1. Study Design: This will be an open-label, randomized control trial of concomitant massage therapy and escitalopram pharmacotherapy for the treatment of major depression in the elderly. All subjects who meet inclusion/exclusion criteria will receive escitalopram and also be randomized 1:1 (by computer) to Swedish massage or light touch. Escitalopram (5.0 mg/day, PO) will initially be administered for one week followed by an increase in dose to 10 mg/day for the remainder of the study; Massage or touch will be performed twice per week for 8 weeks. Every effort will be made to have non-consecutive massage/touch visits. The massage group will receive Swedish massage for 50 minutes twice per week. In the "touch" group, the massage therapist will provide gentle touch to the skin in the same distribution as that of the formal Swedish massage for 50 minutes twice per week. Every effort will be made to have one therapist perform all treatments. All assessments will be obtained prior to the first massage each week.
2. Recruitment and Retention: Subjects will be recruited from our outpatient clinic, attendings, as well as by advertising on radio, television, newspaper and our web site.
3. Treatment
1. Massage: Standard Swedish massage therapy will be employed which includes the systematic manipulation of the soft tissues of the body. It is designed to relax muscles by applying pressure to them against deeper muscles and bones, and rubbing in the same direction as the flow of blood returning to the heart. A routine has been designed whereby the therapist massages different areas of the body in an orderly fashion for specified periods of time. For sham massage (light touch), the massage therapists will provide gentle touch to the skin in the same distribution and duration as that of the formal Swedish massage. As mentioned below, deviations from the structured massage or touch will be recorded in the massage therapist's diary.
2. Quality control for screening and outcome questionnaires: Kappa co-efficients for diagnostic and inter-rater reliability will be determined at six-month intervals. We routinely maintain kappa values of at \>0.85.
4. Outcome Measures: Primary outcome measure for this study will be degree of depression as measured by HAM-D. The secondary outcome measures will include quality of life, HAM-A, Beck Depression Inventory. Raters will be blind to treatment cell and test results.
Quality control for screening and outcome questionnaires: Kappa co-efficients for diagnostic and inter-rater reliability will be determined at six-month intervals. We routinely maintain kappa values of at \>0.85.
5. Statistical Analysis All entries will be blind-verified, and following data entry, results will be checked for obvious outliers and inconsistent values. No single person will have the responsibility for both entry and verification on a given form. Data on computer records will be identifiable by a unique coded identifier to permit matching of records longitudinally. Each record will be logged as it is obtained. An archival record of all data collected, which has passed the above-noted checks, will be maintained on a hard disk. A backup of the archival data will be saved on magnetic tape (DAT tape) and writable CD-ROM archive disks. This copy will be stored in a separate offsite cabinet to ensure the survival of data in case of a natural disaster. A full backup will be made weekly, and stored for at least three months. A CD-ROM backup will be made every three months.
