Journal ofGerontology: MEDICAL SCIENCES Copyright I999 by The Gerontological Society ofAmerica

1999, Vol. 54A, No.9, M456-M466

Measuring Accumulated Health-Related Benefits

of Exercise Participation for Older Adults:
The Vitality Plus Scale
Anita M. Myers,'> Olga W. Malott,' Elana Gray,' CatrineTudor-Locke,'
NancyA. Ecclestone,? Sandra O'Brien Cousins,' and Robert Petrella-

Downloaded from https://academic.oup.com/biomedgerontology/article/54/9/M456/545637 by guest on 02 November 2020

'Department of Health Studies and Gerontology, University of Waterloo, Ontario, Canada.
-Centre for Activity and Ageing, University of Western Ontario, London.
3Faculty of Physical Education and Recreation, University of Alberta, Edmonton.

Background. Existing measures fail to capture the perceived benefits attributed to exercise participation by older adults
themselves. Noticeable improvements in sleep, energy level, bodily aches and pains, constipation, and other psychophysical as-
pects of "feeling good" may represent ongoing sources of motivation for continued participation. The Vitality Plus Scale (YPS)
was developed to measure these potential health-related benefits of exercising. .

Methods. The lO-item VPS was developed using an inductive approach, in collaboration with regularly exercising older
adults and their instructors. Multiple samples of exercisers and nonexercisers ranging in age from 40 to 94 were used to exam-
ine the reliability and validity of the new scale.

Results. The VPS showed good internal consistency and test-retest reliability over one week. Scores were able to discrimi-
nate on the basis of various indicators of health status and self-reported level of physical activity, and were related to two mea-
sures of functional mobility. Convergence was found with several subscales of the SF-36, whereas low correlations emerged
with a measure of episode-specific sensations. Responsiveness to change was found with various types of exercise for individu-
als with low to moderate scores prior to participation.

Conclusions. Improvements in sleep, energy level, mood, and generally feeling good appear to be the most noticeable
benefits of exercising for many adults. These associations are reinforced by sustained exercise participation. Capturing these
interrelated psychophysical constructs in a single, short measure will enable exercise researchers and instructors to measure in-
cremental improvements previously reported only anecdotally.

OMMUNITY exercise programs tailored to older adults
C are proliferating, and evidence is emerging that such pro-
grams have high rates of adherence (1). We are just beginning
to explore what the exercise experience means to older adults.
Like other leisure pursuits, recreational physical activity can
provide a diversion from daily routines and stresses, fun and
enjoyment, companionship, and a sense of accomplishment
(2-4). In addition, exercise can have positive physiological ef-
fects (4-12).
Changes in fitness parameters, such as improved aerobic ca-
pacity or muscle strength, are not directly observable to most
people (5,8,9) and are not predictive of exercise maintenance
for older adults (12). And, while it is widely believed that regu-
lar exercise contributes to overall health, well-being, and qual-
ity of life, measuring such outcomes has proven challenging
(4-15). More than 85 different psychological scales have been
used in exercise studies (6), indicating that no one measure is
considered the standard for the field. Measures such as life sat-
isfaction or self-esteem may be too global, and measures of
negative affect or psychiatric symptomatology such as anxiety
or depression may not be appropriate for psychologically
healthy adults (5,6,9,13,14). Existing psychological scales were
not designed for the exercise experience and fail to capture the
perceived benefits noted by exercisers themselves (6,7,13-15).
This article presents a new scale specifically designed to mea-
sure the accumulated psychological and physical benefits of ex-
ercise participation experienced by older adults.
The most frequently cited reason for engaging in leisure time
physical activity, according to population surveys, is to "feel
better mentally and physically" (3). Personal testimonials of
older participants in community exercise programs (15-17) and
anecdotal reports from older subjects in exercise studies
(13,14,18-20) include statements such as: "feeling better,"
"sleeping better," "moving better," "more relaxed," "more en-
ergy," and "less stiffness." These attributions may represent
sources of motivation for ongoing exercise participation for
many older adults. However, the bias inherent in retrospective,
perceived change ratings underscores the need for pre- and
post-administrations of standardized measures to examine both
immediate and accumulated benefits of exercising (6,7,9,13).
Two measures have been developed with college students to
assess affective states immediately following an exercise ses-
sion. The Subjective Exercise Experiences Scale [SEES (21)]
consists of 12 items grouped into three dimensions: Positive
Well-Being (strong, great, positive, terrific), Psychological
Distress (crummy, awful, miserable, discouraged), and Fatigue
(exhausted, fatigued, tired, drained). The Exercise-Induced
Feeling Inventory [EFI (7,22)] consists of 12 slightly different

M456
 VITALITY PLUS SCALE
items: enthusiastic, upbeat, happy, energetic, refreshed, revived,
fatigued, worn out, tired, calm, peaceful, and relaxed. Scores on
both scales were sensitive to change following an acute bout of
aerobic exercise (7,21). Affective states also appear to be influ-
enced by the social context-exercising in a group versus alone
in the laboratory (22).
Schneider (23) suggests that individuals interpret physio-
logic/somatic (e.g., breathing, perspiration, muscle movement,
and soreness) and cognitive/emotional sensations (e.g., feeling
energized, pleasantly tired, having fun) within the social/environ-
mental context of each exercise episode. Through a feedback
loop, consistent with Leventhal and associates' self-regulation
theory (24), people form more general interpretationsof the exer-
cise experience which, if positive, should foster continued par-
ticipation (23). Interviews with older women indicated that
these regular exercisers were aware of both bodily sensations
(such as feeling "warm and sweaty") and social cues ("cama-
raderie") during exercising. Other themes that emerged sug-
gested a carryover effect beyond the immediate exercise ses-
sion: "It energizes you for the rest of the day"; "It keeps the
joints moving"; "You feel more alert and relaxed." Many of
these women spontaneously reported differences in how they
felt if their exercise routine was disrupted: "When I don't exer-
cise, I get sluggish, tired, don't feel peppy"; "Just icky"; "I get
cranky and irritable" (23).
As adults age, minor somatic complaints including sleep dis-
turbances, digestive problems, assorted aches and pains, and
general lethargy become more common (8,9,25,26). Clearly,
these physical symptoms are interrelated. For example, bodily
discomforts disrupt sleep, and poor sleeping patterns in tum
may lead to lethargy or irritability. Such complaints are exacer-
bated by chronic health problems and a sedentary lifestyle
(5,9,25). Over time, exercise participation may lead to notice-
able changes in such symptoms for older adults who have pre-
viously been inactive. For regularly active individuals, such
symptoms may only become evident if exercise is not main-
tained, as suggested by Schneider's qualitative data (23).
Two recent studies have addressed the relationship between
exercise and subjective appraisals of sleep quality for older
adults using the Pittsburgh Sleep Quality Index (PSQI) and
sleep diaries (8,26). The first, a cross-sectional study, found that
regular exercisers reported less sleep disturbance and fewer
physical complaints in general compared to sedentary persons
(26). The second, a prospective study, found significant im-
provements in a number of sleep parameters for women with
moderate sleep complaints who participated in a 16-week,
moderate-intensity exercise program, compared to a control
group (8). Sleep did not improve in the first eight weeks of ex-
ercising, and changes in aerobic capacity did not predict sleep
outcomes (8). I
For many older adults, improvements in sleep quality and
other psychophysical parameters may underlie the general no-
tion of "feeling good mentally and physically" and represent
observable, accumulated benefits of exercise participation.
Measures do exist that examine sleep quality-such as the 19-
item PSQI (27); or energy and fatigue, bodily pain, and
mood-such as the SF-36 (28,29) or the 65-adjective Profile of
Mood States [POMS (30)]. However, no single measure cap-
tures all these constructs, and concurrent administration of more
than one of these lengthy measures may be frustrating for older
M457
adults (6,29). The present study describes the development of
the Vitality Plus Scale, designed to capture multiple, interre-
lated aspects of "feeling good" relevant to the exercise experi-
ence in a single instrument.
METHODS
Table 1 outlines the sequential process of constructing the
Vitality Plus Scale (VPS) based on established psychometric
guidelines (31-34). As shown in Table 1, several samples com-
prising middle-aged and older adults were used in scale devel-
Downloaded from https://academic.oup.com/biomedgerontology/article/54/9/M456/545637 by guest on 02 November 2020
opment and validation to enhance confidence in the new mea-
sure's psychometric properties (31,32).
Protocol and Subjects
Development phase.-Scale items can be generated through
either a quantitative or a qualitative approach (33,34). The first
approach begins with a large pool of items from existing scales,
and/or based on "expert" opinion, that appear to capture the un-
derlying construct of interest, and then applies item reduction
techniques (31). The alternative, inductive, approach is to gen-
erate items collaboratively with a representative sample of in-
tended test takers (33,34). Using the latter approach, three focus
groups were conducted with 13 men and 15 women from dif-
ferent exercise classes. Three questions were used to stimulate
discussion: "What brought you to this program?" "What keeps
you coming back?" and "What do you get out of it?"
Themes emerging from the focus group discussions, ana-
lyzed using NUD*IST (35), were used to generate initial scale
items. Existing measures were then consulted to refine scale
content. Participants from the focus groups were reassembled
for pilot-testing purposes. Twelve professional instructors who
led fitness classes for middle-aged and older adults were mailed
Table 1. Steps and Samplesin Constructing
the VitalityPlus Scale (VPS)
I. Developmental Phase
Focusgroups withexercisers (n =28; aged56-81)
Initialitemgeneration basedon emerging themes
Pilot-testing andrefinement with28 olderadultsand 24 instructors
Relevance ratings by separate sample (n =81;aged40-82)
II. Reliability Phase
Administration to newsample of 38 (aged40-79) on twooccasions, oneweekapart
m. Validation Phase
Administration to N = 662(aged40-94) from 14different programs
Matching available dataon fitness andperformance measures
Pre-andpost-comparisons on 147participants starting in 5 programs
Administration of the SEES (n =25)andthe SF-36(n =156)
IV. Verification Phase
Refinement ofVPS itemsbasedon findings fromPhaseIII
Newsample (n = 143;aged49-84) usedto verify scaleproperties
M458 MYERSETAL.

the new scale and interviewed by telephone to solicit feedback
on content relevance and ease of administration and scoring. A
focus group was held with 12 volunteer senior fitness instruc-
tors for the same purposes. Similar to the development of the
EPI (7), we also asked an independent sample representative of
the target group to rate the perceived relevance of each item.
Test-retest reliability phase.-To examine stability of scale
scores, we recruited a sample of 28 women and 10 men (mean
age = 57) who were not regular exercisers. This sample was ad-
classes, including Tai Chi, strength training, and various aerobic
conditioning programs, described elsewhere (1). The other exer-
cise samples came from diverse research projects and indepen-
dent community programs. For comparison, we included three
social groups.
The validation pool was administered the new VPS, together
with a background questionnaire to collect demographic,
health, and activity information. For the community exercise
groups, the instruments were administered by either the pro-
gram coordinator or the class instructor. A number of physical

Downloaded from https://academic.oup.com/biomedgerontology/article/54/9/M456/545637 by guest on 02 November 2020

ministered the new VPS on two occasions, one week apart.
Each person was asked whether there was any change in their
normal pattern of activity over the previous week.
Validation phase.-The resulting 10-itemVPS (see Appendix)
was then administered to a large pool of more than 600 adults
(mean age = 68.3, SD = 8.5). As shown in Table 2, the sample
came from a wide variety of exercise groups. Participants from
the Centre for Activity and Ageing came from seven different
measures-aerobic capacity, walking speed (36), and the Timed
"Up and Go" or TUG (37)-were available from some of the
research samples. For research projects and programs starting
during the validation period, the VPS was administered at entry
and at the end of the project (or program session), yielding a
total of 147 completed pre- and post-VPS scales to permit a
preliminary examination of responsiveness to change.
To examine discriminant validity, 25 participants from three
ongoing exercise classes (age range 56-81) were administered

Table2. Vitality Plus ScaleScoresfor DifferentSamples

VPS Score Age HealthProblems Education

Gender
Mean (SD) Mean (SD) Mean number (SD) % LessThan
Sample n Range Range Range Male Female HighSchool
Homecare* 11 34 (6) 82.5 (8.9) 2.7 (1.8) 3 8 55%
25-45 68-94 0-5
Craftclasses'[ 29 36.1 (9) 72.9 (4.9) 2.0 (1.6) 7 22 48
1~50 62-82 D-6
Wellness clinics:j: 55 35.4 (7) 68.6 (7) 2.2 (1.5) 6 49 47
21-48 55-83 D-6
Aquatics:j: 14 36.8 (6) 66.5 (6.8) 2.0 (1.7) 0 14 50
25-46 51-77 0-5
Exerciseclass 12 39.0 (8) 71.2 (8) 1.7 (1.2) 11 31
22-49 5~89 0-4
Activity lab* 11 34.3 (8) 68.8 (4.6) 4 7 27
25-47 63-76
Centrenew:j: 35 36.4 (6) 61.9 (10) 1.9 (1.6) 11 24 16
20-46 41-82 D-6
University class'] 116 37.5 (7) 69.5 (7) 1.7 (1.4) 19 97 3
1~50 53-92 D-6
Walkgroup:j: 28 36.9 (7) 66.4 (6) 1.8 (1.5) 7 21 10
20-49 55-80 D-6
Physicianstudy*:j: 156 38.1 (7) 71.6 (4) 78 78
20-50 61-82
Lions club 29 38.6 (7) 69.6 (4.9) 1.4 (1.3) 7 22 13
22-50 59-79 0-4
Alumniclub 17 40.9 (4) 72.6 (6.9) 1.3 (.8) 5 12 11
33-50 6D-84 0-3
Centregroups 120 37.6 (7) 67.4 (9) 2.2 (1.2) 29 91 8
1~50 4~86 0-5
Socialgroupst 29 37.9 (7) 51.5 (10) 1.2 (2) 9 20 7
20-48 40-77 0-5

*Research project.
tNonexercisegroups.
:j:Baseline and follow-up.
 VITALITY PLUS SCALE
the 12-item SEES and the VPS immediately following an exer-
cise class. The SEES asks respondents to rate (from 1 = not at
all to 7 = very much so) ''the degree to which you are experienc-
ing each feeling now, at this point in time after exercising" (21).
To examine both convergent and discriminant validity, the
SF-36 Health Survey-a general measure of health status
(29)-was administered to 156 subjects entering the physician
study (characteristics shown in Table 2). We were most inter-
ested in Question 9 on the SF-36 ("How much during the past
four weeks have you been feeling ...") comprising the four
items on the subscale labeled "Vitality" or VIT (full of pep, a
lot of energy, worn out, tired), and the five items on the sub-
scale labeled "General Mental Health" or MH (a very nervous
person, so down in the dumps that nothing could cheer you up,
calm and peaceful, downhearted and blue, a happy person).
Each item is rated from 1 = all of the time to 6 = none of the
time (five items were reverse-scored so that low ratings consis-
tently indicated positive feelings). On the Physical Functioning
(PT) subscale of the SF-36, individuals are asked the extent to
which their health now limits them (from 1 =limited a lot, 2 =
limited a little, to 3 = not limited at all) in 10 areas ranging from
"vigorous activities" (such as running, heavy lifting, and partic-
ipating in strenuous sports) to bathing and dressing.
Verificationphase.-Based on findings from the first valida-
tion pool, the VPS was slightly modified (as illustrated in the
Appendix). A subsequent pool of 143 participants (mean age =
69, SD =7; 90% women) from 12 sites of a general condition-
ing "Elderobics" program was recruited to verify the psychome-
tric properties of the revised VPS scale. This sample was also
asked: "How do you usually feel right after this class? [l =not
at all tired, 3 = pleasantly tired, 5 =unpleasantly tired or wiped
out]", and their reasons for joining and continuing with the class.
A Priori Predictions
Regular exercisers should associate exercise with positive
sensations and feelings. Thus, our sample was expected to
score high on the Positive Well-Being and low on the
Psychological Distress subscales of the SEES. Minimal corre-
lations were expected between SEES and VPS scores because
the former measures session-specific feelings (such as
"crummy" or "terrific"), whereas the VPS is intended to mea-
sure more enduring patterns of sleep, energy levels, etc. In con-
trast, we expected more convergence between VPS scores and
specific SF-36 subscales: "Vitality," "General Mental Health,"
and the two bodily pain items.
Persons reporting more health-related limitations, as mea-
sured by the PT subscale on the SF- 36 and the single item on
our background questionnaire ("Are you currently limited in
the type or amount of physical activity, work or leisure, you can
do because of an illness, injury or disability?"), should score
lower on the VPS. In general, adults with poorer health status,
measured through a number of different indicators, were ex-
pected to have lower VPS scores. Adults who are more physi-
cally active should score higher on the VPS. Functional mobil-
ity scores on the walk and TUG tests were expected to correlate
positively with VPS scores. Finally, we expected that individu-
als who had lower VPS scores at entry into an exercise program
or research study would be more likely to show improvement
as a result of participation (5,38).
M459
RESULTS
Instrument Development
Older adults gave a variety of reasons for joining exercise
programs: to get out of the house, to meet people, and to keep
active and healthy. Health-related reasons were sometimes spe-
cific (e.g., "to help my arthritis," "to lose weight," "to reduce
pain," "control high blood pressure," "for my diabetes," "sore
back," "for my bones," "for my joints") and sometimes general
(e.g., "to keep limber," "to keep moving," "to delay the aging
Downloaded from https://academic.oup.com/biomedgerontology/article/54/9/M456/545637 by guest on 02 November 2020
process"). Arthritis and osteoporosis were more frequently
mentioned by the women, whereas cardiac concerns came up
repeatedly in the men's group. Multiple reasons such as cama-
raderie, fun, and "starts the day off right" also emerged for con-
tinued participation. Many people spontaneously said, "Because
I feel better." When explored further, specific themes that
emerged were: sleeping better, more pep and energy, fewer
aches and pains, less fidgety, reduced stiffness, more relaxed
and cheerful, less gas and constipation, and improved appetite.
Each of these themes was developed into a scale item. The
"sleep better" theme was developed into three items to reflect
onset latency, sleep quality, and daytime drowsiness-shown to
be improved through exercise (8,26).
Our pilot test groups recommended using a 14-point Times
New Roman typeface for readability. A 5-point rating, with de-
scriptors of each attribute portrayed as opposite ends of a con-
tinuum, was preferred over a Likert-type (strongly agree to dis-
agree) format. Some people found item reversal confusing, so
all items were worded in the same direction. The initial scale
items, instructions, and rating format are shown in the
Appendix. We chose the title "Vitality Plus Scale" (VPS) to dis-
tinguish this new instrument from previous measures limited to
energy and fatigue content. No label is affixed when the scale is
actually administered.
Instructors who led exercise groups for older adults felt that
the new tool captured perceived improvements often voiced by
their own participants. No item additions or deletions were sug-
gested. An independent sample of older exercises also endorsed
the items. As can be seen in Figure 1, the majority believed that
exercise had a positive influence on the areas captured by the
new scale.
Psychometric Properties,
Temporal stability.-The intraclass correlation coefficient
(ICC) was used to estimate test-retest reliability (31,32). The
ICC was 0.87 (95% confidence interval [CI] = .76, .93) indicat-
ing good temporal stability for VPS scores over a one-week pe-
riod. All 38 subjects confirmed that there had not been any
change in their normal pattern of activity over the week interval.
Scale properties.-The single summary VPS score can range
from 10 to 50 (higher scores are more positive). Examination of
stem and leaf plots indicated some skewness for three of the
items- cheerfulness, appetite, and constipation. Use of para-
metric statistics was justified given that most items showed
good distribution, and the overall sample mean score (35, SD =
7) was close to the center of the possible range (31,32).
Cronbach's alpha indicated good internal consistency for
both the initial (0.83) and the revised version (0.81) of the new
M460
::{¢:f¢';~~::«~: ;~~{~~fH~~~~';·~t;~:;,j;MrN4r$';-x;i;~t:··~(::::-::r»~
*,r.~:~~
RoIIlced :m~~:~:a.*·:i::'.8;r:6r~-::~ ~;st~;:;:~~~~~:::illM~~~m~~,:~~~~~~:;,,~ ~:·~.;? ;
Ene<gy ·:"·;.':f-4,W¥.:P@~t~#;:~,¢r.~~~:~.~~M&:<'::::ii~f~"i{:i:;::::;:t.*~~~::·:""';:1i'.i.:
;:;:;:
:'; :;':>i';::J::::>'~".t~ ~=»""f.~:;;~~iY$:-n.~/t·io4X ~i<~~:;;;;~>~~~R»M~~">:" r-;~;~::M~~~~fi~~~-:~~'~~
:,~::;t{;~~M~F~~;f ~~.:£;:;.,:w:.J~t~:~~':~~<h;*i ~('?;''%<::'~ :~.:i::Lf.:Z;.¥~~:~:tf;;; 1
;%:~:s;~~~;::.r.~'1?;iC::$·:::t:N~'*ji~': ~-*l~~~~WS~W~0~$X;~~~~M$W1~~':?W
~W~~t.$~)@:tr,;,;,)~Y~:~~*:~;".:i.:~.~~~A'":":~~~-:f."~~w~~~:rEh~~'
Aches & PIinI ~ :::;.:·~;:~~x:t""~:t~$t.~W:~~~~f.:~~":~if¥-.i:W-t~.iM~~();i:f';;;~~~~{;.:; ..:.,:HiP;
t
constlpatlon 1 :~~$.w:t:~~Sft. ~~*~ ~~A:;.:t%~~m::lli:1~:,r:«~,ti~~Nr4#f~~:
3O"Ao
Figure I. Perceivedimprovements as a result of exercising.
VPS. Coefficients between 0.70 and 0.80 are acceptable; 0.80
to 0.90 are very good; above 0.90 suggests possi bly shortening
the scale (31). If alpha remains fairly constant across subsam-
ples, one can be more confident that these values are not dis-
torted by chance. Accordingly, we split our sample into persons
aged 65 and over (n = 473) and those under age 65 (n 189).
Alpha was 0.82 and 0.83 for the older and younger subgroups,
respectively.
The rule of thumb is that each item should correlate at least
0.20 with the total score; moderate correlations amo ng items
indicate homogeneity, and high correlations indicate redun-
dancy and possib le loss of content validity (32) . For the initial
version of the VPS, item -total correlations ranged from 0.36 to
0.60, whereas interitem correlations ranged from 0.21 to 0.55.
The tenth item, cheerful, was the most weakly related to the
other items and the total scale score . Post-administration feed-
back suggested some slight modifications to item wording (see
Appendix). Test takers also told us that the item "cheerful" may
be "more of a personality thing" and suggested replacement
with the item "feel good ." Internal consistency was not affected
with the change in this item. For the revised VPS, item-total
correlations ranged from 0.23 (constipation) to 0.58 (the new
"feel good" item).
MYERS ETAL.
Downloaded from https://academic.oup.com/biomedgerontology/article/54/9/M456/545637 by guest on 02 November 2020
So (3-4)
• VII'(
""""
IJ MoclonIloly (1-2)
C NcIAt AI (0)
,
- ,.
.. ~~~Y..m.,.~:;{t::t.t~.t~.;;>! ~,')~.~.;..:::%,
40"1' 5O"A.
PERCENTAGE OF RATINGS
Discriminative abilities.-The total pool of 662 was com -
bined to examine sample characteristics in relation to VPS
scores. Univariate analyses presented in Table 3 show that VPS
scores differed significantly based on demographics and vari-
ous indicators of health status. Being overweight, using medi -
= cations , being limited in the type or amount of physical activity
(work or leisure) due to illness/injury /disability, and experienc-
ing shortness of breath while walking a distance equal to one
city block were associated with lower VPS scores. Current level
of physical activity and perceived importance of physical activ-
ity to one's regular routine , on the other hand, were associated
with higher VPS scores.
Total number of self-reported health problems was inversely
=
and significantly related to VPS scores (r -.45, P < .000) for
the entire pool. Given that 68 was the average age of the sam-
ple, it is not surprising that the majority (almost 80 %) had at
least one chronic health problem, most notably arthritis. The 82
people (or 21 % of the 394 subjects who filled out this question)
who reported no chronic, diagnosed health conditions had su-
perior VPS scores (mean =40.7) compared to persons with sin-
gle and multiple health problems (Table 4).
To determine the relative contribution of these variables to
VPS scores, a stepwise multiple regression analysis was per-
 VITALITY PLUS SCALE M461

Table 3. Relationship of VitalityPlus ScaleScoresto Sample significantly but weakly correlated with VPS scores (r = -.21,
Characteristics and Ratings p < .05), but the number of self-reported health problems was
more strongly related (r = -.35, p < .000). Persons without
VPS
Characteristics n Mean (SD) Statistic p Value
physical activity limitations (t = 3.79, p < .001) had superior
VPS scores, as did those with no chronic health problems
Gender (Table4).
Men 178 38.7 (7) t = 3.12 < .002
Women 464 36.7 (7)
Convergent and discriminant validity.-The sample of 25
Education regular exercisers who were also administered the SEES had a
College graduate 260 38.4 (6) F = 4.90 < .002 mean VPS score of 38 (SD = 8). Scores on each of the SEES
Post secondary 75 36.1 (8)

Downloaded from https://academic.oup.com/biomedgerontology/article/54/9/M456/545637 by guest on 02 November 2020

subscales (possible range 4-28) were skewed and high on
High school graduate 87 36.8 (7) Tukey B *1 vs 4
Less than high school 92 35.4 (8)
Positive Well-Being (mean = 21.4, SD = 5, range 13-28),
skewed and low on Psychological Distress (mean = 5.8, SD =
Perceived Financial Status
4, range 4-19), and near the scale midpoint on Fatigue (mean =
More than sufficient 293 38.1 (7) t= 3.91 <.000
9.8, SD = 4, range 4-16). Correlations between VPS scores and
Sufficient/barely 122 35.1 (7)
each of the three SEES subscales were low and nonsignificant:
Perceived Health Status r = .15 with the Positive Well-Being subscale, r =-.21 with the
Excellent 145 41.1 (6) =
F 59.1 < .000 Psychological Distress subscale, and r = -.22 with the Fatigue
Good 301 36.7 (7) Tukey B *1 vs 2
subscale. Because exercise-induced fatigue can be interpreted
Fair/poor 66 30.8 (7) *1 vs 3 *2 vs 3
either positively or negatively (6,7), our Elderobics sample was
Current Smoker asked the following question on their background question-
No 553 37.4 (7) t= .31 NS naire: "How do you usually feel right after this class? [1 = not
37.1 (8)
Yes 83
at all tired, 3 =pleasantly tired; 5 =unpleasantly tired or wiped
Weight Perception out]". The majority of this sample (97%) said they were either
About right 295 38.4 (7) t= 4.32 <.000 not tired or pleasantly tired after their class.
Overweight 202 35.6 (8) Our subsample of 156 subjects scored very positively (low)
Medication Use on both the SF-36 Vitality (VIT) subscale (mean = 10.1, SD =
No 131 38.9 (6) t = 3.22 <.001 3.7, range 4-24), and the Mental Health (MH) subscale (mean
Yes 269 36.4 (8) = 9.8, SD = 3.8, range 5-22). The VPS score was more highly
Health Limitations correlated with the SF-36 VIT sub scale (r = -.65, p < .001)
No/temporary 412 38.3 (7) t= 7.49 <.000 than with the MH subscale (r = -.48, p < .000). Both bodily
Yes 97 32.6 (7) pain items on the SF-36 (amount and extent of interference with
Perceived Physical Activity normal work over the past 4 weeks) were inversely related to
Very much 273 38.0 (7) F= 5.56 <.004 the VPS score (r = -.49, r = -.47, p < .000), and the SF-36
Somewhat 39 35.3 (6) Tukey B *1 vs 2 VIT (r = .45, r = .48, p < .000) and MH scores (r = .36, r = .43,
Not at all 53 35.1 (8) *1 vs 3 p < .0(0) as expected.
Importance of Exercise This sample also scored positively on PT Functioning sub-
Extremely 87 40.1 (7) F= 5.78 <.004 scale (mean =25.2, SD =4.4, range 10-30). The correlation be-
Moderately 106 36.8 (6) Tukey B *1 vs 3 tween the overall score on PT subscale was r = -.27 (p < .001)
Not important 22 36.5 (11) *2 vs 3 with age, and r = -.19 (p < .02) with the VPS. A breakdown on
Shortness of Breath the individual PT subscale items showed that the majority of
No 326 37.8 (7) t= 3.27 <.001 our sample said they were not limited at all concerning:
Yes 54 34.5 (8) bathing/dressing (94%), walking one block (93%), climbing
one flight of stairs (80%), walking several blocks (80%), lift-
Note: Background information was missing for some respondents.
*Significant group comparison via Tukey B test.
ing/carrying groceries (71%), doing moderate activities-such
as pushing a vacuum cleaner or playing golf (67%), and walk-
ing more than 1 kilometer (65%). In contrast, a roughly equal
formed with 307 subjects who provided complete data for: age, percentage (44%-47%) chose either "not limited at all" or "lim-
gender, education, perceived financial status, perceived health ited a little" for the items bendinglkneeling/stooping and climb-
status, overweight/at right weight, total number of diagnosed ing several flights of stairs. Not surprisingly, our sample per-
health problems, and self-reported physical activity level. Both ceived the most limitation concerning the "vigorous activities"
entry and removal criteria were applied, and partial regression item on the PT subscale (50% of our sample said "limited a lit-
coefficients were adjusted statistically for the other variables in tle"; 32% said "limited a lot").
the equation. Three variables emerged in the final model, ac- VPS scores were significantly lower for persons who re-
counting for a total of 30% of the variance (F = 43.0, p < .(00). ported they were "limited a lot" (vs "limited a little" or "not
The best predictor of VPS scores was total number of health limited at all") on 8 of the 10 items of the SF-36 Physical
problems (Beta = -2.19), followed by perceived health status Functioning subscale. The most significant difference was
(Beta = -3.35), and age (Beta = .07). found for "moderate activities" (F = 10.9, p < .000), followed
Similar relationships emerged between scores on the revised by "bending/stoopinglkneeling" (F:= 9.2,p < .000), and "vigor-
VPS and characteristics of the second pool (n = 143). Age was ous activities" (F = 4.9, P < .009).
M462 MYERSETAL.

Table4. Vitality Plus ScaleScoresfor PersonsWith ChronicHealthConditions

Pool I (n =394) Pool II (n = 143)

Age VPS Score Age VPS Score

Health Condition(s) n Mean (SD) Mean (SD) n Mean (SD) Mean (SD)
None 82 59 (17) 40.7 (6) 16 68 (18) 43.5 (5)
Arthritis 170 69 (9) 35.1(7) 61 72 (7) 37.1 (6)
Back 121 66 (12) 34.5 (8) 45 70 (7) 37.8 (6)
Arthritis and back 71 68 (10) 33.0 (8) 28 70 (8) 35.9 (6)
Osteoporosis 55 69 (7) 34.5 (7) 22 72 (6) 38.5 (6)

Downloaded from https://academic.oup.com/biomedgerontology/article/54/9/M456/545637 by guest on 02 November 2020

Osteoporosis and arthritis 32 68 (7) 34.1 (8) 12 73 (6) 36.2 (5)
Osteoporosis and back 23 69 (9) 31.4 (8) 7 75 (5) 37.2 (5)
Heart 56 72 (7) 34.9 (7) 16 75 (6) 38.6 (6)
High blood pressure 113 69 (9) 35.8 (8) 52 70 (7) 39.0 (5)
Heart and high BP 24 74 (8) 32.9 (7) 11 74 (7) 38.4 (6)
Diabetes 25 69 (7) 35.0 (8) 11 72 (5) 36.8 (5)
COPD 43 69 (11) 34.4 (7) 10 71 (8) 39.2 (5)
Foot 93 65 (13) 34.3 (7) 25 73 (7) 36.7 (7)
Back and foot 49 65 (13) 32.7 (8) 14 71 (7) 34.5 (6)
Hearing 82 71 (9) 34.5 (7) 20 71 (6) 39.8 (7)
Vision 69 65 (16) 36.1 (8) 15 74 (5) 39.0 (5)
Bladder 20 64 (10) 36.5 (7) 0
Overweight 214 64 (11) 35.6 (8) 30 68 (6) 38.4 (6)
Overweight and osteoporosis 18 69 (7) 36.0 (8) 5 68 (4) 33.8 (3)
Overweight and arthritis 71 67 (8) 33.8 (7) 13 67 (7) 36.4 (5)

Relationship to Physical Measures
Data on aerobic capacity, TUG, and walk speed were avail-
ablefrom someof our research samples. V02max (ml-kg t-mirr')
scores for individuals in the physician study ranged from 10.3
to 41.2 (mean =23.3, SD =6) and were not found to be signifi-
cantly correlatedwith VPS scores.
In contrast,both TUG scoresand fast-paced walk speed were
significantly related to VPS scoresin a number of samples. Not
surprisingly, TUG scores were poorer for the group receiving
home support (mean = 21.6, SD =4.5, range 14-29) in com-
parison to adults attending the Centre's exercise classes (mean
= 9.3, SD = 1.6,range 7-12). TUG scores were found to be sig-
nificantly correlated with both age (r = .68, p < .000) and VPS
scores (r = -.58, p < .05). Fast-paced walk scores ranged from
1.28 to 2.05 meters/second (mean = 1.63, SD = .22) in the
Centre group, and from 1.04 to 1.67 mls (mean = 1.37, SD =
.16) in the Wellness Clinics. Walk speed correlated with VPS
scores in both samples (Centre: r = .48, p < .02: Clinics: r =
.43,p < .07).
Responsivenessto Change
We obtained a total of 147 completed pre- and post-VPS
scales from individuals who began their program during the
study validation period. The number who stayed with their re-
spective project or program, and completed the VPS at both
baseline and follow-up, is shown in Table 5. Unfortunately,
baseline level of physical activity and extent of participation
were not systematically recorded in all of the projects. These
factors, together with baseline level of functioning on the mea-
sure in question, will influence the extent of improvement that
can be expected from exercise participation (5,38). Available
start-upprogramsdifferedin both frequency of weekly sessions
and total duration (from the 3-week walk group to the ongoing
Centre classes). For the above reasons, our preliminary exami-
nation focused on the extent of individual change, using an ap-
proach suggested by Lord and colleagues (38). Change was
calculated for each person using the formula: ([follow-up-
baseline score] -;- baselinescore) X 100 (ref. 38).
Table 5 shows that half the total sample (76 of 147) showed
some positive change in VPS scores. Of these 76 individuals,
37 (or 48.7%) improved theirVPS score by over 10% (ranging
up to 78%). Similar to Lord and colleagues' work, we com-
pared individuals who improved by at least 10% (n = 37) to
those showing no positive change (n = 71) and found signifi-
cantly lower averagebaselineVPS scores in the former group (t
= 6.84, p < .000). Individuals were far more likely to improve
by at least 10% (X2 = 31.5,p < .000) if they scored at or below
the total sample mean of 37, as compared to above the mean.
Persons who scoredbelow the scale mean of 30 (aboutone fifth
of this sample) were extremely likely to improve (71 % im-
proved by 10% or more).
Extent of change was also examined for each of the 10VPS
items for the 37 individuals who improved their overall scores
by at least 10%. The highest percentage of change emerged for
the following items: pep and energy (66%), fall asleep (54%),
aches and pains (54%), feel rested (51%), stiff and sore (50%),
and sleep well (40%).The items showing less change were: ap-
petite (28%),calm and relaxed (28%), constipated (22%), and
cheerful (17%).
While a proportion of individualsin each of the fiveexercise
programsimproved, between-and within-sample differences are
noteworthy. Participants in the 3-monthWellness Clinicsgeneral
exercise program offered twice a week showed the most im-
provement.The clinic coordinatorreported that many attendees
were previously sedentary,and the low baseline VPS score for
this sample is consistentwith this report.Attendance was higher
 VITALITY PLUS SCALE M463

Table 5. Extent of Improvement in Vitality Plus Scale Scores From Baseline to Follow-up by Sample

Extentof Improvement
Baseline No Positive SomePositive
Sample n VPS Changesf) Change>0 0.1 to 4.9% 5 to 9.5% ~.6%

Total 147 71 (48%) 76(52%) 20 19 37

Mean (SD) 37.0 (7) 39.5 (6) 34.7 (7) 40.1 (6) 36.4 (7) 30.9 (6)
Mean Change (SD) -9.3 (9) 13.5 (14) 3.5 (1) 6.9 (2) 22.3 (15)
Range -39-0 2.2-78.3 2.2-4.9 5-9.5 9.6-78.3
WalkGroup 16 10 (62.5%) 6 (37.5%) 3 1 2
Mean (SD) 38.9 (7) 39.6 (7) 37.2 (9) 44.3 (3) 26.0 32.0 (7)

Downloaded from https://academic.oup.com/biomedgerontology/article/54/9/M456/545637 by guest on 02 November 2020

Mean Change (SD) -6.6 (8) 9.9 (10) 2.9 (1) 7.7 21.6 (11)
Range -20-0 2.3-29.6 2.3-4.2 13.5-29.6
Aquatics 10 6 (60%) 4 (40%) 1 2 1
Mean (SD) 37.1 (7) 36.0 (9) 38.8 (3) 40.0 40.5 (1) 34.0
Mean Change(SD) -11.6(12) 9.6 (10) 2.5 6.2 (2) 23.5
Range -26-0 2.5-23.5 5-7.3
Centre 23 11 (48%) 12 (52%) 4 4 4
Mean (SD) 36.8 (6) 38.1 (5) 35.6 (7) 36.5 (10) 38.3 (4) 32 (3)
Mean Change (SD) -9.2 (8) 9.3 (7) 4.1 (1) 6.4 (2) 17.3 (4)
Range -26-0 2.7-21.9 2.7-4.8 5.1-9.3 13-21.9
Physician 57 28 (49%) 29 (51%) 7 7 15
Mean (SD) 37.8 (7) 40.2 (6) 35.6 (7) 40.3 (5) 39.6 (3) 31.5 (7)
Mean Change (SD) -9.3 (10) 12.0 (10) 3.5 (1) 6.4 (2) 18.7 (10)
Range -38-0 2.5-40 2.5-4.8 5.0-8.9 9.8-40
Wellness Clinics 41 16 (39%) 25(61%) 5 4 16
Mean (SD) 35.3 (8) 40.3 (5) 32.1 (8) 40.2 (6) 33.3 (8) 29.3 (7)
Mean Change (SD) -10.3 (9) 18.7 (19) 3.5 (1) 7.8 (2) 26.2 (20)
Range -28-0 2.2-78.3 2.2-4.9 95-9.1 9.6-78.3

in the two rural clinics than in the urban clinic (67% and 53% vs
43%), and VPS scores improved for a greater proportion of par-
ticipants in the former settings (64% and 72% vs 42%).
In contrast, the 3-week walk group had the lowest proportion
of improvers (37.5%). According to their coordinator, many
had previously been regular walkers. Similarly, the instructor of
the 2-month aquatics session noted that over half the partici-
pants had attended previous sessions; only 40% of this group
improved their VPS scores. Subjects in the physician study
were prescribed various protocols for exercising on their own
over a 3-month period; 51% improved their VPS scores. Our
sample of new Centre participants, meanwhile, had been exer-
cising anywhere from one to 10 months before the VPS was
readministered (average 3.8 months, SD = 2). We found a sig-
nificant correlation (r = .39, p < .05) in this group between
number of months in the program and VPS scores.
DISCUSSION
The Vitality Plus Scale is a promising new measure for ex-
amining accumulated benefits of exercise participation for
older adults. While many items on the VPS are similar to those
found on previous measures such as the SF-36, the POMS, the
PSQI, and the EFI, the VPS captures a number of interrelated
aspects of "feeling good" in a single instrument that takes less
than 5 minutes to complete. The alternative is to administer a
battery of lengthy scales that may be frustrating to older re-
spondents (6,29).
Minor aches and pains, lethargy, constipation, poor appetite,
and sleep problems increase with age. Regular exercise is an at-
tractive alternative to pharmacological remedies such as anal-
gesics, laxatives, and sleeping pills. VPS scores were strongly
related to various indicators of health status--diagnosed health
problems, medication use, shortness of breath, and perceived
limitations in functioning. While designed for older adults,
middle-aged adults who are sedentary may also show change
on this measure as a result of exercising.
The present study indicates that the new Vitality Plus Scale
has good psychometric properties. Alpha values and item-total
correlations support the homogeneity of the scale. Replicating
findings with multiple and split samples greatly increases the
confidence in a measure (31). Most scale developers report in-
ternal consistency with a single sample, but many fail to exam-
ine temporal stability (31,32). Test-retest reliability is critical to
demonstrate the reproducibility of an instrument and to lay the
foundation for detecting real change as a result of an interven-
tion (32). VPS scores showed good internal consistency and re-
producibility.
Individuals who reported being more physically active, and
rated exercise as important to their regular routine, had higher
VPS scores. Participation may be as important as the exercise
itself (5,9). There is some evidence that energy expenditure
through housework is not associated with the same degree of
positive affect as recreational physical activity (3). Fun and en-
joyment, mastering new skills, camaraderie, and getting fresh
air may all contribute to improvements in sleep, appetite, and
mood. Exercise also has physiological effects on various bodily
M464 MYERS ETAL.
systemsthat may be perceivedsubjectively as tension releaseor
enhancedenergy (5-8).
Sensations and feelings associated with a single exercise
episode may not be the same as more generalized attributions
developed and reinforced through continued participation
(23,24). We found very low correlations between SEES and
VPS scores. Regular exercisers scored high on Positive Well-
Being,low on PsychologicalDistress,and near the midpointon
the Fatigue subscales of the SEES. Exercise-induced fatigue is
more likely to be associated with vigorous exercise (22). The
majority of our large sample from the Elderobics program re-
ported that they did not feel tired or felt pleasantly tired imme-
diatelyafter their class.
Community exercise programs for older adults are typically
low to moderate in level of intensity, progressive in nature, and
may be offered only once or twice a week (15-17). It may be
unrealisticto expect measurable changes in fitness parameters,
such as aerobic capacity or muscle strength, from such pro-
grams. In any case, such changes are not directly observable to
most participants (5,8,12). Similarto previous findings withpsy-
chological measures (6), the SF-36 (10), and sleep (8), a signifi-
cant relationship did not emerge between aerobic capacity and
VPS scores. On the other hand,VPS scoreswere associatedwith
gait speedand mobility-more functional, physicaltests.
The present study provides preliminaryevidence that scores
on the VPS may be responsive to change as a result of exercis-
ing based on pre- and post-comparisons of individuals begin-
ning exercise programs or research projects. Some of the scale
items (pep and energy,fall asleep quickly, aches and pains, feel
rested, stiff and sore) showed more change than others (such as
appetite and constipation), suggesting that these areas may be
more sensitive to exercise, or effects may simply occur sooner
in these areas. Similar to King and colleagues' findings with re-
spect to sleep (8), only people with moderatecomplaints can be
expectedto show improvementas a result of exercise, and such
improvements take time. For already activepersons, changes in
energy level, mood, sleep, and such may only be noticeable if
their exercise routine is disrupted (23). The somewhat skewed
ratings for appetite and constipationindicate that fewer individ-
uals, at least in our fairly healthy and active samples, experi-
enced problems in these areas. In our focus groups, some peo-
ple said they had never been bothered by constipation. Others
said they were "regular" because of exercise and got consti-
pated when they did not exercise.
With respect to total VPS scores, groups who were more
sedentary(accordingto their instructors), and individualswithin
these groups whohad lower baselinescores, were more likelyto
improve by the end of the exercise session or research project.
For ongoing classes, number of months in the program was sig-
nificantlycorrelated with VPS scores.However, we discovered
that obtaining truly "sedentary" individuals from community
programs or even research projects was difficult. It was not un-
common to findthat "new enrollers" had previously participated
in otherprogramsor walkedon theirown or with a companion.
Further research is necessary to examine the incremental ef-
fects of various types of exercise and modes of participation
(solitary, single companion, group-based) on VPS scores for
different populations of adults. Frail adults of advanced age
may be particularly likely to show improvement in the areas of
appetite and constipation, in addition to the other areas mea-
sured by the VPS. Few suitablemeasures currentlyexist for this
population.
Randomized studies are needed to determine whether
changes can be attributed to a particular exercise intervention.
Our findings suggestthat VPS scores remain stable over a one-
week period for individuals who reportedly had not changed
their normal pattern of activity. VPS scores may be affectednot
only by changes in physical activity (becoming more active or
less active),but also by other lifestyle changes (such as diet and
smoking), changes in health, and life events. These influences
Downloaded from https://academic.oup.com/biomedgerontology/article/54/9/M456/545637 by guest on 02 November 2020
need to be controlled for, or taken into consideration, particu-
larly in studiesof longer duration.
As noted repeatedlyin the exercise literature, it is often diffi-
cult to disentangle the effects of exercise per se from other as-
pects of participation (5,9,15,23). People begin and continue to
exercisefor a varietyof reasons.Enjoying the particularactivity
is important for sustained participation, and recent evidence
shows that older adults who try out a number of different exer-
cise options are more likely to remain active (1). In any case, a
given program or exercise regimen will be relativelybeneficial
depending on individualneeds, baseline level' of activity, extent
of participation, and initial scores on the outcome measure in
question (5).
The widespread use of common measurement tools would
greatly facilitate the collection of evidence pertaining to the
benefitsof exerciseparticipation for older adults (5,6,9). We be-
lieve that the Vitality Plus Scale helps fill this void and provides
researchers with a psychometrically credible tool. The scale is
acceptable to and easily administered and scored by fitness in-
.structors and coordinators, providing an outcome measure for
the evaluation of exercise programs for older adults. Most im-
portantly, the scale was developed in collaboration with older
exercisers themselvesand appears to capture the exercise expe-
rience from their perspective.
ACKNOWLEDGMENTS
This study was supported by grants from Searle Canada, Inc., Health
Canada/Fitness, and the Medical Research Council of Canada.
We gratefully acknowledge the cooperation of all the program coordinators,
instructors, and individuals who participated in this project.
Address correspondence to Dr. Anita Myers, Department of Health Studies
and Gerontology, University of Waterloo, Waterloo, Ontario N2L 3Gl, Canada.
E-mail: amyers@healthy.uwaterloo.ca
REFERENCES
1. Ecclestone NA, Myers AM, Paterson DH. Tracking older participants of
twelve physical activity classes over a three year period. 1 Aging Phys
Activ. 1998;6:70-82.
2. Coleman D, Iso-Ahola SE. Leisure and health: the role of social support
and self-determination. 1 LeisureRes. 1993;25:111-128.
3. Stephens T. Physical activity and mental health in the United States and
Canada: evidence from four population surveys. Prev Med. 1988;17:
35-47.
4. World Health Organization. The Heidelberg guidelines for promoting
physical activity among older persons. J Aging PhysActiv. 1997;5:2-8.
5. Rejeski WJ, Brawley LR, Shumaker SA. Physical activity and health-re-
lated quality oflife. ExercSport Sci Rev. 1996;24:71-108.
6. McAuley E, Rudolph D. Physical activity, aging, and psychological well-
being. J Aging PhysActiv. 1995;3:67-96.
7. Gauvin L, Rejeski WJ. The Exercise-Induced Feeling Inventory: develop-
ment and initial validation. 1 Sport Exer Psycho!. 1993;15:403-423.
8. King AC, Oman RF, Brassington GS, Bliwise DL, Haskell WL.
Moderate-intensity exercise and self-rated quality of sleep in older adults:
a randomized controlled trial. lAMA. 1997;277:32-37.
 VITALITY PLUS SCALE
9. Stewart AL, King AC. Evaluating the efficacy of physical activity for in-
fluencing quality of life outcomes in older adults. Ann Behav Med. 1991;
13:108-116.
10. Stewart AL, King AC, Haskell WL. Endurance exercise and health-related
quality oflife in 50-65 year-old adults. Gerontologist. 1993;33:782-789.
11. Stewart AL, Hays RD, Wells KB, Rogers WH, Spritzer KL, Greenfield S.
Long-term functioning and well-being outcomes associated with physical
activity and exercise in patients with chronic conditions in the Medical
Outcomes Study. J Clin Epidemiol. 1994;47:719-730.
12. Minor MA, Brown JD. Exercise maintenance of persons with arthritis
after participation in a class experience. Health Educ Q. 1993;20:83-95.
13. Blumenthal JA, Emery CF, Madden DJ, et al. Long-term effects on psy-
chological functioning in older men and women. J Gerontol Psych Sci.
1991;46:P352-P361.
14. Emery CF, Gatz M. Psychological and cognitive effects of an exercise
program for community-residing older adults. Gerontologist. 1990;30:
184-188.
15. Myers AM, Gonda G. Research on physical activity in the elderly: practi-
cal implications for program planning. Can J Aging. 1986;5: 175-187.
16. Myers AM, Hamilton N. Evaluation of the Canadian Red Cross Society's
Fun and Fitness program for seniors. Can J Aging. 1985;4:201-212.
17. Sager K. Commentary: Senior fitness-for the health of it. Phys Sports
Med.1983;11:31-36.
18. Blumenthal JA, Emery CF, Madden DJ, et al. Cardiovascular and behav-
ioral effects of aerobic exercise training in healthy older men and women.
J Gerontol Med Sci. 1989;44:M 147-M157.
19. Gueldner SH, Spradley J. Outdoor walking lowers fatigue. J Occup
Nursing. 1988;14:6-12.
20. Binder EF, Brown M, Craft S, Schechtman KB, Birge SJ. Effects of a
group exercise program on risk factors for falls in frail older adults. J
Aging PhysActiv. 1994;2:25-37.
21. McAuley E, Courneya KS. The Subjective Exercise Experiences Scale
(SEES): development and validation. J Sport Exer Psycho/. 1994;16:
163-177.
22. Rejeski WJ, Gauvin L, Hobson ML, Norris JL. E±!::.;ts of baseline re-
ponses, in-task feelings, and duration of activity on exercise-induced feel-
ing states in women. Health Psychol. 1995;14:350-359.
23. Schneider JK. Qualitative descriptors of exercise in older women. J Aging
Phys Activ. 1996;4:251-263.
24. Leventhal H, Nerenz DR, Straus A. Self-regulation and the mechanisms
for appraisal. In: Mechanic D, ed. Symptoms, Illness Behavior, and Help-
Seeking. New York: Prodist; 1982:55- 87.
M465
25. Myers AM. Advising your elderly patients concerning safe exercising.
Can Fam Phys. 1987;33:195-205.
26. Brassington GS, Hicks RA. Aerobic exercise and self-reported sleep qual-
ity in elderly individuals. J Aging Phys Activ. 1995;3: 120-134.
27. Buysse DJ, Reynolds CF, Monk TH, Berman SR, Kupfer DJ. The
Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice
and research. Psychiatry Res. 1989;28: 193-213.
28. McHorney CA, Ware JE, Raczek AE. The MOS 36-item short form
health survey (SF-36): II. Psychometric and clinical tests of validity in
measuring physical and mental health constructs. Med Care. 1993;
31:247-263.
29. McHomey CA. Measuring and monitoring general health status in elderly
Downloaded from https://academic.oup.com/biomedgerontology/article/54/9/M456/545637 by guest on 02 November 2020
persons: practical and methodological issues in using the SF-36 Health
Survey. Gerontologist. 1996;36:571-583.
30. McNair DM, Lorr M, Droppleman LF. EDITS Manual: Profile of Mood
States. San Diego, CA: Educational and Industrial Testing Service; 1981.
31. DeVillis RF. Scale Development: Theory and Applications. Newbury
Park, CA: Sage; 1991.
32. Streiner DL, Norman GR. Health Measurement Scales. Toronto: Oxford
University Press; 1989.
33. Williams Jl, Naylor CD. How should health status measures be assessed?
Cautionary notes on procrustean frameworks. J Clin Epidemiol. 1992;45:
1347-1351.
34. Feinstein AR. Clinimetrics. New Haven, CT: Yale University Press; 1987.
35. Richards TJ, Richards L. The NUD*IST qualitative data analysis system.
Qual Sociol. 1991;14:307-324.
36. Himann J E, Cunningham DA, Rechnitzer PA, Paterson D. Age-related
changes in speed of walking. Med Sci Sports Exer. 1988;20:161-166.
37. Podsiadlo D, Richardson S. The Timed "Up and Go": a test of basic func-
tional mobility for frail elderly persons. J Am Geriatr Soc. 1991 ;39:
142-148.
38. Lord SR, Lloyd 00, Nirui M, Raymond J, Williams P, Stewart RA. The
effect of exercise on gait patterns in older women: a randomized trial. J
Gerontol Med Sci. 1996;51A:M64-M70.
Received April 14, 1998
Accepted November 18, 1998
See Appendix next page
M466 MYERS ETAL.

Appendix
Originaland RevisedItems on the Vitality Plus Scale

Instructions: This scale looks at how you are currently feeling. For each statement, circle a number from 1 to 5 that best describes you. For
instance, if you usually fall asleepquickly when you want to, circle 5. Otherwise,circle a numberfrom 1 to to 4, dependingon the extent to which
you usuallyhave difficulty fallingasleep.

Sampleratingformat: Takesa long time _ _ _ _ _ _ _ _ _ Fall asleep quickly

to fall asleep 1 2 3 4 5

Origin8lItems Revised Items

Downloaded from https://academic.oup.com/biomedgerontology/article/54/9/M456/545637 by guest on 02 November 2020

1. Takes a long time to fall asleep 1. same
Fall asleepquickly

2. Sleep very poorly,restlessly 2. Sleeppoorly Sleep well

Sleep well

3. Tiredor drowsyduring the day 3. same

Feel rested

4. Rarelyfeel hungry 4. Rarelyhungry ~xcellent appetite

Excellentappetite

5. Often constipated 5. Oftenconstipated Do not get constipated

Rarely constipated

6. Often have aches & pains 6. Oftenhave aches & pains _

Rarely have aches & pains _ _ _ _ _ Have no aches & pains

7. Easilyplayed out 7. Low energylevel Full of pep and energy

Full of pep and energy

8. Often feel stiff & sore 8. Often stiffin the morning _

Rarely stiff & sore _ _ _ _ _ Not stiffin the morning

9. Oftenrestlessand fidgety 9. Oftenrestlessor agitated Feel relaxed

Usuallycalm and relaxed

to. Oftendown in the dumps, blue 10. Often do not feel good Feel good
Usually cheerful