RESISTANCE TRAINING AS A PRECONDITIONING STRATEGY FOR ENHANCING AEROBIC EXERCISE TRAINING OUTCOMES IN COPD

Abstract

Purpose

Aerobic exercise training is a recognized approach for improving functional capacity in COPD. People with greater disease severity often have difficulty achieving higher aerobic exercise training intensity. The effects of resistance training prior to aerobic training were examined to determine if this sequential approach was associated with greater gains in functional status than aerobic training alone or concurrent aerobic and resistance training.

Methods

Patients were randomized to: 1) sequential resistance then aerobic training (RT-then-AT) (8 weeks resistance training followed by 8 weeks aerobic exercise training), 2) control group (CE-then-AT+RT) (8 weeks of ‘sham’ training followed by 8 weeks concurrent aerobic and resistance training), 3) control group (CE-then-AT) (8 weeks ‘sham’ training followed by 8 weeks aerobic training). Outcomes were assessed at study entry, after week 8, and after week 16: aerobic exercise performance; muscle strength and endurance.

Results

75 patients completed training: FEV₁ %pred 40±10, V̇O_2peak %predicted, 71±22, fat-free mass index 19.5±3.1. RT-then-AT had greater acquisition of peripheral muscle endurance than CE-then-AT+RT and CE-then-AT, but improvements in aerobic exercise performance were similar. Improvements in muscle strength were similar between RT-then-AT and CE-then-AT+RT. Sarcopenia was associated with poorer attendance, and lower aerobic and resistance training volumes.

Conclusion

Although the sequential approach to resistance and aerobic training yielded a greater increase in muscle endurance and higher resistance training volume compared to concurrent resistance and aerobic training, other training outcomes were similar between the two groups, thus the sequential approach is not clearly superior to the concurrent approach in severe COPD. ClinicalTrials.gov Identifier NCT01058213.

Introduction

People with COPD experience peripheral muscle dysfunction of the lower body which impairs functional capacity for aerobic exercise and contributes to premature disability.¹ Pulmonary rehabilitation, in particular aerobic exercise training, is a recognized approach for improving functional capacity and health-related quality of life, and lessening distressing symptoms.² Despite the well documented efficacy of aerobic exercise training, controversy exists regarding the optimal approach for applying this therapy. Evidence suggests that higher training intensities are needed to produce the physiologic training effects (eg, increased aerobic capacity, reduced ventilatory requirement for exercise).^{3, 4} High intensity training is feasible for people with mild to moderate COPD; but, severely debilitated COPD patients may be unable to participate in aerobic training at a level sufficient to reap its maximum benefits.

People with greater disease severity have a very limited ventilatory capacity for exercise, are extremely deconditioned, and experience substantial muscle dysfunction of the lower extremities.⁵ For severely detrained individuals, initiating pulmonary rehabilitation with resistance training prior to aerobic training may be beneficial. Quadriceps strength correlates with peak aerobic capacity,¹ and resistance training (either alone or in combination with aerobic training) has beneficial effects on exercise capacity.⁶ There are no reports on the effects of a sequential approach to training, performing resistance training first in order to build individuals’ strength prior to initiating aerobic training. The use of a sequential training approach could enable people with severe COPD to train at a higher aerobic intensity with less perceived fatigue, in effect, the resistance training would be used to ‘precondition’ debilitated people for aerobic training. Some pulmonary rehabilitation programs do incorporate resistance training, but it is applied concurrently with aerobic training rather than as a preconditioning strategy.

The aims of this study were: 1) To examine whether resistance training prior to aerobic training is associated with greater gains in functional capacity for aerobic exercise (assessed by the symptom-limited incremental exercise test) and for common daily activities (assessed with tests of peripheral muscular strength and endurance, timed walking and chair rise tests), than aerobic training alone or concurrent aerobic training plus resistance training; 2) To examine whether resistance training prior to aerobic training is associated with greater improvements in breathlessness and fatigue during exercise and daily life, than aerobic training alone or concurrent aerobic training plus resistance training; and 3) To examine whether resistance training prior to aerobic training is associated with greater gains in functional status (assessed by functional performance and physical activity questionnaires) than aerobic training alone or concurrent aerobic training plus resistance training.

Methods

Participants

People with COPD who reported dyspnea on exertion, had greater disease severity, and had no other health problems that would interfere with exercise training were recruited from the Chicago area through advertisements and the Veterans Administration. The eligibility criteria included: forced expiratory volume in one second (FEV₁)/forced vital capacity<0.7 and FEV₁≤55% predicted, age≥45 years, and currently in stable clinical condition (eg, no exacerbations within two months of enrollment or recent change in medical therapy). Screening procedures included: pulmonary function tests, medical history and physical examination, chest x-ray, resting electrocardiogram, blood chemistries, hematology and urinalysis. Patients who met eligibility criteria based on the above tests underwent a symptom-limited incremental cycle ergometer test on a separate day. During the screening period, patients performed practice runs of each test of physical performance (exercise test, tests of muscle strength and endurance, timed walking test) in order to account for potential learning effects. The research was approved by Institutional Review Boards at the University of Illinois at Chicago and the Hines Veterans Administration Hospital, and all patients gave written informed consent.

Study Design

This was a prospective randomized trial with one experimental and two control groups that performed 16 weeks of supervised training in the laboratory. The experimental group received eight weeks of resistance training focused on the lower body followed by 8 weeks of aerobic training on a stationary cycle ergometer (RT-then-AT group). One control group received 8 weeks of sham training (gentle chair exercise) followed by eight weeks of the combination of aerobic training on a stationary cycle ergometer and resistance training focused on the lower body (CE-then-AT+RT group). The other control group received 8 weeks of sham training (gentle chair exercise) followed by eight weeks of aerobic training on a stationary cycle ergometer alone (CE-then-AT group). Randomization to group was stratified by gender (strata: male, female) and disease severity (strata: FEV₁ 30%–55% predicted, FEV₁<30% predicted) with a software program (biased coin algorithm to ensure equivalent groups).⁷ Data collectors were blinded to group assignment and patients were not informed of the intent of the three group research design or the expected outcomes of the study. Measures of dependent variables were taken at baseline, after 8 weeks of supervised training, and after 16 weeks of supervised training.

Interventions

Patients trained in small groups three times per week at either the University of Illinois at Chicago or at the Hines Veterans Administration Hospital. Exercise training was supervised by exercise specialists.

Aerobic training was performed on a stationary cycle ergometer, calibrated with a 4 kg weight (Monark 828E, Varberg, Sweden) using an interval training protocol. For the interval training protocol patients performed four work sets of five minutes duration separated by rest intervals of unloaded cycling lasting 2–4 minutes. This approach lessens symptoms of dyspnea and fatigue during training⁸ and enables even extremely dyspneic patients to train at progressively higher intensities without stopping or reducing training intensity. The initial work sets were at 50% of the peak work rate and were evaluated weekly with progressive increases targeted to achieve the highest work rate tolerated.⁹ The typical progression was: 50% peak work rate for weeks 1–2, 60% peak work rate for weeks 3–4, 70% peak work rate for weeks 5–6, and 80% peak work rate for weeks 7–8.

Resistance training was performed with fitness equipment (Body-Solid Inc., Forest Park, IL, United States of America) using 6 lifts: leg press, knee extension, knee flexion, calf raise, hip adduction, and hip abduction. Training was initiated at an intensity of 70% of the one repetition maximum (1RM) performed at baseline with a training volume of 2 sets of 8–10 repetitions for 2 weeks, followed by 2 weeks of training at 80% of the baseline 1RM at a volume of 2 sets. For the remaining 4 weeks the intensity was 80% of the 1RM (re-assessed after 4 weeks of training) at a volume of 3 sets of 8–10 repetitions.

Gentle chair exercises were a form of ‘sham’ training and these seated exercises incorporated stretching of all major joints; this intervention was based on the video “Armchair Fitness: Gentle Exercise”.¹⁰ The chair exercises were conducted at a slow pace to minimize any aerobic stimulus and with only gravity as resistance (e.g., no elastic bands or hand weights).

Outcome measures

Exercise test

Patients performed symptom-limited incremental exercise tests on an electrically braked cycle ergometer (one Lode CORIVAL-V2 bicycle ergometer, Groningen, The Netherlands). Patients breathed 30% oxygen during the test and exhaled gases were collected on a breath-by-breath basis with a metabolic cart (Vmax Encore 29, Viasys Healthcare Inc., Yorba Linda, CA, United States of America). Patients warmed up by cycling for 1 minute without resistance, thereafter the work rate increased by 10 W every one minute until the symptom-limited end-point was achieved. Ratings of breathlessness and leg fatigue were obtained at the end of each minute of exercise and at peak (Borg Category Ratio Scale).¹¹

Muscle strength and endurance

Lower-body strength was assessed using the one repetition maximum (1RM) with the same 6 lifts employed for resistance training. The 1RM represents the highest weight that can be lifted one time over the full range of motion using proper lifting technique. Lower-body strength was reported as the sum of the 1RM for all 6 lifts. The leg extension endurance time was used to measure muscular endurance of the quadriceps, a muscle essential for mobility (e.g. walking, rising from a chair). For the leg extension endurance time (LE-ET), patients performed repeated leg extensions at 60% of the baseline 1RM at a cadence of 12 repetitions per minute until task failure (defined as inability to complete the full range of motion for the lift and/or inability to maintain the required cadence).¹² The 30-second chair rise was performed using an armless straight back chair placed against the wall (to avoid shifting during the test). Patients were seated with their arms folded across the chest, their back firmly against the seat rest, and their feet flat on the floor. Patients were instructed to rise to a full stand then sit back down and to complete this maneuver as many times as possible within 30 seconds.¹³

Timed walking test

The 6-minute distance walk test (6MD) was conducted with patients walking back and forth along a measured corridor.¹⁴ Patients were instructed to establish a speed that would allow them to cover as much distance as possible within 6 minutes. Data collectors provided the verbal reassurance “you are doing well” every minute and let the patient know how much time was remaining,¹⁵ but did not formally coach patients or encourage them to alter their speed.¹⁶

Symptoms and functional status

Symptoms of dyspnea and fatigue were assessed with the Chronic Respiratory Disease Questionnaire (CRQ).¹⁷ The CRQ was administered during screening to familiarize patients with the dyspnea scale and then administered again at the baseline assessment. At subsequent administrations patients were not reminded of their previous responses. The Functional Performance Inventory assesses patients’ perceived difficulty in performing a variety of day-to-day activities. The instrument contains 67 items organized into eight subscales (body care, maintaining the household, physical exercise, recreation, spiritual activities, social interaction-family and friends, and work or school).¹⁸ The instrument was scored as the mean of the items (possible range 0–3) for the total instrument.

Physical activity questionnaire

The CHAMPS Activities Questionnaire for Older Adults (CHAMPS) is a self report instrument of physical activity over the last 4 weeks.¹⁹ The questionnaire assesses weekly duration of physical activities relevant to older adults and was scored as self-reported duration of physical activity per week for the total physical activities (both light and moderate).

Tests to characterize the sample

Disease severity

Comprehensive pulmonary function testing (lung volume measurements, spirometry, and diffusion capacity) was performed (VMAX Encore 22, Viasys Healthcare, Inc., Yorba Linda, CA, United States of America) according to established standards.^20–22 The COPD severity score developed by Eisner et al²³ was used as an additional assessment of COPD severity (possible range 0–36). The Functional Comorbidity Index (FCI)²⁴ was used to assess the effects of comorbid conditions on physical function (possible range 0–18).

Body composition

Whole body dual-energy x-ray absorptiometry (DXA) scans (Discovery Wi, Hologic Inc, Bedford, MA, United States of America) were performed to estimate total and regional body composition (arms, legs, trunk)^{25, 26} using a 3-compartment model: lean (non-osseous) tissue, fat mass, and bone mineral content. Primary body composition indexes were body mass index (BMI), fat-free mass index (FFMI), and body fat mass index (BFMI).

Historical physical activity

The Swedish Mammography Cohort Physical Activity Questionnaire (SMC-PAQ) is a brief self-administered questionnaire designed to assess historical (a lifetime of physical activity) and current (over the last year) leisure and occupational physical activity.^{27, 28} Incorporated are 5 scales: Work/Occupation, Walking/Bicycling, Home/Housework, Leisure time, and Exercise. Scores were calculated as MET·h·day⁻¹ for each scale and for total physical activity at four different time points of life (ages: 15, 30, 50 years, and the past year).

Exercise adherence

Adherence to exercise training can have major effects on the fidelity of the intervention. Adherence was evaluated by examining attrition (percent of patients who withdrew from training for each group), attendance (percent of sessions attended by each patient by group), and the training volume and progression of training intensity achieved. Aerobic training volume was defined as the percent of the total minutes of exercise at the prescribed work rate completed during training divided by the total possible minutes, and resistance training volume was defined as the percentage of total repetitions completed during training divided by the possible repetitions.

Data analysis

Intermittent missing data points were replaced using the last observation carried forward. Descriptive statistics are reported as mean ± SD unless specified otherwise (SPSS version 20.0 for Windows, Chicago, IL, United States of America). Data were checked for normality and where needed data transformation was performed to achieve normality. The following variables required data transformation: LE-ET (natural log), chair rise (square root), CHAMPS total physical activity (square root). In this sample of people with severe to very severe disease 70% of the sample reported ≤ 2 hours/week of moderate activity on the CHAMPS. The moderate physical activity data was severely skewed and unable to be transformed to achieve a normal distribution. However a square root transformation achieved normality for the total physical activity data, thus the total rather than moderate activity was used to report physical activity.

Baseline characteristics of the three groups were compared using multivariate analysis of variance. Bivariate relationships were examined with Pearson correlations. Treatment outcomes were examined using multivariate analysis of variance for repeated measures. Statistical significance was set at P<0.05. Where multivariate analysis of variance main effects were significant, ANOVA statistics for repeated measures were examined to determine differences between groups for individual variables (SPSS GLM LMatrix).

The peak exercise gas exchange and ventilation data represent the highest 20-second average of the breath-by-breath data collected at the symptom-limited end-point of the test. Additionally ‘iso-time’ data were calculated for the highest work rate performed for one minute (stage of exercise) that was completed at all three tests (baseline, after 8 weeks of training, after 16 weeks of training).

The percent of the predicted norms were calculated for the following: spirometry,²⁹ lung volumes,³⁰ diffusion capacity,³¹ and oxygen consumption.³²

Results

Participants

Three hundred fifty four people were screened for eligibility, 113 were enrolled, 99 initiated and 75 completed 16 weeks of laboratory-based training (n=20 RT-then-AT, n=28 CE-then-AT+RT and n=27 CE-then-AT). There were no significant differences in sample characteristics between the three groups (Table 1). Attrition is summarized in the Consort diagram (Figure 1); attrition rates were similar between groups (P=0.116). Additionally, there were no significant differences in demographic characteristics between those who completed the study compared to those who did not (P=0.263).

Table 1.

Descriptive characteristics of the sample by group. ^*

	RT-then-AT (n=20)	CE-then-AT+RT (n=28)	CE-then-AT (n=27)
Gender ratio, male/female	18/2	24/4	25/2
Age, years	68 ± 6	68 ± 8	68 ± 7
BMI, kg/m2	28.1 ± 7.1	29.2 ± 6.7	28.4 ± 6.2
FFMI, kg/m2	19.2 ± 3.5	19.8 ± 3.4	19.3 ± 2.5
FEV1, % predicted	42 ± 10	41 ± 10	39 ± 9
RV/TLC, ratio	0.55 ± 0.10	0.57 ± 0.09	0.56 ± 0.12
DLco, % predicted	59 ± 19	70 ± 27	66 ± 24
Smoking history, pack·years	61 ± 34	57 ± 32	40 ± 29
COPD Severity Scale	12 ± 7	12 ± 6	11 ± 4
Functional Comorbidity Index	4 ± 2	4 ± 2	4 ± 2
Historical PA last year, MET·h·day−1	36 ± 4	35 ± 3	36 ± 5

Definition of abbreviations: AT, aerobic training; BMI, body mass index; CE, chair exercise; FFMI, fat-free mass index; PA, physical activity; RT, resistance training.

^*Data are presented as mean ± SD, with the exception of the gender ratio.

MANOVA between-subjects effects: F(2,72)=1.017, P=0.448.

Figure 1.

Consort Diagram.

Training progression and adherence

Progression of training intensity was similar among the groups for both aerobic (P=0.628) and resistance training (P=0.079) (Figures 2A and 2B). Aerobic training volume (percent of the total minutes of exercise at the prescribed work rate completed during training divided by the total possible minutes) was similar (P=0.947) between the three groups (range of group means: 92–93%). Lower aerobic training volume was associated with worse airflow obstruction, greater hyperinflation, and lower body mass indexes: FEV₁% predicted (r=0.242, P=0.037), residual volume/total lung capacity (RV/TLC) ratio (r=−0.303, P=0.008), FFMI (r=0.356, P=0.002), and BMI (r=0.303, P=0.008). Resistance training volume (percentage of total repetitions completed during training divided by the possible repetitions) was higher (P=0.013) in the RT-then-AT group (88%) compared to the CE-then-AT+RT group (80%). Lower resistance training volume was associated with greater hyperinflation and lower lean mass index: RV/TLC ratio (r=−0.342, P=0.017), FFMI (r=0.310, P=0.032).

Figure 2.

Aerobic (panel A) and Resistance (panel B) Training Progression by Group. Values are mean ± SEM. Definition of abbreviations: AT = aerobic training, CE = chair exercise, RT = resistance training.

There were no differences in either the attrition during training (P=0.116) or attendance rates (P=0.669) between the three groups. Although attrition was not statistically different between the groups, the RT-then-AT had more people (n=6) who withdrew due to lack of interest (Figure 1). A review of the training notes revealed that all six tolerated the resistance training well and only one of those who stopped attending expressed that he didn’t like the resistance training. Although randomization is intended to equalize the characteristics of groups, it does not assure equivalence on all characteristics, such as motivation to persevere with training. Overall attendance was 95% for those that finished the study. Attendance did not correlate with airflow obstruction (FEV₁ % predicted) or diffusion limitation (DLco % predicted), but did correlate with hyperinflation (RV/TLC ratio, r=−0.299, P=0.009). Additionally, lower attendance was related to lower values of body mass indexes: r=0.294, P=0.011 for FFMI, r=0.362, P=0.001 for BFMI, r=0.367, P=0.001 for BMI. The reasons for missed training sessions were: health related (45%), vacation/other family obligations (22%), work/volunteer activities (6%), musculoskeletal complaints (1%), no reason given (12%), miscellaneous other reasons (14%).

Training outcomes

Table 2 summarizes the aerobic training outcomes. There were no differences between the groups at baseline (P=0.311) and no interaction of time-by-group assignment (P=0.372), but there were significant changes with time (P<0.001) in some of the peak and iso-time variables. Specifically all three groups demonstrated improvements in peak work rate (P<0.001), V̇O_2peak (P=0.002), V̇E_peak (P=0.045), and RPLF_peak (P<0.001). At iso-time all three groups demonstrated a reduction in the ventilatory requirement for exercise (P=0.009), RPB (P<0.001), and RPLF (P<0.001). For the RT-then-AT group changes in RPLF at peak and isotime occurred between baseline and week 8 with no additional improvement between weeks 8 and 16. For the CE-then-AT+RT and CE-then-AT groups, RPLF at peak and isotime did not change between baseline and week 8 but did improve between weeks 8 and 16 (Figure 3). Changes in aerobic capacity were associated with symptom reduction at iso-time. Specifically, changes in peak work rate were associated with the reduction in iso-time V̇E (r=−0.413, P<0.001), RPB (r=−0.436, P<0.001), and RPLF (r=−0.435, P<0.001) and changes in V̇O_2peak were associated with the reduction in iso-time RPB (r=−0.360, P=0.002) and RPLF (r=−0.293, P=0.011).

Table 2.

Aerobic exercise outcome measures by group. ^*

	Work rate (watts)	V̇E (L·min−1)	V̇O2 (L·min−1)	Heart rate (bpm)	Borg (score)
					Breathlessness	Leg fatigue
Peak data:
Baseline
RT-then-AT	55 ± 17	36.3 ± 8.9	1.44 ± 0.42	125 ± 17	4.6 ± 1.6	5.6 ± 2.3
CE-then-AT+RT	65 ± 24	38.6 ± 10.9	1.60 ± 0.48	116 ± 17	4.5 ± 2.3	5.2 ± 2.2
CE-then-AT	49 ± 23	32.1 ± 9.3	1.27 ± 0.39	112 ± 20	4.5 ± 2.5	5.3 ± 2.1
Week 8
RT-then-AT	56 ± 21	38.5 ± 8.1	1.39 ± 0.41	125 ± 19	4.5 ± 1.4	4.4 ± 2.8
CE-then-AT+RT	67 ± 23	38.5 ± 11.6	1.52 ± 0.51	116 ± 20	5.3 ± 2.7	5.8 ± 2.8
CE-then-AT	51 ± 20	34.0 ± 6.2	1.45 ± 0.38	116 ± 18	4.7 ± 2.1	5.7 ± 2.1
Week 16
RT-then-AT	61 ± 25	38.0 ± 11.0	1.50 ± 0.52	127 ± 18	4.5 ± 1.2	4.4 ± 2.2
CE-then-AT+RT	74 ± 28	37.8 ± 12.5	1.66 ± 0.57	120 ± 19	4.5 ± 1.5	4.5 ± 2.3
CE-then-AT	62 ± 22	36.2 ± 8.7	1.50 ± 0.36	116 ± 18	4.1 ± 2.0	4.2 ± 2.0
Change score: baseline to week 16
RT-then-AT	6.6 ± 13.2	1.7 ± 5.1	0.06 ± 0.22	2 ± 12	−0.1 ± 1.6	−1.2 ± 1.9
CE-then-AT+RT	9.5 ± 16.7	−0.7 ± 6.3	0.06 ± 0.28	4 ± 10	−0.1 ± 1.6	−0.6 ± 1.9
CE-then-AT	12.3 ± 12.9	4.1 ± 6.7	0.23 ± 0.34	4 ± 10	−0.4 ± 1.4	−1.1 ± 1.8
Iso-time data:
Baseline
RT-then-AT	48 ± 17	32.5 ± 9.2	1.28 ± 0.41	116 ± 17	3.3 ± 1.8	4.0 ± 2.4
CE-then-AT+RT	59 ± 22	34.5 ± 10.0	1.37 ± 0.45	108 ± 18	3.3 ± 1.6	4.0 ± 1.7
CE-then-AT	42 ± 20	29.9 ± 7.0	1.17 ± 0.33	103 ± 17	2.8 ± 2.0	3.3 ± 2.0
Week 8
RT-then-AT	48 ± 17	31.7 ± 6.9	1.21 ± 0.38	114 ± 16	3.2 ± 1.4	3.1 ± 1.6
CE-then-AT+RT	59 ± 22	34.8 ± 10.6	1.36 ± 0.47	108 ± 17	3.7 ± 2.4	4.2 ± 2.4
CE-then-AT	42 ± 20	30.2 ± 5.8	1.19 ± 0.33	103 ± 15	2.8 ± 1.8	3.4 ± 1.8
Week 16
RT-then-AT	48 ± 17	31.6 ± 8.9	1.22 ± 0.45	111 ± 14	2.9 ± 1.4	2.8 ± 1.7
CE-then-AT+RT	59 ± 22	31.7 ± 9.1	1.34 ± 0.45	106 ± 15	2.7 ± 1.9	2.9 ± 1.9
CE-then-AT	42 ± 20	29.2 ± 6.5	1.12 ± 0.28	100 ± 15	1.8 ± 1.4	2.1 ± 1.4
Change score: baseline to week 16
RT-then-AT	--	−1.0 ± 4.6	−0.07 ± 0.22	−5 ± 18	−0.4 ± 1.8	−1.2 ± 1.8
CE-then-AT+RT	--	−2.9 ± 5.2	−0.03 ± 0.19	−3 ± 11	−0.7 ± 1.4	−1.1 ± 1.6
CE-then-AT	--	−0.8 ± 4.3	−0.06 ± 0.24	−3 ± 12	−0.9 ± 1.5	−1.2 ± 1.6

Definition of abbreviations: AT, aerobic training; CE, chair exercise; V̇E, minute ventilation; V̇O₂, oxygen consumption; RT, resistance training.

^*Data are mean ± SD.

MANOVA between-subjects: Group assignment F(2,70) = 1.145, P = 0.311.

MANOVA within-subjects effects: time F(2,70) = 3.563, P = 0.000; interaction of time-by-group assignment F(2,70) = 1.078, P = 0.372.

Figure 3.

RPLF at Peak (panel A) and Iso-time (panel B) Exercise by Group. Values are mean ± SEM. Definition of abbreviations: AT = aerobic training, CE = chair exercise, RPLF = rate of perceived leg fatigue, RT = resistance training. *baseline compared to week 8, P<0.05; +week 8 compared to week 16, P<0.01.

Table 3 summarizes peripheral muscle strength and endurance, and walking test and chair rise outcomes. There were no differences between the groups at baseline (P=0.441), but there were significant time (P<0.001) and time-by-group assignment interaction effects (P<0.001). All three groups demonstrated improvements in the 1RM (P<0.001), leg extension endurance time (P<0.001), 30-second chair rise (P=0.027) and 6MD (P=0.006) over time, but only the 1RM (P<0.001) and leg extension endurance time (P=0.001) demonstrated significant time-by-group differences. For the acquisition of strength, post hoc testing revealed that the RT-then-AT and CE-then-AT+RT groups had greater gains in the 1RM than the CE-then-AT group, but there was no difference in the gain in the 1RM between the RT-then-AT and CE-then-AT+RT groups. For muscular endurance, post hoc testing revealed that the RT-then-AT group had a slightly greater gain in the leg extension endurance time than the CE-then-AT+RT group (P=0.050) and a substantially greater gain than the CE-then-AT group (P=0.001); the CE-then-AT+RT group had a greater gain than the CE-then-AT group (P=0.039). Changes in the 1RM were associated with changes in the 6MD (r=0.373, P=0.001) and LE-ET (r=0.416, P<0.001).

Table 3.

Muscle strength and endurance, and walking test outcomes by group. ^*

	1RM† (kg)	LE-ET (s)	30-second chair rise‡	6MD (m)
Baseline
RT-then-AT	300 ± 69	237 ± 190	10 ± 3	381 ± 88
CE-then-AT+RT	303 ± 102	290 ± 172	9 ± 3	369 ± 97
CE-then-AT	301 ± 74	262 ± 239	10 ± 5	346 ± 104
Week 8
RT-then-AT	363 ± 107	565 ± 451	10 ± 4	382 ± 91
CE-then-AT+RT	327 ± 100	379 ± 271	10 ± 3	389 ± 95
CE-then-AT	330 ± 86	350 ± 306	10 ± 5	348 ± 107
Week 16
RT-then-AT	380 ± 111	649 ± 456	11 ± 4	391 ± 87
CE-then-AT+RT	387 ± 118	570 ± 359	10 ± 4	398 ± 89
CE-then-AT	341 ± 97	385 ± 351	11 ± 6	356 ± 115

Definition of abbreviations: 1RM, one repetition maximum; 6MD, six minute distance; AT, aerobic training; CE, chair exercise; LE-ET, leg extension endurance time; RT, resistance training.

^*Data are mean ± SD.

^†1RM represents the sum of 6 lifts.

^‡30-second chair rise represents the number of completed stands.

MANOVA between-subjects: Group assignment F(2,72) = 0.997, P = 0.441.

MANOVA within-subjects effects: time F(2,72) = 22.490, P = 0.000; interaction of time-by-group assignment F(2,72) = 4.719, P = 0.000.

Post hoc (P value is one-tailed.):

1RM: RT-then-AT > CE-then-AT (P=.006), CE-then-AT+RT > CE-then-AT (P=.0015), no difference RT-then-AT vs CE-then-AT+RT (P=.415).

LE-ET: RT-then-AT > CE-then-AT (P=.001), CE-then-AT+RT > CE-then-AT group (P=.0385), RT-then-AT marginally > CE-then-AT+RT (P=.050).

Table 4 summarizes the outcomes with respect to symptoms of dyspnea for activities important in daily life, fatigue, functional status, and physical activity. There were no differences between the groups at baseline (P=0.857) and no interaction of time-by-group assignment (P = 0.215), but there were significant changes with time for CRQ Dyspnea (P=0.001) and physical activity (P=0.001), but not for CRQ fatigue (P=0.099) or functional status (P=0.588). Improvement in CRQ dyspnea was not related to the reduction in the RPB at iso-time exercise (r=−0.110, P=0.349), but was related to gains in the 1RM (r=0.255, P=0.027), 6MD (r=0.316, P=0.006), and the 30-second chair rise (r=0.324, P=0.005).

Table 4.

Symptoms, functional status, and physical activity outcomes by group. ^*

	CRQ Dyspnea (scale score)	CRQ Fatigue (scale score)	FPI (total score)	CHAMPS† (hours/week)
Baseline
RT-then-AT	3.7 ± 0.9	4.6 ± 1.1	2.1 ± 0.4	7.5 ± 6.3
CE-then-AT+RT	3.9 ± 1.1	4.3 ± 1.0	2.0 ± 0.5	8.2 ± 8.1
CE-then-AT	3.6 ± 0.9	4.3 ± 1.3	2.0 ± 0.4	8.3 ± 7.6
Week 8
RT-then-AT	4.1 ± 1.0	4.6 ± 1.1	2.1 ± 0.4	9.6 ± 6.3
CE-then-AT+RT	4.1 ± 1.1	4.4 ± 1.1	2.0 ± 0.4	11.4 ± 8.8
CE-then-AT	4.0 ± 1.1	4.3 ± 1.2	2.0 ± 0.4	10.8 ± 7.3
Week 16
RT-then-AT	4.1 ± 1.1	4.6 ± 0.9	2.0 ± 0.4	8.2 ± 7.2
CE-then-AT+RT	4.5 ± 1.2	4.4 ± 0.9	2.0 ± 0.4	10.9 ± 10.7
CE-then-AT	4.0 ± 1.2	4.9 ± 1.1	2.1 ± 0.4	9.6 ± 7.7

Definition of abbreviations: AT, aerobic training; CE, chair exercise; CHAMPS, CHAMPS Activities Questionnaire for Older Adults; CRQ, Chronic Respiratory Disease Questionnaire; FPI, Functional Performance Inventory; RT, resistance training.

^*Data are mean ± SD.

^†Reports hours per week of the total physical activity.

MANOVA between-subjects: Group assignment F(2,72) = 0.496, P = 0.857.

MANOVA within-subjects effects: time F(2,72) = 3.763, P = 0.001; interaction of time-by-group assignment F(2,72) = 1.287, P = 0.215.

Discussion

This is the first controlled trial to examine the influence of the temporal sequence of resistance and aerobic training on pulmonary rehabilitation outcomes. Performing resistance training as a preconditioning strategy compared to concurrent aerobic and resistance training demonstrated an advantage with respect to the acquisition of greater muscle endurance, but did not improve the acquisition of aerobic capacity or muscular strength. Although the experimental group did not lead to consistent benefits over the control groups, all three groups experienced benefits in important pulmonary rehabilitation outcomes. The three groups were similar with respect to disease severity, age, and body composition, thus differing sample characteristics do not explain the lack of treatment effect on aerobic capacity.

Performing aerobic and resistance training concurrently had no effect on aerobic training tolerance, but did result in a lower resistance training volume compared to those that completed the resistance training prior to initiating aerobic training suggesting an advantage for the treatment group with respect to tolerance of resistance training. Both the resistance training and cycling engaged the quadriceps muscle; it is possible that the concurrent training was fatiguing especially for patients with lower FFMI which may explain the lower resistance training volume they achieved and the blunted improvement in muscle endurance compared to the treatment group who performed training sequentially. Other factors which influenced training tolerance; sarcopenia and greater hyperinflation were associated with poorer attendance, and both lower aerobic and resistance training volumes. Neither airflow obstruction nor diffusion limitation was associated with attendance or training volume.

The postulation that the sequential approach would improve aerobic training outcomes, the primary outcome of the study, compared to concurrent training was not supported. One possible explanation for this could be that only 20% of the sample had an FEV₁ less than 30% predicted and only 10% demonstrated low FFMI. It is impossible to know if we had been able to recruit more people at the more severe end of the disease spectrum whether more of the specific aims would have been supported. Greater than 1,300 potential patients responded to study recruitment efforts and only 4 people cited severity of their lung disease as a reason for declining. This raises the possibility that people with greater disease severity choose not to respond to efforts to engage them in pulmonary rehabilitation, even in this case where it was offered at no cost to participants. Little work has been done on why people with more severe COPD choose not to participate in pulmonary rehabilitation. Taylor et al³³ examined reasons for refusal in patients with diverse disease severity and found that almost half refused due to barriers related to travel and location of the training site and approximately a fourth refused due to their belief that pulmonary rehabilitation would negatively affect their health status and increase dyspnea. More work is needed to refine the reasons that patients with very severe disease do not participate in pulmonary rehabilitation and also to remove barriers to their participation in pulmonary rehabilitation.

Although the sequential approach to resistance and aerobic training yielded almost a 30% greater increase in muscle endurance and an 8% higher resistance training volume compared to concurrent resistance and aerobic training, other training outcomes were similar between the two groups. Given the limited advantages with respect to muscle endurance and resistance training volume, the sequential approach to training (16 weeks) in severe COPD is less efficient in terms of the duration of pulmonary rehabilitation compared to concurrent resistance and aerobic training (8 weeks). Therefore, the sequential approach cannot be generally recommended for people with severe COPD. But for specific individuals who are unable to tolerate concurrent resistance and aerobic training, separating resistance and aerobic training by using a sequential approach or alternate day training is an option to be explored.

Comparisons of aerobic training alone to the combination of concurrent aerobic and resistance training have yielded mixed results.^34–36 Bernard et al³⁴ found that both the combination of concurrent aerobic plus resistance training and aerobic training alone improved 6MD and CRQ Dyspnea and Fatigue scores to a similar extent which suggests that the addition of resistance training had no advantage. In contrast, Panton et al³⁵ and Phillips et al³⁶ found the combination of aerobic and resistance training improved functional outcomes (timed walking tests, activities of daily living) whereas aerobic training alone did not which suggests resistance training may contribute to the improvement in functional outcomes like walking capacity. In the present study improvement in peripheral muscle strength was associated with improvement in walking capacity, which supports the use of resistance training for improving functional outcomes. However changes in walking capacity in the present study were modest and similar to those reported by Maltais et al,³⁷ whose primary aerobic training intervention was also cycle ergometry. The inclusion of a constant work rate cycle endurance test may have strengthened the association between gains in muscle strength and gains in exercise endurance. Bernard et al³⁴ found similar changes in peak and iso-time aerobic training outcomes between the two groups, which is similar to the findings of the present study that the addition of resistance training (whether concurrent or sequential) does not improve aerobic capacity. Bernard et al³⁴, Panton et al³⁵ and Phillips et al³⁶ only found improvements in muscular strength in the group that performed resistance training. The present study supports findings that aerobic training alone has a negligible effect on the acquisition of peripheral muscle strength.

Two studies compared aerobic training alone, resistance training alone, and the combination of aerobic plus resistance training with differing results.^{6, 38} Interestingly, Ortega at al³⁸ found only aerobic training alone produced statistically significant improvements in peak aerobic training outcomes. Peak work rate improved by 28% with aerobic training alone while resistance training alone and the combination of aerobic plus resistance training produced non-significant changes of 10% and 12% respectively. In the present study, although statistically the three groups achieved similar improvements in peak work rate, the group that only performed aerobic training improved by 25% (12 watts) which was slightly better than the groups that did both resistance and aerobic training who improved by 12–15% (6–9 watts). In contrast, Vonbank et al⁶ found significant increases in peak work rate for all three groups, but only aerobic training alone and aerobic plus resistance training produced improvements in peak oxygen consumption and V̇E/V̇O₂. Sample characteristics differed between studies; disease severity was greater and baseline aerobic capacity was lower in the sample of Ortega et al³⁸ compared to Vonbank et al⁶. A potential explanation for these disparate findings would be that COPD patients with greater disease severity might not tolerate concurrent aerobic and resistance training. Ortega et al³⁸ found all three groups improved with respect to muscular strength, constant work rate endurance, and CRQ Dyspnea. Aerobic training alone and resistance training alone produced statistically significant improvements in CRQ Fatigue, while the combination of aerobic plus resistance training did not. In contrast to the findings of Ortega et al,³⁸ Vonbank et al⁶ found improvements in muscular strength only for the 2 groups that performed resistance training. The results of the current study did show a small increase (13%) in muscular strength in the group that did not perform resistance training (CE-then-AT group) indicating that the increasing resistance on the cycle had a small effect on quadriceps strength, but the two groups that performed resistance training improved to a much greater extent (approximately 27%). Taken as a whole, the state of the science does not support the presence of an additive effect for the combination of aerobic plus resistance training for improving aerobic capacity.

Greater hyperinflation and lower FFMI were associated with lower aerobic and resistance training volumes and poorer attendance, all of which would interfere with the goals of pulmonary rehabilitation. Additionally lower baseline FFMI was related to lower gains in both aerobic and resistance training outcomes. An implication of this finding is that addressing sarcopenia prior to pulmonary rehabilitation may positively impact training outcomes. Severe home-bound COPD patients that received 12 weeks of amino acid supplementation demonstrated improved physical activity, muscle strength, and FFMI.³⁹ In muscle-wasted COPD patients concurrent pulmonary rehabilitation and dietary supplementation improved FFMI, but improvement in dyspnea and functional capacity was similar to pulmonary rehabilitation alone, which raises the possibility that nutritional supplementation may be more effective if applied prior to pulmonary rehabilitation.⁴⁰ In patients without fat-free mass depletion dietary supplementation did not improve pulmonary rehabilitation outcomes compared to pulmonary rehabilitation alone.⁴¹ To date, there are no data available in more severe patients suffering from sarcopenia as to whether improving FFMI prior to initiating pulmonary rehabilitation would enhance pulmonary rehabilitation outcomes. Fat-free mass is an important factor which influences pulmonary rehabilitation outcomes and further study is warranted to determine how this factor can be manipulated to improve the efficacy of pulmonary rehabilitation.

In conclusion, whether resistance training and aerobic training are conducted sequentially or concurrently produces similar improvements in aerobic training outcomes. The sequential approach to resistance and aerobic training has a significant advantage with respect to muscle endurance and resistance training tolerance compared to concurrent resistance and aerobic training, but the acquisition of muscle strength was similar. In COPD patients with severe airflow obstruction, sarcopenia and hyperinflation have a negative impact on pulmonary rehabilitation outcomes.