PREDICT OLDER ADULTS WITH MILD COGNITIVE IMPAIRMENT BY MOBILITY AND COGNITIVE DUAL TASKS COMBINED WITH RISK FACTORS

P.-S. Lin^1,2,3, S.-M. Chang¹, Y.-H. Chiu⁴

¹Chang Gung University, Department of Physical Therapy & Graduate Institute of Rehabilitation Science, Taoyuan, Taiwan, ²Chang Gung University, Master Degree Program in the Health & Long-term Care Industry, Taoyuan, Taiwan, ³Chang Gung University, Healthy Aging Research Center, Taoyuan, Taiwan, ⁴Chang Gung University, Department of Health Care Management, Taoyuan, Taiwan

Background: Early detection and intervention to delay the onset of dementia are beneficial for patients and their caregivers and can result in substantial cost savings for healthcare systems. However, current cognition screening tools show low to moderate predictive ability with large variability, especially for early-stage screening. Since studies have shown parallel and highly correlations of cognition and motor function decline, possible motor tests or markers have been studied to try to detect persons with mild cognitive impairment (MCI) and mild dementia without being noticed.

Purpose: The purpose of the study is to try to find significantly correlated demographic and motor risk factors in detecting early cognitive impairment and to tentatively develop a cognitive risk index.

Methods: Subjects with mild cognitive impairment (N = 31) and healthy subjects (N=30) were recruited from the communities. Risk factors including sex, age, education level, body mass index (BMI), smoking, alcohol consumption, physical activity level, drugs, comorbidity, fall risk, frailty, and quality of life, and the Lawton & Brody Instrumental Activities of Daily Living Scale (IADL) were collected by questionnaire. The gait parameters of the single timed up and go task (TUG-ST), Timed up and go dual task (TUG-DT) were evaluated by the gait-up device. Comparisons between the two groups on TUG-ST and TUG-DT were analyzed by ANCOVA with adjustment of significant demographic factors. Binary logistic regression was used to combine the significant outcomes of mobility and personal characteristic risk factors into models to predict the status of the MCI.

Results: The results have shown that age (p=.004), educational level (p=.002), living status (p=.030), level of physical activity(p=.014), frailty (p=.009), and most gait parameters during TUG-ST and TUG-DT are significantly different between the 2 groups. Further multivariate logistic regression analyzes in TUG-ST shown, compared to healthy, MCI showing a significantly lower education level (> middle school vs. none, OR=0.120, p =0.029), slower walking speed (OR = 0.956, p = 0.031) and smaller foot pitch angle at heel strike (HSP) (OR=0.817, p=0.053) with AUROC = 0.80 (95% CI=0.790 - 0.969). In the TUG-DT, MCI showed significantly lower HSP (OR=0.747, p=0.031) and fewer countbacks in 10 sec. (TUG_CB) (OR=0.538, p=0.001) with AUROC=0.956 (95% CI=0.907 – 1.000) in the 100-1 task, while older (OR=1.096, p=0.022), lower level of education (> middle school v.s. none, OR=0.086, p =0.029), smaller HSP (OR = 0.802, p = 0.033) and fewer number of fruit names in 10 sec. (OR=0.337, p=0.003) with AUROC=0.922 (95% CI=0.857 – 0.986) in the dual task of naming fruits.

Conclusions: Gait parameters, especially heel strike angle and walking speed in mobility and cognitive dual tasks, and number of counting and education level were significant predictors of MCI in older adults. Further study with a larger sample size is needed.

Implications: The aforementioned parameters, walking speed, heel strike angle, number of counting, and education, could be applied in screening the high risk, and the function of the binary logistic regression may be applied to predict the probability of community-dwelling older adults becoming MCI.

Funding acknowledgements: None

Keywords:
Mild Cognitive Impairment
Timed up and go test
Dual-task

Topics:
Older people
Mental health
Community based rehabilitation

Did this work require ethics approval? Yes

Institution: Chang Gung Medical Foundation

Committee: Institutional Review Board

Ethics number: IRB No.: 202100720B0