Joumal of Advanced Nursing, I991,16,299-310
The relationships between physical activity and health-promoting behaviours in mid-life women Angela GilliS RN BSeN MAdEd Associate Profasor mid Chair and Anne Perry RN BN MSe(A) Assistant Profe^or, Department of Nursing, St Francis Xavier Untverstty, Anhgontsh, Nova Scotia, Canada
Accepted for publicahon 27 June 1990
GILLIS A k PERRY A (1991) Joumal of Advanced Nursing 16,299-310 Hie relationships between physical activity and health-promoting behaviours in mid-life women The purpose of this study is to examine the mter-relationships among partiapation between physical activity, perceived well-being, self-esteem, health status, health locus of control and health-promoting lifestyle activities of middleaged women in a rural context Smce women compnse the largest (hving) segment of the older population, it is extremely important to study the predictors of health promotmg activities in this age group pnor to old age Recent research has noted that partiapation in physical activity acrts as an entry pomt to reflect on health behaviour practices, creahng a npple efifect to other areas of lifestyle behaviours The research questions were framed withm the Health Promotion Model The study utilized a longitudinal experimental design.
Thefindmgsof this study contnbute to the sntall, but growmg, research base on physical activity as a health-promotmg lifestyle behaviour and physical activity as a hfestyle practice of women m the mid-hfe years Findmgs also provide a base for the development of commuruty physical-activity programmes and other intervention strategies to enhance the health and well-being of women, particularly those residing in a rural settmg DEARTH OF RESEARCH Until recently, httk research has been available on the health or lifiestyk prachces of mid-life women, parhcularly those resiciing m a rural sethng The mid-hfe penod, defmed as the span of hme fi'om 35 to 65 years, is a umque tune penod m which mdividuals evaluate their assets of health, money,relahonships,etc, as they prepare for the 'pnvileged years' (Ddfy 1988). Smce women compnse the lamest livuw segment of the older populahon, it is ° ^ °
extremely unportant to study the predictors of healthpromotmg achvihes in this group pnor to old age so that nursmg strategies to promote health can be faahtated and implemented The purpose of tbs study was to examme the mter-relahonsbps among parhapation between physical aehvity, perceived weU-bemg, self-esteem, health locus of control and health promotmg lifestyle achvities of nudcile-aged women m a rural context jhe foUowing questions were mvestigatecL r-^ _^ . f i . i ^ « _ 1 E>oes parhapahon m a programnwot physical achvity prociuce change m tlw lifestyle practices of mid-hfe rural wcHiiai? 299
A GiUtsimdA. Perry FigiB« 1 Health Promotion Model (produced by permissicm of Pender N J
Cognitive-perceptual factors
Modifying factors
Partictpation in healtti-promottng tiehaviour
Health Prowiohtm m Nursing
Pradice 2nd edn Appleton k Lange, Norwalk, Connecticut, 1987)
Importance of health
Demographic characteristics
Perceived control of health
Biological characteristics
Perceived self-efficacy
Interpersonal influences
Definition of health
Situational factors
Perceived health status
Behavioural factors
Likelihood of engaging in health-promoting behaviours
Cues to action Perceived benefits of health-pronnoting t>ehaviours
Perceived barriers to health-promoting behaviours
2
To what extent does parhapahon m a programme of physical achvity influence the well-bemg, health status, self-esteem and health locus of conhol of rural mid-bfe wcnnai? 3 To what extent do self-esteen, perceived well-bemg, health status, health locus of conhol and healthpromoting behaviours predid adheraice to a programme of phystcal achvity by rural mid-bfe women over time? 4 What are the perceived factors that influence adherence to prc^rammes erf physical achvity fOT rural mid-bfe women? Pender's (1987) Health Prcwnohon Model provided the theorehcalframeworkfor the study Hie model postulates that people's percephons of thonselves and their health (cogmhve-perceptual factors) and modifymg factors play an influential role m determmmg their lifestyle achvihes Cogmhve-perceptual fadors indude the importaiKe of health perceived conhol of health, self-eflficacy, debmhon of heekh health status, perceived benefits (rf healthfS-OTnohng behaviours ard perceived bamers to healthpromohng behaviours The modifymg factors indude
demograpbcs, biological charadenshcs, mterpersonal mfluences, situahonal facton and behavioural fadors The model has two phases deasion makmg and takmg achon The cogmhve-perceptual aiKl modifymg factors conshtute the deasion-makmg phase The tatang-achon phase mdudes cues that mohvate an mdividual to engage m health prc»notmg behaviour (Figure 1)
SELECTED LITERATURE REVIEW Health stahshcs durmg ^ e past several decades refled changes in tiie health stahis of women (Rice & Augbam 1980) Accordmg to the Surgeon General's report Healthy People (US Deparhnent of Health, Educahon and Welfare 1979), less tihan 25% (rf non-inshhitionabz^l adults are mvolved m regular exerase The Can»la Health Survey (I98I) rujted that less than 22% of females 65 years and olda- are mvolved m regular exeiose The tandoicy towards reduced physical iK:hvity as age mo^ases is a potenti^ concern, as tt has been noted that individuals over the ^ e of 45 who are modendely to very achve are significantly less likely to display bdiaviour related to tli
Physical achvity
(Stewart et al 1985) Pnce & Luther (1980) mdicate that older adults can minimize the losses associated with agemg through physical achvity Although there has been a recent profusion of bterature on health-promohng behaviours, bttle is known about what mohvates mdividuals to make changes in their bfestyle and to mamtam these changes over time Studies over the past 30 years have explored the benefits of physical adivity in relahon to prevenhon of disease processes (Paff«ibarger 1978) More recently, studies have ldenhfied the relahonship between physical activity even at minimal levels, and posihve health status (Wiley & Comacho 1980, Allan 1985, Gillett 1988, Gillett & Eisenman 1987) Recent researdi has noted that parhapahon m physical achvity acts as an entry pomt to refled on health behaviour prachces creatmg a npple effed to other areas of lifestyle behaviours (O'Hagan 1984, Shestowsky 1983) Pender (1987) noted that a sedentary bfestyle often snerges early m females and becomes highly resistant to diange These years of machvity in females may he either a major cause or a contnbutmg factor to degenerahve dianges that are assoaated with ageing and chronic lllness In recent bterature, regular physical exercise has been mcreasmgly recognized as a lifestyle parameter to be assoaated with a state of enhanced health and weU-bemg (Gillett k Eisenman 1987, Walker et al 1988) Mounting evidence suggests that lifestyle choices such as regular partiapahon m physical achvity influences both health and longevity (Berkman & Breslow 1983) Despite mcreasmg support for the relationsbp between bfestyle pradices and health, and the reabty of epidemiological data which reflect bgher morbidity for women than for men, mid-bfe women have only recently begun to be mduded as a target of bfestyle-onented health-promoting programmmg and research (Woods 1981) AlHiough there IS a vast amount of bterature available on the benefits of partiapahon m physical achvity (Lenskyj 1988), bttle IS known about the speafic relahonships tbs study was designed to test There exists a need to explore the role of physical achvity as a health-promohng, rather than disease-preventmg lifestyle behaviour for mid-bfe women Factors related to the promc^on of health seekmg behavltHirs m mid-bfe woa«n remam to be dearly tdenttfied m tt^ bt^ature.
O,
O.
O3 O3
X
R=random assignment, O = observahon, X=expenmental treatmait Figure 2 Expenmental Model
promotmg bfestyle behaviours m mid-bfe women residmg m a mral settmg (Figure 2) Sample and setting One hundred and twenty-six women, aged 35-65 years, residmg m a rural eastem Canadian provmce, volunteered to parhapate m the study Rural was defined as an area with a population of less than 7500 Women were recnuted fi-om a broad vanety of sources, mdudmg women's organizahons. Church groups, soaal organizahons, home and school associations, etc Indusion cntena for sample selechon mduded females between 35-65 years of age, able to speak and understand Engbsh, residing m mral eastem Nova Scotia, wilbng to partiapate m a 12-week programme of physical adivity, and with an absence of any cardiovascular health problems Subjects were randomly assigned to an expenmental (n = 63) and a conhol group (n = 63) usmg a schedule generatedfi-oma table of random numbers Eighty-seven per cent of the women m the experimental group (n = 52) and 66% of the women m the conhol group (« = 40) completed the study questionnaires for time one, time two and time tbee The remaining 13% and 34% of the respechve cases who failed to complete any one of the three sets of questionnaires were removed fi-om the data analysis Data analysis was therefore based on a sample of 92 cases Table 1 summarizes the major background charadenshcs of the sample Experimental manoeuvre
Subjects m the expenmental group parhapated m three 60mmute exerase dasses eadi week for 12 weeks. Exerase dasses were m the form of choreographed dance rouhnes to music that followed the guidelmes outlmed by Pender (1987) The programme mduded a warm-up penod, shetdimg exerase, aerobic exerases and a cool-down METHOD pencKl The exerase programme was under the direchcMi of A kmgih«bial jaetest-post-test expenmaital design was an expenenced and appropnately quabfied fitness lnstmcused to shidy the mter-rekhonsbps among parhapahcm tor /Ul dasses were held m a ccHnmooalfitnesscentre. The beh¥e«i jAyacal achvity, perceived well-bang, health exerase programme was made available to the ccmhol ^atus, setf-esteeoi, health locus of control and health- group followmg annpletion of data ailledKxt 301
AGilhsmdAPeny Table 1 Distnbuhon of bacjcground cJuraetenshes
Group
Vanable
Age
Expenmoital (« = 52)
Conhol
Mean
Mean
(« = 40)
SD 845 3 28
4451 14-64
Years of sehoohng
n Employmoit status Employed Unot^loyed Annual lneome* Bebw 8000 8000-12 000 Above 12 000 Ethme bacjcground White Blaek Otlwr Mantal status Single Mamed Divoreed/separated Widowed Parhapation in exerase at least onee per wedc dunng past 3 mcmths
Yes No
%
4451 1414
707
2 76
n
%
214
78-6
40 12
769 231
32 8
9 3 36
173 58 692
8 4 23
24 3 10-8
52 0 0
lOOO 0-0 OO
38 1 I
95 2
3
4 32 3 1
11-9 78-6
3 2
58 846 58 38
40 12
76-9 231
30 10
244
36 16
69-2 30-8
29 11
275
44
649
24 24
71 24
75-6
Family partieipahcm m exereise at least onee per week, dunng past 3 months
Yes No
72 5
'Mratmgdata.
Aciherence to the exerase programme was measured duectly by a self-rq>ortmg attendance record Tbs consisted of a senes of 36 blanks w b d i represented the 36 exercise dasses held dunng the poiodtrf the study Parhapants placed a cJ:Kdc mark m the blsik if diey attended dass and \ek the blank empty if they were absentfrcrniclass for any reason.
A ktter ctf mtroduchcai was forwarckd to women on the m a ^ g lists erf vanous Churcit groiqps, womoi's organizatu»is, ccHnmuttty groups, etc, m rural easton Nova 302
Scotia. Women mterested m parhapatmg m the study retumed a form of mtoit to the researdiers by mml Those who indicated a wdlmgness to partiajoAe were contacted by phcHie, by the researches, and invited to attend a goieral meehng for the puqxjse of danfyu^ details of the study At tlK meeto^ a veri»l and wntten descrq7tic»i of the study was provicied and wntten cxm^nt obbuned botn aU partiapants Partiapants were then raiuicnnly assigned to either the experunental or ccmtrol group Data were cdlected from both grcH^^s at dvee distinct tune pemyds m tiie study at ^ K genend swetmg erf nitrodudKHi (time one), inuiKdiidely fblloimng caBq»btKm of ihe 12-week exeiase programnme (tone hvo) md a§mn 6 mcHiths kto- (tmne three)
Physical adtvtty
INSTRUMENTS
Convergent vabdity has been estabbshed witti the CPI Self-Acceptance Scale {r—0-27, w=643), and with a OTieFive self-reporhng mstruments were admmistered m the ltem self-esteem scale (r=0 45,« = 643, r = 0 66, H = 101) same order dunng each data collechon penod and retumed For discnminant vabdity, correlations with measures of m stamped, addressed envelopes to the researchers The self-stabibty are moderate (r=021 to 053) (Robmson & instruments mduded Rosenberg's Self-Esteem Scale, the Shaver 1973) Flemmmg & Courtney (1984) also comHealth Locus of Conhol bistrument, Canhil's Well-Bemg puted correlations on the Rosenberg scale and other selfLadder, the Health Promohng Lifestyle Profile and the ratmg scales, and they conduded that the psychometnc Health Percephons Queshonnaire Socio-demograpbc data properhes of the Rosenberg scale are sahsfadory and an measunng age, race, mcome, educahon, mantal status and acceptable measure of self-esteem (Flemmmg & Courtney employment were colleded on all subjeds for time one 1984) These vanables have been ldentifred m the Pender model and mrelatedresearch as bemg correlated with the health of adult women (Duffy 1988) Two queshons were mduded Healtii Locus of Control (HLQ on the demograpbc data sheet wbch required the subjects Tbs 11-item instrument developed by Wallston d al to mdicate from a bst of pre-determmed vanables those (1978) measures the extent to wbch mdividuals bebeve that promot«l and those that prevented adherence to a they conhol certam important health-related issues m their programme of planned exerase Subjeds could check more bves Each item is a bebef statement with which the subjed than one choice if appropnate IS asked to agree or disagree, usmg a six-pomt Likert scale Cantril's Weil-Being Ladder (CWBL) This instrument provides a global measure of perceived well-bemg (Cantnl 1965) Subjects were asked to seled from a self-andiormg lO-rung ladder, the rung on the ladder wbdi best represented the degree of well-bemg expenenced at that hme The top of the ladder (10) represents the best well-bemg possible As you go down the ladder, each rung represents less and less well-bemg The bottom of the ladder (I)representsthe worst possible wellbemg Pabnore & Kevett (1977) reported a 2-year testretest rebability (n = 378) for the Cantnl's Ladder for bfe satisfachon as r=0'65 Tbs modest rebabibty mdicates some stabibty m bfe sahsfadion scores over time
rangmgfrom(1) 'shongly disagree' to (6) 'shongly agree' lnformahon on the rebabibty and vabdity of the HLC has been well documented (Wallston et al 1976) Health Promoting Lifestyle Profile (HPLP)
This 48-item mstrument developed by Walker d al (1987) measures an mdividual's current prachce of behaviours that serve to mamtam or mcrease levels of wellness, selfactuabzahon and fulfilment It is a four-pomt summated rahng scale wbdi contams six subscales self-actuabzahon, health responsibibty, exerase, shess management, mterpersonal support and nutnhon All items are scores on the scale from I to 4 (I) 'never', (2) 'sometimes', (3) 'often' and (4) 'routinely' A bgh score mdicates a bgher health-promohng level of behaviour m all instances Item analysis, fador analysis and rebabibty estimates have beoi calculated on the mstrument The tool Roseiriwrg's Self-Esteem Scale QHSES) has been found to have sahsfadory empmcal e^adence of This 10-item instnonoit measures the self-acceptance rebabibty and vabdity The alpha rebability coeffiaent for aspect of self-esteem (Rosenberg 1965) The subjeds were the total scale was 0-92, n=952, alpha coef&aents for the asked to uKiicate tl* degree of agreemrait with each of the subscales ranged frcan 0 70 to 090, n=952 The test10 items on a four-pomt Likert scale ranging from (I) retest rebabibty over a 2-week penod was bgh (r=0^2) shoi^ly agree' to (4) 'strongly disagree' Lower scores reflect b g ^ self-esteem Rosed^erg (1965) rqKMted a Health Percepdcms Questionnaire (HPQ) reproeated-me»ures ANOVA was &iniadted to determme if parhapahcm in m exsTOse inx>gramme mad^ a difiference m sc(»%s cm the spealic
Phyacal achvity
14 Experfmentol
12
tbee The total mean scores for the confrol group were 106 76 (SD 10 58), 107 51 (SD 9 73) and 108 10 (SD 10-31) respectively The potenhal score range on the total scale was 32 to 160 A repeated-measure ANOVA revealed there was no sigmfieant mteraehon or mam effect of group and/or time on the dependent vanables
Control
10
Health Locus of Control
Time
Figure 3 The mteraetion effeet of group and tune on exerase
Percephons of health locus of eontrol m the expenmental and eonhol group remained fairly stable for hmes one, two and three The means and standard deviahons for this vanable are shown m Table 5 Seores for botb groups melieate that subjeets believe health outeomes are under their own eontrol Seores are m the direehon of supporhng an intemal loeus of eontrol The expenmental manoeuvre of attendjmce at exerase dass had no sigmficant impact on subjeets' scores on tbs vanable F(l,90) = 2 04, P< 0157 Self-Esteem
Control
Experimental
16
The impact of exerase on tbe subjects' basie feelmg of self-worth was determmed by comparmg scores on the Rosenberg Self-Esteem Scale for times one, two and tbee The subjects' measures of self-esteem improved over time for both groups (Table 6), but a repeated-measure ANOVA revealed that the differenee was not sigmfieant F(l,90) = 005,P
The relationships between physical activity and health-promoting behaviours in mid-life women Angela GilliS RN BSeN MAdEd Associate Profasor mid Chair and Anne Perry RN BN MSe(A) Assistant Profe^or, Department of Nursing, St Francis Xavier Untverstty, Anhgontsh, Nova Scotia, Canada
Accepted for publicahon 27 June 1990
GILLIS A k PERRY A (1991) Joumal of Advanced Nursing 16,299-310 Hie relationships between physical activity and health-promoting behaviours in mid-life women The purpose of this study is to examine the mter-relationships among partiapation between physical activity, perceived well-being, self-esteem, health status, health locus of control and health-promoting lifestyle activities of middleaged women in a rural context Smce women compnse the largest (hving) segment of the older population, it is extremely important to study the predictors of health promotmg activities in this age group pnor to old age Recent research has noted that partiapation in physical activity acrts as an entry pomt to reflect on health behaviour practices, creahng a npple efifect to other areas of lifestyle behaviours The research questions were framed withm the Health Promotion Model The study utilized a longitudinal experimental design.
Thefindmgsof this study contnbute to the sntall, but growmg, research base on physical activity as a health-promotmg lifestyle behaviour and physical activity as a hfestyle practice of women m the mid-hfe years Findmgs also provide a base for the development of commuruty physical-activity programmes and other intervention strategies to enhance the health and well-being of women, particularly those residing in a rural settmg DEARTH OF RESEARCH Until recently, httk research has been available on the health or lifiestyk prachces of mid-life women, parhcularly those resiciing m a rural sethng The mid-hfe penod, defmed as the span of hme fi'om 35 to 65 years, is a umque tune penod m which mdividuals evaluate their assets of health, money,relahonships,etc, as they prepare for the 'pnvileged years' (Ddfy 1988). Smce women compnse the lamest livuw segment of the older populahon, it is ° ^ °
extremely unportant to study the predictors of healthpromotmg achvihes in this group pnor to old age so that nursmg strategies to promote health can be faahtated and implemented The purpose of tbs study was to examme the mter-relahonsbps among parhapation between physical aehvity, perceived weU-bemg, self-esteem, health locus of control and health promotmg lifestyle achvities of nudcile-aged women m a rural context jhe foUowing questions were mvestigatecL r-^ _^ . f i . i ^ « _ 1 E>oes parhapahon m a programnwot physical achvity prociuce change m tlw lifestyle practices of mid-hfe rural wcHiiai? 299
A GiUtsimdA. Perry FigiB« 1 Health Promotion Model (produced by permissicm of Pender N J
Cognitive-perceptual factors
Modifying factors
Partictpation in healtti-promottng tiehaviour
Health Prowiohtm m Nursing
Pradice 2nd edn Appleton k Lange, Norwalk, Connecticut, 1987)
Importance of health
Demographic characteristics
Perceived control of health
Biological characteristics
Perceived self-efficacy
Interpersonal influences
Definition of health
Situational factors
Perceived health status
Behavioural factors
Likelihood of engaging in health-promoting behaviours
Cues to action Perceived benefits of health-pronnoting t>ehaviours
Perceived barriers to health-promoting behaviours
2
To what extent does parhapahon m a programme of physical achvity influence the well-bemg, health status, self-esteem and health locus of conhol of rural mid-bfe wcnnai? 3 To what extent do self-esteen, perceived well-bemg, health status, health locus of conhol and healthpromoting behaviours predid adheraice to a programme of phystcal achvity by rural mid-bfe women over time? 4 What are the perceived factors that influence adherence to prc^rammes erf physical achvity fOT rural mid-bfe women? Pender's (1987) Health Prcwnohon Model provided the theorehcalframeworkfor the study Hie model postulates that people's percephons of thonselves and their health (cogmhve-perceptual factors) and modifymg factors play an influential role m determmmg their lifestyle achvihes Cogmhve-perceptual fadors indude the importaiKe of health perceived conhol of health, self-eflficacy, debmhon of heekh health status, perceived benefits (rf healthfS-OTnohng behaviours ard perceived bamers to healthpromohng behaviours The modifymg factors indude
demograpbcs, biological charadenshcs, mterpersonal mfluences, situahonal facton and behavioural fadors The model has two phases deasion makmg and takmg achon The cogmhve-perceptual aiKl modifymg factors conshtute the deasion-makmg phase The tatang-achon phase mdudes cues that mohvate an mdividual to engage m health prc»notmg behaviour (Figure 1)
SELECTED LITERATURE REVIEW Health stahshcs durmg ^ e past several decades refled changes in tiie health stahis of women (Rice & Augbam 1980) Accordmg to the Surgeon General's report Healthy People (US Deparhnent of Health, Educahon and Welfare 1979), less tihan 25% (rf non-inshhitionabz^l adults are mvolved m regular exerase The Can»la Health Survey (I98I) rujted that less than 22% of females 65 years and olda- are mvolved m regular exeiose The tandoicy towards reduced physical iK:hvity as age mo^ases is a potenti^ concern, as tt has been noted that individuals over the ^ e of 45 who are modendely to very achve are significantly less likely to display bdiaviour related to tli
Physical achvity
(Stewart et al 1985) Pnce & Luther (1980) mdicate that older adults can minimize the losses associated with agemg through physical achvity Although there has been a recent profusion of bterature on health-promohng behaviours, bttle is known about what mohvates mdividuals to make changes in their bfestyle and to mamtam these changes over time Studies over the past 30 years have explored the benefits of physical adivity in relahon to prevenhon of disease processes (Paff«ibarger 1978) More recently, studies have ldenhfied the relahonship between physical activity even at minimal levels, and posihve health status (Wiley & Comacho 1980, Allan 1985, Gillett 1988, Gillett & Eisenman 1987) Recent researdi has noted that parhapahon m physical achvity acts as an entry pomt to refled on health behaviour prachces creatmg a npple effed to other areas of lifestyle behaviours (O'Hagan 1984, Shestowsky 1983) Pender (1987) noted that a sedentary bfestyle often snerges early m females and becomes highly resistant to diange These years of machvity in females may he either a major cause or a contnbutmg factor to degenerahve dianges that are assoaated with ageing and chronic lllness In recent bterature, regular physical exercise has been mcreasmgly recognized as a lifestyle parameter to be assoaated with a state of enhanced health and weU-bemg (Gillett k Eisenman 1987, Walker et al 1988) Mounting evidence suggests that lifestyle choices such as regular partiapahon m physical achvity influences both health and longevity (Berkman & Breslow 1983) Despite mcreasmg support for the relationsbp between bfestyle pradices and health, and the reabty of epidemiological data which reflect bgher morbidity for women than for men, mid-bfe women have only recently begun to be mduded as a target of bfestyle-onented health-promoting programmmg and research (Woods 1981) AlHiough there IS a vast amount of bterature available on the benefits of partiapahon m physical achvity (Lenskyj 1988), bttle IS known about the speafic relahonships tbs study was designed to test There exists a need to explore the role of physical achvity as a health-promohng, rather than disease-preventmg lifestyle behaviour for mid-bfe women Factors related to the promc^on of health seekmg behavltHirs m mid-bfe woa«n remam to be dearly tdenttfied m tt^ bt^ature.
O,
O.
O3 O3
X
R=random assignment, O = observahon, X=expenmental treatmait Figure 2 Expenmental Model
promotmg bfestyle behaviours m mid-bfe women residmg m a mral settmg (Figure 2) Sample and setting One hundred and twenty-six women, aged 35-65 years, residmg m a rural eastem Canadian provmce, volunteered to parhapate m the study Rural was defined as an area with a population of less than 7500 Women were recnuted fi-om a broad vanety of sources, mdudmg women's organizahons. Church groups, soaal organizahons, home and school associations, etc Indusion cntena for sample selechon mduded females between 35-65 years of age, able to speak and understand Engbsh, residing m mral eastem Nova Scotia, wilbng to partiapate m a 12-week programme of physical adivity, and with an absence of any cardiovascular health problems Subjects were randomly assigned to an expenmental (n = 63) and a conhol group (n = 63) usmg a schedule generatedfi-oma table of random numbers Eighty-seven per cent of the women m the experimental group (n = 52) and 66% of the women m the conhol group (« = 40) completed the study questionnaires for time one, time two and time tbee The remaining 13% and 34% of the respechve cases who failed to complete any one of the three sets of questionnaires were removed fi-om the data analysis Data analysis was therefore based on a sample of 92 cases Table 1 summarizes the major background charadenshcs of the sample Experimental manoeuvre
Subjects m the expenmental group parhapated m three 60mmute exerase dasses eadi week for 12 weeks. Exerase dasses were m the form of choreographed dance rouhnes to music that followed the guidelmes outlmed by Pender (1987) The programme mduded a warm-up penod, shetdimg exerase, aerobic exerases and a cool-down METHOD pencKl The exerase programme was under the direchcMi of A kmgih«bial jaetest-post-test expenmaital design was an expenenced and appropnately quabfied fitness lnstmcused to shidy the mter-rekhonsbps among parhapahcm tor /Ul dasses were held m a ccHnmooalfitnesscentre. The beh¥e«i jAyacal achvity, perceived well-bang, health exerase programme was made available to the ccmhol ^atus, setf-esteeoi, health locus of control and health- group followmg annpletion of data ailledKxt 301
AGilhsmdAPeny Table 1 Distnbuhon of bacjcground cJuraetenshes
Group
Vanable
Age
Expenmoital (« = 52)
Conhol
Mean
Mean
(« = 40)
SD 845 3 28
4451 14-64
Years of sehoohng
n Employmoit status Employed Unot^loyed Annual lneome* Bebw 8000 8000-12 000 Above 12 000 Ethme bacjcground White Blaek Otlwr Mantal status Single Mamed Divoreed/separated Widowed Parhapation in exerase at least onee per wedc dunng past 3 mcmths
Yes No
%
4451 1414
707
2 76
n
%
214
78-6
40 12
769 231
32 8
9 3 36
173 58 692
8 4 23
24 3 10-8
52 0 0
lOOO 0-0 OO
38 1 I
95 2
3
4 32 3 1
11-9 78-6
3 2
58 846 58 38
40 12
76-9 231
30 10
244
36 16
69-2 30-8
29 11
275
44
649
24 24
71 24
75-6
Family partieipahcm m exereise at least onee per week, dunng past 3 months
Yes No
72 5
'Mratmgdata.
Aciherence to the exerase programme was measured duectly by a self-rq>ortmg attendance record Tbs consisted of a senes of 36 blanks w b d i represented the 36 exercise dasses held dunng the poiodtrf the study Parhapants placed a cJ:Kdc mark m the blsik if diey attended dass and \ek the blank empty if they were absentfrcrniclass for any reason.
A ktter ctf mtroduchcai was forwarckd to women on the m a ^ g lists erf vanous Churcit groiqps, womoi's organizatu»is, ccHnmuttty groups, etc, m rural easton Nova 302
Scotia. Women mterested m parhapatmg m the study retumed a form of mtoit to the researdiers by mml Those who indicated a wdlmgness to partiajoAe were contacted by phcHie, by the researches, and invited to attend a goieral meehng for the puqxjse of danfyu^ details of the study At tlK meeto^ a veri»l and wntten descrq7tic»i of the study was provicied and wntten cxm^nt obbuned botn aU partiapants Partiapants were then raiuicnnly assigned to either the experunental or ccmtrol group Data were cdlected from both grcH^^s at dvee distinct tune pemyds m tiie study at ^ K genend swetmg erf nitrodudKHi (time one), inuiKdiidely fblloimng caBq»btKm of ihe 12-week exeiase programnme (tone hvo) md a§mn 6 mcHiths kto- (tmne three)
Physical adtvtty
INSTRUMENTS
Convergent vabdity has been estabbshed witti the CPI Self-Acceptance Scale {r—0-27, w=643), and with a OTieFive self-reporhng mstruments were admmistered m the ltem self-esteem scale (r=0 45,« = 643, r = 0 66, H = 101) same order dunng each data collechon penod and retumed For discnminant vabdity, correlations with measures of m stamped, addressed envelopes to the researchers The self-stabibty are moderate (r=021 to 053) (Robmson & instruments mduded Rosenberg's Self-Esteem Scale, the Shaver 1973) Flemmmg & Courtney (1984) also comHealth Locus of Conhol bistrument, Canhil's Well-Bemg puted correlations on the Rosenberg scale and other selfLadder, the Health Promohng Lifestyle Profile and the ratmg scales, and they conduded that the psychometnc Health Percephons Queshonnaire Socio-demograpbc data properhes of the Rosenberg scale are sahsfadory and an measunng age, race, mcome, educahon, mantal status and acceptable measure of self-esteem (Flemmmg & Courtney employment were colleded on all subjeds for time one 1984) These vanables have been ldentifred m the Pender model and mrelatedresearch as bemg correlated with the health of adult women (Duffy 1988) Two queshons were mduded Healtii Locus of Control (HLQ on the demograpbc data sheet wbch required the subjects Tbs 11-item instrument developed by Wallston d al to mdicate from a bst of pre-determmed vanables those (1978) measures the extent to wbch mdividuals bebeve that promot«l and those that prevented adherence to a they conhol certam important health-related issues m their programme of planned exerase Subjeds could check more bves Each item is a bebef statement with which the subjed than one choice if appropnate IS asked to agree or disagree, usmg a six-pomt Likert scale Cantril's Weil-Being Ladder (CWBL) This instrument provides a global measure of perceived well-bemg (Cantnl 1965) Subjects were asked to seled from a self-andiormg lO-rung ladder, the rung on the ladder wbdi best represented the degree of well-bemg expenenced at that hme The top of the ladder (10) represents the best well-bemg possible As you go down the ladder, each rung represents less and less well-bemg The bottom of the ladder (I)representsthe worst possible wellbemg Pabnore & Kevett (1977) reported a 2-year testretest rebability (n = 378) for the Cantnl's Ladder for bfe satisfachon as r=0'65 Tbs modest rebabibty mdicates some stabibty m bfe sahsfadion scores over time
rangmgfrom(1) 'shongly disagree' to (6) 'shongly agree' lnformahon on the rebabibty and vabdity of the HLC has been well documented (Wallston et al 1976) Health Promoting Lifestyle Profile (HPLP)
This 48-item mstrument developed by Walker d al (1987) measures an mdividual's current prachce of behaviours that serve to mamtam or mcrease levels of wellness, selfactuabzahon and fulfilment It is a four-pomt summated rahng scale wbdi contams six subscales self-actuabzahon, health responsibibty, exerase, shess management, mterpersonal support and nutnhon All items are scores on the scale from I to 4 (I) 'never', (2) 'sometimes', (3) 'often' and (4) 'routinely' A bgh score mdicates a bgher health-promohng level of behaviour m all instances Item analysis, fador analysis and rebabibty estimates have beoi calculated on the mstrument The tool Roseiriwrg's Self-Esteem Scale QHSES) has been found to have sahsfadory empmcal e^adence of This 10-item instnonoit measures the self-acceptance rebabibty and vabdity The alpha rebability coeffiaent for aspect of self-esteem (Rosenberg 1965) The subjeds were the total scale was 0-92, n=952, alpha coef&aents for the asked to uKiicate tl* degree of agreemrait with each of the subscales ranged frcan 0 70 to 090, n=952 The test10 items on a four-pomt Likert scale ranging from (I) retest rebabibty over a 2-week penod was bgh (r=0^2) shoi^ly agree' to (4) 'strongly disagree' Lower scores reflect b g ^ self-esteem Rosed^erg (1965) rqKMted a Health Percepdcms Questionnaire (HPQ) reproeated-me»ures ANOVA was &iniadted to determme if parhapahcm in m exsTOse inx>gramme mad^ a difiference m sc(»%s cm the spealic
Phyacal achvity
14 Experfmentol
12
tbee The total mean scores for the confrol group were 106 76 (SD 10 58), 107 51 (SD 9 73) and 108 10 (SD 10-31) respectively The potenhal score range on the total scale was 32 to 160 A repeated-measure ANOVA revealed there was no sigmfieant mteraehon or mam effect of group and/or time on the dependent vanables
Control
10
Health Locus of Control
Time
Figure 3 The mteraetion effeet of group and tune on exerase
Percephons of health locus of eontrol m the expenmental and eonhol group remained fairly stable for hmes one, two and three The means and standard deviahons for this vanable are shown m Table 5 Seores for botb groups melieate that subjeets believe health outeomes are under their own eontrol Seores are m the direehon of supporhng an intemal loeus of eontrol The expenmental manoeuvre of attendjmce at exerase dass had no sigmficant impact on subjeets' scores on tbs vanable F(l,90) = 2 04, P< 0157 Self-Esteem
Control
Experimental
16
The impact of exerase on tbe subjects' basie feelmg of self-worth was determmed by comparmg scores on the Rosenberg Self-Esteem Scale for times one, two and tbee The subjects' measures of self-esteem improved over time for both groups (Table 6), but a repeated-measure ANOVA revealed that the differenee was not sigmfieant F(l,90) = 005,P
