Using Voice Assistant Systems to Improve Dietary Recall among Older Adults: Perspectives of Registered Dietitians.

Dietary assessments are important clinical tools used by Registered Dietitians (RDs). Current methods pose barriers to accurately assess the nutritional intake of older adults due to age-related increases in risk for cognitive decline and more complex health histories. Our qualitative study explored whether implementing Voice assistant systems (VAS) could improve current dietary recall from the perspective of 20 RDs. RDs believed the implementing VAS in dietary assessments of older adults could potentially improve patient accuracy in reporting food intake, recalling portion sizes, and increasing patient-provider efficiency during clinic visits. RDs reported that low technology literacy in older adults could be a barrier to implementation. Our study provides a better understanding of how VAS can better meet the needs of both older adults and RDs in managing and assessing dietary intake.

Feasibility of using the Automated Self-Administered 24-hour (ASA-24) dietary assessment tool in older adults.

Background: Dietary patterns can impact the trajectories of healthy aging. However, dietary assessment tools can be challenging to use. With the increased use of technology in older adults, we aimed to evaluate the feasibility of older adults completing the online, Automated Self-Administered 24-h (ASA-24) dietary assessment tool. Methods: We conducted a randomized, two-period, two-sequence, crossover design of twenty community-dwelling older adults (≥65 years) comparing their preference for completing the ASA-24 alone versus with a research assistant (RA). Participants were recruited via ResearchMatch.com and randomly allocated 1:1 to a sequence of completing both an ASA-24 alone or with an RA, separated by one week. After each session, participants completed an online 11-item feasibility survey (Likert-scale range of 1-5, strongly disagree to strongly agree). Mean and standard deviations were reported for each question. Results: Mean age was 69 ± 3.5 years (90% females), with no differences were observed for sex, age, race, ethnicity, education, or income. Neither group felt a need for RA assistance (p = 0.34). However, both groups felt the system was easier to follow with the help of an RA (RA: 4.4 ± 1.3, vs. SA 4.6 ± 0.5, p = 0.65), particularly when they completed the ASA-24 alone, first (p = 0.04). When conducting the ASA-24 alone, there was less confidence the system could be learned quickly (SA 4.5 ± 0.5→3.4 ± 1.0 vs RA 3.4 ± 1.0→3.4 ± 0.7, p = 0.001). The ASA-24 was thought to be less cumbersome after repeated exposure in those concluding with the RA. Conclusion: While older adults were able to complete the ASA-24 independently, the use of an RA led to improved confidence. Enhancing the sample diversity in a larger number of participants could provide helpful data to improve the science of dietary assessment.

Obesity, multiple chronic conditions, and the relationship with physical function: Data from the national health and aging trends survey.

BACKGROUND: The population of older adults living with multiple chronic conditions (MCC) continues to grow. MCC is independently associated with functional limitation and obesity. The aim of our study was to evaluate the association between obesity and MCC, and secondarily, the combined presence of obesity and functional limitations with MCC. METHODS: We analyzed cross-sectional survey data from the National Health and Aging Trends Survey (NHATS) 2011 baseline data, a nationally representative Medicare beneficiary cohort of adults in the United States. We evaluated the coexistent prevalence of obesity and MCC overall, and by standard body mass index (BMI) categories. We then evaluated the prevalence of functional limitations (mobility, self-care, and household activities) and Fried-defined frailty status in persons with a BMI ≥ 30 kg/m2. Logistic regression was used to measure the association between MCC and BMI, and functional limitations and MCC among those with obesity. RESULTS: In the 6,600 participants, the prevalence of concurrent obesity and MCC was 30.4%. Of those with obesity, the prevalence of MCC was 84.0%, and were more likely to have MCC (adjusted OR: 2.17, 95% CI 1.86, 2.54) compared to a normal BMI. Obesity and functional limitations or frailty were more likely have MCC than individuals with obesity alone. CONCLUSIONS: We found that individuals with obesity is strongly associated with MCC and that functional limitations and frailty status have a greater association with having MCC than individuals with obesity without MCC. Future longitudinal analyses are needed to ascertain this relationship.

Evaluating Voice-Assistant Commands for Dementia Detection.

Early detection of cognitive decline involved in Alzheimer's Disease and Related Dementias (ADRD) in older adults living alone is essential for developing, planning, and initiating interventions and support systems to improve users' everyday function and quality of life. In this paper, we explore the voice commands using a Voice-Assistant System (VAS), i.e., Amazon Alexa, from 40 older adults who were either Healthy Control (HC) participants or Mild Cognitive Impairment (MCI) participants, age 65 or older. We evaluated the data collected from voice commands, cognitive assessments, and interviews and surveys using a structured protocol. We extracted 163 unique command-relevant features from each participant's use of the VAS. We then built machine-learning models including 1-layer/2-layer neural networks, support vector machines, decision tree, and random forest, for classification and comparison with standard cognitive assessment scores, e.g., Montreal Cognitive Assessment (MoCA). Our classification models using fusion features achieved an accuracy of 68%, and our regression model resulted in a Root-Mean-Square Error (RMSE) score of 3.53. Our Decision Tree (DT) and Random Forest (RF) models using selected features achieved higher classification accuracy 80-90%. Finally, we analyzed the contribution of each feature set to the model output, thus revealing the commands and features most useful in inferring the participants' cognitive status. We found that features of overall performance, features of music-related commands, features of call-related commands, and features from Automatic Speech Recognition (ASR) were the top-four feature sets most impactful on inference accuracy. The results from this controlled study demonstrate the promise of future home-based cognitive assessments using Voice-Assistant Systems.

Impact of whey protein supplementation in a weight-loss intervention in rural dwelling adults: A feasibility study.

BACKGROUND & AIMS: Weight loss in older adults enhances physical function, but may lead to sarcopenia and osteoporosis. Whey protein is a low cost rich source of essential amino acids, may improve physical function. We evaluated the feasibility and acceptability of consuming whey protein in the context of a weight-loss intervention in older adults with obesity. METHODS: A 12-week pilot feasibility, non-randomized weight loss study of 28 older adults was conducted, consisting of individualized, weekly dietitian visits with twice weekly physical therapist-led group strengthening classes. Half consumed whey protein, three times weekly, following exercise. Preliminary efficacy measures of body composition, sit-to-stand, 6-min walk and grip strength and subjective measures of self-reported health and function were also evaluated. RESULTS: Of the 37 enrolled, 28 completed the study (50 % in the protein group). Attendance rates for protein vs. non-protein groups were 89.9 ± 11.1 % vs. 95.6 ± 3.4 % (p = 0.08). Protein consumption was high in those attending classes (90.3 %) as was compliance at home (82.6 %). Whey was pleasant (67.3 ± 22.1, range 30-100, above average), had little aftertaste, and was neither salty or sticky. All were compliant (0.64 ± 0.84, range 0-5, low = higher compliance). Both groups lost significant weight (protein vs. no protein, -3.45 ± 2.86 vs. -5.79 ± 3.08, p = 0.47); Sit-to-stand, 6-min walk, and gait speed were no different, grip strength was improved in the protein compared to the non-protein group (-2.63 kg vs. 4.29 kg; p < 0.001). CONCLUSIONS: Our results suggest that whey protein is a low-cost and readily available nutritional supplement that can be integrated into a weight loss intervention.