Activity Preferences of persons with dementia: An examination of reports by formal and informal caregivers

Abstract

Objectives

The aims of this study are (1) to describe the types of leisure activities preferred by persons with dementia in the past and present, as reported by family members and therapeutic recreation staff members, and (2) to examine the influence of demographic and functional abilities variables on the perceived current interest of persons with dementia.

Method

Family members of persons with dementia and therapeutic recreation staff involved in the care of those persons completed an activities preference assessment concerning the persons with dementia. Participants were recruited from a nursing home and a senior day center. Family members completed the assessment for the past and present, and staff members completed the assessment for the present.

Results

For present preferences, music was the most often preferred activity according to both family and staff, followed by reminiscence and reading, while the least desirable activity was cooking/baking according to family and trivia games according to the staff. Current preferences were significantly related to past preferences. Gender was also related to preference for some activities. A significant reduction in preference from the premorbid past to the present was found for music, reading, and cooking. Additionally, the number of activities persons preferred as reported by both caregiver groups was affected by dementia. Family members tended to report lower preference levels for the activities in the present in comparison to the staff, and they were also more likely to report not knowing the present level of preference than therapeutic recreation staff members.

Conclusion

The assessments by both therapeutic recreation staff members and family members shed light on leisure preferences of persons with dementia, while the varying reports reflect a possibly different perspective of the two caregiver groups.

Keywords

dementia leisure activities formal caregivers informal caregivers family caregivers,person centerd care

Introduction

Boredom and loneliness are common among people with dementia in nursing homes (Cohen-Mansfield, Marx, & Werner, 1992). One tool to prevent this is engaging this population through group activities. Activities that are similar or related to premorbid interests, which can be evaluated using tools such as the Pleasant Events Schedule – Alzheimer’s Disease (Teri & Logsdon, 1991), can be particularly engaging. For example, one study found that persons with dementia (PwD) spent more time actively participating in an activity when they were able to reminisce about similar activities from their past, compared to activities where there was no such connection to their past (Kovach & Henschel, 1996). Another study individualized activities according to past interests and found high levels of engagement (Buettner & Fitzsimmons, 2003). Thus, past preferences can be used to guide the selection of engaging activities.

Activities that account for present preferences can enhance engagement. Cohen-Mansfield, Marx, Thein, and Dakheel-Ali (2010b) collected reports from caregivers on both past and present preferences for various interests (e.g. art, music, reading, television). Individuals who had past interests in music and pets were more engaged when they were presented with stimuli related to these interests (e.g. a live dog) than when they were presented with unrelated stimuli. Those with reported current preferences in music, art, and pets were more engaged by stimuli related to these interests than by unrelated stimuli. Thus, both past and present preferences can be of predictive value when examining engagement.

Unfortunately, evaluating the preferences of PwD can be challenging as the disease progresses. During early stages of the disease, PwD can participate in interviews and provide self-reports of their activity preferences. Participants mention that activities provide a sense of continuity in their lives. Participants noted that activities were more meaningful when they provided a sense of connection and ties to personal identity (Phinney, Chaudhury, & O’Connor, 2007). Participants also reported that they missed their old hobbies (Tak, Kedia, Tongumpun, & Hong, 2015). However, such in-depth reports may not be possible during later stages of the disease. Instead, evaluations about activity preferences may need to be collected from caregivers such as family members (Tsunaka & Chung, 2012) and nursing home staff. Evaluations can, however, differ across respondents, as demonstrated by a qualitative study that found discrepancies between residents, their caregivers, and their nursing staff regarding activities that they considered meaningful for PwD (Harmer & Orrell, 2008). For example, activities that involved reminiscence were reported as meaningful by PwD and staff members but were not emphasized by family members. Personal characteristics and background variables, such as current level of functioning and physical limitations, may impact the differences in evaluations (Harmer & Orrell, 2008). These functional characteristics could affect the ability of caregivers to give accurate evaluations of interest in activity and should be examined more systematically.

Prior studies examining past or present activity preferences have often been conducted with small sample sizes of 8–30 participants (Phinney et al., 2007; Tak et al., 2015) or as qualitative analyses of interviews with relatively high functioning individuals (Menne, Johnson, Whitlatch, & Schwartz, 2012). Only one study compared past and present preferences (Cohen-Mansfield et al., 2010b), but they did not examine the factors (e.g. personal characteristics) that may have impacted caregivers’ reports of preferences. This study aims to advance this area by further exploring the past and present activity preferences of PwD with varying levels of functioning, as reported by family members and therapeutic recreation (TR) staff. The specific aims of the study are outlined below:

To describe the past and present preferences of PwD for different activities.

To examine the similarities and differences between preferences in the past and the present among PwD.

To compare the evaluations of family members to those of staff members regarding present preferences for activities.

To examine the impact of background variables on the ratings of current preferences, on the number of activities PwD were interested in, and on the ability of caregivers to specify preferences.

This was part of a larger study on engagement of PwD during group activities (Cohen-Mansfield, Hirshfeld, Gavendo, Corey, & Hai, 2016).

Methods

Participants. This study was approved by the Research Ethics Board of Baycrest Health Sciences. We recruited PwD from seven nursing home units and a senior day center that were part of one large facility which provided multiple services. Inclusion criteria were a diagnosis of dementia (from the medical chart) and informed consent. Exclusion criteria were: (1) diagnosis of bipolar disorder or schizophrenia; (2) Cognitive Performance Scale (CPS19) score 1 or lower (higher scores indicate lower function), because persons with higher levels of cognitive function can articulate their interests and needs; (3) no dexterity movement in either hand; (4) cannot be comfortably seated in a chair or wheelchair, or cannot be moved to the location of the group activity; and (5) never spoke any English.

Informed consent was obtained for relatives of 91 PwD. Twenty family members refused to consent.

Procedure. In preparation for conducting group activities, TR staff members and family members completed the preference assessment for each individual PwD prior to running the groups. A research assistant contacted family members via telephone or email to complete the preference assessment, and 81 responded.

Assessment of PwD preferences by family and TR staff members. Family members were asked to rate how much their relative liked group activities in general and specific activities, in the past (i.e. premorbidly) and in the present. The scale included the ratings ‘1 – Dislike very much, 2 – Dislike somewhat, 3 – Neutral, 4 – Somewhat like, 5 – Like it very much.’ They could also indicate that they did not know the past preference. Specific topics conducive to group settings were chosen including reading, music, cooking/baking, art, exercise, playing word/trivia games, playing active games, and reminiscing (e.g. did your relative like listening to music in the past?). TR staff members answered the same questions for the present using the same scale.

Background information. Once consent was obtained for eligible PwD, background information (e.g. age, sex) was gathered from the electronic health record system (Meditech) and also using Minimum Data Set (MDS) version 2. Background information derived from the charts included age, gender, and total number of medical diagnoses. Cognitive functioning was assessed via the CPS (Morris et al., 1994) through the MDS, which has the scale of 0 – ‘intact’ to 6 – ‘severely impaired.’ Activities of Daily Living (ADL) performance was assessed via the ADL-Long form (Morris, Fries, & Morris, 1999), with a scale of 1 – ‘independent’ to 28 – ‘complete dependence.’ Other background variables obtained from the MDS included speech impairment (0= clear speech, 1= unclear speech, 2= no speech); mobility difficulties (1 = mobile [i.e. walking], 2 = use of cane/walker/crutch, 3 = wheelchair independently, 4= wheelchair with assistance from others, 5 = immobile [i.e. bedridden]; English language comprehension difficulties (1 = English as first language, 2 = English as second language, 3 = minimal English, 4 = no English); hearing impairment (scale: 0 = hears adequately to 3 = highly impaired); and vision impairment (scale: 0 = adequate to 4 = severely impaired). Given that the MDS is generally updated quarterly, the functional ratings should be considered approximations of the level of function around the beginning of the group activity.

Results

Of the total 91 potential PwD, 90 people are included in the analysis due to unavailability of both TR and family ratings for one PwD. Over half the PwD were female (60.0%); average age was 86.8 years (SD 8.81, range from 59 to 101 years); over a half (60%) were not married at the time of study. ADL-Long form, obtained through the MDS (Morris et al., 1999), averaged 18.7 (SD 7.82, range 0–28; 0 – ‘independent’ 28 – ‘total dependence’). Cognitive functioning assessed via the CPS (Morris et al., 1994), averaged 4.0 (SD =1.31, range 2–6; scale: 0 –‘intact’ to 6 – ‘very severe’). PwD had an average of 6.06 medical diagnoses (SD = 3.03, range 0–18); 17.6% had primary school education, 39.2% had completed secondary school, and 43.2% had completed post-secondary education.

A total of 81 family members completed the assessments of preferences. Missing information for the other family members is due to our inability to contact them despite multiple attempts. TR staff members completed the questionnaire for 86 PwD. For the others, TR staff members did not complete the questionnaire because they were not sufficiently acquainted with the PwD. Ratings were available from both family members and TR staff for 77 PwD.

Family members included spouses (17), children (55), sibling (1), nephews (3), nieces (2), friend (1), cousin (1), and family friend (1). Nine TR staff members took part in the study, eight of whom worked in the nursing home and one in the senior day center. TRs had a variety of training, some with Diplomas, Bachelor’s degrees, and two had certification as a Certified Therapeutic Recreation Specialist (CTRS). TR staff members were all female, with a median age in the 30 to 39 category. They had a mean of 10.7 years in the profession (median of 8 years) and a median of 11 months working with the PwD in the study (range 4 months–25 years). They completed the assessment on Excel spreadsheets.

Preferences of PwD for different group activities

The preferences of the PwD as reported by family and by TR staff are presented in Table 1. Listening to music was ranked highest by both TR staff and family for the present as well as for family ratings of past preferences. ‘Reminiscing,’ ‘reading,’ ‘art,’ and ‘participating in groups’ tended to receive relatively high ratings by TR and family, whereas ‘playing trivia games’ tended to receive the lowest ratings. ‘Cooking/baking’ had very low ratings according to family members – the second lowest in the past and the lowest for the present – and it was the third lowest according to staff. In order to examine whether the differences between past and present and between TR staff and family members were statistically significant, we examined the paired comparisons for those cases where we had both ratings. These comparisons are presented in Table 2.

Table 1.

Preferences for activities: TR staff and family reports for the full sample.

Ratings by	Family ratings				TR ratings
Pertaining to preferences in the:	Present		Past		Present
Topics rated for level of liking	N	Mean	N	Mean	N	Mean
Listening to music	73	4.45	81	4.63	79	4.62
Reminiscing	47	4.15	79	4.22	59	4.29
Reading	36	3.72	79	4.33	53	4.13
Participating in groups	66	3.59	78	3.76	80	4.05
Exercising	51	3.49	80	3.70	64	3.91
Viewing and/or producing art, painting, etc.	52	3.42	74	3.49	57	4.14
Playing physical games, like ball or bowling	40	3.08	77	3.38	68	4.00
Playing trivia/word games	33	2.7	75	3.03	57	3.75
Cooking/baking	47	2.47	76	3.28	47	3.94

The table is organized in descending order based on family ratings of present preferences.

Table 2.

Comparison of past and present preferences based on family ratings for those who provided both past and present preferences, and percentages of those family members who said they did not know the preferences based on the full sample.

		Those who provided ratings for both past and present						N = 81
Topic rated for level of liking	N^a	Mean past preferences	Mean present preferences	Correlations of past and present preferences	t-test comparing past and present preferences	% Increased interest from past to present	% Decreased interest from past to present	% Don’t know (past)	% Don’t know (present)^b
Listening to music	73	4.64	4.45	0.27*	t₍₇₂₎=1.90* one tailed	9.6	26.0	0	9.9
Reminiscing	46	4.26	4.20	0.29*	NS	26.1	26.1	2.5	42.0***
Reading	36	4.25	3.72	0.65***	t₍₃₅₎=3.26**	8.3	38.9	2.5	55.6***
Participating in groups	65	3.68	3.62	0.31*	NS	27.7	40.0	3.7	18.5**
Exercising	50	3.72	3.54	0.43**	NS	24.0	40.0	1.2	37.0***
Viewing and/or producing art, painting, etc.	47	3.4	3.4	0.38**	NS	27.7	31.9	8.6	35.8***
Playing physical games, like ball or bowling	40	3.2	3.08	0.234	NS	30.0	32.5	4.9	50.6***
Playing trivia/word games	49	2.87	2.74	0.626***	NS	40.8	18.4	7.4	59.3***
Cooking/baking	47	2.79	2.47	0.719***	t(46)=2.09*	8.5	23.4	6.2	42.0***

Note: Rating scale: 1 – Dislike very much, 2 – Dislike somewhat, 3 – Neutral, 4 – Somewhat like, 5 – Like it very much.

^aN does not include the ‘don’t know’ ratings which were counted as missing.

^bAsterisks refer to McNemar tests comparing past and present numbers of ‘don’t know’ ratings.

*p≤.05 two tailed, **p≤.01 two tailed ***p≤.001 two tailed; NS: not significant.

Similarities and differences between preferences in the past and the present of PwD

The preferences of PwD for different group activities in the past and in the present, as reported by family, are described in Table 2. The activities rated as most liked in the present and in the past are ‘listening to music’ and ‘reminiscing.’ The general order of ranking is similar for the past and the present (the correlation between rankings is 0.97). Similarly, there were positive correlations between past and present levels of liking for all the activities, with eight of the nine correlations being statistically significant. Despite the similarity in overall ranking, a general reduction in mean level of liking of the activities was reported, though this reduction was statistically significant only for reading, cooking/baking, and music (Table 2). Although there were generally more reports of decreased liking for an activity over time, there was a substantial percentage of PwD for whom rating for some activities increased. This was most pronounced for the categories of ‘playing trivia games,’ ‘participating in groups,’ ‘physical games,’ ‘reminiscing,’ and ‘art activities,’ where over a quarter of PwD for whom information is available were reported to have an increased interest in the topic. However, the number of activities rated as ‘somewhat liked’ or ‘liked very much’ in the past (mean = 5.80, SD = 1.96) was significantly higher than the number reported for the present (mean = 3.17, SD = 2.30), t₈₀=8.14, p<.001). The number of liked activities in the past was not significantly correlated with the present number of liked activities (r = .074).

We also examined the ‘don’t know’ responses to the questions. Family reported not knowing the preferences of the PwD in the past for different activities for between 0% and 8.6% of the PwDs. The highest levels of not knowing past preferences occurred for ‘art’ (8.6%) and ‘trivia’ (7.6%), and the lowest were for ‘listening to music’ (0%) and ‘exercising’ (1.2%). Rates of not knowing were higher for ratings pertaining to the present. The categories with the highest rates of not knowing were ‘trivia’ (59.3%), ‘reading’ (55.6%), and ‘physical games’ (50.6%), while rates were lowest for ‘music’ (9.9%) and ‘participating in groups’ (18.5%). These rates for lack of knowledge reported by family members regarding preferences in the present were significantly higher (using the McNemar test) than those for preferences in the past for all activities for which this could be examined. (Note that the McNemar test could not be calculated for music because of lack of variance: all relatives could rate degree of liking music in the past.)

Comparison of the evaluations of family members to those of staff members regarding preferences for activities

The comparison of the evaluations of family members to those of staff members regarding preferences for activities in the present is displayed in Table 3. While correlations between TR staff members’ and relatives’ ratings of liking were mostly positive, only three of them (‘reading,’ ‘participating in group activities,’ and ‘cooking or baking’) were statistically significant. When examining the number of activities persons liked (i.e. a rating of ‘somewhat liking’ or ‘liking very much’), there was a significant correlation between ratings by the two groups (r=.52, p<.001). Family members tended to report lower levels of liking of the activities in the present than TR staff members. The differences were statistically significant for most activities (reading, art, physical games, trivia games, participating in groups, and cooking/baking). Family members were also more likely to report not knowing the present level of preference than TR staff members. This trend was true for most activity topics and was statistically significant using the McNemar test for reading, physical games, and trivia type games (Table 3). Family members reported fewer activities liked by the PwD (mean = 3.30, SD=2.28) than TR staff (mean 4.36, SD=2.67), t₇₆=3.82, p<.001), yet there was a significant correlation between the number of preferred activities reported by the two groups (r=.52, p<.001). Similarly, family reported not knowing whether the person liked the activity for more activities (mean = 3.27, SD=2.99) than TR staff (mean = 2.38, SD=2.58), t₇₆=2.47, p=.02). Again, this correlation was significant (r=.355, p<.01).

Table 3.

Differences in ratings of present preferences between TR and family – means and t-test for those with ratings by both.

	Both TR and family valid ratings were available					N = 77^a,b
Topics rated for level of liking	N^a,b	Family present ratings mean	TR present ratings mean	r of TR and family ratings of present preferences	Paired t-test: family vs. TR	% Don’t know for family in present	% Don’t know for TR in present^c
Music	66	4.55	4.71	−0.03	NS	7.8	7.8
Reminiscing	38	4.34	4.53	0.01	NS	42.0	34.4
Reading	28	3.93	4.43	0.51**	t₂₇=2.65*	55.6	36.4*
Art, painting, etc.	39	3.67	4.39	0.22	t₃₈=3.51***	33.8	32.5
Playing physical games, like ball	33	3.24	4.21	0.08	t₃₂=4.10***	48.1	20.8***
Playing trivia/word games	25	2.76	4.2	0.10	t₂₄=4.88***	59.3	32.5**
Participating in groups	61	3.61	4.05	0.34**	t₆₀=2.72**	15.6	6.5
Exercise	41	3.66	3.93	0.16	NS	39.0	24.7
Cooking or baking	29	2.48	3.76	0.39*	t₂₈=5.05***	39.0	44.2

Rating: 1 – Dislike very much, 2 – Dislike somewhat, 3 – Neutral, 4 – Somewhat like, 5 – Like it very much.

^aIncluding all those for whom both family and TR ratings were available.

^bN does not include the ‘don’t know’ ratings which were counted as missing.

^cAsterisks refer to McNemar tests comparing ratings of ‘don’t know’ for relatives vs. TR staff.

*p≤.05 two-tailed, **p≤.01 two-tailed, ***p≤.001 two-tailed.

The relationship between background variables and present preferences, and the ability of caregivers to estimate the preferences

Predicting number of liked activities in the present

Total number of activities liked in the present was negatively and significantly correlated with all cognitive and functional impairment variables for both respondents, except for hearing impairment with family ratings (Table 4). Regression analysis showed that cognitive functioning (CPS) explained 21.4% of the variance in number of liked activities based on family ratings, whereas the combination of cognitive impairment, speech impairment, and hearing impairment explained 42.5% of the variance in TR ratings (Table 4).

Table 4.

The relationship between number of activities liked in the present and background variables.

	Correlations		Stepwise regressions (Beta)
	Number of activities person likes according to family	Number of activities person likes according to TR	Number of activities person likes according to family	Number of activities person likes according to TR
Cognitive impairment – Cognitive Performance Scale	−.463***	−.551***	−.463***	−.318**
Mobility impairment	−.291**	−267*
ADL impairment – Long form	−.432***	−.478***
Speech impairment	−.348**	−.564***		−.357**
Hearing impairment	.025	−.248*		−.191*
			R²=.214	R²=.425

Note: Number of activities the person likes was not significantly related to other background variables nor to the number of activities the person liked in the past.

*p≤.05, **p≤.01, ***p≤.001.

Preference for specific activities

In examining whether background and functional variables explained the ratings of liking for the different activities, we found liking for ‘cooking/baking’ was higher among females, both according to family (t₄₅=2.56, p=.01) and to TR staff (t₄₅=3.62, p=.001). Although men were not reported to like exercising more than women, they liked physical games, such as playing with a ball, more than women, according to family reports (t₃₈=1.71, p<.05 one-tailed). Significant correlations between preferences for specific activities and cognitive and functional variables occurred primarily for TR staff ratings. Family ratings only showed a significant association between preference for exercise and speech impairment (r=−.354, p<.05). For TR staff members, ratings showed significant correlations between cognitive impairment and preferences for ‘music’ (r=−.303, p<.01), ‘reminiscing’ (r=−.329, p=.01), ‘playing physical games’ (r=−.248, p<.05), ‘trivia games’ (r=−.444, p=.001), and ‘reading’ (r=−.418, p<.001); between speech impairment and liking music (r=−.304, p<.05), liking to ‘reminisce’ (r=−.314, p<.05), liking physical games (r=−.264, p<.05), liking trivia games (r=−.366, p<.05), and liking reading (r=−.394, p<.01); and between hearing difficulties and liking to participate in groups (r=−.234, p<.05) and liking to reminisce (r=−348, p<.01). Despite the common sense appeal of these results, they should be examined cautiously because they are based on a large number of multiple comparisons.

Predicting ‘don’t know’ responses

The correlations presented in Table 5 suggest that family members are less able to describe present preferences – as manifested by the number of ‘don’t know’ responses – with increasing impairment of the older person, especially in the level of cognitive (r=.596, p<.001) and ADL (r=.568, p<.001) functioning, but also with other measures of function such as speech, mobility, and vision. As expected, the ability to rate past preferences was not significantly related to current functioning, nor was the ability to describe past or present preferences related to demographic variables. Similar to the family reports, the TR staff’s inability to rate the level of liking was also related to functional and cognitive decline and to sensory deficits. Linear regression models indicated that cognitive function and ADL function explained 40.7% of the variance in family lack of knowledge of preferences, while 52% of the variance in TR staff’s not knowing was explained by speech impairment, ADL, difficulties with using English, and hearing impairment (Table 5).

Table 5.

Relationships between number of ‘don’t know’ responses and background variables.

	Pearson correlations			Stepwise regressions (Beta)
Number of ‘don’t know’ responses reported by:	Family (past)	Family (present)	TR staff (present)	Family (present)	TR staff (present)
Cognitive Performance Scale	0.120	.596***	.546***	.390**
ADL-Long form	0.116	.568***	.538***	.306*	.320***
Speech impairment	0.173	.394***	.587***		.427***
Mobility difficulties	0.065	.379***	.356***
Vision impairment	0.102	.249*	.210*
Difficulties with English language	0.083	0.111	.221*		.197*
Hearing impairment	0.074	−0.013	.288**		.187*
				R²=.407	R²=.520

Note: Correlations with age, sex, marital status, and education were not statistically significant.

Descriptive statistics for 'don't know' responses: M(SD), n: relatives re-past: .37(1.09), 81; relatives re-present: 3.51(3.11), 81; TR staff: 2.44(2.61), 86; relatives’ past vs. present paired t-test: t₈₀=9.147***; TR vs. relative re-present: t₇₆=2.472*.

*p≤.05, **p≤.01, ***p≤.001.

Discussion

This paper aimed to elucidate the past and present preferences of PwD for leisure activities and to explore the ability of family and TR staff to describe those preferences, as well as demographic and functional factors that might impact the level of preferences. Understanding such preferences and the ability to elicit them is important for planning both programs and spaces for such programs. Activities that interest PwD are crucial for alleviating boredom, providing meaning, and through these, also decreasing sad affect and inappropriate behaviors (Cohen-Mansfield et al., 2010a; Cohen-Mansfield, Marx, Thein, & Dakheel-Ali, 2011). Knowledge of the preferences of older PwD can be useful in predicting those activities that would result in higher levels of engagement (Cohen-Mansfield et al., 2010b). Yet, since the accuracy of caregivers’ reports of preferences of PwD has been questioned (Mesman, Buchanan, Husfeldt, & Berg, 2011), we plan to conduct a separate investigation evaluating the predictive value of family and staff reports in terms of actual engagement of older PwD in the group activities queried.

With respect to the first goal, that of clarifying the preferences of PwD for activity topics, we found ‘music’ to be the most highly rated by both family and staff, followed by ‘reminiscence’ and ‘reading,’ for present preferences. The least desirable activity was ‘cooking/baking’ according to family and it was ‘trivia games’ according to TR staff. These results reflect to some extent those reported in Menne et al. (2012), who interviewed PwD with mild cognitive impairment (mini-mental state examination (MMSE) = 22). Their sample was, on average, 10 years younger than ours, and they used different assessments, yet some similarities exist. Although the most commonly endorsed activity in Menne’s study was ‘socializing,’ the second category included ‘music,’ which was rated highest in our findings. It should be noted, though, that their category also included TV and radio, which were not included with ‘music’ in ours. Additionally, Menne’s category of ‘cook/eat’ was seventh, below ‘exercise’ and ‘cognitive stimulation,’ reflected in our finding that ‘cooking/baking’ was below ‘exercise’ and ‘cognitive games.’ These similarities provide some support for the generalizability of our findings.

In addition to examining the preferences themselves, we also examined factors that can affect ratings of preferences. Current preferences, as rated by relatives, were significantly related to past preferences, but the number of activities liked in the present was not related to the number of activities the person had liked in the past. In addition, impairment, represented by cognitive and functional status, affected the number of activities rated as interesting for the older person. TR’s ratings seemed to be influenced by the person’s cognitive and functional status to a greater extent than was true for relatives. It is possible that in cases where dementia might compromise interest in activities, relatives may be more likely to say they did not know rather than guess the level of interest or acknowledge lack of interest, and this was supported by the significant increase in ‘don’t know’ responses with increases in different types of impairment. Relatives have known the person for a long time, and they may base their ratings of present preferences on their knowledge of past preferences, tempered by the shift in abilities of the older person. Gender roles also had an impact on specific preferences, as evidenced by women’s higher ratings for ‘cooking/baking’ than men, and in men’s greater liking for ‘physical games’ than women; this was seen in the family reports and also in the Menne et al. (2012) findings.

As with the number of activities liked, the number of activities reported as unknown by informants was also correlated with cognitive and functional decline of the PwDs as well as by sensory impairment – but in the opposite direction. Regression analyses on both the number of ‘don’t know’ responses and the number of preferred activities for TR staff explained more of the variance than similar regressions on family ratings. TR staff may be better aware of current functional and sensory status of the PwDs. They may also have more experience with actual trials of such activities with the PwDs. The extent to which staff and family base their ratings on actual recent experience with activities, as opposed to their conclusions based on the person’s functional status, needs to be examined in future research. Furthermore, TR staff in this facility are expected to inquire about residents’ interests upon admission and to observe residents’ responsiveness to available activities. We did not inquire about the extent to which these factors affected TR responses in this study, nor did we check which activities were offered prior to the study and whether any were noted in residents’ charts in terms of their responsiveness. However, the range of activities used in the study seemed to exceed those offered by many of the participating TR staff members (Cohen-Mansfield et al., 2016). Future research should examine the actual experiences affecting staff ratings, and it also needs to clarify the extent to which such ratings actually predict responsiveness to specific activities beyond what would be predicted by cognitive function.

Staff and family ratings were significantly correlated for the number of liked activities and the number of activities they could not specify. However, there was less agreement regarding preferences for individual activities, and this information is important for planning activities that are liked by the PwD. It is unclear why there was significant agreement for ‘reading,’ ‘participating in groups,’ and ‘cooking and baking,’ whereas there was no agreement for ‘music’ or ‘reminiscing.’ In a qualitative study, Harmer and Orrell (2008) discuss the differences between PwD, relatives, and staff in what they perceive as a meaningful activity for the person with dementia. For example, nursing home residents mentioned past interests and the need to be useful as rendering activities meaningful, whereas family members mentioned maintaining skills and abilities. When both family members and TR staff provided ratings for preferences, staff tended to rate interest higher than family. This could be due to different viewpoints for each group, i.e. family viewing current interest against past interest, while TR staff viewed interest in comparison to other people with dementia. It is also possible that TR staff are more adept in eliciting interest in activities that are adapted to the current abilities of PwDs. Future research needs to ascertain the extent to which different attitudes and perceptions of the role of activities impact ratings of preferences for activities.

This paper pioneers the study of perceptions of preferences for activities in persons with advanced dementia, focusing on factors that impact those preferences and how they may be determined. It is also the only paper we know of to systematically compare past and present preferences of PwD. It indicated a profound impact of dementia and other impairments on the ability of caregivers to describe preferences and thus also on the number of activities liked according to both family and staff. The picture emerging for specific preferences is more complex. Despite a general decrease in interest in activities, changes were reported in both directions, with increases and decreases in liking for specific activities. The study is limited by the use of only one, albeit large, facility. However, the congruence with the limited prior literature and the similarity of the background characteristics of our population to other studies of nursing home populations supports the generalizability of findings. In the future, we hope to study the utility of such knowledge of preferences for activities to help predict actual engagement with those activities.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was made possible by the generous support of the Rotman Family and the Morris Justein Visiting Scholars Program, and by support of the Minerva Foundation.

Jiska Cohen-Mansfield is a professor at the Department of Health Promotion at the School of Public Health, Sackler Faculty of Medicine, and the director of the Minerva Center for Interdisciplinary Study of End of Life at Tel Aviv University. She is the Igor Orenstein Chair for the Study of Geriatrics at Tel Aviv University Medical School.

Rachel Gavendo is a Certified Therapeutic Recreation Specialist working in the Apotex Long Term Care facility at Baycrest Health Sciences.

Erin Blackburn is a Certified Therapeutic Recreation Specialist working in the Residential and Aging team at Baycrest Health Sciences.

References

Buettner

L. L.

Fitzsimmons

(2003). Activity calendars for older adults with dementia: What you see is not what you get. American Journal of Alzheimer’s Disease and Other Dementias, 18(4), 215–226. DOI:10.1177/153331750301800405.

Cohen-Mansfield

Hirshfeld

Gavendo

Corey

Hai

(2016). Activity-in-a-Box for engaging persons with dementia in groups: Implications for therapeutic recreation practice. American Journal of Recreation Therapy, 15(3), 8–18. DOI: https://dx-doi-org.web.bisu.edu.cn/10.5055/ajrt.2016.0107.

Cohen-Mansfield

Marx

M. S.

Dakheel-Ali

Regier

N. G.

Thein

Freedman

(2010a). Can agitated behavior of nursing home residents with dementia be prevented with the use of standardized stimuli? Journal of the American Geriatrics Society, 58(8), 1459–1464. DOI:10.1111/j.1532-5415.2010.02951.x.

Cohen-Mansfield

Marx

M. S.

Thein

Dakheel-Ali

(2010b). The impact of past and present preferences on stimulus engagement in nursing home residents with dementia. Aging & Mental Health, 14(1), 67–73. DOI: 10.1080/13607860902845574.

Cohen-Mansfield

Marx

M. S.

Thein

Dakheel-Ali

(2011). The impact of stimuli on affect in persons with dementia. The Journal of Clinical Psychiatry, 72(4), 480–486. DOI: 10.4088/JCP.09m05694oli.

Cohen-Mansfield

Marx

M. S.

Werner

(1992). Observational data on time use and behavior problems in the nursing home. Journal of Applied Gerontology, 11(1), 111–121. https://doi.org/10.1177/073346489201100109

Harmer

B. J.

Orrell

(2008). What is meaningful activity for people with dementia living in care homes? A comparison of the views of older people with dementia, staff and family carers. Aging & Mental Health, 12(5), 548–558. DOI:10.1080/13607860802343019.

Kovach

C. R.

Henschel

(1996). Planning activities for patients with dementia: A descriptive study of therapeutic activities on special care units. Journal of Gerontological Nursing, 22(9), 33–38.

Menne

H. L.

Johnson

J. D.

Whitlatch

C. J.

Schwartz

S. M.

(2012). Activity preferences of persons with dementia. Activities, Adaptation & Aging, 36(3), 195–213. DOI: 10.1080/01924788.2012.696234.

10.

Mesman

G. R.

Buchanan

J. A.

Husfeldt

J. D.

Berg

T. M.

(2011). Identifying preferences in persons with dementia: Systematic preference testing vs. caregiver and family member report. Clinical Gerontologist, 34(2), 154–159. DOI: https://dx-doi-org.web.bisu.edu.cn/10.1080/07317115.2011.539516

11.

Morris

J. N.

Fries

B. E.

Mehr

D. R.

Hawes

Phillips

Mor

Lipsitz

L. A.

(1994). MDS Cognitive Performance Scale. Journal of Gerontology, 49(4), M174–M182. DOI: https://doi.org/10.1093/geronj/49.4.M174

12.

Morris

J. N.

Fries

B. E.

Morris

S. A.

(1999). Scaling ADLs within the MDS. The Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 54(11), M546–M553. DOI: https://doi.org/10.1093/gerona/54.11.M546.

13.

Phinney

Chaudhury

O’Connor

D. L.

(2007). Doing as much as I can do: The meaning of activity for people with dementia. Aging & Mental Health, 11(4), 384–393. DOI: 10.1080/13607860601086470.

14.

Tak

S. H.

Kedia

Tongumpun

T. M.

Hong

S. H.

(2015). Activity engagement: Perspectives from nursing home residents with dementia. Educational Gerontology, 41(3), 182–192. DOI: 10.1080/03601277.2014.937217.

15.

Teri

Logsdon

R. G.

(1991). Identifying pleasant activities for Alzheimer’s disease patients: The Pleasant Events Schedule-AD. The Gerontologist, 31, 124–127. DOI: https://doi.org/10.1093/geront/31.1.124.

16.

Tsunaka

Chung

J. C. C.

(2012). Care-givers’ perspectives of occupational engagement of persons with dementia. Ageing & Society, 32, 543–560. DOI: 10.1017/S0144686X11000365.