Community-dwelling elderly subjects (n=48) participated in this study. The subjects were recruited through an advertisement in the local press. The following selection criteria were used: age ≥65 years, community dwelling, independent ambulation, willingness to participate in the measurement of physical fitness, and minimal hearing and vision impairment. Exclusion criteria were as follows: inability to complete the tasks because of reduced cognitive functions, evaluated by Rapid Dementia Screening Test [14] scored 8 or greater; severe cardiac, pulmonary, or musculoskeletal disorders; pathologies associated with increased risk of falls, such as Parkinson’s disease or stroke; osteoporosis; and psychotropic drug use. We obtained written informed consent from each participant. This study was approved (protocol approval E- 880) by the Ethical Review Board of Kyoto University Graduate School of Medicine, Kyoto, Japan.

The DTTC test [8-10] requires users to solve a number placement problem (Sudoku) by controlling a stick figure with the movement of their entire body. The user’s full-body motion is captured using Kinect and is translated into movements for the stick figure on a screen. The cognitive task is to fill in 3 boxes chosen at random from a 4 × 4 grid with digits ranging from 1 to 4. The user selects a digit using his or her right hand and left foot and points to a box with his or her left hand. In addition, the user must move his or her right hand to the left hand to fill the indicated box with the selected digit. As such, full-body motion, similar to Tai Chi Chuan movements, is required. We recorded the time taken to fill in all 3 boxes, as our evaluation index.

To begin with, the user stands 3 m in front of the Kinect sensor with his or her right foot in front of the sensor (Figure 1). The following instructions were provided:

All subjects participated in group training sessions lasting 75-80 minutes once a week for 12 weeks. The participants were divided into two training groups according to their participation in the exercise class: (1) the control group (CG), 75-minute standardized training, and (2) the training group (TG), 5-minute DTTC training, in addition to the standardized training [10].

The exercise classes were individualized for each group and were supervised by physical therapists. Each exercise class used a standardized format that included 15 minutes of moderate-intensity aerobic exercise, 15 minutes of progressive strength training, 10 minutes of flexibility and balance exercises, and 10 minutes of cool-down activities, followed by exercises known to improve muscle strength and balance [15,16]. In addition, the class included a 25-minute rhythmic stepping exercise involving cognitive ability, which is an exercise intended to improve DT ability [11], before cool-down.

The participants in the TG were additionally asked to solve DTTC problems and mirror-reversed DTTC problems, alternately, as many times possible in 5 minutes.

All participants underwent evaluation upon entry into the study (pre-intervention) and at the end of the study (post-intervention), using the results of 2 cognitive performance tests.

Cognitive functions were evaluated using the trail-making test (TMT) [17] and verbal fluency test (VFT) [18-20]. The TMT is a well-established psychomotor test originally developed as part of the Army Individual Test Battery. The TMT has been widely used in clinical evaluations to assess deficits in executive cognitive functions. The test consists of 2 parts: part A is a visual-scanning task and part B is a measure of cognitive flexibility. For this analysis, we used a different score defined as delta-TMT, calculated as the difference between the times for each part (part B-part A). The delta-TMT score is used to control for the effect of motor speed on TMT performance and is considered a more accurate measure of executive functions than performance on part B alone [20,21].

The VFT has a letter fluency component and a category fluency component. Participants were asked to think of as many animal names as possible in 1 minute (category fluency). Verbal fluency is an evaluation of expressive language ability and executive functions. The score was the number of successful words (except for some proper nouns).

We compared baseline characteristics between the participants in each group to examine the comparability between the 2 groups using Student’s t test or the chi-square test. Repeated-measures, mixed-linear, two-way ANCOVA (analysis of covariance) was used to analyze the effect of exercise on outcome measurements while adjusting for each cognitive performance, at pre-intervention, as a covariate.

Data were entered and analyzed using the Statistical Package for the Social Sciences (Windows version 20.0, SPSS Inc., Chicago, IL, USA). For all analyses, P<.05 was considered statistically significant.

A total of 41 of the 48 selected subjects (85.4%) completed the study protocols and returned for their exit interviews and final testing (TG: n=26, CG: n=15). The participants’ baseline data did not differ significantly between the two groups. Thus, the groups were comparable and well matched with regard to their baseline characteristics.

During the 12-week intervention phase, 10 exercise sessions were scheduled and all took place. Excluding the 4 participants who dropped out, the TG subjects had an overall attendance rate of 82% and the CG subjects had an overall attendance rate of 81% over the 12 weeks. No health problems, including cardiovascular or musculoskeletal complications, occurred during training sessions or testing. Moreover, almost all participants seemed to have enjoyed the DTTC training. They shared many positive opinions after each session and seemed to look forward to playing DTTC once a week.

Pre- and post-intervention group statistics and group × time interactions are shown in Table 1. There was a significant difference between the groups regarding the changes (intervention to baseline) in delta-TMT. There were no significant differences among the other measures.

The delta-TMT score was significantly improved after DTTC training. The results suggested that DTTC training was effective in improving executive cognitive functions. In our unpublished data, DTTC training was useful for improving balance ability and mobility among elderly people [10]. Thus, DTTC training has the capability of improving both physical and cognitive functions. Executive cognitive functions are closely related to DT performance and are good predictors of falling [20]. Thus, an improvement in executive functions, by DTTC training, has a positive impact on DT ability and the prevention of falling.

Conversely, the TMT-A and -B scores were not significantly improved. In the TMT-A, both groups improved similarly. On the other hand, in the TMT-B, only the TG had a tendency to improve the score, while CG showed little change. That is why the score of delta-TMT in CG increased and in TG decreased. The TMT-A and -B are used to assess visual scanning, cognitive flexibility, and executive functions [17]. The delta-TMT score is considered a more accurate measurement of executive functions [21,22]. That is, the results reflect specific improvement in executive cognitive functions. The reason is that DTTC training involves executive tasks that are the control of goal-oriented actions and adaptive behaviors. Additionally, VFT was also not significantly improved after DTTC training. It includes recalling tasks as their main elements; however, DTTC training does not involve recalling the task. This is considered the reason why the result of VFT was not significantly changed.

Previous studies reported that other “exergames,” based on Nintendo Wii Fit (Nintendo, Kyoto, Japan) or other devices, improved balance and leg muscle functions [23,24]. However, in these cases, the user needs control devices to capture motions and longer and more frequent exercise is required, compared with DTTC training, to obtain physical effects. DTTC overcomes these weaknesses and can offer comparable cognitive and physical benefits to users. Other previous studies have shown that a home-based dance device, using videogame technology, also improved the physical and cognitive parameters of fall risk in elderly people [25]. These home-based training tools are expected to increase the improvement of various functions and lead to fall prevention; our DTTC device is similar to these tools.

There are several limitations to this study. First, the intervention effects in this study were not due solely to the DTTC training. That means the rhythmic stepping exercise involving cognitive ability in both groups also improves cognitive ability. Second, participants in both groups may have had higher motivation and interest in health issues and fall risk minimization than the general elderly population. An investigation into the effects of DTTC training on functions in frail, elderly adults is necessary, in the future.

In this study, the ANCOVA results of delta-TMT reveal that group × time interactions were statistically significant. They suggest that DTTC training is effective at improving executive cognitive functions in particular.