Frequently Asked Questions

In this randomized controlled trial, Mindfulness-Based Stress Reduction (MBSR) was associated with greater improvement than an active control on several anxiety outcomes in adults with generalized anxiety disorder. Both groups improved on the primary Hamilton Anxiety Rating Scale (HAMA) outcome, but MBSR showed significantly greater benefit on the Clinical Global Impressions-Severity and -Improvement scales and on the Beck Anxiety Inventory (all P values < .05). The authors concluded that MBSR may have a beneficial effect on anxiety symptoms in GAD.

MBSR outperformed the time- and attention-matched Stress Management Education (SME) control on several measures, but not on every outcome. Both groups had significant reductions in HAMA scores at endpoint, and the between-group difference on HAMA was not significant (treatment arm × time interaction: F1,87 = 1.38, P = .244). However, MBSR had significantly greater improvement on the Beck Anxiety Inventory (F1,79 = 4.31, P = .041), CGI-Severity (F1,84 = 4.51, P = .0366), and CGI-Improvement responder status.

The response rate was 66% with MBSR versus 40% with SME. Response was defined as a CGI-I rating of 1 (very much improved) or 2 (much improved) at endpoint. The difference was statistically significant (29/44 vs 14/35; P = .025), with a risk ratio of 1.65 (95% CI, 1.04-2.60) and a number needed to treat of 3.9 (95% CI, 2.23-26.19).

Yes. Compared with SME, MBSR was associated with greater reductions in anxiety and distress during the Trier Social Stress Test, a standardized laboratory stress challenge. State anxiety scores (STAI-S) decreased from 53.9 to 40.8 with MBSR versus 52.2 to 45.2 with SME (treatment arm × time interaction: F1,73 = 4.37, P = .040), and distress ratings (SUDS) decreased from 53.2 to 28.7 versus 50.5 to 39.4 (F1,68 = 5.25, P = .025).

Yes. MBSR was associated with a greater increase in positive self-statements during the Trier Social Stress Test than SME. Positive self-statement scores increased from 15.9 to 18.6 with MBSR but changed from 16.9 to 16.5 with SME, and the between-group difference was significant (F1,72 = 8.64, P = .004).

Negative self-statements decreased in both groups, but the treatment arm × time interaction was not significant (F1,72 = 1.32, P = .26).

Among participants not taking psychiatric medications, sleep improved more with MBSR than with SME. In that subgroup (n = 61), Pittsburgh Sleep Quality Index scores improved in both groups, but the treatment arm × time interaction was significant (F1,59 = 4.69, P = .035), and the mean change score was larger with MBSR than SME (2.6 [3.6] vs 1.0 [1.9]).

The authors restricted this analysis to participants not taking psychiatric medications because allowed medications such as selective serotonin reuptake inhibitors and benzodiazepines can affect sleep and potentially obscure treatment-related effects.

The primary HAMA outcome improved significantly in both groups, but the between-group difference was not significant. The authors suggested one possible explanation: the HAMA heavily weights somatic symptoms and may have been less sensitive to changes in psychological symptoms. They also noted that MBSR participants may have experienced reduced overall distress and impairment that was better captured by global ratings and self-report measures than by the HAMA.

The intervention was an 8-week group Mindfulness-Based Stress Reduction program with weekly 2-hour classes, a single 4-hour weekend retreat, and 20 minutes of daily home practice guided by audio recordings. In-class practices included breath awareness, body-scan meditation, and gentle Hatha yoga, along with instruction in informal mindfulness practice during daily activities.

For this study protocol, the standard MBSR format was shortened from 2.5-hour classes to 2-hour classes, from a day-long retreat to 4 hours, and from 45 minutes of homework to 20 minutes. Metta (loving-kindness) was also introduced in the first class, and a metta audio recording was included for home practice.

The control group was Stress Management Education, a time- and attention-matched active comparison condition that did not include mindfulness components. It also consisted of 8 weekly 2-hour classes, 20-minute homework exercises, and a 4-hour weekend session, with the total number of class and home-practice minutes matched to MBSR.

SME covered topics such as stress physiology, time management, sleep, nutrition, caffeine, exercise, stress hardiness, humor, altruism, and volunteering. It also included gentle strength and posture exercises to more closely match the yoga portion of MBSR.

The study enrolled adults aged 18 years or older with DSM-IV current primary generalized anxiety disorder who identified GAD as their primary problem and had a Hamilton Anxiety Rating Scale score of 20 or higher. Ninety-three individuals were randomly assigned, and the modified intent-to-treat analysis included 89 participants who attended at least 1 class: 48 in MBSR and 41 in SME.

Major exclusions included psychotic disorders, bipolar disorder, posttraumatic stress disorder, obsessive-compulsive disorder, recent substance abuse or dependence, significant suicidality, concurrent psychotherapy directed toward GAD, substantial recent meditation training, pregnancy or lactation, serious medical instability, and likely interfering personality disorder.

Adverse events were uncommon and occurred at the same rate in both groups. The reported adverse event rate was 2% for MBSR, with 1 case of muscle soreness, and 2% for SME, with 1 case of sleep disruption.

The authors identified several limitations. The sample was relatively small, so larger replications are needed. They also noted that comorbidity rates in the sample may limit generalizability compared with large epidemiologic GAD samples, that some participants were taking psychiatric medications but numbers were too small to examine medication effects on response, and that the study did not include a clinical diagnostic assessment at endpoint.