Multisensory Stimulation as an Intervention Strategy for Elderly Patients With Severe Dementia: A Pilot Randomized Controlled Trial

A pilot randomized controlled trial was conducted among older participants aged 65 years or older, stratified according to their cognitive status being afterward randomly assigned to one of the 3 groups (MSSE, activity, and control).

The sample was selected among the residents of a specialized dementia elderly center in A Coruña (Spain). The inclusion criteria were a diagnosis of dementia and the presence of severe or very severe cognitive decline (GDS, 6-7)¹⁹. Dementia diagnoses was noted on the medical history and provided by a neurologist before placement in the gerontological complex, being corroborated by the elderly center’s medical doctor. GDS was applied by a clinical psychologist with experience in assessing people to determine level of severity: severe (GDS 6) or very severe (GDS 7) cognitive decline. The exclusion criteria were the presence of a sensory disorder that would adversely affect interactions with the multisensory stimulation objects (eg, severe vision and hearing impairment) and be bedridden.

After the clinical psychologist checked the eligibility of participants according to the inclusion and exclusion criteria, the sample consisted of 32 participants. A computer-based random number generator was used to randomly divide the sample into 3 groups of 11, 11, and 10 individuals according to GDS. The initial sample size decreased to 27 during the follow-up period due to participant deaths (n = 4) and dropouts (n = 1). The patients’ progress through the trial is shown in a Consolidated Standards of Reporting Trial (CONSORT) diagram (Figure 1).

The study protocol was approved by the Ethics Committee at the University of A Coruña followed the principles of the Declaration of Helsinki. Before beginning data collection, all participants’ proxies were informed about the study. Proxies were used as legally authorized representatives to provide informed consent for the elderly individuals having dementia to participate in the research.

People from the MSSE group participated in multisensory sessions in a Snoezelen room. This room, that stimulate all the senses except taste, included several elements such as alternating colors fiber-optic cables, 2 water bubble columns within 2 mirrors, a water bed, a rotating mirror ball with a color light projector, a video, an interactive projecting system, musical selections, aroma therapy equipment with fragrant oils, and a tactile board with various textures, among others.

The activity group participated in a series of one-to-one activity sessions, in which intellectual and/or physical demands were placed on the individual, and the approach used is directive, being the therapist the person responsible to choose the activities to be performed. Participants were asked to take part in simple activities such as looking at photographs or playing games. This group was included in the study to differentiate the specific benefits of the multisensory stimulation from those derived from attending one-to-one therapy sessions.¹³

The control group did not participate in any of the aforementioned activities; rather, this group continued with the daily routines of the center, including cognitive stimulation group sessions (GDS 6), training on activities of daily living (GDS 6), education and training of nursing assistants in dementia knowledge, acknowledgment of resident’s experiences, and communication techniques and behavior management (GDS 6-7).

The design of the sessions followed by the MSSE and the activity groups was based on the protocol suggested by Baker et al.²⁰ Participants from both groups were required to take part in 2 weekly sessions, for a period of 16 weeks, until they complete 32 sessions. These sessions lasted 30 minutes, unless the participant expressed a desire to leave.

The difference between MSSE and activity sessions was given by the characteristics that define the MSSE. In the MSSE group, multisensory unpatterned stimuli were used, the therapist followed a nondirective approach, and the therapy required few intellectual or physical demands. In contrast, during the activity sessions, no intentional special multisensory experiences were introduced, the therapist followed a directive approach, and intellectual and/or physical demands were placed on the patient.

Data on participants’ sensorial preferences and interests were previously collected to design the content of the sessions and to minimize the behavioral problems that some participants could present within the MSSE and the activity contexts. In the MSSE group, sensorial preferences in the Snoezelen room were assessed based on the procedure suggested by Pace et al.²¹ Furthermore, relatives of participants of both groups were interviewed to identify participants’ hobbies and interests.

Behavior, mood, cognition, and dementia severity were assessed at baseline (pretrial, week 0), in the middle (midtrial, week 8), at the end of the intervention (posttrial, week 16), and 8 weeks after the intervention (follow-up). Due to ethical reasons, the control group did not remain without intervention for more than 16 weeks; therefore, this group did not participate in the 8-week follow-up.

The validated Spanish version²² of the Cohen-Mansfield Agitation Inventory (CMAI)²³ was used to assess the frequency of agitated behaviors in the participants. The CMAI consists of 30 items that are each rated on a 7-point scale of frequency, 1 meaning never and 7 several times per hour. The total score is calculated by summing the scores of each of the individual items. Through a factor analysis, Cohen-Mansfield et al²³ found the following 3 factors of agitation in the nursing home: “aggressive behavior” (hitting, kicking, pushing, scratching, tearing things, cursing or verbal aggression, and grabbing); “physically nonaggressive behavior” (pacing, inappropriate robing or disrobing, trying to get to a different place, handling things inappropriately, general restlessness, and repetitious mannerisms); and “verbally agitated behavior” (complaining, constant requests for attention, negativism, repetitious sentences or questions, and screaming). In this study, for each factor the total score was obtained by summing the scores of the corresponding items. The CMAI interrater reliability²³ ranges from .88 to .92 and the internal consistency reliabilities (Cronbach’s α)²⁴ range from .86 to .91.

Behavior was assessed using the Spanish version²⁵ of the Neuropsychiatric Inventory (NPI).²⁶ This scale was developed to assess a wide range of behaviors in patients with Alzheimer’s disease and other dementias. The NPI questionnaire evaluates 12 neuropsychiatric disturbances including delusions, hallucinations, agitation, dysphoria/depression, anxiety, euphoria, apathy, disinhibition, irritability, aberrant motor behavior, sleep and night-time behavior disturbances, and changes in appetite and eating behaviors. Neuropsychiatric Inventory is completed according to the answers of the caregivers. A group of screening questions are asked first, which are followed by a series of subsequent questions if the response to the initial screening indicates the presence of neuropsychiatric alterations. The caregiver rates the frequency of the symptoms using scores from 1 to 4 (1 = occasionally, less than once per week; 4 = very frequently, once or more per day or continuously) and also rates the severity using scores from 1 to 3 (1 = mild, 2 = moderate, and 3 = severe). The total score ranges from 0 to 144, with higher values indicating more behavioral and psychological alterations. The Spanish version of the NPI²⁵ has shown good internal consistence (α = .85) and interrater reliability (.63 to 1.00).

The Spanish version²⁷ of the Cornell Scale for Depression in Dementia (CSDD)²⁸ was used to assess mood. This scale was specifically developed to assess signs and symptoms of major depression in patients with dementia. Information is elicited through 2 semistructured interviews: an interview with an informant and an interview with the patient. The CSDD consists of 19 items that are rated for severity on a scale of 0 to 2 (0 = absent, 1 = mild or intermittent, 2 = severe). Total score is obtained by summing the scores for all items, being the minimum score 0 and the maximum score 38. Scores above 10 indicate probable major depression. Scores above 18 indicate definite major depression. In the Spanish population, CSDD has shown good test–retest reliability (.61 to .84) and good internal consistency (α = .81).²⁷

To assess cognitive function, the Spanish version²⁹ of the Severe Mini-Mental State Examination (SMMSE)³⁰ was used. The SMMSE was designed for assessment of severe dementia preventing the floor effect found when using the MMSE.³¹ This simple instrument does not require specialized training or foreign material, and it is not tiring for the patient with dementia (takes less than 5 minutes to administer). It consists of 10 items related with autobiographical knowledge (complete name and birth date), constructional praxis tests, phonological loop (spelling), and semantic verbal fluency step (animal category generation). The SMMSE also tests receptive and expressive language skills, along with elementary executive functions and visual–spatial abilities, which are likely to be preserved in patients with severe impairment. The total score ranges from 0 to 30 points, with lower values indicating lower cognitive function. The SMMSE has shown both construct and criterion validity for assessing patients with severely impaired Alzheimer’s disease.³⁰ In the Spanish population, SMMSE has shown high internal consistency (α = .88), test–retest reliability (.64 to 1.00), interrater reliability (.69 to 1.00), and construct validity in correlation with the Spanish version of the Mini-Mental State Examination (Pearson r coefficient = .59).²⁹

The overall severity of dementia was measured by the Bedford Alzheimer Nursing Severity Scale (BANS-S).³² The BANS-S is an observational scale that can be used even with persons who are unable to follow simple commands, uncooperative, or unable to communicate. This is a 7-item scale that combines ratings of interaction ability (speech, eye contact), functional deficits (dressing, eating, and ambulation), and occurrence of pathological symptoms (sleep–wake cycle disturbance, muscle rigidity).³³ Each item is scored on a 4-point scale where a scoring system is specified for each item. The total score ranges from 7 (no impairment) to 28 (most severe impairment). The BANS-S is more sensitive to detecting disease progression beyond the severe stage than scales that measure only cognitive or functional deficits.³² BANS-S has shown good internal consistence (α = .80), convergent validity with other cognitive and functional scales (r = .62-.79), and discriminant validity in comparison with the NPI (r = .36).^32,34

Sample characteristics were summarized as mean and standard deviation (SD) for the continuous variables and as frequency and percentage for the categorical variables. The Shapiro-Wilk test was used to evaluate the normality of the sample. This test is more appropriate for small sample sizes (< 50 samples).³⁵ Differences between groups at baseline were compared using chi-square test for proportions and the nonparametric Mann Whitney U test for nonnormally distributed continuous variables. Statistical significance was set at a P value of less than .05.

Finally, repeated measures two-way analysis of variance (two-way mixed ANOVAs) (two-way mixed ANOVAs) was used to assess performance differences in behavior, mood, and cognitive status and dementia severity over the pre-, mid-, and posttrial assessment points. In the first analysis, the within-subject variable was the measures over time (pre-, mid-, and posttrial assessment) and the between-subject variable included the group (MSSE and activity). In the second analysis, the within-subject variable was the measures overtime (pre-, mid-, and posttrial assessment) and the between-subject variable included the group (MSSE and control).

In addition, repeated measures two-way mixed ANOVAs were used to assess performance differences in behavior, mood, and cognition, and dementia severity between the posttrial and the 8-week follow-up. In this case, the within-subject variable was the measures over time (posttrial assessment and follow-up) and the between-subject variable included the group (MSSE and activity).

Differences between groups were tested by a group–time interaction. Eta-squared values (η²) were reported as indicators of effect size. We interpreted the importance of the effect size using the benchmarks for “small” (η² of .02), “medium” (η² of .13) and “large” (η² of .26) offered by Cohen (1988).³⁶ Statistical significance was set at a P value of less than .05. Statistical analysis was performed using the SPSS version 20.

With regard to agitation, as assessed by the CMAI, when comparing the MSSE group and the activity group, an improvement in both groups in the aggressive behavior factor (F_2,38 = 8.200, P = .001, η² = .300) and in the total score (F_2,36 = 6.990, P = .003, η² = .277) between pre-, mid-, and postintervention (Figure 2) was observed. However, no significant differences were found between groups. For both groups, the scores worsened in the follow-up period compared to the posttrial assessment in the aggressive behavior factor (F_1,16 = 11.983, P = .003, η² = .420) and in the CMAI total score (F_1,16 = 14.031, P = .002, η² = .467), with no significant differences between the groups. With regard the verbally agitated factor, the scores improved from the pretrial to the posttrial in the 2 groups, although the results were not statistically significant. Between the posttrial assessment and the follow-up period, the scores worsened in both groups (F_1,16 = 15.623, P = .042, η² = .456).

When comparing the MSSE group and the control group, group-time interactions in verbally agitated behavior and in the CMAI total score were observed. Specifically, an improvement in verbally agitated behavior (F_2,36 = 3.460, P = .042, η² = .155) and in the CMAI total score (F_2,36= 11.755, P < .001, η² = .301) from the pretrial to the posttrial in the MSSE in comparison with the control group was found. For the aggressive behavior factor, significant time effects (F_2,36 = 3.632, P = .037, η² = .160) were also observed among pre-, mid-, and postintervention results, with a decrease in the scores for both groups.

With regard to physically nonaggressive behavior, no significant time effects or intergroup differences were found.

A significant group–time interaction effect was also found when comparing the NPI scores in the MSSE group and in the activity group (Figure 3). Patients in the MSSE group showed a significant higher improvement than the activity group from pretrial to posttrial (F_2,36 = 6.2121, P = .005, η² = .238), with no significant differences between MSSE and control groups.

The CSDD scores remained stable during the intervention period in the 3 groups (Figure 4). In the follow-up period, the CSDD scores in the MSSE and in the activity group worsened compared to posttrial assessment. However, the results were not significant.

With regard to SMMSE, when comparing the MSSE group and the activity group (Figure 5), both groups displayed an improvement in their scores during intervention. Between the posttrial and the follow-up period, a significant time effect was observed, with an important decrease in the scores of both groups (F_1,15 = 7.276, P = .017, η² = .324), with no significant differences between the 2 groups.

When comparing the MSSE group and the control group, no significant time effects or intergroup differences were found.

For the BANS-S (Figure 6), there was an improvement in the scores between pre-, mid-, and postintervention assessments in the MSSE group but not in the other groups. Significant interaction effects between group and time when comparing the MSSE group with the activity group (F_2,36 = 4.126, P = .024, η² = .171) and with the control group (F_2,36 = 11.578, P < .001, η² = .334) were observed.

In the current study the NPI scores improved significantly after 16 weeks of intervention in the MSSE group in comparison with the individualized activities group. For interpreting the effect sizes, we have not found previous studies to compare with, so the benchmarks of Cohen (1988)³⁶ have been used. According to them, the effect size was large. Previous studies^10,13,17 found that both MSSE sessions as one-to-one preference-based activity interventions could have positive long-term effects (that is, the maintenance of the positive effects of the MSSE outside of the session time and place) on neuropsychiatric symptoms of older adults with moderate to severe cognitive impairment living in nursing homes. In patients with moderate cognitive impairment benefits of intervention can be attributed to the individualized attention rather a specific treatment modality. However, people in the later stages of the disease, with very low level of functioning, may benefit more from sensory interventions, like MSSE, than from more cognitive demanding activities.

As cognitive function deteriorates, the world is experienced at a sensory level and the environment needs to be managed carefully to make it understandable and comfortable.³⁷ Consequently, individuals with severe dementia specially need an appropriate environmental structure and stimulation, which can be achieved by sensory interventions.³⁸ The Guideline on Supporting People with Dementia and Their Carers in Health and Social Care developed by the National Institute for Health and Clinical Excellence and the Social Care Institute for Excellence (NICE-SCIE)³⁹ highlighted that in the late stages of dementia, sensory stimulation is the primary form of psychological intervention to reduce neuropsychiatric symptoms. In this sense, Baker et al¹³ studied the effects of an Snoezelen intervention in comparison with a control group that followed one-to-one activity sessions and found that, in the severe cognitive range, the Snoezelen group was significantly less apathetic after 8 weeks of intervention. Indeed, there is evidence that others forms of sensory interventions, like music interventions⁴⁰ or aromatherapy with essential oils,⁴¹ also show good results in reducing neuropsychiatric symptoms of people with severe dementia.

Regarding agitated behavior, an improvement in the verbally agitated behavior factor and in the CMAI total score in the MSSE group in comparison with the control group was found. An improvement with the intervention in both MSSE and activity groups in the aggressive behavior factor and in the CMAI total score was also observed. As stated earlier, according to the benchmarks of Cohen (1988),³⁶ the effect sizes were large. These results indicate that although MSSE may be an appropriate intervention to reduce agitation in advanced dementia, its efficacy is equivalent to the individualized one-to-one activities.

In this study, the improvements observed in the behavior during the intervention disappeared in the follow-up in both groups. This is consistent with the previous studies comparing the effect of MSSE and activity groups in moderated to severe dementia,^13,17. In this regard, O’Connor et al^42,43 in a systematic review about the psychosocial treatments in people with dementia concluded that the benefits of the intervention on behavioral symptoms of dementia defrayed quickly with time. Therefore, to hope that repeated exposures to a treatment would consolidate benefits in the follow-up could be a too ambitious approach.

Although intervention in Snoezelen rooms has demonstrated to improve the mood of people with dementia at short term, long-term effects were not so evident.¹⁶ The current results found neither benefit after 16 weeks of intervention in the CSDD scores of people with severe dementia nor significant differences between the MSSE group and the other 2 groups. In another previous study with people with moderate to severe dementia,¹⁷ neither an improvement in the CSDD scores during the intervention, nor differences between the groups was found. In the study of Baker et al,¹³ MSSE was not found to be more effective that one-to-one activities in changing the mood neither in people with moderate dementia nor in those with severe dementia.

However, with the MSSE integrated in the daily care, better results have been observed on the mood of people with dementia. In the MSSE integrated in the daily care, the intervention was carried out during morning care, when the staff were engaged with residents in activities relating to bathing, grooming, dressing, and toileting.⁴⁴ Van Weert et al,⁴⁵ observed that patients with moderate to severe dementia receiving this type of intervention demonstrated a significant improvement in their level of depression after 15 months of treatment in comparison to the control group who received the usual care. Therefore, for future work it would be interesting to investigate the effect of MSSE integrated in daily care specifically in people with severe dementia.

Previous studies in people with moderate to severe dementia^13,17 have not found significant effects of the MSSE on the cognitive level or significant differences between MSSE and one-to-one activity sessions. The present study showed an improvement in the SMMSE scores during the intervention in MSSE and activity groups, although it was not significant. However, for both groups, a significant worsening in the scores was found between the posttrial and the follow-up period. Consequently, in people with severe dementia, one-to-one attention could have a positive efficacy on the maintenance of cognitive function.

With regard to BANS-S scores, patients in MSSE group showed a significant improvement during the intervention, which was not observed in the other 2 groups. According to the benchmarks of Cohen,³⁶ the effect sizes were large. There are limited studies that analyze the effect of MSSE on functional status of people with dementia. In patients with moderate to severe dementia, it has also been observed that a more structured approach of multisensory stimulation^46,47 has short-term positive effects on the functional status, however there are very room on the balance in individuals with dementia after 6 weeks of intervention few data on the long-term effectiveness. Klages et al⁴⁸ found no significant effects of a Snoezelen nor significant differences compared with the control group, who received one-to-one visits by volunteers. In another previous study,¹⁷ long-term benefits of MSSE on the functional status (Barthel index score) in people with moderate to severe dementia were not found. A possible reason for that could be that the Barthel index no assesses well the impact of the intervention in the functional status, especially in people with severe dementia. In the current study, this limitation was solved by focusing in people with severe dementia, using the BANS-S, a tool more sensitive to detect disease progression for this group of people.

These results highlight that MSSE could be an appropriate intervention in people with severe dementia, and support the MIS, according to which balancing time periods of high arousal and low arousal in people with dementia can delay their functional decline.

The sample may seem small but it should be considered the difficulty of getting more participants, especially taking into account their homogeneity and the possibility to be randomized. Therefore, future empirical studies with larger samples are needed.

The use of MSSE in a Snoezelen room requires the investment of economic resources greater than those required in other sensory therapies for people with dementia. The future research should conduct randomized control trials to examine whether the Snoezelen benefits are better than those provided by other sensory interventions, like individualized music or aromatherapy interventions.

Further, it would be interesting to compare the traditional MSSE in a Snoezelen room, which follows a nondirectional approach, with a more structured MMSE, like the functional analytic multisensory environmental therapy,⁴⁹ for people with severe dementia.

The MSSE integrated in daily care could be also an adequate intervention for people with severe dementia, because of the high frequency of neuropsychiatric symptoms during morning care in patients in the final phase of dementia.⁵⁰ Therefore, it would be interesting that future studies analyze the effect of MSSE integrated in daily care in people with severe dementia and compare it with the MSSE carried out in a Snoezelen room.^51,52