The development of multidimensional screening model of geriatric syndrome for community-dwelling older adults: results of the Taiwan Integration Health and Welfare (TIHW) Study

Background: Comprehensive geriatric assessment (CGA) is a multidimensional and multidisciplinary diagnostic and treatment process that identifies geriatric syndrome in older adults. However, CGA program is not appropriate in community screening. To the best of our knowledge, there is no applicable multidimensional screening model to evaluate geriatric syndrome in community-dwelling older adults. This study aimed to identify the risk factors of geriatric syndrome among physical function tests, socioeconomic status, medical history, and healthy behaviours in community-dwelling older adults and develop a multidimensional prediction model for community screening. Methods: A total of 1313 community-dwelling older adults aged 60 years or above were recruited from 58 communities in four aging cities in Taiwan. Geriatric syndrome was defined by disability using Instrumental Activities of Daily Livings, cognitive impairment using Short Portable Mental Status Questionnaire, depression using Geriatric Depression Scale, or by receiving mild disability card. The cutoff values of the physical function tests were calculated using receiver operating characteristic analysis. Multivariate logistic regression was used to evaluate the risk factors of geriatric syndrome, and the risk model was developed using stepwise logistic regression. Results: We developed the new cutoff values in predicting geriatric syndrome for dominant handgrip strength test, 6-meter walk, and timed up-and-go tests, which were significantly associated with geriatric syndrome. Moreover, male sex, obesity, absence of labour activities, and participants who cannot report personal information, had depressive mood for the past 2 weeks, and a history of heart disease were associated with geriatric syndrome. Finally, we developed Taiwan Risk Scores for Geriatric Syndrome (TRSGS) with the cutoff value of 6 (sensitivity, 77.2%; specificity, 75.5%). Conclusions: Most of the screening tools focus on specific problems such as sarcopenia,

dementia, or frailty. The TRSGS model demonstrated a multidimensional prediction model, which could be applied in community screening for geriatric syndrome. Management of risk factors to prevent geriatric syndrome in the community is important.

Background
Geriatric syndrome is composed of multifactorial health problems resulting in multiple systems, which is common in older adults [1].The four independent predisposing factors for geriatric syndrome were older age, baseline cognitive impairment, baseline functional and sensory impairment, and impaired mobility, which often lead to an increased risk of poor health outcomes, including fall incidence, disability, hospitalisation, and mortality [2][3][4]. The prevalence of the geriatric syndrome with multimorbidity is about 10.4% [5]. The phenotype of geriatric syndrome includes frailty, sarcopenia, falls, weight loss, delirium, depression, dementia, urinary incontinence, oral problems, malnutrition, pressure ulcer, sexual dysfunction, visual loss, and functional impairment [1,6,7]. Management focusing on geriatric syndrome is effective in improving outcomes, but physicians often fail to recognise geriatric syndrome and appropriately perform therapeutic strategies [8,9]. This implies that we can stop this vicious circle by the early identification of geriatric syndrome and prevention of further disability.
Several screening tools have been developed and validated in predicting frailty, depression, dementia, or sarcopenia [10][11][12]. However, each model only predicted certain parts of the geriatric syndrome. In clinical practice, comprehensive geriatric assessment (CGA) is a multidimensional and multidisciplinary diagnostic and treatment process that identifies geriatric syndrome in older adults [13]. Moreover, a brief assessment tool was specifically designed for geriatric syndromes identification in general practice for general practitioners in the previous study [14]. The brief assessment tool was constructed, based on simple validated tests to detect each of these geriatric syndromes [15]. Additionally, a previous study also tried to use online geriatric assessment in acutely hospitalised patients [16]. However, these assessment tools were not appropriate in community screening. To the best of our knowledge, there is no applicable multidimensional screening model that predicts geriatric syndrome. Therefore, the present study aimed to develop an applicable multidimensional screening model for community-dwelling older adults in predicting geriatric syndrome.

Community and participant enrolment
To manage the health issues of the aging society in Taiwan [17], the Ministry of Taiwan supported the Taiwan Integration of Health and Welfare (TIHW) study to conduct a formal survey in 2017, which aimed to establish a screening strategy for geriatric syndrome in the community. First, of the 14 aging cities in Taiwan, 4 cities (Pingtung County, Tainan City, Changhua County, and Miaoli County) were selected from southern to northern part of Taiwan. Second, 7-14 communities in each city were selected, resulting in a total of 58 communities. Finally, the invitation (phone call and publicity leaflets) was sent to community-dwelling older adults (≧60 years) within the selected communities by the chief of the community. The participants were gathered in their local affiliated community centres for face-to-face interviews and assessments by trained investigators from Pingtung, Tainan, Changhua, or Miaoli Hospital, Ministry of Health and Welfare.
Ethical approval and study protocol the study and provided informed consent for inclusion in the study. Participants who could read and write signed the written informed consents, and those who could not read and write impressed the name stamps or handprint with the assistance of their family members.
The evaluation in the TIHW study comprised multi-dimensional questionnaires, physical function tests, and blood examinations. The TIHW study questionnaire (Approved Case No. Definition of geriatric syndrome and exclusion criteria Geriatric syndrome was defined by disability using IADLs, cognitive impairment using SPMSQ, depression using GDS or by receiving mild disability card. The mild disability card was given to patients with mild disability, such as mild cognitive impairment, low eye vision, or mild hearing impairment (auditory brainstem evoked response threshold between 55 and 69 dB) [18]. We considered that IADLs, SPMSQ, GDS, and disability card could integratedly evaluate cognitive impairment, baseline functional and sensory impairment, and impaired mobility of geriatric syndrome. We excluded participants with a history of cerebrovascular disease, aged < 60 years, and who received moderate to severe disability card.
The IADL scale is a valuable tool to evaluate older individuals with early-stage disease. In the current study, the IADL disability status was assessed with 5 activities from an IADL scale proposed by Lawton and Brody [19]. The sum of IADL score ≧3 was defined as IADL disability [20]. Cognitive function was measured using the nine-item SPMSQ with total scores ranging from 0 to 9 [21]. Participants with four or more errors were defined as having cognitive impairment [22]. Incident depression was considered to be present when the score of GDS-15 was greater than 5 [23].

Questionnaires and physical function tests
Questions related to socioeconomic status and medical history in the questionnaires aimed to collect data on participants' educational level, marital status, living status, medical history (hypertension, diabetes, liver disease, hyperlipidaemia, and chronic kidney disease), and socioeconomic status. Socioeconomic status was self-reported and was not based on real personal income. Moreover, the evaluation of healthy behaviours included exercise habits, smoking, and alcohol and areca nut consumption. Additionally, questions such as 'Do you have low interest in doing things for the past 2 weeks?' and 'Do you have depressive mood for the past 2 weeks?' were also evaluated. We also documented whether participants could remember their personal information correctly, including age, address, and telephone number.
Physical performance was assessed through the dominant handgrip strength test, TUG test, 6MWT, and FTSST. Dominant handgrip strength was evaluated using a handheld dynamometry, and the result was expressed in kilogrammes (kg) [24]. The TUG test assesses transitional function, strength, agility, and dynamic balance [25]. In the TUG test, the participant was sitting with hands at starting position resting on the arms of the chair. The participants crossed the line, which was 3 m distant to the starting point, before turning around and walked back to sit down in the chair again. The participants were instructed to walk at their normal speed, and they performed one trial before they were timed. The timing of the TUG started when the participants' back came off the back of the chair and stopped when their buttocks touched the seat of the chair again [25]. The FTSST assesses functional lower extremity strength, transitional movements, balance, and fall risk [26]. The 6MWT was recommended as a functional assessment tool that predicts sarcopenia and fall [11,27]. The participants completed a six-meter walk measured in seconds at their normal walking speed. The study aimed to assess the validity and establish the cutoff values of the functional performance tests in predicting geriatric syndrome.

Statistical analyses
Descriptive statistics were calculated for the demographic characteristics of the participants. The receiver operating characteristic (ROC) curve and Youden's index were used to evaluate the diagnostic accuracy and to determine the cutoff values for handgrip strength, TUG test, FTSST, and 6MWT in predicting geriatric syndrome [28]. Multivariate regression models were used to investigate the risk factors of geriatric syndrome among physical function tests, socioeconomic status, healthy behaviours, and medical history.
Risk scores were selected by stepwise logistic regression, and the validity of the prediction model was established by ROC analysis. All these analyses were performed using the Statistical Analysis System (SAS) version 9.4 (SAS Institute Inc., Cary, NC, USA).

Results
A total of 1449 participants were recruited in the TIHW study. We excluded 136 participants with a history of cerebrovascular disease, aged < 60 years, and who received moderate to severe disability card. Of the 1313 enrolled participants, 167 (12.7%) and 8 non-geriatric syndrome group, respectively. Participants in the geriatric syndrome group had older age, poorer educational level, higher rate of living with family (76.0% vs. 64.6%), lesser exercise habits (47.3% vs. 72.9%), more possibility to receive social assistance pension (10.8% vs. 3.4%), and higher rate of heart disease history than that of the non-geriatric syndrome group (Table 1)  According to the ROC analysis, the 6MWT had the largest AUC among the tests. The figures associated with the statistics and cutoff values for each test are shown in Figure 1 and 2. We applied our cutoff values for each test in the further risk evaluation. Table 2 shows the adjusted odds ratio for each potential risk factor calculated by multivariate logistic regression. In a fully adjusted model, male, obesity, depressive mood for the past 2 weeks, poor remembering in personal information, absence of exercise habit, heart disease history, low handgrip strength, and abnormal 6MWT and TUG test were significantly associated with geriatric syndrome (Table 2). Living situation, marital status, age, self-reported economic status, educational level, and FTSST were not significantly associated with geriatric syndrome.

Risk factors of geriatric syndrome
Development of the Taiwan Risk Scores for Geriatric Syndrome (TRSGS) model Using stepwise logistic regression, 7 items were selected from the risk factors of geriatric syndrome, including 'cannot report personal information', 'depressive mood for the past 2 weeks', a history of heart disease, absence of labour activities, low handgrip strength, and abnormal 6MWT and TUG test. The weighting of each item was calculated by each adjusted odds ratio. Finally, the cutoff value of TRSGS model was 6 (AUC, 0.826; sensitivity, 77.2%; specificity, 75.5%) (Figure 3).

Discussion
To the best of our knowledge, there is no easily applicable screening tool in predicting geriatric syndrome in the community. Most of the screening tools focus on specific problems such as sarcopenia, dementia, or frailty [29, 30]. The current study assessed the risk factors and developed a multidimensional risk model in predicting geriatric syndrome in community-dwelling older adults. We found out that TRSGS score ≧6 was significantly associated with geriatric syndrome, which comprised items to evaluate cognitive impairment, depressive disorder, baseline functional and sensory impairment, and impaired mobility in community-dwelling older adults. Therefore, the TRSGS can be used in predicting geriatric syndrome for community-dwelling older adults. Further cohort study will be designed to validate the predictors of geriatric syndrome and support the implementation of preventive strategies in community-dwelling older adults.
In the previous studies, physical function tests were considered as indicators for sarcopenia or frailty [11]. For example, low handgrip strength was defined as <26 kg for men and <18 kg for women by the Asian Working Group for Sarcopenia (AWGS) in the screening of sarcopenia. The suggested cutoff values (<27.3 kg for men and <19.2 kg for women) in the current study were higher than that of the previous studies. Using the cutoff value suggested by the AWGS might decrease the sensitivity and increase the specificity because sarcopenia is only part of the geriatric syndrome ( Figure 1A and 2A). Questionnaire, were available [33]. We tried to use one screening item to evaluate cognitive function in the current study. We have proven that if participants could not report their personal information, they were significantly associated with geriatric syndrome in our study. We assumed that a simple item can be a good predictor for the screening of cognitive impairment in community-dwelling older adults.
Our study has some limitations. First, our study was a cross-sectional study that reported the association between risk factors and geriatric syndrome. The incidence of geriatric syndrome cannot be estimated. Second, although we recruited participants from 58 communities among 4 cities, the generalisation of the results in this study is still questionable. Third, the records of socioeconomic status, medical history, and healthy behaviours were self-reported and self-recalled, which may have bias. Finally, although we included several covariates associated with geriatric syndrome, there may be other potential confounders that were unavailable.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and analysed during the current study are not publicity available, but are available from the corresponding author on reasonable request with the permission of the Ministry of Health and Welfare, Taiwan.

Competing interests
The authors declare that they have no competing interests.     The validity of the Taiwan Risk Scores for Geriatric Syndrome model. 6MWT, 6meter walk test; TUG test, timed up-and-go test.