Validation of Each Category of Kihon Checklist for Assessing Physical Functioning, Nutrition and Cognitive Status in a Community-Dwelling Older Japanese Cohort

Background: Kihon Checklist is a self-reported comprehensive health checklist used as a screening tool to identify frailty. The Kihon Checklist is a 25-item questionnaire including seven categories: daily life, physical ability, nutrition, oral condition, the extent to which one is housebound, cognitive status, and depression risk. We aimed to clarify the consistency in assessments of three important categories: physical strength, nutritional status, and cognitive function of Kihon Checklist using assessments of actual physical, nutritional, and cognitive statuses. Methods: The study sample consisted of 5341 elderly individuals aged ≥ 65 years who participated in the Japanese Long-Term Care Prevention Project. We evaluated the Kihon Checklist scores except the depression risk. Physical functioning was evaluated using handgrip strength, one-leg standing-balance time, the Timed Up & Go test, and a walking test at usual or maximum speed. Nutritional status was assessed using the Mini Nutritional Assessment questionnaire. Cognitive functioning was evaluated using Sweet 16. Associations between each category of Kihon Checklist and physical, nutritional, and cognitive functioning assessments were analyzed. Results: There were significant differences in all categories of Kihon Checklist between participants with and without functional decline in physical, nutritional, and cognitive functioning. Multivariate analyses showed that the Kihon Checklist physical strength category correlated with physical functioning assessments, the Kihon Checklist nutritional status category correlated with the Mini Nutritional Assessment, and the Kihon Checklist physical strength category correlated with the Sweet 16 scores. Moreover, the analysis of receiver operating characteristic curve exhibited a moderately accurate relationship of the Kihon Checklist physical strength category with overall physical functioning assessments. Conclusions: We found significant associations of the Kihon Checklist physical strength, nutritional status, and cognition categories, with assessments of physical, nutritional, and cognitive functioning, respectively. Especially, the Kihon Checklist physical strength category is a valid tool for predicting physical functioning for general frailty aspects in older adults.


Introduction
The Japanese population is aging (it is the most aged society in the world with 25% of the population being over 65 years old in 2013) and this rate will increase further to 40% in 2060 [1]. Aging is a continuous and multidimensional process involving an interaction of the effects of personal lifestyle such as physical, nutritional, cognitive, and social factors. Advancing age is associated with increased frailty and decline in the ability to perform activities of daily living (ADLs) in elderly individuals [2]. Even though geriatric frailty is described as global impairment of physiological reserves involving multiple organ systems, ADL decline in the elderly is due in large part to decrements in physical function [3,4]. Several studies have identified factors associated with physical function in elderly individuals [5][6][7][8]. In particular, nutritional state [7] and cognitive functioning [8] greatly influence physical functioning.
Long-term care insurance (LTCI) is a form of mandatory social insurance system that assists frail and disabled older adults with impairments in ADL. In this system, the Kihon Checklist (KCL), a selfreported comprehensive health checklist designed by a study group from the Ministry of Health, Labour and Welfare (MHLW), is used as a screening tool to identify community-dwelling older adults who are vulnerable to frailty and have a higher risk of becoming dependent [9]. Based on the results regarding impairment in specific categories, municipalities provide intervention programs to prevent future disability and promote the need for care among older adults. The KCL is a 25-item questionnaire including seven categories: daily life (five points), physical strength (five points), nutrition (two points), oral condition (three points), the extent to which one is housebound (two points), cognitive function (three points), and depression risk (five points) (Appendix Table 1). Each category is rated on a pass/fail basis and the time required for older adults to answer the KCL is approximately 15 min.  The KCL, originally developed in Japan, has been used in several studies carried out in multiple countries with distinct purposes. Sewo Sampaio et al. [10] reported that the KCL was suitable to address frailty demands among both elderly individuals who are community-dwelling and those who use daycare centers and is adequate for cross-cultural studies. However, evaluations of validity of each category of KCL with respect to assessments of actual physical, nutritional, and cognitive statuses are still unknown. In addition, the samples in such studies were small in size and limited to home and community-based services. The purpose of this study was to clarify the validity of KCL using assessments of actual physical, nutritional, and cognitive statuses. We evaluated muscle strength, walking, and both static and dynamic balance as actual physical measures. Nutritional and cognitive statuses were evaluated using the gold standard assessments, Mini Nutritional Assessment-Short Form (MNA-SF) and the 16-point Brief Cognitive Assessment Tool (Sweet 16), respectively.

Participants
Participants included 5,597 elderly individuals (mean age: 80.1 years) who consented to provide data. We collected data from the database of the Kumamoto Prefecture Community-based Rehabilitation Support System Promotion Project. This project included 17 community-based rehabilitation centers and community general-support centers in 11 regions (31 municipalities) from April 2012 to March 2013. We excluded participants aged <65 years. The Community-based Rehabilitation Support System Promotion Project was officially started in 2000 and revised in 2006, led by the Office of Elderly Health Care of the MHLW. The project participants were both healthy and in the "assistance required" category of elderly individuals, which form the lowest of the seven levels of frailty in elderly individuals who need public LTCI support because of physical and mental disabilities [11,12]. Elderly individuals certified in the "assistance required" category use community care or preventive services to lead self-supporting lives while maintaining their present physical condition as long as possible. In contrast, those certified in other "care required" categories belonging in the higher five LTCI levels can receive home-based, community-based, or institutional care services.
We recorded participants' details such as gender, age, whether they lived alone, presence of lower back pain or knee pain, history of falls and fractures, previous medical history (e.g.: hip fractures, cerebral stroke, heart disease, diabetes), and KCL scores. We evaluated KCL scores using all items except the five points for depression risk as per the MHLW criteria to identify elderly individuals who may be eligible for Japanese LTCI in the near future as a lifestyle category. Scores <10 points indicated no functional decline and 10-20 indicated functional decline (Appendix Table 1). We checked each category of physical strength (Q.6-10), nutritional status (Q. [11][12], and cognition (Q. [18][19][20]. The scores indicate functional decline in case of ≥ 3 negative answers for physical strength, 2 negative answers for nutritional status, and ≥ 1 negative answer for cognitive function. We also measured height, weight, and body mass index (BMI); evaluated physical functioning, nutritional status, and cognitive function; and conducted a geriatric assessment. Data including survey results and evaluations were stored without participants' names in the Department of Rehabilitation of Kumamoto University Hospital. This study was approved by the Institutional Review Board of Kumamoto University Hospital and was conducted in accordance with the Declaration of Helsinki.

Physical functioning
Physical functioning was evaluated according to the physical function improvement manual issued by the MHLW. Muscle strength was evaluated using handgrip strength (HGS). Physical performance was evaluated in terms of one-leg standing time with eyes open (OLS), the Timed Up & Go (TUG) test, a five-meter walking test at usual speed (WTU), and a five-meter walking test at maximum speed (WTM). HGS and OLS were measured on both sides and the better values were used in this study.

Nutritional status
Participants' nutritional state was evaluated using the MNA-SF [13] which has a total score of 14 points. Participants with scores of 0-7 were considered malnourished, 8-11 were considered at risk for malnutrition, and 12-14 were considered well nourished.

Cognitive function
Cognitive function was evaluated using the Sweet 16 including eight orientation items, three registration items, two digit spans, and three recall items [14] (Appendix Table 2). A score of 0-13 indicates cognitive impairment and 14-16 indicates no cognitive impairment. Although the items overlap with those of Mini-Mental State Examination (MMSE), all the cognitive subtests used in the Sweet 16 are widely applied independent of the MMSE, separately, and in other batteries.

Statistical analyses
Differences in continuous variables between genders and in each KCL category in the different physical strength and cognitive functioning groups were analyzed using Mann-Whitney U tests. An analysis of variance (ANOVA) and post-hoc comparisons using the Scheffé test were applied to evaluate differences in each KCL category by nutritional status. The associations between KCL lifestyle category and other participant characteristics were tested with univariate analyses using Pearson's correlation coefficients or Spearman's rankcorrelation coefficients and multivariate stepwise regressions. Using the independent variables with a significance level of 0.05, multivariate stepwise regression analysis was performed. We generated a standard receiver operating characteristic (ROC) curve for each KCL category, plotting sensitivity versus 1-specificity. The area under the ROC curve (AUC) was used to evaluate the discriminatory ability of each system to detect postoperative morbidity. Statistical tests were performed with SPSS statistics 16 software package (SPSS Inc., Chicago, IL) and the EZR (Saitama Medical Center, Jichi Medical University). http:// www.jichi.ac.jp/saitama-sct/SaitamaHP.files/statmedEN.html) [15].The probability threshold for significance was <5%.

Participant characteristics
We included 5,341 participants (82.6% female) with a mean age of 80.3 years (range, 65-102 years; Table 1). The scores in the KCL lifestyle category of participants without functional decline in all physical, nutritional, and cognitive functioning assessments were significantly inferior to those with decline in physical and cognitive status, and malnourished participants or those at risk for malnutrition ( Table 2). There were similar statistical differences in the KCL physical strength category in physical functioning and nutritional status. Additionally, the KCL scores in the nutritional status category and cognitive category were statistically inferior in the participants without functional decline in all physical functioning evaluations compared to the participants with functional decline in physical functioning. There were significant differences in scores on both KCL nutritional status and cognitive categories between well-nourished participants and those at risk for malnutrition.   Concerning each category of KCL, the multivariate analysis of participants' background showed that age, height, weight, lower back pain, knee pain, history of falls, and cerebral stroke were significantly correlated with lower scores in KCL physical strength category (Table  4). There were significant correlations of age, BMI, and history of falls with KCL nutritional status category (Table 5); gender, age, weight, and heart disease correlated with KCL cognitive category (    The ROC curves exhibited a moderately accurate relationship of the KCL lifestyle category with all physical functioning assessments ( Table  7, Figure 1). The physical strength, nutritional status, and cognitive categories of KCL were also correlated with all physical functioning assessments (Figures 2-4). However, AUCs were very low in the KCL nutritional status and cognitive categories. In addition, there were significant correlations between Sweet 16 and the KCL lifestyle, nutritional status, and cognitive categories.

Discussion
We evaluated the validity of KCL including 20 items of its lifestyle category and each category of physical strength, nutritional status, and cognitive function in community-living elderly. There were significant differences in scores in each KCL category between participants with and without functional decline seen in physical, nutritional, and cognitive functioning assessments. Multivariate analyses showed that the KCL lifestyle category was associated with age, lower back pain, history of falls, cerebral stroke, each individual physical functioning assessment (HGS, OLS, and TUG), MNA-SF, and Sweet 16. The KCL physical strength category correlated with physical functioning assessment except for WTM; the KCL nutritional status category correlated with MNA-SF; and the KCL physical strength category correlated with Sweet 16. The ROC curves exhibited a moderately accurate relationship of the KCL physical strength category with overall physical functioning assessments.
The KCL physical strength category contains 5 questionnaires including upstairs, standing, walking, and fall. We studied 4 kinds of physical functioning in this research. The OLS test is a clinical tool that assesses postural steadiness in a static position using a quantitative measurement [16]. The TUG involves components of walking, turning, and transferring from sitting to standing. The KCL questionnaires of physical ability are directly concerning OLS, TUG, and two kinds of walking tests. HGS was also associated with the KCL physical strength category. Muscle strength of handgrip was associated with the TUG, functional balance measured by the Berg Balance Scale score, and walking speed [17][18][19]. The above results may have been obtained for these reasons. Concerning the relationships between the KCL physical strength category and participants' background, the multivariate analysis showed that age, lower back pain, history of falls, and cerebral stroke were significant predictors of the KCL physical strength category. The decline in physical functioning among elderly individuals may be explained by increasing age and previous trauma history because these complications can exacerbate age-related decline in physical, social, and psychological functioning, creating a vicious circle [20].
The KCL nutritional status category has 2 questionnaires of weight loss and BMI, both of which are included in the MNA-SF that basically comprises of 5 questionnaires. Our multivariate analysis indicates that the KCL nutritional status category correlated with BMI. There were also significant associations of the KCL nutritional status category with TUG, WTU, and WTM by the multivariate analysis and with all physical functioning assessments by the ROC analyses. These results are consistent with those of previous studies [7].
The KCL cognitive category includes 3 questionnaires about memory. Sweet 16 is an interactive and actual memory test [14]. These two evaluations differ in terms of subjective and objective assessment, although both evaluate memory function. The multivariate analyses showed that the KCL cognitive category was associated with Sweet 16. There were no correlations between the KCL cognitive category and all measures of physical functioning or the NMA-SF. Atkinson et al. [21] reported that baseline global cognitive functioning and changes in global cognitive functioning were associated with changes in physical performance, but baseline physical performance was not associated with cognitive changes in their elderly sample. On the other hand, several previous studies have shown that cognitive functioning greatly influence physical functioning [8,22,23]. Furthermore, the nutritional status correlated with cognitive function [24]. Only 3 questionnaires in the KCL cognitive category may have been inadequate to predict both, physical functioning and nutritional status.
KCL lifestyle category including daily life, physical strength, nutritional status, oral function, the extent to which one is housebound, and cognitive status are useful to objectively assess frailty among elderly individuals. The geriatric assessment is a multidimensional and multidisciplinary assessment designed to evaluate an older person's functional ability, physical health, cognition, mental health, and socioenvironmental circumstances. In a crosssectional study, Fukutomi et al. [9] showed that at-risk groups in all KCL categories exhibited lower ADLs, lower subjective quality of life scores, and higher scores on a geriatric depression scale. The results of our multivariate analyses are consistent with previous studies reporting that age [19], lower back pain [25], falls [17], previous fractures [26], and cerebral stroke [27], are all factors associated with poorer scores in the KCL lifestyle category.
There are several limitations in the present study. First, there is an over-representation of women in this project. This difference may be because the population of women aged 65 and older is 1.5 times higher than that of men [28] in Japan; in addition, women tended to more actively participate in this project despite the efforts of the office of Kumamoto Prefecture Community-based Rehabilitation Support System Promotion Project to recruit all elderly residents in the prefecture. Second, as this investigation is a cross-sectional study, causal relationships could not be determined. In our cohort, longitudinal studies of changes in physical functioning may provide further information. Third, as this was an observational study, there were instances of missing data. We used a pairwise-deletion method for handling missing data [29].

Conclusion
We investigated validity of the KCL in physical strength, nutritional status, and cognitive categories against the physical, nutritional, and cognitive assessments in a group of healthy elderly participants and those requiring long-term care. We confirmed significant associations of the KCL categories of physical strength, nutritional status, and cognitive functioning with the assessments of physical functioning, nutrition, and cognitive status, respectively. The KCL physical strength category is especially a valid tool for predicting physical functioning and general frailty aspects in older adults. The KCL is recommended for use in community and clinical practice as a screening tool to assess frailty status or a higher risk of dependence because of its short questionnaire and easy administration.