2MST. One-way ANOVA, Pearson or Spearman correlation, and Stepwise multiple linear regression analysis were performed. ITMT induced a greater metabolic, ventilatory, cardiovascular, and perceived exertion demand when compared to the 6MWT and 2MST (p < 0.05). In addition, 2MST elicited a higher chronotropic (HR) and metabolic (V̇O2) demand when compared to the 6MWT (p < 0.05). Significant correlations were found between tests and body composition, metabolic and lipid profile. Fat mass and low-density lipoprotein can explain 30% of the V̇O2 variance in the ITMT; and fat mass, glucose, and performance in the 2MST can explain 42% of the variance of the distance walked in the ITMT. Obesity and its metabolic impairments are capable of influencing responses to exercise. ITMT generated greater demand due to the high stress imposed, however, 2MST demanded greater metabolic and chronotropic demand when compared to the 6MWT.


Materials and methods
Design and study population. This was a cross-sectional study and followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) 21 . This study was approved by the Federal University of Sao Carlos (UFSCar) Ethics Committee (966.613). The subjects received guidance on the procedures, and all participants gave written informed consent before the study's initiation. All methods were carried out in accordance with relevant guidelines and regulations (Declaration of Helsinki). The evaluations were conducted in the Cardiopulmonary Physiotherapy Laboratory at the UFSCar, Sao Carlos, SP, Brazil over a 2-year period (2016)(2017)(2018).
This study recruited individuals with obesity and comorbidities (body mass index (BMI) ≥ 35 kg/m 2 ) and individuals with morbid obesity (BMI 40 ≥ kg/m 2 ), from both genders and aged between 18 and 50 years, awaiting bariatric surgery in three different medical teams, with a non-random selection of convenience. We contacted them by telephone, explaining our procedures and asking about their interest in participating in our experimental. All selected individuals would perform the Roux-en-Y technique and all patients underwent the same routine protocol before surgery (exams, nutritional and psychological follow-up). The non-inclusion criteria were: patients with orthopedic or neurological impairments that impeded exercise testing; Baecke physical activity questionnaire score greater than 8; history consistent with heart disease; uncontrolled hypertension; uncontrolled and/or insulin-dependent diabetes mellitus; use of beta-blocker; respiratory diseases; the presence of any contraindications to exercise testing 15 ; cognitive deficit; difficulty in understanding and/or lack of adherence to the study procedures; users of illicit drugs; pregnant or postmenopausal women.
Evaluations. All evaluations were performed at the same time of day, to avoid different physiological responses, in a climate-controlled room with relative air humidity between 40-60% and temperature between 22-24 °C. Before the evaluations, the patients performed an adaptation and familiarization with the procedures and received guidance on the tests. The patients were instructed not to ingest any stimulant, and they did not perform strenuous activities within 24 h before the evaluations.
Patients underwent four days of evaluations, and on the first day, an anamnesis with a clinical history was carried out, and a questionnaire on the level of physical activity. On the second day, the individuals were submitted to blood collection, followed by an assessment of body composition. The participants underwent the ITMT on the third day. Respecting a minimum interval of 48 h, patients returned for the fourth visit, and performed 2MST and 6MWT, with a minimum interval of thirty minutes between tests. The tests on the fourth day were randomly chosen to avoid interference in the individuals' performance.
1st visit: anamnesis and Baecke's physical activity questionnaire. On the first day, an anamnesis was carried out with the individual's clinical history, with information on past history, medications in use and related comorbidities. The level of physical activity was assessed by information related to occupation, sports activities and leisure habits through the Baecke questionnaire 22 , which was validated and translated for Brazilian adults 23 . This questionnaire consists of a scale from one to five (five representing the most active), with eight questions related to occupation, four addressing sports activities, and four addressing habitual leisure habits. The results are presented with the sum of the points (minimum score of five and maximum of fifteen). www.nature.com/scientificreports/ instructed to fast for 12 to 14 h. Glucose, insulin resistance index by the Homeostasis Model assessment method (HOMA-IR), insulin sensitivity (QUICKI), in addition to the lipid profile: total cholesterol, triglycerides, high density lipoprotein (HDL-c) and low lipoprotein density (LDL-c) were collected. Subsequently, body composition assessment was performed with the reference standard, already validated for individuals with obesity 24 , the Dual Energy X-Ray Absortiometry (DXA) device (Discovery A, Hologi), and the variables used were bone mineral content (BMC), lean mass (LM) and fat mass (FM), in kilogram. This technique allows estimating body composition as a whole and by body segment.
Patients underwent body composition assessment in the morning, wearing light clothing, barefoot, and without any metal in contact with the body. Everyone was instructed not to perform intense physical exercises on the day before the exam; to fast for four hours before the assessment, and to eliminate urine before the evaluation 25 , as recommended by the manufacturer. The equipment's software automatically defined the body estimation areas, and the results were printed and tabulated in Excel® (Microsoft Excel, 2016). 3rd visit: incremental exercise testing. The maximum aerobic capacity assessed by ITMT was based on the Bruce protocol, which shows progressive increases in speed and inclination every three minutes 26 , on a treadmill (Super Inbramed ATL, Rio Grande do Sul, Brasil). The volunteers were encouraged to perform the test until exhaustion. The criteria for stopping/completing the test followed the recommendations of the American Thoracic Society 15 .
The test was performed by two physiotherapists and a physician, enabling potential cardiac arrhythmias detection and also for assistance in case of complications. Volunteers were continuously monitored by a 12-lead electrocardiogram (WinCardio, Microme, Brasília, Brazil) and a cardiofrequencimeter (Polar® S810i, Kempele, Oulu, Finland) fixed on the chest. For the measurement of subjective responses of dyspnea and fatigue in the lower limbs, the Borg scale from 0 to 10 was used 27 , integrating different information in physiological measures of physical performance and work ability. Systolic (SBP) and diastolic (DBP) blood pressure were measured using the auscultatory method. The patients performed two minutes of rest in a sitting position, followed by two minutes in a standing position; after the test, they performed three minutes of active recovery with a speed of 3 km/h and without inclination, followed by three minutes of passive recovery in a sitting position.
The test included gas analysis, and the recording of metabolic and ventilatory parameters was performed using the portable ergospirometry system Oxycon Mobile® (Mijnhardt/Jäger, Würzburg, German) with breathto-breath measures. All volunteers used a face mask as an interface to collect expired gases during exercise tests.
4th visit: six-minute walk test (6MWT) and two-minute step test (2MST). Both tests were performed according to the recommendations of the American Thoracic Society 28 . Patients were permitted to slow down if necessary and even stop the test for rest. The patients performed two minutes in the sitting position, followed by two minutes in the orthostatic position, and performed six minutes of recovery at the end of each test. Standardized verbal encouragement command was given in both tests. For the 6MWT, patients were instructed to walk as far as possible, without running on a flat surface of 30 m, for six minutes 28 . The 2MST was performed according to the protocol previously described 19 , using a single, portable step, 15 cm high, and without hand rest. The volunteers were instructed to go up and down the step as many times as possible for two minutes (free cadence), with manual recording by the evaluator. An evaluator was responsible for counting up-and-down step cycles (UDS).
As with ITMT, dyspnea perception, and fatigue in the legs as well as HR, SBP, DBP and recording of metabolic and ventilatory parameters were continuously monitored in both tests. In addition, subjective symptoms as dyspnea and fatigue in the lower limbs were also obtained by Borg CR-10.

Data analysis.
For the three exercise tests performed, the peak values were defined as the highest 15-s averaged values. The oxygen uptake (VȮ 2 ), carbon dioxide production (VĊO 2 ), minute ventilation (V̇E), breathing frequency (BF), and respiratory exchange ratio (RER) were obtained and exported to Excel® (Microsoft Excel, 2016).

Statistical analysis.
A posteriori power analysis was performed using the GPower statistical package (Version 3.1.3 -Franz Faul Universität Kiel, Germany). Considering our study total sample size of 44 individuals, an α error probability of 0.05, and an effect size of 0.50, the statistical power was calculated to be 98%. The data were analyzed using the statistical program SPSS Statistics, Version 20.0, USA. Data were expressed as mean and 95% of confidence interval. Data normality was tested by the Shapiro-Wilk.
The categorical variables were compared with the chi-square test. Between gender differences were evaluated by means of independent-samples t test or Mann-Whitney U test. One-way ANOVA with Tukey post hoc or Kruskall-Wallis were applied to evaluate the difference between tests (ITMT, 2MST and 6MWT). The Pearson's or Spearman's correlation coefficients were analyzed to investigate the associations between the main variables. The magnitude of the correlation was determined considering the classification of the values of r: 0.00 to 0.19 = none to slight, 0.20 to 0.39 = low, 0.40 to 0.69 = moderate, 0.70 to 0.89 = strong or high, and 0.90 to 1.00 = very high. Besides that, stepwise multiple linear regression analysis was performed to evaluate the best predictors of VȮ 2 (mL·kg −1 ·min −1 ) and distance walked on ITMT. The variables included in the model were: gender, body composition, metabolic and lipid profile, as well as performance data from the submaximal tests (UDS cycles and distance walked). The statistical significance level was set at p < 0.05.
We had 1 missing value for BMC, LM and FM, and 1 missing for HR on 2MST. These variables were regarded as missing data, not being replaced by an average value.  Figure 1 illustrates the recruitment flowchart for individuals who participated in the study. We initially recruited 268 individuals with obesity, and the final sample was composed by 44 patients. Table 1 shows patients general characteristics and body composition. Most of the sample was composed of women, as expected, and the most reported comorbidity was arterial hypertension. Considering an absolute total of 15 points, the results of the Baecke questionnaire revealed low physical activity patterns for the evaluated   Table 2 shows that ITMT required greater metabolic, ventilatory and cardiovascular demand, when compared with 6MWT and 2MST. However, only DBP was not significantly different between ITMT and 2MST. In addition, it is worth mentioning that 2MST required a higher VȮ 2, and HR, when compared to the 6MWT, with no significant difference in the other variables.
Moreover, correlations between 6MWT and 2MST were found: UDS cycles × distance walked (r = 0.56; p = < 0.001) and VȮ 2, mL·min −1 (r = 0.60; p = < 0.001). The responses of VȮ 2, mL.min -1 on ITMT, 6MWT and 2MST are shown in Fig. 2. Table 2. Responses in incremental treadmill test (ITMT), two-minute step test (2MST) and six-minute walk test (6MWT) at the peak of the tests. Data are reported as mean or median, according to data distribution, and CI (confidence interval). VȮ 2 : oxygen uptake, VĊO 2 : carbon dioxide production, RER: respiratory exchange rate, V E : minute ventilation, BF: breathing frequency, HR: heart rate, HRmax: maximal heart rate, SBP: systolic blood pressure, DBP: diastolic blood pressure. Intragroup differences (one-way ANOVA). a ITMT versus 6MWT. b ITMT versus 2MST. c 6MWT versus 2MST; p < 0.05. Leg fatigue (0-10 score Correlations of performance between tests and body composition are shown in Fig. 3.   Table 3. Regarding the distance walked, the model considered the FM, glucose and UDS cycles performed in the 2MST, which were able to explain 42% of the variance of the distance walked in the ITMT, as shown in Table 4. The gender was not significant variable to include in the predictive models.

Discussion
Main findings. This study presents some important results for this population: (1) ITMT elicited greater metabolic, ventilatory and cardiovascular demand when compared to the 6MWT and 2MST in patients with obesity; (2) even with the time difference between the 6MWT and the 2MST, and the activity performed between them (horizontal and vertical, respectively), the 2MST, imposed a greater chronotropic (HR) and metabolic (VȮ 2 ) demand, since these variables differed from each other, and that did not seem noticeable from a ventilatory, blood pressure, dyspnea and leg fatigue point of view; (3) body composition (LM and FM) is able to influence the functional capacity of individuals with obesity, observed by the responses in the three different tests presented, and (4) the estimation model based on FM and LDL can explain 30% of VȮ 2 variance at the peak of the ITMT, and the model based on the FM, glucose and UDS cycles, is able to explain 42% of the distance walked in the ITMT.

Comparison between cardiovascular and metabolic ventilatory variables between maximal and submaximal tests. The maximal incremental cardiopulmonary exercise may have limitations in its
performance because it needs an adequate space and a highly trained team (with a doctor), and it would not be feasible to be applied mediately after a surgical procedure, due to the high demand required. Moreover, individuals with obesity and comorbidities have a substantial limitation in activities due to their inability to perform high-intensity exercise 29 . The use of the exercise test with the assessment of cardiopulmonary variables is extremely important, since general health status better correlates with exercise tolerance than with measures at rest 15 . In addition, this evaluation provides us with valuable information, which optimizes the appropriate intensity finding for an intervention program, and, in this way, the systems' integrative responses can be evaluated in submaximal tests, assisting in clinical decision making 30 . Thus, activities where the level of effort is related to the effort required for day-to-day activities seem adequate to assess functional capacity. Da Costa and collaborators 31 compared the 6MWT and the six-minute step test (6MST) and observed that both are safe and produce submaximal efforts in healthy and sedentary individuals. Nonetheless, they concluded that 6MST requires greater oxygen demands due to different body movements when compared to 6MWT. Still, 6MST is not always a submaximal test when used in some populations, as it is known that the stress to perform activities that require vertical body displacement causes greater fatigue 32 . In this sense, some authors suggest that a shorter step test may be appropriate in individuals who have cardiopulmonary or musculoskeletal limitations, since the frequency of UDS cycles may be constant after the second or third minute [31][32][33] .
Pessoa and collaborators 34 demonstrate that the 2MST is a sensitive and reliable test when analyzing the functional capacity in individuals with severe chronic obstructive pulmonary disease, as it is a short-term test and was sufficient to cause metabolic, ventilatory, cardiovascular and of effort perception. In addition, the 2MST also allows a reliable and safe assessment in individuals with heart failure, as it demands a greater effort, which the 6MWT did not demonstrate. Besides that, the authors performed the same tests performed in the present study www.nature.com/scientificreports/ (ITMT, 2MST and 6MWT), and it is highlighted that VȮ 2 peak of 2MST correlated with ITMT, corroborating our result, suggesting that 2MST can be used for evaluation of the global integrated response to exercise, in a way correlated to the ITMT, for example 35 . Therefore, 2MST proves to be a viable option for several populations, and when compared to 6MWT in the current study, it shows that even with a shorter testing time and a different activity, it required a greater chronotropic (HR) and metabolic (VȮ 2 ) demand, with no difference in dyspnea and leg fatigue perception. Unlike other studies that carry out walking in place during 2MST (the individual needs to raise the knees, one at a time, to the height between the middle of the patella and the iliac crest as many times as possible) 35,36 , we performed the test on an ergometer in order to obtain the vertical and horizontal displacement of the patient, quite similarly to the 6MST.
In comparison to ITMT, 2MST differed from all variables, except for DBP. However, it presented a moderate correlation in the test performance, and in VȮ 2 . Although 6MWT is widely used and practical on a day-to-day basis, we could observe in this study that the demand is much lower, even with a test time longer than 2MST, in addition to needing a space (30 m corridor) for its realization.
Relationship between fat mass and functional capacity obtained in the maximal and submaximal tests. It is clear in the literature that obesity causes much damage to health 37,38 , and that these individuals have significant limitations when performing physical exercise, as this has been clearly investigated 6,39 and compared with eutrophic individuals 40,41 . Excess body mass affects motor function, causing individuals with obesity to have greater energy expenditure 42,43 . Physiologically, it is more difficult for the individual with obesity to do the same amount of work as a eutrophic person, since the excess fat does not contribute to the work performed, hindering performance 42 . Uranga et al. 11 observed that many of the risk factors that are related to obesity depend mainly on the distribution of fat, since the adipose tissue intensifies the state of inflammation 44,45 . In our study, it was possible to observe that FM negatively affected VȮ 2 peak in both maximum and submaximal tests.
Excessive infiltration of fat in the muscle tissue of the lower limbs may explain the impairment in performance and difficulty in daily physical functions 11,43,46 , and it has already been associated with metabolic disorders, such as insulin resistance, and changes in glucose metabolism 43 . In this sense, impairment of motor function in individuals with obesity seems to depend on mechanical and metabolic factors that significantly reduce global motor performance 11 .

Predictors of functional capacity obtained in the maximal and submaximal tests.
In the multiple regression analysis of VȮ 2 in ITMT, FM and LDL were considered. In addition to the implications of body fat for performance already mentioned in the previous paragraphs, individuals with obesity have high levels of LDL 47 and elevations of this lipoprotein have already been associated with a high risk of developing cardiovascular disease 48 . The guidelines of the Adult Treatment Panel III 49 recommend that the ideal LDL level be < 100 mg/ dL, and in our study, we observed that individuals have an average of 125.1 mg/dL. Although not evaluated in the present study, endothelial dysfunction, a consequence of the formation of metabolic products derived from lipids, hormones and pro-inflammatory cytokines, is present in individuals with obesity 50,51 . There are several related mechanisms attributed to the progression of endothelial dysfunction in individuals with obesity, and it includes increased levels of LDL and triglycerides, increased oxidative stress, elevated levels of inflammatory factors and unbalanced hemodynamic activities 51 , and it can be a factor that may explain the contribution of LDL in the predictive model.
In the analysis for the distance walked in the ITMT, however, it was also considered the FM, the performance in the 2MST (obtained through the UDS cycles), in addition to the glucose. The 2MST UDS cycles correlated with the distance walked in the ITMT, and the responses in the 2MST seem to be closer to the ITMT, instead of the 6MWT. Obesity is also associated with dyslipidemia, which can be explained by the expansion of visceral adipose tissue 44 . Insulin resistance, diabetes mellitus, and impaired glucose metabolism are commonly found in individuals with obesity 10,52 , and are associated with poor quality of skeletal muscles 43 . Additionally, it is worth mentioning that the individuals in our study showed altered values in the HOMA and QUICK tests (valid estimates for insulin resistance) 53 .
In view of all the aforementioned changes, there are several benefits of physical exercise for individuals with obesity 38,54 , for atherogenic reduction 55,56 , in the improvement of endothelial dysfunction 50,57 , with potent cardiovascular effects 51 and in reducing mortality 56 . However, the prescription of physical exercises often becomes a challenge for the individuals with obesity 58 . In this sense, effective methods for assessing functional capacity must be applied, such as for prognostic purposes before bariatric surgery 16 , and even after the surgical procedure, to start an intervention program.
In this sense, individuals with obesity, often with comorbidities, are candidates for bariatric surgery, and a test for the assessment of effective, simple and quick functional capacity can be beneficial in a preoperative moment, and right after the surgical procedure, since it can help to monitor possible functional declines and assist in rehabilitation strategies, as it has been carried out in the studies of our group 59,60 . Limitations of study. The present study has limitations that must be considered. According to our exclusion criteria, these findings cannot be extrapolated to individuals who have neurological, cardiac, respiratory disorders, etc. Unfortunately, we did not evaluate the endothelial function in this study, in addition to the infiltration of intramuscular fat, which should be performed through magnetic resonance imaging. Our sample consisted mostly of women, however, gender was not significant in the predictive models in the statistical analysis, so we can consider that this was not a confounding factor. Nonetheless, the population of obese patients under-