Clin Physiol Funct Imaging (2015)
doi: 10.1111/cpf.12275
Can geometric indices of heart rate variability predict improvement in autonomic modulation after resistance training in chronic obstructive pulmonary disease? Ana Alice Soares dos Santos, Ana Laura Ricci-Vitor, Vanessa Santa Rosa Bragatto, Ana Paula Soares dos Santos, Ercy Mara Cipulo Ramos and Luiz Carlos Marques Vanderlei Departamento de Fisioterapia, Faculdade de Ci^encias e Tecnologia da Universidade Estadual Paulista (UNESP), Presidente Prudente, Sa˜o Paulo, Brazil
Summary Correspondence Ana Alice Soares dos Santos, Departamento de Fisioterapia, Faculdade de Ci^encias e Tecnologia – Universidade Estadual Paulista (FCT/UNESP), Rua Roberto Simonsen, n° 305 – Centro Educacional, 19060–900 Presidente Prudente, Brazil E-mail: [email protected]
Accepted for publication Received 20 September 2014; accepted 1 June 2015
Key words autonomic nervous system; exercise; heart rate; muscle strength; nonlinear dynamics; respiratory tract diseases
Chronic obstructive pulmonary disease (COPD) is associated with autonomic dysfunctions that can be evaluated through heart rate variability (HRV). Resistance training promotes improvement in autonomic modulation; however, studies that evaluate this scenario using geometric indices, which include nonlinear evaluation, thus providing more accurate information for physiological interpretation of HRV, are unknown. This study aimed to investigate the influence of resistance training on autonomic modulation, using geometric indices of HRV, and peripheral muscle strength in individuals with COPD. Fourteen volunteers with COPD were submitted to resistance training consisting of 24 sessions lasting 60 min each, with a frequency of three times a week. The intensity was determined as 60% of one maximum repetition and was progressively increased until 80% for the upper and lower limbs. The HRV and dynamometry were performed at two moments, the beginning and the end of the experimental protocol. Significant increases were observed in the RRtri (481 160 versus 655 269, P = 0033), TINN (6536 3549 versus 10107 6334, P = 0028), SD1 (748 317 versus 1104 645, P = 0038) and SD2 (2230 856 versus 3292 1878, P = 0022) indices after the resistance training. Visual analysis of the Poincare plot demonstrated greater dispersion beat-to-beat and in the longterm interval between consecutive heart beats. Regarding muscle strength, there was a significant increase in the shoulder abduction and knee flexion. In conclusion, geometric indices of HRV can predict improvement in autonomic modulation after resistance training in individuals with COPD; improvement in peripheral muscle strength in patients with COPD was also observed.
Introduction Chronic obstructive pulmonary disease (COPD), besides the pulmonary complications, is associated with musculoskeletal dysfunctions (Rabe et al., 2007) and autonomic changes (Reis et al., 2010); these being important in controlling the body visceral functions (Carvalho et al., 2011a). The changes could be evaluated using a non-invasive tool denominated heart rate variability (HRV). Among the methods used for HRV analysis, the geometric indices stand out, as they allow presentation of the interval between consecutive heart beats (RR) in geometric patterns and the use of various approaches to calculate the measures of HRV (Rajendra Acharya et al., 2006). The Poincare plot is one example of an index which represents graphically the correlation between each RR and the following one. The
analysis can also be assessed qualitatively, through evaluation of the figure created by this attractor, or quantitatively, through the adjustment of the ellipse of the figure created by this attractor, from which is obtained the SD1, SD2 and SD1/ SD2 indices (Vanderlei et al., 2010). The Poincare plot is considered by some authors as a nonlinear index (Voss et al., 2007). Nonlinear dynamics have been gaining increasing interest as the mechanisms of cardiovascular system regulation interact in a nonlinear way (Vanderlei et al., 2009) and the nonlinear indices provide more precise information for physiological interpretation of HRV (Vanderlei et al., 2010). Due to the peripheral muscle and autonomic damage in COPD individuals, the insertion of these patients in pulmonary rehabilitation programmes is important (Rabe et al., 2007). There is evidence that aerobic exercise (Borghi-Silva et al., 2009), resistance exercise (Ricci-Vitor et al., 2013) and the
© 2015 Scandinavian Society of Clinical Physiology and Nuclear Medicine. Published by John Wiley & Sons Ltd
1
2 Geometric indices post-resistence training in COPD, A. A. S. Santos et al.
combination of both aerobic and resistance exercise (Camillo et al., 2011) promote an increase in autonomic modulation, although studies that evaluate the influence of resistance exercise on geometric indices of autonomic modulation are unknown. The hypothesis is that these indices, which include nonlinear evaluation, are able to predict autonomic modulation changes after resistance training. Therefore, this study aimed to investigate the influence of resistance training on autonomic modulation, through analysis of geometric HRV indices, and peripheral muscle strength in individuals with COPD.
Materials and methods Subjects A total of 14 volunteers with a medical and spirometric diagnosis of COPD (Rabe et al., 2007) were analysed, they reported an absence of severe cardiovascular diseases and comorbidities that could influence the performance of the experimental protocol or any parameter evaluated. In Fig. 1, the sample losses are displayed (Martins et al., 2009). The individuals who composed the final sample were informed about the objectives and procedures of the research and, after signing an informed consent form, participated voluntarily and effectively in the study. All procedures used in this study were accepted by the Research Ethics Committee of the Faculty of Sciences and Technology, UNESP, Presidente Prudente, S~ao Paulo, Brazil (Protocol No. 42/2010). Study design The experimental protocol consisted of an initial evaluation, an exercise programme and a final evaluation. The initial
evaluation was composed of anamnesis, to confirm compliance with the inclusion criteria of the investigation; anthropometric and spirometric evaluations; and autonomic and strength evaluations. The volunteers were then submitted to a one repetition maximum test (1RM) and the resistance exercise programme, consisting of 24 sessions, before the final evaluation which consisted of autonomic and strength evaluations. Anamnesis Information on the volunteer was identified through the medical history with questions regarding any severe cardiovascular, neuromuscular, skeletal or lung associated diseases, and any history of diseases and current clinical state (medications currently prescribed, and the presence of diabetes, arterial hypertension and smoking risk factors). Anthropometry was performed to characterize the population and to investigate the presence of excess weight and obesity (Lohman et al., 1988). Spirometry was performed to confirm the diagnosis of COPD and to classify the severity of the airflow (American Thoracic Society, 2005). Autonomic evaluation Heart rate variability evaluations in the initial and final period were performed to verify the autonomic modulation. The volunteers were assessed individually in a room at a temperature between 21°C and 23°C and humidity between 40% and 60% in the morning, between 8.00 and 12.00 to standardize the interference of circadian rhythm. The volunteers were instructed to avoid alcoholic and/or stimulant drinks such as coffee or tea, and to suspend their medication for a period of 12 h prior to the autonomic evaluation. During the autonomic
Figure 1 Diagram representing the flow of participants for the steps of the study. © 2015 Scandinavian Society of Clinical Physiology and Nuclear Medicine. Published by John Wiley & Sons Ltd
Geometric indices post-resistence training in COPD, A. A. S. Santos et al. 3
evaluation, the volunteers were instructed to remain alert, in silence, breathing spontaneously, resting in the supine position for 30 min on a couch. After receiving an explanation of the data collection procedures, an electrode was placed on the volunteer’s chest at the sternal angle using an elastic strap, and the heart rate receiver (Polar Electro, model S810i, Kempele, Finland) was attached to the patient’s wrist. The equipment had been previously validated (Vanderlei et al., 2008). To analyse the HRV indices, 256 intervals of consecutive cardiac beats were used, selected after digital filtering, complemented by manual filtering to eliminate artefact and ectopic beats, and only series with more than 95% sinus beats were included in the study (Godoy, 2005). The analysis of geometric indices was performed from the data provided by the frequency counter and processed using Kubios software (Biosignal Analysis and Medical Image Group, Department of Physics, Universidade de Kuopio, Finland) (Tarvainen et al., 2014). The indices used were triangular index, triangular interpolation of RR (TINN) and the Poincare plot which was analysed quantitatively (SD1, SD2 and the SD1/SD2 ratio indices) and qualitatively (analysis of the figure created by the attractor). The triangular index and TINN were calculated from the construction of a density histogram of RR, which demonstrates all possible values of RR on the horizontal axis and the frequency with which each occurred on the vertical axis. The union of the points in histogram columns created a figure similar to a triangle, from which the indices were extracted (Vanderlei et al., 2010). The triangular index consisted of the integral of the histogram (namely the total number of RR) divided by the maximum density distribution (modal frequency of RR), measured on a discrete scale with boxes of 78125 ms (1/128 s) (Task Force of the European Society of Cardiology et al., 1996). The TINN index consisted of the width of the base of the distribution measured as the base of the triangle, approaching the distribution of all RR, being the difference in least squares used to determine the triangle (Task Force of the European Society of Cardiology et al., 1996). The Poincare plot is one of the most important techniques to visually represent HRV, providing the capacity to evidence nonlinear aspects in a sequence of data. In this method, the duration of RRn is represented on the ‘x’ axis and the subsequent interval (RRn + 1) on the ‘y’ axis; therefore, each point on the graph (RRn, RRn + 1) corresponds to two successive beats (Manzano et al., 2011). For the quantitative analysis, the following indices were calculated: SD1 (standard deviation of instant variability beat to beat), SD2 (standard deviation of long-term RR) and the SD1/SD2 ratio. The qualitative analysis was performed from the figure created by the attractor. The figures were classified as (i) a normal plot in which a figure presents an increase in the dispersion of RR, (ii) a figure with less global dispersion beat to beat and without an increase in dispersion of RR in the long-term, characterized as a plot with less variability (Tulppo et al., 2005).
Strength evaluation The measurement of strength was performed unilaterally (using the dominant member) at the initial and final assessments, using a digital dynamometer (Force Gaugeâ, model FG-100 kg, Brazil) with results expressed in Newtons (N). The patients were instructed to execute maximum voluntary isometric contractions for 6-s, resisted by a non-stretch strap attached to the dynamometer. One end of the strap was attached to the equipment and the other one to the body segment executing the movement (Ricci-Vitor et al., 2013). The measurement was repeated three to five times with an interval of 1 min between repetitions, the values were allowed to differ up to 20%, and the mean of the higher values was registered. The positions of the individuals according to the movements evaluated are described below: 1 Knee flexion and extension: sitting position, with hip flexion and knees at 90°. The strap was attached using the ankle adapter, and the patient was instructed to perform knee flexion against the resistance; extensor and flexor sitting chairs were used for the lower limbs; 2 Shoulder flexion and abduction: standing position, with the shoulder at 70° and the elbow in the prone position. The strap was attached to a hand grip, and the patient was instructed to perform shoulder flexion against the resistance; 3 Elbow flexion: standing position, with the arm attached to the lateral region of the body. The strap was attached to a hand grip, and the patient was instructed to perform elbow flexion to 90° in the supine position, against the resistance; a simple pulley device was used for the upper limbs. The devices were adjusted according to each volunteer for the correct execution of the exercises (Ricci-Vitor et al., 2013).
One repetition maximum test The 1RM test was performed prior to the resistance training to determine the workload, using a pulley system (Ipiranga, Academia Hard, S~ao Paulo, Brazil). The initial load was stipulated as 20% of body mass for the lower limbs and 5% for the upper limbs, with progressive 5% increments, according to the perception of the subjects. The rest interval between each attempt was 5 min. The test was considered concluded when the volunteer attained the maximum load without mechanical failure. No more than five attempts were permitted for the establishment of the maximum load (Brown & Weir, 2001). The muscle groups tested to determine 1RM were the same as those tested by dynamometry: knee flexion, knee extension, shoulder flexion, shoulder abduction and elbow flexion (Ricci-Vitor et al., 2013). Resistance exercise programme The resistance training was performed using the same equipment as in the 1RM test and was conducted over 24 morning sessions, each session lasting 60 min, with a frequency of three times a week. The resistance training protocol structure
© 2015 Scandinavian Society of Clinical Physiology and Nuclear Medicine. Published by John Wiley & Sons Ltd
4 Geometric indices post-resistence training in COPD, A. A. S. Santos et al.
was as follows: (i) global stretching (muscles of body trunk, upper and lower limbs) in the initial and final session; (ii) lower limb strength training (knee flexion and extension on leg bend or leg extension); (iii) upper limb strength training (shoulder flexion and extension and elbow flexion on simple pulley equipment). The initial training intensity was 60% of 1RM and was progressively increased every five sessions up to 80%. Three sets of ten repetitions were performed for each of the training muscle groups with 2–3 min between the sets (Silva & Dourado, 2008). Statistical analysis Data distribution was checked using the Shapiro–Wilk test and, depending on the data distribution, the paired t-test or Wilcoxon test was used. The level of statistical significance was set at P
doi: 10.1111/cpf.12275
Can geometric indices of heart rate variability predict improvement in autonomic modulation after resistance training in chronic obstructive pulmonary disease? Ana Alice Soares dos Santos, Ana Laura Ricci-Vitor, Vanessa Santa Rosa Bragatto, Ana Paula Soares dos Santos, Ercy Mara Cipulo Ramos and Luiz Carlos Marques Vanderlei Departamento de Fisioterapia, Faculdade de Ci^encias e Tecnologia da Universidade Estadual Paulista (UNESP), Presidente Prudente, Sa˜o Paulo, Brazil
Summary Correspondence Ana Alice Soares dos Santos, Departamento de Fisioterapia, Faculdade de Ci^encias e Tecnologia – Universidade Estadual Paulista (FCT/UNESP), Rua Roberto Simonsen, n° 305 – Centro Educacional, 19060–900 Presidente Prudente, Brazil E-mail: [email protected]
Accepted for publication Received 20 September 2014; accepted 1 June 2015
Key words autonomic nervous system; exercise; heart rate; muscle strength; nonlinear dynamics; respiratory tract diseases
Chronic obstructive pulmonary disease (COPD) is associated with autonomic dysfunctions that can be evaluated through heart rate variability (HRV). Resistance training promotes improvement in autonomic modulation; however, studies that evaluate this scenario using geometric indices, which include nonlinear evaluation, thus providing more accurate information for physiological interpretation of HRV, are unknown. This study aimed to investigate the influence of resistance training on autonomic modulation, using geometric indices of HRV, and peripheral muscle strength in individuals with COPD. Fourteen volunteers with COPD were submitted to resistance training consisting of 24 sessions lasting 60 min each, with a frequency of three times a week. The intensity was determined as 60% of one maximum repetition and was progressively increased until 80% for the upper and lower limbs. The HRV and dynamometry were performed at two moments, the beginning and the end of the experimental protocol. Significant increases were observed in the RRtri (481 160 versus 655 269, P = 0033), TINN (6536 3549 versus 10107 6334, P = 0028), SD1 (748 317 versus 1104 645, P = 0038) and SD2 (2230 856 versus 3292 1878, P = 0022) indices after the resistance training. Visual analysis of the Poincare plot demonstrated greater dispersion beat-to-beat and in the longterm interval between consecutive heart beats. Regarding muscle strength, there was a significant increase in the shoulder abduction and knee flexion. In conclusion, geometric indices of HRV can predict improvement in autonomic modulation after resistance training in individuals with COPD; improvement in peripheral muscle strength in patients with COPD was also observed.
Introduction Chronic obstructive pulmonary disease (COPD), besides the pulmonary complications, is associated with musculoskeletal dysfunctions (Rabe et al., 2007) and autonomic changes (Reis et al., 2010); these being important in controlling the body visceral functions (Carvalho et al., 2011a). The changes could be evaluated using a non-invasive tool denominated heart rate variability (HRV). Among the methods used for HRV analysis, the geometric indices stand out, as they allow presentation of the interval between consecutive heart beats (RR) in geometric patterns and the use of various approaches to calculate the measures of HRV (Rajendra Acharya et al., 2006). The Poincare plot is one example of an index which represents graphically the correlation between each RR and the following one. The
analysis can also be assessed qualitatively, through evaluation of the figure created by this attractor, or quantitatively, through the adjustment of the ellipse of the figure created by this attractor, from which is obtained the SD1, SD2 and SD1/ SD2 indices (Vanderlei et al., 2010). The Poincare plot is considered by some authors as a nonlinear index (Voss et al., 2007). Nonlinear dynamics have been gaining increasing interest as the mechanisms of cardiovascular system regulation interact in a nonlinear way (Vanderlei et al., 2009) and the nonlinear indices provide more precise information for physiological interpretation of HRV (Vanderlei et al., 2010). Due to the peripheral muscle and autonomic damage in COPD individuals, the insertion of these patients in pulmonary rehabilitation programmes is important (Rabe et al., 2007). There is evidence that aerobic exercise (Borghi-Silva et al., 2009), resistance exercise (Ricci-Vitor et al., 2013) and the
© 2015 Scandinavian Society of Clinical Physiology and Nuclear Medicine. Published by John Wiley & Sons Ltd
1
2 Geometric indices post-resistence training in COPD, A. A. S. Santos et al.
combination of both aerobic and resistance exercise (Camillo et al., 2011) promote an increase in autonomic modulation, although studies that evaluate the influence of resistance exercise on geometric indices of autonomic modulation are unknown. The hypothesis is that these indices, which include nonlinear evaluation, are able to predict autonomic modulation changes after resistance training. Therefore, this study aimed to investigate the influence of resistance training on autonomic modulation, through analysis of geometric HRV indices, and peripheral muscle strength in individuals with COPD.
Materials and methods Subjects A total of 14 volunteers with a medical and spirometric diagnosis of COPD (Rabe et al., 2007) were analysed, they reported an absence of severe cardiovascular diseases and comorbidities that could influence the performance of the experimental protocol or any parameter evaluated. In Fig. 1, the sample losses are displayed (Martins et al., 2009). The individuals who composed the final sample were informed about the objectives and procedures of the research and, after signing an informed consent form, participated voluntarily and effectively in the study. All procedures used in this study were accepted by the Research Ethics Committee of the Faculty of Sciences and Technology, UNESP, Presidente Prudente, S~ao Paulo, Brazil (Protocol No. 42/2010). Study design The experimental protocol consisted of an initial evaluation, an exercise programme and a final evaluation. The initial
evaluation was composed of anamnesis, to confirm compliance with the inclusion criteria of the investigation; anthropometric and spirometric evaluations; and autonomic and strength evaluations. The volunteers were then submitted to a one repetition maximum test (1RM) and the resistance exercise programme, consisting of 24 sessions, before the final evaluation which consisted of autonomic and strength evaluations. Anamnesis Information on the volunteer was identified through the medical history with questions regarding any severe cardiovascular, neuromuscular, skeletal or lung associated diseases, and any history of diseases and current clinical state (medications currently prescribed, and the presence of diabetes, arterial hypertension and smoking risk factors). Anthropometry was performed to characterize the population and to investigate the presence of excess weight and obesity (Lohman et al., 1988). Spirometry was performed to confirm the diagnosis of COPD and to classify the severity of the airflow (American Thoracic Society, 2005). Autonomic evaluation Heart rate variability evaluations in the initial and final period were performed to verify the autonomic modulation. The volunteers were assessed individually in a room at a temperature between 21°C and 23°C and humidity between 40% and 60% in the morning, between 8.00 and 12.00 to standardize the interference of circadian rhythm. The volunteers were instructed to avoid alcoholic and/or stimulant drinks such as coffee or tea, and to suspend their medication for a period of 12 h prior to the autonomic evaluation. During the autonomic
Figure 1 Diagram representing the flow of participants for the steps of the study. © 2015 Scandinavian Society of Clinical Physiology and Nuclear Medicine. Published by John Wiley & Sons Ltd
Geometric indices post-resistence training in COPD, A. A. S. Santos et al. 3
evaluation, the volunteers were instructed to remain alert, in silence, breathing spontaneously, resting in the supine position for 30 min on a couch. After receiving an explanation of the data collection procedures, an electrode was placed on the volunteer’s chest at the sternal angle using an elastic strap, and the heart rate receiver (Polar Electro, model S810i, Kempele, Finland) was attached to the patient’s wrist. The equipment had been previously validated (Vanderlei et al., 2008). To analyse the HRV indices, 256 intervals of consecutive cardiac beats were used, selected after digital filtering, complemented by manual filtering to eliminate artefact and ectopic beats, and only series with more than 95% sinus beats were included in the study (Godoy, 2005). The analysis of geometric indices was performed from the data provided by the frequency counter and processed using Kubios software (Biosignal Analysis and Medical Image Group, Department of Physics, Universidade de Kuopio, Finland) (Tarvainen et al., 2014). The indices used were triangular index, triangular interpolation of RR (TINN) and the Poincare plot which was analysed quantitatively (SD1, SD2 and the SD1/SD2 ratio indices) and qualitatively (analysis of the figure created by the attractor). The triangular index and TINN were calculated from the construction of a density histogram of RR, which demonstrates all possible values of RR on the horizontal axis and the frequency with which each occurred on the vertical axis. The union of the points in histogram columns created a figure similar to a triangle, from which the indices were extracted (Vanderlei et al., 2010). The triangular index consisted of the integral of the histogram (namely the total number of RR) divided by the maximum density distribution (modal frequency of RR), measured on a discrete scale with boxes of 78125 ms (1/128 s) (Task Force of the European Society of Cardiology et al., 1996). The TINN index consisted of the width of the base of the distribution measured as the base of the triangle, approaching the distribution of all RR, being the difference in least squares used to determine the triangle (Task Force of the European Society of Cardiology et al., 1996). The Poincare plot is one of the most important techniques to visually represent HRV, providing the capacity to evidence nonlinear aspects in a sequence of data. In this method, the duration of RRn is represented on the ‘x’ axis and the subsequent interval (RRn + 1) on the ‘y’ axis; therefore, each point on the graph (RRn, RRn + 1) corresponds to two successive beats (Manzano et al., 2011). For the quantitative analysis, the following indices were calculated: SD1 (standard deviation of instant variability beat to beat), SD2 (standard deviation of long-term RR) and the SD1/SD2 ratio. The qualitative analysis was performed from the figure created by the attractor. The figures were classified as (i) a normal plot in which a figure presents an increase in the dispersion of RR, (ii) a figure with less global dispersion beat to beat and without an increase in dispersion of RR in the long-term, characterized as a plot with less variability (Tulppo et al., 2005).
Strength evaluation The measurement of strength was performed unilaterally (using the dominant member) at the initial and final assessments, using a digital dynamometer (Force Gaugeâ, model FG-100 kg, Brazil) with results expressed in Newtons (N). The patients were instructed to execute maximum voluntary isometric contractions for 6-s, resisted by a non-stretch strap attached to the dynamometer. One end of the strap was attached to the equipment and the other one to the body segment executing the movement (Ricci-Vitor et al., 2013). The measurement was repeated three to five times with an interval of 1 min between repetitions, the values were allowed to differ up to 20%, and the mean of the higher values was registered. The positions of the individuals according to the movements evaluated are described below: 1 Knee flexion and extension: sitting position, with hip flexion and knees at 90°. The strap was attached using the ankle adapter, and the patient was instructed to perform knee flexion against the resistance; extensor and flexor sitting chairs were used for the lower limbs; 2 Shoulder flexion and abduction: standing position, with the shoulder at 70° and the elbow in the prone position. The strap was attached to a hand grip, and the patient was instructed to perform shoulder flexion against the resistance; 3 Elbow flexion: standing position, with the arm attached to the lateral region of the body. The strap was attached to a hand grip, and the patient was instructed to perform elbow flexion to 90° in the supine position, against the resistance; a simple pulley device was used for the upper limbs. The devices were adjusted according to each volunteer for the correct execution of the exercises (Ricci-Vitor et al., 2013).
One repetition maximum test The 1RM test was performed prior to the resistance training to determine the workload, using a pulley system (Ipiranga, Academia Hard, S~ao Paulo, Brazil). The initial load was stipulated as 20% of body mass for the lower limbs and 5% for the upper limbs, with progressive 5% increments, according to the perception of the subjects. The rest interval between each attempt was 5 min. The test was considered concluded when the volunteer attained the maximum load without mechanical failure. No more than five attempts were permitted for the establishment of the maximum load (Brown & Weir, 2001). The muscle groups tested to determine 1RM were the same as those tested by dynamometry: knee flexion, knee extension, shoulder flexion, shoulder abduction and elbow flexion (Ricci-Vitor et al., 2013). Resistance exercise programme The resistance training was performed using the same equipment as in the 1RM test and was conducted over 24 morning sessions, each session lasting 60 min, with a frequency of three times a week. The resistance training protocol structure
© 2015 Scandinavian Society of Clinical Physiology and Nuclear Medicine. Published by John Wiley & Sons Ltd
4 Geometric indices post-resistence training in COPD, A. A. S. Santos et al.
was as follows: (i) global stretching (muscles of body trunk, upper and lower limbs) in the initial and final session; (ii) lower limb strength training (knee flexion and extension on leg bend or leg extension); (iii) upper limb strength training (shoulder flexion and extension and elbow flexion on simple pulley equipment). The initial training intensity was 60% of 1RM and was progressively increased every five sessions up to 80%. Three sets of ten repetitions were performed for each of the training muscle groups with 2–3 min between the sets (Silva & Dourado, 2008). Statistical analysis Data distribution was checked using the Shapiro–Wilk test and, depending on the data distribution, the paired t-test or Wilcoxon test was used. The level of statistical significance was set at P
![[Body composition and heart rate variability in patients with chronic obstructive pulmonary disease pulmonary rehabilitation candidates].](/assets/img/hold.png)
