Title
Author
DOI
Article Type
Special Issue
Volume
Issue
Physiological indexes, psychological resilience, sensory functions, and sleep quality on the cognitive function of older adults with pre-frailty: a predictive study
1Administration department, New Taipei City Health Center, 242011 New Taipei City, Taiwan
2School of Nursing, National Taipei University of Nursing and Health Sciences, 112 Taipei, Taiwan
3Outpatient department, Cardinal Tien Hospital, 23155 New Taipei City, Taiwan
*Corresponding Author(s):linda@ntunhs.edu.tw (Shu-Fang Chang); gara600815@gmail.com (Hsiao-Chi Tsai)
| History | Submitted: 21 June 2022 | Accepted: 12 October 2022 | Published: 30 December 2022 |
| Copyright: | ©2022 The Author(s). Published by MRE Press. |
Population ageing has increased the prevalence of prefrailty comorbid with cognitive impairment among older adults. However, few studies have explored the risk factors common to both prefrailty and cognitive impairment. This study determined the predictive accuracy of demographic characteristics, physiological indexes, psychological resilience, sensory function, and sleep quality on the cognitive function of older adults with prefrailty. In this cross-sectional study, the physiological indexes, psychological resilience, sensory function, sleep quality, and cognitive function of 167 community-dwelling older adults with prefrailty recruited through purposive sampling were measured. SPSS software was used for data coding and compilation. Data analysis involved the use of descriptive statistics, the independent samples t test, the chi-square test, and logistic regression. Overall, in cognitive function, there was no difference in gender but were in age, were incapable of text messaging, had a greater number of chronic diseases, were less able to perform activities of daily living, had low psychological resilience, and had depressive tendencies. In addition, Text messaging capability and depression status can all predict the cognitive impairment state of prefrail older elderly. Physiological indexes, psychological resilience, sensory function, and sleep quality can affect cognitive function in older adults with prefrailty. Meanwhile, depressive tendencies and the inability to send text messages on a mobile device constituted critical predictors of cognitive function in the participants.
Cite this article
Lu-Ling Hsieh, Shu-Fang Chang, Hsiao-Chi Tsai. Physiological indexes, psychological resilience, sensory functions, and sleep quality on the cognitive function of older adults with pre-frailty: a predictive study.Journal of Men's Health,2022,18(12):12-23 DOI:10.22514/jomh.2022.006
1. Introduction
According to “Healthy People 2020”, a white paper published by the Taiwanese government, older adults are expected to account for 24.5% and 35.5% of the national population by 2030 and 2050, up from the current 16.1%. This white paper considered older adults as one of the focus groups. A longitudinal study on the health status of middle-aged and older adults in Taiwan indicates that the prevalence of frailty among community-dwelling older adults is 7.4%, and that 38.1% of these individuals are in the prefrailty stage [1]. Dementia, the chronic and progressive decline of cognitive functions, including attention, memory, language, spatial orientation, and the ability to make and execute decisions, can induce brain damage, impairing memory, thinking, behaviour, and the ability to perform activities of daily living. Therefore, mild cognitive impairment has been recognised as a sign of very early dementia [2]. On the basis of the three-stage, five-level model, Chang et al. [3] designed a frailty prevention programme for older adults with dementia in which secondary prevention involved early detection and early treatment through health examination. Older adults with frailty were identified as a group potentially at high risk of cognitive impairment. The incidence of prefrailty comorbid with cognitive impairment among community-dwelling older adults is expected to increase with population ageing. Consequently, health problems induced by the cognitive decline associated with ageing and extended life expectancy have received considerable scholarly attention.
According to Chang and Lin [3], who conducted a systematic literature review and meta-analysis of frailty among 35,538 older people. The prevalence of frailty is higher in female than in male. Notably, the increased risk of mortality (frail compared to robust) was significantly higher for male compared to female (summary HR for male and female: 2.656 and 1.875, respectively; Q between ÷ 2 = 4.92, p = 0.027). It showed that male frail older people have a higher risk of death and need to be paid to the prevention of frailty.
Sleep problem is a common and serious status in the older population. About 50% of older adults have sleep problems [4]. Poor sleep quality was reported to be related to prevalent and incident frailty at follow-up in U.S. community-dwelling older people [5]. Sleep disturbances have been found to be associated with an increased risk of different adverse outcomes, such as cognitive impairment, depression, and death. Frailty is defined as “a state of reduced resilience to stress”. Poor sleep quality and low resilience were main related factors for frailty [6].
The presence of risk factors common to both prefrailty and cognitive impairment can be determined through health examination, after which clinical referral can proceed to control disease progression, thus reducing both medical costs and caregiving burden. This study investigated the predictability of prefrailty among older adults through the assessment of physiological indexes, psychological resilience, sensory function, sleep quality, and cognitive function.
2. Method
2.1 Participants
This cross-sectional study was conducted from September 2019 to May 2020. Community-dwelling older adults were recruited through purposive sampling. The inclusion criteria were as follows: aged ≥65 years; having sufficient ability to communicate in either Mandarin or Taiwanese; diagnosed as having prefrailty according to the Study of Osteoporotic Fractures (SOF) index; having sufficiently good hearing and oral communication abilities, as well as sufficiently clear vision, to not hinder communication with researchers or impede the data collection process (the responses of illiterate individuals were entered by researchers on their behalf). Individuals diagnosed as having dementia or determined to be nonfrail or frail through assessment were excluded.
2.2 Sample Size
G*Power version 3.1.7 software (Uiversität Kiel, Kiel, Germany) was used with α at 0.05, power at 0.80, and effect size at 0.15, and 16 variables were considered. The minimum sample size was 152. The estimated sample loss rate is about 10%, therefore, the number of samples was increased to 167.
2.3 Informed Consent
To protect the privacy as well as the rights and interests of the study participants, the study objectives, methods, and amount of time to be spent were explained to the older people. The older people could decide whether to take part in the study or withdraw midway without it having an adverse effect on the care that they received in the institution. All the written consent was obtained from participants. The research data were solely used for research purposes and strictly kept confidential. The names of the participants and the questionnaire numbers were replaced with codes prior to analyses to ensure the participants’ data security.
2.4 Research Assessment
2.4.1 Frailty
The SOF index was used to assess frailty. Its three components served as the basis on which the questionnaire items were developed. They are as follows: (1) unintentional weight loss of more than 5% in the past year; (2) inability to rise from a chair five times without using the arms; and (3) feeling poor in energy more than three times in the past week. Responses of “yes” were given 1 point, whereas responses of “no” were give 0 points, with a total score of 1 point from the three components combined indicating prefrailty [4]. Individuals with a total score of at least 1 point were enrolled as participants. Luciani et al. [5] noted that the sensitivity, specificity, and accuracy of the SOF index on older adults aged 65 years and over with early-stage cancer was 89.0%, 81.1%, and 86.5%, respectively. The predictive power of the SOF index surpasses that of the Cardiovascular Health Study index of frailty. Moreover, as mentioned, the reliability and validity of the SOF index are favourable. In addition, SOF-based approaches are easy to use, can be applied in regular community venues, and are not restricted to clinical settings. In essence, the SOF index is an ideal indicator of frailty among older adults [4]. A study conducted in Taipei, Taiwan Hu et al. [6] applied both the SOF index and Fried’s frailty phenotype on 471 community-dwelling older adults aged 65 years and over, observing that the two were moderately correlated (p < 0.001). The AUC of the SOF index was significantly correlated with the ability of the index to predict falls (p < 0.001). Furthermore, Fried’s frailty phenotype had greater accuracy in predicting disability (p < 0.001).
2.4.2 Demographic Characteristics
A structured questionnaire was developed and used to collect data on the participants’ age, sex, marital status, residential status, community activity participation, text messaging capability, self-rated health, number of chronic diseases, height, weight, and waist circumference.
2.4.3 Pain
Pain was assessed using the numeric rating scale (NRS), which rates pain from 0 to 10, with 0 and 10 indicating no and the worst pain imaginable, respectively. It can be administered verbally or on paper [7]. Ferraz et al. [7] applied the NRS on 91 patients, reporting a Pearson product-moment correlation coefficient of 0.937 between the first and second tests. Alghadir et al. [8] indicated that the NRS had excellent test–retest reliability (0.95).
2.4.4 Physical Activity
Physical leisure activity was assessed using a 2-item questionnaire, and the participants were grouped according to their responses. Regarding the first item (“Do you exercise on a regular basis?”), the response options were “no”, “less than twice a week”, “three to five times a week”, and “six or more times a week”. As for the second item (“How long do you exercise at a time?”), the response options were “15 minutes or fewer”, “16 to 29 minutes”, “30–59 minutes”, and “60 or more minutes”. The respondents answering “no” or “less than twice a week” were assigned to the low physical activity group; those answering “three to five times a week” or “16 to 29 minutes” were assigned to the moderate physical activity category; and those answering more than six times a week, 30 to 59 minutes, or 60 or more minutes were assigned to the high physical activity group. The Godin-Shephard Leisure-Time Physical Activity Questionnaire (GSLTPAQ), developed in 1985, a simple assessment comprising four items that measure the frequency of mild, moderate, and strenuous physical activity, as well as sweat-inducing exercise. Lee et al. [9] administered the GSLTPAQ to 39 community-dwelling women (mean age 51 years). Strenuous exercise score, moderate exercise score, and total score were significantly correlated with maximum oxygen uptake (r = 0.35, 0.36, and 0.58, respectively). Moreover, total score was significantly negatively correlated with body fat percentage (r = 0.37, p < 0.01), and strenuous exercise score and total score were significantly correlated with muscle endurance (r = 0.35 and 0.39, respectively; p < 0.05). The strenuous exercise score and total score had a 4-week test-retest reliability of 0.85 and 0.93, respectively.
2.4.5 Independent Living Skills
The Lawton Instrumental Activities of Daily Living (IADL) Scale, developed by Lawton and Brody in 1969, consists of eight items, namely ability to use telephone, shopping, food preparation, housekeeping, laundry, mode of transportation, responsibility for own medications, and ability to handle finances. The scoring model is dichotomous; for each item, respondents are given 1 point if they are able to partially or complete the activities described. If they partially or completely rely on others, 0 points are given. Higher scores indicate a better functional level (maximum score: 8 points). The scale has a scorer reliability of 0.85 [10]. A 2014 study evaluated the reliability and validity of the IADL scale on dementia, using the data of 60 patients with dementia obtained from the Iran Dementia & Alzheimer’s Association. Content validity was rated on a 4-point Likert scale. Isik et al. [11] used the IADL scale to evaluate 80 participants’ independent living skills (mean age 71.6 years). Internal consistency and (Cronbach α) inter-item correlation were of 0.843 and 0.915, respectively. Reliability was satisfactory, as indicated by the high inter-scorer correlation (r = 0.961, p < 0.001) between the items correlated with the total score (0.427 < r < 0.606; p < 0.001) and the intraclass correlation coefficient.
2.4.6 Psychological Resilience
The Resilience Scale for Adults (RSA), developed by Friborg et al. [12], is a semantic differential scale with excellent reliability that assesses psychological resilience and recoverability in the face of adversity on a 7-point Likert scale. It comprises 33 items that can be categorised under five factors. Chang et al. [3] translated the RSA into traditional Chinese with the developer’s consent. The traditional Chinese version contains 29 items (4 items were removed according to factor analysis results) that are categorised into five dimensions. Specifically, the personal strength, family unity, social resources, social skills, and future organisational style dimensions contain 6, 7, 8, 4, and 4 items, respectively. Scores range from 0 to 203 points, with higher scores indicating higher psychological resilience. The traditional Chinese version has favourable reliability and internal consistency, with a Cronbach α of 0.88 and a 4-month test-retest reliability between 0.69 and 0.84. In the original study by Friborg et al. [12], the participants were grouped according to their scores on the scale. Mean score differed significantly (t = 10.58, p < 0.001) between the high-score group (6.02; standard deviation (SD) = 0.45) and the low-score group (4.61; SD = 0.77). Specifically, the RSA was used to examine the relationship between psychological resilience and symptoms of posttraumatic stress disorder. The Chinese version of the RSA has excellent reliability, with an internal consistency (Cronbach α) of 0.89 and a 4-week test-retest reliability of 0.89.
2.4.7 Depression
As its name implies, the Geriatric Depression Scale is used to evaluate depressive symptoms in older adults. Sheikh and Yesavage [13] developed the Geriatric Depression Scale-Short form (GDS-SF), which consists of 15 highly correlated items from the original 30. The GDS-SF allows quick and easy measurement of depression status in the past week, and the administrator is not required to receive special training in advance. All the items are yes/no questions, and the total score ranges between 0 and 15 points. Scores from 0 to 4, 5 to 9, and ≥10 indicate no depression; mild depression, and moderate-to-severe depression, respectively. With a sensitivity of 92% and a specificity of 89%, the GDS-SF can accurately differentiate between individuals with and without depression (r = 0.84, p < 0.001). Liao et al. [10] divided 136 patients into a depression and a non-depression group (containing 45 and 91 patients, respectively) according to the Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised, reporting a correlation coefficient of 0.86, an AUC of 0.97, a sensitivity of 93.3%, and a specificity of 92.3%. In short, its excellent reliability and validity make the GDS-SF it an ideal instrument for assessing depression in older adults.
2.4.8 Sensory Function
Vision and hearing were assessed with the 4-item questionnaire used in the Taiwan Longitudinal Study on Aging (TLSA). Affirmative and negative responses are given 1 and 0 points, respectively, with a lower score indicating better sensory function [14]. In a study conducted from 1987 to 1988, Nondahl et al. [15] administered the Hearing Handicap Inventory-Screening Version (HHIE-S) to 3556 older adults who were on average aged 65.8 years and had eye diseases and hearing impairment. The sensitivity, specificity, positive predictive value, and negative predictive value of the questionnaire and four self-rated items were 0.67, 0.80, 0.87, and 0.56, respectively. In their study on 2696 participants aged 55 to 59 years, Sindhusake et al. [16] reported that the item on hearing had a sensitivity of 71%, specificity of 72%, positive likelihood ratio (PLR) of 71%, and negative likelihood ratio (NLR) of 69%. The results were confirmed using air conduction pure tone audiometry.
2.4.9 Sleep Quality
Buysse et al. [17] initially developed the Pittsburgh Sleep Quality Index (PSQI) for clinical use in patients with psychotic disorders. This self-rated questionnaire, which assesses sleep quality and sleep disturbances over a 1-month time interval, comprises 10 items under seven components: subjective sleep quality, sleep latency, sleep duration, sleep efficiency, sleep disturbance, use of sleep medication, and daytime dysfunction. In the original study, which included 168 participants, the PSQI achieved a Cronbach α of 0.83, a specificity of 89.5%, and a sensitivity of 89.6% (κ = 0.75, p < 0.001). Scores on each item range from 0 to 3 points, and the total score ranges between 0 and 21 points. Scores exceeding 5 points indicate unsatisfactory sleep quality, and the higher the score, the poorer the sleep quality.
2.4.10 Dementia
The Ascertain Dementia 8-item Informant Questionnaire (AD-8) was developed by researchers at Washington University in St. Louis. A Chinese version is used in Taiwan. Affirmative responses to more than two of the eight items indicates a high risk of dementia [18]. Yang et al. [19] tested the predictive accuracy of the AD-8 in discriminating between patients with Alzheimer disease from healthy individuals in Taiwan. Under a threshold for cognitive impairment of ≥2, the AUC, sensitivity, specificity, PLR, and NLR were 0.961, 97.6%, 78.1%, and 0.03, respectively.
2.5 Statistical Analysis
The present analyses were performed using IBM SPSS Statistics for Windows, version 22 (IBM Corp., Armonk, NY, USA). The demographic characteristics were analysed using descriptive statistics, and the questionnaire items were analysed by examining the relationship between linked items. The cognitive function data are presented as means and standard deviations. The independent samples t test was conducted to compare cognitive function among participants with different demographic characteristics. The chi-square test was applied to data on cognitive function and the physiological, psychological, sensory, and sleep variables. Logistic regression was performed to identify predictors of cognitive function. Nagelkerke’s R2 was obtained as an estimation of the total log likelihood explained by a summation of the significant occlusal factors.
3. Results
3.1 Demographic Characteristics
The 167 participants were aged between 65 and 89 years, with a mean age of 74.31 years (SD = 6.408 years). The participants aged 65 to 74 years constituted the largest group (88 individuals), and men and women (69 and 98 individuals) accounted for 41.3% and 58.7% of the participants, respectively. Most of the participants were married (105; 62.9%), lived with family members (152; 91.0%), and capable of text messaging (104; 62.3%). Constituting the largest proportion, 68 and 81 of the participants (40.7% and 48.5%) rated their health as moderately good and reported having two to three chronic diseases, respectively (Table 1).
| Variable | Number | Percentage | Mean | SD | |
| Age (years) | 74.31 | 6.408 | |||
| 65–74 | 88 | 52.7 | |||
| 75–84 | 68 | 40.7 | |||
| ≥85 | 11 | 6.6 | |||
| Sex | |||||
| Male | 69 | 41.3 | |||
| Female | 98 | 58.7 | |||
| Marital status | |||||
| Married | 105 | 62.9 | |||
| Divorced | 7 | 4.2 | |||
| Unmarried | 8 | 4.8 | |||
| Widowed | 47 | 28.1 | |||
| Residential status | |||||
| Living alone | 15 | 9.0 | |||
| Living with family | 152 | 91.0 | |||
| Community activity participation | |||||
| Yes | 142 | 85.0 | |||
| No | 25 | 15.0 | |||
| Text messaging capability | |||||
| Yes | 104 | 62.3 | |||
| No | 63 | 37.7 | |||
| Health (Self-rated) | |||||
| Very good | 33 | 19.8 | |||
| Good | 55 | 32.9 | |||
| Moderately good | 68 | 40.7 | |||
| Not very good | 11 | 6.6 | |||
| Very bad | 0 | 0 | |||
| Number of chronic diseases | |||||
| ≤1 | 78 | 46.7 | |||
| 2 to 3 | 81 | 48.5 | |||
| ≥4 | 8 | 4.8 | |||
| Physiological indexes | |||||
| Height | 159.31 | 8.933 | |||
| Weight | 64.49 | 12.74 | |||
| Waist circumference | 61.5 | 10.252 | |||
| Normal (men ≤90 cm, women ≤80 cm) | 109 | 65.3 | 1.53 | 0.501 | |
| Excessive (men ≥90 cm, women ≥80 cm) | 58 | 34.7 | 1.69 | 0.467 | |
| Instrumental activities of daily living | |||||
| Yes | 14 | 8.4 | 3.71 | 1.858 | |
| No | 153 | 91.6 | 7.62 | 0.639 | |
| Physical activity | |||||
| Low | 51 | 30.5 | 2.00 | 0.000 | |
| Moderate | 92 | 55.1 | 3.43 | 0.498 | |
| High | 24 | 14.4 | 5.29 | 0.464 | |
| Pain (Numeric rating scale score) | 1.99 | 1.668 | |||
| 0 | 42 | 25.1 | |||
| 1 | 21 | 12.6 | |||
| 2 | 46 | 27.5 | |||
| 3 | 36 | 21.6 | |||
| 4 | 10 | 6.0 | |||
| 5 | 6 | 3.6 | |||
| 6 | 2 | 1.2 | |||
| 7 | 3 | 1.8 | |||
| 8 | 1 | 0.6 | |||
| Psychological resilience | |||||
| Low score group | 2 | 1.2 | 1.00 | 0.000 | |
| Moderate score group | 56 | 33.5 | 0.66 | 0.478 | |
| High score group | 109 | 65.3 | 0.35 | 0.479 | |
| Geriatric depression | |||||
| No depression | 124 | 74.3 | 2.23 | 0.989 | |
| Mild depression | 34 | 20.4 | 6.35 | 1.252 | |
| Moderate-to-severe depression | 9 | 5.4 | 11.78 | 1.481 | |
| Sensory function | |||||
| Wearing glasses | 111 | 66.5 | |||
| Not wearing glasses | 56 | 33.5 | |||
| Able to see clearly | 110 | 65.9 | |||
| Unable to see clearly | 57 | 34.1 | |||
| Wearing hearing aids | 16 | 9.6 | |||
| Not wearing hearing aids | 151 | 90.4 | |||
| Able to hear clearly | 117 | 70.1 | |||
| Unable to hear clearly | 50 | 29.9 | |||
| Sleep quality (Pittsburgh sleep quality index score) | |||||
| Good sleep quality (0–5 points) | 38 | 22.8 | 2.84 | 1.079 | |
| Poor sleep quality (6–21 points) | 129 | 77.2 | 8.29 | 3.284 | |
| Cognitive function | |||||
| Normal | 90 | 53.89 | 0.30 | 0.461 | |
| Impaired | 77 | 46.11 | 2.84 | 1.319 |
The average height, weight, and waist circumference was 159.31 cm (SD = 8.933), 64.49 kg (SD = 12.74), and 81.44 cm (SD = 10.252), respectively. Pain scores of 2 were reported by the largest group—46 participants (27.5%). The IADL scores revealed that the vast majority of the participants (153; 91.6%) experienced no difficulty in performing activities of daily living. The mean IADL score was 7.62 (SD = 0.639). The RSA scores ranged between 29 and 203. The mean score was 159.59 (SD = 30.633). The mean GDS-SF score was 3.59 (SD = 2.777), with 74.3%, 20.4%, and 5.4% of the participants being categorised as normal (0–4 points), having mild depression (5–9 points), and moderately to severely depressed, respectively (Table 1).
Regarding vision, 111 (65.9%) of the participants noted that they wore glasses and 110 (65.9%) reported being able to see clearly. As for hearing, 16 (9.6%) of the participants noted that they wore hearing aids, and 117 (70.1%) reported being able to hear clearly. Mean score on sleep quality was 7.05 (SD = 3.717). Most of the participants (129; 77.2%) had poor sleep quality. The mean AD-8 score was 1.47 (SD = 1.590). Among the participants, 90 and 77 (53.89%) were considered having normal and impaired cognitive function, respectively (Table 1).
3.2 Participant Differences in Demographic Characteristics, Physiological Indexes, Psychological Resilience, Sensory Function, Sleep Quality, and Cognitive Function
Age (χ2 = 20.540, p < 0.05), text messaging capability (χ2 = 22.932, p < 0.05), and number of chronic diseases (χ2 = 6.196, p < 0.05) were significantly correlated with cognitive function (p < 0.05), indicative of a correlation with cognitive function. Moreover, this suggests that individuals with prefrailty who are older or have more chronic diseases are more likely to have cognitive impairment, and that those capable of text messaging are more likely to have normal cognitive function than those of who not. In terms of physiological indexes, IADL score correlated significantly with cognitive function (χ2 = 3.943, p < 0.05), indicating that a high IADL score (disability) is associated with cognitive impairment. Psychological resilience (χ2 = 16.867, p < 0.05), depression tendency (χ2 = 24.352, p < 0.05), and sleep quality (χ2 = 7.735, p < 0.05) correlated significantly with cognitive function, clearly indicating that the psychological indexes and sleep of older adults with prefrailty are related to their cognitive function. With the correlation coefficient of psychological resilience and geriatric depression attending statistical significance, those older adults with prefrailty whose sleep quality has a low score are more prone to cognitive impairment (Table 2).
| Variable | Cognitive function | χ2 | p | ||
| Normal | Impaired | ||||
| n (%) | n (%) | ||||
| Age (years) | 20.540 | <0.001 | |||
| 65–74 | 62 (37.1) | 26 (15.6) | |||
| 75–84 | 24 (14.4) | 44 (26.3) | |||
| ≥85 | 4 (2.4) | 7 (4.2) | |||
| Sex | 1.741 | 0.187 | |||
| Male | 33 (19.8) | 36 (21.6) | |||
| Female | 57 (34.1) | 41 (24.6) | |||
| Marital status | 3.569 | 0.313 | |||
| Married | 61 (36.5) | 44 (26.3) | |||
| Divorced | 2 (1.2) | 5 (3.0) | |||
| Unmarried | 3 (1.8) | 5 (3.0) | |||
| Widowed | 24 (14.4) | 23 (13.8) | |||
| Residential status | 1.280 | 0.258 | |||
| Living alone | 6 (3.6) | 9 (5.4) | |||
| Living with family | 84 (50.3) | 68 (40.7) | |||
| Community activity participation | 0.053 | 0.819 | |||
| Yes | 76 (45.5) | 66 (39.5) | |||
| No | 14 (8.4) | 11 (6.6) | |||
| Text messaging capability | 22.932 | <0.001 | |||
| Yes | 71 (42.5) | 33 (19.8) | |||
| No | 19 (11.4) | 44 (26.3) | |||
| Health (Self-rated) | 8.582 | 0.055 | |||
| Very good | 22 (13.2) | 11 (6.6) | |||
| Good | 34 (20.4) | 21 (12.6) | |||
| Moderate | 31 (18.6) | 37 (22.2) | |||
| Not very good | 3 (1.8) | 8 (4.8) | |||
| Very bad | 0 (0.0) | 0 (0.0) | |||
| Number of chronic diseases | 6.196 | 0.040* | |||
| ≤1 | 51 (30.5) | 29 (17.4) | |||
| 2 to 3 | 36 (21.6) | 43 (25.7) | |||
| ≥4 | 3 (1.8) | 5 (3.0) | |||
| Waist circumference | 54.366 | 0.064 | |||
| Men, normal (≤90 cm) | 22 (13.2) | 29 (17.4) | |||
| Men, excessive (>91 cm) | 11 (6.6) | 7 (4.2) | |||
| Women, normal (≤80 cm) | 28 (16.8) | 30 (18.0) | |||
| Women, excessive (>81 cm) | 29 (17.4) | 11 (6.6) | |||
| Impaired | 4 (2.4) | 10 (6.0) | |||
| Not impaired | 86 (51.5) | 67 (40.1) | |||
| Physical activity | 24.352 | 0.982 | |||
| Low | 28 (16.8) | 23 (13.8) | |||
| Moderate | 49 (29.3) | 43 (25.7) | |||
| High | 13 (7.8) | 11 (6.6) | |||
| Pain | 3.499 | 0.176 | |||
| Mild (≤3) | 28 (16.8) | 23 (13.8) | |||
| Moderate (4–6) | 49 (29.3) | 43 (25.7) | |||
| Severe (≥7) | 13 (7.8) | 11 (6.6) | |||
| Psychological resilience | 16.867 | <0.001 | |||
| Low score group (score: <86) | 0 (0.0) | 2 (1.2) | |||
| Moderate score group (score: 87–145) | 19 (11.4) | 37 (22.2) | |||
| High score group (score: >146) | 71 (42.5) | 38 (22.8) | |||
| Geriatric depression | 24.352 | <0.001 | |||
| No depression (score: 0–4) | 80 (47.9) | 44 (26.3) | |||
| Mild depression (score: 5–9) | 10 (6.0) | 24 (14.4) | |||
| Moderate-to-severe depression (score: ≥10) | 0 (0.0) | 9 (5.4) | |||
| Wearing glasses | 0.514 | 0.474 | |||
| Yes | 62 (37.1) | 49 (29.3) | |||
| No | 28 (16.8) | 28 (16.8) | |||
| Able to see clearly | 3.505 | 0.061 | |||
| Yes | 65 (38.9) | 45 (26.9) | |||
| No | 25 (15.0) | 32 (19.2) | |||
| Wearing hearing aids | 0.733 | 0.392 | |||
| Yes | 7 (4.2) | 9 (5.4) | |||
| No | 83 (49.7) | 68 (40.7) | |||
| Able to hear clearly | 2.811 | 0.094 | |||
| Yes | 68 (40.7) | 49 (29.3) | |||
| No | 22 (13.2) | 28 (16.8) | |||
| Sleep quality | 7.755 | 0.005** | |||
| Good | 28 (16.8) | 10 (6.0) | |||
| Poor | 62 (37.1) | 67 (40.1) | |||
| Note: *p < 0.05, **p < 0.01. |
3.3 Predictors of Cognitive Function in Community-Dwelling Older Adults with Prefrailty
Logistic regression analysis was performed to determine the predictive accuracy of the independent variables—that is, the demographic characteristics (age, text messaging capability, and number of chronic diseases), physiological indexes (IADL score), psychological indexes (psychological resilience and geriatric depression), and sleep quality—regarding cognitive function. Text messaging capability attained statistical significance (β = −0.981, OR: 0.375, 95% CI: 0.149–0.941, p < 0.05) with cognitive impairment. In other words, the participants incapable of using mobile devices had a higher risk of cognitive impairment than those capable. The participants with depression were 4.188 times more likely to have cognitive impairment than those without depression (β = 1.432, OR: 4.188, 95% CI: 1.838–9.539) (Table 3), with a mean of 43.75% for specificity, 52.95% for sensitivity and an accuracy of 52.25%. The percentage of the total log likelihood for disc displacement with reduction explained by the significant occlusal factors was low, with a Nagelkerke’s R2 = 0.124.
| Independent variable | β | S.E. | Wald | p | OR | 95% CI of OR |
| Constant | –3.581 | 3.106 | 1.329 | 0.249 | 0.028 | |
| Age | 0.060 | 0.037 | 2.599 | 0.107 | 1.061 | 0.987–1.141 |
| Text messaging capability | –0.981 | 0.469 | 4.370 | 0.037* | 0.375 | 0.149–0.941 |
| Number of chronic diseases | 0.099 | 0.197 | 0.251 | 0.616 | 1.104 | 0.751–1.622 |
| Instrumental activities of daily living | –0.500 | 0.769 | 0.422 | 0.516 | 0.607 | 0.134–2.740 |
| Psychological resilience | –0.005 | 0.007 | 0.549 | 0.459 | 0.995 | 0.981–1.009 |
| Geriatric depression | 1.432 | 0.420 | 11.625 | 0.001*** | 4.188 | 1.838–9.539 |
| Sleep quality | –0.361 | 0.488 | 0.546 | 0.460 | 0.697 | 0.268–1.814 |
| Note: β is the unstandardised regression coefficient. S.E., standard error; OR, odds ratio; CI, confidence interval. *p < 0.05, ***p < 0.001. |
4. Discussion
As mentioned, the mean age of the participants was 74.31 years, and most of them were female, married, and living with family. Moreover, most of them participated in community activities, were capable of text messaging, considered themselves moderately healthy, and reported having 2 to 3 chronic diseases. In a study conducted in 2015 in Miaoli, Taiwan, 94 of the 275 middle-aged and older community-dwelling adults had prefrailty [20]. Those participants’ age and number of chronic diseases were relatively similar to those of the present participants. Studies have confirmed that ageing, history of stroke, and depression tendency can result in a high risk of cognitive impairment, and that community-dwelling older adults who frequently participate in community activities, which present opportunities for physical activity and social interaction, have significantly better cognitive function [21]. In a study by Nakai et al. [22], the number of participants with high RSA scores was similar to that of those considered to have no depression. In a study on older adults by Chen and Chang [23], most (52.3%) of the participants had poor sleep quality (PSQI score >5), and sleep duration was found to affect body mass index and sleep quality. On average, the present participants went to bed between 9 and 10 PM, fell asleep within 30 minutes, and slept for 6 hours. These results are in line with those from another study on community-dwelling older adults in Taiwan [24], in which the mean PSQI score and mean sleep duration were 5.73 and 6 hours, respectively, and 55% of their participants slept poorly (PSQI score >5).
The AD-8 results revealed that 90 and 77 (53.89% and 46.11%) of the present participants had normal and impaired cognitive function, respectively. Similarly, in a cross-sectional study on the relationship between frailty, cognition, and self-perceived decline in cognitive function, 174 and 132 participants (43%) were determined to have normal and mildly impaired cognitive function, respectively, suggesting correlations between frailty, cognitive impairment, and self-perceived decline in cognitive function. Robertson et al. [25] reported that cognitive function was significantly poorer in older adults with prefrailty than in their nonfrail counterparts (p < 0.05). Furthermore, self-rated cognitive function was correlated with actual cognitive function. The proportion of older adults with prefrailty and cognitive impairment (46.11%) in the present study is comparable to corresponding results from studies conducted in other countries.
4.1 Participant Differences in Demographic Characteristics, Physiological Indexes, Psychological Resilience, Sensory Function, Sleep Quality, and Cognitive Function
In the present study, age, text messaging capability, and number of chronic diseases were correlated with cognitive function. Specifically, the participants who were more advanced in age, incapable of text messaging, and had more chronic diseases were more likely to have cognitive impairment. In a study by Turner Goins et al. [26], older age, being female, a higher number of chronic diseases, and a low IADL score were identified as factors significantly correlated with prefrailty in community-dwelling older adults. Few studies have explored the relationship between mobile device use and cognitive function of older adults with prefrailty. Navabi et al. [27] conducted a questionnaire survey on 328 older adults, noting that 80% and 20% of the participants used regular mobile phones and smartphones, respectively. Moreover, the participants mostly used mobile devices to make calls, and only 7% knew how to send text messages online. These results differ considerably from what was observed in the present study—that 62.3% of the participants were capable of text messaging. The obvious connection between physiological indexes and cognitive function that the lower the IADL scores, the higher the likelihood that they were cognitively impaired. A higher proportion of the participants who scored high in depression were cognitively impaired. Cognitive impairment and depression are similar in pattern. The prevalence of geriatric depression and dementia, both highly complex conditions requiring extensive care and a high number of medical resources, expected to increase in the future [28]. The correlation coefficient of psychological resilience was negative and attained statistical significance (p < 0.05), indicating that the participants had low psychological resilience overall. A cross-sectional study on 121 institutionalised older adults found that psychological resilience was moderately high and significantly positively correlated with physiological and psychological health [29]. By contrast, most of the present participants had high psychological resilience, but a similar trend was observed in that individuals with low psychological resilience had poorer cognitive function.
In the present study, the participants who slept poorly were more likely to have cognitive impairment than those who slept well. Ma et al. [30] indicated that overall PSQI score was not correlated with cognitive function, but in the component of sleep quality, poor sleep efficiency was significantly correlated with poor cognitive function. This contrasts with the present finding of a correlation between sleep quality and cognitive function. Notably, the greatest differences in the seven components of the PSQI between the participants with and without cognitive impairment were in their sleep duration.
4.2 Demographic Characteristics, Physiological Indexes, Psychological Resilience, Sensory Function, and Sleep Quality in Predicting Cognitive Function
Older age, text messaging capability, number of chronic diseases, IADL score, psychological resilience, geriatric depression, and sleep quality were significantly correlated with the participants’ cognitive function. Predictor analysis revealed that the participants who were incapable of text messaging and had depressive tendencies were more likely to have cognitive impairment. The literature indicates that text messaging capability and geriatric depression are significant predictors of cognitive impairment in older adults with prefrailty. For example, Almeida [31] noted that social media use lowered the risk of dementia in older adults. However, few studies have investigated the relationship between mobile device use and cognitive function in older adults with prefrailty. In the present study, 62% of the participants were capable of text messaging, and not being capable of this task was associated with an increased risk of cognitive impairment. Therefore, courses on using the Internet and mobile devices should be promoted, both to increase older adults’ exposure to social media and familiarise them with these platforms and their functions. The present results indicate that most of the participants did not have depression, but that those with depressive tendencies were more likely to be cognitively impaired, in line with the findings of Nakai et al. [22]. According to studies on prefrailty, frailty, and geriatric depression, prefrailty can aggravate depressive symptoms [32, 33, 34]. The simplification of scales for geriatric depression and cognition would facilitate the identification of residents with these conditions. The presentation of assessment results and referral information in residents’ files can expedite access to relevant resources. In addition, attending courses on Internet or mobile device use can help older adults socialise with others and receive mental support.
This study investigated the accuracy the demographic characteristics, physiological indexes, psychological resilience, sensory function, and sleep quality in predicting cognitive function in older adults with prefrailty. According to the AD-8 results, 77 (46.11%) of the 167 participants had cognitive impairment. Statistical analysis revealed that factors affecting participants’ cognitive state include older age, text messaging capability, number of chronic diseases, IADL score, physiological indexes, and sleep quality. As mentioned, the participants were aged between 65 and 89 years. Those aged 65 to 74 years accounted for the largest proportion of the total: 52.7%. Cognitive impairment was more common among the participants of more advanced age, consistent with the evidence that the prevalence of cognitive impairment increases with age. Participants with 2 to 3 chronic diseases constituted the largest proportion, and only eight individuals had four or more chronic diseases. Cognitive impairment was more common among participants with more chronic diseases than among those with fewer [35, 36]. A total of 104 participants (62.3%) were capable of text messaging, and cognitive impairment was more common among those who were not. Regarding the scores on the IADL scale, most of the participants (153; 91.6%) did not experienced difficulty in performing activities of daily living, and those that did were more likely to be cognitively impaired. Based on the demographic characteristics and physiological indexes, cognitive impairment clearly correlated with the structural or functional decline of body organs or systems. Cognitive impairment is also caused by brain degeneration. Functional and systemic decline that occur with age cannot be reversed. However, adequate systemic protection and the introduction of a mental or physical training programme in middle age or older age can delay the damage attributable to ageing [37, 38].
In the present study, psychological resilience and geriatric depression were correlated with cognitive impairment. The literature indicates that enhancing psychological resilience can delay the development of frailty and improve both health and quality of life. This premise supported by the present findings; most of the participants had high psychological resilience, of whom 71 did not have cognitive impairment. Moreover, a high score in geriatric depression indicated tendencies to develop cognitive impairment. In addition, poor mental health, lack of support from friends or family, and lack of interpersonal interactions were determined to be critical factors affecting cognitive function. Social support and positive interpersonal interactions may constitute the best mental health protection for community-dwelling older adults [39, 40, 41]. Care facilities are advised to provide courses held in various venues to allow gatherings among residents and thereby increase the care and support they receive from those around them.
Most of the participants with poor sleep quality had cognitive impairment (67; 40.1%). The most substantial differences in the seven components of the PSQI between the participants with and without cognitive impairment were in their sleep duration. This supports the evidence that sleep impairment can considerably affect both frailty and cognitive function. In this regard, the influence of sleep quality cannot be underestimated. Older adults tend to doze off or take naps during the day if they have no arranged activities; some of them also like to watch television in bed or on the couch. Increasing the frequency of daytime activities and outings can help older adults expend their energy, thereby benefiting sleep duration and sleep quality in general [42, 43, 44]. The present findings serve as a reference for the Taiwanese government with regard to the prevention and mitigation of cognitive impairment in older adults with prefrailty and minimizing the costs borne by individuals, families, and society as a whole.
Predictor analysis indicated that older adults who were incapable of text messaging and had depressive tendencies were more likely to be cognitively impaired. For reference, 62% of the participants were capable of text messaging. Courses on the use of mobile devices and the Internet should be promoted to encourage older adults to use mobile devices in daily life and equip them with the necessary skills. Familiarisation with social media allows older adults to socialise and receive mental support, thereby reducing their risk of developing cognitive impairment.
4.3 Limitations
There are some limitations. This a cross sectional study and one of the aims of this study was to explore the relationship between text messaging capability and cognitive function. We found they are high correlation. Due to it is not cohort study, the causal relationship is not test. Restrictions imposed by the lack of human resources and the COVID-19 pandemic meant that participants could only be recruited from a single care facility in New Taipei City, Taiwan. Thus, the generalizability of the present findings may be limited.
5. Conclusions
Physiological indexes, psychological resilience, sensory function, and sleep quality can affect cognitive function in older adults with prefrailty. Also, depressive tendencies and the inability to send text messages on a mobile device constituted critical predictors of cognitive function in the participants.
Author contributions
SFC made substantial contributions to research conception. SFC also designed the draft of the research process and submitted the manuscript as corresponding author. LLH made substantial contributions to analysis and interpretation of data. SFC, LLH and HCT had been involved in revising manuscript critically for important intellectual content. SFC, LLH and HCT modified the manuscript format, discussed reviewer opinions, and clarified the professional name. All authors read and approved the final manuscript.
Ethics approval and consent to participate
This study passed the review and ethical approval by the Behavioral and Social Sciences Research Ethics Office of National Taiwan University (IRB-Reference Code: 202002ES013) in Taiwan. All methods were performed in accordance with the relevant guidelines and regulations. All the written consent was obtained from participants. The research data were solely used for research purposes and strictly kept confidential.
Acknowledgment
We are thankful to all respondents of this study and workers from the community dwelling older people for their cooperation.
Funding
This study was founding by Cardinal Tien Hospital Grant CTH111AK-NHS-2232, National Science and Technology Council Grant MOST 111-2314-B-227-003 and Saint Paul’s Hospital Grant 110D003-3 in Taiwan.
Conflict of interest
The authors declare no conflict of interest.
