Osteoporosis International

, Volume 18, Issue 11, pp 1547–1556

Complementary and alternative medicine use by osteoporosis clinic patients

Authors

  • C. A. K. Y. Chong
    • Division of General Internal Medicine and Clinical EpidemiologyUniversity Health Network and Mount Sinai Hospital, University of Toronto
  • N. Diaz-Granados
    • Division of General Internal Medicine and Clinical EpidemiologyUniversity Health Network and Mount Sinai Hospital, University of Toronto
  • G. A. Hawker
    • Division of Rheumatology, Department of MedicineWomen’s College Hospital, University of Toronto
  • S. Jamal
    • Division of Endocrinology, Department of MedicineSt. Michael’s Hospital, University of Toronto
  • R. G. Josse
    • Division of Endocrinology, Department of MedicineSt. Michael’s Hospital, University of Toronto
    • Division of General Internal Medicine and Clinical EpidemiologyUniversity Health Network and Mount Sinai Hospital, University of Toronto
    • Division of Endocrinology, Department of MedicineUniversity Health Network and Mount Sinai Hospital, University of Toronto
Original Article

DOI: 10.1007/s00198-007-0417-x

Cite this article as:
Chong, C.A.K.Y., Diaz-Granados, N., Hawker, G.A. et al. Osteoporos Int (2007) 18: 1547. doi:10.1007/s00198-007-0417-x

Abstract

Summary

We describe complementary and alternative medicine use (CAM) in 360 patients attending osteoporosis clinics. Of these patients, 57% were CAM users. Predictors of CAM use included lower mental quality of life, younger age and higher education. Less than half of CAM use was disclosed to physicians, despite potential adverse interactions.

Introduction

The prevalence of complementary and alternative medicine (CAM) use in osteoporosis clinics is not known. The objective of this study was to describe the pattern of CAM use in this population.

Methods

We performed a cross-sectional study of 360 patients attending academic osteoporosis clinics in Toronto, Canada in 2001. Subjects completed a self-administered questionnaire on CAM use. Health-related quality of life (HQL) was measured with the Short-Form 36v2. Comparative statistics and logistic regression were performed to identify sociodemographic, HQL and clinical correlates of CAM use.

Results

More than 80% of participants were women, Caucasian and had at least a high school education. Of subjects, 57% used CAM in the previous year. Only 44% of CAM use was disclosed to a medical doctor. CAM users and non-users did not differ in clinical characteristics such as bone mineral density, level of comorbidity and fracture history. In univariate analysis, CAM users were less satisfied with conventional medicine. However, when we explored patient satisfaction, comorbidities and sociodemographic as predictors for CAM use, the multivariable analyses showed that lower mental HQL, younger age, and post-secondary education were the only significant predictors. We identified 35 cases in which the utilization of CAM supplements could possibly exacerbate existing medical conditions.

Conclusion

Patients attending osteoporosis clinics frequently use CAM. Conceptually, the predictors of use identified in this study may fit into a socio-behavioral model that helps explain why people turn to CAM. Physicians may need to elicit a history of CAM use more vigilantly so as to better screen for possible adverse clinical interactions.

Keywords

Alternative medicineBone diseases metabolicDrug interactionsQuality of life

Introduction

Complementary and alternative medicine (CAM) use has become mainstream in North America. Over a decade ago, in the 1995 National Population Health Survey, 15% of Canadians reported having seen an alternative health care practitioner [1]. From 1990 to 1997, the number of Americans using CAM rose from 33.8% to 42.1% [2], a prevalence which has since stabilized in 2002 [3]. As more individuals turn to unconventional therapies, there is a growing need to learn how and why people use CAM.

Studies exploring how specific patient groups use CAM highlight the importance of understanding alternative medicine. The percentage of patients using CAM is high, ranging anywhere from 24% in those with brain tumors in 1999 [4] to 91% in those with fibromyalgia in 1996 [5]. Many factors prompt CAM usage - recently, a model has been proposed that suggests physical symptoms, chronic illness, loss of empowerment, and psychological stressors all play a role [6]. Empirically, studies have found that factors predicting CAM use include female sex, higher income, higher education level and younger age [7, 8]. There have been hypotheses that CAM users are less satisfied with conventional medicine, but this has not been a consistent finding [2, 7, 9]. Overall, CAM users tend to fall into a group of “cultural creatives” who can be identified by their commitment to the environment, feminism and spiritual and personal growth [10].

Research in CAM helps clinicians better appreciate their patients’ health care practices and provides direction for how to improve care. CAM use by osteoporosis and osteopenia patients, however, has yet to be described. Among patients who are at risk for or have osteoporosis, understanding CAM use may be of special importance given the high representation of women and the wide variety of CAM products marketed for promoting post-menopausal health [11, 12]. People attending osteoporosis clinics might span the spectrum from a healthy post-menopausal woman who is concerned about her bone health and actively looking for a variety of modalities to prevent osteoporosis, to an ill elderly man on multiple medications referred for prolonged glucocorticoid use. In all cases, it would be important to know CAM therapies patients may be employing.

Thus, the primary purpose of this study was to describe the pattern of CAM use in osteoporosis clinic patients (OPCP). Secondarily, we also sought to identify correlates of CAM use and potential undesired clinical interactions of CAM in this population.

Methods

Study design

This was a cross-sectional study with a primary objective of measuring the prevalence of CAM use in OPCPs. Using an estimated prevalence of 35% [2, 7] with a desired precision of 5%, we calculated a required sample size of 350 [13]. Our secondary objective included determining which sociodemographic, clinical, patient satisfaction and quality of life factors were associated with CAM use. We also explored potential adverse clinical interactions of CAM with co-morbidities in this population.

Subject recruitment

Male and female patients of all ages from three osteoporosis clinics were invited to complete a self-administered questionnaire from January 15 to August 16, 2001. The clinics were located at academic tertiary-care hospitals in downtown Toronto, Canada. With a population of about 2.5 million, Toronto is Canada’s largest city covering 650 km2 in southern Ontario [14]. The only criteria excluding patients from eligibility were the inability to read English and known cognitive impairment. Ethics approval was obtained from the ethics review boards of the participating hospitals.

At the University Health Network (UHN), of 278 consecutive outpatients asked, 27 were ineligible and 44 declined to participate, for a participation rate of 82% among eligible subjects. At St. Michael’s Hospital (SMH), we only approached consecutive follow-up patients as new patients were already being asked to participate in a separate study. Of 142 patients approached, 16 were ineligible and 30 declined to participate, for a participation rate of 76% among eligible subjects. The third site was at the Women’s College Hospital (WCH). Here, patients were sent the study questionnaire by mail, in conjunction with a medical information form all WCH patients are required to complete and bring to their upcoming appointments. Of 92 patients who were mailed the study questionnaire and attended a visit at the WCH clinic, 57 returned a completed questionnaire for a participation rate of 62%. The overall participation rate from all three sites was 77% for a total sample size of 360 subjects.

Questionnaire design

The questionnaire given to consenting participants consisted of the following items:
  1. 1)

    Prevalence of general CAM use. The NIH National Center for Complementary and Alternative Medicine (NCCAM) has defined CAM as “a group of diverse medical and health care systems, practices, and products that are not presently considered to be part of conventional medicine.” The NCCAM further defines conventional medicine as “…medicine as practiced by holders of MD…or DO…degrees and by their allied health professionals…” and divides CAM into five domains: alternative medical systems, mind-body interventions, biologically based therapies, manipulative and body-based methods and energy therapies [15]. Shaping these categories into a practical definition for research purposes can be difficult, and not surprisingly, there are no widely-accepted criteria for what constitutes CAM. As this project is, to the best of our knowledge, the first to explore CAM use in OPCPs, we wanted to use a definition that (a) had been previously validated in other studies; (b) was sufficiently general to be useful for most medical conditions; and (c) was broad enough to encompass many different potential treatments, i.e., a definition with a high sensitivity for detecting CAM use. Therefore, we asked patients if, in the past year, they had used any of 16 general CAM therapies, including relaxation techniques, herbal medicine, massage therapy, chiropractic services, spiritual healing by others, megavitamins, self-help groups, imagery, commercial diet, folk remedies, lifestyle diet, energy healing, homeopathy, hypnosis, biofeedback and acupuncture. Together, these 16 treatments form a standard definition of CAM that has been widely used in many other studies [2, 3, 7, 10, 16, 17]. We included the following CAM explanations suggested by Eisenberg et al. [2]: megavitamins were defined as high-dose vitamins or megavitamin therapies, not a small daily vitamin or something prescribed by a medical doctor; examples of lifestyle diets included vegetarian or macrobiotics; commercial diet programs were described as “the kind you have to pay for, not losing/gaining weight on your own.” Examples of energy healing included magnets, energy-emitting machines or the “laying on of hands.” Examples of relaxation techniques included meditation and the relaxation response. We inquired whether patients used the above therapies for their bone health and related conditions such as fracture pain, whether they had informed their medical doctor about their CAM use, and whether they had any insurance coverage for the treatments.

    While this standard CAM definition has been widely used, it is potentially too liberal in what it includes as alternative treatments. For example, some may reasonably argue that relaxation techniques are more of a recreational choice employed without specific expected health benefits. As well, Eisenberg et al. have suggested that six of the therapies listed may be considered fairly conventional, namely biofeedback, hypnosis, imagery, relaxation techniques, lifestyle diet and megavitamins [2]. Thus, all analyses were repeated using a more stringent CAM definition that did not include these six items, allowing an additional more conservative estimate of CAM use.

     
  2. 2)

    Prevalence of specific types of CAM use. In addition to obtaining an estimate of general CAM use, we also wanted to ask about certain specific therapies, such as a particular mineral or herbal supplement, that might be highly used in an OPCP population. A literature search, interviews with alternative care practitioners, and visits to downtown Toronto natural food and herbal stores helped generate a list of common CAM bone and joint treatments. These included chondroitin, glucosamine, soy isoflavones, ipriflavone, magnesium, black cohosh, primrose oil, vitamin E, dong quai, spinal manipulation, heat/cold therapy, electrotherapy and joint mobilization.

     
  3. 3)

    Patient satisfaction. A previous US national study suggested that dissatisfaction with conventional medicine and need for control are two reasons why patients turn to CAM [10]. From this national survey, we took four items that measured patient satisfaction with conventional medicine on a four-point Likert scale, (e.g., How much confidence do you have in the medical doctor you see most often for your health care?), as well as an item asking participants how much control they preferred to have over their health care decisions [10].

     
  4. 4)

    Health-related quality of life (HQL). We incorporated the Short-Form-36v2 (SF-36v2), a general health questionnaire that measures HQL on eight domains, including physical functioning, role physical, bodily pain, general health, vitality, social functioning, role emotional and mental health [18]. Scores on these eight scales can be aggregated to form Physical Component Summary (PCS) and Mental Component Summary (MCS) scores. We transformed the scores based on Canadian population norms (not age-matched), with 50 being the average and higher scores indicating better HQL [19].

     
  5. 5)

    Demographic and clinical information. Subjects provided socio-demographic information, including age, sex, birth country, ethnicity, marital status, employment, education, and income. Data on lifestyle such as current smoking, alcohol, and exercise behaviors were also collected. Subjects were also asked about fracture history (any fractures, including fragility), use of conventional bone and joint treatments, and comorbidities. We measured the severity of each subject’s co-existing diseases with the Charlson comorbidity index, in which a higher score indicates more serious comorbid burden [20]. Studies have shown that patient self-administed Charlson indices reliably predict mortality [21] and can be used as a measure of comorbidity.

     

The entire 163-item questionnaire was pilot-tested on 12 UHN outpatients for readability, after which a few minor word changes were made. These pilot participants were included in the final sample as the questionnaire alterations were not major. Subjects who gave inconsistent answers or left large sections of the questionnaire blank were contacted by telephone to clarify their responses.

Bone mineral density

A systematic clinic chart review was undertaken to abstract each subjects’ height, weight and most recent bone mineral density (BMD). In accordance with World Health Organization (WHO) criteria, patients with BMD T-scores ≥ −1.0 were categorized as normal BMD, < −1.0 and > −2.5 as osteopenia, and ≤ −2.5 as osteoporosis [22]. We did not adjust the T-scores for machine and reference population differences among the BMD reports because whether patients were told they had osteoporosis, osteopenia or normal BMD would have been based on these unaltered T-scores. However, to compare the actual BMD in g/cm2, we standardized all readings to Hologic 4500A densitometer equivalents [23].

Potential interactions of CAM

To explore one of our secondary objectives of whether CAM use might have adverse clinical interactions, we searched MEDLINE for potential side effects of the specific alternative treatments listed in our questionnaire. Using the subjects’ self-reported comorbidities, we then identified patients whose underlying medical condition might make them vulnerable to CAM side effects.

Analysis

Data were entered in Microsoft Excel database and statistics computed using SPSS version 10.0. Total ranges for each variable were calculated to identify outliers potentially entered in error. Statistical significance was set at a p-value < 0.05.

For our primary objective, prevalence of CAM use was simply calculated as the percentage of patients endorsing use. For our secondary objective of examining which factors were associated with CAM use, we first performed univariate analysis to assess if there were any differences in CAM users and non-users with respect to their sociodemographics, satisfaction with conventional medicine, clinical status or HQL as assessed in our questionnaire. One-way ANOVAs, post-hoc Tukey tests, and Students’ t-tests were used to compare the means of normally distributed continuous variables. If continuous variables were not normally distributed, we also employed Kruskal-Wallis tests to validate the results of the parametric tests. Then, we conducted hierarchical multivariate logistic regression modeling with CAM use as the outcome. For independent correlates of CAM use, we first entered all variables that we considered to be independent main effects based on our review of the literature (age, satisfaction with conventional medicine, education, HQL), starting with the most significant factors uncovered in our univariate analyses. We then entered independent potential confounders (employment status, income, lifestyle habits, comorbidity) to construct a final model. Males were not included in this final regression model as they were too few in number and might have had significantly different predictors of CAM use that we would have been unable to control for properly.

Results

Sociodemographic characteristics of sample

Table 1 outlines the sample’s sociodemographic characteristics. The average age was 61 years and the vast majority of patients were female and Caucasian. Over one-third of the sample was born outside Canada. Most subjects were married, had post-secondary education, and either worked full-time or were retired. Fifty-six percent reported a family income less than $60 000/year, comparable to the Toronto average of about $57 900/year in 2001 [24].
Table 1

Sociodemographic characteristics of osteoporosis clinic patients (n = 360)

Sociodemographic characteristic

Mean age (range)

60.6 (15.7–90.8)

Female

320 (89.1)

Caucasian (%)

294 (83.8)

Born outside Canada (%)

137 (38.7)

Mean age migration (range)

23.8 (0.5–69.7)

Marital status

Single (%)

61 (17.3)

Married/common law/engaged (%)

215 (61.1)

Separated/divorced (%)

33 (9.4)

Widowed (%)

43 (12.2)

Employment status

Full-time homemaker (%)

51 (14.7)

On disability (%)

22 (6.3)

Retired (%)

118 (33.9)

Unemployed (%)

12 (3.4)

Working part-time (%)

48 (13.8)

Working full-time (%)

97 (27.9)

Education

No formal schooling (%)

3 (0.9)

Elementary (%)

25 (7.1)

High school (%)

111 (31.7)

University/college (%)

211 (60.3)

Gross family income

< 20 000 (%)

38 (12.6)

20 000–40 000(%)

63 (20.9)

40 000–60 000(%)

68 (22.6)

60 000–100 000 (%)

80 (26.6)

> 100 000(%)

52 (17.3)

Live alone (%)

105 (29.9)

Non-smoker during past year

186 (90.7)

≤ 3 alcoholic drinks/week during past year

162 (79.0)

Percentages calculated out of total number of respondents to each question.

Participants were similar across the 3 sites, with few exceptions. Subjects from WCH were more likely to be single or divorced (p < 0.001) and SMH patients were about 5 years older on average (p = 0.025). There were no other significant demographic or clinical differences among the three sites.

Patterns of CAM use

The patterns of CAM use are detailed in Table 2. Fifty-seven percent of respondents used at least one kind of CAM treatment in the past year. Even using a stringent CAM definition, 45% of subjects could still be categorized as having used unconventional medicine. Patients most commonly used herbs (23%), relaxation techniques (20%) and massage therapy (19%).
Table 2

Patterns of complementary and alternative medicine (CAM) use in osteoporosis clinic patients (OPCPs)

Type of therapy

% OPCP used for any reason past yr (n) n = 360

% OPCP used for bone-related health past yr (n) n = 360

% therapies medical doctor aware OPCP used (n)

% therapies with some insurance (n)

Any CAM therapy, standard CAM definition

57.2 (206)

36.4 (131)

44.3 (231)

21.6 (107)

Any CAM therapy, stringent CAM definition

45.0 (162)

21.4 (77)

41.3 (128)

27.8 (81)

Megavitamins

16.9 (61)

13.6 (49)

77.4 (48)

13.6 (8)

Herbal medicine

23.3 (84)

7.5 (27)

42.0 (34)

1.3 (1)

Lifestyle diet

12.5 (45)

8.1 (29)

50.0 (23)

0.0 (0)

Folk medicine

2.5 (9)

0.3 (1)

20.0 (2)

0.0 (0)

Commercial diet

2.8 (10)

0.8 (3)

45.5 (5)

0.0 (0)

Massage

19.2 (69)

11.1 (40)

47.8 (33)

53.7 (36)

Chiropractic

12.5 (45)

7.8 (28)

52.2 (24)

67.4 (29)

Acupuncture

4.2 (15)

2.2 (8)

50.0 (8)

40.0 (6)

Homeopathy

5.0 (18)

1.1 (4)

36.8 (7)

12.5 (2)

Energy healing

6.4 (23)

1.9 (7)

30.4 (7)

13.6 (3)

Relaxation techniques

19.7 (71)

7.5 (27)

30.4 (21)

14.7 (10)

Spiritual healing by others

6.1 (22)

1.7 (6)

15.0 (3)

0.0 (0)

Imagery

8.1 (29)

3.3 (12)

21.4 (6)

15.4 (4)

Biofeedback

1.4 (5)

0.8 (3)

60.0 (3)

40.0 (2)

Hypnosis

0.6 (2)

0.3 (1)

0.0 (0)

100.0 (2)

Self-help group

4.2 (15)

0.8 (3)

33.3 (5)

26.7 (4)

percentage calculated out of total number of patients using each treatment and responding to question

standard definition includes all therapies listed in table; stringent definition excludes biofeedback, hypnosis, imagery, relaxation techniques, lifestyle diet and megavitamins

Under the standard CAM definition, 36% of OPCPs employed alternative medicine at least in part for their bone/joint health. This prevalence drops to 21% under a stringent CAM definition. The most common therapies used for bone/joint health were megavitamins (14%), massage therapy (11%) and lifestyle diets (8%). With respect to the megavitamin users, three of the 61 (4.9%) listed vitamin D as the type of vitamin; none of these three reported vitamin D deficiency as a co-morbidity, although it is possible that they may have neglected to report this condition.

Overall, only 44% of CAM use was disclosed to a medical doctor. This percentage varies widely depending on the therapy, ranging from 15% for spiritual healing, 42% for herbal medicine and 77% for megavitamins. Twenty-two percent of CAM treatments had at least partial insurance coverage, varying from 0% for lifestyle diets, 27% for self-help groups and 67% for chiropractic services.

Univariate comparison of clinical characteristics and HQL between CAM users and nonusers

The clinical features of CAM users and non-users were similar. The non-users did not differ significantly by age, mean number of comorbidities and Charlson scores, prevalence of osteoporosis versus osteopenia and BMD values, fracture history, body mass index, or physical disability (SF-36 Physical Component score). However, users had significantly lower scores on the SF-36 Mental Component score (p = 0.002) (Table 3).
Table 3

Clinical characteristics and health-related quality of life in CAM users and nonusers attending osteoporosis clinics

 

CAM users n = 206

CAM nonusers n = 154

p value

BMD categorization

  

0.963

Osteoporosis (%)

87 (49.4)

62 (48.4)

 

Osteopenia (%)

80 (45.5)

60 (46.9)

 

Normal (%)

9 (5.1)

6 (4.7)

 

Average BMD (g/cm2)

Lumbar L1 to L4 (SE)

0.837 (0.010)

0.837 (0.018)

0.993

Femoral neck (SE)

0.641 (0.008)

0.626 (0.008)

0.208

Total hip (SE)

0.770 (0.010)

0.760 (0.011)

0.474

Comorbidity

Mean # of comorbidities (SE)

2.62 (0.07)

2.77 (0.09)

0.187

Mean Charlson score (SE)

0.69 (0.07)

0.79 (0.09)

0.396

Fracture history

Vertebral fracture (%)

19 (9.5)

19 (12.5)

0.378

Hip fracture (%)

9 (4.5)

7 (4.6)

0.963

Wrist fracture (%)

32 (16.2)

19 (12.5)

0.336

Body mass index in kg/m2 (SE)

23.8 (0.4)

24.0 (0.4)

0.640

Short Form-36v2 scores

Physical component summary (SE)

42.3 (0.9)

41.8 (1.0)

0.742

Mental component summary (SE)

45.7 (1.0)

49.9 (0.9)

0.002

Univariate comparison of satisfaction with conventional medicine between CAM users and nonusers

In univariate analysis, OPCPs who used alternative medicine were less satisfied with conventional medicine - only 50.5% reported being very satisfied, compared to 61.7% of CAM non-users (p = 0.04). Of CAM users, 14.2% preferred to keep control over their own health care rather than have an equal relationship with their doctors, compared to 8.2% of CAM nonusers (p = 0.08). However, there was no difference in the use of prescription drugs for bone health, such as bisphosphonates, raloxifene, calcitonin and hormone replacement therapy, between CAM users and nonusers (67.0% vs 69.3% respectively, p NS).

Multivariate logistic regression for predictors of CAM use

Logistic regression identified three variables that were significantly associated with CAM use after controlling for potential clinical confounders: post-secondary education (odds ratio (OR) 1.96, 95% confidence interval (CI) 1.15 − 3.35); higher mental HQL (OR 0.47, 95% CI 0.23–0.95); and younger age (OR 0.98, 95% CI 0.96 to 1.00) (Hosmer-Lemeshow statistic for global fit of the model was good with a Chi-square of 7.52, p = 0.48).

Potential adverse clinical interactions with CAM

Table 4 outlines specific CAM treatments used by OPCPs. The most common megavitamin supplements included vitamin E (39%), magnesium (20%) and glucosamine (14%). Commonly used CAM services included electrotherapy (5.6%) and spinal manipulation (4.4%).
Table 4

Number of osteoporosis clinic outpatients using specific complementary and alternative medicine (CAM) supplements, and number of patients (pts) who could potentially have an adverse clinical interaction

CAM supplement/service

Used supplement (%) n = 360

Possible adverse clinical interactions of supplement

Number of patients potentially at risk

Number of at-risk patients using supplement (%)

Vitamin E

140 (39.3)

May increase prothrombin time, INR, and bleeding risk in patients using anticoagulant drugs [2527]

Pts who may use anticoagulants:

i) 12 pts angina

i) 9 (75.0)

ii) 13 pts atrial fibrillation

ii) 7 (53.8)

iii) 15 pts stroke/near stroke

iii) 3 (20.0)

iv) 6 pts peripheral vascular disease possibly treated with anticoagulants

iv) 2 (33)

Magnesium

72 (20.4)

Renal disease at risk for hypermagnesemia [30]

41 pts kidney disease

4 (9.7)

Glucosamine

51 (14.3)

May increase insulin resistance [28, 29]

17 pts diabetic

4 (23.5)

Plant estrogens such as soy isoflavones

31 (8.8)

Uncertain effect on breast cancer risk [31, 32]

27 pts history breast cancer

5 (16.1)

Chondroitin

20 (5.7)

Electrotherapy

20 (5.7)

Primrose oil

18 (5.2)

May lower seizure thresholds in patients treated with anticonvulsants [33]

Spinal manipulation

16 (4.5)

Unclear incidence of vertebrovascular accidents [34]

15 pts stroke/near stroke

1 (6.7)

Black cohosh

9 (2.6)

Synthetic phytoestrogens like ipriflavone

6 (1.7)

May inhibit cytochrome P450 system [35]

26 pts liver disease

0 (0)

  

May cause leukopenia [36]

30 transplant pts who may use immunosuppressants

0 (0)

There were 35 cases in which underlying clinical conditions could be worsened by the use of unconventional therapies. For example, we found that many OPCPs who may be using anticoagulant drugs are also using vitamin E, a supplement that may increase the risk of bleeding, particularly in high doses [2527]. As well, glucosamine, which may increase insulin resistance [28, 29], was being used by almost a quarter of diabetic subjects. Further details and potential interactions are outlined in Table 4.

Discussion

Our study of 360 subjects suggests that well over half of patients seen in academic tertiary-care osteoporosis programs use some form of alternative medicine, often to improve bone and joint-related health. This percentage is higher than in studies of other patient groups which have employed the same CAM definition - for instance, prevalence has been estimated at 38.8% in women with early stage breast cancer [16], 42.1% in the US general population [2], 49.8% in patients with systemic lupus erythematosus [17], and 52.5% in HIV patients [37]. Even using a stringent CAM definition, just under half of our sample could still be categorized as employing unconventional therapies. While the literature indicates that younger people are more likely to utilize CAM [2], it is important to recognize that this group of mostly older, post-menopausal women also frequently uses CAM. Thus, for physicians who care for those at risk for osteoporosis, it may be useful to become familiar with alternative bone and post-menopausal treatments.

Why do so many OCPCs turn to CAM? Unfortunately, it is difficult to appreciate what predictors of CAM use mean without an established theoretical framework [38]. There has been some work trying to understand CAM use through a health self-management model with four domains: self care (e.g., taking individual action to research and use a supplement), informal support (e.g., having friends who support CAM use), formal support (e.g., having access to subsidized CAM practitioners), and medical care (having a medical doctor who is non-judgmental about CAM use) [39]. Others have suggested CAM use can be understood as a cultural shift towards holistic health and openness to new ideas [10]; in this sense, the sociodemographic predictors are simply characterizing a group of people that tends to be more aligned with these philosophies.

Yet another approach has been to comprehend CAM use in a socio-behavioral model. Here, the choice to use a health care modality depends on three types of factors that influence decision-making: predisposing, need and enabling [40]. A study specifically using this model found it helpful in interpreting predictors of CAM use by categorizing them into these three domains. For example, health-aware behaviors and dissatisfaction can be viewed as predisposing factors, degree of self-reported medical illness as a need factor, and income as an enabling factor [9]. Our results are probably most congruent with this model. Our univariate analysis did suggest that dissatisfaction and a desire to have more control may have been predisposing factors that made patients more likely to explore alternative therapies. From our multivariate analysis, having both post-secondary education and a higher income may have been enabling factors by giving patients the ability to research and afford CAM. In our sample, might mental health be a need factor?

It is interesting that we found CAM use was associated with decreased mental HQL. Burstein et al. found a similar trend in a longitudinal study of breast cancer patients, in which women who reported new CAM use also reported an increased fear of disease recurrence and greater depression compared with CAM nonusers [16]. Likewise, Moore et al. demonstrated that among SLE patients, those who reported CAM use had lower self-rated health than those who did not despite having similar disease severity [17]. We found that the lower HQL scores also do not seem to be due to worse physical health, as CAM users and nonusers were essentially identical on all clinical parameters including comorbidity and bone status. Furthermore, the physical HQL scores were similar between both groups, suggesting that CAM users were not more physically impaired. A recent study of CAM use in a Canadian academic gastroenterology clinic showed comparable results using the SF-12 [6]. There thus appears to be a common trend across diverse diseases indicating that CAM use correlates with lower mental HQL. Why this pattern exists is unclear, although it has been hypothesized that physical symptoms and psychological stressors lead to poorer perceived HQL and in an attempt to regain control over their illness, patients may turn to CAM [6, 10]. It is also possible that CAM use may be a sign of distress, and it might be helpful for physicians to be cognizant of this potential relationship, but future studies are needed to clarify the association and what clinical import it may have.

Our study also indicates that the use of about half of CAM treatments is not brought to the medical doctor’s attention. This frequency is surprisingly high - for instance, in a study of HIV patients 73.8% of total CAM use was made aware to a health care provider [37]. Physicians caring for osteoporosis patients may benefit from being more vigilant in eliciting a CAM history. When a positive history is noted, it is important for clinicians not to view the use of CAM as a rejection of conventional therapies. CAM users still employed doctor-prescribed bone medications at the same rate as CAM nonusers. Such findings are similar to previous research suggesting that people turn to alternative treatments not as a replacement for conventional medicine but rather as an adjunct [2, 7]. For instance, multiple sclerosis patients who use CAM also utilize outpatient services more frequently than CAM nonusers [41].

Perhaps one of the key difficulties clinicians face, however, is what to advise when one discovers that a patient uses CAM. Given the paucity of research into many of these unconventional medicines, it is difficult for clinicians to make helpful recommendations. The lack of studies precludes accurate estimates of benefit and harm. Certainly, many of the commonly used CAM treatments identified in this study, such as relaxation techniques and massage, seem largely innocuous and likely beneficial. Nonetheless, we do identify several cases in which an apparently harmless supplement could potentially cause an adverse clinical interaction. For example, many OPCPs who may have used anticoagulant drugs for cardiovascular disease also took vitamin E, a supplement that could potentiate their bleeding risks [2527]. Diabetes is quite prevalent in older populations, and some of our subjects with diabetes used glucosamine, an agent that may worsen blood glucose control [28, 29]. Of special concern for tertiary care centres are patients who develop low BMDs secondary to severe kidney or liver disease, subsequently receive organ transplants and then take immunosuppressant therapy. Although none of our 30 transplant subjects took ipriflavone, it is important to note that this bone supplement was found to cause leukopenia in a well-designed randomized trial [36]. We emphasize, however, that we do not know if any of the possible dangers we list actually occurred, as exploring potential side effects was not a primary objective of this study. Our research is unable to calculate the risk for these adverse events. We also strongly stress that many of the possible dangers we list have yet to be rigorously confirmed and that unconventional treatments may have very positive effects. Nonetheless, physicians who are aware that interactions could exist may have the opportunity to prevent an undesired outcome.

This study does have limitations. Most importantly, our sample was drawn from academic hospitals, is likely not fully representative of osteoporosis clinic populations at large and has limited generalizibility to men with osteoporosis. Second, the sampling sites were located in downtown Toronto, a multi-ethnic city whose population may not reflect that of most cities across Canada. Finally, in order to keep the questionnaire a reasonable length, we were unable to explore how often subjects used CAM and the precise costs of various treatments.

Despite these limitations, our data does indicate that a large proportion of patients attending osteoporosis clinics use CAM. Given this, future studies into CAM may be a worthwhile research venture, especially considering that alternative therapies can be very costly. Expenditures on CAM rival those spent on conventional care [2], and in this study, about three-quarters of alternative treatments had no insurance coverage. Further work on developing and testing conceptual models of CAM use may also prove beneficial and help clarify the association between CAM and mental HQL. As well, it may be helpful to explore strategies for encouraging better elicitation and disclosure of CAM use. Perhaps most important, however, is the need to better quantify the potential risks and benefits of commonly employed alternative bone treatments. OPCP patients would benefit from evidence-based CAM advice. Future government measures should build towards a comprehensive health policy, guided by evidence-based data, that ensures the high quality and appropriate practice of alternative medicine.

Acknowledgments

The authors are indebted to Ms. Cherry Mendoza and Ms. Susanne Stewart for their assistance in administering the questionnaire, Ms. Kathryn Herridge for obtaining patient information for us, Ms. Irene Ho for her assistance with the ethics applications, and Ms. Donna Cheung, Mr. Adrian Lau and Ms. Judy Lew for data collection and data entry. The authors would also like to thank the CAM practitioners who provided information on common alternative bone treatments.

Copyright information

© International Osteoporosis Foundation and National Osteoporosis Foundation 2007