INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

CHRONIC SYSTOLIC HEART FAILURE (CHF) is the only cardiac diagnosis that is continuing to increase in prevalence in the United States due to the prolonged survival of those with hypertension and ischemic heart disease, as well as improved average life span of the population (35, 42). The clinical hallmark of the disease is exercise intolerance, manifested as fatigue and dyspnea during increasingly minimal activities (21).

Mounting evidence suggests that peripheral skeletal muscle abnormalities figure prominently in the exercise intolerance associated with CHF (15-17, 26, 40, 53, 54), whereas central hemodynamic parameters, such as ejection fraction, are far less predictive of clinical symptoms or mortality (34). These peripheral abnormalities include a skeletal muscle myopathy characterized by preferential loss and atrophy of type I (slow, oxidative) fibers, decreased oxidative enzyme capacity and mitochondrial volume density, early activation of glycolytic pathways of ATP generation during work, and reduced muscle endurance, strength, power, and overall exercise tolerance compared with healthy, age-matched individuals (39). The selective loss of oxidative fibers distinguishes this condition from type II fiber selectivity of muscle atrophy common to aging and disuse syndromes (29, 47, 66) and suggests that a different pathogenetic mechanism may be operative.

The degree of exercise intolerance in CHF correlates strongly with both mortality and quality of life (69, 70); yet current medical therapies often fail to specifically address many of the potential mechanisms that underlie these symptoms (25). Aerobic exercise interventions have been shown to be tolerable in middle-aged and young-old patients with stable CHF and result in improved aerobic capacity without improving central hemodynamic features (contractility, stroke volume, ejection fraction, chamber diameter, filling times) of the disease (1, 18-20, 24, 62), as well as improved survival, quality of life, and reduced rate of hospitalizations (5, 27). However, the limited experience reported from the European Heart Failure Training Group in women (n = 9) and patients over 70 yr (n = 14) suggests that the improvements in exercise capacity seen in younger men after aerobic training may not extend to older individuals (27).

Therefore, we first conducted a cross-sectional comparison of older patients with CHF and subjects with chronic illnesses but no CHF to test our hypothesis that muscle function would be more specifically impaired than aerobic capacity in elderly patients with CHF. Next, we conducted a randomized controlled clinical trial to determine the efficacy of high-intensity progressive resistance training (PRT) in older patients with CHF. We hypothesized that strength training in these patients would be well tolerated and result in improved overall exercise performance without changes in central cardiac function. Moreover, we hypothesized that the adaptations in skeletal muscle following strength training would explain much of the improvement in overall functional capacity.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Design

Subject Recruitment and Eligibility

Intervention Protocols

PRT group. Subjects in the experimental group participated in high-intensity PRT 3 days/wk for 10 wk under the supervision of a research assistant. High-intensity PRT was defined as training with 80% of the maximal weight that could be lifted in good form [the one repetition maximum (1 RM)]. Strength was retested every tenth session to establish a new 1 RM. Dynamic contractions of the large upper and lower body muscle groups (seated leg press, chest press, knee extension, triceps and knee flexion) were performed on pneumatic-resistance training equipment (Keiser Sports Health Equipment, Fresno, CA). Training sessions averaged 60 min and were preceded by 2 min of walking or cycling for warm-up and ended with 5 min of stretching. Subjects performed three sets of eight repetitions on each machine before moving on to the next. Each repetition lasted 6-9 s with a 2- to 3-s rest between repetitions and a 60- to 90-s rest between sets. As needed, rest periods of 1-3 min were inserted between machines.

Placebo control group. Subjects in the placebo control group participated 2 days/week in supervised, low-intensity stretching exercises, without resistance other than that imposed by gravity, that were designed as a "sham exercise" condition. Exercise sessions lasted 1 h and included stretching exercises of the neck, trunk, and extremities.

Outcome Measures

Overall functional performance. The 6-min walk, as described by Guyatt et al. (36), was used to measure overall functional exercise performance. The distance walked was recorded to the nearest foot by a rolling measuring wheel (RediMeasure, Redington, Windsor, CT). The self-perceived exertion scale defined by Borg and Linderholm (9) was obtained from the subject on completion of each test.

Aerobic capacity. Peak aerobic capacity (O₂) and time to exhaustion were measured on a motorized treadmill (Woodway, Waukesha, WI) by a symptom-limited, graded test to exhaustion. Treadmill speed was held constant at 80% of the subject's habitual gait speed, as determined by an ultrasonic gait monitor (Ultratimer, DCPB Electronics, Glasgow, Scotland), a device that measures speed (m/s) via an ultrasonic transmitter. Beginning at 0% grade for the first 2 min, the grade was then increased 2% every minute until subject exhaustion was reached. The subject's expired gases were obtained through a Hans Rudolph face mask (series 8930 no. 5, Hans Rudolph, Kansas City, MO) custom-fitted with a thin compliant sealant (Elastogel) around the mask to prevent air leakage, a method validated in older populations (52). Expired gases were continuously analyzed by a Beckman LB-2 CO₂ analyzer and an Applied Electrochemistry S-3A O₂ analyzer using standard procedures (31). The higher of the two values for peak O₂ obtained during screening and baseline testing was used in the analyses.

Musculoskeletal performance indexes: strength and endurance. Dynamic concentric muscle strength was determined using the same type of pneumatic resistance equipment as that used in the training intervention by the 1-RM method and expressed in newtons (N). To minimize improvement related to repeated testing, the best of two measurements taken 1 wk apart was used as the baseline value. 1-RM measurements were reliable with test; retest differences ranged from 5.0 ± 2.2% (leg press) to 9.1 ± 3.2% (chest press).

Skeletal muscle biopsy procedures. Percutaneous needle biopsies of the nondominant vastus lateralis (knee extensor) muscle were obtained under local anesthetic (1% Xylocaine hydrogen chloride) using the method described by Evans et al. (28). To determine citrate synthase activity (CSA), biopsies were quick-frozen in liquid nitrogen, homogenized, and assayed spectophotometrically using the method described by Srere (64), and values were normalized to the total protein content of each muscle homogenate obtained using the Bradford colorimetric assay (Bio-Rad, Sunnyvale, CA) (10). The mean of duplicate values was used for analyses.

Total body muscle mass. Total body skeletal muscle mass was estimated from a 24-h urine creatinine collection obtained in an outpatient setting as described by Heymsfield et al. (41). Urine collections were obtained following a 3-day meat-free diet that was continued throughout the urine collection. Creatinine was measured by the Jaffe kinetic reaction on a clinical chemistry analyzer (Roche Diagnostic Systems, Montclair, NJ), according to the procedures of Jaffe (43) as modified by Roche Diagnostic Systems. CV of this assay in our laboratory is 4.0%. Urinary 3-methylhistidine excretion was measured from the same urine samples used for creatinine analyses as an index of myofibrillar protein breakdown, according to the method of Wassner et al. (68). The mean of two 24-h urine collections was used in analyses.

Cardiac function. Resting cardiac function was determined in the left lateral supine position by transthoracic M-mode and two-dimensional Doppler echocardiography (Sonos 2500, Hewlett-Packard, Palo Alto, CA). Left ventricular (LV) function was assessed using ejection fraction measured from the apical images. LV diastolic function was determined by mitral valve inflow from the apical four-chamber image in all patients who were not in atrial fibrillation. LV morphology was measured using M-mode measurements of the LV posterior wall, anterior wall, and the intraventricular diameter in the parasternal long view. All echo images were analyzed blindly by a cardiologist on an off-line workstation (Tomtec, CA).

Statistical Analysis

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subject Characteristics

Compliance, Safety, and Health Status

After a baseline 1 RM determination was made, one control subject developed an exacerbation of underlying trochanteric bursitis that resolved with rest. One control subject developed exertional angina with ischemic electrocardiogram (ECG) changes during the screening exercise treadmill test but not during 1 RM testing or control exercise sessions. No potential subjects were excluded for ECG abnormalities during monitored screening weight lifting tests. Four experimental subjects required minor modifications in the resistance training regimens due to intermittent mild musculoskeletal symptoms. Overall, 48 peak exercise treadmill tests, 91 maximal strength tests, and 410 exercise training sessions (experimental and control) were completed without cardiovascular complications.

Exercise Capacity in Normal Controls and CHF Patients

Clinical status of heart failure patients, categorized by NYHA class, was related to several musculoskeletal indexes and performance measures. Patients with increasingly severe CHF had smaller mean type I muscle fiber area (P = 0.028) and lower urinary 3-methylhistidine excretion (187.9, 112.0 ± 12.1, and 70.1 ± 7.8 µmol/g creatinine for class I, II, and III, respectively; P = 0.028; see Fig. 1). By contrast, mean type II muscle fiber area was unrelated to clinical disease severity. Increasing severity of CHF was also related to slower habitual gait speed (1.5, 1.0 ± 0.1, 0.6 ± 0.0 m/s for class I, II, and III, respectively; P = 0.003), lower peak aerobic capacity (18.5, 16.1 ± 0.9, and 11.0 ± 1.0 ml · kg¹ · m¹ for class I, II, and III, respectively; P = 0.02), and poorer performance on the 6-min walk (341, 426 ± 24, and 219 ± 41 m for class I, II, and III, respectively; P = 0.003). Ejection fraction and other cardiac indexes were unrelated to NYHA class.

View larger version (54K): [in this window] [in a new window]   Fig. 1.   Relationship between disease severity and estimates of myofibrillar protein degradation in skeletal muscle in women with chronic heart failure (CHF) at baseline. New York Heart Association class vs. urinary 3-methylhistidine excretion (µmol/g creatinine) at baseline. P = 0.028.

Intervention Outcomes

Overall exercise performance. There were no baseline differences in peak exercise measurements between groups. Baseline 6-min walk distance was significantly correlated with both knee extensor strength (r = 0.559; P = 0.02) and peak O₂ (r = 0.768; P = 0.0005). The results of strength training on overall exercise performance are summarized in Table 3.

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Musculoskeletal indexes that independently predict functional performance measured by 6-min walk in stepwise multiple regression analysis. A: change in 6-min walk as a function of change in type I muscle fiber area. B: change in 6-min walk as a function of change in muscle oxidative capacity.

Changes in musculoskeletal performance. Musculoskeletal performance between treatment groups was comparable at baseline (see Table 5). Significant strength and endurance gains were achieved for all resistance-trained subjects in every muscle group tested, whereas no significant changes were seen in the controls. Gains in muscle strength ranged from 33.5 ± 7.3% (leg press) to 68.0 ± 13.2% (knee extension) for a mean gain of 43.4 ± 8.8% compared with 1.7 ± 2.8% in the controls (P = 0.001). Submaximal muscle endurance increased 299 ± 66% in the resistance trained group compared with 1 ± 3% in the controls (P = 0.001) (see Fig. 3).

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Changes in muscle endurance for each muscle group in response to 10 wk of high-intensity progressive resistance training (PRT) or placebo exercise condition (control). *P < 0.04; **P < 0.001.

Skeletal muscle adaptations. Metabolic and morphological musculoskeletal adaptations are summarized in Table 6.

Muscle Mass

Oxidative Capacity

1

1

1

1

Mean Fiber Area and Fiber-type Distribution

Relationships between musculoskeletal performance and other muscle characteristics. Baseline knee extensor strength was significantly related to baseline muscle mass (r = 0.644; P = 0.007), and similar trends were observed with other muscle groups (data not shown). Changes in knee extensor strength after training were directly associated with changes in muscle mass (r = 0.489; P = 0.056) but not with changes in oxidative enzyme capacity (Fig. 4A). Changes in knee extensor muscle endurance were, by contrast, significantly correlated with changes in oxidative capacity in that muscle group (r = 0.693; P = 0.012) (see Fig. 4B).

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Relationship between changes in musculoskeletal performance and skeletal muscle adaptations following 10 wk of PRT. A: changes in knee extensor (KE) strength are significantly correlated to changes in muscle mass measured by 24 h urine creatinine. B: changes in KE endurance are related to changes in muscle oxidative capacity measured by citrate synthase activity in the vastus lateralis muscle.

Cardiac function. Systolic ejection fraction was not related to baseline exercise capacity, NYHA class, Minnesota Living with Heart Failure Scale score, or musculoskeletal morphology or function. Neither resting indexes of systolic nor diastolic function changed during the study in either group, as summarized in Table 7. Moreover, there were no associations between changes in any measure of resting cardiac function and muscle performance or overall exercise capacity.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This is the first randomized controlled trial of whole body, high-intensity PRT in elderly women with CHF. We studied PRT based on the theoretical advantages of this exercise modality for the skeletal muscle derangements associated with this condition, as well as the lack of clear efficacy of aerobic exercise in this subpopulation of CHF patients.

There were four major findings in this study. First, at baseline, muscle dysfunction was evident from both the cross-sectional study comparing controls without CHF to CHF patients, as well as from the relationships between disease severity and muscle parameters. Women with CHF had ~60% of the muscle strength of women with functional impairment and other chronic diseases but comparable aerobic capacity. Previous cross-sectional studies have generally compared healthy age-matched individuals to patients with CHF rather than to other populations with chronic diseases, perhaps explaining the deficits found in both strength and aerobic capacity (39). As disease severity increased within our CHF cohort, the women were characterized by increased type I fiber skeletal muscle atrophy and lower 3-methylhistidine excretion, an index of myofibrillar protein breakdown, as well as lower aerobic capacity. Compared with cross-sectional fiber areas in elderly nursing home residents of 87 ± 1 yr that we have previously reported (29), the patients with CHF in this study had larger type II fiber areas (2,456 ± 178 µm³ in CHF vs. 2,229 ± 146 µm³) but smaller type I fiber areas (3,268 ± 192 µm³ in CHF vs. 3,603 ± 244 µm³). This decreasing myofibrillar protein degradation with increasing severity of heart failure has not been previously reported. Because our 3-methylhistidine values are normalized to muscle mass estimated from creatinine excretion, they do not simply reflect lower muscle mass but suggest other changes in muscle metabolism related to the progressive myopathy of CHF in these women.

Second, resistance training improved both peripheral musculoskeletal as well as overall exercise performance. Third, the improvements in musculoskeletal performance were closely related to the improved overall exercise tolerance after training. Fourth, our data suggest that the improvements in musculoskeletal function and overall exercise performance were mediated through beneficial adaptations in peripheral muscle without improving or adversely affecting central resting cardiac function.

Collectively, the above four findings extend the previous literature, attesting to the central role of peripheral skeletal muscle pathology in the exercise intolerance of CHF, highlight the oxidative, slow-twitch fiber specificity of this myopathy (16), and identify a potential treatment for the pathophysiology demonstrated.

The observed changes in submaximal muscle endurance may be more important from the patient's perspective than maximal force production since most daily activities are done at submaximal force levels. This result is parallel to many aerobic training interventions in CHF, which have shown more significant improvements in submaximal aerobic capacity or anaerobic threshold than in peak O₂ or maximal work load achievable (14). The adaptations leading to such improvements in both aerobic and anaerobic work capacity are clearly inducible with targeted training stimuli despite the central and peripheral pathology of CHF.

As hypothesized, these improvements in skeletal muscle metabolism and function following strength training were accompanied by clinically significant improvements in overall functional performance measured by the 6-min walk, a test widely used to assess the effectiveness of treatment regimens in CHF and other chronic diseases. Strength-trained subjects increased their walking distance by nearly 50 m compared with controls participating in low-intensity exercise. This relative gain of 13% is somewhat less than the gains reported after intensive aerobic training protocols (19, 44, 46) but comparable to lower-intensity aerobic protocols (6, 45), as well as pharmacological management of CHF (4, 33, 71), and virtually identical to benefits reported in a recent meta-analysis of 18 controlled trials of aerobic training programs in chronic airways disease (49 ± 26 m) (62).

Previous reviews of the benefits of aerobic exercise in CHF indicate that patients over 70 had little or no improvement in exercise tolerance compared with younger patients [e.g., increases in treadmill duration of 0.8 ± 1.2 m (older) vs. 1.8 ± 3.2 m (younger); and increases in peak O₂ of 0.2 ± 1.8 ml · kg¹ · m¹ (older) vs. 1.7 ± 1.8 ml · kg¹ · m¹ (younger)] (54). Also, patients who had a baseline maximal exercise test duration of >7 min did not respond to aerobic training. By contrast, our CHF patients responded to resistance training despite the fact that they ranged in age from 67 to 85 yr, and 50% of them had a baseline maximal exercise test of >7-min duration. This suggests that resistance training may be applicable in CHF subpopulations in whom the efficacy of aerobic exercise has not been adequately demonstrated.

Importantly, the improvement in overall functional performance in this study was related to a number of peripheral factors, including muscle strength, endurance, type I fiber area, and oxidative enzyme capacity but not to changes in any resting central hemodynamic factors or changes in peak O₂. Seventy-eight percent of the variance in improved exercise capacity was explained by adaptations within slow-twitch fibers (increase in type I fiber area and oxidative enzyme capacity). These findings suggest a mechanism by which resistance training may be providing benefit in CHF and document the metabolic and morphological adaptations still possible within these affected fibers with targeted training regimens. It is also possible that the resistive exercise was associated with alterations in habitual activity patterns and/or status of chronic diseases, such as arthritis, which indirectly could have improved exercise performance in these subjects. Future studies should include detailed assessments of these possible mechanistic factors and the physiological factors evaluated in this study.

Despite the wealth of evidence that skeletal muscle abnormalities contribute to symptoms in CHF (2, 16, 17, 40, 60), previous interventions generally have not been designed with anabolism in mind and thus the full potential of exercise in this condition has likely not yet been defined. Peripheral muscle adaptations in CHF patients have been demonstrated following nonaerobic exercise interventions in three previous uncontrolled or nonrandomized studies. Minotti and colleagues (55) reported increases in localized muscle endurance of 260% following a 4-wk, single-arm, isotonic handgrip training program. This improvement was associated with improved muscle energetics at submaximal work loads without changes in cardiac hemodynamics. Magnussen et al. (48) compared aerobic vs. strengthening exercises for the quadriceps muscle in an uncontrolled study of 11 CHF patients. Both forms of exercise improved work capacity, although the adaptations were specific to the mode of training. Strength and muscle area improved after the resistance training, whereas capillary density and oxidative enzyme capacity improved after endurance training only. Recently, Delagardelle et al. (23) reported the results of an uncontrolled study over 6 mo of hospital-based aerobic and strengthening exercise in 14 middle-aged (41-68 yr) patients (3 women) with CHF. Strength training of upper leg, shoulders and abdomen was low to moderate intensity (60-80% of the 10 RM; equivalent to ~48-64% of the 1 RM). There were significant improvements in cardiopulmonary indexes, muscle endurance (18-25%), and NYHA classification but minimal changes in strength, as might be expected from the relatively low intensity of the strengthening component. Overall, the design limitations inherent in these uncontrolled studies do not allow definitive conclusions to be drawn about the significance of their findings.

In conclusion, the morbidity, mortality, and societal costs of CHF in the elderly are substantial, and our current treatment paradigms do not optimally address what appears to be a primary derangement in muscle morphology, metabolism, and function. We have reported, for the first time in this trial, the safety of isolated high-intensity PRT in frail, older women with CHF, including the absence of symptomatic ischemia, arrhythmias, hemodynamic compromise, and unaltered resting cardiac function after 10 wk of exercise. These findings extend the previous literature on the lack of acute adverse hemodynamic consequences during single exposures to weight-lifting exercise in patients with CHF (50, 51) as well as other conditions (49). The plasticity of the musculoskeletal system, previously demonstrated even in the face of severe frailty and advanced age (29), now has been shown to extend to the skeletal myopathy accompanying CHF in the elderly. Ongoing research will further target the type I specificity of this myopathy, define the mechanistic features most amenable to intervention, and monitor the long-term adaptations in other subpopulations with CHF.