May 2011  << Back  

  Can't open the PDF? Click here for help.

 

 RESEARCH ARTICLE

12-month herbal medicine use for mental health from the national Comorbidity survey replication (NCS-R)

Simha E. Ravven, MD

Department of Psychiatry, Harvard Medical School, The Cambridge Hospital, Cambridge, MA, USA

M. Bridget Zimmerman, PhD

Department of Biostatistics, The University of Iowa College of Public Health, Iowa City, IA, USA

Susan K. Schultz, MD

Department of Psychiatry, University of Iowa Hospitals and Clinics, Iowa City, IA, USA

Robert B. Wallace, MS, MD

Department of Internal Medicine. University of Iowa Hospitals and Clinics, Department of Epidemiology, The University of Iowa College of Public Health, Iowa City, IA, USA

BACKGROUND: Herbal medicine is widely used by individuals with mental health problems, although research on their health characteristics and health care utilization patterns—including concomitant treatment with conventional mental health care and psychotropic medication—remains limited.

METHODS: We gathered data from the National Comorbidity Survey Replication (NCS-R), a representative survey of US adults in which diagnoses of mental disorders were based on a fully structured diagnostic interview.

RESULTS: Our analysis found that NCS-R respondents with mental disorders were significantly more likely to have used herbal medicines for mental health problems than respondents who did not meet criteria for a mental disorder. Users of herbal medicines for mental health problems were likely to utilize conventional health care as well, particularly conventional psychiatric medication. Herbal use also was associated with having multiple comorbid medical problems.

CONCLUSIONS: A substantial proportion of US adults use herbal medicine to treat mental health problems. Herbal medicine is frequently used concomitantly with conventional health care, including prescription psychotropic medication. Herbal use also is associated with having multiple chronic medical problems. These factors increase the potential for interactions between herbal medicines and psychiatric and nonpsychiatric medications.

KEYWORDS: complementary and alternative medicine, herbal medicine, mental illness

ANNALS OF CLINICAL PSYCHIATRY 2011;23(2):83-94

  INTRODUCTION

Use of complementary and alternative medicine in general, and herbal medicine in particular, has grown increasingly popular in the United States over the past 2 decades. Alternative medicine was used by 60 million US adults in 1990 and by 83 million in 1997 (representing 33.8% and 42.1% of the adult population, respectively), according to a nationally representative study by Eisenberg et al.1 Herbal medicine use increased >4-fold in this time period, from 2.5% to 12.1% of the US adult population.

Adults with mental illness are more likely than those without psychiatric illness to utilize multiple unconventional medicine modalities: herbal medicines, alternative practitioners (such as acupuncturists, chiropractors, hypnotists, and massage therapists), and self-practices such as meditation.2-6 The term “unconventional medicine” encompasses uses in conjunction with conventional medical care (complementary), as an alternative to conventional medicine, and sequentially with conventional medicine. In this paper, “unconventional medicine” refers to health practices outside the conventional medical system.

US adults use herbs to treat various mental health symptoms, including depression, anxiety, memory problems, sleep problems, and symptoms associated with menopause. Evidence-based research on the efficacy and mechanisms of action of herbal preparations in treating psychiatric conditions is expanding. Notably, St. John’s wort has shown efficacy in treating mild to moderate depression when compared with placebo and various prescription antidepressants (such as imipramine, desipramine, amitriptyline, and maprotiline).7,8

Even so, concerns remain about the standardization and quality of commercially available herbal preparations and the potential for herb users to suffer side effects and herb-herb or herb-drug interactions.9 In addition, herb users often do not discuss this use with their health care providers, which complicates the clinical recognition of herbal medicine’s adverse effects.1,9,10 Thus, an understanding of the scope of use, reasons for use, and sociodemographic characteristics of individuals who use herbal medicines for mental health problems is relevant to the care of patients within all medical specialties.

The National Comorbidity Survey Replication (NCSR) is a nationally representative community household survey of the prevalence and correlates of mental disorders in the United States. Using data from the NCS-R, our study examined herb use for mental health problems among US adults. We explored which herbs are most commonly used, sociodemographic and health characteristics of herb users, and the association between specific psychiatric disorders and herb use for psychiatric symptoms. We also examined mental health and primary care utilization among herb users and conventional psychotropic medication use.

  METHODS

NCS-R survey of US adults

The NCS-R is a nationally representative survey of English-speaking household residents age ≥18 years in the contiguous United States. In face-to-face interviews from 2001 to 2003, researchers administered core diagnostic assessments to 9,282 respondents selected from a probability sample of US community-dwelling adults.11 Before each interview, interviewers explained study procedures and obtained verbal informed consent. Overall response rate was 70.9%. Recruitment and consent were approved by the Human Subjects Committees of Harvard Medical School and the University of Michigan.

A subsample of 5,692 respondents was questioned about risk factors, consequences, other correlates, and additional disorders, as well as utilization of health services and treatments. This subsample included all respondents with a lifetime psychiatric disorder plus a probability subsample of other respondents.11

The NCS-R sample “was weighted to adjust for differential probabilities of selection within households and for differences in intensity of recruitment effort among hard-to-recruit cases.”11 It then was stratified to match census population distribution on geographic and sociodemographic variables.11

Diagnostic criteria and categories. Diagnoses were based on the World Mental Health Survey Initiative version of the World Mental Health Composite International Diagnostic Interview (WMH-CIDI). The WMH-CIDI is a fully structured diagnostic interview that generates ICD-10 and DSM-IV diagnoses. It assesses and accounts for organic exclusions and employs diagnostic hierarchy rules in making diagnoses. NCS-R interviewers received extensive study-specific training before administering the WMH-CIDI.11

Diagnostic categories and disorders. Diagnostic categories were employed to describe groups of disorders. Anxiety disorders included agoraphobia, generalized anxiety disorder, separation anxiety, panic attack, panic disorder, specific phobia, social phobia, and posttraumatic stress disorder. Mood disorders included major depressive disorder, major depressive episode, dysthymia, and bipolar I and II disorders. Impulse control disorders included 4 disorders that share a common feature of difficulty with impulse control: intermittent explosive disorder, oppositional-defiant disorder, conduct disorder, and attention-deficit/hyperactivity disorder. Substance use disorders included alcohol abuse, drug abuse, alcohol dependence, drug dependence, and tobacco dependence.

Survey participants whom interviewers suspected were cognitively impaired were screened for dementia and excluded from the study if they failed this screen.12

Demographic characteristics. Demographic characteristics of all respondents were assessed, including sex, age, race, years of formal education, and employment status. Interview location was used to determine geographic region of residence within the United States (Northeast, Midwest, South, and West). Insurance coverage and type of health insurance were determined. Private insurance, employer provided insurance, and coverage provided by the military were grouped into 1 category called “private insurance.” Public insurance included Medicare and Medicaid as well as other governmental insurance. Insurance that did not fit into either of these categories was listed as “other,” and “no health insurance” comprised a fourth category.

Chronic medical conditions. Respondents were asked about chronic medical conditions. “The next few questions are about health problems you might have had at any time in your life. Have you ever had any of the following: Chronic neck or back problems? Frequent or severe headaches? Any other chronic pain? Seasonal allergies? A stroke? A heart attack?” “Did a doctor ever tell you that you had any of the following illnesses: Heart disease? High blood pressure? Asthma? Tuberculosis? Any other chronic lung disease like COPD or emphysema? High blood sugar? An ulcer in your stomach or intestine? HIV infection or AIDS? Epilepsy or seizures? Cancer?”

Utilization of health services. Respondents were asked about visits to health professionals in the 12 months preceding the interview: “How many visits did you make to each of the following types of health professionals in the past 12 months: A doctor, hospital, or clinic for a routine physical check-up or gynecological exam?”

Utilization of mental health services. Respondents who met criteria for 12-month DSM-IV mental disorders were asked about the number of visits they had in the previous 12 months to a health care provider for “problems with emotions or nerves or mental health or use of alcohol or drugs use.” Respondents were then asked if they had seen a psychiatrist, general practitioner/other medical doctor, psychologist, psychotherapist, mental health nurse, nurse, occupational therapist, or other health professional for mental health reasons.

Use of psychiatric medications. Use of psychotropic medication was assessed in participants who met diagnostic criteria for mental disorders. Respondents were given a list of medications that included sedatives, anti-depressants, tranquilizers, stimulants, and antipsychotic medications and asked: “Which of the medicines on this list did you take in the past 12 months for any of the following problems: problems with your emotions, nerves, mental health, substance use, energy, concentration, sleep, or ability to cope with stress? Include medicines even if you took them only once. Any others?”

Use of herbal medicines. Respondents were given a list of herbal medications including chamomile, kava, lavender, St. John’s wort, valerian, chasteberry, and black cohosh. Respondents were asked, “What types of herbal medicines did you use for your emotions or nerves or mental health or your use of alcohol or drugs?” Respondents who reported use of herbal therapy were asked if they had used additional herbs in the previous 12-month period. When multiple herbs were used, all responses were recorded.

Statistical analyses

Our analyses of data from the NCS-R examined various aspects of 12-month herbal medicine use for mental health in a representative sample of US adults. We compared the demographic characteristics of NCS-R respondents who had used herbal therapy for mental illness with those who had not. We also compared the use of herbal medicine—and particularly St. John’s wort—among respondents with 12-month DSM-IV-endorsed disorders vs respondents without those disorders.

We compared the distribution of the number of visits to a doctor, hospital, or clinic in the past 12 months among respondents who had used herbal therapy vs those who had not. Our logistic regression model of none, 1, and 2 or more annual routine check-up or gynecologic exam was adjusted for race, sex, education, age, and marital status.

NCS-R respondents had been questioned about chronic medical conditions they had in the past or currently. We grouped these conditions into 9 categories: 1) musculoskeletal problems; 2) chronic pain and severe or frequent headaches; 3) asthma and seasonal allergies; 4) stroke, epilepsy/seizure, heart attack, high blood pressure; 5) heart disease; 6) lung disease; 7) diabetes; 8) ulcer/intestinal problems; and 9) cancer. We then compared the number of comorbid conditions reported among those who had used herbal therapy for mental illness in the previous 12 months and those who had not.

We also examined the use vs nonuse of herbal medicine in the previous 12 months in NCS-R respondents with diagnoses of mental disorders who reported using conventional psychiatric medications in that time. Logistic regression models including race, sex, age, education, marital status, number of comorbidities, and region of residence within the United States were constructed.

Finally, we analyzed the number of visits to a non-psychiatrist mental health care provider or other provider for mental health problems in NCS-R respondents with mental disorders, comparing those who had used herbal therapy in the previous 12 months vs those who had not. A logistic regression model that included race, sex, age, education, marital status, number of comorbidities, and region of residence in the United States was constructed.

  RESULTS

TABLE 1 shows 12-month use of all unconventional therapies for problems with mental health among NCS-R respondents. More than one-quarter reported using ≥1 unconventional therapy for mental health problems. Herbal therapy was used for mental health problems by 351 (3.7%) respondents in the 12 months preceding their NCS-R interview, making herbals the most commonly used oral unconventional therapy for mental health problems. Among other unconventional therapies taken by mouth, 1.5% of respondents used special diets, 1.6% used high-dose megavitamins, and 0.5% used homeopathic medicine for mental health problems.

Demographic characteristics of respondents who had used herbal medicines for mental health problems were compared with those who had not (TABLE 2). Participants who used herbs specifically for problems with mental health were more likely than nonherb users to be female and to describe their race as white. Herb users also were somewhat younger than those who did not use herbs (mean age 40.7 years vs 44.9 years, respectively), and they were more likely to never have been married or to be separated, widowed, or divorced. Herb users also had greater educational attainment and were more likely to describe themselves as working, as opposed to being students, retired, or homemakers.

Geographic variations also were associated with herb use. Herbal medicine users were more likely to live in the Western United States and less likely to live in the South. A similar percentage of respondents in both the herb-using group and nonherb-using group hailed from the Northeast and the Midwest.


TABLE 1

Use of unconventional therapies in the past 12 months by NCS-R respondents for problems with emotions or nervesa

Unconventional health practitioner/practice used Frequencyb Percentage (SE of %)
Acupuncturist 65 0.68% (0.09)
Biofeedback specialist 34 0.34% (0.05)
Chiropractor 212 2.24% (0.16)
Energy healing specialist 69 0.72% (0.08)
Exercise/movement therapist 746 7.79% (0.46)
Herbal therapy 351 3.72% (0.22)
High-dose megavitamins 153 1.58% (0.14)
Homeopath 44 0.48% (0.08)
Hypnotist 30 0.30% (0.07)
Guided imagery specialist 90 0.86% (0.10)
Masseuse 329 3.28% (0.22)
Prayer/other spiritual practice 1577 16.24% (0.54)
Relaxation/meditation techniques 662 6.73% (0.37)
Special diets 157 1.48% (0.10)
Spiritual healing by others 161 1.70% (0.16)
a25.89%±0.65 (mean±standard error of the mean).
bNumber of respondents who reported using each therapy.
SE: standard error.
Source: National Comorbidity Survey Replication (NCS-R), 2001-2003. Use of unconventional therapies in past 12 months for problems with emotions or nerves (n=9,271; 11 with missing data for alternative therapies).

Medical insurance coverage differed in the 2 groups: 18.1% of those who used herbs for mental health carried no health insurance, compared with 13.8% of nonherb users (P=.045). Herb users also were less likely than nonherb users to carry public health coverage (Medicare, Medicaid, or another public insurance plan) than those who had not used herbal medicines (7.9% vs 12.5%, respectively).

Herb use was associated with greater illness burden (TABLE 2). Multiple comorbid nonpsychiatric conditions were reported more frequently by herbal medicine users. Herbal medicine users also were less likely to report having no comorbid medical problems.

When psychiatric disorders were grouped into 4 broad diagnostic categories (anxiety disorders, mood disorders, impulse control disorders, substance use disorders) respondents with disorders in each of these categories were significantly more likely to have used herbal medicines for their mental health problems than those who did not meet criteria for a mental disorder (TABLE 3).

We also examined herb use of respondents with specific psychiatric disorders (TABLE 4) and compared herb use among those with and those without these diagnoses. A trend toward increased herb use was associated with each psychiatric diagnosis, with the exception of hypomania. This trend toward increased herb use was most pronounced in respondents with agoraphobia, dysthymia, generalized anxiety disorder, intermittent explosive disorder, major depression and major depressive episode, panic attack, panic disorder, posttraumatic stress disorder, social phobia, specific phobia, and nicotine dependence.

The herbs used most frequently for mental health problems were chamomile and St. John’s wort (TABLE 5). Use of kava, lavender, valerian, chasteberry, and black cohosh also was reported. Many respondents reported using multiple different herbs in the preceding year. We found that St. John’s wort, although generally used for symptoms of depression, was not limited to respondents with mood disorders (TABLE 6). It was used by respondents with disorders in all 4 diagnostic categories: anxiety disorders, mood disorders, impulse control disorders, and substance use disorders.

We found no significant differences in number of mental health and primary care visits between herbal users and nonusers (TABLES 7, 8, AND 9). All respondents reported similar numbers of primary care visits in the previous year. Similarly, among respondents with mental disorders, herb use was not associated with variation in the number of psychiatric visits in the previous 12 months. When we examined utilization of mental health care provided by psychiatrists and nonspecialist providers, however, we found that respondents who had used herbal medicine for mental health were more likely to have seen any provider for mental health problems in the previous year.

Concomitant use of conventional psychotropic medication and herbal medicine was extensive in this 12-month sample (TABLE 10). Nearly one-half of respondents with mental disorders who had used herbal medicines for mental health reported that they also had taken conventional psychiatric medication. Herb use among respondents with a psychiatric disorder was, in fact, associated with greater use of conventional psychiatric medications (>44% of herbal medicine users vs 30.7% of nonherbal medicine users). This association was particularly strong among respondents with anxiety disorders, although a trend toward increased use of conventional psychiatric medication in conjunction with herbal preparations for mental health problems also was observed among respondents with impulse control and substance use disorders. Use of conventional psychiatric medication was highest among respondents with mood disorders (nearly one-half of respondents with mood disorders had used psychiatric medications in the previous year), although no difference in rates of psychiatric medication use was observed between herbal users and nonusers.


TABLE 2

Demographic characteristics of NCS-R respondents who used herbal therapy in past 12 months for problems with emotions or nerves vs those who did not

Demographic variable Used herbal therapy n=351 Did not use herbal therapy n=8,920 P value
Sex (% female [SE of %]) 77.00% (3.34) 51.12% (0.55) <.0001
Race (% [SE of %])
  White
  Hispanic
  Black
  Other

81.44% (2.89)
9.86% (2.06)
5.75% (1.23)
2.95% (1.02)

72.93% (1.96)
10.89% (1.03)
11.77% (1.13)
4.41% (0.42)
.004
Age (mean±SEM) 40.65±0.83 44.92±0.41 <.001
Marital status (% [SE of %])
  Married
  Separated/widowed/divorced
  Never married

50.52% (2.94)
22.00% (1.97)
27.48% (2.61)

56.07% (1.15)
20.37% (0.62)
23.56% (1.14)
.105
Education (% [SE of %])
  0 to 11 years
  12 years
  13 to 15 years
  ≥16 years

7.36% (0.97)
25.83% (2.38)
31.31% (2.65)
35.51% (3.06)

16.46% (0.69)
32.51% (1.23)
27.53% (0.75)
23.49% (1.12)
<.0001
Employment category (% [SE of %])
  Working
  Student
  Homemaker
  Retired
  Other

78.04% (2.32)
3.13% (1.84)
6.95% (1.55)
4.59% (1.14)
7.29% (1.24)

67.37% (0.88)
3.19% (0.45)
5.63% (0.39)
14.86% (0.60)
8.96% (0.58))
.0001
Region of country (% [SE of %])
  Northeast
  Midwest
  South
  West

18.28% (3.95)
21.85% (2.57)
28.18% (3.49)
31.69% (4.86)

19.33% (3.35)
23.21% (1.75)
36.14% (2.04)
21.32% (2.10)
<.0001
Insurance (% [SE of %])
  Private/employer/military
  Medicare/Medicaid/other government
  Other
  None
(n=304)
72.53% (2.05)
7.89% (1.49)
1.53% (0.71)
18.05% (2.16)
(n=5,402)
72.29% (1.34)
12.45% (0.82)
1.42% (0.30)
13.84% (0.74)
.045
Comorbiditya (% [SE of %])
  None
  1
  2
  3
  ≥4
(n=304)
15.28% (2.47)
23.44% (3.42)
24.53% (2.43)
22.13% (2.45)
14.62% (2.10)
(n=5,408)
21.37% (0.75)
27.09% (0.75)
24.56% (0.76)
15.77% (0.75)
11.21% (0.52)
.008
aComorbidities include chronic conditions that the respondent had at any time (past or current). There are 9 types: 1) musculoskeletal (head/neck, arthritis/rheumatism); 2) chronic pain/severe or frequent headaches; 3) asthma/seasonal allergies; 4) stroke, epilepsy/seizure, heart attack, high blood pressure; 5) heart disease; 6) lung disease; 7) diabetes; 8) ulcer/intestinal problems; and 9) cancer.
SE: standard error; SEM: standard error of the mean.
Source: National Comorbidity Survey Replication (NCS-R), 2001-2003.

TABLE 3

NCS-R respondents who used herbal therapy in past 12 months for problems with emotions or nerves, comparing those with vs those without a general mental disorder

Diagnostic categories % with DSM-IV diagnosis who reported herbal use (SE of %) % without DSMIV diagnosis who reported herbal use (SE of %) P value
Anxiety disorders 8.54% (0.73) n=2,198 2.24% (0.16) n=7,073 <.0001
Mood disorders 10.94% (1.26) n=854 3.02% (0.20) n=8,417 <.0001
Disorders related to impulse control 6.65% (0.96) n=580 3.52% (0.22) n=8,691 <.0001
Substance use disorders 8.18% (1.19) n=543 3.41% (0.22) n=8,728 <.0001
SE: standard error.
Source: National Comorbidity Survey Replication (NCS-R), 2001-2003.

TABLE 4

NCS-R respondents who used herbal therapy in past 12 months for problems with emotions or nerves, comparing those with vs those without a specific mental disorder

Specific diagnosis % with DSM-IV diagnosis who reported herb use (SE of %) % without DSM-IV diagnosis who reported herb use (SE of %) P value
Attention-deficit/ hyperactivity disorder 4.16% (1.40), n=190 3.71% (0.22), n=9,081 .732
Agoraphobia 11.85% (3.00), n =142 3.59% (0.21), n=9,129 <.0001
Alcohol abuse 8.12% (3.03), n=213 3.59% (0.23), n=9,058 .038
Alcohol dependence 7.43% (2.81), n=106 3.67% (0.23), n=9,165 .071
Adult separation anxiety 6.57% (2.64), n=156 3.66% (0.22), n=9,115 .158
Bipolar I 8.29% (3.62), n=65 3.69% (0.22), n=9,206 .061
Bipolar II 4.82% (2.56), n=74 3.71% (0.21), n=9,197 .611
Bipolar subthreshold 8.24% (2.77), n=121 3.65% (0.23), n=9,150 .021
Conduct disorder 5.99% (4.16), n=33 3.71% (0.22), n=9,238 .496
Drug abuse 5.22% (1.54), n=102 3.70% (0.22), n=9,169 .239
Drug dependence 13.46% (6.06), n=36 3.68% (0.21), n=9,235 .002
Dysthymia 12.07% (2.56), n=226 3.52% (0.21), n=9,045 <.0001
Generalized anxiety disorder 12.73% (2.20), n=393 3.36% (0.19), n=8,878 <.0001
Hypomania 3.44% (3.11), n=32 3.72% (0.22), n=9,239 .932
Intermittent explosive disorder 7.19% (1.19), n=404 3.56% (0.23), n=8,867 <.0001
Mania 7.29% (2.20), n=188 3.64% (0.22), n=9,083 .025
Major depression 12.55% (1.66), n=656 3.08% (0.21), n=8,615 <.0001
Major depressive episode 11.38% (1.36), n=802 3.03% (0.20), n=8,469 <.0001
Oppositional defiant disorder 5.35% (3.39), n=55 3.70% (0.22), n=9,216 .556
Panic attack 9.51% (1.09), n=992 3.02% (0.20), n=8,279 <.0001
Panic disorder 10.70% (1.83), n=262 3.52% (0.22), n=9,009 <.0001
Posttraumatic stress disorder 12.59% (2.01), n=326 3.41% (0.20), n=8,945 <.0001
Social phobia 9.34% (1.20), n=679 3.28% (0.21), n=8,592 <.0001
Specific phobia 9.16% (1.06), n=873 3.17% (0.19), n=8,398 <.0001
Nicotine dependence 7.58% (1.31), n=312 3.57% (0.22), n=8,959 <.0001
SE: standard error.
Source: National Comorbidity Survey Replication (NCS-R), 2001-2003. Use of alternative therapies in past 12 months for problems with emotions or nerves (n=9,271, 11 with missing data for alternative therapies).

TABLE 5

NCS-R respondents’ use of specific herbs in past 12 months for problems with emotions or nerves (n=9266, 16 missing)

Herb used Frequencya Percentage (SE of %)
Chamomile 215 2.26% (0.14)
St. John’s wort 212 2.26% (0.16)
Kava 76 0.75% (0.11)
Valerian 56 0.54% (0.07)
Lavender 47 0.47% (0.08)
Black cohosh 21 0.19% (0.04)
Chasteberry 4 0.04% (0.02)
*number of respondents who reported using each therapy.
SE: standard error.
Source: National Comorbidity Survey Replication (NCS-R), 2001-2003. Use of alternative therapies in past 12 months for problems with emotions or nerves (n=9271, 11 with missing data for alternative therapies).

TABLE 6

NCS-R respondents who used St. John’s wort in past 12 months for problems with emotions or nerves, comparing those with vs those without a general mental disorder

Diagnostic categories % with DSM-IV diagnosis who reported St. John’s wort use (SE of %) % without DSM-IV diagnosis who reported St. John’s wort use (SE of %) P value
Anxiety disorders 5.21% (0.56) n=2193 1.36% (0.15) n=7073 <.0001
Mood disorders 6.88% (1.00) n=852 1.82% (0.12) n=8414 <.0001
Disorders related to impulse control 4.82% (0.98) n=579 2.09% (0.16) n=8687 <.0001
Substance use disorders 6.02% (1.18) n=542 2.00% (0.15) n=8724 <.0001
SE: standard error.
Source: National Comorbidity Survey Replication (NCS-R), 2001-2003. Use of alternative therapies in past 12 months for problems with emotions or nerves (n=9271, 11 with missing data for alternative therapies).

  DISCUSSION

This study of a nationally representative sample of US adults found substantial herbal medicine use for mental health problems. Previous yearly estimates of herbal medicine use for all reasons range from 10% to 14% of adults using herbal medicine in a given year.1,4,13 In their 1997 survey, Eisenberg et al1 found that 12.1% of all adults had used herbal medicine. In the National Health Interview Survey of 2002, Ni et al13 found that 9.6% of US adults had used herbal medicines in the preceding year. The 2006 Slone Survey, a population-based study of medication use in US adults, found that 22% of respondents had used herbal supplements in the week preceding the interview.4

Our study found that approximately 4% of US adults had used herbal medicines specifically for mental health problems in the previous year. Putting this figure into the context of previous studies, one could infer that approximately one-quarter to one-third of herbal medicine is used with the primary intention of treating mental health symptoms.

Our study found that herb use for emotional problems is more common among persons who meet diagnostic criteria for mental disorders. This confirmed the findings of previous studies showing an association between mental illness, particularly anxiety and depression, and herbal medicine use.3,5,6,14 Existing studies have found higher rates of unconventional medicine use among persons with mental illness to treat a wide range of health problems. Although herbal medicines for mental health are more often employed by persons with mental disorders, they also are used by individuals without current mental disorders. Possible reasons for use include secondary prevention of a lifetime mental disorder or treatment of symptoms that do not meet diagnostic criteria for a mental disorder. Some DSM-IV diagnoses with low prevalence in the population were not included in this study, including nonaffective psychotic disorders. This may account for some of the respondents who used herbal medicine for mental health problems but did not meet diagnostic requirements for a mental illness.

The demographic characteristics of persons who used herbal medicine to treat mental health problems in our study more closely resembled the demographic profile of herb users established in the literature than the sociodemographic characteristics associated with serious mental illness. Serious mental illness has been associated with low income and low attainment of formal education, as well as other markers of low socioeconomic status.15,16 In contrast, higher socioeconomic status, characterized by higher educational attainment and income, has been widely associated with unconventional medicine and herbal medicine use.1,6,17 Our study found that female sex, higher education level, greater medical illness, and residence in the Western United States were associated with using herbal medicine for mental health problems. Greater burden of medical illness has been associated with both mental illness and herbal medicine use in the literature.

The NCS-R included only community-dwelling US adults; homeless and institutionalized persons were not represented. It also did not include all DSM-IV diagnoses. Notably, nonaffective psychoses, including schizophrenia, were not included. The NCS-R researchers stated: “Nonaffective psychoses were excluded from the NCS-R core because previous studies have shown they are dramatically overestimated in lay-administered interviews.”18 In fact, other NCS-R analyses have shown that the majority of 12-month cases (63.7%) of nonaffective psychosis met criteria for other DSM-IV hierarchy-free disorders—primarily anxiety, mood, or substance disorders—and were consequently captured as cases. These study characteristics may skew the demographics of this sample toward higher functioning participants and may decrease the number with serious mental illness in the sample.

We found variation in medical insurance coverage of respondents who used herbs for mental health problems compared with those who did not. Herb users were more likely to be uninsured and less likely to carry public insurance. This may be explained in part by the lower mean age of herb users compared with nonusers (41 years vs 45 years), suggesting that fewer herb users might have reached age eligibility for Medicare coverage. Nearly three-quarters of respondents in both groups reported having private health insurance coverage.

Our finding that herb users are more likely to have multiple chronic medical conditions is consistent with the literature on unconventional medicine use.1,3,5,6 Astin3 found in a sample of US adults that alternative medicine use increased as health status declined. Increased illness burden has been widely associated with both unconventional medicine use and use of herbal medicine.

We found that herbal therapies are largely employed in concert with conventional care. Greater than three-quarters of herb users had had at least 1 primary care visit in the past year, and herb users were no less likely to utilize mental health clinical services. Of particular interest, users of herbal medicine also used prescription psychotropic medications at high rates (nearly one-half of those who used herbal medicine for mental health problems had used psychotropic medications as well in the same 1-year period). These findings are in keeping with previous literature that has found that herbs and other modalities of unconventional medicine are largely used in conjunction with conventional care.1,4,6,9,14

Eisenberg et al1 found in a national sample of US adults that “among those who used unconventional therapy for serious medical conditions, the vast majority (83%) also sought treatment for the same condition from a medical doctor.” The Slone Survey found in a sample of US adults that concomitant herbal and conventional medication use was highest with antidepressant use. This study reported that 22% of US adults used herbals/natural supplements, and 32% of prescription drug users also took herbals or natural supplements.4 They found that the rate of concurrent use was highest with fluoxetine (an antidepressant) at 22%.19

The association between herb use and utilization of conventional care appears to be particularly pronounced in the treatment of mental illness. Our findings also confirm previous studies’ assertions that unconventional medicine use is not motivated by a rejection of conventional care. In fact, the high burden of chronic medical illness among herbal medicine users may motivate them to seek out both conventional and unconventional medical care. Concomitant use of conventional and unconventional care may reflect greater health care needs.

Coincident use of herbal and conventional psychotropic medication has significant potential for side effects and herb-drug interactions. There are multiple potential and documented mechanisms for St. John’s wort interaction with several classes of psychiatric and nonpsychiatric medications. St. John’s wort is thought to work by inhibiting monoamine oxidase and the reuptake of serotonin and norepinephrine. These mechanisms of action underlie the therapeutic effects of several conventional antidepressants as well. St. John’s wort can have an additive effect with several classes of antidepressants, and its use has been associated with serotonin syndrome.20

St. John’s wort also is an inducer of several cytochrome P450 (CYP) isoenzymes and would decrease plasma concentrations of medications metabolized by these isoenzymes. Commonly prescribed psychiatric compounds metabolized by these CYP isoenzymes include norfluoxetine—an active metabolite of the anti-depressant fluoxetine—and the antipsychotic medications aripiprazole and haloperidol.20 St. John’s wort—like other drugs with antidepressant properties—also has the potential to precipitate mania or hypomania, and cases have been documented.20 Several case studies have reported regular St. John’s wort use precipitating acute psychosis among patients with a pre-existing psychotic disorder.21

The association between herbal medicine use and multiple chronic medical conditions suggests that herb users also may take multiple medications for chronic medical conditions. St. John’s wort has been shown to interact with medications commonly prescribed for a wide range of conditions. It has been shown to reduce plasma concentration of warfarin, oral contraceptives, cyclosporine, theophylline, digoxin, and the protease inhibitor indinavir.20

Research into the mechanism and use of St. John’s wort is more extensive than that of the other 2 most commonly used herbs we studied: chamomile and kava. Kava is commonly used to alleviate anxiety and is an effective anxiolytic. It also is thought to potentiate the effects of benzodiazepines, which are commonly prescribed for anxiety disorders.20

Previous studies have found that two-thirds or fewer patients who use unconventional medicine do not inform their doctors of this use.1,9,10 Patients may be unaware of or underestimate the potential for side effects, herb-herb interactions, and interactions with conventional medication, psychiatric and otherwise, and may not see herb use as relevant to discuss with physicians. Patients also may be concerned about stigma associated with herb use.


TABLE 7

Distribution of number of visits to a doctor, hospital, or clinic in past 12 months: NCS-R respondents who used herbal therapy in past 12 months vs those who did not

Type of visit Number of visits % who used herbal therapy (SE of %) % who did not use herbal therapy (SE of %) P value
Routine physical
check-up or
gynecological exam

0
1
2
3 to 4
≥5
(n=303)
22.92% (2.82)
34.21% (2.31)
21.24% (2.37)
11.38% (2.23)
10.25% (1.42)
(n=5394)
25.39% (0.95)
36.59% (1.05)
15.58% (0.57)
11.78% (0.61)
10.66% (0.51)
.381
SE: standard error.
Source: National Comorbidity Survey Replication (NCS-R), 2001-2003. Utilization of Conventional Health Care—all subgroup respondents.

TABLE 8

Distribution of number of psychiatrist visits in past 12 months in NCS-R respondents with diagnosis of specific psychiatric disorders: Respondents who used herbal therapy in past 12 months vs those who did not

Diagnostic category Number of psychiatrist visits % who used herbal therapy (SE of %) % who did not use herbal therapy (SE of %) P value
Anxiety disorders
0
1
2 to 5
6 to 9
≥10
(n=179)
88.30% (2.64)
3.23% (1.56)
2.89% (1.38)
1.78% (1.24)
3.81% (1.59)
(n=1847)
88.28% (0.91)
2.39% (0.70)
4.50% (0.55)
1.56% (0.30)
3.28% (0.73)
.860
Mood disorders
0
1
2 to 5
6 to 9
≥10
(n=77)
81.93% (3.92)
4.31% (2.42)
4.50% (2.13)
3.77% (2.57)
5.50% (2.92)
(n=729)
78.27% (1.86)
3.85% (1.86)
8.59% (1.06)
2.30% (0.57)
7.00% (1.08)
.978
Disorders related to impulse control
0
1
2 to 5
6 to 9
≥10
(n=39)
93.90% (4.52)
2.61% (2.58)


3.49% (3.64)
(n=512)
87.31% (1.43)
1.42% (0.82)
5.62% (1.33)
1.59% (0.57)
4.06% (0.94)
.155
Substance use disorders
0
1
2 to 5
6 to 9
≥10
(n=42)
89.82% (5.73)
4.85% (4.84)
3.01% (2.22)

2.32% (2.25)
(n=468)
88.68% (1.24)
2.26% (1.03)
4.44% (1.05)
1.73% (0.56)
2.89% (0.92)
.714
SE: standard error.
Source: Utilization of Conventional Health Care—among those with specific DSM-IV endorsed disorders.

TABLE 9

Distribution of number of visits to mental health/other health professionala in past 12 months for problems with emotions, nerves, or alcohol or drug use in NCS-R respondents with diagnosis of specific psychiatric disorders: Respondents who used herbal therapy in past 12 months vs those who did notb

Diagnostic category Number of visits % who used herbal therapy (SE of %) % who did not use herbal therapy (SE of %) P value
Any psychiatric disorder
0
1
2 to 5
6 to 9
≥10
(n=210)
57.19% (3.69)
10.07% (4.29)
15.31% (2.36)
3.17% (1.13)
14.26% (2.87)
(n=2414)
73.24% (1.05)
7.41% (0.68)
9.93% (0.84)
3.10% (0.36)
6.33% (0.60)
.0004
Anxiety disorders
0
1
2 to 5
6 to 9
≥10
(n=181)
57.77% (3.97)
8.62% (3.65)
15.29% (2.58)
3.66% (1.29)
14.66% (2.95)
(n=1882)
71.41% (1.18)
7.66% (0.72)
10.73% (1.03)
3.21% (0.43)
6.99% (0.78)
<.001
Mood disorders
0
1
2 to 5
6 to 9
≥10
(n=78)
45.32% (6.61)
16.03% (8.89)
13.03% (3.81)
4.99% (2.60)
20.63% (5.39)
(n=738)
54.62% (2.41)
10.84% (1.35)
14.75% (1.46)
6.29% (1.02)
13.51% (1.41)
.109
Disorders related to impulse control
0
1
2 to 5
6 to 9
≥10
(n=40)
64.35% (6.85)
8.44% (4.33)
21.42% (6.71)
2.36% (2.34)
3.43% (3.57)
(n=515)
74.76% (2.24)
5.55% (1.25)
8.30% (1.67)
3.79% (0.97)
7.60% (1.20)
.341
Substance use disorders
0
1
2 to 5
6 to 9
≥10
(n=42)
67.65% (6.82)
7.90% (5.63)
17.00% (5.80)
3.35% (2.40)
4.09% (2.80)
(n=471)
74.55% (1.77)
6.75% (1.42)
7.75% (1.48)
4.32% (0.84)
6.63% (1.18)
.539
aPsychiatrist, general practitioner/other medical doctor, psychologist, psychotherapist, mental health nurse/other mental health professional, nurse, occupational therapist, other health professional.
bFrom logistic regression that included race, sex, age, education, marital status, number of comorbidities, and region of residence in the United States as covariates.
SE: standard error.
Source: Utilization of Conventional Health Care—among those with specific DSM-IV endorsed disorders.

TABLE 10

Use of conventional psychiatric medications (sedatives/sleeping pills, antidepressants, tranquilizers, amphetamines/stimulants, antipsychotics, other medications for emotions, substance abuse, and energy) in the past 12 months in those with diagnosis of specific psychiatric disorders, comparing those who used herbal therapy in past 12 vs those who did nota

Diagnostic category % who did use herbal therapy (SE of %) % who did not use herbal therapy (SE of %) χ2P value
Any psychiatric disorder 44.18% (5.23), n=218 30.69% (1.32), n=2565 .008
Anxiety disorders 47.67% (5.13), n=186 32.36% (1.32), n=2010 .003
Mood disorders 46.15% (8.47), n=82 47.76% (2.20), n=771 .859
Disorders related to impulse control 45.02% (7.75), n=41 30.94% (2.37), n=539 .076
Substance use disorders 40.58% (7.51), n=44 31.29% (2.47), n=499 .197
*From logistic regression that included race, sex, age, education, marital status, number of comorbidities, and region of residence in the United States as covariates.
SE: standard error.
Source: National Comorbidity Survey Replication (NCS-R), 2001-2003. Utilization of Conventional Health Care—among those with specific DSM-IV endorsed disorders.

  CONCLUSIONS

Our analysis of data from the NCS-R demonstrates substantial use of herbal medicines by US adults to treat mental health problems and other medical conditions. Based on this representative sample, we found that herbal medicine largely is used concomitantly with conventional mental health care and primary care, including prescription psychotropic medication. Herb use is also associated with having multiple chronic medical problems.

Significant potential for herb-drug interaction exists with psychiatric and nonpsychiatric medications, given the increased medical illness burden of herb users and the prevalence of coincident use of psychiatric medication and herbs. Health care providers who care for persons with mental illness should systematically ask their patients about herbal medicine use.

ACKNOWLEDGMENTS: This publication was made possible by the National Institute of Environmental Health Sciences (NIEHS) and the Office of Dietary Supplements (ODS), NIH (grant P01 ES012020), Doris Duke Charitable Foundation, and the American Federation for Aging Research.

DISCLOSURES: The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

    REFERENCES

  1. Eisenberg DM, Davis RB, Ettner SL, et al Trends in alternative medicine use in the United States, 1990-1997: results of a follow-up national survey. JAMA. 1998;280:1569–1575.
  2. Alderman CP, Kiepfer B. Complementary medicine use by psychiatry patients of an Australian hospital. Ann Pharmacother. 2003;37:1779–1784.
  3. Astin J. Why patients use alternative medicine: results from a national study. JAMA. 1998;279:1548–1553.
  4. Patterns of medication use in the United States 2006. A report from the Slone Survey. Boston MA: Slone Epidemiology Center at Boston University; 2007.
  5. Bausell RB, Lee W, Berman BM. Demographic and health-related correlates of visits to complementary and alternative medical providers. Med Care. 2001;39:190–196.
  6. Unützer J, Klap R, Strum R, et al Mental disorders and the use of alternative medicine results from a national survey. Am J Psychiatry. 2000;157:1851–1857.
  7. Wong AH, Smith M, Boon HS. Herbal remedies in psychiatric practice. Arch Gen Psychiatry. 1998; 55:1033–1044.
  8. Linde K, Ramirez G, Mulrow CD, et al St John’s wort for depression—an overview and meta-analysis of randomised clinical trials. BMJ. 1996;313:253–258.
  9. Druss BG, Rosenheck RA. Association between use of unconventional therapies and conventional medical services. JAMA. 1999;282:651–656.
  10. Kennedy J. Herb and supplement use in the US adult population. Clin Ther. 2005;27:1847–1858.
  11. Kessler RC, Berglund P, Chiu WT, et al The US National Comorbidity Survey Replication (NCS-R): design and field procedures. Int J Methods Psychiatr Res. 2004;13:69–92.
  12.  NCS-R Dementia note to all users, Available at: http://www.hcp.med.harvard.edu/ncs/notes_dementia.php. Accessed January 1 2009.
  13. Ni H, Simile C, Hardy AM. Utilization of complementary and alternative medicine by United States adults: results from the 1999 National Health Interview Survey. Med Care. 2002;40:353–358.
  14. Kessler RC, Soukup J, Davis RB, et al The use of complementary and alternative therapies to treat anxiety and depression in the United States. Am J Psychiatry. 2001;158:289–294.
  15. Muntaner C, Eaton WW, Miech R, et al Socioeconomic position and major mental disorders. Epidemiol Rev. 2004;26:53–62.
  16. Fryers T, Melzer D, Jenkins R, et al The distribution of common mental disorders: social inequalities in Europe. Clin Pract Epidemiol Ment Health. 2005;1:14.
  17. Ness J, Cirillo DJ, Weir DR, et al Use of complementary medicine in older Americans: results from the Health and Retirement Study. Gerontologist. 2005;45:516–524.
  18. Kessler RC, Chiu WT, Demler O, et al Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry. 2005;62:617–627.
  19. Kaufman DW, Kelly JP, Rosenberg L, et al Recent patterns of medication use in the ambulatory adult population of the United States: the Slone survey. JAMA. 2002;287:337–344.
  20. Izzo AA, Ernst E. Interactions between herbal medicines and prescribed drugs: a systematic review. Drugs. 2001;61:2163–2175.
  21. Lal S, Iskandar H. St. John’s wort and schizophrenia. CMAJ. 2000;163:262–263.

CORRESPONDENCE: Simha E. Ravven, MD The Cambridge Hospital Department of Psychiatry 1493 Cambridge Street Cambridge, MA 02193 USA E-MAIL: sravven@cha.harvard.edu