When Conventional Medical Providers Recommend Unconventional Medicine: Results of a National Study

In 2007, 38% of Americans used complementary and alternative medicine (CAM).¹ Rates of CAM use have increased since 2002, with mind-body therapies (MBT) composing 75% of this rise.¹ Evidence to support the therapeutic use of MBT (including yoga, tai chi, qi gong, meditation, guided imagery, progressive muscle relaxation, and deep-breathing exercises)² is growing.³ Little is known about the use of MBT by patients as a result of conventional medical provider recommendation. Our study objective was to compare patients using MBT as a result of conventional medical provider referral with those who self-referred for MBT.

We obtained data from the 2007 National Health Interview Survey (NHIS),⁴ which uses a cross-sectional, multistage, stratified sampling design to question randomly selected households within the United States. The final sample included 23 393 respondents (response rate, 67.8%).⁴ We were interested in respondents who had used MBT (n = 4296) who were asked “In the past 12 months, did you use [an MBT] because it was recommended by a health care provider?” Respondents answering “yes” were classified as provider-referred mind-body therapy (P-MBT) users and those answering “no” were classified as self-referred mind-body therapy (S-MBT) users.

To account for the complex sampling scheme used by NHIS, we conducted weighted analyses using SAS-callable SUDAAN (version 10.0; SAS Institute Inc, Cary, North Carolina). Potential correlates of P-MBT use were (1) sociodemographic characteristics (age, sex, race, education, income, region, marital status, and insurance status), (2) health status (current health, comparison with prior year, number of days in bed because of illness, and the Charlson Comorbidity Index), (3) pre-existing medical conditions (using the 16 most prevalent comorbidities in the United States), (4) health behaviors (smoking, alcohol use, exercise, and body mass index), and (5) health care utilization (office visits in the past year, emergency department visits in the past year, and encounters with general physicians, medical specialists, and mental health professionals). Multivariable logistic regression modeling identified those factors independently associated with P-MBT use.

Nearly 1 in 30 Americans (2.9% of respondents), representing 6.36 million Americans, reported using P-MBT (n = 668) compared with 15.5% of respondents, representing 34.8 million Americans, who reported using S-MBT (n = 3628) in the past 12 months. The mean age of P-MBT users was 46.8 years compared with 43.4 years among S-MBT users. Deep-breathing exercises were the most common P-MBT used (84.4%), followed by meditation (49.3%), yoga (22.6%), progressive muscle relaxation (19.9%), and guided imagery (13.9%); similar trends were seen in the S-MBT group. The total percentage exceeded 100% because more than 1 MBT modality was used by some respondents.

Our adjusted multivariable analyses identified factors independently associated with P-MBT use (Table). No sociodemographic characteristics were independently associated with P-MBT use. Of our health status markers, higher Charlson scores were associated with a greater likelihood of P-MBT use. Respondents with more chronic conditions, quantified by a Charlson score of 4 or higher and composing 11.3% of all P-MBT users, were more likely to use P-MBT. Of our 16 comorbid conditions, only chronic obstructive pulmonary disease and anxiety were associated with P-MBT use. Greater health care use was associated with a greater use of P-MBT. We observed a “dose-response” relationship with the number of office visits and the use of P-MBT: as the number of office visits increased over a 12-month period, so did the likelihood of using P-MBT. Use of P-MBT was associated with an encounter with a mental health professional over the past 12 months. Finally, respondents with heavy alcohol use were less likely to use P-MBT.

To our knowledge, our study is the first to examine patient factors associated with the use of MBT as a result of conventional medical provider referral. We found that individuals who used P-MBT tended to have a greater illness burden and use the health care system more than their counterparts who self-referred for MBT. This is consistent with prior literature showing that increasing comorbidities correspond to greater rates of overall CAM use,⁵ and CAM users are high users of conventional health care services.^5,6 Our data suggest that conventional health care providers treating sicker patients with more frequent office visits may offer referrals for MBT as a last resort once conventional therapeutic options have been exhausted or have failed.

Both anxiety and visits to a mental health professional in the past year was associated with P-MBT use. Recent data suggest that the majority of patients who have seen a psychiatrist for treatment of anxiety or depression have also used CAM,⁷ and the association between MBT use and anxiety is well documented.^2,8

Although MBT shows promise in the treatment of substance abuse,⁹ heavy alcohol users compose the smallest proportion of MBT users overall.² Possibly, physicians refer patients who drink heavily to MBT, but a variety of barriers prevent their use of P-MBT.

Whether MBT referrals could result in improved patient outcomes or decreased health care use if offered earlier in the course of illness remains to be seen. Physicians' referrals for MBT may inform recommendations for use, highlight areas of underuse or overuse, or may suggest areas for future research and intervention.

Correspondence: Dr Nerurkar, Osher Research Center, Harvard Medical School, 77 Avenue Louis Pasteur, Ste 1030, Boston, MA 02115 (Aditi_Nerurkar@hms.harvard.edu).

Author Contributions:Study concept and design: Nerurkar, Yeh, Birdee, and Phillips. Acquisition of data: Nerurkar. Analysis and interpretation of data: Nerurkar, Yeh, Davis, Birdee, and Phillips. Drafting of the manuscript: Nerurkar, Birdee, and Phillips. Critical revision of the manuscript for important intellectual content: Nerurkar, Yeh, Davis, Birdee, and Phillips. Statistical analysis: Nerurkar, Davis, Birdee, and Phillips. Obtained funding: Phillips. Administrative, technical, and material support: Phillips. Study supervision: Yeh and Phillips.

Financial Disclosure: None reported.

Funding/Support: Dr Nerurkar is supported by an Institutional National Research Service Award (T32AT000051-11) from the National Institutes of Health. Dr Phillips and Davis are supported by a Mid-Career Investigator Award (K24-AT000589) from the National Center for Complementary and Alternative Medicine, National Institutes of Health.

Disclaimer: The analyses, interpretations, and conclusions are of the authors and do not reflect the views of the National Center for Health Statistics, the Centers for Disease Control and Prevention, the National Center for Complementary and Alternative Medicine, or the National Institutes of Health.

Previous Presentation: A portion of this study was presented at the annual meeting of the Society of General Internal Medicine; April 29, 2010; Minneapolis, Minnesota.