- Email: [email protected]
Return to work following an aquafitness and muscle strengthening program for the low back injured
1247
Return to Work Following an Aquafitness and Muscle Strengthening Program for the Low Back Injured Sandra M. LeFort, PhD, RN, T. Edward
Hannah,
PhD
ABSTRACT. LeFort SM, Hannah TE. Return to work following an aquafitness and muscle strengthening program for the low back iqjured. Arch Phys Med Rehabil 1994;75:1247-55. a Our purpose, in thii prospective clinical study, was to identify the best predictors of 2-month return to work or retraining for a group of low back injured subjects (n = 40) who completed at least 8 weeks of a communitybased rehabilitation program that combined aerobic and flexibiity exercise conducted in the water (aquafitness) with muscle strength and endurance training. Baselhte demographic characteristics and changes in physical fitness, pain, disability, and psychological well-being during the course of program participation were compared between two groups of low back injured subjects: those who returned to work (RTW) [n = 241, and those who did not (N-RTW) [n = 161. Subjects in both groups showed comparable improvement in measures of physical fitness at 8 weeks. However, multivariate analyses showed significant between-group differences in self-report measures. The RTW group showed sign&ant improvement in measures of pain, disability, anxiety, and vigor while self-esteem and a&& remained stable. The N-RTW group displayed no change in pain and disability variables and had sigrdticant deterioration in mean overall psychological well-being over time. The best predictors of return to work using logistic regression analyses were a first injury rather than a repeat injury to the lower back, and stability in self-esteem. Suggestions are offered for further research to examine the benefits of aqua& exercise for the low back injured, for additional interventions for those with a reinjury, and for maintaining or enhancing self-esteem as a treatment goal. 0 1994 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation
Low back pain (LBP), the so-called ‘ ‘nemesis of medicine and the albatross of industry,” is of increasing concern to society.’ Studies indicate that LBP and its related disability is of epidemic proportions particularly in the working population. In both Canada and the United States, the total number of reported job-related back injuries, the average time off work, and the amount awarded in compensation payments continue to increase at an alarming rate.2V3Similarly, health care costs associated with this condition are increasing. This escalating situation has led to expanded research and clinical interest in finding effective approaches to the rehabilitation of the low back injured that will return them to productive employment. One approach receiving increased attention is the role of activity, exercise, and aerobic fitness in comprehensive treatment programs for both subacute and chronic LBP.“-6 A basic premise of these programs is that low back injured individuals become deconditioned as a result of excessive protection and disuse of spinal joints and muscles, as well as general inactivity.’ Movement leading to improved fitness is thought to provide stability to joints and injured ligaments, to overcome muscle spasm and tension, and to reduce pain.3 Treatment programs that emphasize active rehabilitation and From the School of Nuning, McGill University; and the Department of Psychology and Division of Community Medicine, Memorial University of Newfoundland, Canada. Submitted for publication December 10, 1993. Accepted in revised form May 5, 1994. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the authors or upon any ornanization with which the authors are associated. Reprint requests to Sandra M. LeFort, PhD (Cand), RN, 111 Strawberry Marsh Road, St. John’s, Newfoundland, AlB 2V7, Canada. 8 1994 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation ooO3-9993/94/7511-2537$3.00/O
physical conditioning have been reported to significantly enhance cardiorespiratory fitness; increase muscular strength and endurance; improve flexibility; decrease perceptions of pain, depression, and disability; im$r;ve functional status; and, facilitate early return to work. However, despite these successes a significant subset of individuals with LBP who participate in physical conditioning programs do not return to work within the expected time if they return at a11.i4To date, few research reports have directly compared the physical and psychological outcomes of participation in exercise programs between LBP patients who subsequently return to work and those who do not. Such comparisons may be particularly valuable in identifying the subset of individuals for whom exercise as a treatment modality provides maximum benefits and in discriminating which articular benefits of exercise lead to positive outcomes. B This study reports on an 8- to 1Zweek community-based rehabilitation program that was designed as an adjunct to traditional physical therapy treatment. The program combines aerobic and flexibility exercise conducted in the water (aquafit exercise as distinct from swimming), with muscle strengthening and endurance exercises on weight-training machines. Aquafit aerobic exercise programs are now considered corn arable with land exercise in cardiorespiratory outcomes.“* P6 Such programs provide a safe alternative to land exercise especially for a deconditioned population such as those with LBP. To our knowledge however, no other studies of aquafit exercise in the rehabilitation of LBP patients have been reported with the exception of previous reports by Hannah and colleagues.‘7~20 This prospective study focused on low back injured individuals as they progressed through this aquafit and muscle strengthening program and compared two groups of participants: those who returned to work or began retraining two Arch Phys Med Rehabil Vol75, November 1994
1248
RETURN TO WORK AFTER AQUAFITNESS PROGRAM, LeFort
months after program completion (RTW group); and those who did not return to work (N-RTW group). A particular interest of the investigators was whether improvements in psychological well-being that are associated with physical exercise in the healthy population (ie, improved self-esteem, lowered state anxiety, and improved mood especially vigor) would also be experienced by the low back injured subjects in this study.21-23In our review of the literature, no studies were found that investigated the impact of physical conditioning in the LBP population on this set of variables, nor of the potential significance of these variables on early return to work. Three specific research questions were addressed: (1) Were there significant differences in physical fitness achievement between the RTW group and the N-RTW group over 8 weeks in the program? (2) Were there significant differences in self-reported pain, disability, and psychological well-being between the RTW group and the N-RTW group over 8 weeks in the program? (3) What were the best predictors of 2-month RTW or retraining for this group of subjects?
METHOD Subjects The study sample comprised individuals consecutively admitted over an g-month period to a rehabilitation program located in a fitness facility in a medium-sized Canadian city. Individuals were included if they met the following criteria: a primary diagnosis of idiopathic or mechanical LBP with no current evidence of herniated disc, fracture, or neurological complication; referred by a physical therapist and/or physician; and, receiving standard physical therapy treatment. Standard treatment included one or more of the following: transcutaneous-electrical nerve stimulation (TENS), ultrasound, heat/cold treatments, massage, and stretching/extension exercises. None of the subjects was receiving any additional psychological or structured education interventions. The vast majority of subjects were referred to the program from three physical therapy clinics. The caseloads of these clinics included virtually all of the Workers’ Compensation recipients in the city and surrounding area who were recovering from work-related injuries amenable to physical therapy. The program, initially funded by the provincial Worker’s Compensation Commission, was considered part of the rehabilitation of work-injured patients. Thus, we are reasonably certain that subjects are representative of low back injured individuals who were seeking treatment at the time of data collection.
Rehabilitation Program Description The program consists of two supervised components: aquafn classes combining aerobic and flexibility exercises, and muscle-strengthening workouts on training equipment for primary muscle groups, such as hip/lower back, quadriceps, hamstrings, abductorladductor, trapezius, latissimus dorsi, pectoralis major, deltoids, abdominal, biceps, and triceps. Normally, subjects attended two aquafit classes and three muscle-strengthening sessions per week. The individualized nature of the program, however, meant that some Arch Phys Med Rehabll Vol75,
November 1984
subjects did not begin muscle-strength training until the second or third week. All subjects were monitored biweekly by a physical therapist or physician. The aquafit component of the program was conducted in a swimming pool in water waist-deep to shoulder-height. Those who feared the water were assured that the program need not include swimming and were introduced to the water slowly. Ordinarily, 10 to 12 subjects participated in a class with a minimum of two rehabilitation instructors in attendance. Each aquafit session included a lo-minute warm-up, a 25-minute aerobics workout consisting of variations of walking and jogging lengths in the pool, a 20minute series of specific exercises (range of motion, flexibility, and stretching), and ended with a 5-minute cool-down. Subjects progressively increased their speed, number of repetitions, and level of exercise in a programmed fashion week by week. The longer subjects were in the program, the more the instructors encouraged them to work harder and longer. The muscle-strengthening program consisted of 14 possible exercises on 11 weight-training machines. Subjects began by lifting light weights, with a gradual increase of repetitions to 20, before adding extra weight. Subjects were consistently monitored by the rehabilitation staff throughout their weight-lifting sessions for correct warm-up (lo-minutes on a stationary bicycle), correct body alignment while using the machines, and appropriate increases in weights. Although most individuals used all machines, not all of them did so. Exercises were prescribed by their physical therapists based on the nature of the injury and the subject’s stage of recovery. For both the aquafit and muscle-strengthening components of the program, subjects were cautioned not to continue a particular exercise if pain increased significantly. The consultant physical therapist was notified in such cases. Other aspects of the program have been described elsewhere.“-*’
Measures The following two sets of measures were taken: indices of physical fitness; and self-report of demographic characteristics, pain, disability, and psychological well-being.
Physical Fitness Measures Pulse rates. Resting pulse rate before exercise and 2minute recovery pulse rate after aerobic exercise were used as crude indicators of change in aerobic fitness. The pulse rate taken during aerobic exercise was used as an index of change in work intensity.24 Subjects’ pulse rates were taken by palpating the radial pulse for a full 60 seconds and counting the beats per minute (bpm) using a wall clock. Pulse rates were monitored at the first complete aquafit session and monthly thereafter by trained program staff. Five heartrate measures were taken at each monitored session. Resting pulse rate was taken after subjects sat by the edge of the pool for 3 to 4 minutes before the aquafit class. They then entered the pool and participated in the warm-up and aerobics section of the aquafit class. During the final third of the aerobics section, following a specified exercise, the duringexercise pulse rate was measured. At the end of the final
RETURN TO WORK AFTER AQUAFITNESS PROGRAM, Lefort
aerobic exercise, the subject was asked to stop and the immediately-after pulse rate was taken followed by the l-minuteafter and 2-minute-after recovery pulse rates. Total weights &led. Total weights lifted on the training machines at the first versus the 15th session served as the index of muscular strength and endurance. Fifteen sessions represented between 6 to 8 weeks of muscle-strength training. At each weight-lifting session, the amount of weight lifted on each machine was recorded. These data were reviewed on an ongoing basis by the program staff.
Self-Report Measures General background information. Demographic characteristics, current treatment, history of the low back injury, and other pain-related information were collected through a structured patient interview. A second source of data were the referral notes from physicians and physical therapists. Pain. The Pain. Rating Index (PRI) of the McGill Pain Questionnaire (MPQ) was used to assess the subjective experience of painz5 The PRI consists of a list of 20 sets of words that describe pain quality and intensity. Subjects were asked to select the word sets that best described their pain in the past 24 hours and then to choose the most appropriate word within each selected set. The reliability and validity of the MPQ has been widely demonstrated and it remains the most commonly used measure of self-reported pain.26 Disability. Perceived functional impairment was measured using the Oswestry Low Back Pain Disability Questionnaire.27 It contains 10 sections on different activities of daily living relevant to the individual with LBP. Scores range from 0% to 100% disability. A 2-day test-retest reliability of r = 0.99 with good internal consistency has been reported. A number of investigators11*‘3~27 have found that scores are correlated to improvement in LBP patients during rehabilitation. Self-esteem. Self-esteem was measured using the Rosenberg Self-Esteem Scale, a IO-item Like&type format.28 The scale is regarded as a unidimensional measure of the self-acceptance aspect of self-esteem and all items revolve around liking or approving of the self. Scores range from 10 to 40, with lower scores indicating higher self-esteem. A 2-week test-retest reliability of r = 0.85 is reported; data on convergent, discriminant, and predictive validity have been described.29*30 Anxiety. The 20-item state portion of the State/Trait Anxiety Inventory (STAB, Form Yl was used as a measure of situational anxiety.31 Stability coefficients for the state anxiety scale are relatively low (r = 0.16 to 0.62), whereas internal consistency is high (Cronbach (Y= 0.92). Extensive evidence of the concurrent, convergent, divergent, and construct validity of the STAI scale is provided by the tool’s developers.31 Mood. Mood was measured using the 23-item version of the Memorial University Mood Scale (MUMS). Subjects are asked to answer “yes,” “no,” or “don’t know” to 7 positive and 7 negative mood adjectives, and 9 vigor adjectives.32 The measure consists of two subscales, affect and vigor, which are reported to be affected by participation in exercise.*9.32The scale has high internal consistency (Cronbath (Y = 0.80), and relatively low stability (r < 0.50 over
1249
3 days; r < 0.30 over 2 years), as expected for a valid measure of short-term mood. Construct validity of the tool has also been established.32
Procedure The study protocol and measurement instruments received ethical approval from a university-based Human Subjects Review Committee. Consenting subjects were interviewed by the first author a few days before beginning the rehabilitation program and monthly thereafter to collect demographic and other self-report data. The measures of physical fitness were collected by three trained rehabilitation program staff as part of standard record keeping. Pulse rates were measured at the first aquafit session and monthly thereafter. Weights lifted were recorded at each weight-lifting session. Only after, all subjects were discharged from the program did the first author gain access to the physical fitness data. Data on return to work or retraining were collected by questionnaire mailed to all subjects 2 months after completing the fitness program. Subjects were asked if they had returned to either full-time or part-time work or had started retraining. If they had not gone back to work, they were asked to explain the reasons for not returning and to indicate when they expected to be able to do so. If the survey was not returned within 2 weeks of mailing, subjects were contacted by the first author, and questions were answered over the telephone.
Statistics Data analyses were carried out in three steps using the Statistical Package for the Social Sciences (SPSSx). First, equivalence of groups on demographic, injury-related characteristics, and baseline measures of pain, disability and psychological well-being was assessed by means of &i-square analyses and 2-tailed Student t tests for independent groups. Second, two separate 2 X 2 repeated measures multivariate analysis of variance (RM MANOVA) procedures were conducted. Variables in the first RM MANOVA consisted of the physical fitness indicators: resting pulse rate, duringexercise pulse rate, 2-minute-after pulse rate, and total weights lifted on weight-training machines. The five selfreport measures (pain, disability, self-esteem, anxiety, and the affect and vigor subscales of mood) were included in the second RM MANOVA. For each RM MANOVA, outcome group (RTW versus N-RTW) served as the between-subjects factor, whereas time of assessment (baseline versus final assessment) served as the repeated within-subjects factor. The multivariate approach to repeated measures was used in these analyses for two reasons: to account for the interdependency of measures at the two assessment periods, and to assess how combinations of dependent variables changed together over time. The multivariate results are reported using Pillais’ F value because this criterion is reported to be the most robust with unequal sample size.33 Univariate results that contribute to significant multivariate effects are also reported. The univariate results are equivalent to 2 x 2 mixed model analysis of variance (ANOVAs) of the individual variables. To identify the best predictors of return to work, logistic Arch Phys Mad Fkhabil Vol75,
November 1904
1250
RETURN TO WORK AFTER AQUAFITNESS PROGRAM, LeFort
Table 1: Reasons Subjects (n = 10) Were Excluded From the Study Dropouts from program before 4 weeks Increased back pain Moved to another city Accident (unrelated to the program) Private-paying subjects decided that the program was not for them Returned to work after 3 weeks Completed program but excluded from the study Loss to follow-up Not eligible for work (maternity leave) Already working full-time while attending program
1 1 2
regression analyses were applied using a forward stepwise procedure. Return to work or retraining (0 = yes; 1 = no) served as the dependent variable. Because of the small sample size, predictor variables were limited to four. Only those variables that differed significantly between the RTW and the N-RTW groups were allowed into the equation. For each variable in the final model, odds ratios and 95% confidence intervals are reported. For all analyses, the 0.05 alpha level was set to determine statistical significance.
subjects took regular medication for pain, most commonly nonnarcotic analgesics and muscle relaxants. Subjects were employed in a wide variety of occupations, the most common classifications being: clerical and sales (n = 8); service (n = 8); scientific/technical (nursing and teaching) (n = 7); transportation/handling (n = 7); production/processing (n = 5); and natural resources (n = 5). Eighty-eight percent of subjects (n = 35) were receiving some kind of disability benefit, and no one was in litigation. To assess the equivalence of the RTW group compared with the N-RTW group, demographic and other injury-related characteristics were compared (table 2). Whether subjects were experiencing a first injury or a repeat injury was the only background characteristic that differed significantly between groups [x2 (1, n = 40) = 6.9, p < .Ol]. Thirty-three percent (n = 8) of RTW subjects versus 75% of N-RTW subjects (n = 12) stated that their current back injury was a repeat injury. Most subjects reporting a repeat injury had experienced only one other significant episode of LBP; however, 3 subjects Table 2: Characteristics at Program Entry of the RTW Group, the N-RTW Group, and Those Excluded From the Study Included
RESULTS
Subject Characteristics Fifty of 52 eligible subjects consented to participate. Of these, 10 subjects had to be excluded from the study for a variety of reasons (table 1). The remaining 40 subjects were not working because of their injury, but had full- or parttime jobs waiting for them. This group completed at least 8 weeks in the program and were able to be recontacted for assessment of return-to-work or retraining status. Two months after program completion, 60% (n = 24) of the subjects had returned to work (n = 23) or retraining (n = 1), whereas 40% (n = 16) had not. Reasons cited for not returning to work were continuing back pain (n = 14), or awaiting further medical investigation (n = 2). To test the equivalence of the 40 subjects included in the study compared with the 10 excluded subjects, betweengroup comparisons were made of demographic and injuryrelated characteristics (table 2), and baseline measures of pain, disability, and psychological well-being (table 3). Radiation of pain was the only variable that differed significantly between the included and the excluded group [x2 (1, n = 50) = 5.89, p < .02]. All excluded subjects reported that their pain was localized in the lower back and did not radiate down the leg, whereas 40% (n = 16) of the included group reported pain radiated down the leg. Radiating pain may indicate a more severe low back problem.34 As a group, the excluded subjects may have had less deep structure or neurogenic involvement than those included in the study. The 40 subjects included consisted of 25 men and 15 women whose ages ranged from 21 to 57 years (mean + SD = 37.1 ? 9.28). Seventy-eight percent (n = 31) were married or partnered, but no one lived alone. Only 13% (n = 5) had experienced pain for less than 3 months, thus 87% (n = 35) could be classified as having chronic LBP, with seven having had previous back surgery. Less than half of the Arch Phys Mad Rehabil Vol75,
November 1994
Variable
Excluded
RTW (n = 24)
N-RTW (n = 16)
Total (n = 40)
Total (n = 10)
37.6 9.0
36.0 10.1
36.9 9.4
35.4 9.6
16 (67%) a (33%)
9 (56%) 7 (44%)
25 (63%) 15 (37%)
6 (60%) 4 (40%)
20 (83%) 4 (17%)
11 (69%) 5 (31%)
31 (78%) 9 (22%)
8 (80%) 2 (20%)
16 (67%)* 8 (33%)*
4 (25%)* 12 (75%)*
20 (50%) 20 (50%)
4 (40%) 6 (60%)
(17%) (29%) (33%) (21%)
1 (6%) 8 (50%) 5 (31%) 2 (13%)
5 (13%) 15 (38%) 13 (32%) 7 (17%)
3 (30%) 3 (30%) 4 (40%)
17 (71%) 2 (8%) 5 (21%)
7 (44%) 4 (25%) 5 (31%)
24 (60%)+ 6 (15%)+ 10 (25%)+
8 (33%) 16 (67%)
9 (56%) 7 (44%)
17 (42%) 23 (58%)
4 (40%) 6 (60%)
6 (25%) 18 (75%)
1 (6%) 15 (94%)
7 (18%) 33 (82%)
3 (30%) I (70%)
20 (83%) 4 (17%)
11 (69%) 5 (31%)
31 (78%) 9 (22%)
6 (60%) 4 (40%)
21 (88%) 3 (12%)
14 (88%) 2 (12%)
35 (88%) 5 (12%)
6 (60%) 4 (40%)
Age (years) Mean SD Gender Men Women Marital status Married/partnered Single First or repeat injury First Repeat Pain duration <3 months 3-6 months 6 12 months > 12 mouths Radiation of pain None Above knee Below knee Pain medication Yes No Previous back surgery Yes No Program status WCB-sponsored* Private-paying Disability income” Yes No
4 7 8 5
*p s .Ol +p c .02 t Worker’s Compensation Board paying for cost of the rehabilitation proQ”. Includes both Worker’s Compensation payments and private insurance income.
RETURN
Table 3: Base-Line Scores of Self-Reported Pain, DisabMty, and Paychokqical Well-Being for Those Included and Those Excluded From the Study Excluded (II = 10)
Included (n = 40)
1251
TO WORK AFTER AQUAFITNESS PROGRAM, LeFort
Variable
Mean
SD
Mean
SD
Pain rating Disability Self-esteem State anxiety Affect Vigor
21.23 37.85 17.65 40.55 6.55 7.08
8.98 13.12 4.81 10.47 6.41 5.62
23.03 31.50 15.83 37.54 7.94 6.12
7.41 12.34 4.47 12.46 6.69 3.67
(12.5%) in the RTW group and 5 subjects (31%) in the NRTW group reported that they had suffered three or more previous injuries affecting the lower back. To further assess the equivalence of these two groups at program entry, baseline measures of pain, disability, and psychological well-being were compared (table 4). No significant between-group differences were found on any of these dependent variables. All p values were greater than 0.10. In summary, baseline characteristics of subjects who returned to work were not significantly different from the characteristics of those who did not return to work with the exception of history of prior back injury. Those in the RTW group were more likely to be experiencing a first back injury rather than a repeat back injury.
Physical Fitness To assess between-group differences in change in physical fitness over time, scores of resting pulse rate, during-exercise pulse rate, 2-minute recovery pulse rate, as well as total weights lifted were analyzed using 2 X 2 RM MANOVA. Outcome group (RTW versus N-RTW) served as the between-subjects factor and time of assessment (baseline versus final assessment) served as the repeated within-subjects factor. Final assessment for pulse rates was the I&week measure, whereas the 15th session was the final assessment for total weights lifted. Data from only 36 subjects were included in this analysis. Four subjects (two from each outcome group) were lost to analysis because of incomplete data sets on the physical fitness variables. From a review of the physical fitness records, missing data seemed to be a random occurrence. Hence, it is unlikely that dropping these four subjects affected the representativeness of the sample. A significant multivariate main effect for time (F(4,31) = 46.46, p = .OOO)was found with univariate results from all four dependent variables contributing to this effect: total weights lifted (F( 1,34) = 133.73, p = .OOO);during exercise pulse rate (F( 1,34) = 63.44 p = BOO); resting pulse rate (F(1,34) = 14.57, p = .OOl), and 2-minute recovery pulse rate (F( 1,34) = 4.7, p = .038). The multivariate main effect of outcome group (RTW versus N-RTW) was not significant (F(3,32) = 0.91 p = .47), nor was the group by time interaction (F(3,32) = 0.47, p = .76). For descriptive purposes, the means of pulse rates (broken down by outcome group and time of assessment) taken at the five monitored intervals during the aerobics component of aquafit class is presented in figure 1. After 8 weeks in
the program, subjects in both groups had, on average, decreased their mean resting pulse rate by 9bpm and decreased their mean postexercise 2minute recovery rate (RTW by 4bpm; N-RTW by Bbpm). Correspondingly, pulse rate during exercise increased substantially over time (RTW mean increase: 29bpm; N-RTW mean increase: 22bpm) indicating that subjects were working to a greater intensity by 8 weeks in the program and were probably elevating their heart rates into the cardiac training sensitive zone (70% of maximal heart rate).35 For water exercise, the target heart rate needs to be adjusted from 13 to 20bpm lower than for dry land exercise to take account of the cooling effect of the water as well as the lessened gravitational pull on the heart.‘5S24 These significant results suggest mean improvement in cardiac fitness for subjects in both groups. In muscle strength and endurance, subjects in the RTW group had a mean percentage increase of 130% in total weights lifted on 11 strength-training machines, whereas the N-RTW group had a 134% increase from the first to the 15th session. Figure 2 shows the means and standard deviations of total weights lifted broken down by outcome group and time of assessment. In summary, no significant differences were found in mean physical fitness achievement between the RTW and N-RTW groups. On average, these subjects were improving their level of physical fitness over time despite their perceived level of pain and disability.
Self-Report of Pain, Disability, and Psychological Well-Being To assess between-group differences in change in selfreport measures over time, scores of pain, disability, selfesteem, state anxiety, and the affect and vigor subscales of mood were analyzed together using a 2 X 2 RM MANOVA (table 4). Results showed a significant multivariate group by time interaction (F(6,33) = 2.49, p = .042), a significant Table 4: Scores of Self-Reported Pain, Disability, and Psychological Well-Being by Outcome Group (RlW or NRTW) and Time of Assessment (Base-Line and 8 Weeks)
Variable (possible range of scores) Pain rating (0 to 78) Baseline 8 weeks Disability (0% to 100%) Baseline 8 weeks Self-esteem (40 to lo)* Baseline 8 weeks State anxiety (20 to 80) Baseline 8 weeks Affect (-14 to +14) Baseline 8 weeks Vigor (0 to 18) Baseline 8 weeks
RTW (n = 24)
N-RTW (n = 16)
Mean
SD
Mean
SD
20.00 10.71
9.71 7.86
23.06 20.81
7.66 9.92
34.29 24.25
12.47 13.64
39.19 37.25
12.60 12.17
17.67 17.25
5.49 5.73
17.63 20.69
3.72 3.72
38.63 34.46
10.27 11.02
43.44 42.81
10.42 9.64
7.08 7.96
6.57 7.71
5.75 5.44
7.07 6.78
7.96 11.00
5.93 6.69
5.75 4.19
4.50 4.01
* Higher score indicates lower self-esteem.
Arch Phys Mod Rehabil Vol76,
November 19Q4
RETURN TO WORK AFTER AQUAFITNESS PROGRAM, LeFort
1252
Predictors of Return To Work The logistic regression equation developed to predict RTW for this sample of 40 subjects included the following four variables: first or repeat back injury, and the change scores (baseline score minus 8.week score) of those variables that contributed to the significant multivariate group by time interaction, namely pain, self-esteem, and vigor. The final regression model contained two variables that were associated with return to work 2 months after discharge from the fitness program. For this study sample, the only significant predictor of return to work was first or repeat back injury (odds ratio [OR] = 2.5,95% confidence interval [CI] = 1.17 to 5.23,~ = .02). The next best predictor, which approached statistical significance, was self-esteem (OR = 0.80, 95% CI = 0.64 to 1.01, p = .059). In the N-RTW group, 62% (n = 10) reported worse self-esteem scores by week 8, whereas 38% (n = 6) reported the same or improved scores. By contrast, only 29% (n = 7) of the RTW-group reported worse scores, whereas 71% (n = 17) reported the same or improved scores. Mean change in self-esteem for those in the N-RTW group was 3.06 units + 5.79, and for the RTW group was -0.42 units + 3.24. In summary, those experiencing a first low back injury
80
350 70 RESTING
I DURING
I IMMED. AFTER
I I-MIN. AFTER
I 2.MIN. AFTER
Fig l-This figure compares pulse rates measured at 5 intervals during the aerobic component of aquafit class at two time periods, the first complete aquafit class (baseline) and 8 weeks. By 8 weeks, both the RTW group (n = 22) and the N-RTW group (n = 14) had decreased their resting pulse rate, increased during exercise pulse rate, and had a steeper slope of pulse rate recovery from immediately after exercise to l-minute and 2.minutes after aerobic exercise. These results suggest mean improvement in aerobic exercise for both groups. - n RTW Baseline; - A - N-RTW Baseline; - 4 - RTW 8 Weeks; - 0 - N-RTW 8 Weeks.
multivariate main effect of group (F(6,33) = 2.74, p = .028), and a significant multivariate main effect of time (F(6,33) = 3.37, p = .Ol 1). The univariate results contributing to the significant multivariate effects are reported in table 5. The significant multivariate group by time interaction indicates that the overall pattern of change in these variables during the 8 weeks in the program was different in the RTW group compared with the N-RTW group. Subjects who subsequently returned to work had a significant mean reduction in pain and disability as well as significant mean improvement in the vigor subscale of mood and state anxiety. Selfesteem and overall affect remained stable. Conversely, subjects in the N-RTW group, with no significant change in their pain or disability scores, significantly deteriorated in scores of self-esteem and vigor. Mean state anxiety for this group, which was elevated compared with published population norms (43 v 35) but consistent with other chronic LBP populations,31’36remained unchanged during the 8 weeks in the program as did overall affect. Arch Phys Med Rehabil Vol75, November 1994
300
v) 250 z d 8 200 $ z g 150 Q z L
100
50
0 1st SESSION
15th SESSION
Fig 2-This figure compares total weights lifted in kilograms by the RTW group (n = 22) and the N-RTW group (n = 14) on 11 weight-training machines at the 1st and 15th muscle strengthening and endurance sessions. S RTW q N-RTW.
RETURN TO WORK AFTER AQUAFITNESS PROGRAM, LeFort Table 5: Uniwriate Tests of Significance for Selfand Psychologiil Well-Being Reported Pain, Di for Outcome Group (lb or N-RTW) and Time of Asserrsment (Bese-Line and 8 Weeks) hivariate Group by time Pain rating Disability Self-esteem State anxiety Affect Vigor Group Pain rating Disability Self-esteem State anxiety Affect Vigor Time Pain rating Disability Self-esteem State anxiety Affect Vigor
F (1,38)
P
4.12 2.73 5.93 1.79 0.37 5.52
.O49* .107 .020* .189 .547 .024*
8.43 10.45 1.41 4.54 0.87 7.53
.006+ .003+ ,243 .040* .356 .009+
11.06 7.30 3.43 3.28 0.08 0.57
.002+ .010+ ,072 ,078 ,775 ,455
*p 5 .05. +JJ 5 .Ol.
were 2.5 times more likely to return to work compared with subjects with a repeat injury. In addition, the results suggest that with each unit increase on the self-esteem scale (a deterioration in self-esteem), there was a 0.80 decrease in the odds of returning to work. DISCUSSION The results of this intervention study must be evaluated within the limitations of the design. The lack of a control group and the relatively small sample size limits both the generalizability of the results and the conclusions that can be directly attributed to the treatment program. Therefore, these findings are best considered preliminary or suggestive in nature. Within these limitations however, the results indicate several interesting observations. First, the analysis of the physical fitness measures indicates that, regardless of subsequent RTW status, low back injured subjects who participated in this aquafitness and muscle-strengthening program were not significantly different in terms of physical fitness achievement by 8 weeks. Both groups (RTW and NRTW) had significantly decreased their resting and 2minute recovery pulse rates, increased their work intensity during exercise, and increased their muscle strength and endurance. These results support findings of earlier studies that investigated fitness benefits for Workers’ Compensation recipients with a variety of musculoskeletal conditions, who were enrolled in the same rehabilitation program described in this article. “-‘O These findings also add to the growing body of evidence that low back injured individuals are able to improve their general mobility and fitness despite their perceived pain and disability.7*8*“*37‘41 Finally, these preliminary results suggest that aqualit exercise may be a safe and effective alternative or adjunct to land-based exercise for the low
1253
back injured and consequently deserves further investigation as a treatment modality. Of special interest are the findings suggesting that despite comparable levels of physical fitness achievement, improvement in pain, disability, and psychological well-being occurred only for those who subsequently returned to work. On average, those who did not return to work showed either no significant improvement or actually deteriorated on these measures during the 8 weeks in the program. Studies have documented a strong association between chronic back pain and lower levels of psychological well-being, including lower self-esteem,42‘45 higher state anxiety,36*46S47 and lower levels of mood especially vigor.36739Y48 Although studies of both healthy and chronically ill populations report that these variables tend to improve with participation in physical training, mechanisms of action are not fully understood.2’-23*49-51 In addition, how a pain problem might modify the association between exercise and psychological well-being has not been examined. In light of these unresolved issues, one possible interpretation of the results of this study might be that pain is mediating the responsiveness of these psychological variables to exercise (ie, only if pain is reduced during the course of an exercise program will the psychological benefits of exercise be experienced). For example, for some, but not all, low back injured subjects (ie, for most first injury compared with repeat injury subjects), exercise improves circulation to the site of injury, promotes healing, reduces muscle spasm and tension, and along with beta-endorphin release, results in a decreased perception of pain. Whether by physiological, psychological, and/or social processes, the alleviation of pain works together with exercise participation to enhance overall psychological well-being. However, when pain and disability persist despite increased efforts to improve fitness, psychological well-being deteriorates over time. This scenario may be exacerbated by other psychosocial processes such as social comparison.48 Those whose pain is improving may begin to feel better, to experience increased vigor and less anxiety, and report these positive changes to their exercise companions. Correspondingly, those who feel no better become more alarmed, wonder why their pain persists, continue feeling anxious, and begin to feel worse about themselves. As appealing as this interpretation might be, there may also be a case to support the alternative proposition that selfreported pain and disability tend to improve only when some of the psychological benefits of exercise are experienced. Recent findings on pain perception and stress-induced changes in the immune system that are associated with such variables as low self-esteem and perceived helplessness/ hopelessness need further investigation to test these relationships.52 The potential importance of the psychological benefits of exercise programs for the low back injured is highlighted by the results of the analyses of predictors of early retumto-work. Change in self-esteem was one of two predictor variables for this group of 40 subjects. Although not statistically significant, deterioration in self-esteem may be potentially clinically important and warrants further evaluation Arch Phys Med Rehabil Vol75, November 1994
RETURN TO WORK AFTER AQUAFITNESS PROGRAM, LeFort
and testing as an outcome of active rehabilitation programs for the LBP population.43 The finding that those with a repeat injury were at significantly greater risk of not returning to work by 2 months post treatment may suggest physiological differences in healing, or may indicate different expectations, beliefs, or fears in the individual as a result of having sustained a reinjury. Interestingly, this was only the second injury for most individuals in this group suggesting that poor response to treatment (ie, entrenched chronicity) may occur much earlier than commonly thought. Overall, this combined aquafit and muscle-strengthening intervention seems to have benefitted study subjects as measured by physical fitness indices. It did not reduce pain and disability ratings for all subjects however. Subjects whose pain did not decrease were likely to have had a repeat injury and had not returned to work or retraining. This finding may indicate the need for additional interventions targeted specifically to the repeat injury group to maximize treatment benefits of active rehabilitation programs. Acknowledgments: We wish to thank Lorraine Vardy, DPT, Dan Mosher, and Gail Innes for their assistance and support during this study. The authors gratefully acknowledge the contributions of Carey Norman, Robin Atkinson and others at the Lifestyles Program for data collection of the physical fitness measures. Thanks are extended also to Mary Ellen Jeans, PhD, RN, School of Nursing, McGill University, Montreal and Bonnie Stevens, PhD, RN, Faculty of Nursing, University of Toronto for critiquing earlier drafts of this article. References 1. Nachemson AL. Advances in low-back pain. Clin Grthop 1985; 2. Andersson G, Pope M, Frymoyer J, Snook S. Epidemiology and cost. In: Pooe M. Andersson G. Frvmover J. Chaffin D, editors. Occupational low back p’ain. St. Louis; M&by Year Book, 1991:95-113. _ 3. Mitchell R, Carmen G. Results of a multicenter trial using an intensive active exercise program for the treatment of acute soft tissue and back injuries. Spine 1990; 15:514-21. 4. BattiC M. Aerobic fitness and its measurement. Spine 1991; 16:677-8. 5. Frymoyer J. Back pain and sciatica. N Engl J Med 1988;318:291-300. 6. Nutter P. Aerobic exercise in the treatment and prevention of low back pain. &cup Med: State Art Rev 1988;3:137-45. 7. Mayer T, Gatchell R, Mayer H, Kishino N, Keeley J, Mooney V. A prospective two-year study of functional restoration in industrial low back injury. An objective assessment procedure. JAMA 1987; 258:1763-7. 8. Altmaier E, Lehmann T, Russell D, Weinstein J, Kao C. The effectiveness of psychological interventions for the rehabilitation of low back pain: a randomized controlled trial evaluation. Pain 1992;49:329-35. 9. Beekman E, Axtell L. Ambulation, activity level and pain. Outcomes of a program for spinal pain. Phys Ther 1985;65:1649-55. 10. Fredrickson B, Trief P, VanBeveren P, Yuan H, Baum G. Rehabilitation of the patient with chronic back pain. A search for outcome predictors. Spine 1988;13:351-3. 11. Hazard R, Fenwick J, Kalisch S, Redmond J, Reeves V, Reid S, et al. Functional restoration with behavioral support. A one-year prospective study of patients with chronic low-back pain. Spine 1989;14:157-61. 12. Lindstrom I, Ohlund C, Eek C, Wallin L, Peterson L, Nachemson A. Mobility, strength, and fitness after a graded activity program for patients with subacute low back pain. A randomized prospective clinical study with a behavioral therapy approach. Spine 1992;17:641-9. 13. Maver T. Gatchell R. Kishino N. Keelev J, Caura P, Maver H, et al. Objective assessment of spine function following indus&l injury. A prospective study with comparison group and one-year follow-up. Spine 1985; 10:482-93. 14. Oland G, Tveiten GA. Trial of modem rehabilitation for chronic lowback pain and disability. Vocational outcome and effect of pain modulation. Spine 1991; 16:457-9.
Arch whys Med Rehebil Vol75,
November 1994
15. Cassady S, Nielsen D. Cardiorespiratory responses of healthy subjects to calisthenics performed on land versus in water. Phys Ther 1992;72:532-8. 16. Whitley J, Schoene L. Comparison of heart rate responses. Water walking versus treadmill walking. Phys Ther 1987;67:1501-4. 17. Hannah TE, Hannah E, Mosher D, Vardy L. Fitness benefits of a lifestyles program for the rehabilitation of workers’ compensation recipients: effects of heartrate, muscle strength, and weight. J Rehabil 1988;54:37-41. 18. Hannah TE, Hannah E, Mosher D, Vardy L. Psychological benefits of a lifestyles modification program for workers’ compensation recipients. Physiother Can 1988;40:10-5. 19. Hannah TE, Kozma A, Stones M, Mosher D, Vardy L. Effects on mood of a lifestyles program for the rehabilitation of injured workers. J Gccup Med 1989;31:454-7. 20. Hannah TE, Hannah E, Mosher D, Vardy L. Confirmation of the physical fitness benefits of a lifestyle program for injured workers. Can J Rehabil 1989;2:163-8. 21. LaFontaine T, DiLorenzo T, Frensch P, Stucky-Kopp R, Bargman E, McDonald D. Aerobic exercise and mood. A brief review, 1985-1990. Sports Med 1992; 13:160-70. 22. Petmzzello S, Landers D, Hatfield B, Kubitz K, Salazar W. A_ _____memanalysis on the anxiety-reducing effects of acute and chronic exercise. Outcomes and mechanisms. Sports Med 1991: 11: 143-82. 23. Raglin J. Exercise and mental health. Beneficial and detrimental effects. Sports Med 1990;9:323-9. 24. Astrand P, Rodahl K. Textbook of work physiology: physiological basis of exercise. New York: McGraw Hill, 1986. 25. Melzack R. The McGill pain questionnaire.: major properties and scar ing methods. Pain 1975; 1:277-99. 26. Melzack R, Katz J. The McGill pain questionnaire: appraisal and current status. In: Turk D, Melzack R, editors. Handbook of pain assessment. New York Guilford, 1992:152-68. 27. Fairbank J, Couper J, Davies J, O’Brien J. The Oswestry low back disability questionnaire, Physiother 1980;66:271-3. 28. Rosenberg M. Society and the adolescent self-image. Princeton: Princeton University Press, 1965. 29. Duffy ME. Determinants of health status in employed women. Health Values 1989; 13:50-7. 30. Robinson J, Shaver P. Measures of social psychological attitudes. Ann Arbor: Institute of Social Research, University of Michigan, 1973. 31. Spielberger C, Gorsuch R, Lushene R, Vagg P, Jacobs G. Manual for the state-trait anxiety inventory. STAI (Form Y). Palo Alto: Consulting Psychologist Press, 1983. 32. McNeil J. Mood: measurement, diurnal variation and age effects [dissertation]. St. John’s (NF): Memorial University of Newfoundland, 1986. 33. Tabachnick B, Fidel1 L. Using multivariate statistics. New York: Haroer Collins, 1989. 34. Spitzer WO, LeBlanc FE, Dupuis M. Scientific approach to the assessment and management of activity-related spinal disorders. A monograph for clinicians. Spine 1987; 12:Sl-59. 35. McArdle W, Katch F, Katch V. Exercise physiology. Energy, nutrition and human performance. Philadelphia: Lea & Febiger, 1986:358-9. 36. Fuerstein M, Carter R, Papciak A. A prospective analysis of stress and fatigue in recurrent low back pain. Pain 1987;31:333-44. 37. Dolce J, Cracker M, Moletteire C, Doleys D. Exercise quotas, anticipatory concern and self-efficacy expectancies in chronic pain: a preliminary report. Pain 1986;24:365-72. 38. Fordyce W, McMahon R, Rainwater G, Jackins S, Quested K, Murphy T, et al. Pain complaint-exercise performance relationship in chronic pain. Pain 1981;10:311-21. 39. Kleinke C, Spangler A. Predicting treatment outcomes of chronic back pain patients in a multidisciplinary pain clinic: methodological issues and treatment implications. Pain 1988:33:41-8. 40. Linton SJ, Bradley LA, Jensen I, Spangfort E, Sundell L. The secondary prevention of low back pain: a controlled study with follow-up. Pain 1989;36:197-207. 41. Manniche C, Lundberg E, Christensen I, Bentzen L, Hesselsoe G. Intensive dynamic back exercises for chronic low back pain: a clinical trial. Pain 1991;47:53-63. 42. Armentrout D. The impact of chronic pain on self-concept. J Clin Psycho1 1979;35:517-21.
RETURN TO WORK AFTER AQUAFITNESS PROGRAM, LeFort 43. Beekman E, Axtell L, Noland K, West J. Self-concept: an outcome of a program for spinal pain. Pain 1985;22:59-66. 44. Coates R, Femmi P. Sexual dy&nction and marital d&harmony as a comequence of chtonic lumbar spinal pain. Sex Ma&al l&r 1991$X5-9. 45. Vikari-Juntura E, Vuori J, Silverstein B, Kalimo R, Kuosma E, Videman T. A life-long prospective study on the role of psychosocial factors in neck-shoulder and low-back pain. Spine 1991; 16:1056-61. 46. Ackerman M, Stevens M. Acute and chronic pain: pain dimensions and psychological status. J Clin Psycho1 1989;45:223-8. 41. Garron D, Leavitt F. Chronic low back pain and depression. J Clin Psycho1 1983;39:486-93.
12%
48. Mechanic D, Angel R. Some factors associated with the report and evaluation of back pain. J Health Sot Behav 1987;28:131-9. 49. Phelps JR. Physical activity and health maintenance. Exactly what is known? West J Med 1987; 146:200-6. 50 Sonstroem R, Morgan W. Exercise and self-esteem: rationale and ’ model. Med Sci Sports Exert 1989;21:329-37. 51. Wilfley D, Kunce J. Differential physical and psychological effects of exercise. J Counsel Psycho1 1986;33:337-42. 52. Marbach J, Schleifer S, Keller S. Facial pain, distress, and immune function. Brain Behav Immun 1990;4:243-54.
NOTICE TO AUTHORS The Archives of Physical Medicine and Rehabilitation is changing its address effective December 10, 1994. Manuscripts and correspondence should be addressed to: Editorial Board Archives of Physical Medicine and Rehabilitation Suite 2510 330 North Wabash Avenue Chicago, IL 6061 l-3604 Telephone (3 12) 464-9550
Return to Work Following an Aquafitness and Muscle Strengthening Program for the Low Back Injured Sandra M. LeFort, PhD, RN, T. Edward
Hannah,
PhD
ABSTRACT. LeFort SM, Hannah TE. Return to work following an aquafitness and muscle strengthening program for the low back iqjured. Arch Phys Med Rehabil 1994;75:1247-55. a Our purpose, in thii prospective clinical study, was to identify the best predictors of 2-month return to work or retraining for a group of low back injured subjects (n = 40) who completed at least 8 weeks of a communitybased rehabilitation program that combined aerobic and flexibiity exercise conducted in the water (aquafitness) with muscle strength and endurance training. Baselhte demographic characteristics and changes in physical fitness, pain, disability, and psychological well-being during the course of program participation were compared between two groups of low back injured subjects: those who returned to work (RTW) [n = 241, and those who did not (N-RTW) [n = 161. Subjects in both groups showed comparable improvement in measures of physical fitness at 8 weeks. However, multivariate analyses showed significant between-group differences in self-report measures. The RTW group showed sign&ant improvement in measures of pain, disability, anxiety, and vigor while self-esteem and a&& remained stable. The N-RTW group displayed no change in pain and disability variables and had sigrdticant deterioration in mean overall psychological well-being over time. The best predictors of return to work using logistic regression analyses were a first injury rather than a repeat injury to the lower back, and stability in self-esteem. Suggestions are offered for further research to examine the benefits of aqua& exercise for the low back injured, for additional interventions for those with a reinjury, and for maintaining or enhancing self-esteem as a treatment goal. 0 1994 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation
Low back pain (LBP), the so-called ‘ ‘nemesis of medicine and the albatross of industry,” is of increasing concern to society.’ Studies indicate that LBP and its related disability is of epidemic proportions particularly in the working population. In both Canada and the United States, the total number of reported job-related back injuries, the average time off work, and the amount awarded in compensation payments continue to increase at an alarming rate.2V3Similarly, health care costs associated with this condition are increasing. This escalating situation has led to expanded research and clinical interest in finding effective approaches to the rehabilitation of the low back injured that will return them to productive employment. One approach receiving increased attention is the role of activity, exercise, and aerobic fitness in comprehensive treatment programs for both subacute and chronic LBP.“-6 A basic premise of these programs is that low back injured individuals become deconditioned as a result of excessive protection and disuse of spinal joints and muscles, as well as general inactivity.’ Movement leading to improved fitness is thought to provide stability to joints and injured ligaments, to overcome muscle spasm and tension, and to reduce pain.3 Treatment programs that emphasize active rehabilitation and From the School of Nuning, McGill University; and the Department of Psychology and Division of Community Medicine, Memorial University of Newfoundland, Canada. Submitted for publication December 10, 1993. Accepted in revised form May 5, 1994. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the authors or upon any ornanization with which the authors are associated. Reprint requests to Sandra M. LeFort, PhD (Cand), RN, 111 Strawberry Marsh Road, St. John’s, Newfoundland, AlB 2V7, Canada. 8 1994 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation ooO3-9993/94/7511-2537$3.00/O
physical conditioning have been reported to significantly enhance cardiorespiratory fitness; increase muscular strength and endurance; improve flexibility; decrease perceptions of pain, depression, and disability; im$r;ve functional status; and, facilitate early return to work. However, despite these successes a significant subset of individuals with LBP who participate in physical conditioning programs do not return to work within the expected time if they return at a11.i4To date, few research reports have directly compared the physical and psychological outcomes of participation in exercise programs between LBP patients who subsequently return to work and those who do not. Such comparisons may be particularly valuable in identifying the subset of individuals for whom exercise as a treatment modality provides maximum benefits and in discriminating which articular benefits of exercise lead to positive outcomes. B This study reports on an 8- to 1Zweek community-based rehabilitation program that was designed as an adjunct to traditional physical therapy treatment. The program combines aerobic and flexibility exercise conducted in the water (aquafit exercise as distinct from swimming), with muscle strengthening and endurance exercises on weight-training machines. Aquafit aerobic exercise programs are now considered corn arable with land exercise in cardiorespiratory outcomes.“* P6 Such programs provide a safe alternative to land exercise especially for a deconditioned population such as those with LBP. To our knowledge however, no other studies of aquafit exercise in the rehabilitation of LBP patients have been reported with the exception of previous reports by Hannah and colleagues.‘7~20 This prospective study focused on low back injured individuals as they progressed through this aquafit and muscle strengthening program and compared two groups of participants: those who returned to work or began retraining two Arch Phys Med Rehabil Vol75, November 1994
1248
RETURN TO WORK AFTER AQUAFITNESS PROGRAM, LeFort
months after program completion (RTW group); and those who did not return to work (N-RTW group). A particular interest of the investigators was whether improvements in psychological well-being that are associated with physical exercise in the healthy population (ie, improved self-esteem, lowered state anxiety, and improved mood especially vigor) would also be experienced by the low back injured subjects in this study.21-23In our review of the literature, no studies were found that investigated the impact of physical conditioning in the LBP population on this set of variables, nor of the potential significance of these variables on early return to work. Three specific research questions were addressed: (1) Were there significant differences in physical fitness achievement between the RTW group and the N-RTW group over 8 weeks in the program? (2) Were there significant differences in self-reported pain, disability, and psychological well-being between the RTW group and the N-RTW group over 8 weeks in the program? (3) What were the best predictors of 2-month RTW or retraining for this group of subjects?
METHOD Subjects The study sample comprised individuals consecutively admitted over an g-month period to a rehabilitation program located in a fitness facility in a medium-sized Canadian city. Individuals were included if they met the following criteria: a primary diagnosis of idiopathic or mechanical LBP with no current evidence of herniated disc, fracture, or neurological complication; referred by a physical therapist and/or physician; and, receiving standard physical therapy treatment. Standard treatment included one or more of the following: transcutaneous-electrical nerve stimulation (TENS), ultrasound, heat/cold treatments, massage, and stretching/extension exercises. None of the subjects was receiving any additional psychological or structured education interventions. The vast majority of subjects were referred to the program from three physical therapy clinics. The caseloads of these clinics included virtually all of the Workers’ Compensation recipients in the city and surrounding area who were recovering from work-related injuries amenable to physical therapy. The program, initially funded by the provincial Worker’s Compensation Commission, was considered part of the rehabilitation of work-injured patients. Thus, we are reasonably certain that subjects are representative of low back injured individuals who were seeking treatment at the time of data collection.
Rehabilitation Program Description The program consists of two supervised components: aquafn classes combining aerobic and flexibility exercises, and muscle-strengthening workouts on training equipment for primary muscle groups, such as hip/lower back, quadriceps, hamstrings, abductorladductor, trapezius, latissimus dorsi, pectoralis major, deltoids, abdominal, biceps, and triceps. Normally, subjects attended two aquafit classes and three muscle-strengthening sessions per week. The individualized nature of the program, however, meant that some Arch Phys Med Rehabll Vol75,
November 1984
subjects did not begin muscle-strength training until the second or third week. All subjects were monitored biweekly by a physical therapist or physician. The aquafit component of the program was conducted in a swimming pool in water waist-deep to shoulder-height. Those who feared the water were assured that the program need not include swimming and were introduced to the water slowly. Ordinarily, 10 to 12 subjects participated in a class with a minimum of two rehabilitation instructors in attendance. Each aquafit session included a lo-minute warm-up, a 25-minute aerobics workout consisting of variations of walking and jogging lengths in the pool, a 20minute series of specific exercises (range of motion, flexibility, and stretching), and ended with a 5-minute cool-down. Subjects progressively increased their speed, number of repetitions, and level of exercise in a programmed fashion week by week. The longer subjects were in the program, the more the instructors encouraged them to work harder and longer. The muscle-strengthening program consisted of 14 possible exercises on 11 weight-training machines. Subjects began by lifting light weights, with a gradual increase of repetitions to 20, before adding extra weight. Subjects were consistently monitored by the rehabilitation staff throughout their weight-lifting sessions for correct warm-up (lo-minutes on a stationary bicycle), correct body alignment while using the machines, and appropriate increases in weights. Although most individuals used all machines, not all of them did so. Exercises were prescribed by their physical therapists based on the nature of the injury and the subject’s stage of recovery. For both the aquafit and muscle-strengthening components of the program, subjects were cautioned not to continue a particular exercise if pain increased significantly. The consultant physical therapist was notified in such cases. Other aspects of the program have been described elsewhere.“-*’
Measures The following two sets of measures were taken: indices of physical fitness; and self-report of demographic characteristics, pain, disability, and psychological well-being.
Physical Fitness Measures Pulse rates. Resting pulse rate before exercise and 2minute recovery pulse rate after aerobic exercise were used as crude indicators of change in aerobic fitness. The pulse rate taken during aerobic exercise was used as an index of change in work intensity.24 Subjects’ pulse rates were taken by palpating the radial pulse for a full 60 seconds and counting the beats per minute (bpm) using a wall clock. Pulse rates were monitored at the first complete aquafit session and monthly thereafter by trained program staff. Five heartrate measures were taken at each monitored session. Resting pulse rate was taken after subjects sat by the edge of the pool for 3 to 4 minutes before the aquafit class. They then entered the pool and participated in the warm-up and aerobics section of the aquafit class. During the final third of the aerobics section, following a specified exercise, the duringexercise pulse rate was measured. At the end of the final
RETURN TO WORK AFTER AQUAFITNESS PROGRAM, Lefort
aerobic exercise, the subject was asked to stop and the immediately-after pulse rate was taken followed by the l-minuteafter and 2-minute-after recovery pulse rates. Total weights &led. Total weights lifted on the training machines at the first versus the 15th session served as the index of muscular strength and endurance. Fifteen sessions represented between 6 to 8 weeks of muscle-strength training. At each weight-lifting session, the amount of weight lifted on each machine was recorded. These data were reviewed on an ongoing basis by the program staff.
Self-Report Measures General background information. Demographic characteristics, current treatment, history of the low back injury, and other pain-related information were collected through a structured patient interview. A second source of data were the referral notes from physicians and physical therapists. Pain. The Pain. Rating Index (PRI) of the McGill Pain Questionnaire (MPQ) was used to assess the subjective experience of painz5 The PRI consists of a list of 20 sets of words that describe pain quality and intensity. Subjects were asked to select the word sets that best described their pain in the past 24 hours and then to choose the most appropriate word within each selected set. The reliability and validity of the MPQ has been widely demonstrated and it remains the most commonly used measure of self-reported pain.26 Disability. Perceived functional impairment was measured using the Oswestry Low Back Pain Disability Questionnaire.27 It contains 10 sections on different activities of daily living relevant to the individual with LBP. Scores range from 0% to 100% disability. A 2-day test-retest reliability of r = 0.99 with good internal consistency has been reported. A number of investigators11*‘3~27 have found that scores are correlated to improvement in LBP patients during rehabilitation. Self-esteem. Self-esteem was measured using the Rosenberg Self-Esteem Scale, a IO-item Like&type format.28 The scale is regarded as a unidimensional measure of the self-acceptance aspect of self-esteem and all items revolve around liking or approving of the self. Scores range from 10 to 40, with lower scores indicating higher self-esteem. A 2-week test-retest reliability of r = 0.85 is reported; data on convergent, discriminant, and predictive validity have been described.29*30 Anxiety. The 20-item state portion of the State/Trait Anxiety Inventory (STAB, Form Yl was used as a measure of situational anxiety.31 Stability coefficients for the state anxiety scale are relatively low (r = 0.16 to 0.62), whereas internal consistency is high (Cronbach (Y= 0.92). Extensive evidence of the concurrent, convergent, divergent, and construct validity of the STAI scale is provided by the tool’s developers.31 Mood. Mood was measured using the 23-item version of the Memorial University Mood Scale (MUMS). Subjects are asked to answer “yes,” “no,” or “don’t know” to 7 positive and 7 negative mood adjectives, and 9 vigor adjectives.32 The measure consists of two subscales, affect and vigor, which are reported to be affected by participation in exercise.*9.32The scale has high internal consistency (Cronbath (Y = 0.80), and relatively low stability (r < 0.50 over
1249
3 days; r < 0.30 over 2 years), as expected for a valid measure of short-term mood. Construct validity of the tool has also been established.32
Procedure The study protocol and measurement instruments received ethical approval from a university-based Human Subjects Review Committee. Consenting subjects were interviewed by the first author a few days before beginning the rehabilitation program and monthly thereafter to collect demographic and other self-report data. The measures of physical fitness were collected by three trained rehabilitation program staff as part of standard record keeping. Pulse rates were measured at the first aquafit session and monthly thereafter. Weights lifted were recorded at each weight-lifting session. Only after, all subjects were discharged from the program did the first author gain access to the physical fitness data. Data on return to work or retraining were collected by questionnaire mailed to all subjects 2 months after completing the fitness program. Subjects were asked if they had returned to either full-time or part-time work or had started retraining. If they had not gone back to work, they were asked to explain the reasons for not returning and to indicate when they expected to be able to do so. If the survey was not returned within 2 weeks of mailing, subjects were contacted by the first author, and questions were answered over the telephone.
Statistics Data analyses were carried out in three steps using the Statistical Package for the Social Sciences (SPSSx). First, equivalence of groups on demographic, injury-related characteristics, and baseline measures of pain, disability and psychological well-being was assessed by means of &i-square analyses and 2-tailed Student t tests for independent groups. Second, two separate 2 X 2 repeated measures multivariate analysis of variance (RM MANOVA) procedures were conducted. Variables in the first RM MANOVA consisted of the physical fitness indicators: resting pulse rate, duringexercise pulse rate, 2-minute-after pulse rate, and total weights lifted on weight-training machines. The five selfreport measures (pain, disability, self-esteem, anxiety, and the affect and vigor subscales of mood) were included in the second RM MANOVA. For each RM MANOVA, outcome group (RTW versus N-RTW) served as the between-subjects factor, whereas time of assessment (baseline versus final assessment) served as the repeated within-subjects factor. The multivariate approach to repeated measures was used in these analyses for two reasons: to account for the interdependency of measures at the two assessment periods, and to assess how combinations of dependent variables changed together over time. The multivariate results are reported using Pillais’ F value because this criterion is reported to be the most robust with unequal sample size.33 Univariate results that contribute to significant multivariate effects are also reported. The univariate results are equivalent to 2 x 2 mixed model analysis of variance (ANOVAs) of the individual variables. To identify the best predictors of return to work, logistic Arch Phys Mad Fkhabil Vol75,
November 1904
1250
RETURN TO WORK AFTER AQUAFITNESS PROGRAM, LeFort
Table 1: Reasons Subjects (n = 10) Were Excluded From the Study Dropouts from program before 4 weeks Increased back pain Moved to another city Accident (unrelated to the program) Private-paying subjects decided that the program was not for them Returned to work after 3 weeks Completed program but excluded from the study Loss to follow-up Not eligible for work (maternity leave) Already working full-time while attending program
1 1 2
regression analyses were applied using a forward stepwise procedure. Return to work or retraining (0 = yes; 1 = no) served as the dependent variable. Because of the small sample size, predictor variables were limited to four. Only those variables that differed significantly between the RTW and the N-RTW groups were allowed into the equation. For each variable in the final model, odds ratios and 95% confidence intervals are reported. For all analyses, the 0.05 alpha level was set to determine statistical significance.
subjects took regular medication for pain, most commonly nonnarcotic analgesics and muscle relaxants. Subjects were employed in a wide variety of occupations, the most common classifications being: clerical and sales (n = 8); service (n = 8); scientific/technical (nursing and teaching) (n = 7); transportation/handling (n = 7); production/processing (n = 5); and natural resources (n = 5). Eighty-eight percent of subjects (n = 35) were receiving some kind of disability benefit, and no one was in litigation. To assess the equivalence of the RTW group compared with the N-RTW group, demographic and other injury-related characteristics were compared (table 2). Whether subjects were experiencing a first injury or a repeat injury was the only background characteristic that differed significantly between groups [x2 (1, n = 40) = 6.9, p < .Ol]. Thirty-three percent (n = 8) of RTW subjects versus 75% of N-RTW subjects (n = 12) stated that their current back injury was a repeat injury. Most subjects reporting a repeat injury had experienced only one other significant episode of LBP; however, 3 subjects Table 2: Characteristics at Program Entry of the RTW Group, the N-RTW Group, and Those Excluded From the Study Included
RESULTS
Subject Characteristics Fifty of 52 eligible subjects consented to participate. Of these, 10 subjects had to be excluded from the study for a variety of reasons (table 1). The remaining 40 subjects were not working because of their injury, but had full- or parttime jobs waiting for them. This group completed at least 8 weeks in the program and were able to be recontacted for assessment of return-to-work or retraining status. Two months after program completion, 60% (n = 24) of the subjects had returned to work (n = 23) or retraining (n = 1), whereas 40% (n = 16) had not. Reasons cited for not returning to work were continuing back pain (n = 14), or awaiting further medical investigation (n = 2). To test the equivalence of the 40 subjects included in the study compared with the 10 excluded subjects, betweengroup comparisons were made of demographic and injuryrelated characteristics (table 2), and baseline measures of pain, disability, and psychological well-being (table 3). Radiation of pain was the only variable that differed significantly between the included and the excluded group [x2 (1, n = 50) = 5.89, p < .02]. All excluded subjects reported that their pain was localized in the lower back and did not radiate down the leg, whereas 40% (n = 16) of the included group reported pain radiated down the leg. Radiating pain may indicate a more severe low back problem.34 As a group, the excluded subjects may have had less deep structure or neurogenic involvement than those included in the study. The 40 subjects included consisted of 25 men and 15 women whose ages ranged from 21 to 57 years (mean + SD = 37.1 ? 9.28). Seventy-eight percent (n = 31) were married or partnered, but no one lived alone. Only 13% (n = 5) had experienced pain for less than 3 months, thus 87% (n = 35) could be classified as having chronic LBP, with seven having had previous back surgery. Less than half of the Arch Phys Mad Rehabil Vol75,
November 1994
Variable
Excluded
RTW (n = 24)
N-RTW (n = 16)
Total (n = 40)
Total (n = 10)
37.6 9.0
36.0 10.1
36.9 9.4
35.4 9.6
16 (67%) a (33%)
9 (56%) 7 (44%)
25 (63%) 15 (37%)
6 (60%) 4 (40%)
20 (83%) 4 (17%)
11 (69%) 5 (31%)
31 (78%) 9 (22%)
8 (80%) 2 (20%)
16 (67%)* 8 (33%)*
4 (25%)* 12 (75%)*
20 (50%) 20 (50%)
4 (40%) 6 (60%)
(17%) (29%) (33%) (21%)
1 (6%) 8 (50%) 5 (31%) 2 (13%)
5 (13%) 15 (38%) 13 (32%) 7 (17%)
3 (30%) 3 (30%) 4 (40%)
17 (71%) 2 (8%) 5 (21%)
7 (44%) 4 (25%) 5 (31%)
24 (60%)+ 6 (15%)+ 10 (25%)+
8 (33%) 16 (67%)
9 (56%) 7 (44%)
17 (42%) 23 (58%)
4 (40%) 6 (60%)
6 (25%) 18 (75%)
1 (6%) 15 (94%)
7 (18%) 33 (82%)
3 (30%) I (70%)
20 (83%) 4 (17%)
11 (69%) 5 (31%)
31 (78%) 9 (22%)
6 (60%) 4 (40%)
21 (88%) 3 (12%)
14 (88%) 2 (12%)
35 (88%) 5 (12%)
6 (60%) 4 (40%)
Age (years) Mean SD Gender Men Women Marital status Married/partnered Single First or repeat injury First Repeat Pain duration <3 months 3-6 months 6 12 months > 12 mouths Radiation of pain None Above knee Below knee Pain medication Yes No Previous back surgery Yes No Program status WCB-sponsored* Private-paying Disability income” Yes No
4 7 8 5
*p s .Ol +p c .02 t Worker’s Compensation Board paying for cost of the rehabilitation proQ”. Includes both Worker’s Compensation payments and private insurance income.
RETURN
Table 3: Base-Line Scores of Self-Reported Pain, DisabMty, and Paychokqical Well-Being for Those Included and Those Excluded From the Study Excluded (II = 10)
Included (n = 40)
1251
TO WORK AFTER AQUAFITNESS PROGRAM, LeFort
Variable
Mean
SD
Mean
SD
Pain rating Disability Self-esteem State anxiety Affect Vigor
21.23 37.85 17.65 40.55 6.55 7.08
8.98 13.12 4.81 10.47 6.41 5.62
23.03 31.50 15.83 37.54 7.94 6.12
7.41 12.34 4.47 12.46 6.69 3.67
(12.5%) in the RTW group and 5 subjects (31%) in the NRTW group reported that they had suffered three or more previous injuries affecting the lower back. To further assess the equivalence of these two groups at program entry, baseline measures of pain, disability, and psychological well-being were compared (table 4). No significant between-group differences were found on any of these dependent variables. All p values were greater than 0.10. In summary, baseline characteristics of subjects who returned to work were not significantly different from the characteristics of those who did not return to work with the exception of history of prior back injury. Those in the RTW group were more likely to be experiencing a first back injury rather than a repeat back injury.
Physical Fitness To assess between-group differences in change in physical fitness over time, scores of resting pulse rate, during-exercise pulse rate, 2-minute recovery pulse rate, as well as total weights lifted were analyzed using 2 X 2 RM MANOVA. Outcome group (RTW versus N-RTW) served as the between-subjects factor and time of assessment (baseline versus final assessment) served as the repeated within-subjects factor. Final assessment for pulse rates was the I&week measure, whereas the 15th session was the final assessment for total weights lifted. Data from only 36 subjects were included in this analysis. Four subjects (two from each outcome group) were lost to analysis because of incomplete data sets on the physical fitness variables. From a review of the physical fitness records, missing data seemed to be a random occurrence. Hence, it is unlikely that dropping these four subjects affected the representativeness of the sample. A significant multivariate main effect for time (F(4,31) = 46.46, p = .OOO)was found with univariate results from all four dependent variables contributing to this effect: total weights lifted (F( 1,34) = 133.73, p = .OOO);during exercise pulse rate (F( 1,34) = 63.44 p = BOO); resting pulse rate (F(1,34) = 14.57, p = .OOl), and 2-minute recovery pulse rate (F( 1,34) = 4.7, p = .038). The multivariate main effect of outcome group (RTW versus N-RTW) was not significant (F(3,32) = 0.91 p = .47), nor was the group by time interaction (F(3,32) = 0.47, p = .76). For descriptive purposes, the means of pulse rates (broken down by outcome group and time of assessment) taken at the five monitored intervals during the aerobics component of aquafit class is presented in figure 1. After 8 weeks in
the program, subjects in both groups had, on average, decreased their mean resting pulse rate by 9bpm and decreased their mean postexercise 2minute recovery rate (RTW by 4bpm; N-RTW by Bbpm). Correspondingly, pulse rate during exercise increased substantially over time (RTW mean increase: 29bpm; N-RTW mean increase: 22bpm) indicating that subjects were working to a greater intensity by 8 weeks in the program and were probably elevating their heart rates into the cardiac training sensitive zone (70% of maximal heart rate).35 For water exercise, the target heart rate needs to be adjusted from 13 to 20bpm lower than for dry land exercise to take account of the cooling effect of the water as well as the lessened gravitational pull on the heart.‘5S24 These significant results suggest mean improvement in cardiac fitness for subjects in both groups. In muscle strength and endurance, subjects in the RTW group had a mean percentage increase of 130% in total weights lifted on 11 strength-training machines, whereas the N-RTW group had a 134% increase from the first to the 15th session. Figure 2 shows the means and standard deviations of total weights lifted broken down by outcome group and time of assessment. In summary, no significant differences were found in mean physical fitness achievement between the RTW and N-RTW groups. On average, these subjects were improving their level of physical fitness over time despite their perceived level of pain and disability.
Self-Report of Pain, Disability, and Psychological Well-Being To assess between-group differences in change in selfreport measures over time, scores of pain, disability, selfesteem, state anxiety, and the affect and vigor subscales of mood were analyzed together using a 2 X 2 RM MANOVA (table 4). Results showed a significant multivariate group by time interaction (F(6,33) = 2.49, p = .042), a significant Table 4: Scores of Self-Reported Pain, Disability, and Psychological Well-Being by Outcome Group (RlW or NRTW) and Time of Assessment (Base-Line and 8 Weeks)
Variable (possible range of scores) Pain rating (0 to 78) Baseline 8 weeks Disability (0% to 100%) Baseline 8 weeks Self-esteem (40 to lo)* Baseline 8 weeks State anxiety (20 to 80) Baseline 8 weeks Affect (-14 to +14) Baseline 8 weeks Vigor (0 to 18) Baseline 8 weeks
RTW (n = 24)
N-RTW (n = 16)
Mean
SD
Mean
SD
20.00 10.71
9.71 7.86
23.06 20.81
7.66 9.92
34.29 24.25
12.47 13.64
39.19 37.25
12.60 12.17
17.67 17.25
5.49 5.73
17.63 20.69
3.72 3.72
38.63 34.46
10.27 11.02
43.44 42.81
10.42 9.64
7.08 7.96
6.57 7.71
5.75 5.44
7.07 6.78
7.96 11.00
5.93 6.69
5.75 4.19
4.50 4.01
* Higher score indicates lower self-esteem.
Arch Phys Mod Rehabil Vol76,
November 19Q4
RETURN TO WORK AFTER AQUAFITNESS PROGRAM, LeFort
1252
Predictors of Return To Work The logistic regression equation developed to predict RTW for this sample of 40 subjects included the following four variables: first or repeat back injury, and the change scores (baseline score minus 8.week score) of those variables that contributed to the significant multivariate group by time interaction, namely pain, self-esteem, and vigor. The final regression model contained two variables that were associated with return to work 2 months after discharge from the fitness program. For this study sample, the only significant predictor of return to work was first or repeat back injury (odds ratio [OR] = 2.5,95% confidence interval [CI] = 1.17 to 5.23,~ = .02). The next best predictor, which approached statistical significance, was self-esteem (OR = 0.80, 95% CI = 0.64 to 1.01, p = .059). In the N-RTW group, 62% (n = 10) reported worse self-esteem scores by week 8, whereas 38% (n = 6) reported the same or improved scores. By contrast, only 29% (n = 7) of the RTW-group reported worse scores, whereas 71% (n = 17) reported the same or improved scores. Mean change in self-esteem for those in the N-RTW group was 3.06 units + 5.79, and for the RTW group was -0.42 units + 3.24. In summary, those experiencing a first low back injury
80
350 70 RESTING
I DURING
I IMMED. AFTER
I I-MIN. AFTER
I 2.MIN. AFTER
Fig l-This figure compares pulse rates measured at 5 intervals during the aerobic component of aquafit class at two time periods, the first complete aquafit class (baseline) and 8 weeks. By 8 weeks, both the RTW group (n = 22) and the N-RTW group (n = 14) had decreased their resting pulse rate, increased during exercise pulse rate, and had a steeper slope of pulse rate recovery from immediately after exercise to l-minute and 2.minutes after aerobic exercise. These results suggest mean improvement in aerobic exercise for both groups. - n RTW Baseline; - A - N-RTW Baseline; - 4 - RTW 8 Weeks; - 0 - N-RTW 8 Weeks.
multivariate main effect of group (F(6,33) = 2.74, p = .028), and a significant multivariate main effect of time (F(6,33) = 3.37, p = .Ol 1). The univariate results contributing to the significant multivariate effects are reported in table 5. The significant multivariate group by time interaction indicates that the overall pattern of change in these variables during the 8 weeks in the program was different in the RTW group compared with the N-RTW group. Subjects who subsequently returned to work had a significant mean reduction in pain and disability as well as significant mean improvement in the vigor subscale of mood and state anxiety. Selfesteem and overall affect remained stable. Conversely, subjects in the N-RTW group, with no significant change in their pain or disability scores, significantly deteriorated in scores of self-esteem and vigor. Mean state anxiety for this group, which was elevated compared with published population norms (43 v 35) but consistent with other chronic LBP populations,31’36remained unchanged during the 8 weeks in the program as did overall affect. Arch Phys Med Rehabil Vol75, November 1994
300
v) 250 z d 8 200 $ z g 150 Q z L
100
50
0 1st SESSION
15th SESSION
Fig 2-This figure compares total weights lifted in kilograms by the RTW group (n = 22) and the N-RTW group (n = 14) on 11 weight-training machines at the 1st and 15th muscle strengthening and endurance sessions. S RTW q N-RTW.
RETURN TO WORK AFTER AQUAFITNESS PROGRAM, LeFort Table 5: Uniwriate Tests of Significance for Selfand Psychologiil Well-Being Reported Pain, Di for Outcome Group (lb or N-RTW) and Time of Asserrsment (Bese-Line and 8 Weeks) hivariate Group by time Pain rating Disability Self-esteem State anxiety Affect Vigor Group Pain rating Disability Self-esteem State anxiety Affect Vigor Time Pain rating Disability Self-esteem State anxiety Affect Vigor
F (1,38)
P
4.12 2.73 5.93 1.79 0.37 5.52
.O49* .107 .020* .189 .547 .024*
8.43 10.45 1.41 4.54 0.87 7.53
.006+ .003+ ,243 .040* .356 .009+
11.06 7.30 3.43 3.28 0.08 0.57
.002+ .010+ ,072 ,078 ,775 ,455
*p 5 .05. +JJ 5 .Ol.
were 2.5 times more likely to return to work compared with subjects with a repeat injury. In addition, the results suggest that with each unit increase on the self-esteem scale (a deterioration in self-esteem), there was a 0.80 decrease in the odds of returning to work. DISCUSSION The results of this intervention study must be evaluated within the limitations of the design. The lack of a control group and the relatively small sample size limits both the generalizability of the results and the conclusions that can be directly attributed to the treatment program. Therefore, these findings are best considered preliminary or suggestive in nature. Within these limitations however, the results indicate several interesting observations. First, the analysis of the physical fitness measures indicates that, regardless of subsequent RTW status, low back injured subjects who participated in this aquafitness and muscle-strengthening program were not significantly different in terms of physical fitness achievement by 8 weeks. Both groups (RTW and NRTW) had significantly decreased their resting and 2minute recovery pulse rates, increased their work intensity during exercise, and increased their muscle strength and endurance. These results support findings of earlier studies that investigated fitness benefits for Workers’ Compensation recipients with a variety of musculoskeletal conditions, who were enrolled in the same rehabilitation program described in this article. “-‘O These findings also add to the growing body of evidence that low back injured individuals are able to improve their general mobility and fitness despite their perceived pain and disability.7*8*“*37‘41 Finally, these preliminary results suggest that aqualit exercise may be a safe and effective alternative or adjunct to land-based exercise for the low
1253
back injured and consequently deserves further investigation as a treatment modality. Of special interest are the findings suggesting that despite comparable levels of physical fitness achievement, improvement in pain, disability, and psychological well-being occurred only for those who subsequently returned to work. On average, those who did not return to work showed either no significant improvement or actually deteriorated on these measures during the 8 weeks in the program. Studies have documented a strong association between chronic back pain and lower levels of psychological well-being, including lower self-esteem,42‘45 higher state anxiety,36*46S47 and lower levels of mood especially vigor.36739Y48 Although studies of both healthy and chronically ill populations report that these variables tend to improve with participation in physical training, mechanisms of action are not fully understood.2’-23*49-51 In addition, how a pain problem might modify the association between exercise and psychological well-being has not been examined. In light of these unresolved issues, one possible interpretation of the results of this study might be that pain is mediating the responsiveness of these psychological variables to exercise (ie, only if pain is reduced during the course of an exercise program will the psychological benefits of exercise be experienced). For example, for some, but not all, low back injured subjects (ie, for most first injury compared with repeat injury subjects), exercise improves circulation to the site of injury, promotes healing, reduces muscle spasm and tension, and along with beta-endorphin release, results in a decreased perception of pain. Whether by physiological, psychological, and/or social processes, the alleviation of pain works together with exercise participation to enhance overall psychological well-being. However, when pain and disability persist despite increased efforts to improve fitness, psychological well-being deteriorates over time. This scenario may be exacerbated by other psychosocial processes such as social comparison.48 Those whose pain is improving may begin to feel better, to experience increased vigor and less anxiety, and report these positive changes to their exercise companions. Correspondingly, those who feel no better become more alarmed, wonder why their pain persists, continue feeling anxious, and begin to feel worse about themselves. As appealing as this interpretation might be, there may also be a case to support the alternative proposition that selfreported pain and disability tend to improve only when some of the psychological benefits of exercise are experienced. Recent findings on pain perception and stress-induced changes in the immune system that are associated with such variables as low self-esteem and perceived helplessness/ hopelessness need further investigation to test these relationships.52 The potential importance of the psychological benefits of exercise programs for the low back injured is highlighted by the results of the analyses of predictors of early retumto-work. Change in self-esteem was one of two predictor variables for this group of 40 subjects. Although not statistically significant, deterioration in self-esteem may be potentially clinically important and warrants further evaluation Arch Phys Med Rehabil Vol75, November 1994
RETURN TO WORK AFTER AQUAFITNESS PROGRAM, LeFort
and testing as an outcome of active rehabilitation programs for the LBP population.43 The finding that those with a repeat injury were at significantly greater risk of not returning to work by 2 months post treatment may suggest physiological differences in healing, or may indicate different expectations, beliefs, or fears in the individual as a result of having sustained a reinjury. Interestingly, this was only the second injury for most individuals in this group suggesting that poor response to treatment (ie, entrenched chronicity) may occur much earlier than commonly thought. Overall, this combined aquafit and muscle-strengthening intervention seems to have benefitted study subjects as measured by physical fitness indices. It did not reduce pain and disability ratings for all subjects however. Subjects whose pain did not decrease were likely to have had a repeat injury and had not returned to work or retraining. This finding may indicate the need for additional interventions targeted specifically to the repeat injury group to maximize treatment benefits of active rehabilitation programs. Acknowledgments: We wish to thank Lorraine Vardy, DPT, Dan Mosher, and Gail Innes for their assistance and support during this study. The authors gratefully acknowledge the contributions of Carey Norman, Robin Atkinson and others at the Lifestyles Program for data collection of the physical fitness measures. Thanks are extended also to Mary Ellen Jeans, PhD, RN, School of Nursing, McGill University, Montreal and Bonnie Stevens, PhD, RN, Faculty of Nursing, University of Toronto for critiquing earlier drafts of this article. References 1. Nachemson AL. Advances in low-back pain. Clin Grthop 1985; 2. Andersson G, Pope M, Frymoyer J, Snook S. Epidemiology and cost. In: Pooe M. Andersson G. Frvmover J. Chaffin D, editors. Occupational low back p’ain. St. Louis; M&by Year Book, 1991:95-113. _ 3. Mitchell R, Carmen G. Results of a multicenter trial using an intensive active exercise program for the treatment of acute soft tissue and back injuries. Spine 1990; 15:514-21. 4. BattiC M. Aerobic fitness and its measurement. Spine 1991; 16:677-8. 5. Frymoyer J. Back pain and sciatica. N Engl J Med 1988;318:291-300. 6. Nutter P. Aerobic exercise in the treatment and prevention of low back pain. &cup Med: State Art Rev 1988;3:137-45. 7. Mayer T, Gatchell R, Mayer H, Kishino N, Keeley J, Mooney V. A prospective two-year study of functional restoration in industrial low back injury. An objective assessment procedure. JAMA 1987; 258:1763-7. 8. Altmaier E, Lehmann T, Russell D, Weinstein J, Kao C. The effectiveness of psychological interventions for the rehabilitation of low back pain: a randomized controlled trial evaluation. Pain 1992;49:329-35. 9. Beekman E, Axtell L. Ambulation, activity level and pain. Outcomes of a program for spinal pain. Phys Ther 1985;65:1649-55. 10. Fredrickson B, Trief P, VanBeveren P, Yuan H, Baum G. Rehabilitation of the patient with chronic back pain. A search for outcome predictors. Spine 1988;13:351-3. 11. Hazard R, Fenwick J, Kalisch S, Redmond J, Reeves V, Reid S, et al. Functional restoration with behavioral support. A one-year prospective study of patients with chronic low-back pain. Spine 1989;14:157-61. 12. Lindstrom I, Ohlund C, Eek C, Wallin L, Peterson L, Nachemson A. Mobility, strength, and fitness after a graded activity program for patients with subacute low back pain. A randomized prospective clinical study with a behavioral therapy approach. Spine 1992;17:641-9. 13. Maver T. Gatchell R. Kishino N. Keelev J, Caura P, Maver H, et al. Objective assessment of spine function following indus&l injury. A prospective study with comparison group and one-year follow-up. Spine 1985; 10:482-93. 14. Oland G, Tveiten GA. Trial of modem rehabilitation for chronic lowback pain and disability. Vocational outcome and effect of pain modulation. Spine 1991; 16:457-9.
Arch whys Med Rehebil Vol75,
November 1994
15. Cassady S, Nielsen D. Cardiorespiratory responses of healthy subjects to calisthenics performed on land versus in water. Phys Ther 1992;72:532-8. 16. Whitley J, Schoene L. Comparison of heart rate responses. Water walking versus treadmill walking. Phys Ther 1987;67:1501-4. 17. Hannah TE, Hannah E, Mosher D, Vardy L. Fitness benefits of a lifestyles program for the rehabilitation of workers’ compensation recipients: effects of heartrate, muscle strength, and weight. J Rehabil 1988;54:37-41. 18. Hannah TE, Hannah E, Mosher D, Vardy L. Psychological benefits of a lifestyles modification program for workers’ compensation recipients. Physiother Can 1988;40:10-5. 19. Hannah TE, Kozma A, Stones M, Mosher D, Vardy L. Effects on mood of a lifestyles program for the rehabilitation of injured workers. J Gccup Med 1989;31:454-7. 20. Hannah TE, Hannah E, Mosher D, Vardy L. Confirmation of the physical fitness benefits of a lifestyle program for injured workers. Can J Rehabil 1989;2:163-8. 21. LaFontaine T, DiLorenzo T, Frensch P, Stucky-Kopp R, Bargman E, McDonald D. Aerobic exercise and mood. A brief review, 1985-1990. Sports Med 1992; 13:160-70. 22. Petmzzello S, Landers D, Hatfield B, Kubitz K, Salazar W. A_ _____memanalysis on the anxiety-reducing effects of acute and chronic exercise. Outcomes and mechanisms. Sports Med 1991: 11: 143-82. 23. Raglin J. Exercise and mental health. Beneficial and detrimental effects. Sports Med 1990;9:323-9. 24. Astrand P, Rodahl K. Textbook of work physiology: physiological basis of exercise. New York: McGraw Hill, 1986. 25. Melzack R. The McGill pain questionnaire.: major properties and scar ing methods. Pain 1975; 1:277-99. 26. Melzack R, Katz J. The McGill pain questionnaire: appraisal and current status. In: Turk D, Melzack R, editors. Handbook of pain assessment. New York Guilford, 1992:152-68. 27. Fairbank J, Couper J, Davies J, O’Brien J. The Oswestry low back disability questionnaire, Physiother 1980;66:271-3. 28. Rosenberg M. Society and the adolescent self-image. Princeton: Princeton University Press, 1965. 29. Duffy ME. Determinants of health status in employed women. Health Values 1989; 13:50-7. 30. Robinson J, Shaver P. Measures of social psychological attitudes. Ann Arbor: Institute of Social Research, University of Michigan, 1973. 31. Spielberger C, Gorsuch R, Lushene R, Vagg P, Jacobs G. Manual for the state-trait anxiety inventory. STAI (Form Y). Palo Alto: Consulting Psychologist Press, 1983. 32. McNeil J. Mood: measurement, diurnal variation and age effects [dissertation]. St. John’s (NF): Memorial University of Newfoundland, 1986. 33. Tabachnick B, Fidel1 L. Using multivariate statistics. New York: Haroer Collins, 1989. 34. Spitzer WO, LeBlanc FE, Dupuis M. Scientific approach to the assessment and management of activity-related spinal disorders. A monograph for clinicians. Spine 1987; 12:Sl-59. 35. McArdle W, Katch F, Katch V. Exercise physiology. Energy, nutrition and human performance. Philadelphia: Lea & Febiger, 1986:358-9. 36. Fuerstein M, Carter R, Papciak A. A prospective analysis of stress and fatigue in recurrent low back pain. Pain 1987;31:333-44. 37. Dolce J, Cracker M, Moletteire C, Doleys D. Exercise quotas, anticipatory concern and self-efficacy expectancies in chronic pain: a preliminary report. Pain 1986;24:365-72. 38. Fordyce W, McMahon R, Rainwater G, Jackins S, Quested K, Murphy T, et al. Pain complaint-exercise performance relationship in chronic pain. Pain 1981;10:311-21. 39. Kleinke C, Spangler A. Predicting treatment outcomes of chronic back pain patients in a multidisciplinary pain clinic: methodological issues and treatment implications. Pain 1988:33:41-8. 40. Linton SJ, Bradley LA, Jensen I, Spangfort E, Sundell L. The secondary prevention of low back pain: a controlled study with follow-up. Pain 1989;36:197-207. 41. Manniche C, Lundberg E, Christensen I, Bentzen L, Hesselsoe G. Intensive dynamic back exercises for chronic low back pain: a clinical trial. Pain 1991;47:53-63. 42. Armentrout D. The impact of chronic pain on self-concept. J Clin Psycho1 1979;35:517-21.
RETURN TO WORK AFTER AQUAFITNESS PROGRAM, LeFort 43. Beekman E, Axtell L, Noland K, West J. Self-concept: an outcome of a program for spinal pain. Pain 1985;22:59-66. 44. Coates R, Femmi P. Sexual dy&nction and marital d&harmony as a comequence of chtonic lumbar spinal pain. Sex Ma&al l&r 1991$X5-9. 45. Vikari-Juntura E, Vuori J, Silverstein B, Kalimo R, Kuosma E, Videman T. A life-long prospective study on the role of psychosocial factors in neck-shoulder and low-back pain. Spine 1991; 16:1056-61. 46. Ackerman M, Stevens M. Acute and chronic pain: pain dimensions and psychological status. J Clin Psycho1 1989;45:223-8. 41. Garron D, Leavitt F. Chronic low back pain and depression. J Clin Psycho1 1983;39:486-93.
12%
48. Mechanic D, Angel R. Some factors associated with the report and evaluation of back pain. J Health Sot Behav 1987;28:131-9. 49. Phelps JR. Physical activity and health maintenance. Exactly what is known? West J Med 1987; 146:200-6. 50 Sonstroem R, Morgan W. Exercise and self-esteem: rationale and ’ model. Med Sci Sports Exert 1989;21:329-37. 51. Wilfley D, Kunce J. Differential physical and psychological effects of exercise. J Counsel Psycho1 1986;33:337-42. 52. Marbach J, Schleifer S, Keller S. Facial pain, distress, and immune function. Brain Behav Immun 1990;4:243-54.
NOTICE TO AUTHORS The Archives of Physical Medicine and Rehabilitation is changing its address effective December 10, 1994. Manuscripts and correspondence should be addressed to: Editorial Board Archives of Physical Medicine and Rehabilitation Suite 2510 330 North Wabash Avenue Chicago, IL 6061 l-3604 Telephone (3 12) 464-9550