Full text of DOI:10.1017/S0714980800000635

Articles
Using Alternative Therapies to Manage Chronic
Illness Among Older Adults: An Examination of the
Health Context, Predisposing and Enabling Processes
Andrew V. Wister, Minda Chittenden, Bonnie McCoy, Kelly Wilson, Trasey Allen,
and Melanie Wong
Simon Fraser University at Harbour Center*
RÉSUMÉ
Cet article traite de l’utilisation de médecines douces pour gérer une maladie chronique chez les adultes âgés qui souf-
frent d’au moins une des trois affections suivantes: arthrite, maladie du cœur et hypertension. En tirant des conclusions
de la littérature reliée à la santé, l’accent est mis sur le contexte de la maladie, les facteurs prédisposants et plusieurs fac-
teurs qui semblent inciter certaines personnes à utiliser des médecines douces. Des données de départ (n = 87) tirées
d’une étude effectuée en 1995–96 à Vancouver, Canada (North Shore Self-Care Study) ont été utilisées. On a analysé deux
variables dépendantes en utilisant des techniques de régression logistique – la première se fonde sur une question glo-
bale concernant l’utilisation des médecines douces (les herbes médicinales, l’acupuncture, les massages thérapeutiques,
etc.) pour gérer un état chronique et la deuxième porte sur une question plus spécifique relative à la méditation ou la
prière. Les résultats de la première analyse montrent que le fait d’être plus jeune, de souffrir d’arthrite par rapport à
l’hypertension, de comorbidité, de consommer moins de médicaments, d’avoir des revenus plus bas, de lire sur son état
chronique et d’agir sur soi à partir de ses lectures sont associés à l’essai des médecines douces. Les conclusions de la
deuxième analyse montrent que le fait d’être une femme, plus jeune, célibataire, d’avoir un état plus grave et une mala-
die plus longue, le lien entre des niveaux modérés d’entraide et le nombre de confidents, incitent vraisemblablement à
avoir recours à la méditation et à la prière. Les répercussions de ces résultats sont abordés sous l’angle de leur significa-
tion théorique et du degré de pertinence de la complémentarité de la médecine conventionnelle et non conventionnelle.
ABSTRACT
This paper examines use of alternative therapies to manage a chronic illness among older adults with at least one of three
major conditions: arthritis, heart disease, and hypertension. Drawing from developments in the health utilization litera-
ture, a focus is placed on the illness context, predisposing factors, and several factors deemed to enable persons to use
complementary medicine. The baseline data (n = 879) from the 1995–96 North Shore Self-Care Study conducted in Van-
couver, Canada were used for this study. Two dependent variables were analysed using logistic regression techniques –
the first is based on a comprehensive question about using alternative therapies (such as herbal remedies, acupuncture,
massage therapy, etc.) to manage a chronic condition; and the second uses a more specific question pertaining to medita-
tion or praying. The results from the first analysis show that being younger, suffering from arthritis compared to hyper-
tension, comorbidity, taking fewer medications, lower income, reading on the chronic condition, and the interaction
between reading and illness self-efficacy are associated with trying alternative therapies. The findings for the second
analysis show that being female, being younger, and not married, as well as reporting a more serious condition, illness
duration and the interaction between having moderate levels of mutual aid and number of confidants result in a greater
likelihood of trying meditation/prayer. Implications of these results are discussed in terms of their theoretical import,
and their relevance for the degree to which unconventional and conventional medicine are complementary.
Manuscript received: 18/10/99
Manuscript accepted: 10/09/00
Keywords: Alternative Therapies, Illness Management, Complementary
Mots clés: médicines douces, gestion de la maladie, complémentaire Requests etc.:
Canadian Journal on Aging/ La Revue canadienne du vieillissement 21 (1)
47

48 Canadian Journal on Aging 21 (1)
Andrew. V. Wister et. al.
Introduction
Review of Literature
The field of self-care has proliferated in recent years in
recognition of the fact that individuals are becoming
more involved in all aspects of their health care (Ory
& Defriese, 1998). Self-care has been defined as a con-
tinuum of caring for the self to enhance health, pre-
vent disease, evaluate symptoms and restore health,
which is organized by the perceptions, decisions and
options available to each individual (Dean & Kick-
busch, 1995, p. 36). One group of self-care practices
that has received considerable attention has been
what is termed “alternative therapies,” also known as
“complementary medicine” and “unconventional
medicine” (Eisenberg, Kessler, Foster et al., 1993).
These terms encompass a range of alternative prac-
tices, including manual healing (such as acupuncture,
massage, and chiropractic); herbal medicine (such as
ginkgo, ginseng, St. Johns’ Wort); vitamin therapy;
mind/body control (such as meditation or praying);
bioelectromagnetic applications; and various pharma-
cological treatments (see OAM–NIH Classification of
Alternative Medicine Practices, 1999). Although the
efficacy of many of these therapies remains controver-
sial, it is clear that they are becoming popular self-care
practices, and ones for which individuals appear to be
willing to pay. In fact, it is estimated that the cost of
unconventional medicine in the United States equals
the annual expenditures on hospitalization by indi-
viduals (Eisenberg et al., 1993).
The Prevalence and Cost of Alternative Therapy Use
The use of alternative therapies in the general popula-
tion appears to be on the rise in many nations. A 1992
national survey conducted in the US revealed a preva-
lence rate of alternative therapy use in the previous
year of approximately 34 per cent, and about one-
third (36%) of these individuals visited providers for
unconventional therapy (Eisenberg et al., 1993). These
“unconventional” practices included: relaxation, chi-
ropractic, massage, imagery, spiritual healing, com-
mercial weight-loss programs, lifestyle diets, herbal
medicines, megavitamin therapy, self-help groups,
energy healing, biofeedback, hypnosis, homeopathy,
acupuncture, folk remedies, exercise, and prayer. The
Canada Health Monitor found that one in five Canadi-
ans had used an alternative therapy within the pre-
ceding six months (CMAJ, 1991). Furthermore, data
from the 1993/94 Canadian National Population
Health Survey indicates that about 15 per cent of
adult Canadians have consulted an alternative health
practitioner in the year prior to the survey (Millar,
1997). European statistics show rates ranging from 20
per cent of the population in the Netherlands to 49 per
cent in France (Fisher & Ward, 1994). And in Aus-
tralia, a high prevalence rate was also found; 48.5 per
cent of South Australians used at least one medically
prescribed alternative medicine and one-fifth visited
an alternative practitioner in the preceding year
(MacLennan, Wilson, & Taylor, 1996). The variability
of the rates is likely dependent on health insurance
coverage, availability and accessibility of various ther-
apies, normative and belief systems, advertising,
acceptability among the medical professions, and
other population characteristics.
The apparent rise in alternative therapy use is espe-
cially significant given population aging, increased
life expectancy, and the concomitant expansion in the
numbers of individuals managing chronic illness,
which may magnify current patterns of use. Yet, little
is known about how and why people are led to try
alternative health practices, especially linking pat-
terns of use for specific types of alternative therapies
to specific types of health problems. Indeed, no stud-
ies have explicitly examined these issues for older
populations managing a chronic illness. In response to
these research gaps, this paper provides an explora-
tory study of alternative therapy use among older
adults aged 50 and over who have been diagnosed
with one of three common chronic conditions – arthri-
tis, heart disease, and hypertension as part of a study
of self-care and aging. The purpose of the present
research is to better understand why alternative ther-
apy is integrated within an illness management sys-
tem, including the elaboration of relationships
between the formal medical system and this type of
self-care behaviour.
Specific studies of chronically ill populations indicate
that the use of alternative therapies is even more prev-
alent in this group than in the general population.
Surveys of arthritis and rheumatism sufferers using
alternative therapies have reported prevalence rates
ranging from 82 to 94 per cent (Boisset & Fitzcharles,
1994; Kestin, Miller, Littlejohn, & Wallgrist, 1985; Kro-
nenfeld & Wasner, 1982). In their Montreal study of
235 rheumatology patients, Boisset and Fitzcharles
(1994) found that, over a 12-month period, 54 per cent
used over the counter products, 39 per cent used
prayer, relaxation and meditation, and 13 per cent had
visited an alternative practitioner. In the 1993/94
National Population Health Survey, the number of
chronic illnesses reported was positively associated
with the use of alternative therapies. Prevalence
increased from 9 per cent in the population reporting
no chronic disease to 26 per cent of subjects reporting
the presence of three or more chronic conditions (Mil-
lar, 1997).

Alternative Therapies
La Revue canadienne du vieillissement 21 (1)
49
The economic costs associated with alternative medi-
cine are also noteworthy, both to the consumer and to
the public health providers. In England and Wales, it
was estimated that the public spent between £150 mil-
lion and £450 million on “unconventional” treatments
from a professional, with an additional £250 million
spent on supplements and £100 million on natural
remedies (Richmond, 1991). In an Australian study, it
was estimated that $AU621 million (Australian dol-
lars) were spent for alternative medicines and $AU309
million for alternative therapists, per annum
(MacLennan et al., 1996). Extrapolating their results to
the entire population, Eisenberg et al. (1993) estimated
that Americans spent a total of $13.7 billion, with
$10.3 billion paid by the consumer (Eisenberg et al.,
1993). This was compared with the $12.8 billion spent
by the consumer, annually, on all hospitalizations in
the United States. Canadian costs have yet to be calcu-
lated. Thus, rates of alternative therapy use are signif-
icant, especially among persons with a chronic illness;
and these patterns have enormous economic implica-
tions.
Although some medical practitioners refer patients to
alternative therapists, one of the most commonly cited
reasons for seeking alternative medicine has been the
perceived failure of allopathic medicine in treating
pain and other symptoms (Marshall et al., 1990; Moore,
Phipps, & Marcer, 1985; Vincent & Furnham, 1996).
Moore et al. (1985) found that 22 per cent of subjects
seeking assistance at the Centre for Alternative Thera-
pies in the United Kingdom were referred by a medical
practitioner, but that 54 per cent had noted that the
main reason for attending was due to the failure of con-
ventional medicine. Marshall et al. (1990), who inter-
viewed practitioners in New Zealand, found that 68.7
per cent of the doctors referred patients to alternative
therapists and that about 39 per cent cited failure of
conventional medicine as the reason for their referral.
In a similar vein, other researchers suggest that alter-
native therapy use is an attempt by individuals to
increase their control over health problems (Montbri-
and & Laing, 1991; Murray & Shepherd, 1993), in part,
due to the anxiety produced by unsuccessful treat-
ments of chronic disease using medications and other
forms of allopathic medicine. However, while scepti-
cism, the need for autonomy and control, and dissatis-
faction with conventional medicine may be sufficient
motivators, people also require information on uses,
sources and costs of alternative therapies, either
through reading on the subject, via information chan-
nels, or through their support networks. Additionally,
use of alternative therapies may be linked to people’s
health orientation (Astin, 1996). An extension of our
understanding about the pathways through which
health beliefs and health experiences affect alternative
therapy use therefore requires consideration of a con-
stellation of decision-making elements (Kart & Engler,
1994; Wister, 1995).
Major Correlates of Alternative Therapy Use
Studies of correlates of alternative therapy use have
met with inconsistent findings. MacLennan et al.
(1996) found that users of alternative therapies were
more likely to be women, better educated, to drink
more, be of normal body weight, and employed com-
pared to non-users. Using data from the 1994–95
National Population Health Survey, it was deter-
mined that alternative health care providers were uti-
lized more often by persons of higher education,
higher income, specific regions (British Columbians
had the highest rate), and those suffering a higher
number of chronic conditions (CMAJ, 1997). Boisset &
Fitzcharles (1994) found that higher income and
French-speaking Quebecers used more bought prod-
ucts, but found no other differences. Comparisons
across surveys, however, are complicated by: 1) differ-
ing sample sizes; 2) variation in the type/definition of
alternative therapy use; 3) differing time periods,
regions or countries; and 4) varying characteristics of
the sample group in terms of chronic disease type(s),
age, etc.
An Application of the Andersen–Newman Model
The original Andersen–Newman (Andersen & New-
man, 1973) model, and the revised model (Andersen,
1995), have been applied extensively to a variety of
health utilization behaviours (Andersen, 1995; Chap-
pell & Blandford, 1987; Fleming, Giachello, Andersen,
& Andrade, 1984; Wister, 1992; Wolinsky, 1978). Draw-
ing from this conceptual frame, we can organize the
socio–demographic correlates, health factors, reasons
for alternative therapy use, and other key explanatory
variables according to whether they are predisposing,
need or enabling factors. This provides a means
whereby prior research on correlates of use and dis-
cussion of substantive issues can be transformed into
the development of specific hypotheses that can be
tested, compared, and placed within the context of a
self–management decision model.
Reasons for Alternative Therapy Use
Preliminary research suggests that people seek out
alternative therapies for a number of reasons (Kelner &
Wellman, 1997). One of these is dissatisfaction or scep-
ticism with conventional medicine (Furnham & Forey,
1994; Sutherland & Verhoef, 1994; Verhoef, Sutherland
& Brkich, 1990). Eisenberg et al. (1993) estimate that
about 25 per cent of people who see their medical doc-
tor for a serious problem also use alternative therapies
in combination with allopathic (traditional) medicine.

50 Canadian Journal on Aging 21 (1)
Andrew. V. Wister et. al.
Predisposing Factors -- Predisposing
factors
are
On one level, doctor visits and prescription medicines
may be deemed to be indicators of need, because per-
sons tend to seek treatment when experiencing symp-
toms of chronic disease. But on another level, they
also directly measure involvement with the formal
health care system. Previous studies have established
that persons who use alternative therapies tend to be
concurrently under the care of a doctor (e.g., Eisen-
berg et al., 1993; Northcott & Bachynsky, 1993). How-
ever, it is not clear whether past utilization of the
formal system affects current use, whether this associ-
ation holds when accounting for other factors affect-
ing patterns of use, nor whether the same relationship
is apparent for prescription drug use. Given the con-
troversy surrounding the complementary nature of
allopathic and unconventional medicine (Lewith,
1997), we will explore whether greater formal care use
facilitates or inhibits the use of alternative therapies
through an examination of the impact of past doctor
visits and medication use on current use of alternative
therapies.
deemed antecedent to the onset of illness and influ-
ence the proclivity of an individual to use health serv-
ices, or in this case, alternative therapies. Gender, age,
marital status, education, and health beliefs are con-
sidered to predispose an individual to use alternative
therapies (Boisset & Fitzcharles, 1994; MacLennan et
al., 1996). In line with correlational studies, we hypothe-
size that being female, being younger, and of higher educa-
tion will predispose a person to trying alternative therapies
(CMAJ, 1997; MacLennan et al., 1996). While we have
no formal hypothesis for marital status, we explore its
impact on patterns of alternative therapy use.
Need Factors -- Need factors pertain to the objective
and subjective health states of individuals, which are
anticipated to increase demand for health services,
including the likelihood of trying alternative therapies
(Gray, 1985; Sutherland & Verhoef, 1994). Illness
dynamics may lead persons to seek out alternative
therapies to manage their condition, or as a response
to dissatisfaction with aspects of the formal health
system in dealing with their health problems (Murray
& Rubel, 1992; Sutherland & Verhoef, 1994; Vincent &
Furnham, 1996). It is expected that persons with
chronic conditions that are typically associated with
more overt symptoms, such as presentation of pain
and limited activities of daily living, would demon-
strate a propensity to use more unconventional cop-
ing strategies than those with less disruptive
symptoms (McDonald-Miszczak, Wister & Gutman,
2001). In this study, three major chronic disease types
have been examined, which can be ordered in terms of
degree of overt symptoms. Persons with arthritis were
expected to present more symptoms disruptive of
daily activities than those reporting hypertension.
Those reporting heart disease as their most serious
chronic condition were expected to fall between the
former two. For example, persons with angina may
present pain and discomfort, but only upon physical
exertion. Among persons with more than one condi-
tion, co-morbidity, or the number of other chronic
conditions, is expected to be positively associated
with alternative therapy use, given that multiple con-
ditions complicate and exacerbate problematic symp-
toms of chronic disease. Similarly, persons who
perceive their main condition as more serious would
also be more likely to use complementary medicine.
In addition, it is anticipated that, the longer the dura-
tion of the major chronic illness (i.e., time since diag-
nosis), the greater the tendency to engage in
alternative responses to health (Gray, 1985). Thus, we
hypothesize that chronic illnesses with more overt symp-
toms, comorbidity, higher perceived seriousness of the ill-
ness, and longer duration of the illness will lead to a greater
likelihood of alternative therapy use.
Enabling Factors -- According to Andersen (1995), ena-
bling factors are conditions affecting access to alterna-
tive therapies. Consistent with other research, greater
income is expected to be associated with trying alternative
medicine (Wolinsky & Johnson, 1991; CMAJ, 1997).
Advancements of behavioural models applied to
older chronically-ill populations have led to an expan-
sion of the enabling component in a number of ways
that one would anticipate are important for alterna-
tive therapy use. These include: support from confi-
dants embedded within the support network, contact
with other individuals with the same problem, cur-
rent levels of confidence/control or self-efficacy that
one has in dealing with health problems, and sources
of information, such as reading on the subject (Ban-
dura, 1986; Lorig & Holman, 1989; Wister, 1992; Zim-
mer & Chappell, 1994).
The social support network of individuals has long
been recognized as a principal facilitator for a variety
of health behaviours (Cohen & Syme, 1985). In addi-
tion, research has identified mutual aid as an impor-
tant form of support that is shared among persons
experiencing the same illness (Romeder, 1999). It is
anticipated that receipt of support through confidants, and
interacting with persons suffering from the same illness
(mutual aid), enables the use of alternative therapies. Fur-
thermore, research on self-management of chronic
disease (see for example, Lorig & Holman, 1989; Clark
et al., 1991) has demonstrated that self-efficacy is an
important factor affecting health behaviour. Further-
more, a major reason for seeking alternative medicine
is to increase control over one’s health (Montbriand &
Laing, 1991; Murray & Shepherd, 1993). It is expected
that lower levels of self-efficacy will be associated with

Alternative Therapies
La Revue canadienne du vieillissement 21 (1)
51
alternative therapy use. Given the growth in popular
self-help and self-care literature, and access to infor-
mation via the Internet, reading about one’s illness
condition would likely expose the person to knowl-
edge about alternative therapies. This comprises an
important facilitator in learning where and how to
obtain and try alternative therapies, especially those
that are less known. It is hypothesized that reading on the
subject will be associated with the use of unconventional
medicine to manage a chronic condition. We also exam-
ined interactions between all four of these enabling
factors because of the likelihood of simultaneous
influences of processes deemed to facilitate use. For
example, mutual aid combined with reading about an
alternative type of therapy may encourage use; and a
desire to increase control over one’s health in combi-
nation with reading on the subject, or social support,
may also act interactively.
cent) with more than one of these chronic conditions
were placed in the chronic condition category that
they self–reported as affecting them most. In keeping
with this theme, the majority of questions about man-
aging the chronic illness focused on the major condi-
tion.
Descriptive statistics show that the sample is com-
prised of 417 (47.4%) persons with arthritis/rheuma-
tism, 229 (26.1%) with a heart problem, and 233
(26.5%) with hypertension. The average age is 69.3,
with a range between 50 and 94. The gender break-
down includes 513 (58.4%) females and 366 (41.6%)
males. Also, 340 (38.7%) live alone, while 538 (61.3%)
live with a spouse and/or others. The socio-economic
profile of the sample shows that the North Shore sam-
ple is better educated, but similar in income level
compared to the general Canadian population aged 50
and over.
Methods
Measurement
The Sample
Dependent Variables -- Respondents were asked several
questions about a variety of self–care behaviours spe-
cifically used to manage their chronic condition. Two
of these questions asked about alternative therapies –
a general question and one specific to prayer/medita-
tion. These were measured using yes/no responses to
the following question predicated by the opening
statement: “People try various things to improve or
cope with their condition.”. . . “Which of these are you
doing now?” The two were: 1) try alternative thera-
pies, such as herbal remedies, acupuncture, massage,
etc; and 2) try meditation or praying. Of the total sam-
ple (n = 879), 133 (15.1%) responded in the affirmative
to the first question about alternative therapy use;
while 251 (28.6%) stated that they had tried medita-
tion or praying to cope with their chronic condition.
Examination of rankings of specific types of alterna-
tive therapies indicates that herbal, massage, and acu-
puncture were the most prevalent, but that many
others are also used, albeit at very low rates. In addi-
tion, meditation/prayer is considered to capture one
type of mind–body experience that falls within the
broad definition of alternative therapies. Overall, it is
argued that these two measures encapsulate impor-
tant dimensions of alternative therapy use, albeit,
some of these are combined rather than measured
separately as would be the ideal.
The first wave of the North Shore Self-Care Study,
conducted in 1995–1996, was used in this research.
Using random digit dialling, trained interviewers
completed telephone surveys with 904 older adults
aged 50 and over living in the municipalities of North
and West Vancouver, British Columbia, Canada.
While we deem persons 50 and over as “older adults,”
it is recognized that both middle-aged and elderly
persons are included in the sample. The response rate
to our request for participation among eligible sub-
jects for this sample was 72.2 per cent and is consist-
ent with telephone surveys. Eligibility based on the
illness criterion required respondents to have been
professionally diagnosed with one of four major
chronic illnesses, including: arthritis or rheumatism,
heart disease, hypertension and stroke. The small
number (25) of persons with stroke have been
removed, resulting in a sub-sample of 879 persons
used in the present analysis. Heart disease included:
arrhythmia (irregular heart beat), angina, myocardial
infarction, congestive heart failure, valvular disease,
and any other diagnosed heart condition. Hyperten-
sion was included because it tends to be linked with
fewer overt symptoms compared to the other types.
Since the purpose of the survey was to understand ill-
ness management behaviours, respondents (54 per

52 Canadian Journal on Aging 21 (1)
Andrew. V. Wister et. al.
Table 1: Frequency distributions, range, means and standard deviations for all variables*
Dependent Variable
Categories
Frequency
Percentage
Alternative Therapies
Doing Now
Not Doing Now
Missing
133
745
1
15.1
84.7
0.1
Total
879
100.0
Meditation/Prayer
Doing Now
Not Doing Now
Missing
251
625
3
28.6
71.1
0.3
Categorical Independent Variables
Gender
Male
Female
366
513
41.6
58.4
Marital Status
Education
Not Married
Married/Common Law
Missing
390
488
1
44.4
55.5
0.1
Some Sec. or Less
Completed Secondary
Post-Secondary
Missing
128
218
529
4
14.6
24.8
60.2
0.5
Income
<$20,000
194
307
142
68
22.1
34.9
16.2
7.7
$20,000–$49,999
$50,000–$69,999
$70,000–$99,999
$100,000 +
54
6.1
Missing
114
13.0
Most Serious Chronic Condition
Perceived Seriousness
Arthritis
Heart Disease
High Blood Pressure
417
229
233
47.4
26.1
26.5
Not At All Serious
Slightly Serious
Moderately Serious
Extremely Serious
Missing
198
255
349
74
22.5
29.0
39.7
8.4
3
0.3
Mutual Aid
Daily–Every 2 Weeks
Once/Month–Few/Year
Never
69
311
499
7.8
35.4
56.8
Reading on Subject
No
Yes
Missing
269
607
3
30.6
69.1
0.3
Interval Independent Variables
Missing Cases
Range
Mean
Standard
Deviation
Age
4
0
0
0
0
0
0
50–95
0–7
69.30
1.56
0.81
1.37
11.73
2.77
2.21
3.99
8.24
Comorbidity
Years Since Diagnosis
Number of Doctor Visits (last 3 months)
Number of Prescriptions
Number of Confidants
Illness Self Efficacy
0.1–83.6
0–30
12.40
2.57
0–21
2.37
0–50
4.20
9–55
36.45
*
Frequencies and percentages are provided only for categorical variables.
Independent variables -- A total of 15 independent vari-
ables were included into the analysis under the cate-
gories of predisposing, need, and enabling factors.
Table 1 shows the frequency distributions for the cate-
gorical independent variables, and selected conde-
scriptive statistics for continuous variables, as well as
missing values. The first set of factors includes the
four predisposing factors: gender, age, marital status,

Alternative Therapies
La Revue canadienne du vieillissement 21 (1)
53
and education. Gender was divided into males and
females. Participants’ chronological age at the time of
the survey was used. Marital status was dichotomized
into not married and married/common-law. Level of
education was collapsed into a three-point scale with
the categories: some secondary or less, completed sec-
ondary, and post–secondary education.
ease, and high blood pressure answered 11, 9, and 8
questions, respectively, to measure illness self-efficacy.
Each response was rated on a five-point Likert scale
ranging from 1 (not confident at all) to 5 (totally confi-
dent). Since the number of questions posed differed
according to the type of illness, a proportional score
was used (total response score/number of scale
items). Cronbach’s Alphas were computed for each
scale: arthritis self-efficacy = .89 (consistent with
Lorig’s research); heart problem self-efficacy = .84;
and hypertension self-efficacy = .74, indicating good
reliability. Reading about the chronic condition was
measured by a yes or no response to the question:
“Once diagnosed with (arthritis/disease/high blood
pressure), did you do some reading on the subject?”.
Although current reading on the main chronic illness
was also available, we chose the measure tied to diag-
nosis time in order to sequentially prioritize this vari-
able. The mean length of time since diagnoses for the
three illness types are as follows: arthritis – 12.7 years;
heart disease – 10.3 years; and hypertension – 12.6
years.
The second set of variables encompassed six measures
of need. The type of chronic illness included: arthri-
tis/rheumatism, heart disease, and high blood pres-
sure (hypertension). The number of co-morbidities
experienced by participants, in addition to the major
chronic condition affecting their lives, was calculated
by counting the total number of reported “other
health problems” as listed, including: asthma/
emphysema, anxiety, depression, cancer, diabetes,
neurological disease, Alzheimer’s disease, osteoporo-
sis, vision problems, hearing problems, and other
problems. Participants were also asked how serious
they perceived their condition to be (ranging from not
at all serious to extremely serious). Duration of the
main chronic condition (time since diagnosis) was
expressed in years and proportions of years, based on
the question: “when were you first professionally
diagnosed (by a doctor, nurse, physiotherapist etc.)
with your specific condition”. Number of doctor visits
was measured by the number of visits within the last
three months. Also, the number of prescriptions cur-
rently being taken was used to measure prescription
drug use.
Missing Data -- Since there were very few missing
cases (between 0 and 4 cases) for all variables except
for income, we used either mean or modal substitution
to recode these cases for continuous and categorical
variables, respectively. The large number (114, 15%) of
missing data for income were substituted using
income distributions across three other variables. We
used three age groups (50–64, 65–74, and 75+), two sex
groups, and three education groups (same as in Table
1) to redistribute the missing cases for income. This
results in more accurate substituted missing data
because of the associations between income and these
three variables, and does not seriously decrease the
standard deviation of the derived income variable.
The final grouping of variables included five enabling
factors – income, confidants, mutual aid, self-efficacy,
and reading on the subject of alternative therapy.
Income was measured using reported total household
income based on a five point scale: less than $20,000;
$20,000–$49,999; $50,000–$69,999; $70,000–$99,999;
and $100,000+. Confidants were identified in terms of
the total number of persons with whom one can con-
fide about personal matters. Mutual aid was based on
the question: “About how often do you discuss your
condition with another person who also has (arthri-
tis/heart disease/high blood pressure)” and was
divided into three categories, daily–every two weeks–
once/month–few times a year, and never. Also, each
specific illness type had a self-efficacy scale. We used
Lorig et al.’s (1989) Arthritis Self-efficacy Scale to
measure arthritis self-efficacy. This scale is comprised
of 11 questions pertaining to the participants' confi-
dence in controlling selected aspects of their condition
(see Appendix 1). In modifying the scale for the heart
and hypertension groups, we removed a few of the
arthritis-specific questions. In addition, some of the
questions were slightly reworded for the heart prob-
lem and hypertension groups in order to be relevant
to their condition. Subjects with arthritis, heart dis-
Logistic Regression Methods
Logistic regression is a statistical technique that has
been developed to perform multivariate analyses
using a dichotomous response variable with various
degrees of skewness (Morgan & Teachman, 1988;
DeMaris, 1995; Menard, 1995). The betas generated by
logistic regression are maximum likelihood estimates
for the linear regression of the latent dependent varia-
ble in log form. A simple transformation of the beta
coefficient exp(Bk) produces the estimated odds ratio
for persons who are a unit apart on Xk (for continuous
variables), or compared to a reference category (for
categorical variables), after statistically controlling for
all other predictors in the model (DeMaris, 1995).
The logistic regression beta coefficient, its level of sta-
tistical significance, an odds ratio, and the model chi-
square value are presented in this analysis. The logis-
tic regression coefficient represents the change in the

54 Canadian Journal on Aging 21 (1)
Andrew. V. Wister et. al.
log odds of using alternative therapies (compared to
not using them) for a one–unit change in an explana-
tory variable, while statistically controlling for all oth-
ers (Menard, 1995). Since the parameter estimate is
difficult to interpret as a log, it can be easily converted
into an arithmetic odds ratio, using the formula
[exp(B)]. The conversion is the estimated odds ratio
for those who are a unit apart on a given explanatory
variable, after other variables in the model have been
statistically controlled. In this study, it is the probabil-
ity of engaging in alternative practices for each incre-
ment of an interval variable or for one category of a
categorical variable compared to the reference cate-
gory. A positive coefficient results in an odds ratio
ranging between 1 and infinity, whereas a negative
coefficient ranges between 1 and zero (but never
reaching zero). For example, an odds ratio of 2.0 for a
categorical variable (e.g., gender) indicates that the
probability of practising alternative therapies is twice
as large for the specified category of that variable
(e.g., females) compared to the reference category
(e.g., males), while statistically controlling for all other
explanatory variables (DeMaris, 1995). The model chi-
square value is the improvement that the proposed
model makes over the model of independence, that is,
a larger chi-square indicates a greater improvement in
fit to the data (Morgan & Teachman, 1988). In addi-
tion, we examine all first order interaction effects
between the four enabling factors – number of confi-
dants, mutual aid, illness self-efficacy and reading on
the illness.
effects, results from the two analyses are presented
separately. Six variables in the main effects model are
found to be statistically significant. An inverse associ-
ation between alternative therapy use and age is sup-
ported. The odds of using alternative therapies to
manage a chronic condition are .97 lower for each
increment or year in age, controlling for all other vari-
ables in the equation. The other three predisposing
variables (gender, education and marital status) were
not found to be statistically significant. Three of the
six need factors, illness type, co-orbidity, and number
of prescription medicines, demonstrated statistically
significant associations with alternative therapy use.
The odds of using alternative therapies is almost half
as low (odds ratio = .45) for the hypertension group as
for the arthritis group. No association is supported for
the heart problem–arthritis comparison. However, the
likelihood of use is increased by a factor of 1.23 times
higher for each additional co-morbid condition. In
addition, use of alternative therapies is inversely asso-
ciated with prescription drug use. The odds of use are
.84 lower for each prescriptive medication taken. Per-
ceived seriousness, duration of condition, and
number of doctor visits result in relationships that are
not statistically significant. Two of the five enabling
factors exhibit statistically significant associations
with alternative therapy use. The odds of use are .60
and .32 lower for the $20,000–$49,999 and the
$100,000+ income groups respectively compared to
the less than $20,000 group. Furthermore, the odds of
alternative therapy use are almost twice (odds ratio =
1.89) as high for persons who read about their chronic
illness after being diagnosed compared to those who
do not.
Results
Logistic Regression for Alternative Therapy Use
The logistic regression results for alternative therapy
use are shown in Table 2. A main effects model and an
interaction model have been estimated, both of which
result in statistically significant Model Chi Square sta-
tistics (71.94, p < .001 and 80.23, p < .001, respectively).
Since the inclusion of the interaction terms can cause
instability in the regression coefficients for the main
Turning to the interaction model, only one of the eight
interaction terms was found to be associated with use
of alternative therapies. Persons who read about their
condition and who report higher self-efficacy are
more likely to use alternative therapies (odds ratio =
1.07).

Alternative Therapies
La Revue canadienne du vieillissement 21 (1)
55
Table 2: Logistic regression for alternative therapies
Full Mode
B
S.E.
Exp (B)
Predisposing Variables
Gender (ref = Male)
Female
.37
.23
.01
NS
.97
Age
–.03**
Education (ref = Some Sec. or Less)
Completed Secondary
Post-Secondary
–.08
–.15
.33
.29
NS
NS
Marital Status (ref = Not Married)
Married/Common Law
–.22
.21
NS
Need Variables
Illness Type (ref = Arthritis)
Heart Disease
Hypertension
–.55
–.81**
.29
.30
NS
.45
Comorbidity
.21*
.09
1.23
Perceived Seriousness (ref = Not At All)
Slightly Serious
Moderately Serious
.01
.31
.26
.30
.28
.43
NS
NS
NS
Extremely Serious
Duration of Condition
–.003
.06
–.10
.03
NS
NS
.84
Number of Doctor Visits
Number of Prescription Medications
–.17**
.06
Enabling Variables
Income (ref = < $20,000)
$20,000–49,999
$50,000–69,999
$70,000–99,999
$100,000 +
–.51*
–.10
.33
.25
.29
.37
.57
.60
NS
NS
.32
–1.12*
Number of Confidants
.03
.02
NS
Mutual Aid (ref = Never)
Daily–Every 2 Weeks
Once Per Month/Few Times Year
–.52
–.20
.38
.22
NS
NS
Illness Self-Efficacy
Reading on Subject
Constant
.009
.64**
–.40
.01
.24
.99
NS
1.89
Model Chi Square
71.94***
Interaction Model
Interaction Terms
Reading x Number of Confidants
Reading x Self-Efficacy
–.02
.07**
.65
.05
.03
NS
1.07
NS
NS
NS
Reading x High Mutual Aid
1.23
.59
Reading x Moderate Mutual Aid
High Mutual Aid x Number of Confidants
.38
–.68
.09
Table 2: Logistic regression for alternative therapies continued on the following page

56 Canadian Journal on Aging 21 (1)
Andrew. V. Wister et. al.
Table 2: Logistic regression for alternative therapies continued
Moderate Mutual Aid x Number of Confidants
High Mutual Aid x Self-Efficacy
Moderate Mutual Aid x Self-Efficacy
Constant
.03
–.52
–.02
.79
.05
.05
NS
NS
NS
.03
1.14
Model Chi Square
80.23***
*p <.05; **p<.01; ***p<.001
B = Parameter Estimate; S.E. = Standard Error; Exp (B) = Odds Ratio
Sec = Secondary; ref = Reference Category
Note: Parameter estimates are not shown for the main effects included in the Interaction Model because of instability in some of
the coefficients.
Logistic Regression for Meditation/Prayer
hood of engaging in meditation or prayer to manage a
chronic condition is higher (odds ratio = 1.77) for
females than for males. The probability of this type of
coping behaviour is lower for each year of age (odds
ratio = .98) and is lower for married (odds ratio = .64)
compared to non–married older adults. No support
was found for an association between education and
alternative therapy use.
Both the full main effects model and the interaction
model also result in statistically significant Model Chi
Square statistics (69.18, p < .001, and 79.50, p < .001,
respectively) for the meditation/prayer analysis. As
shown in table 3, a total of six variables predict this
type of illness management strategy. Three of the four
predisposing factors, age, gender and marital status,
exhibit statistically significant associations. The likeli-
Table 3: Logistic regression for meditation/prayer
Full Model
B
S.E.
Exp (B)
Predisposing Variables
Gender (ref = Male)
Female
.57**
.18
.01
1.77
.98
Age
–.02*
Education (ref = Some Sec. or Less)
Completed Secondary
–.22
–.18
.26
.23
NS
NS
Post–Secondary
Marital Status (ref = Not Married)
Married/Common Law
–.45**
.17
.64
Need Variables
Illness Type (ref = Arthritis)
Heart Disease
.34
–.09
.22
.23
NS
NS
Hypertension
Comorbidity
.07
.07
NS
Perceived Seriousness (ref = Not At All)
Slightly Serious
–.01
.43*
.71*
.24
.22
.33
NS
1.53
2.04
Moderately Serious
Extremely Serious
Duration of Condition
.02**
.005
.01
.03
.04
1.02
NS
Number of Doctor Visits
Number of Prescription Medications
–.006
NS
Table 3: Logistic regression for meditation/prayer continued on the following page

Alternative Therapies
La Revue canadienne du vieillissement 21 (1)
57
Table 3: Logistic regression for meditation/prayer continued
Enabling Variables
Income (ref = < $20,000)
$20,000–49,999
$50,000–69,999
$70,000–99,999
$100,000 +
–.11
.25
–.36
–.51
.20
.24
.34
.39
NS
NS
NS
NS
Number of Confidants
.02
.02
NS
Mutual Aid (ref = Never)
Daily–Every 2 Weeks
–.28
–.20
.31
.17
NS
NS
Once Per Month/Few Times Year
Illness Self–Efficacy
Reading on Subject
Constant
–.002
.29
.01
.18
.79
NS
NS
–.20
Model Chi Square
69.18***
Interaction Model
Interaction Terms
Reading x Number of Confidants
Reading x Self-Efficacy
–.01
–.002
.29
.05
.02
.97
.39
.07
.04
.04
.02
.92
NS
NS
NS
NS
NS
1.10
NS
NS
Reading x High Mutual Aid
Reading x Moderate Mutual Aid
High Mutual Aid x Number of Confidants
Moderate Mutual Aid x Number of Confidants
High Mutual Aid x Self-Efficacy
Moderate Mutual Aid x Self-Efficacy
Constant
–.53
.09
.09*
–.07
–.01
.29
Model Chi Square
79.50***
*p <.05; **p<.01; ***p<.001
B = Parameter Estimate; S.E. = Standard Error; Exp (B) = Odds Ratio
Sec = Secondary; ref = Reference Category
Note: Parameter estimates are not shown for the main effects included in the Interaction Model because of instability in some of
the coefficients.
Among the six need factors, perceived seriousness
and duration of condition were found to be impor-
tant. The likelihood of meditating or praying is higher
for persons perceiving their condition to be moder-
ately serious (odds ratio = 1.53) and extremely serious
(odds ratio = 2.04) compared to not at all serious. In
addition, the odds of engaging in this style of coping
behaviour increased (odds ratio = 1.02) for each year
of illness duration. However, relationships with ill-
ness type, co-morbidity, number of doctor visits and
number of prescription medications were not sup-
ported.
persons with combinations of moderate levels of
mutual aid and a higher number of confidants are
more likely to mediate or pray to cope with their con-
dition.
Discussion
Alternative therapy use has received increasing atten-
tion as one primary form of self-care engaged by per-
sons with a chronic illness, which is typically only
part of a comprehensive and dynamic illness manage-
ment system (Clark et al., 1991; Clark et al., 2000;
Wister & Dean, 1998). This phenomenon is particu-
larly relevant in the face of accelerated population
aging projected for the first several decades of the
new millennium. Our substantive focus entailed
Surprisingly, none of the five enabling factors showed
statistically significant associations with illness man-
agement by means of meditation/prayer. However,

58 Canadian Journal on Aging 21 (1)
Andrew. V. Wister et. al.
chronic illness management among older adults aged
50 and over because of their increased risk of illness
onset. Drawing from the Andersen–Newman Model,
this research investigated sets of predisposing, need
and enabling factors affecting two measures of alter-
native therapy use – a broad measure of use and one
that focuses on prayer and meditation. The principal
research questions guiding this research are whether
and to what degree the propensity to use alternative
therapies as an illness response is affected by predis-
posing factors, dimensions of need embodied in the
illness context, involvement with the formal medical
system, and facilitation through enabling compo-
nents. And further, the question as to whether the use
was affected by an expanded set of enabling factors
deemed to be important was addressed. These factors
included: supportive ties, self-efficacy, and reading
information sources, as well as their interactions.
those reporting annual household incomes of $20,000–
$49,999 and over $100,000 compared to those of low
income (under $20,000). Only the income result is con-
trary to the hypothesized association.
With regard to prayer/meditation, predisposing and
need factors appear to be most important in affecting
the probability of use. Being female, perceiving one’s
condition as more serious and longer duration of the
chronic condition result in a greater tendency to try
these methods of coping. Lower rates are observed for
persons of higher age, and among persons who are
married/common law compared to those who are
unattached. The only enabling effect was found for
the interaction of moderate levels of mutual aid cou-
pled with higher numbers of confidants.
These results indicate that age is an important predis-
posing factor affecting the tendency to engage in alter-
native therapies, and moreover, that rates of use
decrease among older persons. However, it is not
clear whether it is an age, period or cohort effect that
results in people of older ages exhibiting lower rates
of use. It is likely that the alternative therapy move-
ment has targeted younger and middle-aged persons,
and that the medical orientation of older persons is
more traditional, relying more on opinions of doctors
than information or other sources. In addition, unat-
tached women in the pre-retirement ages appear to be
the most likely candidates to manage their chronic ill-
ness using meditation techniques or prayer. Overall,
however, factors that predispose or inhibit individu-
als’ propensities to try alternative forms of illness
management appear only to play a modest role. Fur-
thermore, the absence of a relationship with education
suggests that we should investigate other factors to
explain why some people are trying and adopting
alternative therapies.
Approximately 15 per cent of the older adult subjects
(n = 879) in the North Shore Self-Care Study (NSSCS)
used alternative therapies specifically to manage their
chronic condition, such as herbal remedies, massage,
acupuncture, and other unconventional medicine,
and almost 29 per cent used meditation or prayer as
an illness management strategy. These rates are simi-
lar or a little lower than those in other studies. For
example, the 1993/94 National Population Health
Survey shows that about 15 per cent of adult Canadi-
ans reported that they had consulted an alternative
health practitioner in the year prior to the survey (Mil-
lar, 1997). And in the United States, Eisenberg et al.
(1993) found that 34 per cent of adults had reported
using alternative therapies in the previous year. The
patterns of use identified in our study reinforce that a
substantively important proportion of individuals
adopt unconventional medical practices in their ill-
ness management. For comparative purposes, it was
also found in our study that rates of traditional life-
style self-care practices specifically used to manage
their chronic illness, such as exercise, diet and weight
change, ranged between 50 and 70 per cent, whereas
self-help group participation was approximately 10
per cent.
Examination of a comprehensive set of need factors
capturing the illness context, (including the type of
chronic illness, co-morbidity, perceived seriousness,
duration of the condition, doctor visits and prescrip-
tion medication use), supports research that con-
cludes that the objective and subjective health status
of individuals is a primary factor affecting use (Gray,
1985; Sutherland & Verhoef, 1994). The presentation of
certain illnesses appears to also affect the propensity
to try various alternative therapies, for example, per-
sons with hypertension are half as likely to try them
as those with arthritis or rheumatism. Perhaps this is
partly because some individuals do not perceive
hypertension as a chronic illness and they have fewer
symptoms. Moreover, co-morbidity magnifies the
probability of resorting to alternative practices by a
factor of 1.23 for each additional illness. Indeed, there
appears to be a cumulative impact of illness that fuels
Multivariate findings based on the logistic regression
analyses establish that the use of alternative therapies
is affected by measures of predisposing, need and
enabling factors. The probability of use of unconven-
tional medicine was shown to be higher among per-
sons who report co-morbidity, adults who read on the
subject of their illness, and is altered by the interaction
of reading and higher levels of illness self-efficacy. A
lower likelihood of use was found among persons of
higher age, those persons reporting hypertension as
their major chronic illness (compared to arthritis/
rheumatism), those taking more medications, and

Alternative Therapies
La Revue canadienne du vieillissement 21 (1)
59
the propensity for older adults to turn to unconven-
tional medicine. Whereas, the length and perceived
seriousness of the condition appear to be more impor-
tant illness attributes for determining patterns of
meditation and prayer. This may signify that the
mind–body experience of meditation and the spiritual
experience of prayer become important coping mech-
anisms for persons perceiving their chronic condition
as life threatening and among those whose illness has
continued for a longer period of time.
of various alternative therapies provided by health
professionals (e.g., massage therapy, chiropractic, acu-
puncture, naturopathic medicine, etc.). Legislation of
user fees and limited coverage under provincial insur-
ance programs in Canada may inhibit use of many
therapies, for example, access to massage therapy
requires a referral and a user fee for a limited number
of annual visits under the health plans in several
provinces. Thus, insurance programs can influence
the development and sustainability of alternative
health professions. In this regard, the medical profes-
sion has the strongest lobbying power for influencing
health legislation and for acceptance of new health
professions (Schepers & Hermans, 1999). Wardell
(1994) has developed a hierarchical classification sys-
tem of alternative health practitioners – quasi, mar-
ginal, limited and ancillary types – which has been
used to document the evolution of specific alternative
health professions, typically in relation to the medical
profession and the organization of training and licens-
ing standards. As Schepers & Hermans (1999) con-
tend, health care systems may well become more
pluralistic, even if medical dominance is preserved.
Although characteristics of the illness are particularly
important in our understanding of how and why
older people manage their chronic illness through the
use of unconventional medicine, use of the formal
health system also appears to play a role. It is interest-
ing that individuals who take more prescription med-
ications actually are less likely to use alternative
therapies, as well as the converse. Yet, number of doc-
tor visits was not found to be associated with either
type of unconventional practices. On the one hand,
the inverse association uncovered for prescriptions
suggests that older persons with chronic conditions
who engage in alternative therapies substitute conven-
tional medicine with unconventional medicine to
some degree. This lends support to a growing number
of studies indicating that alternative therapy use may
be a response to unsuccessful allopathic treatments of
disease states obtained through the formal system
(Moore et al., 1985; Vincent & Furnham, 1996). On the
other hand, the absence of a relationship between
number of doctor visits and use of alternative thera-
pies suggests that the conventional and unconven-
tional medical systems are not completely opposed to
one another. This finding is consistent with several
studies showing that persons resorting to alternative
therapies do not stop seeing their doctors and, for
some individuals, are referred to alternative practi-
tioners through their general practitioner or specialist
(see e.g., Marshall et al., 1990; Eisenberg et al., 1993;
Northcott & Bachynsky, 1993).
Furthermore, there is an issue of control over prescrib-
ing, developing, and dispensing of herbal remedies,
which are fast becoming multi-million dollar indus-
tries in most developed countries. For example, the
most commonly prescribed medicine in Germany is
ginkgo, a herbal remedy to which a variety of healing
properties are attributed. Physicians, pharmacists and
pharmaceutical companies have entered this new
arena along with representatives from alternative
medicine (especially naturopathic medicine), and pri-
vate industry. The integration of the conventional and
unconventional systems, therefore, is integrally tied to
the economics of health care, in addition to the evolu-
tion and acceptance of alternative health professions.
With respect to the findings for the enabling compo-
nent, we anticipated that the expanded set of facilita-
tors examined in this research would also shape
alternative therapy use among older adults managing
a chronic illness. Indeed, one of the strongest findings
was that reading about one’s chronic illness (at the
time of diagnosis) doubles the likelihood of using alter-
native therapies to manage that illness. Information
about chronic illness has become increasingly availa-
ble through the various illness societies (e.g., Heart &
Stroke Foundation, Arthritis Society, etc.), books,
other print media, and via the Internet. Regardless of
the source, information about chronic illness typically
includes the use of alternative therapies, albeit the
degree of acceptance of these practices is highly varia-
ble. And while many information sources recommend
that individuals consult their doctor, a significant
number of persons apparently are inhibited to discuss
Indeed, the degree of complementarity of the informal
and formal health systems remains controversial. Not-
ing the salience of the emerging biopsychosocial para-
digm, Longino (1997) calls for an integrated health
delivery system capable of synthesizing the public
health sphere and curative medicine. However, many
other researchers maintain, as does Lewith (1997, p.
48) that, “there are few, if any, formal pathways
through which to implement and integrate the rele-
vant aspects of complementary medicine into conven-
tional medicine”.
In the Canadian and American health care systems, it
would appear that several interrelated issues loom on
the horizon. One of these is health insurance coverage

60 Canadian Journal on Aging 21 (1)
Andrew. V. Wister et. al.
alternative therapies with their family physician
because they feel that he/she lacks a full understand-
ing of their condition and may not condone uncon-
ventional medicine (Murray & Rubel, 1992). Yet, the
mere discussion of alternative therapies as a means to
manage chronic illness may encourage further investi-
gation by individuals into their efficacy, cost, availa-
bility, and access. Interestingly, we also found that
alternative therapy use is influenced by the combina-
tion of having a high level of illness self-efficacy (per-
ceived confidence that one can control aspects of one’s
illness) and reading about one’s illness. Thus, a
greater feeling of confidence to control chronic illness
coupled with reading information about one’s illness
facilitates the use of alternative therapies. This seems
to be indicative of persons striving for greater auton-
omy and personal involvement in their illness man-
agement.
context is critically important in affecting the propen-
sity of individuals to use alternative therapies. So too
are certain predisposing factors (particularly age), use
of prescription medicines for illness maintenance, and
the facilitating forces of health information, income,
and key interactions tapping into self–efficacy and
reading, as well as those pertaining to mutual aid and
social support. These patterns are occurring against
the backdrop of social change, including shifts in the
normative and preference structures underlying
health behaviour; increasing information, including
website developments; availability and access of alter-
native therapies; the evolution of alternative health
professions; and reformulation of the health care sys-
tem.
It is apparent from this study that further research is
required to address the gaps in our understanding of
how and why people engage in unconventional medi-
cine as part of their illness management system. In
particular, it would be useful to gather data that con-
tain further details about the specific type of alterna-
tive therapy use, frequency, duration, and perceived
impact on different populations of healthy and ill
individuals. Furthermore, the complementary and
integrative aspects of conventional and unconven-
tional medicine remain an issue of major importance.
Indeed, further establishing the efficacy and non-effi-
cacy of various alternative approaches to illness man-
agement would probably be the most effective
method to erode the barriers to the integration of
“complementary” medicine.
Income was also found to influence alternative ther-
apy use. Persons accessing total household incomes in
the $20,000–$49,999 and the upper ($100,000+) groups
compared to the lowest (under $20,000) income group
are less likely to use alternative therapies. This is in
contrast to the positive association found in an earlier
Canadian study (CMAJ, 1997). However, our study
focuses only on persons aged 50 and over who have a
diagnosed chronic illness, whereas the previous
research includes all adults in the population. It
would appear, however, that further research is
needed on how income affects the process of trying
alternative therapies for various groups of healthy
and ill groups.
Notes
Turning to meditation and prayer, only one enabling
relationship was found for engaging in these mind-
body techniques to manage chronic illness – the
simultaneous effect of moderate levels of mutual aid
(regular contact with someone with the same illness)
and a higher number of confidants. It appears that
social support, combined with having someone with
whom to speak who has the same illness, enables
mind-body approaches to illness management. While
we recognize that meditation and prayer may be two
relatively distinct coping behaviours, there appears to
be something important about the social milieu that
affects how people deal with their chronic illness on a
psychosomatic level.
1. The North Shore Self-Care Study data were collected as
part of a grant (#6610-2122-602) funded by the National
Health Research and Development Program, Health
Canada (Co-PI’s in alphabetical order are: Larry Green,
Gloria Gutman, Patrick McGowan & Andrew Wister).
References
Andersen, R. (1995). Revisiting the behavioral model and
access to medical care: Does it matter? Journal of Health
and Social Behavior, 36(March),1–10.
Andersen, R., & Newman, J. (1973). Societal and individual
determinants of medical care use in the United States.
Milbank Memorial Fund Quarterly, 51, 91–124.
Astin, J.A. (1996). Why patients use alternative medicine:
Results of a national study. Journal of the American Medi-
cal Association, 279(19), 1548–1553.
In conclusion, there is no doubt that people adopt a
multitude of strategies to manage their chronic illness
and that these are continually changing because of the
dynamics of the illness or co-morbid conditions as
people age. The Andersen–Newman Model remains
useful in providing a conceptual frame for positioning
major dimensions of health utilization and self–care
behaviour. More specifically, we found that the illness
Bandura, A. (1986). Social foundations of thought and action.
NJ: Prentice Hall.
Boisset, M., & Fitzcharles, M-A. (1994). Alternative medicine
use by rheumatology patients in a universal health care
setting. Journal of Rheumatology, 21, 148–152.
Chappell, N., & Blandford, A. (1987). Health service utiliza-
tion by elderly persons. Canadian Journal of Sociology, 12,
195–215.

Alternative Therapies
La Revue canadienne du vieillissement 21 (1)
61
Clark, N., Becker, M., Janz, N., Lorig, K., Rakowski, W., &
Anderson, L. (1991). Self-management of chronic dis-
ease by older adults: A review and questions for
research. Journal of Aging and Health, 3(1), 3–27.
Lorig, K., & Holman, H.R. (1989). Long-term outcomes of
the Arthritis Self-Management Study: Effects of rein-
forcement efforts. Social Science and Medicine, 29, 221–
224.
Clark, N., Janz, N., Dodge, J., Schork, M., Fingerlin, T.,
Wheeler, J., Liang, J., Keteyian, S., & Santinga, J. (2000).
Changes in functional health status of older women
with heart disease: Evaluation of a program based on
self-regulation. Journal of Gerontology: Social Sciences,
55B(2), S117–S126.
CMAJ. (1991). One in five Canadians is using alternative
therapies, survey finds. Canadian Medical Association
Journal, 144(4), 469.
McDonald-Miszczak., L., Wister, A.V., & Gutman, G.M.
(2001). Self-care among older adults: An analysis of the
objective and subjective illness context. Journal of Aging
& Health, 13 (1), 120–145.
MacLennan, A.H., Wilson, D.H., & Taylor, A.W. (1996). Prev-
alence and cost of alternative medicine in Australia. The
Lancet, 347, 569–573.
Marshall, R., Gee, R., Israel, D., Edwards, F., Dumble, J.,
Wong, S., Chan, C., Patel, R., Poon, P. & Tam, G. (1990).
The use of alternative therapies by Auckland general
practitioners. New Zealand Medical Journal, 103, 213–215.
CMAJ. (1997). Higher earners seek more alternative care.
Canadian Medical Association Journal, 157(7), 996.
Cohen, S., & Syme, L. (1985). Social support and health. New
Menard, S. (1995). Applied Logistic Regression Analysis. Quan-
titative Applications in the Social Sciences. London: Sage.
York: Academic Press.
Dean, K., & Kickbusch, I. (1995). Health related behaviour in
health promotion: utilizing the concept of self-care.
Health Promotion International, 10(1), 35–40.
Millar, W.J. (1997). Use of alternative health care practitio-
ners by Canadians. Canadian Journal of Public Health,
88(3), 154–158.
DeMaris, A. (1995). A tutorial in logistic regression. Journal
of Marriage and the Family, 57, 956–968.
Moore, J., Phipps, K., & Marcer, D. (1985). Why do people
seek treatment by alternative medicine? British Medical
Journal, 290, 28–29.
Montbriand, M., & Laing, G. (1991). Alternative health care
as a control strategy. Advanced Journal of Nursing, 16,
325–332.
Morgan, S.P., & Teachman, J.D. (1988). Logistic regression:
Description, examples, and comparisons. Journal of
Marriage and the Family, 50; 929–936.
Eisenberg, D.M., Kessler, R.C., Foster, C., Norlock, F.E.,
Calkins, D.R., & Delbanco,T.L. (1993). Unconventional
medicine in the United States: Prevalence, costs, and
patterns of use. New England Journal of Medicine, 328(4),
246–252.
Fisher, P., & Ward, A. (1994). Complementary medicine in
Europe. British Medical Journal, 309, 107–111.
Fleming, G., Giachello, A., Anderson, A., & Andrade, P.
(1984). Substitute, supplement or stimulus for formal
medical care services? Medical Care, 22, 950–966.
Murray, J., & Shepherd, S. (1993). Alternative or additional
medicine? An exploratory study in general practice.
Social Science and Medicine, 37(8), 983–988.
Furnham, A., & Forey, J. (1994). The attitudes, behaviours
and beliefs of patients of conventional vs. complemen-
tary (alternative) medicine. Journal of Clinical Psychology,
50(3), 458–469.
Murry, R.H., & Rubel, A.J. (1992). Sounding board: Physi-
cians and healers – unwitting patterns in health care.
The New England Journal of Medicine, 326(1), 61–64.
Northcott, H.C., & Bachynsky, J. (1993). Research Note: Con-
current utilization of chiropractic, prescription medi-
cines, nonprescription medicines and alternative health
care. Social Science and Medicine, 37(3), 431–435.
Gray, D. (1985). The treatment strategies of arthritis suffer-
ers. Social Science and Medicine, 12, 507-515
Kart, C.S., & Engler, C.A. (1994). Predisposition to self-
health care: Who does what for themselves and why?
Journal of Gerontology: Social Sciences, 49(6), S301–S308.
OAM–NIH Classification of Alternative Medicine Practices.
(1999). Office of Alternative Medicine. Webpage.
Kelner & Wellman (1997). Health care and consumer choice:
Medical and alternative therapies. Social Science and
Medicine, 45(2), 201–212.
Ory, M., & Defriese, G. (Eds.) (1998). Self-Care in later life:
Research, program & policy Perspectives. New York: Sage.
Richmond, C. (1991). As their health improves, British
broaden quest for alternative health care. Canadian Med-
ical Association Journal, 144(7), 912–914.
Kestin, M., Miller, L., Littlejohn, G., & Wahlqvist, M. (1985).
Medical Journal of Australia, 143, 516–519.
Kronenfeld, J.J., & Wasner, C. (1982). The use of unorthodox
therapies and marginal practitioners. Social Science and
Medicine, 16, 1119–1125.
Lewith, G. (1997). A measure of success. British Journal of
General Practice, 47, 47–49.
Romeder, J.M. (1990). The self-help way: Mutual aid and health.
Ottawa: Canada Council on Social Development.
Schepers, R.M., & Hermans, H.E. (1999). The medical pro-
fession and alternative medicine in the Netherlands: Its
history and recent developments. Social Science & Medi-
cine, 48, 343–351.
Longino (1997). Beyond the body: An emerging medical
paradigm. American Demographics, Dec., 14–19.
Sutherland, L., & Verhoef, M. (1994). Why do patients seek a
second opinion or seek alternative medicine? Journal of
Clinical Gastroenterology, 19(3), 194–197.
Verhoef, M., Sutherland, L., & Brkich, L. (1990). Use of alter-
native medicine by patients attending a gastoenterol-
Lorig, K., Brown, B.W., Ung, E., Chastain, R. Shoor, S., &
Holman, H.R. (1989). Development and evaluation of a
scale to measure the perceived self-efficacy of people
with arthritis. Arthritis and Rheumatism, 32, 37–44.

62 Canadian Journal on Aging 21 (1)
Andrew. V. Wister et. al.
ogy clinic. Canadian Medical Assocaition Journal, 142(2),
121–125.
Vincent, C., & Furnham, A. (1996). Why do patients turn to
complementary medicine? British Journal of Clinical Psy-
chology, 35, 37–48.
Appendix
Lorig Arthritis Self–efficacy Scale
I am going to ask you how confident you are about
your ability to control different aspects of your condi-
tion. On a scale from 1 to 5, where 1 is not at all confi-
dent, and 5 is totally confident, how confident are you
that you can…
Wardell, W.I. (1994). Alternative medicine in the United
States. Social Science & Medicine, 38(8), 1061–1068.
Wister, A.V. (1992). Residential attitudes and knowledge,
use, and future use of home support agencies. Journal of
Applied Gerontology, 11(1), 84–100.
Wister, A.V. (1995) The relationship between self–help group
participation and other health behaviours among older
adults. Canadian Journal of Community Mental Health,
14(2), 23–38.
Wister, A.V., & Dean, K. (1998) A comparative analysis of
graphical interaction and logistic regression modelling:
Self–care and coping with a chronic illness in later life.
Paper presented at the 14th World Congress of Sociol-
ogy, Montreal, Quebec, July 26–Aug. 1, 1998.
1.
2.
Control your fatigue?
Regulate your activities so as to be active without
aggravating your arthritis?
3.
4.
5.
Do something to help yourself feel better if you are
feeling blue?
Manage arthritis pain during your daily activities
(compared to other people with arthritis like yours)?
Manage your arthritis symptoms so that you can do
the things you enjoy doing?
Wolinsky, F.D. (1978). Assessing the effects of predisposing,
enabling and illness-morbidity characteristics on health
service use. Journal of Health and Social Behaviour, 19,
384–396.
Wolinsky, F.D., & Johnson, R.J. (1991). The use of health ser-
vices by older adults. Journal of Gerontology: Social Sci-
ences, 46(6), S345–S357.
Zimmer, Z., & Chappell, N. (1994). Mobility restriction and
the use of devices among seniors. Journal of Aging and
Health, 6(2), 185–208.
6.
7.
8.
9.
Deal with the frustration of arthritis?
Decrease your pain quite a bit?
Continue most of your daily activities?
Keep arthritis pain from interfering with your sleep?
10. Make a small-to-moderate reduction in your arthritis
pain by using methods other than taking extra medica-
tion?
11. Make a large reduction in your arthritis pain by using
methods other than taking extra medications?