A 3-arm randomized controlled trial on the effects of dance movement intervention and exercises on elderly with early dementia

McLaren reviewed interventions for dementia involving aerobic, balance, flexibility and resistance exercise and concluded that exercise benefits both functioning and quality of life [10]. More specifically, the systematic review provided clinical trial evidence that non-pharmacologic interventions can delay the progression of functional impairment or disability among dementia patients [10]. One neurobiological study found that the hippocampus, which is responsible for memory and is one of the first areas to degenerate in dementia, increases in volume as both acute and chronic aerobic exercise increases [14]. This indicates that activating the neuro-hormonal mechanisms through exercise can induce cognitive enhancement [14].

Additionally, a controlled study on physical exercises supported by music for moderate to severe dementia patients found improvements in overall cognition, especially information processing and recall [15].

Similarly, dance movement used as a therapeutic intervention combines the physical benefits of exercise with psychosocial therapeutic benefits [16]. Although dance-movement interventions for dementia patients are only mild/moderate forms of physical activity, the level of exertion prompted by a weekly hour of dance among healthy elderly patients was shown to be sufficient to change brain growth factors supporting brain plasticity, similar to the changes induced by exercise [17]. Overall, Foster [17] illustrated that dancing activates the areas in the brain associated with perception, emotions, executive function, memory and motor skills. Specifically, dance and movement improvisation uses frontal lobe resources while movement relies on path integration and visual tracking, which activate the hippocampal and entorhinal networks associated with spatial memory—an area of profound impairment in dementia [17]. In a randomized controlled study using 9 weeks of dance-movement therapy among dementia patients in nursing homes, improved visuospatial abilities were observed and the deterioration of self-care was prevented [18].

The components of dance-movement therapy such as rhythm and tapping target at different cognitive domains. Dance movement affects procedural memory abilities such as motor and cognitive skills, which is highly applicable because motor learning is comparatively well-preserved in dementia [19, 20]. There is growing empirical evidence that moving to rhythm enhances the body’s ability to encode rhythmical movement [21, 22], which is reflected in one’s gait. Certain therapeutic aims such as improving balance among elderly Parkinson’s patients can be achieved by enhancing rhythmical movement—an inherent component of dance [23]. For elderly dementia patients, rhythmic physical activity appears to improve mood and memory [24]. The behavior of elderly residents in nursing care stabilized, with reductions in agitation and wandering upon completion of a psychosocial-focused dance intervention [25]. Emotionally salient and meaningful dance movement can promote both functional and emotional skills and behavior that remain useful to the elderly, even as they progress into the later stages of dementia [26]. Rhythmic movement will be particularly emphasized in the proposed dance-movement intervention for its potential benefit to the elderly.

Psychologically, there is preliminary evidence that dance movement can promote positive changes by facilitating self-expression and communication among elderly dementia patients [25]. More importantly, a non-verbal approach provides a psychotherapeutic medium of expression for patients with dementia when their verbal abilities decline [3]. Group dance, in particular, supports a sense of integration and connection that counteracts the difficulty of orientating oneself in time and space [26]. Additionally, group dance can ease feelings of fear and isolation, as disorientation often leads to a fragmented sense of self [26]. Elderly patients facing gradual memory loss may also benefit from the propensity of dance to evoke reminiscence and restore long-term memories or associated emotions [3].

Salivary cortisol is a neuroendocrine indicator of hypothalamic-pituitary-adrenal (HPA) axis activity, responsible for stress and immune regulation in the body [27]. An elevated cortisol level has recently been found to correlate with rapid progression of dementia [28]. Prolonged heightening of cortisol levels among the elderly is predictive of reduced hippocampal volume, resulting in poor performance in hippocampus-dependent memory tasks, compared with controls [29]. Abnormal cortisol level patterns have more often been found in moderate dementia, with relatively flat cortisol patterns seen in severe dementia [30, 31]. Elderly individuals who are motivated to strengthen their social ties have exhibited healthier cortisol activity across the day [32]. How healthy cortisol activity induced by moderate exercise extends its effect on dementia symptoms has not been properly investigated. The assessment of cortisol can help to elucidate the interrelationship between immunity, psychological health and dementia-related decline.

Although dance-movement interventions are clinically well-established, evidence remains descriptive and the designs of related studies have limitations [25]. Given the increasing aging population, research on non-pharmacological interventions has remained limited. Consequently, a non-pharmacological intervention adopting a holistic approach to improving the overall physical and mental well-being of people with early dementia should be explored. This study aims to examine the effects of a dance movement based intervention for elderly patients with very mild to mild dementia, compared with an exercise intervention and no intervention control conditions.

The primary objective of the proposed study is to explore the effects of a dance movement based intervention on dementia-related indicators such as neuropsychiatric symptoms (behavior, psychosis, ,apathy), psychosocial well-being (depression, loneliness and current affective state), physiological stress (indicated by cortisol), cognitive functioning (executive function, episodic memory, visuospatial construction) and daily functioning. The secondary objectives of this study are to explore the proposed intervention’s effect on dementia prognoses and examine how cortisol levels relate to neuropsychiatric, psychosocial, cognitive and functional measures in early dementia. Through these objectives we aim to provide insight into possible mechanisms of change and interactions among the aforesaid variables after dance-movement intervention.

We hypothesize that upon completion of the dance-movement intervention, elderly dementia patients will experience a greater delay in the progression of dementia severity compared with the exercise and control groups. Furthermore, those who participate in the dance-movement intervention should experience greater resilience against dementia-related decline, compared with the exercise and control groups. We expect to find an association between cortisol levels and neuropsychiatric, cognitive and daily functioning, in addition to psychosocial measures.

This study will adopt a 3-arm RCT design with waitlist control. After screening and collecting baseline data, computer-generated random numbers will be used to randomize the eligible participants into one of three intervention groups: dance movement, exercise or waitlist control. The waitlist control group will be provided with a dance-movement or exercise intervention upon study completion. All of the participants will continue to receive regular medication and care. Four repeated assessments will be conducted for all of the participants at baseline, post intervention (3 months), and 6 and 12 months after the baseline. The manuscript and flow diagram (Fig. 1) adheres to CONSORT guidelines.

Eligible elderly participants (aged 65 and above) with a clinical diagnosis of dementia (DSM IV) or major/mild neurocognitive disorder (DSM V) will be recruited from the psychogeriatric outpatient departments of local hospitals, elderly daycare centers and community centers by their treating psychiatrists or facility staff. Before baseline assessment, a Clinical Dementia Rating (CDR) evaluation will be conducted by the team’s psychiatrist and a trained CDR rater to screen for very mild to mild dementia (CDR 0.5–1).

A sample size of 131 has been determined. Due to the presence of multiple independent variables, multiple regression modeling with a medium effect size (f2) of 0.15, at 0.8 power and a significance level of 0.05, was used to calculate the sample size. The experimental, clinical and demographic variables anticipated to affect the intervention outcomes were included in the model. Based on prior local trials with early dementia patients, the recruitment of 201 participants is required (67 per arm), assuming an attrition rate of about 35 % [33].

The dance-movement intervention will be led by a registered dance-movement therapist or one in training. The program will be modified from a pre-existing dance/movement therapy program applied within various Chinese populations in the past decade with evidenced benefits. The modified elderly program has also been tested in a group of elderly participants and proved both feasible and well-received. The dance-movement intervention program to be used in the proposed study will include gentle body movements with special emphasis on rhythm and contralateral movement on both sides of the body to enhance coordination and stimulation between right and left cerebral hemispheres [21, 22]. Simple group dances will encourage the participants’ ability to remember steps and sequences, movement games will improve mood and vitality and improvisational dance movement will foster imagination, creativity and personal expression [17, 26]. Movement interactions among group members are expected to enhance communication and social exchanges. Finally, vocal sharing about movement experiences should provide articulation opportunities that may help the participants in building mutual support and care. To cater to the elderly, movements will be conducted both in standing and sitting positions. The structure and schedule of each session will be similar, with the content adjusted to suit the needs and dynamics of the group.

The exercise intervention group will receive a mild to moderate exercise program of comparable length and intensity, conducted by an instructor trained by a qualified fitness coach. Each session of the exercise and dance-movement interventions will include a warm up (15 min), stretching and joint movements (15 min), exercising with towels (15 min) and cool down (15 min). Heart rates will be monitored throughout the entire dance-movement or exercise routine using a portable heart rate monitor to maintain a similar mild to moderate level of exercise exertion in both groups (40–60 % of the VO2 max value). For both the dance-movement and exercise arms, interventions will span 12 weeks, with 1-h sessions twice a week (a total of 2 h per week) in groups of about 6-12 participants each. Based on studies with exercise interventions, a 12-week intervention supports significant result comparison [10]. The waitlist controls will continue with routine care and can join either a dance-movement or exercise intervention after the 12-month assessment.

The assessments will be conducted by a researcher blinded to the randomization, trained in using the instruments. CDRs will be assessed by a psychiatrist and a trained CDR rater blinded to the randomization. Self-rated scales will be completed by the participants under guidance.

To obtain cortisol levels from the saliva samples, the samples will be thawed and then centrifuged at 3000 rpm for 15 min at room temperature. The cortisol levels will be determined using an enzyme-linked immunoabsorbent assay kit (EIA, Salimetrics). The assay sensitivity for the kit is 0.007 g/dl and the intra- and inter-assay coefficients of variation are 3 and 10 %, respectively. Due to the skewed distribution of cortisol data, natural logarithm will be used to transform the raw cortisol data and yield a normal distribution for all of the analyses. An average of the five collected samples across that day will produce the mean cortisol level. Diurnal cortisol rhythm will be calculated by linear regression of the log transformed cortisol on collection time, and total cortisol level will be measured by the area under the curve.

Correlation analyses will be used to explore the interrelationships between baseline diurnal cortisol patterns and psychosocial, cognitive and functioning measures. Using a variant of multiple regression modeling for the nested structure of our data, a two-level individual growth curve model will be used to explore the individual trajectories of changes in cortisol levels over time, and the complex relationships between different variables. The cortisol measures at the five daily time points will be nested within the participants.