Yang et al.BMC Complementary and Alternative Medicine (2015)15:93 Page 3 of 8 and process the electrocardiogram(ECG)signals.The low CMAI scores were significantly higher compared with frequency percentage (LF%)and the low frequency-high the control group.Over time,the scores in the aroma- frequency (LF/HF)ratio were used to represent sympa- acupressure group significantly declined,with the post- thetic nervous activity,and the high frequency (HF)was test score lower compared with the post-three week used to represent parasympathetic nervous activity. score.A similar decline was also observed in the aroma- therapy group,although there was little difference be- Study procedure tween the two scoring times.Accordingly,both the In the pre-test,the research assistant was responsible for aroma-acupressure and aromatherapy groups experi- completing the CMAI and the HRV.The HRV was used enced improvement in agitation. to measure time in the pre-test and every week during Regarding the results of the HRV test (see Table 3)for the four-week interventions following the pre-test.Fol- the sympathetic nervous system,the LF/HF in the lowing the completion of the interventions,the data aroma-acupressure group was significantly higher com- from the post-test and post-three week assessments were pared with the control group.Over time,the LF%and collected.In order to control the environmental factors LF/HF in the aroma-acupressure group did not exhibit a that might affect HRV results,all participants were mea- weekly difference.However,a group interaction with time sured after lunch break in their own rooms. was present,which was demonstrated by the weekly de- cline in LF%every week until the fourth week.A similar Statistical analyses result occurred in the aromatherapy group,and there was All statistical analyses were performed using the software a significant difference in the second week.In addition, SPSS 20.0(IBM Corporation,Armont,NY,USA).Descrip-the weekly LF/HF was significantly reduced each week tive statistics were used to describe the characteristics of until the fourth week;however,this declining trend was the participants.The three groups were compared using not present in the aromatherapy group. one-way ANOVA and chi-square tests.A generalized For the parasympathetic nervous system,there were estimating equation (GEE)for repeated measurements no significant differences between the three groups or was used to assess the outcome indicators.Based on the assessment times.In the group interaction with time, intention-to-treat analyses,which included subjects with the HF in the aroma-acupressure group was significantly missing data points,we used the missing-at-random as- higher in the second,third,and fourth weeks,whereas sumption to conduct the GEE analysis. the HF was significantly higher only in the fourth week for the aromatherapy group.The effects of aroma- Results acupressure on the parasympathetic nervous system Of the 276 qualified participants,21 participants were were also stronger over time. hospitalized,1 participant was deceased,and 68 partici- The participants did not experience side effects at any pants chose not to participate in the study.Thus,186 stage of the interventions. participants were included in the study and were randomly assigned to the three groups.Figure 1 depicts Discussion the flow of the participants through each stage.The The CMAl score was significantly higher in the aroma- demographic data of the three groups are shown in acupressure and aromatherapy groups compared with Table 1.The average age was 85.3 years in the aroma- the control group in the pre-and post-tests;the average acupressure group,83.67 years in the aromatherapy scores in the pre-and post-tests were highest in the group,and 81.56 years in the control group,with signifi- aroma-acupressure group,followed by the aromatherapy cant age differences between the 3 groups.However, group and the control group.Prior to the interventions, there were no significant differences among the groups agitation was significantly more severe in the aroma- in regards to sex.Regarding the types of dementia acupressure and aromatherapy groups compared with present,the majority of the participants in the three the control group,and the CMAI score was significantly groups were AD patients.Two types of restraints were higher in the post-test and post-three-week assessments used:a physical constraint that involved the use of a re- compared with the pre-test.In the aroma-acupressure straint belt,and a chemical restraint that involved the and aromatherapy groups,the pre-test CMAI score use of antipsychotic drugs.There were no significant dif was significantly higher compared with the post-test ferences in the use of restraints among the three groups. and post-three-week scores,which indicates that both aroma-acupressure and aromatherapy can immediately Outcome and estimation and persistently improve agitation.This finding is con- The comparisons of the CMAI among the groups are sistent with previous studies demonstrating that both demonstrated in Table 2.In both the aroma-acupressure aroma-acupressure and aromatherapy can improve agi- and aromatherapy groups in the pre-and post-tests,the tation [6,7].These effects may persist for three weeksand process the electrocardiogram (ECG) signals. The low frequency percentage (LF%) and the low frequency-high frequency (LF/HF) ratio were used to represent sympa￾thetic nervous activity, and the high frequency (HF) was used to represent parasympathetic nervous activity. Study procedure In the pre-test, the research assistant was responsible for completing the CMAI and the HRV. The HRV was used to measure time in the pre-test and every week during the four-week interventions following the pre-test. Fol￾lowing the completion of the interventions, the data from the post-test and post-three week assessments were collected. In order to control the environmental factors that might affect HRV results, all participants were mea￾sured after lunch break in their own rooms. Statistical analyses All statistical analyses were performed using the software SPSS 20.0 (IBM Corporation, Armont, NY, USA). Descrip￾tive statistics were used to describe the characteristics of the participants. The three groups were compared using one-way ANOVA and chi-square tests. A generalized estimating equation (GEE) for repeated measurements was used to assess the outcome indicators. Based on intention-to-treat analyses, which included subjects with missing data points, we used the missing-at-random as￾sumption to conduct the GEE analysis. Results Of the 276 qualified participants, 21 participants were hospitalized, 1 participant was deceased, and 68 partici￾pants chose not to participate in the study. Thus, 186 participants were included in the study and were randomly assigned to the three groups. Figure 1 depicts the flow of the participants through each stage. The demographic data of the three groups are shown in Table 1. The average age was 85.3 years in the aroma￾acupressure group, 83.67 years in the aromatherapy group, and 81.56 years in the control group, with signifi￾cant age differences between the 3 groups. However, there were no significant differences among the groups in regards to sex. Regarding the types of dementia present, the majority of the participants in the three groups were AD patients. Two types of restraints were used: a physical constraint that involved the use of a re￾straint belt, and a chemical restraint that involved the use of antipsychotic drugs. There were no significant dif￾ferences in the use of restraints among the three groups. Outcome and estimation The comparisons of the CMAI among the groups are demonstrated in Table 2. In both the aroma-acupressure and aromatherapy groups in the pre- and post-tests, the CMAI scores were significantly higher compared with the control group. Over time, the scores in the aroma￾acupressure group significantly declined, with the post￾test score lower compared with the post-three week score. A similar decline was also observed in the aroma￾therapy group, although there was little difference be￾tween the two scoring times. Accordingly, both the aroma-acupressure and aromatherapy groups experi￾enced improvement in agitation. Regarding the results of the HRV test (see Table 3) for the sympathetic nervous system, the LF/HF in the aroma-acupressure group was significantly higher com￾pared with the control group. Over time, the LF% and LF/HF in the aroma-acupressure group did not exhibit a weekly difference. However, a group interaction with time was present, which was demonstrated by the weekly de￾cline in LF% every week until the fourth week. A similar result occurred in the aromatherapy group, and there was a significant difference in the second week. In addition, the weekly LF/HF was significantly reduced each week until the fourth week; however, this declining trend was not present in the aromatherapy group. For the parasympathetic nervous system, there were no significant differences between the three groups or the assessment times. In the group interaction with time, the HF in the aroma-acupressure group was significantly higher in the second, third, and fourth weeks, whereas the HF was significantly higher only in the fourth week for the aromatherapy group. The effects of aroma￾acupressure on the parasympathetic nervous system were also stronger over time. The participants did not experience side effects at any stage of the interventions. Discussion The CMAI score was significantly higher in the aroma￾acupressure and aromatherapy groups compared with the control group in the pre- and post-tests; the average scores in the pre- and post-tests were highest in the aroma-acupressure group, followed by the aromatherapy group and the control group. Prior to the interventions, agitation was significantly more severe in the aroma￾acupressure and aromatherapy groups compared with the control group, and the CMAI score was significantly higher in the post-test and post-three-week assessments compared with the pre-test. In the aroma-acupressure and aromatherapy groups, the pre-test CMAI score was significantly higher compared with the post-test and post-three-week scores, which indicates that both aroma-acupressure and aromatherapy can immediately and persistently improve agitation. This finding is con￾sistent with previous studies demonstrating that both aroma-acupressure and aromatherapy can improve agi￾tation [6,7]. These effects may persist for three weeks Yang et al. BMC Complementary and Alternative Medicine (2015) 15:93 Page 3 of 8