Copyright 1998 by
The Gerontological Society of America
The Cerontologist
Vol. 38, No. 5, 602-609
Sedentary, overweight women aged 60-70 years were assigned either to a 16-week health
education group (/? = 70) in which they were instructed to exercise aerobically on their
own three times per week or to a 16-week exercise group (n = 76) that consisted of three
supervised aerobic sessions per week. Regression analyses performed at post-test and at
3-, 6-, and 18-months follow-up revealed that exercise frequency was not explained by
group membership, but often was negatively associated with placing a greater value on
exercising with peers. Those who were exercising more often at 6- and 18-months followup, however, valued the perceived benefits of exercise more highly. The findings of this
study suggest the importance of emphasizing strategies that stress the intrinsically
motivating benefits of exercise, as well as building into programs those aspects of
exercise valued by the participants.
Key Words: Physical activity, Exercise frequency, Perceived benefits of exercise
Older Women's Feelings About Exercise
and Their Adherence to an Aerobic
Regimen Over Time
Michael S. Caserta, PhD,1 and Patricia A. Gillett, PhD, RN:
A regular exercise program can be beneficial for
sedentary older women—especially those who are
overweight—because it can reduce their risks for cardiovascular disease, hypertension, diabetes, and musculoskeletal problems (U.S. Dept. of Health and Human
Services [USDHHS], 1990, 1996). Despite the known
importance of and preponderance of media attention to exercise, more than 60% of women over age
60 participate in little or no sustained physical activity
of at least moderate intensity (Caspersen & Merritt,
1992). When they do initially engage in exercise, adherence often drops dramatically after 6 months (Dishman, 1994; Gillett & Caserta, 1996). This has prompted
investigators to examine the factors associated with
adherence to exercise programs. At the intrapersonal
level, some investigations have revealed that those individuals with a greater sense of mastery or self-efficacy and those who are motivated by enjoyment or
satisfaction are more likely to maintain regular exercise programs (Avers & Wharton, 1991; Clark, 1996;
Lyons & Lachman, 1996; McAuley, 1993; McAuley,
Lox, & Duncan, 1993; Oman & McAuley, 1993).
Opportunities to socialize and enjoy a sense of camaraderie, for instance, can be primary motivating factors to attend exercise classes (Gillett, 1988). Furthermore, exercise behavior is potentially reinforced
when the participant perceives and enjoys its benefits (Emery & Blumenthal, 1990; Emery, Hauck, & Blumenthal, 1992; Gillett, 1988, 1993; Sharpe et al., 1997).
This project was funded by a grant from the National Institute of Nursing
Research (R29 NR02087).
1
Cerontology Center, University of Utah. Address correspondence to
Dr. Michael Caserta, University of Utah Gerontology Center, 10 S 2000 E
Front, Salt Lake City, UT 84112-5880. E-mail: [email protected]
J
College of Nursing, University of Utah, Salt Lake City.
602
Older women are often introduced to regular
exercise through programs such as community or
hospital-based health education classes, where an
exercise regimen is prescribed, or through supervised
exercise programs, such as aerobic dance or walking
clubs (Gillett, White, & Caserta, 1996; Gillett et al.,
1993; King, Haskell, Taylor, Kreamer, & DeBusk, 1991).
Program features that promote participation include
group activities where socialization, support, and a
sense of group cohesion and identity are developed,
and a sense of comfort is generated from exercising
among peers of similar gender, age, and fitness level
(Emery & Blumenthal, 1990; Gillett, 1988,1993; Henry,
1996; Mayer et al., 1994). Comfort is especially important for older overweight women who are reluctant to participate in exercise classes with people they
perceive as different from themselves (Gillett, 1988;
Knapp, 1988). Attendance also improves when classes
are held at convenient times and locations with no or
minimal cost, and the instructors are knowledgeable,
provide adequate feedback, and are seen as peers
with whom the participants can identify (American College of Sports Medicine [ACSM], 1995; Emery & Blumenthal, 1990; Franklin, 1994; Gillett, 1988; Gillett
et al., 1993; Henry, 1996; Mayer et al., 1994; Sepsis
etal., 1995).
Although these features may be applied to a variety of health promotion programs, supervised exercise programs that are time-limited or close-ended (such
as those conducted for research studies) can be especially problematic for the participants who become
attached to them. Participants in programs where homebased exercise is prescribed can continue to follow
the exercise prescription when the program ends; however, participants in supervised exercise programs need
The Gerontologist
to maintain their exercise routine without their previous leader and peer support. Finding group exercise opportunities in community-based exercise classes
usually involves adapting to a social and physical environment that is different from the classes they attended during a time-limited, close-ended program and
may even involve a different form of physical activity.
If any of these differences are perceived as inconvenient, threatening, or otherwise unappealing, it is
questionable whether an older woman will continue
to exercise regularly. Alternatively, once the participants perceive themselves as benefitting from exercise, will they be sufficiently motivated to continue
some form OT regular physical activity to sustain these
benefits if structured classes are not readily available
or accessible? Because each of these factors could
provide the motivation to adhere to a regular exercise
program, it is not entirely clear which are most important to an older woman's continued participation.
The purpose of this study, therefore, was to examine the perceived importance of four domains of the
exercise experience as reflected by short- and longterm adherence to a regular exercise regimen following one of two 16-week organized programs: a weekly
health education class that prescribed a self-selected
regular aerobic routine (e.g., walking) or a structured
exercise class that included three supervised aerobic
sessions per week. The four domains examined were
the perceived importance of exercising with peers, the
structural features of exercise programs, the experience of companionship and support during exercise,
and the perceived benefits of exercise.
excluded. Of 347 respondents, 182 met the study criteria, including 36 controls whose data are not reported here. The women were randomly assigned
to one of the three groups using a table of random
numbers. The participants were nonsmokers, obese
(mean % body fat = 40.7, SD = 4.3; mean body
mass index = 32 kg/m2, SD = 3.9), and reported only
light exercise (less than 60 minutes/week) for the previous 6 months. The mean baseline VO2max, which
is a measure of aerobic fitness, was 23.5 ml/kg"1/
min*1 (SD = 6.6), indicating a low initial fitness level.
The mean age was 64.4 years {SD = 3.0). Approximately 70% were married and 2 1 % were widowed.
Ninety-four percent graduated high school (mean
years of education = 14.0, SD = 2.2) and the median annual household income category was $20,000$29,000. Although a variety of techniques were used
to recruit participants and recruitment covered a wide
geographical area, 98% of the participants were nonHispanic White women (6% of the older women in
the study community are minorities). This predominantly non-Hispanic White sample is typical of many
exercise studies to date (King, Haskell, Young, Oka, &
Stefanick, 1995).
As the data in Table 1 indicate, the EX and ED
groups were highly similar in major background variables. No significant differences were observed with
respect to their ages, educations, household incomes,
marital status, body compositions, or fitness levels.
Table 1. Sociodemographic, Body Composition, and Fitness
Level Comparisons Between Exercise (EX; n = 76)
and Education-Only (ED; n = 70) Groups at Baseline3
Methods
The data reported here are from a randomized, controlled trial that tested the short- and long-term effects
of two 16-week, nurse-delivered exercise/education
programs specifically designed for obese, older women:
one group of participants received health and fitness
education only (ED; n = 70), and the other group
received health and fitness education with aerobic
training (EX; n = 76). The larger study also included a
nonintervention control group whose data were not
used for this article. The controls did not keep physical activity records within the larger study because
there was concern that by doing so they would change
their normal activity patterns (see Measures section).
The intervention's effects on physiological and fitness
outcomes are reported elsewhere (Cillett & Caserta,
1996; Cillett et al., 1996). These earlier articles also
contain detailed descriptions of the overall study design; the major features germane to this report are
highlighted below.
nirttini
Variable/Croup
Age
EX
ED
ir\|
1 IC
M
(SD)
t
64.7
64.1
(2.9)
(3.2)
1.02
14.0
13.9
(2.3)
(2.1)
0.39
40.8
40.7
(4.4)
(4.2)
0.16
31.7
32.4
(4.1)
(3.7)
1.12
23.9
23.1
(5.3)
(7.7)
0.66
%
Square
54.3
56.7
0.08
68.1
71.0
0.15
Years Education
EX
ED
% Body Fat
EX
ED
Body Mass Index (kg/m2)
EX
ED
VO2max (ml/kg~1/min~1)
EX
ED
Annual Household
Income <$30,000
EX
ED
Married
Sample
Community-dwelling, sedentary, obese women aged
60-70 years were recruited through newspaper and
radio advertisements. Women whose chronic health
or mobility problems precluded vigorous walking were
Vol. 38, No. 5,1998
Categorical
/-,Ji ILIIIUUI
603
EX
ED
a
T tests were performed on continuous variables and chi-square
tests were performed on categorical variables. All differences were
nonsignificant.
Study Protocol
Both interventions (EX and ED) included one weekly
health and fitness education class during which the
women were instructed in 16 specific health and fitness topics salient to their age group over the 16-week
study period. A seminar/discussion teaching format
with visual aids, exercise demonstrations, and handouts was used. Women received careful instruction
on principles of warm-up, cool-down, and flexibility.
The EX and ED groups followed identical educational
protocols, but each group followed different exercise
protocols. The EX group received 1 hour of supervised, class-based, low-impact aerobic dance exercise
3 days per week for 16 weeks. Exercise duration and
intensity progressed gradually over the 16 weeks until
women were exercising at a moderate intensity for
about 30 minutes per session. Each member of the
ED group selected an aerobic exercise of her own
choice and exercised on her own. To guide their homebased program, the ED women were given an exercise prescription based upon ACSM (1995) guidelines
and their baseline fitness test results. They were also
taught self-monitoring techniques and how to select
an appropriate exercise intensity level. The ED group
was instructed to increase exercise intensity and
duration gradually so that, at the end of 16 weeks,
they were exercising at least 3 days per week at
moderate intensity for 30 minutes per session. Both
intervention groups were led by experienced geriatric
nurse practitioners.
Data collection occurred over five measurement
periods. These data points were baseline, post-test
following the 16-week intervention, and at 3-, 6-, and
18-months follow-up.
Companionship and support during exercise refers to
the support and approval of others during the exercise experience. The individual items, factor loadings,
and alpha coefficients (ranging from .74. to .83) associated with each subscale are presented in Table 2.
The possible ranges for each of the subscales are as
follows: perceived benefits (6-30), peer group factors
and structural features (both 4-20), and companionship and support (5-25). In each instance, a higher
score indicates greater importance placed on that
particular domain of exercise experience. Pearson
product-moment coefficients generated among the
subscales ranged from .23 (perceived benefits with
companionship and support) to .50 (peer group factors with structural program features). The overall
average correlation among the subscales was .38, which
indicates that although some variance was shared, a
reasonable amount of independence separated each
of the domains.
Activity Records—The extent to which the women
exercised regularly was measured through written, selfreport, 7-day physical activity records (PARs). These
were maintained throughout the intervention period
and through 18 months postintervention. The PARs
were used to record all aerobic exercise. The women
recorded the type, frequency, and duration (in minutes) of aerobic activity (if any) they performed each
week. Aerobic was defined as exercise at 60%-80%
Table 2. Principal Components Factor Analysis of Feelings
About Exercise Scale (GFES)a: Subscale Items,
Factor Loadings, and Alpha Coefficients
Subscale
Measures
Feelings About Exercise—At each data point, the
women completed the Gillett Feelings About Exercise
Scale (GFES), a self-administered scale that consists of
19 Likert-type items representing perceptions about
the exercise experience. The items were derived from
previous work of the principal investigator concerning
the responses of older and middle-aged women to
exercise programs (Gillett, 1988, 1993; Gillett & Eisenman, 1987). For each item, the respondent rated how
important that facet of exercise was for her on a scale
from 1 (not at all important) to 5 (very important).
As a data reduction strategy, the 19 items were
subjected to a principal components factor analysis
with oblique rotation (Direct = Oblimin) that yielded
four factors. Each factor represents a subscale measuring a domain of the exercise experience that one
may feel is important. These four domains are perceived benefits, peer group factors, structural features
of an exercise program, and companionship and
support during exercise. Perceived benefits refers to
feelings of increased health and energy and positive
changes in body composition. Peer group factors refers to the comfort that accompanies exercising with
similar others. Structural features of an exercise program refers to leader characteristics and class format.
604
Items
Loading
Perceived Benefits
(.83)b
Makes me feel strong
Gives me energy
Makes me healthier
Changes my body size
Changes my body shape
Helps with depression
Peer-Group
Factors (.78)b
Exercise with people my
own age
Exercise with people who are
in same physical shape as I am
Exercise with people my own sex
Exercise leader my own age
Structural Features
of Program (.75)b
Leader with health-related
background
Leader with fitness background
Structured exercise classes
Exercise program that is
time-limited
Companionship
Exercise with my spouse
and Support (.74)b Exercise with co-workers
Exercise with friends
Approval of others in class
Support of others in class
.82
.81
.77
.71
.67
.57
.81
.82
.82
.38
.83
.83
.60
.49
.81
.65
.61
.60
.51
a
Oblic<ue Rotation (Direct = Oblimin).
Cronbach alpha coefficient in parentheses.
b
The Gerontologist
maximal heart rate reserve (MHRR). PARs were turned
in weekly and averaged to ascertain a mean activity
level for each data interval. The data intervals were
(a) from the beginning of the classes until post-test,
(b) from post-test to 3-month follow-up, (c) from 3month to 6-month follow-up, and (d) from 6-month
to 18-month follow-up.
Results
Retention was high throughout the intervention
period (91%) but tended to moderate somewhat at
3 and 6 months (84% and 75%, respectively). Retention then declined to 62% by the 18-month followup. These retention rates are similar to those reported
by King and colleagues (1995), who conducted a similar
long-term, community-based exercise study. Although
every effort was made to minimize drop out, reasons
for drop out were recorded and tracked. The reasons
for drop out were: the presence of non-exerciserelated medical problems (n = 22), loss of interest
(n = 19), conflicting time and work pressures (n =
15), relocation (n = 5), family illness (n = 5), vacation (n = 1), and death (n = 1).
Before examining the relative impact of each domain of feelings about exercise on short- and longterm exercise frequency, we determined how these
domains changed or remained stable over time and
whether they were associated with the type of organized program (EX vs ED) to which the women were
exposed. Table 3 contains the subscale means and
standard deviations for the EX and ED groups as well
as the results of the repeated measures analyses of
variance (ANOVAs) conducted over the five data points.
The women tended to place at least moderate importance on each domain of exercise experience, as
the mean scores generally were near or exceeded the
midpoint for each of the scales. The EX and ED groups
differed, however, in the importance they placed on
peer-group factors (F = 13.63, p < .001) and the
structural features of the program (F = 21.20, p <
.001). In both cases, EX participants placed significantly greater emphasis on these aspects of exercise
than did ED participants. Both groups changed over
time with respect to these two factors: the importance of the factors increased immediately following
the 16-week program, but declined to near baseline
levels thereafter. The lack of a significant group by
time interaction indicates that both EX and ED group
means increased and then decreased similarly with
respect to these two domains.
No group effects were observed with respect to the
importance placed on perceived benefits and on companionship and support while exercising, although
changes over time were observed for these factors (both
p < .01). These changes represented a pattern of
increasing importance at 3 months postintervention,
followed by gradual decrease through the remainder
of the study. Again, these changes were independent
of group membership because the group by time interactions were not statistically significant.
Table 4 presents the exercise frequency for both
groups over the four data intervals of the study.
Although there was an overall decline in exercise
frequency by the 18-month follow-up, the decline
for the EX women was greater than for the ED women
(F = 4.49, p < .01). During the 6 months following
the intervention, those in the EX group gradually re-
Table 3. Repeated Measures Analyses of Variance of GFES Subscales: Comparisons Between Exercise (EX)
and Education Only (ED) Groups Across Five Time Points From Baseline to 18 Months Post-Intervention
Follow-up
Effects"
Post-Test
3 months
(SD)
M
(SD)
M
Perceived Benefits3
EX
25.3
ED
23.5
(3.4)
(5.3)
25.4
24.7
(3.2)
(3.2)
Peer Croup Factors6
EX
14.7
ED
12.6
(3.2)
(4.3)
17.1
14.2
Structural Features of Program15
EX
15.9
(3.2)
ED
12.3
(2.8)
Companionship and Support0
EX
13.6
(4.7)
ED
12.8
(4.4)
Baseline
Subscale/Croup
M
6 months
18 months
(SD)
M
(SD)
M
(SD)
Croup
Time
Group x
Time
23.8
24.1
(4.9)
(3.8)
24.4
24.8
(4.4)
(4.0)
23.2
23.6
(4.4)
(4.3)
0.11
4.63*
1.49
(2.4)
(4.0)
16.0
12.6
(3.4)
(4.3)
15.3
12.4
(4.1)
(4.5)
14.9
11.8
(3.8)
(4.9)
13.63**
6.66**
0.52
16.9
14.1
(2.6)
(3.1)
15.9
13.9
(2.5)
(2.7)
15.4
13.2
(3.3)
(3.5)
15.2
11.6
(3.3)
(4.2)
21.20**
4.38*
1.64
15.1
13.5
(4.2)
(4.9)
13.3
13.1
(4.4)
(4.3)
12.6
12.1
(4.4)
(4.6)
13.6
12.8
(4.7)
(4.4)
0.53
3.38*
0.47
*p < .01; * * p < .001.
"Range = 6 (not at all important)-30
b
Range = 4 (not at all important)-20
c
Range = 5 (not at all important)-25
d
F ratio based on repeated measures
Vol. 38, No. 5, 1998
(very important).
(very important).
(very important).
analyses of variance.
605
Table 4. Differences in Average Exercise Frequency (Days/Week) Between Exercise (EX)
and Education-Only (ED) Groups Over Four Data Intervals'
EX
ED
Data Intervals
nb
%b
M
(SD)
nb
%b
M
(SD)
Up to post-test
Post-test to 3-mo. follow-up
3 mo. to 6-mo. follow-up
6 mo. to 18-mo. follow-up
69
57
48
51
90.7
75.0
63.0
67.1
4.0
3.3
2.5
2.6
(0.9)
(1.4)
(1.4)
(0.9)
53
44
41
41
75.7
62.8
58.6
58.6
3.7
3.6
3.1
2.8
(1.3)
(1.4)
(1.6)
(1.4)
a
Based on repeated measures ANOVA: F (Group) = 0.88, n.s.; F(Time) = 47.74, p < .001; F (Croup x
Time) = 4.49, p < .01.
b
Number and percent reporting.
duced their frequency of exercise by 1.5 days per week,
whereas those in the ED group decreased their exercise frequency by only .5 day per week, remaining
virtually unchanged during the initial 3-month period
following post-test. At the 18-month follow-up, however, both groups declined to a similar exercise frequency level of slightly more than 2.5 days per week.
Throughout the study, the most common form of
aerobic exercise reported by the ED group was walking (followed by riding a stationary bicycle). The structured aerobic exercise dance program comprised the
majority of activity reported by the EX group during
the intervention. Thereafter, most of the women in
that group participated in alternate forms of aerobic
exercise, walking being the most common (Gillett &
Caserta, 1996). Regardless of the activity in which they
engaged, the women in both groups reported exercising at their training heart rate an average of 33
minutes per session throughout the study.
To determine how changes in feelings about the
exercise experience influenced the frequency of regular
aerobic exercise over time, hierarchical regressions
were performed at post-test and at 3-, 6-, and 18month follow-ups. In each equation, the dependent
variable was the average exercise frequency for the
data interval concluding with that time point. For example, the frequency in the post-test equation is the
average number of days per week exercised for the
period covering the beginning of the classes to posttest; the 3-month equation represents that interval
between post-test and the 3-month follow-up, and
so on. The independent variables were entered in
two blocks: the treatment variable, which was dummy
coded (EX = 1), was entered into the equations first,
followed by the GFES subscales entered as the
second block. Residuals were analyzed and plotted
against predicted scores for each of the equations and
the assumptions of linearity, normality, and homoscedasticity appeared unviolated. Furthermore, tolerance
and variance inflation factors were at levels that ruled
out problems associated with multicollinearity. The
estimates from the regression equations are presented
in Table 5.
With the exception of a trend at post-test (b = .47,
p = .08), exercise frequency generally was not
dependent upon whether the women were in the EX
or ED group. The trend at post-test is most likely explained by the high attendance of the structured exercise classes (approximately 86%). Also, nearly 78%
of the EX women reported exercising more than 3
days per week (versus 54.5% for ED). This trend only
accounted for 2.6% of the variance at post-test and
did not persist at later points in the study. In three of
the four equations, peer-group factors had an inverse
influence on exercise frequency (all p < .05), where
those who placed greater importance on exercising
Table 5. Effects of Group Assignment (EX vs ED) and Feelings About Exercise on Exercise Frequency Over 18 Months:
Estimates From Hierarchical Regression Equations' at Post-Test, and 3, 6, and 18 Month Follow-Ups
Equations
Post-Test
Independent Variables
Group Assignment15
Perceived Benefits
Companionship & Support
Peer Group Factors
Structural Features of Program
6 months
3 months
18 months
b
beta
b
beta
b
beta
b
beta
A7 +
.01
.05
-.10 *
.03
.19
.04
.19
-.29
.06
-.15
.03
.04
-.12*
.04
-.05
.08
.05
-.35
.10
-.57
.10*
-.03
-.02
-.11
-.19
.29
-.10
-.06
-.27
-.06
.09*
-.05
-.14*
.08
-.02
.35
-.15
-.51
.26
.092
.099
.200
.191
+
p = .08 (trend); *p < .05.
"Order of inclusion: block 1 (group assignment), block 2 (GFES subscales).
b
Dummy coded (1 = EX).
606
The Gerontologist
with similar others tended to exercise less frequently
than those for whom this was less important. Conversely, those who were exercising more often at the
6- and 18-month follow-ups {b = .10 and .09, respectively, p < .05), tended to place a greater importance on the perceived benefits of exercise. At no
data point did the importance of companionship and
support or the structural features of an exercise program explain exercise frequency in this study. The variance in exercise frequency explained by each of the
equations ranged from 9.2% (post-test) to 20% (6month follow-up).
Discussion
Consistent with earlier studies, there was a decrease
in the amount of weekly aerobic activity following
the intervention period (Dishman, 1988). The biggest
decline in exercise frequency was reported by the
women in the EX group during the 6 months following post-test. Much of this was probably due to the
lack of structured classes in the community that were
convenient, accessible, and in an environment comfortable for the women. A typical comment made
by an EX woman was "I don't make time unless I
have a specific class to go to." Consequently, many
in this group took up alternative activities (e.g., walking), which they may not have perceived as enjoyable
("It's hard to do alone"). Following the intervention,
ED women had little difficulty adopting a home-based
exercise regimen, because they all had done so
since the beginning of the study. Irrespective of which
group they were in, the women who appreciated and
valued the benefits of regular exercise had a greater
likelihood of maintaining a regular aerobic routine over
time. Adherers from both groups reported an increased
desire to improve/maintain their health and physical
function as demonstrated by comments such as "I want
to live versus exist" and "I can climb stairs now and
my knees don't hurt." They also reported perceived
emotional benefits ("I feel happier now than I have in
the past 5 years"; "I have a new lease on life") and
positive body changes ("My husband says I look better"; "I'm wearing a smaller dress size").
Both EX and ED interventions were associated with
a temporary increase in how important the women
perceived all facets of the exercise experience to be.
This tended to moderate as time passed. The women
in the EX group consistently valued peer-group and
program-structure factors more highly than those in
the ED group, most likely because these features
played a more integral role in the 16-week exercise
class compared with the 16-week health and fitness
education classes. The supervised exercise classes (EX)
reinforced the sense of importance of peer-group
factors and the structural features of the program. EX
women reported making many new friends and enjoyed "being with women my own age, size and who
are as out of shape as I am." Others reported that the
"women were upbeat and fun to be with." Program
factors such as the gradual exercise progression ("I felt
I was exercising at a safe level") and leader characteristics ("She understands my health and weight probVol. 38, No. 5,1998
607
lems") were particularly attractive to many of the women
in these exercise classes.
The amount of social contact differed between the
two groups during the intervention period: EX women
had contact with each other three times per week,
compared with once a week for those in the ED group.
Despite these differences, both groups placed similar
importance on companionship and support while
exercising whether in a structured class or at home.
Furthermore, this factor did not explain overall exercise frequency at any point in this study nor did the
structural features of exercise programs factor. These
factors probably initially enhanced attendance while
the programs were ongoing, but did not have a lingering effect on subsequent exercise behavior outside
the classes or after they ended. In three out of four
data points, the women who placed greater value on
exercising with peers tended to exercise less than those
for whom this was less important. It is possible that
these were women who felt less comfortable and were
less motivated to exercise when they were either alone
or with others whom they did not perceive as similar
in terms of fitness level, oody shape, gender, or age.
In these instances, the need for exercising with one's
peers appeared to be linked closely to the sense of
insecurity and anxiety many older overweight women
experience as they attempt to exercise among those
who are thinner, younger, or in better physical condition, or when they attempt to exercise by themselves
(Gillett, 1988; Knapp, 1988). It appears that exercising with like others may be an important aspect during the adoption or early phase of exercise behavior
change, particularly for this group (Clark, 1996; Gillett,
1988).
Alternatively, those who valued the perceived
benefits of exercise such as improved health and function tended to exercise more, particularly over the
long term. This fits well within the concept of intrinsic motivation as a predictor of exercise adherence
(Oman & McAuley, 1993). A key component of intrinsic motivation is self-determination where one exercises not because it is imposed (e.g., ordered by a
physician) but out of internal motives. The benefits
such as feeling stronger, healthier, and having more
energy clearly could be perceived as intrinsically motivating. Although the potential health outcomes are
often cited as reasons to initiate exercise, perceived
satisfying benefits are usually better explanations
for why exercise programs are maintained over time
(Oman & McAuley, 1993).
There are some caveats that must be noted in interpreting the findings of this study. First, because the
activity records were self-reported data, they could be
subject to social desirability bias. Some women did
not turn in PARs each week, and we were sometimes
unable to determine if it was because they did not
exercise during that time or if it was simply missing
data. Even if aggregated, however, the weekly recording of physical activity is generally preferable to more
global assessments that cover similar periods of time,
largely because problems associated with recall are
minimized (Ransdell & McMillen, 1997). Following the
intervention, we did not explicitly prescribe a prede-
termined frequency and duration of aerobic exercise.
The purpose of the activity records in this study was
to determine merely if and how much the women
exercised aerobically over the course of the study, rather
than to what extent they followed a set activity prescription. Adherence to an aerobic regimen was more
loosely defined as frequency and duration of any
aerobic activity versus percentage of a prescribed
activity.
Another limitation was that no quantitative measure of self-efficacy, an often-mentioned predictor of
exercise maintenance (McAuley, 1993; McAuley et al v
1993), was used in this study. Comments from women
in both groups suggested that they felt confident in
their ability to exercise aerobically for 25-30 minutes
at a time (Gillett et al., 1996), but it was not possible
to account for these feelings of self-efficacy in the statistical analyses. Future studies should incorporate a
measure of self-efficacy along with the program factors that were examined here especially because a
large proportion of the variance was left unexplained.
Finally, this study was conducted with a predominantly White sample, which precluded examining
ethnic or racial influences on exercise behavior. Although his findings were derived from a small sample
of older African American women who participated
in a focus group, Clark (1996) reported that his participants viewed exercising with similar others and
beneficial health expectations as important considerations, similar to what we found in this study. Consequently, our findings may not be overly constrained
by the paucity of non-Whites in the sample. The need
clearly exists, however, to explore feelings about
exercise and exercise behavior among ethnically and
racially diverse samples in future studies.
The findings in this study do lend themselves to
some recommendations for those who plan to design
organized exercise programs, whether prescriptive or
supervised, for older women. First it is important
to stress the benefits of exercise—not only the wellknown health benefits, but particularly those related
to functioning and quality of life, which the women
in this study found satisfying. Comments such as "I
can do more yard work" exemplify the promising effects that regular exercise can have on older adults'
functioning and quality of life, such as better performance of activities of daily living, greater enjoyment
of hobbies and recreation, feeling more energetic, and
a greater overall sense of well-being (Stewart & King,
1991). Possible strategies to emphasize the intrinsically
motivating features of these benefits could include
forms of self-monitoring and journal-keeping that help
the participants focus on those aspects of regular exercise they find enjoyable and reinforcing (Kanfer &
Caelick, 1986).
For older women who value exercising with peers,
attempts should be made to identify resources in the
community where they may find peer-supported programs. If none are available, one possible approach
could involve implementing a "buddy system" where
women can continue to exercise in groups of two
or three (O'Brien Cousins & Burgess, 1992). In the
absence of other alternatives, home-based programs
608
could be designed so that if no peers are readily available as exercise partners, women can continue some
form of physical activity in the security of their residence (King et al., 1995). To better facilitate women
who move from a group-based exercise program to a
self-directed, home-based regimen, we suggest a transition and maintenance period to ease women into a
different way of exercising.
Finally, it is apparently useful to identify those aspects of exercise programs and the exercise experience that are important to the participants. In this way,
the program can be individualized (Mills, Stewart,
Sepsis, & King, 1997) whenever possible to accent
what is motivating and appealing. Future research
should focus on examining creative mechanisms that
support those features of exercise that are valued by
the participants. Greater understanding of the aspects of the exercise experience that are sources
of motivation would more effectively facilitate older
adults' long-term adherence to regular aerobic exercise programs.
References
American College of Sports Medicine (ACSM). (1995). Guidelines for
exercise testing and prescription (5th ed). Baltimore, M D : Williams
& Wilkins.
Avers, D., & Wharton, M. A. (1991). Improving exercise adherence:
Instructional strategies. Topics in Geriatric Rehabilitation, 6, 62-73.
Caspersen, C. J., & Merritt, R. K. (1992). Trends in physical activity patterns among older adults: Behavior risk factor surveillance system
survey, 1986-90 [Abstract]. Medicine & Science in Sports & Exercise, 24, S26.
Clark, D. O. (1996). Age, socioeconomic status, and exercise selfefficacy. The Gerontologist, 36, 157-164.
Dishman, R. K. (1988). Overview. In R. K. Dishman (Ed.), Exercise adherence (pp. 1-9). Champaign, IL: Human Kinetics.
Dishman, R. K. (1994). Motivating older adults to exercise. Southern
Medical Journal, 87(5), S79-S82.
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., Hauck, E. R., & Blumenthal, J. A. (1992). Exercise adherence or maintenance among older adults: 1-year follow-up study.
Psychology and Aging, 7, 466-470.
Franklin, B. A. (1994). Program factors that influence exercise adherence: Practical adherence skills for clinical staff. In R. K. Dishman
(Ed.), Exercise adherence: Its impact on public health (8th ed., pp.
237-258). Champaign, IL: Human Kinetics.
Gillett, P. A. (1988). Self-reported factors influencing exercise adherence in overweight women. Nursing Research, 37, 25-29.
Gillett, P. A. (1993). Senior women's fitness project: A pilot study. Journal of Women and Aging, 5, 49-66.
Gillett, P. A., & Caserta, M. S. (1996). Changes in aerobic power, body
composition, and exercise adherence in obese, postmenopausai
women six months after exercise training. Menopause, 3, 126-132.
Gillett, P. A., & Eisenman, P. (1987). The effects of intensity-controlled
aerobic dance exercise on aerobic capacity of middle-aged, overweight women. Research in Nursing and Health, 10, 383-390.
Gillett, P. A., Johnson, M., Juretich, M., Richardson, N., Sleagle, L, &
Farikoff, K. (1993). The nurse as exercise leader. Geriatric Nursing,
14, 133-137.
Gillett, P. A., White, A. T., & Caserta, M. S. (1996). Effect of exercise
and/or fitness education on fitness in older, sedentary, obese women.
Journal of Aging and Physical Activity, 4, 42-55.
Henry, J. (1996). Promoting exercise behaviors in older adults. The Health
Education Monograph Series, 14, 38-42.
Kanfer, F. H., & Gaelick, L. (1986). Self-management methods. In F. H.
Kanfer & A. P. Goldstein (Eds.), Helping people change: A textbook
of methods (3rd ed., pp. 283-345). New York: Pergamon Press.
King, A. C , Haskell, W. L, Taylor, B., Kreamer, H. C , & DeBusk, R. F.
(1991). Group- v. home-based exercise training in healthy older men
and women. Journal of the American Medical Association, 266, 15351542.
King, A. C , Haskell, W. L, Young, D. R., Oka, R. K., & Stefanick, M. L.
(1995). Long-term effects of varying intensities and formats of physical activity on participation rates, fitness, and lipoproteins in men
and women aged 50 to 65 years. Circulation, 91, 2596-2604.
The Gerontologist
Knapp, D. N. (1988). Behavioral management techniques and exercise
promotion. In R. K. Dishman (Ed.), Exercise adherence: Its impact
on public health (pp. 203-235). Champaign, IL: Human Kinetics.
Lyons, H. W., & Lachman, M. E. (1996, November). Who exercises in
later life? Paper presented at the 49th Annual Meeting of the Cerontological Society of America, Washington, DC.
Mayer, J., Jermanovich, A., Wright, B. L, Elder, J. P., Drew, J. A., &
Williams, S. J. (1994). Changes in health behaviors of older adults:
The San Diego Medicare preventive health project. Preventive Medicine,
23, 127-133.
McAuley, E. (1993). Self-efficacy and the maintenance of exercise participation in older adults. Journal of Behavioral Medicine, 16, 10SMcAuley, E., Lox, C , & Duncan, T. E. (1993). Long-term maintenance
of exercise, self-efficacy, and physiological change in older adults.
Journal of Gerontology: Psychological Sciences, 48, P218-P224.
Mills, K. M., Stewart, A. L., Sepsis, P. C , & King, A. C. (1997). Consideration of older adults' preferences for format of physical activity.
Journal of Aging & Physical Activity, 5, 50-58.
O'Brien Cousins, S., & Burgess, A. (1992). Perspectives on older adults in
physical activity and sports. Educational Gerontology, 18, 461-481.
Oman, R., & McAuley, E. (1993). Intrinsic motivation and exercise behavior. Journal of Health Education, 24, 232-238.
Ransdell, L. B., & McMillen, B. (1997). Uses and limitations of physical
activity questionnaires in health education. Journal of Health Education, 28, 182-186.
Sepsis, P.C., Stewart, A. L, McLellan, B., Mills, K., King, A. C , & Shoumaker,
W. (1995). Seniors' ratings of helpfulness of various program support
mechanisms utilized in a physical activity promotion program. Journal of Aging and Physical Activity, 3, 193-207.
Sharpe, P. A., Jackson, K. L., White, C , Vaca, V. L, Hickey, T., Cu, J.,
& Otterness, C. (1997). Effects of a one-year physical activity intervention for older adults at congregate nutrition sites. The Gerontologist, 37, 208-215.
Stewart, A. L., & King, A. C. (1991). Evaluating the efficacy of physical
activity for influencing quality-of-life outcomes in older adults. Annals of Behavioral Medicine, 13, 108-116.
U.S. Department of Health and Human Services. (1990). Healthy People
2000: National Health Promotion and Disease Prevention Objectives. Washington DC: U.S. Government Printing Office.
U.S. Department of Health and Human Services. (1996). Physical activity and health: A report of the Surgeon General. Atlanta, CA: Centers for Disease Control and Prevention, National Center of Chronic
Disease Prevention and Health Promotion.
Received October 28, 1997
Accepted July 1, 1998
BE PART OF THE NEW VA
VETERANS INTEGRATED SERVICE NETWORK
(VISN11)
Extended Care Service Line Manager applications are invited for the position of Senior Director for the Veterans Integrated Service Network (VISN 11) located in Ann Arbor, Michigan, and serving medical centers and clinics in Michigan, Indiana, Illinois, and Ohio. The
incumbent leads network and care site managers in coordinating, planning and monitoring all extended care service activities, including service delivery, education, and research. Coordinates quality management activities including risk management, credentialing and privileging, accreditation, and provider profiling. The incumbent will establish
and implement tactical and strategic plans and process, ensuring clinical and administrative services are appropriately integrated to provide cost effective, high quality extended
care service.
Applicants must be U.S. citizens, holding full and unrestricted physician license(s) from any
state and must possess a broad knowledge and understanding of health care policies,
missions and operating programs, and extensive and specialized knowledge of health
care management. Salaries for this position range from $100,000 to $150,000.
Applications will be accepted until the position is filled. For application instructions, contact Paul J. Serefine, Assistant Personnel Officer, VA Medical Center, Ann Arbor, Michigan,
at 734-769-7100, extension 5127. To obtain further information about this position, please
contact Barbara A. McLelland, CHE, Director of Operations, VISN 11, at 734-930-5990.
Vol. 38, No. 5, 1998
609