Copyright 1991 by The Geronlological Society of America
Journal of Gerontology: PSYCHOLOGICAL SCIENCES 1991, Vol. 46, No. 6, P352-361
Long-Term Effects of Exercise on Psychological Functioning in Older Men and Women James A. Blumenthal,1'2 Charles F. Emery,1-3 David J. Madden,1-3 Susan Schniebolk,1 Margaret Walsh-Riddle,1 Linda K. George,1-3 Daphne C. McKee,s Michael B. Higginbotham,2 Frederick R. Cobb,2 and R. Edward Coleman4 'Department of Psychiatry, department of Medicine, 3Center for the Study of Aging and Human Development, and ••Department of Radiology, Duke University Medical Center. 'North Carolina Spine Center, Chapel Hill, North Carolina.
R
ECENTLY, there has been a great deal of interest in the effects of exercise training on psychological functioning among older men and women. To date, results have been inconclusive. Although some studies have reported exercise training to be associated with improvements in cognitive performance and mood (Blumenthal, Williams, Needles, & Wallace, 1982; Dustman et al., 1984; Elsayed, Ismail, & Young, 1980; Powell, 1974; Stones & Kozma, 1988), other studies have failed to find appreciable changes in psychological functioning after exercise training (Barry, Steinmetz, Page, & Rodahl, 1966; Emery & Gatz, 1990; Gitlin, 1985; Hughes, Casal, & Leon, 1986). The inconsistencies in results may be attributed to a variety of factors, including differences in the duration of exercise training, experimental design, statistical analyses, subject characteristics, and selection of assessment instruments. In an initial report (Blumenthal et al., 1989), we described the cardiovascular and behavioral adaptations associated with a 4-month program of aerobic exercise training in a sample of 101 older men and women. Prior to and following the program, subjects completed an extensive battery of tests measuring mood, cognitive functioning, and cardiorespiratory fitness. Results revealed a significant 11.6% improvement in peak oxygen consumption (peak VO2) for subjects who engaged in aerobic exercise, while subjects in nonaerobic exercise control groups (Yoga or Waiting List) did not change. However, there were few significant psychological changes among study participants. Men, but not women, in the aerobic group became less depressed, and subjects in all three groups experienced a reduction in obsessiveP352
compulsive and psychoticism subscales of the SCL-90. Gender differences were observed in neuropsychological test performance such that men achieved more taps than women and had greater grip strength. Women achieved higher scores than men on the Selective Reminding Test, and men recalled more digits than women on the Digit Span Subtest of the WAIS-R and achieved higher scores than women on the Nonverbal Fluency Test. In addition, subjects in all three groups increased their tapping speed; showed less grip strength; and improved their performance on Selective Reminding, Digit Symbol, Nonverbal Fluency, and Stroop tests. These changes were largely attributed to practice effects rather than to the aerobic exercise. Similarly, performance on an information processing task also failed to improve with exercise (Madden, Blumenthal, Allen, & Emery, 1989). This study represents a follow-up report of the psychological changes associated with 8 and 14 months of continued exercise training. We hypothesized that a more extended exercise program may have been necessary in order to show improvement in psychological functioning. METHOD
Subjects. — The Duke Aging and Exercise Study was conducted from 1985 to 1987. Subjects were recruited from advertisements placed in local newspapers, television and radio advertisements, and from the Duke Aging Center Subject Registry. Details of the selection process are described in our earlier publication (Blumenthal etal., 1989). In brief, 101
Downloaded from http://geronj.oxfordjournals.org/ at University of Birmingham on June 6, 2015
The purpose of this study was to determine the psychological, behavioral, and cognitive changes associated with up to 14 months of aerobic exercise training. For the first 4 months of the study, 101 older (> 60 years) men and women were randomly assigned to one of three conditions: Aerobic exercise, Yoga, or a Waiting List control group. Before and following the intervention, all subjects completed a comprehensive assessment battery, including measures of mood and cognitive functioning. A semi-crossover design was employed such that, following completion of the second assessment, all subjects completed 4 months of aerobic exercise and underwent a third assessment. Subjects were given the option of participating in 6 additional months of supervised aerobic exercise (14 months total), and all subjects, regardless of their exercise status, completed a fourth assessment. Results indicated that subjects experienced a 10-15% improvement in aerobic capacity after 4 months of aerobic exercise. Continued exercise, however, was associated with small increments in aerobic capacity. In general, there were relatively few improvements in cognitive performance associated with aerobic exercise, although subjects who maintained their exercise participation for 14 months experienced improvements in some psychiatric symptoms. However, the healthy subjects in this study were functioning at a relatively high level to begin with, and exercise training may produce greater improvements among elderly with concomitant physical or emotional impairments.
EXERCISE AND PSYCHOLOGICAL FUNCTION
subjects (50 men, 51 women) were selected from an initial group of 113 adults. All study participants were at least 60 years of age (range 60-83 years), and were healthy and free from coronary disease as judged from medical history, physical examination, and bicycle ergometry testing performed under continuous electrocardiographic (ECG) recording. All subjects were sedentary and did not engage in regular exercise prior to their enrollment in this study. Although men in the study were slightly younger (mean age 66.4 ± 4.6 vs 67.7 ± 5.1 years) and slightly more educated (mean 15.5 ± 2.4 vs 14.8 ± 2.4 years) than the women, these differences were not statistically significant. The age distribution of the sample is depicted in Figure 1.
Age Distribution of Total Cohort (N=101)
60-64 65-69 70-74 75-80
Age (Years) Group Figure 1. Age distribution of sample.
>80
After 16 weeks, following the completion of the second (Time 2) assessment, all subjects from the three groups participated in an aerobic exercise training regimen. Exercise prescriptions were derived from the Time 2 bicycle ergometry studies and followed the AE training protocol described previously. All subjects participated in the program for 16 weeks and underwent a third (Time 3) assessment, which was identical to the assessments at Times 1 and 2. Following the completion of the Time 3 assessment, in response to numerous requests from study participants, the research was extended and subjects were given an opportunity to participate in 6 additional months of supervised aerobic exercise. The format of the exercise training was the same as that of the first 8 months of the study. At the end of the 6 months (14 months total), a fourth follow-up assessment (Time 4) was performed for all subjects, regardless of whether or not they engaged in supervised exercise. Each of the 4 assessments consisted of measures of physiological and psychological functioning. For the purposes of this report, only the measures of psychological functioning between Time 2-3 and Time 3-4 will be described. Results from the extensive physiological assessment (including bone density studies, lipids, and anaerobic and aerobic capacity) are reported in Blumenthal et al. (1991, in press). The psychological assessments at Times 1-3 were identical. However, due to time constraints, the psychological assessment at Time 4 included only a subset of the neuropsychological measures. Measures of cardiorespiratory function. — In order to measure cardiorespiratory fitness, subjects underwent bicycle ergometry testing. Each subject performed four maximum effort exercise tests following an initial practice test on a Fitron cycle ergometer (F1000750, Cybex Lumex, Ronkonkoma, NY). The graded exercise protocol consisted of 3-minute stages starting at 150 kpm and increasing 150 kpm each stage. Subjects maintained a pedaling rate of 60 rpm. Subjects exercised under continuous electrocardiographic monitoring until exhaustion. Heart rate was recorded every minute. Blood pressure was measured by cuff sphygmomanometry at 3-minute intervals. Respiratory and oxygen consumption measurements were obtained using a System 4400 metabolic system (Alpha Technologies, Laguna Hills, CA). Measurements of VO2, expired ventilation VE), and respiratory exchange ratio were measured breathby-breath and recorded every 15 seconds. Measures ofpsychological function. — A comprehensive psychological test battery, including measures of mood, psychiatric symptoms, and neuropsychological functioning, was administered at baseline (Time 1), at 4 months (Time 2), 8 months (Time 3), and 14 months (Time 4). These instruments have been described previously (Blumenthal et al., 1989) and for convenience are noted in the Appendix. Anxiety, depression, and overall mood were assessed by questionnaires. Anxiety was assessed by the State-Trait Anxiety Inventory (STAI) (Spielberger, Gorsuch, & Luschene, 1970). Depression was measured by the Center for Epidemiological Studies Depression Scale (CES-D)
Downloaded from http://geronj.oxfordjournals.org/ at University of Birmingham on June 6, 2015
Procedure. — For the first 4 months of this study, subjects were randomly assigned to an Aerobic exercise group (AE), a Yoga group (YO), or a Waiting List group (WL). A conditional randomization procedure was employed such that an approximately equal number of men and women were included in each group (AE: 17M/16F; YO: 17M/17F; WL: 16M/18F). The age distribution for the three groups was comparable (AE: M = 66.5, range 60-76; YO: M = 67.8, range 60-83; WL: M = 66.8, range 60-78). Subjects in the AE group performed aerobic exercise three times per week. Exercise sessions consisted of 30 minutes of continuous bicycle ergometry at an intensity equivalent to 70% maximum heart rate reserve determined at the time of an initial bicycle exercise test, and 15 additional minutes of walking, mild jogging, and arm ergometry exercises. A 10minute warm-up period and 5-minute cool-down period consisting of stretching and light calisthenics preceded and followed each AE session. For the YO group, subjects participated in 60 minutes of nonaerobic yoga exercises two times per week. This group provided a control for group support, attention, and social stimulation, while minimizing improvements in cardiorespiratory functioning. The WL group received no intervention.
P353
BLUMENTHAL ET AL.
P354
Data analyses. — The principal mode of data analysis was a repeated measures multivariate analysis of variance (MANOVA). Group (AE, YO, or WL) and Sex (male or female) served as between-subject factors, while Time served as a within-subject factor. For the comparison between Times 3-4, subjects were grouped by Status (i.e., whether or not they participated in the extended 6-month exercise program), rather than by their initial group assignment. In order to control for possible spurious statistical effects resulting from the numerous instruments that were used, variables were clustered into conceptual units (e.g., cardiorespiratory fitness, mood, psychiatric symptoms, motor function, memory, etc.) whenever possible. Univariate analyses were examined when significant multivariate effects were found.
testing. Only four subjects (2 AE, 2 WL) were unable to complete the Time 2 assessments, yielding an overall compliance rate of 97%. Of these 97 subjects, 90 subjects completed the Time 3 assessments (28 AE, 31 YO, 31 WL) for an 8-month compliance rate of 90%. At the conclusion of the Time 3 assessment, 50 subjects volunteered to continue for an additional 6 months of exercise, of whom 49 (98%) completed the program. At Time 4 (14 months after baseline), 84 subjects returned for study, including all of the subjects who had participated in the 6-month exercise extension. Table 1 shows the compliance for the exercise program. Cardiorespiratory fitness. — The changes in cardiorespiratory function over the 14-month period are described in Blumenthal et al. (1989, 1991). Because it is important to understand the psychological changes in the context of cardiorespiratory functioning, the effects of the exercise intervention in cardiorespiratory function will be briefly summarized. Between Times 1 and 2, the AE group had achieved significantly longer times on the bicycle, lower submaximal heart rates (HR), and greater improvements in peak VO2 compared with the YO and WL groups (see Table 2). The AE group increased their aerobic capacity by 11.6%, while the YO and WL groups did not change. Between Times 2 and 3, all subjects engaged in aerobic exercise. The YO and WL groups increased their peak aerobic capacity (V02max) by 10.5 and 15%, respectively, while the AE group only increased by 1.5%. In addition, subjects in the YO and WL groups increased their bicycle time, and their submaximal HR was reduced. Thus, by Time 3, all groups were generally equivalent with respect to their level of aerobic fitness. Between Times 3 and 4, 50 of the original group of 101 subjects elected to continue for 6 more months of supervised aerobic exercise. However, 84 subjects returned for followup at Time 4. At Time 4, the subjects who had participated in the additional 6 months of exercise showed lower submaximal HR and significantly longer times on the bicycle than the subjects who had not exercised. Thus, aerobic exercise was associated with a significant increase in functional capacity, which was maintained and increased among subjects who continued to exercise.
RESULTS
Subject compliance. — Of the 101 subjects who entered the study initially, 97 completed both Time 1 and Time 2
Mood. — Because data from Times 1 and 2 have been reported previously (Blumenthal et al., 1989), the MANOVA for the initial 4 months (Time 1-2) will not be
Table 1. Exercise Compliance Time 2-3 (n = 97)
Time 1-2 (n = 101)
Aerobic Yoga Waiting List Total
Time 3-4 (n = 50)
Compliers
M Sessions
Compliers
M Sessions
Compliers
M Sessions
31 (33) 34 (34) 32 (34) 97/101 (97%)
46 (48) 32 (32)
28(31) 31 (34) 31(32) 90/97 (91%)
44 (48) 45 (48) 44 (48)
23 (23) 12(12) 14(15) 49/50 (98%)
52 (72) 56 (72) 53 (72)
—
Note. The number of subjects complying with the program are depicted (the number of subjects initially in each group is in parentheses). The average (M) number of sessions attended are also included (out of the number possible).
Downloaded from http://geronj.oxfordjournals.org/ at University of Birmingham on June 6, 2015
(Radloff, 1977). Global mood was assessed by the Affect Balance Scale (Bradburn, 1969). Psychiatric symptoms were assessed from the Hopkins Symptom Checklist (SCL-90) (Derogatis, Lipman, Rickels, Uhlenhuth, & Covi, 1974). Five scales were selected in our initial report to measure psychiatric symptoms and included the Somatization, Obsessive-compulsive, Phobic anxiety, Paranoid ideation, and Psychoticism scales. Motor function was assessed by the Finger Tapping Test (Reitan, 1979). Memory functioning was assessed by the Short Story Module of the Randt Memory Test (Randt, Brown, & Osborne, 1980), the Digit Span Subtest of the Wechsler Adult Intelligence Scale-Revised (WAIS-R) (Wechsler, 1981), the Benton Revised Visual Retention Test (Benton, 1974), and the Selective Reminding Test (Buschke & Fuld, 1974). Perceptual motor function was assessed by the Digit Symbol Subtest of the WAIS-R (Wechsler, 1981), the Trail Making Test (Part B) (Reitan, 1979), and the 2 & 7 Test (Ruff, Evans, & Light, 1986). Several miscellaneous tests administered at Times 1-3 only were also included as additional measures of neuropsychological function, including the Nonverbal Fluency Test (Jones-Gotman & Milner, 1977), the Verbal Fluency Test (Benton, 1974), and the Stroop Color-Word Test (Stroop, 1935).
P355
EXERCISE AND PSYCHOLOGICAL FUNCTION
Table 2. Mean Values (±SD) for Cardiorespiratory Function by Group Aerobic Exercise Assessment
Yoga Assessment
Waiting List Assessment
2
3
4
1
2
3
4
1
2
3
4
70 (13)
72
(ID
72 (9)
75 (12)
73 (10)
70 (10)
72 (11)
70 (14)
68 (14)
69 (14)
71 (13)
HR submax (bpm)
105 (18)
98 (14)
99 (16)
96 (15)
106 (16)
104 (16)
97 (16)
98 (16)
103 (23)
105 (25)
99 (20)
100 (23)
Time (min)
11.3 (3.4)
12.1 (3.3)
12.5 (3.7)
12.2 (3.9)
11.0 (2.8)
10.1 (2.2)
11.1 (2.5)
10.6 (2.3)
10.6 (2.7)
9.8 (2.7)
11.1 (3.2)
10.6 (3.2)
VO2(max) (ml/kg/min)
19.5 (5.2)
21.4 (5.8)
21.6 (6.0)
22.7 (6.0)
18.8 (4.5)
18.7 (4.8)
19.8 (4.9)
20.0 (4.7)
18.4 (3.9)
17.9 (4.2)
20.4 (5.1)
20.6 (5.4)
% change in VO2(max)
—
11.6
1.5
4.1
—
.2
10.5
1.5
—
-2.5
15.1
2.4
Note. HR rest, heart rate at rest; HR submax, heart rate at 150 kpm; time, total time on bicycle; V02(max), peak oxygen consumption during bicycle ergometry test; % change in V02(max), percent change from previous assessment. Values are means (±SD). Assessments: 1, baseline; 2, 4 months; 3, 8 months; 4, 14 months. These data have been reported previously (Blumenthal et al., 1989; 1991), but are included to document changes in aerobic fitness.
reviewed in detail. Measures of depression (CES-D), anxiety (STAI), and overall mood (Affect balance) were considered in a MANOVA (see Table 3). There were no significant univariate effects for the STAI or Affect Balance Scale. A significant 3-way interaction was noted for the CES-D, with only men in the aerobic group achieving a significant reduction in depression scores. The MANOVA between Times 2 and 3 revealed a significant multivariate Time x Sex interaction (F(4,75) = 3.31, p < .02). Examination of the univariate effects revealed a single significant Time x Sex interaction for Affect Balance (F(l,78) = 5.55, p < .02), with men exhibiting a slight increase in mood and women exhibiting a slight decrease in mood. Between Times 3 and 4, none of the multivariate main effects or interactions were significant. Psychiatric symptoms. — In order to assess psychiatric symptoms, 5 of 10 scales from the SCL-90 were included in a MANOVA. These scales (somatization, obsessivecompulsive, phobic anxiety, paranoid ideation, and psychoticism) were the same scales that were selected in our initial report. Between Times 1 and 2, all three groups experienced reductions in scores on the obsessivecompulsive and psychoticism subscales. Between Times 2 and 3, there were significant multivariate main effects for sex (F(5,78) = 2.35, p < .05) and time (F(5,83) = 2.64, p < .03), and a marginally significant multivariate Time X Sex interaction (F(5,78) = 2.21, p< .06). The sex main effects occurred because men scored significantly higher than women on the somatization (52.2 ± 5.8 vs 49.8 ± 3.8) (F(l,82) = 8.16, p < .006) and obsessive-compulsive (57.6 ± 8.1 vs53.9 ± 5.1)(F(1,82) = 7.51, p < .008) subscales. In addition, Table 4 shows that the Time x Sex univariate interactions occurred because scores for men tended to increase between Times 2 and 3, while scores for women stayed the same or decreased on the somatization (F(l,82) = 4.04, p < .05), obsessivecompulsive (F(l,82) = 4.14, p < .05), paranoid ideation
(F(l,82) = 4.62, p < .03), and psychoticism (F(l,82) = 8.91, p < .004) subscales. Because all subjects engaged in aerobic exercise between Times 2-3, this differential pattern of change in psychiatric symptoms could not be attributed to exercise. The MANOVA between Times 3 and 4 revealed a significant multivariate sex main effect (F(5,65) = 2.47, p < .04) and a significant Time x Status interaction (F(5,65) = 2.41, p < .05. The Time 4 data presented in Table 4 represent the mean scores for all 84 subjects who completed the follow-up assessment. However, examination of the univariate Time x Status interactions revealed that the subjects who continued with the exercise program exhibited a significant reduction in scores on the paranoid ideation (51.6 ± 4.6 to 50.1 ± 4.3)(F(1,69) = 6.80, p< .01) and psychoticism(52.8 ± 5.6to50.9 ± 3.4) (F(l,69) = 6.80,p q
1
00 ^ "
00 CN
00 CN
Mal
Time 00
ime
mal
u
ro O
EXERCISE AND PSYCHOLOGICAL FUNCTION
failed to demonstrate that exercise training improves cognitive performance. Thus, it would appear that differences between fit and unfit elderly on tests of cognitive performance may be due more to intrinsic factors (e.g., genetic predisposition) or to subject selection factors than to extrinsic factors that may be modifiable with exercise training. Several methodological issues are worth noting. First, our study sample was not necessarily representative of all older persons. All subjects in our study were healthy, which is unusual because 85% of the elderly have been reported to have at least one coexisting chronic illness (George, Landerman, Blazer, & Melville, 1989). Our sample was also more highly educated than the general population, with all subjects having had at least a high school education, and 80% having some form of postsecondary education. Our sample was also highly motivated: only four subjects dropped out of the program after 4 months, and seven subjects after 8 months, for an overall compliance rate of 90%. In addition, half of the subjects elected to participate in 6 months of further supervised exercise, and all but one of those subjects (98%) completed the 6-month (14 months total) program. Indeed, after 14 months, 84% of the original sample returned for follow-up assessment. The high proportion of subjects who completed our program and our statistical approach, which employed the "intention to treat" principle (i.e., all subjects were included in the analyses independent of their physiologic response to the different interventions, which is a standard procedure for clinical trials), suggest that the present findings are likely to be representative of the kinds of changes that highly motivated, healthy older men and women can expect. It should also be noted that our data were not transformed to normalized scores [unlike Dustman et al. (1984) and Stacey et al. (1985)], but were analyzed using the observed raw scores on the various test instruments. Because our subjects were functioning at such a high level from the outset, it also may have been difficult to demonstrate significant improvement in psychological functioning, and an absence of deterioration over 14 months may be meaningful. The use of exercise training among elderly with concomitant health problems may be more beneficial than among healthy older men and women. Although subjects exhibited little objective improvement in psychological functioning, they perceived that they had improved (Emery & Blumenthal, 1990). Thus, while anecdotal accounts and retrospective self-reports indicate that subjects feel better and perceive themselves as performing better on cognitive tasks, objective evidence fails to corroborate these selfperceptions. These perceptions may be important in their own right and may be more sensitive to the effects of exercise than the types of psychological processes measured by the test instruments employed in our study. However, based upon objective data, it would appear that regular exercise is not invariably associated with improved psychological function. Identification of subgroups of elderly who may benefit from exercise training may be useful in future studies.
ACKNOWLEDGMENTS
This research was supported by grants from the National Institute on
Downloaded from http://geronj.oxfordjournals.org/ at University of Birmingham on June 6, 2015
tion. For example, subjects who maintained their participation exhibited reduced scores on the paranoid ideation and psychoticism scales from the SCL-90. However, there was no selective advantage for the individuals who maintained their exercise program, either on the other measures of mood, or on several tests of memory and psychomotor functioning. These data can be contrasted with previous cross-sectional studies that noted fit persons to perform better on various cognitive measures than their unfit counterparts (Clarkson-Smith & Hartley, 1989; Spirduso, 1980; Stacey, Kozma, & Stones, 1985). Stacey et al. (1985) studied 29 new and 37 active volunteers over age 50; cognitive performance after 6 months was compared to baseline. Although time main effects were observed for such cognitive measures as reaction time and Digit Symbol, there was no group membership by time interaction. The absence of a nonexercise control group and the failure to observe improvements in aerobic power among exercisers make it possible to attribute improved cognitive performance to improved fitness levels. Two previous longitudinal studies have suggested that exercise may improve cognitive performance. In the Functional Age and Physical Activity study (Stones & Kozma, 1988), 311 subjects between 50 and 86 years (mean 62.8 years) were initially studied, and 200 subjects were retested after one year. A composite performance index was constructed for each subject from standardized scores on several neuropsychological tests: Critical Flicker Fusion Threshold, Culture Fair Intelligence, Digit Span, Digit Symbol, Dots Estimation, Reaction Time, and Stroop. The authors claimed that the findings showed that the improvement on the aggregate (composite) index was significantly greater for the Aerobic group than for the other groups; furthermore, the Aerobic group improved on five tests, while the other groups improved on only one test. However, it is difficult to interpret these results since the actual test scores and measures of aerobic capacity were not reported in sufficient detail. In another longitudinal study, Dustman et al. (1984) reported significant improvements in cognitive functioning among 13 exercise participants. Compared to the present subjects, participants in the Dustman et al. study were younger (mean 60 vs 67 years) and experienced twice the level of improvement in aerobic capacity (27%) after 4 months of exercise training. Although Dustman et al. alternately assigned subjects to an aerobic exercise and a control group, they did not randomize subjects in a third nonexercise control group, and no follow-up was performed. Despite these problems, it is widely believed that cognitive performance is enhanced by exercise. Stones and Kozma (1988) suggested that data from the Dustman et al. and Functional Age and Physical Activity studies provide "strong evidence for an encompassing improvement in functional capabilities after an aerobic exercise intervention, such that the generalized effects include gains in cognitive performance (p. 298)." With the exception of Dustman et al. and the Functional Age and Physical Activity studies, however, other recent longitudinal studies of middle-aged adults (Blumenthal & Madden, 1988), patients with hypertension (Madden, Blumenthal, & Ekelund, 1988), and healthy elderly (Gitlin, 1985; Madden et al., 1989) have all
P359
P360
BLUMENTHAL ET AL.
Aging (AG-04238) and the National Heart, Lung, and Blood Institute (HL30675). The authors wish to thank Janet Simon, Robin Pomeroy, Sally Schnitz, Julia Whitaker, and Cheri Rich for their technical assistance; Ross Ulmer and Dr. Bruce Burchett for their statistical assistance; and Janet Ivey for her secretarial assistance. Address correspondence to Dr. James A. Blumenthal, Department of Psychiatry, Duke University Medical Center, Box 3119, Durham, NC 27710. REFERENCES
Skills, 62, 407^16. Sidney, K. H., & Shephard, R. J. (1978). Frequency and intensity of exercise training for elderly subjects. Medicine and Science in Sports Exercise, 10, 125-131. Spielberger, C. E., Gorsuch, R. L., & Luschene, R. E. (1970). Manual for the state-trait anxiety inventory. Palo Alto, CA: Consulting Psychologist Press. Spirduso, W. W. (1980). Physical fitness, aging, and psychomotor speed: A review. Journal of Gerontology, 35, 850-865. Stacey, C , Kozma, A., & Stones, M. J. (1985). Simple cognitive and behavioral changes resulting from improved physical fitness in persons over 50 years of age. Canadian Journal of Aging, 4, 67-74. Stones, M. J., & Kozma, A. (1988). Physical activity, age and cognitive/ motor performance. In M. L. Howe & C. J. Brainerd (Eds.), Cognitive development in adulthood: Progress in cognitive development research (pp. 273-321). New York: Springer. Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental Psychology, 18, 643-662. Weschler, D. (1981). WAIS-R Manual. New York: The Psychological Corporation.
Received January 31, 1990 Accepted October 16, 1990
(See Appendix on next page)
Downloaded from http://geronj.oxfordjournals.org/ at University of Birmingham on June 6, 2015
Barry, A. J., Steinmetz, J. R., Page, H. F., & Rodahl, K. (1966). The effects of physical conditioning on older individuals. II. Motor performance and cognitive function. Journal of Gerontology, 21, 192-199. Benton, A. L. (1974). The revised visual retention test (4th ed.). New York: Psychological Corporation. Blumenthal, J. A., Emery, C. F., Madden, D. J., Coleman, R. E., Riddle, M. W., Schniebolk, S., Cobb, F. R., Sullivan, M. J., & Higginbotham, M. B. (1991). Effects of exercise training on cardiorespiratory function in men and women 60 years of age. American Journal of Cardiology, 67, 633-639. Blumenthal, J. A., Emery, C. F., Madden, D. J., George, L. K., Coleman, R. E., Riddle, M. W., McKee, D. C , Reasoner. J., & Williams, R. S. (1989). Cardiovascular and behavioral effects of aerobic exercise training in healthy older men and women. Journal of Gerontology: Medical Sciences, 44, M147-M157. Blumenthal, J. A., Emery, C. F., Madden, D. J., Schniebolk, S., Riddle, M. W., Cobb, F. R., & Coleman, R. E. (In press). Effects of exercise training on bone density in older men and women. Journal of the American Geriatrics Society. Blumenthal, J. A., Williams, R. S., Needles, T. L., & Wallace, A. G. (1982). Psychological changes accompany aerobic exercise in healthy middle-aged adults. Psychosomatic Medicine, 44, 529-536. Bradburn, N. M. (1969). The structure of psychological well-being. Chicago: Aldine, Buschke, H., & Fuld, P. A. (1974). Evaluating storage, retention, and retrieval in disordered memory and learning. Neurology I, 11, 10191025. Clarkson-Smith, L., & Hartley, A. A. (1989). Relationships between physical exercise and cognitive abilities in older adults. Psychology and Aging, 4, 183-189. Cunningham, D. A., Rechnitzer, P. A., Howard, J. H., & Donner, A. P. (1987). Exercise training of men at retirement: A clinical trial. Journal of Gerontology, 42, 17-23. Derogatis, L., Lipman, R. S., Rickels, K., Uhlenhuth, E. H., & Covi, L. (1974). The Hopkins symptom checklist (HSCL): A measure of primary symptom dimensions. In C. Pichot (Ed.), Psychological measurements in psychopharmacology: Modern problems in pharmacopsychiatry (Vol. 7). Basel: S! Karger. Dustman, R. E., Ruhling, R. O., Russell, E. M., Shearer, D. E., Bonekat, H. W., Shigeoka, J. W., Wood, J. S., & Bradford, D. C. (1984). Aerobic exercise training and improved neurophysiological function of older individuals. Neurobiology of Aging, 5, 35-42. Elsayed, M., Ismail, A. H., & Young, R. J. (1980). Intellectual differences of adult men related to age and physical fitness before and after an exercise program. Journal of Gerontology, 35, 383-387. Emery, C. F., & Blumenthal, J. A. (1990). Perceived change among participants in an exercise program for older adults. The Gerontologist, 30, 516-521. Emery, C. F., & Gatz, M. (1990). Psychological and cognitive effects of an exercise program for community-residing older adults. The Gerontologist, 30, 184-188.
George, L. K., Landerman, R., Blazer, D., & Melville, M. L. (1989). Concurrent morbidity between physical and mental illness: An epidemiologic examination. In L. L. Carstensen & J. Neale (Eds.), Mechanisms of psychological influence on physical health, with special attention to the elderly. New York: Plenum Press. Gitlin, L. N. (1985, March). Psychological effects of physical conditioning in the well elderly. Paper presented at the 2nd National Forum on Research in Aging, University of Nebraska, Lincoln, NE. Hughes, J. R., Casal, D. C , & Leon, A. S. (1986). Psychological effects of exercise: A randomized cross-over trial. Journal of Psychosomatic Research, 30, 355-360. Jones-Gotman, M., & Milner, B. (1977). Design fluency: The invention of nonsense drawings after focal cortical lesions. Neuropsychologia, 15, 653-674. Madden, D. J., Blumenthal, J. A., Allen, P. A., & Emery, C. F. (1989). Improving aerobic capacity in healthy older adults does not necessarily lead to improved cognitive performance. Psychology and Aging, 4, 307-320. Madden, D. J., Blumenthal, J. A., & Ekelund L-G. (1988). Effects of beta blockade and exercise on cardiovascular and cognitive functioning. Hypertension, 11, 470-476. Niinimaa, V., & Shephard, R. J. (1978). Training and oxygen conductance in the elderly. I. The respiratory system. Journal of Gerontology, 33, 354-361. Powell, R. R. (1974). Psychological effects of exercise therapy upon institutionalized geriatric mental patients. Journal of Gerontology, 29, 157-161. Radloff, L. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measures, 1, 385-401. Randt, C. T., Brown, E. R., & Osborne, D. P., Jr. (1980). A memory test for longitudinal measurement of mild to moderate deficits. Clinical Neuropsychology, 2, 184-194. Reitan, R. M. (1979). Manual for administration of neuropsychological test batteries for adults and children. Tucson: Reitan Neuropsychological Laboratories, Inc. Ruff, R. M., Evans, R. W., & Light, R. H. (1986). Automatic detection vs controlled search: A paper-and-pencil approach. Perceptual and Motor
EXERCISE AND PSYCHOLOGICAL FUNCTION
P361
Appendix Measures of Psychological Function
9. 10.
11.
12.
13.
14.
15.
to draw from memory a series of geometric shapes following a 10-second exposure. Higher scores indicate better memory. The Selective Reminding Test requires the subject to read a list of words and then to recall as many as possible. Higher scores are associated with better memory. The Digit Symbol Subtest of the WAIS-R is a paper-andpencil task that requires the subject to reproduce as many coded symbols as possible in a blank box beneath the digit with which it is paired within 90 seconds. More digits completed indicate better performance. The Trail Making Test (Part B) requires the subject to connect a series of consecutively numbered and lettered circles in alternating sequence (i.e., 1-A-2-B, etc.) as quickly as possible. Faster times indicate better performance. The 2 & 7 Test is a timed visual scanning task in which subjects must identify "target" stimuli (the digits 2 & 7) from an array of visually presented distractors. Faster times indicate better performance. The Nonverbal Fluency Test requires the subject to draw as many original, nonrepetitive designs as possible within a five-dot matrix within a 2-minute time limit. More items completed indicate better performance. The Verbal Fluency Test involves a series of 1-minute trials in which subjects are asked to generate as many words as possible that begin with a given letter. More items completed indicate better performance. The Stroop Color-Word Test measures the ease with which subjects can shift their perceptual set by accurately naming the color of the ink that the words red, blue, or green are printed in. Each word is printed in a different color ink than the color it is printed in. The number of correct (C) and incorrect (I) responses are recorded. More correct items indicate better performance.
Downloaded from http://geronj.oxfordjournals.org/ at University of Birmingham on June 6, 2015
1. The State-Trait Anxiety Inventory (STAI) is a 40-item questionnaire that measures levels of anxiety at the time of the assessment (state) and in general (trait). Higher scores are associated with greater anxiety. 2. The Center for Epidemiological Studies Depression Scale (CES-D) is a 20-item scale that assesses how many days in the past week the subject experienced symptoms associated with depression. Higher scores are associated with greater depression. 3. The Affect Balance Scale is a 10-item mood survey used to measure global positive and negative emotional states. Higher scores are associated with more positive affect. 4. The Hopkins Symptom Checklist (SCL-90) is a 90-item questionnaire designed to assess the degree of emotional distress and psychopathology. Five of 10 clinical scales were preselected to assess psychiatric symptoms, including somatization, obsessive-compulsive, phobic anxiety, paranoid ideation, and psychoticism. Higher scores are associated with greater emotional disturbance. 5. The Finger (Oscillation) Tapping Test requires the subject to tap the index finger on each hand as quickly as possible for two 30-second trials. Measures were obtained for both dominant and nondominant hands. Data from 15 and 30 seconds were recorded. More taps indicate better performance. 6. The Short Story Module of the Randt Memory Test requires the subject to recall the details of a short story immediately after it has been read to them and after a 30-minute delay. Higher scores are associated with better memory and attention. 7. The Digit Span Subtest of the Wechsler Adult Intelligence Scale-Revised (WAIS-R) requires the subject to repeat a series of digits that have been orally presented to them both forward and, in an independent test, in reverse order. The more digits recalled indicate better memory and attention. 8. The Benton Revised Visual Retention Test requires the subject
Journal of Gerontology: PSYCHOLOGICAL SCIENCES 1991, Vol. 46, No. 6, P352-361
Long-Term Effects of Exercise on Psychological Functioning in Older Men and Women James A. Blumenthal,1'2 Charles F. Emery,1-3 David J. Madden,1-3 Susan Schniebolk,1 Margaret Walsh-Riddle,1 Linda K. George,1-3 Daphne C. McKee,s Michael B. Higginbotham,2 Frederick R. Cobb,2 and R. Edward Coleman4 'Department of Psychiatry, department of Medicine, 3Center for the Study of Aging and Human Development, and ••Department of Radiology, Duke University Medical Center. 'North Carolina Spine Center, Chapel Hill, North Carolina.
R
ECENTLY, there has been a great deal of interest in the effects of exercise training on psychological functioning among older men and women. To date, results have been inconclusive. Although some studies have reported exercise training to be associated with improvements in cognitive performance and mood (Blumenthal, Williams, Needles, & Wallace, 1982; Dustman et al., 1984; Elsayed, Ismail, & Young, 1980; Powell, 1974; Stones & Kozma, 1988), other studies have failed to find appreciable changes in psychological functioning after exercise training (Barry, Steinmetz, Page, & Rodahl, 1966; Emery & Gatz, 1990; Gitlin, 1985; Hughes, Casal, & Leon, 1986). The inconsistencies in results may be attributed to a variety of factors, including differences in the duration of exercise training, experimental design, statistical analyses, subject characteristics, and selection of assessment instruments. In an initial report (Blumenthal et al., 1989), we described the cardiovascular and behavioral adaptations associated with a 4-month program of aerobic exercise training in a sample of 101 older men and women. Prior to and following the program, subjects completed an extensive battery of tests measuring mood, cognitive functioning, and cardiorespiratory fitness. Results revealed a significant 11.6% improvement in peak oxygen consumption (peak VO2) for subjects who engaged in aerobic exercise, while subjects in nonaerobic exercise control groups (Yoga or Waiting List) did not change. However, there were few significant psychological changes among study participants. Men, but not women, in the aerobic group became less depressed, and subjects in all three groups experienced a reduction in obsessiveP352
compulsive and psychoticism subscales of the SCL-90. Gender differences were observed in neuropsychological test performance such that men achieved more taps than women and had greater grip strength. Women achieved higher scores than men on the Selective Reminding Test, and men recalled more digits than women on the Digit Span Subtest of the WAIS-R and achieved higher scores than women on the Nonverbal Fluency Test. In addition, subjects in all three groups increased their tapping speed; showed less grip strength; and improved their performance on Selective Reminding, Digit Symbol, Nonverbal Fluency, and Stroop tests. These changes were largely attributed to practice effects rather than to the aerobic exercise. Similarly, performance on an information processing task also failed to improve with exercise (Madden, Blumenthal, Allen, & Emery, 1989). This study represents a follow-up report of the psychological changes associated with 8 and 14 months of continued exercise training. We hypothesized that a more extended exercise program may have been necessary in order to show improvement in psychological functioning. METHOD
Subjects. — The Duke Aging and Exercise Study was conducted from 1985 to 1987. Subjects were recruited from advertisements placed in local newspapers, television and radio advertisements, and from the Duke Aging Center Subject Registry. Details of the selection process are described in our earlier publication (Blumenthal etal., 1989). In brief, 101
Downloaded from http://geronj.oxfordjournals.org/ at University of Birmingham on June 6, 2015
The purpose of this study was to determine the psychological, behavioral, and cognitive changes associated with up to 14 months of aerobic exercise training. For the first 4 months of the study, 101 older (> 60 years) men and women were randomly assigned to one of three conditions: Aerobic exercise, Yoga, or a Waiting List control group. Before and following the intervention, all subjects completed a comprehensive assessment battery, including measures of mood and cognitive functioning. A semi-crossover design was employed such that, following completion of the second assessment, all subjects completed 4 months of aerobic exercise and underwent a third assessment. Subjects were given the option of participating in 6 additional months of supervised aerobic exercise (14 months total), and all subjects, regardless of their exercise status, completed a fourth assessment. Results indicated that subjects experienced a 10-15% improvement in aerobic capacity after 4 months of aerobic exercise. Continued exercise, however, was associated with small increments in aerobic capacity. In general, there were relatively few improvements in cognitive performance associated with aerobic exercise, although subjects who maintained their exercise participation for 14 months experienced improvements in some psychiatric symptoms. However, the healthy subjects in this study were functioning at a relatively high level to begin with, and exercise training may produce greater improvements among elderly with concomitant physical or emotional impairments.
EXERCISE AND PSYCHOLOGICAL FUNCTION
subjects (50 men, 51 women) were selected from an initial group of 113 adults. All study participants were at least 60 years of age (range 60-83 years), and were healthy and free from coronary disease as judged from medical history, physical examination, and bicycle ergometry testing performed under continuous electrocardiographic (ECG) recording. All subjects were sedentary and did not engage in regular exercise prior to their enrollment in this study. Although men in the study were slightly younger (mean age 66.4 ± 4.6 vs 67.7 ± 5.1 years) and slightly more educated (mean 15.5 ± 2.4 vs 14.8 ± 2.4 years) than the women, these differences were not statistically significant. The age distribution of the sample is depicted in Figure 1.
Age Distribution of Total Cohort (N=101)
60-64 65-69 70-74 75-80
Age (Years) Group Figure 1. Age distribution of sample.
>80
After 16 weeks, following the completion of the second (Time 2) assessment, all subjects from the three groups participated in an aerobic exercise training regimen. Exercise prescriptions were derived from the Time 2 bicycle ergometry studies and followed the AE training protocol described previously. All subjects participated in the program for 16 weeks and underwent a third (Time 3) assessment, which was identical to the assessments at Times 1 and 2. Following the completion of the Time 3 assessment, in response to numerous requests from study participants, the research was extended and subjects were given an opportunity to participate in 6 additional months of supervised aerobic exercise. The format of the exercise training was the same as that of the first 8 months of the study. At the end of the 6 months (14 months total), a fourth follow-up assessment (Time 4) was performed for all subjects, regardless of whether or not they engaged in supervised exercise. Each of the 4 assessments consisted of measures of physiological and psychological functioning. For the purposes of this report, only the measures of psychological functioning between Time 2-3 and Time 3-4 will be described. Results from the extensive physiological assessment (including bone density studies, lipids, and anaerobic and aerobic capacity) are reported in Blumenthal et al. (1991, in press). The psychological assessments at Times 1-3 were identical. However, due to time constraints, the psychological assessment at Time 4 included only a subset of the neuropsychological measures. Measures of cardiorespiratory function. — In order to measure cardiorespiratory fitness, subjects underwent bicycle ergometry testing. Each subject performed four maximum effort exercise tests following an initial practice test on a Fitron cycle ergometer (F1000750, Cybex Lumex, Ronkonkoma, NY). The graded exercise protocol consisted of 3-minute stages starting at 150 kpm and increasing 150 kpm each stage. Subjects maintained a pedaling rate of 60 rpm. Subjects exercised under continuous electrocardiographic monitoring until exhaustion. Heart rate was recorded every minute. Blood pressure was measured by cuff sphygmomanometry at 3-minute intervals. Respiratory and oxygen consumption measurements were obtained using a System 4400 metabolic system (Alpha Technologies, Laguna Hills, CA). Measurements of VO2, expired ventilation VE), and respiratory exchange ratio were measured breathby-breath and recorded every 15 seconds. Measures ofpsychological function. — A comprehensive psychological test battery, including measures of mood, psychiatric symptoms, and neuropsychological functioning, was administered at baseline (Time 1), at 4 months (Time 2), 8 months (Time 3), and 14 months (Time 4). These instruments have been described previously (Blumenthal et al., 1989) and for convenience are noted in the Appendix. Anxiety, depression, and overall mood were assessed by questionnaires. Anxiety was assessed by the State-Trait Anxiety Inventory (STAI) (Spielberger, Gorsuch, & Luschene, 1970). Depression was measured by the Center for Epidemiological Studies Depression Scale (CES-D)
Downloaded from http://geronj.oxfordjournals.org/ at University of Birmingham on June 6, 2015
Procedure. — For the first 4 months of this study, subjects were randomly assigned to an Aerobic exercise group (AE), a Yoga group (YO), or a Waiting List group (WL). A conditional randomization procedure was employed such that an approximately equal number of men and women were included in each group (AE: 17M/16F; YO: 17M/17F; WL: 16M/18F). The age distribution for the three groups was comparable (AE: M = 66.5, range 60-76; YO: M = 67.8, range 60-83; WL: M = 66.8, range 60-78). Subjects in the AE group performed aerobic exercise three times per week. Exercise sessions consisted of 30 minutes of continuous bicycle ergometry at an intensity equivalent to 70% maximum heart rate reserve determined at the time of an initial bicycle exercise test, and 15 additional minutes of walking, mild jogging, and arm ergometry exercises. A 10minute warm-up period and 5-minute cool-down period consisting of stretching and light calisthenics preceded and followed each AE session. For the YO group, subjects participated in 60 minutes of nonaerobic yoga exercises two times per week. This group provided a control for group support, attention, and social stimulation, while minimizing improvements in cardiorespiratory functioning. The WL group received no intervention.
P353
BLUMENTHAL ET AL.
P354
Data analyses. — The principal mode of data analysis was a repeated measures multivariate analysis of variance (MANOVA). Group (AE, YO, or WL) and Sex (male or female) served as between-subject factors, while Time served as a within-subject factor. For the comparison between Times 3-4, subjects were grouped by Status (i.e., whether or not they participated in the extended 6-month exercise program), rather than by their initial group assignment. In order to control for possible spurious statistical effects resulting from the numerous instruments that were used, variables were clustered into conceptual units (e.g., cardiorespiratory fitness, mood, psychiatric symptoms, motor function, memory, etc.) whenever possible. Univariate analyses were examined when significant multivariate effects were found.
testing. Only four subjects (2 AE, 2 WL) were unable to complete the Time 2 assessments, yielding an overall compliance rate of 97%. Of these 97 subjects, 90 subjects completed the Time 3 assessments (28 AE, 31 YO, 31 WL) for an 8-month compliance rate of 90%. At the conclusion of the Time 3 assessment, 50 subjects volunteered to continue for an additional 6 months of exercise, of whom 49 (98%) completed the program. At Time 4 (14 months after baseline), 84 subjects returned for study, including all of the subjects who had participated in the 6-month exercise extension. Table 1 shows the compliance for the exercise program. Cardiorespiratory fitness. — The changes in cardiorespiratory function over the 14-month period are described in Blumenthal et al. (1989, 1991). Because it is important to understand the psychological changes in the context of cardiorespiratory functioning, the effects of the exercise intervention in cardiorespiratory function will be briefly summarized. Between Times 1 and 2, the AE group had achieved significantly longer times on the bicycle, lower submaximal heart rates (HR), and greater improvements in peak VO2 compared with the YO and WL groups (see Table 2). The AE group increased their aerobic capacity by 11.6%, while the YO and WL groups did not change. Between Times 2 and 3, all subjects engaged in aerobic exercise. The YO and WL groups increased their peak aerobic capacity (V02max) by 10.5 and 15%, respectively, while the AE group only increased by 1.5%. In addition, subjects in the YO and WL groups increased their bicycle time, and their submaximal HR was reduced. Thus, by Time 3, all groups were generally equivalent with respect to their level of aerobic fitness. Between Times 3 and 4, 50 of the original group of 101 subjects elected to continue for 6 more months of supervised aerobic exercise. However, 84 subjects returned for followup at Time 4. At Time 4, the subjects who had participated in the additional 6 months of exercise showed lower submaximal HR and significantly longer times on the bicycle than the subjects who had not exercised. Thus, aerobic exercise was associated with a significant increase in functional capacity, which was maintained and increased among subjects who continued to exercise.
RESULTS
Subject compliance. — Of the 101 subjects who entered the study initially, 97 completed both Time 1 and Time 2
Mood. — Because data from Times 1 and 2 have been reported previously (Blumenthal et al., 1989), the MANOVA for the initial 4 months (Time 1-2) will not be
Table 1. Exercise Compliance Time 2-3 (n = 97)
Time 1-2 (n = 101)
Aerobic Yoga Waiting List Total
Time 3-4 (n = 50)
Compliers
M Sessions
Compliers
M Sessions
Compliers
M Sessions
31 (33) 34 (34) 32 (34) 97/101 (97%)
46 (48) 32 (32)
28(31) 31 (34) 31(32) 90/97 (91%)
44 (48) 45 (48) 44 (48)
23 (23) 12(12) 14(15) 49/50 (98%)
52 (72) 56 (72) 53 (72)
—
Note. The number of subjects complying with the program are depicted (the number of subjects initially in each group is in parentheses). The average (M) number of sessions attended are also included (out of the number possible).
Downloaded from http://geronj.oxfordjournals.org/ at University of Birmingham on June 6, 2015
(Radloff, 1977). Global mood was assessed by the Affect Balance Scale (Bradburn, 1969). Psychiatric symptoms were assessed from the Hopkins Symptom Checklist (SCL-90) (Derogatis, Lipman, Rickels, Uhlenhuth, & Covi, 1974). Five scales were selected in our initial report to measure psychiatric symptoms and included the Somatization, Obsessive-compulsive, Phobic anxiety, Paranoid ideation, and Psychoticism scales. Motor function was assessed by the Finger Tapping Test (Reitan, 1979). Memory functioning was assessed by the Short Story Module of the Randt Memory Test (Randt, Brown, & Osborne, 1980), the Digit Span Subtest of the Wechsler Adult Intelligence Scale-Revised (WAIS-R) (Wechsler, 1981), the Benton Revised Visual Retention Test (Benton, 1974), and the Selective Reminding Test (Buschke & Fuld, 1974). Perceptual motor function was assessed by the Digit Symbol Subtest of the WAIS-R (Wechsler, 1981), the Trail Making Test (Part B) (Reitan, 1979), and the 2 & 7 Test (Ruff, Evans, & Light, 1986). Several miscellaneous tests administered at Times 1-3 only were also included as additional measures of neuropsychological function, including the Nonverbal Fluency Test (Jones-Gotman & Milner, 1977), the Verbal Fluency Test (Benton, 1974), and the Stroop Color-Word Test (Stroop, 1935).
P355
EXERCISE AND PSYCHOLOGICAL FUNCTION
Table 2. Mean Values (±SD) for Cardiorespiratory Function by Group Aerobic Exercise Assessment
Yoga Assessment
Waiting List Assessment
2
3
4
1
2
3
4
1
2
3
4
70 (13)
72
(ID
72 (9)
75 (12)
73 (10)
70 (10)
72 (11)
70 (14)
68 (14)
69 (14)
71 (13)
HR submax (bpm)
105 (18)
98 (14)
99 (16)
96 (15)
106 (16)
104 (16)
97 (16)
98 (16)
103 (23)
105 (25)
99 (20)
100 (23)
Time (min)
11.3 (3.4)
12.1 (3.3)
12.5 (3.7)
12.2 (3.9)
11.0 (2.8)
10.1 (2.2)
11.1 (2.5)
10.6 (2.3)
10.6 (2.7)
9.8 (2.7)
11.1 (3.2)
10.6 (3.2)
VO2(max) (ml/kg/min)
19.5 (5.2)
21.4 (5.8)
21.6 (6.0)
22.7 (6.0)
18.8 (4.5)
18.7 (4.8)
19.8 (4.9)
20.0 (4.7)
18.4 (3.9)
17.9 (4.2)
20.4 (5.1)
20.6 (5.4)
% change in VO2(max)
—
11.6
1.5
4.1
—
.2
10.5
1.5
—
-2.5
15.1
2.4
Note. HR rest, heart rate at rest; HR submax, heart rate at 150 kpm; time, total time on bicycle; V02(max), peak oxygen consumption during bicycle ergometry test; % change in V02(max), percent change from previous assessment. Values are means (±SD). Assessments: 1, baseline; 2, 4 months; 3, 8 months; 4, 14 months. These data have been reported previously (Blumenthal et al., 1989; 1991), but are included to document changes in aerobic fitness.
reviewed in detail. Measures of depression (CES-D), anxiety (STAI), and overall mood (Affect balance) were considered in a MANOVA (see Table 3). There were no significant univariate effects for the STAI or Affect Balance Scale. A significant 3-way interaction was noted for the CES-D, with only men in the aerobic group achieving a significant reduction in depression scores. The MANOVA between Times 2 and 3 revealed a significant multivariate Time x Sex interaction (F(4,75) = 3.31, p < .02). Examination of the univariate effects revealed a single significant Time x Sex interaction for Affect Balance (F(l,78) = 5.55, p < .02), with men exhibiting a slight increase in mood and women exhibiting a slight decrease in mood. Between Times 3 and 4, none of the multivariate main effects or interactions were significant. Psychiatric symptoms. — In order to assess psychiatric symptoms, 5 of 10 scales from the SCL-90 were included in a MANOVA. These scales (somatization, obsessivecompulsive, phobic anxiety, paranoid ideation, and psychoticism) were the same scales that were selected in our initial report. Between Times 1 and 2, all three groups experienced reductions in scores on the obsessivecompulsive and psychoticism subscales. Between Times 2 and 3, there were significant multivariate main effects for sex (F(5,78) = 2.35, p < .05) and time (F(5,83) = 2.64, p < .03), and a marginally significant multivariate Time X Sex interaction (F(5,78) = 2.21, p< .06). The sex main effects occurred because men scored significantly higher than women on the somatization (52.2 ± 5.8 vs 49.8 ± 3.8) (F(l,82) = 8.16, p < .006) and obsessive-compulsive (57.6 ± 8.1 vs53.9 ± 5.1)(F(1,82) = 7.51, p < .008) subscales. In addition, Table 4 shows that the Time x Sex univariate interactions occurred because scores for men tended to increase between Times 2 and 3, while scores for women stayed the same or decreased on the somatization (F(l,82) = 4.04, p < .05), obsessivecompulsive (F(l,82) = 4.14, p < .05), paranoid ideation
(F(l,82) = 4.62, p < .03), and psychoticism (F(l,82) = 8.91, p < .004) subscales. Because all subjects engaged in aerobic exercise between Times 2-3, this differential pattern of change in psychiatric symptoms could not be attributed to exercise. The MANOVA between Times 3 and 4 revealed a significant multivariate sex main effect (F(5,65) = 2.47, p < .04) and a significant Time x Status interaction (F(5,65) = 2.41, p < .05. The Time 4 data presented in Table 4 represent the mean scores for all 84 subjects who completed the follow-up assessment. However, examination of the univariate Time x Status interactions revealed that the subjects who continued with the exercise program exhibited a significant reduction in scores on the paranoid ideation (51.6 ± 4.6 to 50.1 ± 4.3)(F(1,69) = 6.80, p< .01) and psychoticism(52.8 ± 5.6to50.9 ± 3.4) (F(l,69) = 6.80,p q
1
00 ^ "
00 CN
00 CN
Mal
Time 00
ime
mal
u
ro O
EXERCISE AND PSYCHOLOGICAL FUNCTION
failed to demonstrate that exercise training improves cognitive performance. Thus, it would appear that differences between fit and unfit elderly on tests of cognitive performance may be due more to intrinsic factors (e.g., genetic predisposition) or to subject selection factors than to extrinsic factors that may be modifiable with exercise training. Several methodological issues are worth noting. First, our study sample was not necessarily representative of all older persons. All subjects in our study were healthy, which is unusual because 85% of the elderly have been reported to have at least one coexisting chronic illness (George, Landerman, Blazer, & Melville, 1989). Our sample was also more highly educated than the general population, with all subjects having had at least a high school education, and 80% having some form of postsecondary education. Our sample was also highly motivated: only four subjects dropped out of the program after 4 months, and seven subjects after 8 months, for an overall compliance rate of 90%. In addition, half of the subjects elected to participate in 6 months of further supervised exercise, and all but one of those subjects (98%) completed the 6-month (14 months total) program. Indeed, after 14 months, 84% of the original sample returned for follow-up assessment. The high proportion of subjects who completed our program and our statistical approach, which employed the "intention to treat" principle (i.e., all subjects were included in the analyses independent of their physiologic response to the different interventions, which is a standard procedure for clinical trials), suggest that the present findings are likely to be representative of the kinds of changes that highly motivated, healthy older men and women can expect. It should also be noted that our data were not transformed to normalized scores [unlike Dustman et al. (1984) and Stacey et al. (1985)], but were analyzed using the observed raw scores on the various test instruments. Because our subjects were functioning at such a high level from the outset, it also may have been difficult to demonstrate significant improvement in psychological functioning, and an absence of deterioration over 14 months may be meaningful. The use of exercise training among elderly with concomitant health problems may be more beneficial than among healthy older men and women. Although subjects exhibited little objective improvement in psychological functioning, they perceived that they had improved (Emery & Blumenthal, 1990). Thus, while anecdotal accounts and retrospective self-reports indicate that subjects feel better and perceive themselves as performing better on cognitive tasks, objective evidence fails to corroborate these selfperceptions. These perceptions may be important in their own right and may be more sensitive to the effects of exercise than the types of psychological processes measured by the test instruments employed in our study. However, based upon objective data, it would appear that regular exercise is not invariably associated with improved psychological function. Identification of subgroups of elderly who may benefit from exercise training may be useful in future studies.
ACKNOWLEDGMENTS
This research was supported by grants from the National Institute on
Downloaded from http://geronj.oxfordjournals.org/ at University of Birmingham on June 6, 2015
tion. For example, subjects who maintained their participation exhibited reduced scores on the paranoid ideation and psychoticism scales from the SCL-90. However, there was no selective advantage for the individuals who maintained their exercise program, either on the other measures of mood, or on several tests of memory and psychomotor functioning. These data can be contrasted with previous cross-sectional studies that noted fit persons to perform better on various cognitive measures than their unfit counterparts (Clarkson-Smith & Hartley, 1989; Spirduso, 1980; Stacey, Kozma, & Stones, 1985). Stacey et al. (1985) studied 29 new and 37 active volunteers over age 50; cognitive performance after 6 months was compared to baseline. Although time main effects were observed for such cognitive measures as reaction time and Digit Symbol, there was no group membership by time interaction. The absence of a nonexercise control group and the failure to observe improvements in aerobic power among exercisers make it possible to attribute improved cognitive performance to improved fitness levels. Two previous longitudinal studies have suggested that exercise may improve cognitive performance. In the Functional Age and Physical Activity study (Stones & Kozma, 1988), 311 subjects between 50 and 86 years (mean 62.8 years) were initially studied, and 200 subjects were retested after one year. A composite performance index was constructed for each subject from standardized scores on several neuropsychological tests: Critical Flicker Fusion Threshold, Culture Fair Intelligence, Digit Span, Digit Symbol, Dots Estimation, Reaction Time, and Stroop. The authors claimed that the findings showed that the improvement on the aggregate (composite) index was significantly greater for the Aerobic group than for the other groups; furthermore, the Aerobic group improved on five tests, while the other groups improved on only one test. However, it is difficult to interpret these results since the actual test scores and measures of aerobic capacity were not reported in sufficient detail. In another longitudinal study, Dustman et al. (1984) reported significant improvements in cognitive functioning among 13 exercise participants. Compared to the present subjects, participants in the Dustman et al. study were younger (mean 60 vs 67 years) and experienced twice the level of improvement in aerobic capacity (27%) after 4 months of exercise training. Although Dustman et al. alternately assigned subjects to an aerobic exercise and a control group, they did not randomize subjects in a third nonexercise control group, and no follow-up was performed. Despite these problems, it is widely believed that cognitive performance is enhanced by exercise. Stones and Kozma (1988) suggested that data from the Dustman et al. and Functional Age and Physical Activity studies provide "strong evidence for an encompassing improvement in functional capabilities after an aerobic exercise intervention, such that the generalized effects include gains in cognitive performance (p. 298)." With the exception of Dustman et al. and the Functional Age and Physical Activity studies, however, other recent longitudinal studies of middle-aged adults (Blumenthal & Madden, 1988), patients with hypertension (Madden, Blumenthal, & Ekelund, 1988), and healthy elderly (Gitlin, 1985; Madden et al., 1989) have all
P359
P360
BLUMENTHAL ET AL.
Aging (AG-04238) and the National Heart, Lung, and Blood Institute (HL30675). The authors wish to thank Janet Simon, Robin Pomeroy, Sally Schnitz, Julia Whitaker, and Cheri Rich for their technical assistance; Ross Ulmer and Dr. Bruce Burchett for their statistical assistance; and Janet Ivey for her secretarial assistance. Address correspondence to Dr. James A. Blumenthal, Department of Psychiatry, Duke University Medical Center, Box 3119, Durham, NC 27710. REFERENCES
Skills, 62, 407^16. Sidney, K. H., & Shephard, R. J. (1978). Frequency and intensity of exercise training for elderly subjects. Medicine and Science in Sports Exercise, 10, 125-131. Spielberger, C. E., Gorsuch, R. L., & Luschene, R. E. (1970). Manual for the state-trait anxiety inventory. Palo Alto, CA: Consulting Psychologist Press. Spirduso, W. W. (1980). Physical fitness, aging, and psychomotor speed: A review. Journal of Gerontology, 35, 850-865. Stacey, C , Kozma, A., & Stones, M. J. (1985). Simple cognitive and behavioral changes resulting from improved physical fitness in persons over 50 years of age. Canadian Journal of Aging, 4, 67-74. Stones, M. J., & Kozma, A. (1988). Physical activity, age and cognitive/ motor performance. In M. L. Howe & C. J. Brainerd (Eds.), Cognitive development in adulthood: Progress in cognitive development research (pp. 273-321). New York: Springer. Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental Psychology, 18, 643-662. Weschler, D. (1981). WAIS-R Manual. New York: The Psychological Corporation.
Received January 31, 1990 Accepted October 16, 1990
(See Appendix on next page)
Downloaded from http://geronj.oxfordjournals.org/ at University of Birmingham on June 6, 2015
Barry, A. J., Steinmetz, J. R., Page, H. F., & Rodahl, K. (1966). The effects of physical conditioning on older individuals. II. Motor performance and cognitive function. Journal of Gerontology, 21, 192-199. Benton, A. L. (1974). The revised visual retention test (4th ed.). New York: Psychological Corporation. Blumenthal, J. A., Emery, C. F., Madden, D. J., Coleman, R. E., Riddle, M. W., Schniebolk, S., Cobb, F. R., Sullivan, M. J., & Higginbotham, M. B. (1991). Effects of exercise training on cardiorespiratory function in men and women 60 years of age. American Journal of Cardiology, 67, 633-639. Blumenthal, J. A., Emery, C. F., Madden, D. J., George, L. K., Coleman, R. E., Riddle, M. W., McKee, D. C , Reasoner. J., & Williams, R. S. (1989). Cardiovascular and behavioral effects of aerobic exercise training in healthy older men and women. Journal of Gerontology: Medical Sciences, 44, M147-M157. Blumenthal, J. A., Emery, C. F., Madden, D. J., Schniebolk, S., Riddle, M. W., Cobb, F. R., & Coleman, R. E. (In press). Effects of exercise training on bone density in older men and women. Journal of the American Geriatrics Society. Blumenthal, J. A., Williams, R. S., Needles, T. L., & Wallace, A. G. (1982). Psychological changes accompany aerobic exercise in healthy middle-aged adults. Psychosomatic Medicine, 44, 529-536. Bradburn, N. M. (1969). The structure of psychological well-being. Chicago: Aldine, Buschke, H., & Fuld, P. A. (1974). Evaluating storage, retention, and retrieval in disordered memory and learning. Neurology I, 11, 10191025. Clarkson-Smith, L., & Hartley, A. A. (1989). Relationships between physical exercise and cognitive abilities in older adults. Psychology and Aging, 4, 183-189. Cunningham, D. A., Rechnitzer, P. A., Howard, J. H., & Donner, A. P. (1987). Exercise training of men at retirement: A clinical trial. Journal of Gerontology, 42, 17-23. Derogatis, L., Lipman, R. S., Rickels, K., Uhlenhuth, E. H., & Covi, L. (1974). The Hopkins symptom checklist (HSCL): A measure of primary symptom dimensions. In C. Pichot (Ed.), Psychological measurements in psychopharmacology: Modern problems in pharmacopsychiatry (Vol. 7). Basel: S! Karger. Dustman, R. E., Ruhling, R. O., Russell, E. M., Shearer, D. E., Bonekat, H. W., Shigeoka, J. W., Wood, J. S., & Bradford, D. C. (1984). Aerobic exercise training and improved neurophysiological function of older individuals. Neurobiology of Aging, 5, 35-42. Elsayed, M., Ismail, A. H., & Young, R. J. (1980). Intellectual differences of adult men related to age and physical fitness before and after an exercise program. Journal of Gerontology, 35, 383-387. Emery, C. F., & Blumenthal, J. A. (1990). Perceived change among participants in an exercise program for older adults. The Gerontologist, 30, 516-521. Emery, C. F., & Gatz, M. (1990). Psychological and cognitive effects of an exercise program for community-residing older adults. The Gerontologist, 30, 184-188.
George, L. K., Landerman, R., Blazer, D., & Melville, M. L. (1989). Concurrent morbidity between physical and mental illness: An epidemiologic examination. In L. L. Carstensen & J. Neale (Eds.), Mechanisms of psychological influence on physical health, with special attention to the elderly. New York: Plenum Press. Gitlin, L. N. (1985, March). Psychological effects of physical conditioning in the well elderly. Paper presented at the 2nd National Forum on Research in Aging, University of Nebraska, Lincoln, NE. Hughes, J. R., Casal, D. C , & Leon, A. S. (1986). Psychological effects of exercise: A randomized cross-over trial. Journal of Psychosomatic Research, 30, 355-360. Jones-Gotman, M., & Milner, B. (1977). Design fluency: The invention of nonsense drawings after focal cortical lesions. Neuropsychologia, 15, 653-674. Madden, D. J., Blumenthal, J. A., Allen, P. A., & Emery, C. F. (1989). Improving aerobic capacity in healthy older adults does not necessarily lead to improved cognitive performance. Psychology and Aging, 4, 307-320. Madden, D. J., Blumenthal, J. A., & Ekelund L-G. (1988). Effects of beta blockade and exercise on cardiovascular and cognitive functioning. Hypertension, 11, 470-476. Niinimaa, V., & Shephard, R. J. (1978). Training and oxygen conductance in the elderly. I. The respiratory system. Journal of Gerontology, 33, 354-361. Powell, R. R. (1974). Psychological effects of exercise therapy upon institutionalized geriatric mental patients. Journal of Gerontology, 29, 157-161. Radloff, L. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measures, 1, 385-401. Randt, C. T., Brown, E. R., & Osborne, D. P., Jr. (1980). A memory test for longitudinal measurement of mild to moderate deficits. Clinical Neuropsychology, 2, 184-194. Reitan, R. M. (1979). Manual for administration of neuropsychological test batteries for adults and children. Tucson: Reitan Neuropsychological Laboratories, Inc. Ruff, R. M., Evans, R. W., & Light, R. H. (1986). Automatic detection vs controlled search: A paper-and-pencil approach. Perceptual and Motor
EXERCISE AND PSYCHOLOGICAL FUNCTION
P361
Appendix Measures of Psychological Function
9. 10.
11.
12.
13.
14.
15.
to draw from memory a series of geometric shapes following a 10-second exposure. Higher scores indicate better memory. The Selective Reminding Test requires the subject to read a list of words and then to recall as many as possible. Higher scores are associated with better memory. The Digit Symbol Subtest of the WAIS-R is a paper-andpencil task that requires the subject to reproduce as many coded symbols as possible in a blank box beneath the digit with which it is paired within 90 seconds. More digits completed indicate better performance. The Trail Making Test (Part B) requires the subject to connect a series of consecutively numbered and lettered circles in alternating sequence (i.e., 1-A-2-B, etc.) as quickly as possible. Faster times indicate better performance. The 2 & 7 Test is a timed visual scanning task in which subjects must identify "target" stimuli (the digits 2 & 7) from an array of visually presented distractors. Faster times indicate better performance. The Nonverbal Fluency Test requires the subject to draw as many original, nonrepetitive designs as possible within a five-dot matrix within a 2-minute time limit. More items completed indicate better performance. The Verbal Fluency Test involves a series of 1-minute trials in which subjects are asked to generate as many words as possible that begin with a given letter. More items completed indicate better performance. The Stroop Color-Word Test measures the ease with which subjects can shift their perceptual set by accurately naming the color of the ink that the words red, blue, or green are printed in. Each word is printed in a different color ink than the color it is printed in. The number of correct (C) and incorrect (I) responses are recorded. More correct items indicate better performance.
Downloaded from http://geronj.oxfordjournals.org/ at University of Birmingham on June 6, 2015
1. The State-Trait Anxiety Inventory (STAI) is a 40-item questionnaire that measures levels of anxiety at the time of the assessment (state) and in general (trait). Higher scores are associated with greater anxiety. 2. The Center for Epidemiological Studies Depression Scale (CES-D) is a 20-item scale that assesses how many days in the past week the subject experienced symptoms associated with depression. Higher scores are associated with greater depression. 3. The Affect Balance Scale is a 10-item mood survey used to measure global positive and negative emotional states. Higher scores are associated with more positive affect. 4. The Hopkins Symptom Checklist (SCL-90) is a 90-item questionnaire designed to assess the degree of emotional distress and psychopathology. Five of 10 clinical scales were preselected to assess psychiatric symptoms, including somatization, obsessive-compulsive, phobic anxiety, paranoid ideation, and psychoticism. Higher scores are associated with greater emotional disturbance. 5. The Finger (Oscillation) Tapping Test requires the subject to tap the index finger on each hand as quickly as possible for two 30-second trials. Measures were obtained for both dominant and nondominant hands. Data from 15 and 30 seconds were recorded. More taps indicate better performance. 6. The Short Story Module of the Randt Memory Test requires the subject to recall the details of a short story immediately after it has been read to them and after a 30-minute delay. Higher scores are associated with better memory and attention. 7. The Digit Span Subtest of the Wechsler Adult Intelligence Scale-Revised (WAIS-R) requires the subject to repeat a series of digits that have been orally presented to them both forward and, in an independent test, in reverse order. The more digits recalled indicate better memory and attention. 8. The Benton Revised Visual Retention Test requires the subject
