items (0-9); (d) aggressiveness, 3 items (0-9); (e) diurnal rhythm disturbances, 1 item (0-3); (f) affective disturbance, 2 items (0-6); (g) anxieties and phobias, 4 items (0-12); and (h) global rating.19 Higher Behave-AD scores represent worse functioning. The BEHAVE-AD was done at baseline before the 6-week intervention and again after intervention. Nurses who took charge of participants in the intervention and control groups performed the BEHAVE-AD evaluations before and after the intervention. Salivary CgA Physiological stress was evaluated using salivary CgA, an acid glycoprotein (439 amino acid residues) isolated from chromaffin granules in the adrenal medulla. Chromogranin A concentration has a direct relationship with salivary catecholamine; therefore, testing for catecholamine is a precise indicator of CgA levels.20,21 To avoid variations due to circadian rhythms, measurements of salivary CgA were taken at fixed times from both the tactile massage therapy and control group. Salivary CgA was collected from both the intervention and control groups at the same time, between at 16:00 and at 17:00 hours, at baseline (before intervention) and after the 30th (final) tactile massage therapy. Collection occurred immediately before the start of tactile massage (16:00 hours) and immediately after completion of tactile massage (17:00 hours). In the control group, saliva was also collected twice daily at 16:00 and 17:00 hours to establish baseline levels (before the start of tactile massage) and 6 weeks later (the time of the last tactile massage). Saliva samples were absorbed using cylindrical tubes containing cotton (Salivette, Aktiengesellschaft Company, Walldorf, Germany). Saliva was collected using a cyclone separator (3000 rpm) and analyzed using a YK070 Chromogranin A EIA kit (Yanaihara Institute Inc, Shizuoka, Japan). For statistical analysis, paired t test was used to compare each item before and after tactile massage therapy intervention using SPSS Ver.17 (SPSS Inc, Chicago, Illinois). The probability of rejecting the null hypothesis (a) was set at .05 throughout this study. Results Although each patient was scheduled to undergo a total of 30 tactile massage sessions, only 2 underwent all 30 sessions. The mean number of sessions received was 24.64 + 4.92 (range 15-30). The intervention and control groups each contained 20 participants at baseline. To facilitate between-group comparisons, when a participant left the study due to hospital transfer, hospital discharge, poor health condition, or other reasons, the corresponding participant in the other group was also removed from the study. Consequently, by the end of the study, this decreased to 14 participants (4 men and 10 women). Of these 28 participants, 24 had Alzheimer’s disease (12 in the Table 1. The Tactile Care Technique 1.The nurse tells the subject that it is time for tactile massage and to get in a comfortable posture. 2. The nurse wraps both hands in a towel before removing one towel to effleurage the hand and turn the palm up 3. The nurse rubs organic olive oil on the subject after warming the oil in her palm. 4. For finger effleurage, the nurse lightly strokes the side of the hand slowly. 5. The nurse wraps her hand around each finger while performing effleurage in a slow circular motion starting from the base of the finger to the tip of the finger 6. The nurse strokes the hand making small clockwise circles on the palm and puts the hands together before stroking the sides of the fingers. 7. The nurse lets her hands slide around the wrists while performing effleurage in a circular motion. 8. The nurse carefully wraps the subject’s hand in a towel and moves to the next hand. 9. The nurse repeats the same procedure for the other hand until the end of the tactile massage after which she thanks the subject. 682 American Journal of Alzheimer’s Disease & Other Dementias® 25(8) intervention group and 12 in the control group) and 4 had cerebrovascular dementia (2 in the intervention group and 2 in the control group). The mean age of the intervention group was 88.71 (+7.28) years, and the mean age of the control group was 88.00 (+6.63) years. Table 2 shows baseline and postintervention changes in MMSE and BEHAVE-AD scores. For the MMSE, the intervention group showed a slight improvement in scores from baseline 5.29 (+6.19) to postintervention 5.43 (+6.73), whereas the control group showed a slight decline from 6.07 (+5.88) to 5.57 (+5.57); however none of these changes were statistically significant. For the GBS, the control group showed a significant decline in the subscale of (B) intellectual function, from 44.50 (+15.48) to 46.93 (+16.03; P ¼ .04), and the intervention group showed a significant decline on (C) emotional function, from 9.79 (+5.77) to 10.97 (+5.54; P ¼ .03), but there were no significant changes in overall GBS score for the intervention group. The intervention group showed significant improvement on the BEHAVEAD scale for (d) aggressiveness, from 2.50 (+1.21) before the intervention to 1.21 (+1.58) after the intervention (P ¼ .048), whereas no changes were seen in the control group. Figures 1 and 2 show pre- to postintervention changes of the intervention group and the control group in salivary CgA levels. We were able to obtain saliva samples at all 4 times, that is at 16:00 and 17:00 hours (before and after the intervention), at baseline, and after