Use of Alpha-Stim Cranial-electrotherapy Stimulation (CES) in the Treatment of Anxiety

The goal of this study is to evaluate the efficacy of Alpha-Stim technology in the treatment of anxiety by using a double-blind clinical trial over a period of five weeks. Cranial-electrotherapy stimulation (CES) is a noninvasive procedure that has been used for decades in the United States to treat anxiety, depression, and insomnia. Although many studies on CES have been published in previous decades, most have used relatively small samples using various frequencies and links of treatment. Given the positive results of the many general studies that have been performed to date, this study seeks to add to the current literature by addressing previous criticisms using a large sample size (n=150)and using participants that represent more than a single type of anxiety along with comorbidity. Such a population is believed to better represent typical treatment complications.

Participants will be recruited via advertisements in a local newspaper and posting of flyers around a local university campus. Participants will be screened through an intake process using the Structured Clinical Interview for Axis I Disorders (SCID-I) to confirm a diagnosis of an anxiety disorder. Comorbid conditions are acceptable, however, an anxiety disorder must be the primary diagnosis. As part of the intake process, participants will complete the Hamilton Anxiety Rating Scale (HAM-A)and the Hamilton Depression Rating Scale (HAM-D17) to establish a baseline. These instruments will serve as the primary efficacy measures. In order to participate, HAM-A scores must equal to or greater than 18 (Moderate Anxiety). Depression will be monitored using The HAM-D17 as depression is a commonly occurring condition with anxiety and the HAM-D17 is also sensitive to anxiety. Current treatment with medications are also acceptable provided there will be no changes in dosage or type during the study. See Inclusions and Exclusions for further information. All participants will pay a fee of $30 to enter the study to cover administration costs and as a way to measure placebo effect. Participants will be assigned a device to take home and self-administer treatment 5-7 days a week. Participants will be required to log their treatment times and dates with logs provided by the researcher. The student population will not be given a take home device but will receive treatment at the Graduate Assistant's Office or Student Care Office.

Electromedical Products International, Inc. has agreed to supply the devices needed to carry out the research to include conductance solution, felt tabs, and batteries. Half of the devices will be set by the manufacturer to deliver treatment at 0.5 Hz and at the lowest therapeutic setting. The remaining devices will be shame and not deliver any therapeutic benefit. The devices will not be able to be altered by the participant.

Participants will meet with the PI or research assistant to complete the HAM-A and HAM-D17 at weeks 1, 3, and 5 at a designated location either in the researchers private practice office or university lab location. Adverse treatment reactions will also be noted. Treatment logs will also be collected at these points. At weeks 1 and 3, participants will be given fresh batteries, conductance solution, and felt tabs as needed. At the completion of the study, participants found to be in the shame treatment group will be offered the normal course of treatment. Participants not wanting to continue the study or do not wish to follow through with normal treatment after being in the shame group will be refunded their fee.

The hypothesis this study seeks to address is:

H°: There is no difference between sham treatment and active CES on level of anxiety as compared to scores on the HAM-A and HAM-D17.

H¹ : The active CES treatment group has significantly lower anxiety scores on the HAM-A and HAM-D17 than the sham group at the endpoint of the study.

Statistical analysis will compare baselines scores (week 1) to midpoint scores (week, 3), and scores at the endpoint of the study (week 5). Response to treatment will be defined as an improvement of 30% or greater on the HAM-A and HAM-D17 at the endpoint of the study. The impact of prescription medication as a confounding variable in relation to the effect of CES on anxiety and depression will also be examined. Data analysis will consist of a repeated measures design that will likely use a general linear model in SPSS and hierarchical linear and quadratic growth models to assess individual change. The reason for this approach is that hierarchical analysis of individual change is advantageous when there are multiple repeated measure data patterns because it affords the research an opportunity to combine into a single analysis (Raudenbush \& Byrk, 2002). We will estimate both the individual growth and mean growth for each point in the time series. Cross-validation of quadratic will occur via the estimation of a piecewise linear growth models to assess whether growth is more variable for one period over another during the study time frame.

Optimal design software has been used to make a preliminary determination that 100 or more participants in a randomized quadratic growth individual change model will achieve a power of .80 to .95 for a study expecting an effect size of between 0.6 to .80, respectively (0.62 is premised upon meta-analysis for CES undertaken by Klawansky, et al., 1995 and .8 is considered a large effect by Cohen, 1988). We will employ power analysis again once sample size is confirmed following recruitment of participants and expect upwards of 150 or more participants. Missing data will be explained using an intent-to-treat analysis whereby we employ redundant analysis procedures (i.e., survival analysis) to understand and context the true treatment effect.