A cluster randomized controlled trial of a train-the-trainer behavioral intervention delivered via respite care centers to improve nutritional outcomes and quality of life persons with dementia and their caregivers.

Objective: To evaluate a behavioral intervention, Partners at Meals (PAM), designed to empower caregivers (CGs) to improve caloric intake, weight, and quality of life (QOL) of persons with dementia (PWD), and to address dysfunctional behaviors during mealtime. The study also aimed to assess differences in caregiver psychosocial outcomes, including depression, burden, and QOL.

Design: A cluster randomized controlled experimental design was used, in which participating Respite Care Centers (RCCs) were randomized to either the PAM intervention group (n = 3) or the control enhanced usual care (EUC) group (n = 3). Enrolled PWD/CG dyads were assigned to PAM or EUC based on the RCCs they attended.

Setting and participants: This longitudinal 6-month clinical trial was conducted across six RCCs in the Southeast coastal region of the United States. A total of 53 PWD/CG dyads provided post-enrollment data: PAM (n = 27) and EUC (n = 26).

Methods: The PAM intervention was delivered using a train-the-trainer approach, based on the C3P (Change the Person, Change the People, Change the Place) model, and was implemented through RCC volunteers who provided adaptive mealtime strategies via telehealth to CGs in the home. Primary outcome measures for PWD were changes in body weight and feeding/dysfunctional behaviors from baseline to 6-month follow-up or end of study. Secondary outcomes included mid-upper arm circumference (MUAC), caloric intake, and QOL. Caregiver outcomes included changes in burden, depression, and QOL. Between- and within-group comparisons were performed using pooled and paired t-tests or chi-square tests as appropriate. Generalized linear mixed models (GLMM) were used to assess outcomes over time.

Results: At enrollment, PWD participants had a diagnosis of mild to moderate Alzheimer's disease or related dementia, with a mean age of 77.6 ± 9.8 years. The mean age of caregivers was 66.3 ± 11.8 years. The PWDs in the PAM group started with greater weight loss prior to study enrollment but showed a slightly lower > 5% weight loss from baseline to end of study (20.8%), compared to the EUC group (22.7%), although not statistically significant (p = 0.275). While the EUC group experienced a slight decrease in MUAC from baseline to follow-up of 0.2 ± 7 cm, those in the PAM group showed an increase of 1.0 ± 2.0 cm, suggesting better maintenance of nutritional status. The estimated mean daily caloric intake between the PWD groups upon enrollment (368 ± 131, p = 0.006) showed the EUC group consumed more calories than the PAM group throughout the study. Although no significant differences were found in the unadjusted changes in mealtime scores for feeding difficulty (-0.8 ± 3.1, p = 0.411), dysfunctional behavior (-1.4, 6.0, p = 0.605) or QOL (0.3 ± 6.6, p = 0.482), the PAM group exhibited a decrease in dysfunctional mealtime behaviors over the study period. No significant differences in the unadjusted means scores at baseline or changes from baseline to follow-up were observed on measures of CG depression (-1.3 ± 4.6, p = 0.371) or burden (-1.4 ± 5.4, p = 0.354). Results show that there was a significant difference in the unadjusted QOL score between groups at end of study follow-up (10.8 ± 15.8, p = 0.031); mean health state scores among caregivers in the PAM group (81.2 ± 3.4) were higher at the end of study compared to the EUC group (68.9 ± 3.7), indicating better general health state and QOL.

Conclusion: Findings suggest that the PAM intervention, using a train-the-trainer approach shows promise in improving nutritional outcomes and reducing dysfunctional mealtime behaviors in PWD, as well as enhancing the QOL of CG. This approach extends the reach of the intervention and fosters the idea that mealtime interventions to enhance in-home caregiving may offer a viable approach to maintaining and improving the nutritional status of the PWD and overall well-being of both PWD and their CG, which is essential for addressing the multifaceted challenges of dementia care.