To Name :
To Email :
From Name :
From Email :
Comments :

Evidence-Based Reviews

CAM for your anxious patient: What the evidence says

Vol. 9, No. 10 / October 2010
This week's quiz:
More »

Dr. Antonacci discusses when CAM might be appropriate for your patient with anxiety

The number of people with psychiatric disorders who use complementary and alternative medicine (CAM) is on the rise. In surveys of patients seeking psychiatric care, estimates of CAM use range from 8% to 57%; the most frequent uses are for depression and anxiety disorders. A population-based study in the United States found that 9% of respondents had anxiety attacks and 57% of these individuals had used CAM.1 Similarly, in a Finnish population-based study (N=5,987) 35% of subjects reported some form of CAM use in the previous year; those with comorbid anxiety and depressive disorders used CAM most frequently.2

Unfortunately, a MEDLINE search shows that the number of studies examining psychotropic medications dwarfs the number of studies on even the most common CAM treatments used for psychiatric disorders. Far more patients with diagnosed mental disorders are studied in trials of standard treatments than CAM treatments. Because very few studies evaluate the cost-effectiveness of CAM treatments for psychiatric disorders, the risk-to-benefit ratio is difficult to calculate. Although several CAM treatments for depressive disorders have enough support to be considered options,3 CAM options for anxiety disorders are fewer and have less evidence of efficacy.

For these reasons, it is hard to recommend any CAM treatment as first line. Despite the relative lack of high quality research on CAM treatment outcomes, high rates of CAM use make it critical for clinicians to understand what treatments are available—or at least which treatments should be favored if patients are intent on trying them. We review the current research for yoga, exercise, bibliotherapy, and the dietary supplements kava and inositol for treating anxiety disorders and suggest those that warrant consideration for patients who do not respond, respond partially, or suffer from side effects from selective serotonin reuptake inhibitors (SSRIs) or benzodiazepines.

Limitations of CAM research

There are several limitations to the research literature on CAM approaches for anxiety disorders.4 First, there is a wide diversity of practices considered alternative or complementary and various ways in which these methods are applied across cultures. Some authors consider complementary medicines to be only herbal remedies, whereas others include individual therapies such as acupuncture, aromatherapy, herbal therapy, homeopathy, iridology, naturopathy, and reflexology.5 This article defines “alternative” treatments as those other than a form of psychotherapy or an FDA-approved medication that substitute for standard psychiatric treatment, and “complementary” approaches as those used to augment standard psychiatric treatments.

Anxiety and stress are ubiquitous, perhaps motivating interest in CAM options and prompting research on heterogeneous groups of individuals with poorly defined clinical syndromes or with isolated symptoms of anxiety or subjective distress. Few studies examine well-defined patient groups with diagnosed anxiety disorders. There are also multiple research design problems, including poorly specified treatments, poorly chosen placebos, and interpreting nonsignificant differences from established treatments as equivalence in underpowered studies.

The CAM treatments reviewed in this article have ≥2 randomized controlled trials (RCTs) that support their use for patients with diagnosed anxiety disorders, and ≥1 study that shows that the treatment can induce remission.


In 2005 Kirkwood et al carried out the first systematic review of research evidence for the effectiveness of yoga in anxiety treatment.6 Of 19 studies identified, 4 RCTs and 1 nonrandomized trial met their inclusion criteria, which were an anxiety disorder diagnosis, use of yoga or yoga-based exercises alone, and anxiety rating scales used as outcome measures. Most found significant improvement in anxiety symptoms with yoga compared with placebo. Details of the 5 trials evaluated in Kirkwood’s review are summarized in Table 1.7-11

Since the 2005 review, 3 additional studies of yoga and anxiety have been published, but none would meet Kirkwood’s inclusion criteria. One that evaluated a heterogeneous group of patients using an intervention with multiple components—only 1 of which was yoga—found the intervention significantly reduced anxiety scores.12 A second study comparing yoga with relaxation in 131 patients with mild-to-moderate stress but no anxiety disorder diagnosis showed yoga was as effective as relaxation in improving anxiety symptoms as measured by the anxiety subscale of the State Trait Personality Inventory.13 In a study of 183 nonrandomized survivors of the 2004 southeast Asia tsunami with posttraumatic stress disorder (PTSD) symptoms, yoga-based breathing either alone or paired with trauma reduction exposure techniques significantly reduced PTSD symptoms compared with wait-list controls.14

Conclusion. Few controlled studies evaluated yoga for anxiety disorders, and all have significant methodologic limitations and/or poor methodology reporting. The diagnostic conditions treated and both yoga interventions and control conditions varied. However, these limited results are encouraging, particularly for treatment of obsessive-compulsive disorder (OCD). There is little information regarding safety or contraindications of yoga. Reported attrition rates were high in most studies, which may raise concerns about patient motivation and compliance.

Table 1

Evidence on the effectiveness of yoga for anxiety disorders




Vahia et al, 19737

36 patients with psychoneurosis randomly assigned to yoga (N=15) or a control intervention of relaxation, postures, breathing, and writing (N=12)

Significant difference between groups in TAS scores after but not before treatment. Reduction in mean TAS score for yoga group but not control group

Vahia et al,19738

39 patients received 6 weeks of yoga (N=21) or medication (amitriptyline and chlordiazepoxide on a variable dosage schedule) (N=18)

Yoga showed significantly greater reductions in TAS in this non-randomized sample

Sahasi et al, 19899

91 patients randomly assigned to yoga practiced daily for 40 minutes (N=38) or diazepam at unspecified frequency or doses (N=53) for 3 months

Mean reduction in IPAT with yoga (3.39, P < .05) vs control group (0.36, P > .05). Attrition rate was 21.1% in yoga group and 66% in controls

Sharma et al, 199110

71 patients with anxiety neurosis randomly assigned to 1-week yoga training, then daily practice (N=41) or control (N=30, placebo capsule)

HAM-A measured at 3 weekly intervals for 12 weeks. Significant between group mean difference at 3 weeks (greater improvement in yoga group compared with controls). Significant improvement in yoga group between 3 and 6 weeks but not for controls

Shannahoff-Khalsa et al, 199911

21 OCD patients randomly assigned to kundalini yoga (N=11) or relaxation and mindfulness meditation (N=10). Multiple outcome measures; Y-BOCS was primary

Seven in each group completed 3 months; patients who practiced yoga demonstrated greater improvements on Y-BOCS. Intent-to-treat analysis (Y-BOCS) for the baseline and 3-month tests showed that only the yoga group improved. Groups were merged for an additional year of yoga; at 15 months, the final group (N=11) improved 71% on the Y-BOCS

HAM-A: Hamilton Anxiety Rating scale; IPAT: Institute for Personality and Ability Testing, Anxiety Scale; OCD: obsessive-compulsive disorder; TAS: Taylor’s Anxiety Scale; Y-BOCS: Yale-Brown Obsessive Compulsive Scale


The literature examining the relationship between exercise and depression is extensive, but much less has been published about exercise in patients with anxiety disorders (Table 2).15-17 In a 10-week trial, Broocks and colleagues compared clomipramine, exercise (running), and placebo in 46 outpatients with panic disorder.15 Both exercise and clomipramine, 112.5 mg/d, significantly reduced panic symptoms compared with placebo, but clomipramine was more effective and faster-acting.

A more recent RCT compared group cognitive-behavioral therapy (GCBT) plus a home-based walking program vs GCBT and in 21 patients with panic disorder, generalized anxiety disorder (GAD), or social phobia.16 Compared with GCBT plus educational sessions, GCBT plus walking had a significant effect on self-reported depression, anxiety, and stress. Results differed by diagnosis; the most marked effects occurred in individuals with social phobia, whereas benefits for those with panic disorder or GAD were questionable.

Fifteen patients with OCD were recruited to participate in a 12-week, moderate-intensity aerobic exercise program added to their standard behavioral and/or pharmacologic treatment.17 Subjects demonstrated improvement in negative mood, anxiety, obsessions, and compulsions after each exercise session. Changes after each session persisted over the 12-week intervention, although the magnitude attenuated over the duration of the intervention.

Conclusion. Although initial results from small trials suggest exercise may help improve anxiety symptoms, further studies are needed to determine how to best use exercise training to treat anxious patients, specifically regarding dose-response relationship, differences in effectiveness between aerobic and resistance training, and the mechanisms by which exercise improves psychiatric symptoms.

Table 2

Exercise for anxiety: More research is needed




Broocks et al, 199915

46 patients with panic disorder randomly assigned to 10 weeks of running, clomipramine, or wplacebo pills

Both exercise and clomipramine resulted in significant decreases in symptoms but clomipramine improved symptoms earlier and more effectively

Merom et al, 200816

21 patients with panic disorder, GAD, or social phobia randomly assigned to GCBT and either a home-based walking program or educational sessions

GCBT plus walking had a significant effect on depression, anxiety, and stress compared with GCBT plus educational sessions

Abrantes et al, 200917

15 patients with OCD assigned to a 12-week exercise intervention that was added to their standard behavioral and/or pharmacologic treatment

Subjects reported improved mood, anxiety, obsessions, and compulsions after each exercise session

GAD: generalized anxiety disorder; GCBT: group cognitive-behavioral therapy


Investigation of bibliotherapy for treatment of anxiety disorders has been limited (Table 3).18-20 A 2009 RCT demonstrated that for 21 patients with mild-to-moderate social phobia, bibliotherapy—in the form of an 8-week self-directed CBT program with minimal therapist involvement—was superior to a wait-list control and induced clinically significant change in approximately one-third of patients.20

Rapee et al randomly assigned 267 children age 6 to 12 with anxiety disorders to bibliotherapy that consisted of parents treating their children in the home with written materials, 9 sessions of GCBT, or a wait-list control condition.19 Bibliotherapy provided by parents demonstrated benefit compared with wait-listing but was not as efficacious as GCBT at post-treatment and 3-month follow-up.

Lidren and colleagues randomly assigned 36 adult patients with panic disorder to bibliotherapy, group therapy combined with bibliotherapy, or a waitlist.18 Both treatments were more effective than wait-listing in reducing the frequency of panic attacks, severity of physical panic symptoms, catastrophic cognitions, agoraphobic avoidance, and depression. Both interventions maintained their effects at 3-and 6-month follow-up and produced clinically significant change in most patients.

Conclusion. Some preliminary evidence supports the effectiveness of bibliotherapy for social anxiety disorder, childhood anxiety disorders, and panic disorder.

Table 3

Preliminary evidence supports bibliotherapy for select anxiety disorders




Lidren et al, 199418

36 adults with panic disorder randomly assigned to bibliotherapy, bibliotherapy plus group therapy, or wait-list control

Both bibliotherapy and bibliotherapy plus group therapy were more effective than wait-listing in reducing the frequency of panic attacks and severity of physical panic symptoms

Rapee et al, 200619

267 children with anxiety disorders randomly assigned to bibliotherapy (parents treating their children in the home with written materials with no therapist contact), 9 sessions of group CBT, or wait-list control

Parent-delivered bibliotherapy was beneficial compared with wait-listing but was not as efficacious as group CBT

Abramowitz et al, 200920

21 patients with mild-to-moderate social phobia underwent an 8-week self-directed CBT program with minimal therapist involvement

Bibliotherapy was superior to wait-listing. One-third of patients experienced clinically significant change

CBT: cognitive-behavioral therapy

Dietary supplements

Many dietary and herbal supplements are purported to have therapeutic efficacy for anxiety symptoms. Because of inadequate FDA regulation of manufacturing and marketing of these agents, most of these supplements have not been tested on patients with anxiety disorders.21 Limited evidence supports the use of kava for GAD and inositol for panic disorder (Table 4).22-28

Kava. Multiple double-blind RCTs found kava (Piper methysticum)—a plant indigenous to South Pacific islands—has effects greater than placebo and comparable to standard treatments for mild to moderately severe GAD. A Cochrane meta-analysis22 of 11 trials with 645 participants concluded that kava is effective for reducing GAD symptoms, with risks comparable to standard treatments for up to 6 months of use.

Did you miss this content?
10 Triggers of inflammation to be avoided, to reduce the risk of depression