A total of 1,275 individuals were referred to MIND at Home Phase II. Of those referred, 664 (52.1%) received a telephone screen, 255 (20.0%) were ineligible due to their demographic characteristics or had died before study contact, 188 (14.8%) were unreachable, 150 (11.8%) declined the screening, and 18 (1.4%) agreed to screening but could not be scheduled. Among those screened (n=664), 371 (55.9%) screened positive. Of those who screened positive (n=371), 57 (15.4%) received a home visit but did not meet eligibility criteria, 11 (3.0%) were eligible but did not complete the baseline assessment, and 303 (81.7%) completed the in-home assessment and were randomized into the trial. The randomized sample included 265 (87.5%) PWD and 38 (12.5%) participants with MCI. The total sample for the current analyses consists of 254 PWD whose study partners met the definition of an informal caregiver (i.e., a relative or friend who knows the person well, has contact with the person regularly and on whom the person relies for assistance).
A majority of the PWD were female (65%), white (68%), and lived with others (81%) (). Most (96%) of the non-white participants were African Americans. Participant mean age was 83.6. Participants were taking an average of 6.4 medications to address chronic health problems that fell into a mean of 3.0 disease group categories. With an average MMSE score of 17.8, the majority (59%) had scores >17, 26% had scores in the 10-17 range, and 15% scored <10 on the MMSE. The most prevalent dementia-related unmet needs related to safety (90%), general medical care (63%), and advance care planning (48%). The majority of caregivers were female (74.7%) and white (67.8%), with caregivers’ average age being 66.2. Caregivers’ relationships to PWD included spouses (40.2%), children (50.4%), other relatives (5.9%), and non-relatives (3.5%).
Characteristics of Persons with Dementia and Their Relationships to Quality of Life (QOL)
The majority (87%) of participants could rate their own QOL based on the QOL-AD; although few (n=12) of those with MMSE scores <10 could do so. The mean QOL-AD_SR score (37.7, SD=6.8) was significantly higher than the mean QOL-AD_CR score (31.2, SD=6.0) (paired t-test=12.61, p<.001), and ADRQL scores averaged 83.1 (SD=13.2). As expected, the three QOL scores were significantly correlated with one another. While QOL-AD_SR scores were correlated moderately with both QOL-AD_CR (r=.36, p<.001) and ADRQL scores (r=.15, p<.05), the highest correlation was between QOL-AD_CR and ADRQL scores (r=.52, p<.001).
Significant bivariate relationships between the PWD’s characteristics and the three QOL measures (see ) reflect similarities and differences across these measures. Few demographic characteristics were associated significantly with QOL in this sample. White participants had higher QOL-AD_SR and QOL-AD_CR scores; higher education was associated with higher QOL-AD_SR scores; and living alone was associated with higher ADRQL scores. Participants taking more medications had lower QOL-AD_SR scores, and those with more health problems had lower QOL-AD_SR and QOL-AD_CR scores. Lower QOL-AD_CR and ADRQL scores were associated with more functional impairments (IADLs, PGDRS) and lower cognitive function. Those with more severe and more distressing NPS and more severe depression had lower QOL on all three measures. Finally, having more dementia-related unmet needs was associated with lower QOL-AD_SR and QOL-AD_CR scores. The range of correlations among the independent variables for PWD was −.005 to .881.
Results of multiple regression models for the three QOL measures using only the PWD’s characteristics as independent variables are in . QOL-AD_SR scores were significantly lower for non-whites, those with more unmet needs, more symptoms of depression (i.e., CSDD), and those taking more medications, with these variables accounting for 22% of the total variance. In contrast, those with more distressing NPS, more unmet needs, more functional impairments (i.e., IADLs, PGDRS scores), and more health problem categories had significantly lower QOL-AD_CR scores, with these variables accounting for 31% of the total variance. Lower ADRQL scores were significantly associated with more distressing NPS, greater IADL impairment, lower cognitive function, and more symptoms of depression, with these variables accounting for 45% of the total variance.
Correlates of Quality of Life Using Characteristics of Persons with Dementia in Multiple Linear Regression Analyses
Bivariate relationships between caregiver characteristics and the two caregiver-rated QOL measures are shown in . Non-white race, the only significant demographic characteristic, was associated with lower QOL-AD_CR scores. Self-rated health, ZBI and GDS scores, CAS time, number of caregiver unmet needs, and SF-12 physical and mental health scores were significantly related to the QOL-AD_CR. Most of these same characteristics were significantly associated with ADRQL scores, with total caregiver unmet needs marginally (p=.067) related to the ADRQL. The range of correlations among the independent variables for the caregivers was .004 to .790.
Characteristics of Caregivers and Their Relationships to Caregiver-rated Quality of Life (QOL) of Persons with Dementia
The multiple regression analyses of the two caregiver-rated QOL measures that included characteristics of both PWD and caregivers are shown in . Caregiver burden and self-rated health along with the PWD’s IADL impairments, unmet needs, depression, and total health problem categories were significantly related to QOL-AD_CR scores, accounting for 36% of the variance. Significant correlates of ADRQL scores included caregiver burden and symptoms of depression along with the PWD’s neuropsychiatric symptom severity, IADL impairments and cognitive function, collectively accounting for half of the variance.
Correlates of Caregiver-Rated Quality of Life Measures Using Characteristics of the Caregivers and Persons with Dementia in Multiple Linear Regression Analyses