The Ministry of Health, Labour and Welfare estimates that 15% of the Japanese population aged >65 years have dementia, i.e., a total of approximately 4.62 million individuals (estimated in 2012) ¹ . Alzheimer's disease (AD) is the most common type of dementia and accounts for 60% or more of dementia cases; vascular dementia is the second most prevalent type of dementia, accounting for 15–20 % of cases ² . However, many individuals have a mixed dementia type, with lesions characteristic of both AD and vascular dementia ³ . AD has increased year-by-year ⁴ , and among adults >65 years of age, the incidence rate doubles with every 5 years of increased age ⁵ . This means that the construction of countermeasures is an urgent issue. For instance, although the amyloid vaccine was developed in 2000, it could not suppress the decline in cognitive function, even after amyloid β is removed from the brain ⁶ .

Since taking preventive measures before the onset of dementia is important, a worldwide project, the Alzheimer's Disease Neuroimaging Initiative (ADNI; http://www.adni-info.org/), has been implemented to establish a test that can indicate AD before onset, and therefore offer an opportunity for intervention to stop the progression of disease ⁷ . J-ADNI was also launched in Japan in 2007 ( http://www.alz.org/research/funding/partnerships/WW-ADNI_japan.asp), though regional efforts in Japan are still at an early stage.

With this in mind, we started a brain training preventive intervention for AD in Kashihara City (Nara, Japan), at all 11 of the city’s public halls. This initiative is a joint project between Nara Medical University and the dementia prevention project of the Kashihara City Council of Social Welfare, and screens individuals for Mild Cognitive Impairment (MCI), with follow-up for those meeting criteria for an MCI diagnosis. Screening is conducted twice a year, using the Montreal Cognitive Assessment (MoCA), a screening scale for MCI ( http://www.mocatest.org/). Public health nurses conduct follow-up visits to participants with low scores. In addition, a brain training class is held once a month, and prevention interventions and cognitive function evaluations are continuously conducted.

The monthly brain training class aims to inform participants how to take precautions against risk factors for AD, and transform their lifestyle. Risk factors for AD that are targeted comprise of the following: high blood pressure, obesity, smoking, dyslipidemia, and lifestyle-related diseases, such as diabetes, and complications of these ⁸ . It is reported that the probability of developing AD in the future is 2 times as high in people with hypertension, 2.1 times in those who are obese (with a BMI of 30 or more), 1.8 times in those who smoke, 2.9 times in those with dyslipidemia (total cholesterol 250mg/dl or more), and 4.6 times in people with diabetes (HbA1c, 7% or higher) ⁹ . Furthermore, in the brains of people with AD, there is an increased quantity of oxidatively modified products ¹⁰ . Therefore, improving diet is an important part of disease prevention.

Aerobic exercise is also essential in preventing AD. For instance, it has been reported that participation in continuous aerobic exercise is linked to increases in brain derived neurotrophic factor (BDNF) and increases in the capacity of the hippocampus ¹¹ . In addition, brain training has shown some positive effects on the brain: the n-back task (a test of memory retention, requiring the participant to identify the item occupying the nth-back position in a sequence of items) has been validated as an effective brain training task, and a meta-analysis indicates that this task is associated with activation of frontal and parietal cortex ¹² . Furthermore, it has been reported that, compared to a single-task (such as exercise or learning only), a dual-task (requiring the participant to perform two activities at the same time) generates more activation in the brain, in particular in the prefrontal cortex ^{13, 14} .

Drawing on the above-mentioned previous studies, in this study we considered that an intervention combining healthy-eating habit guidance, aerobic exercise, an n-back task and a dual-task might be effective. In addition, the majority of non-drug therapies, such as music therapy, for the maintenance of cognitive function promote feelings of comfort. These feelings of comfort activate reward systems in the brain, motivating the individual to continue the task in which they are engaged. For instance, in a study comparing the effects of negative and positive feeling, positive feeling increased an individual’s repertoire of thinking, behavior and attention, whereas negative feeling reduced an individual’s repertoire of thinking and behavior ¹⁵ . In addition, it is reported that improvement in self-efficacy is related to memory improvement ¹⁶ . Therefore, we thought it was important that the intervention assessed herein improved positive mood and feelings of comfort; therefore, we decided to use physical recreation in this intervention. The full intervention consisted of diet guidance, recreation, exercise, an n-back task, and a dual-task. The purpose of this study was to measure the effect of this intervention on improving cognitive function, and to clarify whether there is a correlation between cognitive function and positive mood.

The intervention had a 100% completion rate. The average age of the subjects was 72.3 (± 6.6) years old. The group consisted of 129 men and 253 women. The MoCA scores for each age category (50s, 60s, 70s, and 80s) are shown in Figure 1. This was measured before intervention. The correlation coefficients between age and each cognitive function at the beginning of the intervention are also shown in Figure 1. These are classified by MoCA Assessment items.

As shown in Figure 1, the cognitive function with the strongest negative correlation with age at the start of the intervention was the short-term memory recall task; the MoCA score rapidly decreased in participants in their 70s and 80s compared to participants in their 50s and 60s (r = -0.56). Other tests whose results were negatively correlated with age were the trail making task and the clock-task (together indicating visuospatial cognitive ability, r = -0.38, -0.49), verbal fluency (memory retrieval ability, r = -0.46), attention, concentration and working memory (ability to concentrate and attention and memory, r = 0.53), repetition task (memory, r = -0.49), abstract thinking (r = -0.31), and orientation (r = -0.34). Cognitive functions that were maintained with increasing age were visuoconstructional skills (cube: graphic replication) and naming (animal name recall).

With an intervention once a month, an evaluation can be seen before and six months after the intervention in Table 1. This is a comparison of the results of all ages. The average of the total scores before the intervention was >26 points in both the dual-task group and the single-task group; this met the cut-off value for MCI (26 points). After the intervention, there was significant improvement in both the dual-task group and single-task group, and the average value on the MoCA was above the cutoff value for MCI (p < 0.01).

Comparing the results of the dual-task group and the single-task group, only the dual-task group showed a significant improvement in the trail making and the cube drawing tests (visual-spatial cognitive abilities), and in abstract thinking, speaking in order, speaking in reverse, sustained attention, and calculation (ability to concentrate & attention & memory) (p < 0.01). There were significant improvements in both the dual-task and single-task groups in the verbal fluency, the repetition task (memory), delayed recall task (memory playback capability), and in the overall score on the MoCA (the presence or absence of MCI) (p < 0.05).

Regarding collection of salivary amylase, was taken at a briefing session before the intervention. At that time, the subjects, after receiving the MoCA test, remained in the venue and collected saliva. As it takes time to collect individual saliva and measure alpha amylase, only subjects with time remained in the venue. In addition, there were subjects who were unable to measure α-amylase (measurement error) subject to influence of hypertensive internal medicine and others. For this reason, the number of subjects that were able to collect the salivary α-amylase, is 280 people. In the measurement of salivary α-amylase, the minimum value was 2 KU/L and the maximum value was 216 KU/L (mean, 49.7 ± 47.0). The correlation of salivary α-amylase and MoCA total score was negative (r = -0.31).

Therefore, there was a trend for MoCA scores to be lower in individuals experiencing greater stress ( Figure 2).

Prophylactic interventions of AD, which were carried out in each municipality, are still at the stage of trial and error ( http://www.mhlw.go.jp/file/06-Seisakujouhou-12300000-Roukenkyoku/0000136616.pdf). As shown in this study, dementia preventive measures in Japan, there is a disparity of each municipality. A preventive program for dementia has not been established yet. It can be noted that general cognitive function, which decreased along with age, can be dissociated from specific cognitive functions, some of which were relatively maintained with increasing age. The specific cognitive function that had the strongest negative correlation with age was the delayed recall task, which requires the individual to memorize five nouns and to recall them after about five minutes. The brain’s ability to memorize new things, maintain them, and reproduce them rapidly decreased in proportion to age. Performance on the trail making and clock-drawing tasks (both measuring visuospatial cognitive ability) also rapidly decreased with age. Visuospatial cognition refers to the brain’s ability to process visual information. When this ability declines, people tend to get lost ¹⁸ . There were also correlations between age and word recall (thinking ability), speaking in order, speaking in reverse, sustained attention and calculation (ability to concentrate & attention & memory), and the repetition task (memory). Decreased performance in the word recall task indicates that an individual may have trouble recalling words during a conversation. Decreasing ability to concentrate affects safety and the continuity of actions; when attention is impaired, people struggle to maintain attention on stimuli, so their actions become distracted ¹⁹ . Memory was assessed in this study through tasks requiring the participants to speak in order, to speak in reverse, and to listen to and reproduce sentences. A decline in memory due to aging (age-associated memory impairment) is caused by a decline in the function of cranial nerve ²⁰ and a failure in the network of the brain ²¹ . Disruption to the cognitive functions considered above adversely affects daily life and lowers safety. Therefore, reducing the risks associated with cognitive decline is an important issue. By contrast, some cognitive functions were maintained even with increasing age, as indicated by performance on the tasks requiring shape replication, animal name recall. These tasks require the reproduction of familiar forms, learned from a young age, and are not tasks in which the participant memorizes new things. Abstract thinking is also a common task of the situations that require common knowledge. In tasks based on familiar stimuli, cognitive function is maintained with increasing age. Due to this, familiar experiences are easier to retrieve ²² .

Comparing scores before and after the intervention, the cognitive tasks that improved significantly in the course of the invention were the trail making and clock drawing tests (visual-spatial cognitive ability), abstract thinking, speaking in order, speaking in reverse, sustained attention and calculation (ability to concentrate & attention & memory), the repetition task (memory), the delayed recall task (memory playback capability). All these cognitive functions decrease with age, but here showed an improvement throughout the intervention; importantly, these functions are associated with the ability of an older person to keep safe and secure in their everyday life.

On comparing the results of the dual-task group and the single-task group, a greater number of cognitive functions showed significant improvement in the dual-task group. Comparing the results after intervention, visuospatial cognition, abstract thinking, concentration, attention, and memory, only those within the dual-task group showed significant improvement. To run two tasks at the same time, the frontal lobe, and in particular the most anterior region, the prefrontal cortex, is essential ²³ . Based on this understanding, the dual-task is considered to train the frontal lobe, and this view has been supported by neuroscience studies that show frontal activation during the dual-task ^{24, 25} . The prefrontal cortex ages earlier than other brain regions ²⁶ , meaning that cognitive training for the elderly needs to be considered a priority on maintaining the function on the prefrontal cortex. Interestingly, it has been found that the number of neural stem cells, which are needed for the regeneration of nerve cells, is maintained, even with increasing age ^{27, 28} . Thus, it is reasonable to expect that cognitive training can maintain cognitive function by drawing on these neural stem cells.

The n-back task, first introduced by Wayne Kirchner in 1958 ²⁹ , measures the capacity of an individual’s working memory. It has been demonstrated that the n-back task not only measures working memory, but also can improve it when structured in a brain training intervention. Improvements in fluid intelligence ³⁰ , and increased density of dopamine ³¹ have been demonstrated. A synergistic effect in combination with dual-task can be expected.

In addition, the present intervention improved positive mood, presumably activating the brain’s reward system. An increase in positive feelings is reported to have effects on body and mind, such as increasing satisfaction and success ³² , improving immune function ³³ , increasing confidence towards others and fostering closer relationships ³⁴ , positively affecting physical and mental health ³⁵ , and speeding up recovery from disease ³⁶ . Thus, by targeting an improved positive mood, other positive effects might be expected. In the present study, a correlation was found between salivary α-amylase, which reflects mood, and cognitive function; when negative stress was high, the MoCA was low. Our research group, in a study in the last year, demonstrated a correlation between the ability to cope with stress (Sense of Coherence SOC) and cognitive functions ³⁷ , and the current study supports this finding.

An intervention, combining exercise, an n-back task, and a dual-task, improved performance in a greater number of areas of cognitive function, compared with a similar intervention in which a single learning task was substituted for the dual-task. The cognitive functions that decreased most with increasing age were delayed recall, visuospatial cognition, thinking, concentration, attention, and memory. The cognitive tasks that had no correlation with age, and were maintained even with age, were graphic replication, animal name recall, abstract thinking, and orientation – all of which require the reproduction of familiar forms or names. The results of this study show that it is possible to improve cognition by a structured intervention. In addition, a correlation between cognitive function and positive mood has been demonstrated by the present study. It would be interesting to investigate whether improvement in positive feeling directly improves the effectiveness of brain training.

The data referenced by this article are under copyright with the following copyright statement: Copyright: � 2017 Sawami K et al.

Data associated with the article are available under the terms of the Creative Commons Zero "No rights reserved" data waiver (CC0 1.0 Public domain dedication).

Dataset 1. MoCA scores and age. Sheet 1 is the data before age and MoCA test intervention and after intervention. Sheet 2 is the data before and after the intervention divided into the dual-task group and the single-task group. doi, 10.5256/f1000research.10584.d150825 ³⁸

Dataset 2. MoCA and salivary amylase. This is the data showing the score of the MoCA test before the intervention and the measured value of salivary α-amylase. doi, 10.5256/f1000research.10584.d150826 ³⁹