Diagnostic accuracy of Instrumental Activities of Daily Living for dementia in community-dwelling older adults

Diagnostic accuracy of Instrumental Activities of Daily Living for dementia in community-dwelling... Abstract Background many people living with dementia remain underdiagnosed and unrecognised. Screening strategies are important for early detection. Objective to examine whether the Lawton’s Instrumental Activities of Daily Living (IADL) scale, compared with other cognitive screening tools—the Mini-Mental State Examination (MMSE), and the Ascertain Dementia 8-item Informant Questionnaire (AD8)—can identify older (≥ 65 years) adults with dementia. Design population-based cross-sectional observational study. Setting all 19 counties in Taiwan. Participants community-dwelling older adults (n = 10,340; mean age 74.87 ± 6.03). Methods all participants underwent a structured in-person interview. Dementia was identified using National Institute on Aging-Alzheimer’s Association core clinical criteria for all-cause dementia. Receiver operator characteristic curves were used to determine the discriminant abilities of the IADL scale, MMSE and AD8 to differentiate participants with and without dementia. Results we identified 917 (8.9%) participants with dementia, and 9,423 (91.1%) participants without. The discriminant abilities of the MMSE, AD8 and IADL scale (cutoff score: 6/7; area under curve = 0.925; sensitivity = 89%; specificity = 81%; positive likelihood ratio = 4.75; accuracy = 0.82) were comparable. Combining IADL with AD8 scores significantly improved overall accuracy: specificity = 93%; positive likelihood ratio = 11.74; accuracy = 0.92. Conclusions our findings support using IADL scale to screen older community-dwelling residents for dementia: it has discriminant power comparable to that of the AD8 and MMSE. Combining the IADL and the AD8 improves specificity. Alzheimer’s disease, dementia, screen, instrumental activity of daily living, diagnostic accuracy, older people Introduction Early detection and timely diagnosis of dementia, a healthcare priority in many countries and regions [1], are a gateway for providing healthcare and social services [2]. Studies [3, 4] have reported that early detection and timely diagnosis reduce the patient’s functional decline, the caregiver’s burden and the cost of care. Others [5–7] have reported that diagnostic coverage is only 40–50% in high-income countries, and that it is unlikely to exceed 5–10% in low- and middle-income countries. Many people living with dementia remain underdiagnosed and unrecognised. Screening strategies are important for early detection [8]. Valid and reliable cognitive tests are frequently used for early dementia detection, including performance-based measures, e.g. the Mini-Mental State Examination (MMSE), the Mini-Cog™ and informant-based measures like the Ascertain Dementia 8-item Informant Questionnaire (AD8) [4, 9–12]. However, only 20% of patients are screened using cognitive tests before being referred to a memory specialist [13]. Using cognitive tests for screening seems difficult in clinical and community settings [14]. A high prevalence of illiteracy and low-level education might be barriers to cognitive screening in middle and low-income countries [15]. The space, time and training required to properly administer conventional screening tools can also be a burden for community workers [16]. Furthermore, the cost-effectiveness of large-scale screening in detecting cognitive impairment has been debated, but identifying cognitive impairment during routine care is becoming critical [1, 4]. Therefore, it is necessary to find an alternative and commonly used screening tool that has no educational bias and is easy to administer in a community-based setting [8, 16]. Using the instrumental activities of daily living (IADL) as a screening tool for dementia has been proposed as an alternative strategy [17–23]. There is robust evidence that difficulty with IADL is one early feature of dementia. People with early-stage dementia often show declines in IADL performance, which requires complex cognitive processing [24, 25]. The National Institute on Aging-Alzheimer’s Association (NIA-AA) [26], Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition: DSM-IV-TR (DSM-IV) [27] and DSM-V [28] criteria for diagnosis of dementia require not only deficits in cognitive function, but also decreases in daily, social and occupational function. Because it is an early feature of dementia syndromes [25], IADL disability might be a good screening indicator for dementia. Using IADL as a screening tool for dementia has not been supported consistently by empirical studies, despite the fair specificity found in population-based studies [29]. Moreover, the diagnostic accuracy—especially specificity—was not as high in clinical settings as in community settings [19, 20]. Common findings of physical impairments or comorbidities in older (≥65 years) adults in clinical settings might contribute to IADL dependence. The Lawton IADL assessment is one of the most popular IADL scales. However, it was not developed to screen people with dementia [30]. If used in dementia screening, a cutoff score needs to be examined and established. The purpose of this study was to use a cross-sectional nationwide population to (i) evaluate the discriminant validity (accuracy and utility) of Lawton’s IADL scale to identify community-dwelling older adults with dementia; and (ii) to compare the discriminant validity of the IADL scale, MMSE and AD8, and the combined AD8 and IADL scale for dementia screening. Methods Study design and sampling A door-to-door survey of 10,432 potential participants was done by well-trained field interviewers between December 2011 and March 2013. To assess the interviewees, interviewers used the MMSE [31], the AD8, the Clinical Dementia Rating Scale (CDR) and a structured questionnaire that asked about demographics, medical history, cognitive and functional status (IADL and ADL), and lifestyle factors. The sampling process, training of interviewers, home visiting procedures, quality control and protocol approvals are described elsewhere [32]. Diagnosing dementia: procedures Dementia was diagnosed using the core clinical criteria for all-cause dementia from the NIA-AA [26]. The detailed diagnosis criteria are described elsewhere [32]. IADL performance Participants or informants were asked about each participant’s ability during the previous month to perform eight activities listed in Lawton’s IADL scale (Table 1). The ranges of the raw scores of the Lawton’s IADL items were 1–3, 1–4 and 1–5. The raw score of each item was converted to a dichotomised score (1 = independent, 0 = not independent). A summary score ranges from 0 (low function: dependent) to 8 (high function: independent) [30]. Data from participants who answered ‘unable to judge’ or ‘missing’ were scored as missing, and they were excluded from the ROC curves analyses. The AD8 is described in Appendix 1, and other measures of covariates are described in Appendix 2 (both appendices available at Age and Ageing online). Table 1. Demographics, and IADL, AD8 and MMSE performances (n = 10,360) and comparisons of characteristics of participants. Parameters Non-Dementiaa (n = 9,443) Dementiab (n = 917) P Male (n, %) 4,590 (48.6) 348 (37.9) <0.001 Age (years) (mean ± SD) 75.63 ± 6.39 81.63 ± 7.45 <0.001 Living area (n, %) <0.001  Rural 2,358 (25.0) 395 (43.1)  Urban 7,085 (75.0) 522 (56.9) Marital and living status (n, %) <0.001  Live alone 973 (10.3) 71 (7.7)  Live with partner only 38,69 (41.0) 264 (28.8)  Live with family 4,541 (48.1) 540 (58.9)  Other 60 (0.6) 42 (4.6) Educational level (n, %) <0.001  Illiterate 2,879 (30.5) 445 (48.5)  Literate 926 (9.8) 105 (11.5)  Elementary school 3,421 (36.2) 241 (26.3)  Junior high school 745 (7.9) 43 (4.7)  Senior high school 814 (8.6) 46 (5.0)  College 595 (6.3) 34 (3.7)  Graduate school and above 63 (0.7) 3 (0.3) MMSE (mean ± SD) 24.54 ± 4.72 12.84 ± 6.28 <0.001 AD8 (mean ± SD) 0.71 ± 1.52 4.97 ± 2.75 <0.001 AD8 ≥ 2 (n, %) 1,464 (15.5) 804 (87.7) <0.001 IADL, independent (n, %)  Shopping 7,571 (81.4) 77 (8.7) <0.001  Transportation 8,261 (89.1) 229 (25.6) <0.001  Finances 8,159 (90.5) 214 (24.7) <0.001  Telephone use 8,924 (95.2) 414 (46.0) <0.001  Medication 9,077 (98.5) 403 (63.9) <0.001  Food preparation 7,275 (81.8) 152 (17.3) <0.001  Household chores 8,147 (90.6) 296 (33.5) <0.001  Laundry 7,410 (83.7) 220 (25.1) <0.001 IADL (mean ± SD) 7.19 ± 1.63 2.76 ± 2.51 <0.001 IADL < 7 (n, %) 1,541 (18.7) 520 (89.0) <0.001 Parameters Non-Dementiaa (n = 9,443) Dementiab (n = 917) P Male (n, %) 4,590 (48.6) 348 (37.9) <0.001 Age (years) (mean ± SD) 75.63 ± 6.39 81.63 ± 7.45 <0.001 Living area (n, %) <0.001  Rural 2,358 (25.0) 395 (43.1)  Urban 7,085 (75.0) 522 (56.9) Marital and living status (n, %) <0.001  Live alone 973 (10.3) 71 (7.7)  Live with partner only 38,69 (41.0) 264 (28.8)  Live with family 4,541 (48.1) 540 (58.9)  Other 60 (0.6) 42 (4.6) Educational level (n, %) <0.001  Illiterate 2,879 (30.5) 445 (48.5)  Literate 926 (9.8) 105 (11.5)  Elementary school 3,421 (36.2) 241 (26.3)  Junior high school 745 (7.9) 43 (4.7)  Senior high school 814 (8.6) 46 (5.0)  College 595 (6.3) 34 (3.7)  Graduate school and above 63 (0.7) 3 (0.3) MMSE (mean ± SD) 24.54 ± 4.72 12.84 ± 6.28 <0.001 AD8 (mean ± SD) 0.71 ± 1.52 4.97 ± 2.75 <0.001 AD8 ≥ 2 (n, %) 1,464 (15.5) 804 (87.7) <0.001 IADL, independent (n, %)  Shopping 7,571 (81.4) 77 (8.7) <0.001  Transportation 8,261 (89.1) 229 (25.6) <0.001  Finances 8,159 (90.5) 214 (24.7) <0.001  Telephone use 8,924 (95.2) 414 (46.0) <0.001  Medication 9,077 (98.5) 403 (63.9) <0.001  Food preparation 7,275 (81.8) 152 (17.3) <0.001  Household chores 8,147 (90.6) 296 (33.5) <0.001  Laundry 7,410 (83.7) 220 (25.1) <0.001 IADL (mean ± SD) 7.19 ± 1.63 2.76 ± 2.51 <0.001 IADL < 7 (n, %) 1,541 (18.7) 520 (89.0) <0.001 MMSE, Mini-Mental State Examination; AD8, The Ascertain Dementia 8-item Informant Questionnaire; IADL, Lawton’s Instrumental Activities of Daily Living scale with modified scoring method. aNon-Dementia group includes the non-dementia control, the unclassified older adults and the mild cognitive impairment (MCI) subgroups. bDementia group includes older adults with very mild dementia and all-cause dementia. Table 1. Demographics, and IADL, AD8 and MMSE performances (n = 10,360) and comparisons of characteristics of participants. Parameters Non-Dementiaa (n = 9,443) Dementiab (n = 917) P Male (n, %) 4,590 (48.6) 348 (37.9) <0.001 Age (years) (mean ± SD) 75.63 ± 6.39 81.63 ± 7.45 <0.001 Living area (n, %) <0.001  Rural 2,358 (25.0) 395 (43.1)  Urban 7,085 (75.0) 522 (56.9) Marital and living status (n, %) <0.001  Live alone 973 (10.3) 71 (7.7)  Live with partner only 38,69 (41.0) 264 (28.8)  Live with family 4,541 (48.1) 540 (58.9)  Other 60 (0.6) 42 (4.6) Educational level (n, %) <0.001  Illiterate 2,879 (30.5) 445 (48.5)  Literate 926 (9.8) 105 (11.5)  Elementary school 3,421 (36.2) 241 (26.3)  Junior high school 745 (7.9) 43 (4.7)  Senior high school 814 (8.6) 46 (5.0)  College 595 (6.3) 34 (3.7)  Graduate school and above 63 (0.7) 3 (0.3) MMSE (mean ± SD) 24.54 ± 4.72 12.84 ± 6.28 <0.001 AD8 (mean ± SD) 0.71 ± 1.52 4.97 ± 2.75 <0.001 AD8 ≥ 2 (n, %) 1,464 (15.5) 804 (87.7) <0.001 IADL, independent (n, %)  Shopping 7,571 (81.4) 77 (8.7) <0.001  Transportation 8,261 (89.1) 229 (25.6) <0.001  Finances 8,159 (90.5) 214 (24.7) <0.001  Telephone use 8,924 (95.2) 414 (46.0) <0.001  Medication 9,077 (98.5) 403 (63.9) <0.001  Food preparation 7,275 (81.8) 152 (17.3) <0.001  Household chores 8,147 (90.6) 296 (33.5) <0.001  Laundry 7,410 (83.7) 220 (25.1) <0.001 IADL (mean ± SD) 7.19 ± 1.63 2.76 ± 2.51 <0.001 IADL < 7 (n, %) 1,541 (18.7) 520 (89.0) <0.001 Parameters Non-Dementiaa (n = 9,443) Dementiab (n = 917) P Male (n, %) 4,590 (48.6) 348 (37.9) <0.001 Age (years) (mean ± SD) 75.63 ± 6.39 81.63 ± 7.45 <0.001 Living area (n, %) <0.001  Rural 2,358 (25.0) 395 (43.1)  Urban 7,085 (75.0) 522 (56.9) Marital and living status (n, %) <0.001  Live alone 973 (10.3) 71 (7.7)  Live with partner only 38,69 (41.0) 264 (28.8)  Live with family 4,541 (48.1) 540 (58.9)  Other 60 (0.6) 42 (4.6) Educational level (n, %) <0.001  Illiterate 2,879 (30.5) 445 (48.5)  Literate 926 (9.8) 105 (11.5)  Elementary school 3,421 (36.2) 241 (26.3)  Junior high school 745 (7.9) 43 (4.7)  Senior high school 814 (8.6) 46 (5.0)  College 595 (6.3) 34 (3.7)  Graduate school and above 63 (0.7) 3 (0.3) MMSE (mean ± SD) 24.54 ± 4.72 12.84 ± 6.28 <0.001 AD8 (mean ± SD) 0.71 ± 1.52 4.97 ± 2.75 <0.001 AD8 ≥ 2 (n, %) 1,464 (15.5) 804 (87.7) <0.001 IADL, independent (n, %)  Shopping 7,571 (81.4) 77 (8.7) <0.001  Transportation 8,261 (89.1) 229 (25.6) <0.001  Finances 8,159 (90.5) 214 (24.7) <0.001  Telephone use 8,924 (95.2) 414 (46.0) <0.001  Medication 9,077 (98.5) 403 (63.9) <0.001  Food preparation 7,275 (81.8) 152 (17.3) <0.001  Household chores 8,147 (90.6) 296 (33.5) <0.001  Laundry 7,410 (83.7) 220 (25.1) <0.001 IADL (mean ± SD) 7.19 ± 1.63 2.76 ± 2.51 <0.001 IADL < 7 (n, %) 1,541 (18.7) 520 (89.0) <0.001 MMSE, Mini-Mental State Examination; AD8, The Ascertain Dementia 8-item Informant Questionnaire; IADL, Lawton’s Instrumental Activities of Daily Living scale with modified scoring method. aNon-Dementia group includes the non-dementia control, the unclassified older adults and the mild cognitive impairment (MCI) subgroups. bDementia group includes older adults with very mild dementia and all-cause dementia. Statistical analyses For each measure and the combined AD8 and IADL scale (‘and’ rule: the participant was considered positive only when both tests were positive) the area under the receiver-operating characteristic (ROC) curve (AUC) was calculated for each gender and educational subgroup. A cutoff score for each measure that best differentiated diagnostic groups (Dementia versus Non-Dementia) was determined using the Youden index [34]. A combination of the IADL and the MMSE was not evaluated because of copyright problems. Other details of the statistical analysis are provided in Appendix 3 (available at Age and Ageing online). Declaration of ethics and sources of funding This study was approved by National Taiwan University Hospital’s Institutional Review Board (201012021RB), and supported in part by grant DOH101-TD-M-113-100001 from the Taiwan Ministry of Health and Welfare, and in part by the Taiwan Alzheimer Disease Association. The sponsors were not involved in executing the study, analysing or interpreting the data, or writing the manuscript. Results Participant characteristics Of the 10,432 interviewees, 72 living in institutions were excluded, which left us 10,340 community-dwelling older adults to analyse. There were 917 older adults (8.85%) in the Dementia group, and 9,443 (91.15%) in the Non-Dementia group, including 2,021 (19.51%) with mild cognitive impairment (MCI) and 410 (3.96%) unclassified; 7,012 (67.68%) with intact cognitive function were the Non-Dementia Control group. Demographic information, IADL, MMSE and AD8 performance are presented in Table 1. Lifestyle characteristics and medical and clinical information are presented in Appendix 4. IADL disabilities independently discriminating between older people with and without dementia even after controlling for cognitive function are presented in Appendix 5 and Appendix 6 (all appendices available at Age and Ageing online). The discriminative ability of IADL, MMSE, AD8, and the combination of IADL and AD8 The IADL scale discriminated extremely well between the Dementia and Non-Dementia groups: cutoff IADL scores were 6/7 (AUC = 0.925; 95% CI = 0.915–0.935; sensitivity = 89%; specificity = 81%; LR+ = 4.75; LR− = 0.13; and classification accuracy = 0.82) (Figure 1 and Table 2). Those unable to perform independently six or more IADL items from Lawton’s scale (scored < 7) were 4.75 times more likely to be diagnosed with dementia, which is considered a fair LR+ [36]. Using the IADL cutoff score of 6/7, 89% of those with dementia (true-positives) and 81% of those without dementia (true-negatives) were expected to be correctly identified. Table 2. Cutoff scores for optimal sensitivity and specificity of MMSE, AD8 and IADL for dementiaa by gender and educational level. Index and subgroups Cutoff AUC (95% CI) P Youden indexb Sensitivity Specificity LR+ LR− Accuracy MMSE  All 20.5 0.931 (0.922–0.940) <0.001 0.71 0.89 0.82 5.03 0.14 0.83  Gender   Male 21.5 0.929 (0.915–0.943) <0.001 0.73 0.87 0.85 5.93 0.15 0.85   Female 17.5 0.937 (0.925–0.948) <0.001 0.74 0.84 0.90 8.21 0.18 0.89  Educational level   Illiterate 15.5 0.950 (0.939–0.961) <0.001 0.79 0.89 0.91 9.49 0.13 0.90   Elementary school 21.5 0.937 (0.925–0.950) <0.001 0.74 0.89 0.85 6.02 0.13 0.85   Junior high school or above 24.5 0.944 (0.919–0.969) <0.001 0.79 0.90 0.89 8.31 0.12 0.89 AD8  All 1.5 0.905 (0.894–0.917) <0.001 0.72 0.88 0.84 5.65 0.15 0.85  Gender   Male 1.5 0.912 (0.893–0.931) <0.001 0.74 0.87 0.87 6.79 0.15 0.87   Female 1.5 0.897 (0.883–0.912) <0.001 0.70 0.88 0.82 4.88 0.15 0.83  Educational level   Illiterate 1.5 0.891 (0.875–0.907) <0.001 0.66 0.92 0.74 3.56 0.10 0.77   Elementary school 1.5 0.898 (0.878–0.919) <0.001 0.71 0.84 0.87 6.66 0.19 0.87   Junior high school or above 1.5 0.913 (0.879–0.947) <0.001 0.74 0.82 0.92 10.65 0.20 0.92 IADL  All 6.5 0.925 (0.915–0.935) <0.001 0.70 0.89 0.81 4.75 0.13 0.82  Gender   Male 6.5 0.926 (0.910–0.942) <0.001 0.71 0.93 0.78 4.22 0.09 0.79   Female 6.5 0.928 (0.915–0.940) <0.001 0.71 0.86 0.84 5.50 0.16 0.84  Educational level   Illiterate 6.5 0.913 (0.898–0.929) <0.001 0.64 0.89 0.75 3.60 0.14 0.77   Elementary school 6.5 0.916 (0.898–0.934) <0.001 0.69 0.87 0.82 4.86 0.16 0.82   Junior high school or above 6.5 0.947 (0.917–0.978) <0.001 0.82 0.94 0.88 7.60 0.06 0.88 AD8+IADL  All 1.5/6.5 0.947 (0.939–0.955) <0.001 0.74 0.81 0.93 11.74 0.21 0.92 Index and subgroups Cutoff AUC (95% CI) P Youden indexb Sensitivity Specificity LR+ LR− Accuracy MMSE  All 20.5 0.931 (0.922–0.940) <0.001 0.71 0.89 0.82 5.03 0.14 0.83  Gender   Male 21.5 0.929 (0.915–0.943) <0.001 0.73 0.87 0.85 5.93 0.15 0.85   Female 17.5 0.937 (0.925–0.948) <0.001 0.74 0.84 0.90 8.21 0.18 0.89  Educational level   Illiterate 15.5 0.950 (0.939–0.961) <0.001 0.79 0.89 0.91 9.49 0.13 0.90   Elementary school 21.5 0.937 (0.925–0.950) <0.001 0.74 0.89 0.85 6.02 0.13 0.85   Junior high school or above 24.5 0.944 (0.919–0.969) <0.001 0.79 0.90 0.89 8.31 0.12 0.89 AD8  All 1.5 0.905 (0.894–0.917) <0.001 0.72 0.88 0.84 5.65 0.15 0.85  Gender   Male 1.5 0.912 (0.893–0.931) <0.001 0.74 0.87 0.87 6.79 0.15 0.87   Female 1.5 0.897 (0.883–0.912) <0.001 0.70 0.88 0.82 4.88 0.15 0.83  Educational level   Illiterate 1.5 0.891 (0.875–0.907) <0.001 0.66 0.92 0.74 3.56 0.10 0.77   Elementary school 1.5 0.898 (0.878–0.919) <0.001 0.71 0.84 0.87 6.66 0.19 0.87   Junior high school or above 1.5 0.913 (0.879–0.947) <0.001 0.74 0.82 0.92 10.65 0.20 0.92 IADL  All 6.5 0.925 (0.915–0.935) <0.001 0.70 0.89 0.81 4.75 0.13 0.82  Gender   Male 6.5 0.926 (0.910–0.942) <0.001 0.71 0.93 0.78 4.22 0.09 0.79   Female 6.5 0.928 (0.915–0.940) <0.001 0.71 0.86 0.84 5.50 0.16 0.84  Educational level   Illiterate 6.5 0.913 (0.898–0.929) <0.001 0.64 0.89 0.75 3.60 0.14 0.77   Elementary school 6.5 0.916 (0.898–0.934) <0.001 0.69 0.87 0.82 4.86 0.16 0.82   Junior high school or above 6.5 0.947 (0.917–0.978) <0.001 0.82 0.94 0.88 7.60 0.06 0.88 AD8+IADL  All 1.5/6.5 0.947 (0.939–0.955) <0.001 0.74 0.81 0.93 11.74 0.21 0.92 AUC, area under the curve; LR+, positive likelihood ratio; LR−, negative likelihood ratio; accuracy, classification accuracy; MMSE, Mini-Mental State Examination; AD8, The Ascertain Dementia 8-item Informant Questionnaire; IADL, Lawton’s Instrumental Activities of Daily Living scale. aNumber of participants: Dementia group, n = 584; Non-Dementia group, n = 8,221 (participants with any missing data in the IADL scale were excluded). bA suggested cutoff was decided using the optimised Youden Index, if the differences of the Youden Index among multiple candidates were within 0.01. Table 2. Cutoff scores for optimal sensitivity and specificity of MMSE, AD8 and IADL for dementiaa by gender and educational level. Index and subgroups Cutoff AUC (95% CI) P Youden indexb Sensitivity Specificity LR+ LR− Accuracy MMSE  All 20.5 0.931 (0.922–0.940) <0.001 0.71 0.89 0.82 5.03 0.14 0.83  Gender   Male 21.5 0.929 (0.915–0.943) <0.001 0.73 0.87 0.85 5.93 0.15 0.85   Female 17.5 0.937 (0.925–0.948) <0.001 0.74 0.84 0.90 8.21 0.18 0.89  Educational level   Illiterate 15.5 0.950 (0.939–0.961) <0.001 0.79 0.89 0.91 9.49 0.13 0.90   Elementary school 21.5 0.937 (0.925–0.950) <0.001 0.74 0.89 0.85 6.02 0.13 0.85   Junior high school or above 24.5 0.944 (0.919–0.969) <0.001 0.79 0.90 0.89 8.31 0.12 0.89 AD8  All 1.5 0.905 (0.894–0.917) <0.001 0.72 0.88 0.84 5.65 0.15 0.85  Gender   Male 1.5 0.912 (0.893–0.931) <0.001 0.74 0.87 0.87 6.79 0.15 0.87   Female 1.5 0.897 (0.883–0.912) <0.001 0.70 0.88 0.82 4.88 0.15 0.83  Educational level   Illiterate 1.5 0.891 (0.875–0.907) <0.001 0.66 0.92 0.74 3.56 0.10 0.77   Elementary school 1.5 0.898 (0.878–0.919) <0.001 0.71 0.84 0.87 6.66 0.19 0.87   Junior high school or above 1.5 0.913 (0.879–0.947) <0.001 0.74 0.82 0.92 10.65 0.20 0.92 IADL  All 6.5 0.925 (0.915–0.935) <0.001 0.70 0.89 0.81 4.75 0.13 0.82  Gender   Male 6.5 0.926 (0.910–0.942) <0.001 0.71 0.93 0.78 4.22 0.09 0.79   Female 6.5 0.928 (0.915–0.940) <0.001 0.71 0.86 0.84 5.50 0.16 0.84  Educational level   Illiterate 6.5 0.913 (0.898–0.929) <0.001 0.64 0.89 0.75 3.60 0.14 0.77   Elementary school 6.5 0.916 (0.898–0.934) <0.001 0.69 0.87 0.82 4.86 0.16 0.82   Junior high school or above 6.5 0.947 (0.917–0.978) <0.001 0.82 0.94 0.88 7.60 0.06 0.88 AD8+IADL  All 1.5/6.5 0.947 (0.939–0.955) <0.001 0.74 0.81 0.93 11.74 0.21 0.92 Index and subgroups Cutoff AUC (95% CI) P Youden indexb Sensitivity Specificity LR+ LR− Accuracy MMSE  All 20.5 0.931 (0.922–0.940) <0.001 0.71 0.89 0.82 5.03 0.14 0.83  Gender   Male 21.5 0.929 (0.915–0.943) <0.001 0.73 0.87 0.85 5.93 0.15 0.85   Female 17.5 0.937 (0.925–0.948) <0.001 0.74 0.84 0.90 8.21 0.18 0.89  Educational level   Illiterate 15.5 0.950 (0.939–0.961) <0.001 0.79 0.89 0.91 9.49 0.13 0.90   Elementary school 21.5 0.937 (0.925–0.950) <0.001 0.74 0.89 0.85 6.02 0.13 0.85   Junior high school or above 24.5 0.944 (0.919–0.969) <0.001 0.79 0.90 0.89 8.31 0.12 0.89 AD8  All 1.5 0.905 (0.894–0.917) <0.001 0.72 0.88 0.84 5.65 0.15 0.85  Gender   Male 1.5 0.912 (0.893–0.931) <0.001 0.74 0.87 0.87 6.79 0.15 0.87   Female 1.5 0.897 (0.883–0.912) <0.001 0.70 0.88 0.82 4.88 0.15 0.83  Educational level   Illiterate 1.5 0.891 (0.875–0.907) <0.001 0.66 0.92 0.74 3.56 0.10 0.77   Elementary school 1.5 0.898 (0.878–0.919) <0.001 0.71 0.84 0.87 6.66 0.19 0.87   Junior high school or above 1.5 0.913 (0.879–0.947) <0.001 0.74 0.82 0.92 10.65 0.20 0.92 IADL  All 6.5 0.925 (0.915–0.935) <0.001 0.70 0.89 0.81 4.75 0.13 0.82  Gender   Male 6.5 0.926 (0.910–0.942) <0.001 0.71 0.93 0.78 4.22 0.09 0.79   Female 6.5 0.928 (0.915–0.940) <0.001 0.71 0.86 0.84 5.50 0.16 0.84  Educational level   Illiterate 6.5 0.913 (0.898–0.929) <0.001 0.64 0.89 0.75 3.60 0.14 0.77   Elementary school 6.5 0.916 (0.898–0.934) <0.001 0.69 0.87 0.82 4.86 0.16 0.82   Junior high school or above 6.5 0.947 (0.917–0.978) <0.001 0.82 0.94 0.88 7.60 0.06 0.88 AD8+IADL  All 1.5/6.5 0.947 (0.939–0.955) <0.001 0.74 0.81 0.93 11.74 0.21 0.92 AUC, area under the curve; LR+, positive likelihood ratio; LR−, negative likelihood ratio; accuracy, classification accuracy; MMSE, Mini-Mental State Examination; AD8, The Ascertain Dementia 8-item Informant Questionnaire; IADL, Lawton’s Instrumental Activities of Daily Living scale. aNumber of participants: Dementia group, n = 584; Non-Dementia group, n = 8,221 (participants with any missing data in the IADL scale were excluded). bA suggested cutoff was decided using the optimised Youden Index, if the differences of the Youden Index among multiple candidates were within 0.01. Figure 1. View largeDownload slide Receiver operating characteristic curve for discriminating between elderly adults with and without dementia. (A) IADL; (B) MMSE; (C) AD8; and (D) IADL+AD8. Figure 1. View largeDownload slide Receiver operating characteristic curve for discriminating between elderly adults with and without dementia. (A) IADL; (B) MMSE; (C) AD8; and (D) IADL+AD8. An MMSE cutoff score of 20/21 resulted in an AUC value of 0.931. An AD8 cutoff score of 1/2 resulted in an AUC value of 0.905 (Table 2). There were no significant differences in the diagnostic and classification accuracy between the IADL scale, the MMSE, and the AD8 (P > 0.05). When the IADL and AD8 were combined, the AUC improved to 0.947 (95% CI = 0.939–0.955), sensitivity decreased by 8% (81%), specificity increased by 12% (93%), utility rose to excellent [LR+ (11.74)], and classification accuracy significantly increased by 10% [χ2] (1) = 58.36, P < 0.0001. IADL was gender- and education-independent The cutoff scores of the IADL and AD8 scales did not differ by gender or educational subgroup. However, different genders and educational subgroups had different MMSE cutoff scores (Table 2). Discussion Our most important finding was that Lawton’s IADL scale discriminated between older adults with and without dementia well as did the MMSE and the AD8. We also found that the IADL scale discriminated between those with and without dementia independently of the MMSE and AD8, even after comorbidities and lifestyle risk factors had been adjusted for. Several longitudinal studies [25, 37–39] reported that IADL performance predicts incident dementia. Our findings support the notion that declines in functional performance can be an independent predictor for dementia; thus, we conclude that IADL disabilities can be used as a screening index. Using IADL as a dementia screening strategy has been overlooked in favour of focusing on cognitive assessments [4]. This study supports that the IADL is a valid tool for identifying older adults with dementia in communities with a relatively low-prevalence rate (about 8–10%). An IADL scale score of 6 or less detected 92.5% of dementia among community-dwelling older adults (sensitivity = 0.89, specificity = 0.81), which was comparable with other, similar studies (estimated summarised AUC = 0.88, weighted average sensitivity = 90.18%, weighted average specificity = 78.79%) [29]. However, the diagnostic accuracy reported by some studies [19, 20] done in settings with higher prevalence rates, such as memory clinics, was not satisfactory, particularly the specificity. Therefore, we hypothesise that IADL measures are more effective for identifying community-dwelling older adults with dementia. This study is the first to compare the three frequently used tools to screen for dementia in community-dwelling older adults: Lawton’s IADL scale, the MMSE, and the AD8. All three detected more than 90% of the cases. The major concern is that factors unrelated to dementia, e.g. physiological degeneration and comorbidities, can also lead to functional deterioration in the older adults. Thus, using the IADL scale to screen for dementia might lead to less satisfactory specificity and a relatively high false-positive rate [29]. In the current study, the related discriminant indices of the IADL scale were as good as those of the AD8 and the MMSE. This might be because each screening tool has its own shortcomings. For example, the MMSE yielded a high false-positive rate in the older adults and in participants with lower education levels [4, 10, 16, 40]. We also found that the MMSE yielded gender and educational effects, but that the IADL and AD8 did not. Therefore, we can reasonably assume that although the IADL as a dementia screening scale did not have an ideal specificity, its discriminant validity as a screening tool was as good as that of the AD8 and the MMSE, with the additional benefit of easier administration. Furthermore, the IADL scale is frequently used in routine health examinations to provide information about a person’s functional status, which can be used to assess the care services they need [38]. Thus, its use for dementia detection will not add an extra burden for care providers. Moreover, evidence that IADL disabilities indicate a high risk of dementia should increase public awareness of dementia. It is interesting to note that the cutoff scores of Lawton’s IADL scale were identical for both genders, and that similar diagnostic accuracy was reported. Traditionally, gender differences would be considered when assessing IADL disabilities. A few studies [24, 41, 42] found that the patterns of functional abilities vary for men and women. However, the current study and Juva [23] found no substantial gender differences in IADLs. Therefore, the IADL scale can be used for both genders with the same standard. Recent research [13, 43, 44] found that combining two or more tools for dementia screening might increase the specificity and lower the probability of mislabeling a typically aging older adult as someone with dementia. This might prevent the expensive and potentially stressful series of tests for dementia, and minimise the possible adverse psychological effects of screening. Our findings showed that the correlations between the MMSE, AD8, and IADL were only moderate, and that each could independently predict dementia, which implied that each screening tool might evaluate different traits of dementia, and therefore complement each other. We suggest combining the IADL and AD8 results to reduce the false-positive rate. Those who screen positive on both will need a comprehensive evaluation. The current study has several strengths. It analysed a large representative sample from a nationwide epidemiology study that used face-to-face interviews. All participants underwent a detailed assessment to ensure a reliable diagnosis of dementia based on NIA-AA core clinical criteria, a primary diagnostic guideline for dementia research. Additionally, this is the first study that extensively compares the three most frequently used screening assessments (Lawton’s IADL, the MMSE, and the AD8). Lastly, we propose valid screening cutoff scores that allow immediate clinical application in community settings. This study has some limitations. First, we did not assess other factors that influence IADL, e.g. hand function, depression, and environmental factors. Second, there might be disagreement between self-reports and informant-based IADL data [45], but the data collected using these two methods were not differentiated in the study. Third, the results can be affected by ‘incorporation bias’, whereby the IADL disabilities, MMSE, and subjective complaints of cognitive problems (i.e. the AD8) form a part of the reference standard (the NIA-AA criteria). This bias is common in dementia diagnostic studies, according to the reporting standards for studies of diagnostic test accuracy in dementia [46]. Therefore, screening accuracy is associated with a tendency to overestimate the value of the three measures. In addition, validation has not been established for another population. Because of the likelihood of overfitting the data in this study, diagnostic cutpoints in an exploratory analysis in other cohorts are likely to be less accurate. Finally, the results of the current study can be generalised to settings with a low-prevalence rate of dementia and for people with similar cultural backgrounds, such as Asian communities. Additional prospective (longitudinal) studies aimed at establishing the predictive validity of the IADL scale in discriminating people with and without dementia are warranted. Because advanced activities of daily living (AADLs) (e.g. interests in hobbies, driving and trip planning) are more complex than are IADLs, future studies might want to investigate whether AADLs are more sensitive for detecting early-stage dementia [47]. Conclusion IADL disabilities independently discriminate between older people with and without dementia even after controlling for cognitive function. This study supports using Lawton’s IADL scale with a modified scoring method as an alternative dementia screening tool for community-dwelling older adults. Using the IADL as an initial screening tool can decrease the bias associated with gender and with educational levels. The discriminant ability of the IADL for dementia in community-dwelling older adults is comparable to that of the MMSE and AD8, the conventional cognitive-based screening tools. We recommend combining the IADL and the AD8 to improve specificity. Key points The discriminant ability of Lawton’s Instrumental Activities of Daily Living (IADL) scale to identify persons with dementia is comparable to that of Mini-Mental State Examination (MMSE), and Ascertain Dementia 8-item Informant Questionnaire (AD8). The combination of two informant-based tools, the Instrumental Activities of Daily Living (IADL) and Ascertain Dementia 8-item Informant Questionnaire (AD8), reduces the false-positive rate and improves specificity. The Instrumental Activities of Daily Living (IADL) is feasible to detect cognitive impairment in a routine healthcare examination. Supplementary data Supplementary data mentioned in the text are available to subscribers in Age and Ageing online. Acknowledgements We thank the study participants and their families for their cooperation and support, the staff members of Taiwan Alzheimer Disease Association, and the interviewers for their valuable contributions. Funding This work was supported in part by the Taiwan Ministry of Health and Welfare (grant number DOH101-TD-M-113-100001), and in part by the Taiwan Alzheimer Disease Association. The sponsors were not involved in executing the study, analyzing or interpreting the data, or writing the manuscript. Conflicts of interest None declared. References The very long list of references supporting this review has meant that only the most important are listed here and are represented by bold type throughout the text. The full list of references is available in Supplementary data available in Age and Ageing online. 1 Pot AM , Petrea I . Improving dementia care worldwide: ideas and advice on developing and implementing a National Dementia Plan. London: Bupa/ADI. 2013. Available at <www.alz.co.uk/sites/default/files/pdfs/global-dementia-plan-report-ENGLISH.pdf> (accessed 14 Feb, 2014). 2 Alzheimer’s Association . 2016 Alzheimer’s disease facts and figures . Alzheimers Dement 2016 ; 12 : 459 – 509 . CrossRef Search ADS PubMed 3 Olazarán J , Reisberg B , Clare L et al. . Nonpharmacological therapies in Alzheimer’s disease: a systematic review of efficacy . Dement Geriatr Cogn Disord 2010 ; 30 : 161 – 78 . Google Scholar CrossRef Search ADS PubMed 4 Lin JS , O’Connor E , Rossom RC et al. . Screening for cognitive impairment in older adults: a systematic review for the U.S. Preventive Services Task Force . Ann Intern Med 2013 ; 159 : 601 – 12 . Google Scholar PubMed 8 Borson S , Frank L , Bayley PJ et al. . Improving dementia care: the role of screening and detection of cognitive impairment . Alzheimers Dement 2013 ; 9 : 151 – 9 . Google Scholar CrossRef Search ADS PubMed 12 Tsoi KK , Chan JY , Hirai HW et al. . Cognitive tests to detect dementia: a systematic review and meta-analysis . JAMA 2015 ; 175 : 1450 – 8 . 13 Menon R , Larner AJ . Use of cognitive screening instruments in primary care: the impact of national dementia directives (NICE/SCIE, National Dementia Strategy) . Fam Pract 2011 ; 28 : 272 – 6 . Google Scholar CrossRef Search ADS PubMed 14 Fowler NR , Boustani MA , Frame A et al. . Effect of patient perceptions on dementia screening in primary care . J Am Geriatr Soc 2012 ; 60 : 1037 – 43 . Google Scholar CrossRef Search ADS PubMed 15 Prince M , Acosta D , Ferri CP et al. . The association between common physical impairments and dementia in low and middle income countries, and, among people with dementia, their association with cognitive function and disability. A 10/66 Dementia Research Group population-based study . Int J Geriatr Psychiatry 2011 ; 26 : 511 – 9 . Google Scholar CrossRef Search ADS PubMed 17 Barberger‐Gateau P , Commenges D , Gagnon M et al. . Instrumental activities of daily living as a screening tool for cognitive impairment and dementia in elderly community dwellers . J Am Geriatr Soc 1992 ; 40 : 1129 – 34 . Google Scholar CrossRef Search ADS PubMed 18 Cromwell DA , Eagar K , Poulos RG . The performance of instrumental activities of daily living scale in screening for cognitive impairment in elderly community residents . J Clin Epidemiol 2003 ; 56 : 131 – 7 . Google Scholar CrossRef Search ADS PubMed 19 De Lepeleire J , Aertgeerts B , Umbach I et al. . The diagnostic value of IADL evaluation in the detection of dementia in general practice . Aging Ment Health 2004 ; 8 : 52 – 7 . Google Scholar CrossRef Search ADS PubMed 20 Hancock P , Larner AJ . The diagnosis of dementia: diagnostic accuracy of an instrument measuring activities of daily living in a clinic-based population . Dement Geriatr Cogn Disord 2007 ; 23 : 133 – 9 . Google Scholar CrossRef Search ADS PubMed 21 Iavarone A , Milan G , Vargas G et al. . Role of functional performance in diagnosis of dementia in elderly people with low educational level living in Southern Italy . Aging Clin Exp Res 2007 ; 19 : 104 – 9 . Google Scholar CrossRef Search ADS PubMed 22 Koskas P , Henry-Feugeas MC , Feugeas JP et al. . The Lawton Instrumental Activities Daily Living/Activities Daily Living scales: a sensitive test to Alzheimer disease in community-dwelling elderly people? J Geriatr Psychiatry Neurol 2014 ; 27 : 85 – 93 . Google Scholar CrossRef Search ADS PubMed 23 Juva K , MÄKELÄ M , Erkinjuntti T et al. . Functional assessment scales in detecting dementia . Age Ageing 1997 ; 26 : 393 – 400 . Google Scholar CrossRef Search ADS PubMed 24 Hesseberg K , Bentzen H , Ranhoff AH et al. . Disability in instrumental activities of daily living in elderly patients with mild cognitive impairment and Alzheimer’s disease . Dement Geriatr Cogn Disord 2013 ; 36 : 146 – 53 . Google Scholar CrossRef Search ADS PubMed 29 Castilla‐Rilo J , López‐Arrieta J , Bermejo‐Pareja F et al. . Instrumental activities of daily living in the screening of dementia in population studies: a systematic review and meta‐analysis . Int J Geriatr Psychiatry 2007 ; 22 : 829 – 36 . Google Scholar CrossRef Search ADS PubMed 30 Lawton MP , Brody EM . Assessment of older people: self-maintaining and instrumental activities of daily living . Gerontologist 1969 ; 9 : 179 – 86 . Google Scholar CrossRef Search ADS PubMed 32 Sun Y , Lee HJ , Yang SC et al. . A nationwide survey of mild cognitive impairment and dementia, including very mild dementia, in Taiwan . PloS One 2014 ; 9 : e100303 . Google Scholar CrossRef Search ADS PubMed 37 Péres K , Helmer C , Amieva H et al. . Natural history of decline in instrumental activities of daily living performance over the 10 years preceding the clinical diagnosis of dementia: a prospective population‐based study . J Am Geriatr Soc 2008 ; 56 : 37 – 44 . Google Scholar CrossRef Search ADS PubMed 38 Fauth EB , Schwartz S , Tschanz JT et al. . Baseline disability in Activities of Daily Living predicts dementia risk even after controlling for baseline global cognitive ability and depressive symptoms . Int J Geriatr Psychiatry 2013 ; 28 : 597 – 606 . Google Scholar CrossRef Search ADS PubMed 39 Luck T , Luppa M , Angermeyer M et al. . Impact of impairment in instrumental activities of daily living and mild cognitive impairment on time to incident dementia: results of the Leipzig Longitudinal Study of the Aged . Psychol Med 2011 ; 41 : 1087 – 97 . Google Scholar CrossRef Search ADS PubMed 40 Milan G , Iavarone A , Vargas NF et al. . Effects of demographic and environmental variables on cognitive performance in a rural community sample of elderly people living in Southern Italy . Aging Clin Exp Res 2004 ; 16 : 398 – 402 . Google Scholar CrossRef Search ADS PubMed 41 Lechowski L , de Stampa M , Denis B et al. . Patterns of loss of abilities in instrumental activities of daily living in Alzheimer’s disease: the REAL cohort study . Dement Geriatr Cogn Disord 2008 ; 25 : 46 – 53 . Google Scholar CrossRef Search ADS PubMed 42 Pérès K , Helmer C , Amieva H et al. . Gender differences in the prodromal signs of dementia: Memory complaint and IADL-restriction. A prospective population-based cohort . J Alzheimers Dis 2011 ; 27 : 39 – 47 . Google Scholar CrossRef Search ADS PubMed 43 Narasimhalu K , Lee J , Auchus AP et al. . Improving detection of dementia in Asian patients with low education: combining the Mini-Mental State Examination and the Informant Questionnaire on Cognitive Decline in the Elderly . Dement Geriatr Cogn Disord 2007 ; 25 : 17 – 22 . Google Scholar CrossRef Search ADS PubMed 44 Chan QL , Shaik MA , Xu J et al. . The combined utility of a brief functional measure and performance-based screening test for case finding of cognitive impairment in primary healthcare . J Am Med Dir Assoc 2016 ; 17 ( 372 ): e9 – 11 . 46 Noel-Storr AH , McCleery JM , Richard E et al. . Reporting standards for studies of diagnostic test accuracy in dementia: The STARDdem Initiative . Neurology 2014 ; 83 : 364 – 73 . Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the British Geriatrics Society. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Age and Ageing Oxford University Press

Diagnostic accuracy of Instrumental Activities of Daily Living for dementia in community-dwelling older adults

Loading next page...
 
/lp/ou_press/diagnostic-accuracy-of-instrumental-activities-of-daily-living-for-orSsc5FcEn
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the British Geriatrics Society. All rights reserved. For permissions, please email: journals.permissions@oup.com
ISSN
0002-0729
eISSN
1468-2834
D.O.I.
10.1093/ageing/afy021
Publisher site
See Article on Publisher Site

Abstract

Abstract Background many people living with dementia remain underdiagnosed and unrecognised. Screening strategies are important for early detection. Objective to examine whether the Lawton’s Instrumental Activities of Daily Living (IADL) scale, compared with other cognitive screening tools—the Mini-Mental State Examination (MMSE), and the Ascertain Dementia 8-item Informant Questionnaire (AD8)—can identify older (≥ 65 years) adults with dementia. Design population-based cross-sectional observational study. Setting all 19 counties in Taiwan. Participants community-dwelling older adults (n = 10,340; mean age 74.87 ± 6.03). Methods all participants underwent a structured in-person interview. Dementia was identified using National Institute on Aging-Alzheimer’s Association core clinical criteria for all-cause dementia. Receiver operator characteristic curves were used to determine the discriminant abilities of the IADL scale, MMSE and AD8 to differentiate participants with and without dementia. Results we identified 917 (8.9%) participants with dementia, and 9,423 (91.1%) participants without. The discriminant abilities of the MMSE, AD8 and IADL scale (cutoff score: 6/7; area under curve = 0.925; sensitivity = 89%; specificity = 81%; positive likelihood ratio = 4.75; accuracy = 0.82) were comparable. Combining IADL with AD8 scores significantly improved overall accuracy: specificity = 93%; positive likelihood ratio = 11.74; accuracy = 0.92. Conclusions our findings support using IADL scale to screen older community-dwelling residents for dementia: it has discriminant power comparable to that of the AD8 and MMSE. Combining the IADL and the AD8 improves specificity. Alzheimer’s disease, dementia, screen, instrumental activity of daily living, diagnostic accuracy, older people Introduction Early detection and timely diagnosis of dementia, a healthcare priority in many countries and regions [1], are a gateway for providing healthcare and social services [2]. Studies [3, 4] have reported that early detection and timely diagnosis reduce the patient’s functional decline, the caregiver’s burden and the cost of care. Others [5–7] have reported that diagnostic coverage is only 40–50% in high-income countries, and that it is unlikely to exceed 5–10% in low- and middle-income countries. Many people living with dementia remain underdiagnosed and unrecognised. Screening strategies are important for early detection [8]. Valid and reliable cognitive tests are frequently used for early dementia detection, including performance-based measures, e.g. the Mini-Mental State Examination (MMSE), the Mini-Cog™ and informant-based measures like the Ascertain Dementia 8-item Informant Questionnaire (AD8) [4, 9–12]. However, only 20% of patients are screened using cognitive tests before being referred to a memory specialist [13]. Using cognitive tests for screening seems difficult in clinical and community settings [14]. A high prevalence of illiteracy and low-level education might be barriers to cognitive screening in middle and low-income countries [15]. The space, time and training required to properly administer conventional screening tools can also be a burden for community workers [16]. Furthermore, the cost-effectiveness of large-scale screening in detecting cognitive impairment has been debated, but identifying cognitive impairment during routine care is becoming critical [1, 4]. Therefore, it is necessary to find an alternative and commonly used screening tool that has no educational bias and is easy to administer in a community-based setting [8, 16]. Using the instrumental activities of daily living (IADL) as a screening tool for dementia has been proposed as an alternative strategy [17–23]. There is robust evidence that difficulty with IADL is one early feature of dementia. People with early-stage dementia often show declines in IADL performance, which requires complex cognitive processing [24, 25]. The National Institute on Aging-Alzheimer’s Association (NIA-AA) [26], Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition: DSM-IV-TR (DSM-IV) [27] and DSM-V [28] criteria for diagnosis of dementia require not only deficits in cognitive function, but also decreases in daily, social and occupational function. Because it is an early feature of dementia syndromes [25], IADL disability might be a good screening indicator for dementia. Using IADL as a screening tool for dementia has not been supported consistently by empirical studies, despite the fair specificity found in population-based studies [29]. Moreover, the diagnostic accuracy—especially specificity—was not as high in clinical settings as in community settings [19, 20]. Common findings of physical impairments or comorbidities in older (≥65 years) adults in clinical settings might contribute to IADL dependence. The Lawton IADL assessment is one of the most popular IADL scales. However, it was not developed to screen people with dementia [30]. If used in dementia screening, a cutoff score needs to be examined and established. The purpose of this study was to use a cross-sectional nationwide population to (i) evaluate the discriminant validity (accuracy and utility) of Lawton’s IADL scale to identify community-dwelling older adults with dementia; and (ii) to compare the discriminant validity of the IADL scale, MMSE and AD8, and the combined AD8 and IADL scale for dementia screening. Methods Study design and sampling A door-to-door survey of 10,432 potential participants was done by well-trained field interviewers between December 2011 and March 2013. To assess the interviewees, interviewers used the MMSE [31], the AD8, the Clinical Dementia Rating Scale (CDR) and a structured questionnaire that asked about demographics, medical history, cognitive and functional status (IADL and ADL), and lifestyle factors. The sampling process, training of interviewers, home visiting procedures, quality control and protocol approvals are described elsewhere [32]. Diagnosing dementia: procedures Dementia was diagnosed using the core clinical criteria for all-cause dementia from the NIA-AA [26]. The detailed diagnosis criteria are described elsewhere [32]. IADL performance Participants or informants were asked about each participant’s ability during the previous month to perform eight activities listed in Lawton’s IADL scale (Table 1). The ranges of the raw scores of the Lawton’s IADL items were 1–3, 1–4 and 1–5. The raw score of each item was converted to a dichotomised score (1 = independent, 0 = not independent). A summary score ranges from 0 (low function: dependent) to 8 (high function: independent) [30]. Data from participants who answered ‘unable to judge’ or ‘missing’ were scored as missing, and they were excluded from the ROC curves analyses. The AD8 is described in Appendix 1, and other measures of covariates are described in Appendix 2 (both appendices available at Age and Ageing online). Table 1. Demographics, and IADL, AD8 and MMSE performances (n = 10,360) and comparisons of characteristics of participants. Parameters Non-Dementiaa (n = 9,443) Dementiab (n = 917) P Male (n, %) 4,590 (48.6) 348 (37.9) <0.001 Age (years) (mean ± SD) 75.63 ± 6.39 81.63 ± 7.45 <0.001 Living area (n, %) <0.001  Rural 2,358 (25.0) 395 (43.1)  Urban 7,085 (75.0) 522 (56.9) Marital and living status (n, %) <0.001  Live alone 973 (10.3) 71 (7.7)  Live with partner only 38,69 (41.0) 264 (28.8)  Live with family 4,541 (48.1) 540 (58.9)  Other 60 (0.6) 42 (4.6) Educational level (n, %) <0.001  Illiterate 2,879 (30.5) 445 (48.5)  Literate 926 (9.8) 105 (11.5)  Elementary school 3,421 (36.2) 241 (26.3)  Junior high school 745 (7.9) 43 (4.7)  Senior high school 814 (8.6) 46 (5.0)  College 595 (6.3) 34 (3.7)  Graduate school and above 63 (0.7) 3 (0.3) MMSE (mean ± SD) 24.54 ± 4.72 12.84 ± 6.28 <0.001 AD8 (mean ± SD) 0.71 ± 1.52 4.97 ± 2.75 <0.001 AD8 ≥ 2 (n, %) 1,464 (15.5) 804 (87.7) <0.001 IADL, independent (n, %)  Shopping 7,571 (81.4) 77 (8.7) <0.001  Transportation 8,261 (89.1) 229 (25.6) <0.001  Finances 8,159 (90.5) 214 (24.7) <0.001  Telephone use 8,924 (95.2) 414 (46.0) <0.001  Medication 9,077 (98.5) 403 (63.9) <0.001  Food preparation 7,275 (81.8) 152 (17.3) <0.001  Household chores 8,147 (90.6) 296 (33.5) <0.001  Laundry 7,410 (83.7) 220 (25.1) <0.001 IADL (mean ± SD) 7.19 ± 1.63 2.76 ± 2.51 <0.001 IADL < 7 (n, %) 1,541 (18.7) 520 (89.0) <0.001 Parameters Non-Dementiaa (n = 9,443) Dementiab (n = 917) P Male (n, %) 4,590 (48.6) 348 (37.9) <0.001 Age (years) (mean ± SD) 75.63 ± 6.39 81.63 ± 7.45 <0.001 Living area (n, %) <0.001  Rural 2,358 (25.0) 395 (43.1)  Urban 7,085 (75.0) 522 (56.9) Marital and living status (n, %) <0.001  Live alone 973 (10.3) 71 (7.7)  Live with partner only 38,69 (41.0) 264 (28.8)  Live with family 4,541 (48.1) 540 (58.9)  Other 60 (0.6) 42 (4.6) Educational level (n, %) <0.001  Illiterate 2,879 (30.5) 445 (48.5)  Literate 926 (9.8) 105 (11.5)  Elementary school 3,421 (36.2) 241 (26.3)  Junior high school 745 (7.9) 43 (4.7)  Senior high school 814 (8.6) 46 (5.0)  College 595 (6.3) 34 (3.7)  Graduate school and above 63 (0.7) 3 (0.3) MMSE (mean ± SD) 24.54 ± 4.72 12.84 ± 6.28 <0.001 AD8 (mean ± SD) 0.71 ± 1.52 4.97 ± 2.75 <0.001 AD8 ≥ 2 (n, %) 1,464 (15.5) 804 (87.7) <0.001 IADL, independent (n, %)  Shopping 7,571 (81.4) 77 (8.7) <0.001  Transportation 8,261 (89.1) 229 (25.6) <0.001  Finances 8,159 (90.5) 214 (24.7) <0.001  Telephone use 8,924 (95.2) 414 (46.0) <0.001  Medication 9,077 (98.5) 403 (63.9) <0.001  Food preparation 7,275 (81.8) 152 (17.3) <0.001  Household chores 8,147 (90.6) 296 (33.5) <0.001  Laundry 7,410 (83.7) 220 (25.1) <0.001 IADL (mean ± SD) 7.19 ± 1.63 2.76 ± 2.51 <0.001 IADL < 7 (n, %) 1,541 (18.7) 520 (89.0) <0.001 MMSE, Mini-Mental State Examination; AD8, The Ascertain Dementia 8-item Informant Questionnaire; IADL, Lawton’s Instrumental Activities of Daily Living scale with modified scoring method. aNon-Dementia group includes the non-dementia control, the unclassified older adults and the mild cognitive impairment (MCI) subgroups. bDementia group includes older adults with very mild dementia and all-cause dementia. Table 1. Demographics, and IADL, AD8 and MMSE performances (n = 10,360) and comparisons of characteristics of participants. Parameters Non-Dementiaa (n = 9,443) Dementiab (n = 917) P Male (n, %) 4,590 (48.6) 348 (37.9) <0.001 Age (years) (mean ± SD) 75.63 ± 6.39 81.63 ± 7.45 <0.001 Living area (n, %) <0.001  Rural 2,358 (25.0) 395 (43.1)  Urban 7,085 (75.0) 522 (56.9) Marital and living status (n, %) <0.001  Live alone 973 (10.3) 71 (7.7)  Live with partner only 38,69 (41.0) 264 (28.8)  Live with family 4,541 (48.1) 540 (58.9)  Other 60 (0.6) 42 (4.6) Educational level (n, %) <0.001  Illiterate 2,879 (30.5) 445 (48.5)  Literate 926 (9.8) 105 (11.5)  Elementary school 3,421 (36.2) 241 (26.3)  Junior high school 745 (7.9) 43 (4.7)  Senior high school 814 (8.6) 46 (5.0)  College 595 (6.3) 34 (3.7)  Graduate school and above 63 (0.7) 3 (0.3) MMSE (mean ± SD) 24.54 ± 4.72 12.84 ± 6.28 <0.001 AD8 (mean ± SD) 0.71 ± 1.52 4.97 ± 2.75 <0.001 AD8 ≥ 2 (n, %) 1,464 (15.5) 804 (87.7) <0.001 IADL, independent (n, %)  Shopping 7,571 (81.4) 77 (8.7) <0.001  Transportation 8,261 (89.1) 229 (25.6) <0.001  Finances 8,159 (90.5) 214 (24.7) <0.001  Telephone use 8,924 (95.2) 414 (46.0) <0.001  Medication 9,077 (98.5) 403 (63.9) <0.001  Food preparation 7,275 (81.8) 152 (17.3) <0.001  Household chores 8,147 (90.6) 296 (33.5) <0.001  Laundry 7,410 (83.7) 220 (25.1) <0.001 IADL (mean ± SD) 7.19 ± 1.63 2.76 ± 2.51 <0.001 IADL < 7 (n, %) 1,541 (18.7) 520 (89.0) <0.001 Parameters Non-Dementiaa (n = 9,443) Dementiab (n = 917) P Male (n, %) 4,590 (48.6) 348 (37.9) <0.001 Age (years) (mean ± SD) 75.63 ± 6.39 81.63 ± 7.45 <0.001 Living area (n, %) <0.001  Rural 2,358 (25.0) 395 (43.1)  Urban 7,085 (75.0) 522 (56.9) Marital and living status (n, %) <0.001  Live alone 973 (10.3) 71 (7.7)  Live with partner only 38,69 (41.0) 264 (28.8)  Live with family 4,541 (48.1) 540 (58.9)  Other 60 (0.6) 42 (4.6) Educational level (n, %) <0.001  Illiterate 2,879 (30.5) 445 (48.5)  Literate 926 (9.8) 105 (11.5)  Elementary school 3,421 (36.2) 241 (26.3)  Junior high school 745 (7.9) 43 (4.7)  Senior high school 814 (8.6) 46 (5.0)  College 595 (6.3) 34 (3.7)  Graduate school and above 63 (0.7) 3 (0.3) MMSE (mean ± SD) 24.54 ± 4.72 12.84 ± 6.28 <0.001 AD8 (mean ± SD) 0.71 ± 1.52 4.97 ± 2.75 <0.001 AD8 ≥ 2 (n, %) 1,464 (15.5) 804 (87.7) <0.001 IADL, independent (n, %)  Shopping 7,571 (81.4) 77 (8.7) <0.001  Transportation 8,261 (89.1) 229 (25.6) <0.001  Finances 8,159 (90.5) 214 (24.7) <0.001  Telephone use 8,924 (95.2) 414 (46.0) <0.001  Medication 9,077 (98.5) 403 (63.9) <0.001  Food preparation 7,275 (81.8) 152 (17.3) <0.001  Household chores 8,147 (90.6) 296 (33.5) <0.001  Laundry 7,410 (83.7) 220 (25.1) <0.001 IADL (mean ± SD) 7.19 ± 1.63 2.76 ± 2.51 <0.001 IADL < 7 (n, %) 1,541 (18.7) 520 (89.0) <0.001 MMSE, Mini-Mental State Examination; AD8, The Ascertain Dementia 8-item Informant Questionnaire; IADL, Lawton’s Instrumental Activities of Daily Living scale with modified scoring method. aNon-Dementia group includes the non-dementia control, the unclassified older adults and the mild cognitive impairment (MCI) subgroups. bDementia group includes older adults with very mild dementia and all-cause dementia. Statistical analyses For each measure and the combined AD8 and IADL scale (‘and’ rule: the participant was considered positive only when both tests were positive) the area under the receiver-operating characteristic (ROC) curve (AUC) was calculated for each gender and educational subgroup. A cutoff score for each measure that best differentiated diagnostic groups (Dementia versus Non-Dementia) was determined using the Youden index [34]. A combination of the IADL and the MMSE was not evaluated because of copyright problems. Other details of the statistical analysis are provided in Appendix 3 (available at Age and Ageing online). Declaration of ethics and sources of funding This study was approved by National Taiwan University Hospital’s Institutional Review Board (201012021RB), and supported in part by grant DOH101-TD-M-113-100001 from the Taiwan Ministry of Health and Welfare, and in part by the Taiwan Alzheimer Disease Association. The sponsors were not involved in executing the study, analysing or interpreting the data, or writing the manuscript. Results Participant characteristics Of the 10,432 interviewees, 72 living in institutions were excluded, which left us 10,340 community-dwelling older adults to analyse. There were 917 older adults (8.85%) in the Dementia group, and 9,443 (91.15%) in the Non-Dementia group, including 2,021 (19.51%) with mild cognitive impairment (MCI) and 410 (3.96%) unclassified; 7,012 (67.68%) with intact cognitive function were the Non-Dementia Control group. Demographic information, IADL, MMSE and AD8 performance are presented in Table 1. Lifestyle characteristics and medical and clinical information are presented in Appendix 4. IADL disabilities independently discriminating between older people with and without dementia even after controlling for cognitive function are presented in Appendix 5 and Appendix 6 (all appendices available at Age and Ageing online). The discriminative ability of IADL, MMSE, AD8, and the combination of IADL and AD8 The IADL scale discriminated extremely well between the Dementia and Non-Dementia groups: cutoff IADL scores were 6/7 (AUC = 0.925; 95% CI = 0.915–0.935; sensitivity = 89%; specificity = 81%; LR+ = 4.75; LR− = 0.13; and classification accuracy = 0.82) (Figure 1 and Table 2). Those unable to perform independently six or more IADL items from Lawton’s scale (scored < 7) were 4.75 times more likely to be diagnosed with dementia, which is considered a fair LR+ [36]. Using the IADL cutoff score of 6/7, 89% of those with dementia (true-positives) and 81% of those without dementia (true-negatives) were expected to be correctly identified. Table 2. Cutoff scores for optimal sensitivity and specificity of MMSE, AD8 and IADL for dementiaa by gender and educational level. Index and subgroups Cutoff AUC (95% CI) P Youden indexb Sensitivity Specificity LR+ LR− Accuracy MMSE  All 20.5 0.931 (0.922–0.940) <0.001 0.71 0.89 0.82 5.03 0.14 0.83  Gender   Male 21.5 0.929 (0.915–0.943) <0.001 0.73 0.87 0.85 5.93 0.15 0.85   Female 17.5 0.937 (0.925–0.948) <0.001 0.74 0.84 0.90 8.21 0.18 0.89  Educational level   Illiterate 15.5 0.950 (0.939–0.961) <0.001 0.79 0.89 0.91 9.49 0.13 0.90   Elementary school 21.5 0.937 (0.925–0.950) <0.001 0.74 0.89 0.85 6.02 0.13 0.85   Junior high school or above 24.5 0.944 (0.919–0.969) <0.001 0.79 0.90 0.89 8.31 0.12 0.89 AD8  All 1.5 0.905 (0.894–0.917) <0.001 0.72 0.88 0.84 5.65 0.15 0.85  Gender   Male 1.5 0.912 (0.893–0.931) <0.001 0.74 0.87 0.87 6.79 0.15 0.87   Female 1.5 0.897 (0.883–0.912) <0.001 0.70 0.88 0.82 4.88 0.15 0.83  Educational level   Illiterate 1.5 0.891 (0.875–0.907) <0.001 0.66 0.92 0.74 3.56 0.10 0.77   Elementary school 1.5 0.898 (0.878–0.919) <0.001 0.71 0.84 0.87 6.66 0.19 0.87   Junior high school or above 1.5 0.913 (0.879–0.947) <0.001 0.74 0.82 0.92 10.65 0.20 0.92 IADL  All 6.5 0.925 (0.915–0.935) <0.001 0.70 0.89 0.81 4.75 0.13 0.82  Gender   Male 6.5 0.926 (0.910–0.942) <0.001 0.71 0.93 0.78 4.22 0.09 0.79   Female 6.5 0.928 (0.915–0.940) <0.001 0.71 0.86 0.84 5.50 0.16 0.84  Educational level   Illiterate 6.5 0.913 (0.898–0.929) <0.001 0.64 0.89 0.75 3.60 0.14 0.77   Elementary school 6.5 0.916 (0.898–0.934) <0.001 0.69 0.87 0.82 4.86 0.16 0.82   Junior high school or above 6.5 0.947 (0.917–0.978) <0.001 0.82 0.94 0.88 7.60 0.06 0.88 AD8+IADL  All 1.5/6.5 0.947 (0.939–0.955) <0.001 0.74 0.81 0.93 11.74 0.21 0.92 Index and subgroups Cutoff AUC (95% CI) P Youden indexb Sensitivity Specificity LR+ LR− Accuracy MMSE  All 20.5 0.931 (0.922–0.940) <0.001 0.71 0.89 0.82 5.03 0.14 0.83  Gender   Male 21.5 0.929 (0.915–0.943) <0.001 0.73 0.87 0.85 5.93 0.15 0.85   Female 17.5 0.937 (0.925–0.948) <0.001 0.74 0.84 0.90 8.21 0.18 0.89  Educational level   Illiterate 15.5 0.950 (0.939–0.961) <0.001 0.79 0.89 0.91 9.49 0.13 0.90   Elementary school 21.5 0.937 (0.925–0.950) <0.001 0.74 0.89 0.85 6.02 0.13 0.85   Junior high school or above 24.5 0.944 (0.919–0.969) <0.001 0.79 0.90 0.89 8.31 0.12 0.89 AD8  All 1.5 0.905 (0.894–0.917) <0.001 0.72 0.88 0.84 5.65 0.15 0.85  Gender   Male 1.5 0.912 (0.893–0.931) <0.001 0.74 0.87 0.87 6.79 0.15 0.87   Female 1.5 0.897 (0.883–0.912) <0.001 0.70 0.88 0.82 4.88 0.15 0.83  Educational level   Illiterate 1.5 0.891 (0.875–0.907) <0.001 0.66 0.92 0.74 3.56 0.10 0.77   Elementary school 1.5 0.898 (0.878–0.919) <0.001 0.71 0.84 0.87 6.66 0.19 0.87   Junior high school or above 1.5 0.913 (0.879–0.947) <0.001 0.74 0.82 0.92 10.65 0.20 0.92 IADL  All 6.5 0.925 (0.915–0.935) <0.001 0.70 0.89 0.81 4.75 0.13 0.82  Gender   Male 6.5 0.926 (0.910–0.942) <0.001 0.71 0.93 0.78 4.22 0.09 0.79   Female 6.5 0.928 (0.915–0.940) <0.001 0.71 0.86 0.84 5.50 0.16 0.84  Educational level   Illiterate 6.5 0.913 (0.898–0.929) <0.001 0.64 0.89 0.75 3.60 0.14 0.77   Elementary school 6.5 0.916 (0.898–0.934) <0.001 0.69 0.87 0.82 4.86 0.16 0.82   Junior high school or above 6.5 0.947 (0.917–0.978) <0.001 0.82 0.94 0.88 7.60 0.06 0.88 AD8+IADL  All 1.5/6.5 0.947 (0.939–0.955) <0.001 0.74 0.81 0.93 11.74 0.21 0.92 AUC, area under the curve; LR+, positive likelihood ratio; LR−, negative likelihood ratio; accuracy, classification accuracy; MMSE, Mini-Mental State Examination; AD8, The Ascertain Dementia 8-item Informant Questionnaire; IADL, Lawton’s Instrumental Activities of Daily Living scale. aNumber of participants: Dementia group, n = 584; Non-Dementia group, n = 8,221 (participants with any missing data in the IADL scale were excluded). bA suggested cutoff was decided using the optimised Youden Index, if the differences of the Youden Index among multiple candidates were within 0.01. Table 2. Cutoff scores for optimal sensitivity and specificity of MMSE, AD8 and IADL for dementiaa by gender and educational level. Index and subgroups Cutoff AUC (95% CI) P Youden indexb Sensitivity Specificity LR+ LR− Accuracy MMSE  All 20.5 0.931 (0.922–0.940) <0.001 0.71 0.89 0.82 5.03 0.14 0.83  Gender   Male 21.5 0.929 (0.915–0.943) <0.001 0.73 0.87 0.85 5.93 0.15 0.85   Female 17.5 0.937 (0.925–0.948) <0.001 0.74 0.84 0.90 8.21 0.18 0.89  Educational level   Illiterate 15.5 0.950 (0.939–0.961) <0.001 0.79 0.89 0.91 9.49 0.13 0.90   Elementary school 21.5 0.937 (0.925–0.950) <0.001 0.74 0.89 0.85 6.02 0.13 0.85   Junior high school or above 24.5 0.944 (0.919–0.969) <0.001 0.79 0.90 0.89 8.31 0.12 0.89 AD8  All 1.5 0.905 (0.894–0.917) <0.001 0.72 0.88 0.84 5.65 0.15 0.85  Gender   Male 1.5 0.912 (0.893–0.931) <0.001 0.74 0.87 0.87 6.79 0.15 0.87   Female 1.5 0.897 (0.883–0.912) <0.001 0.70 0.88 0.82 4.88 0.15 0.83  Educational level   Illiterate 1.5 0.891 (0.875–0.907) <0.001 0.66 0.92 0.74 3.56 0.10 0.77   Elementary school 1.5 0.898 (0.878–0.919) <0.001 0.71 0.84 0.87 6.66 0.19 0.87   Junior high school or above 1.5 0.913 (0.879–0.947) <0.001 0.74 0.82 0.92 10.65 0.20 0.92 IADL  All 6.5 0.925 (0.915–0.935) <0.001 0.70 0.89 0.81 4.75 0.13 0.82  Gender   Male 6.5 0.926 (0.910–0.942) <0.001 0.71 0.93 0.78 4.22 0.09 0.79   Female 6.5 0.928 (0.915–0.940) <0.001 0.71 0.86 0.84 5.50 0.16 0.84  Educational level   Illiterate 6.5 0.913 (0.898–0.929) <0.001 0.64 0.89 0.75 3.60 0.14 0.77   Elementary school 6.5 0.916 (0.898–0.934) <0.001 0.69 0.87 0.82 4.86 0.16 0.82   Junior high school or above 6.5 0.947 (0.917–0.978) <0.001 0.82 0.94 0.88 7.60 0.06 0.88 AD8+IADL  All 1.5/6.5 0.947 (0.939–0.955) <0.001 0.74 0.81 0.93 11.74 0.21 0.92 Index and subgroups Cutoff AUC (95% CI) P Youden indexb Sensitivity Specificity LR+ LR− Accuracy MMSE  All 20.5 0.931 (0.922–0.940) <0.001 0.71 0.89 0.82 5.03 0.14 0.83  Gender   Male 21.5 0.929 (0.915–0.943) <0.001 0.73 0.87 0.85 5.93 0.15 0.85   Female 17.5 0.937 (0.925–0.948) <0.001 0.74 0.84 0.90 8.21 0.18 0.89  Educational level   Illiterate 15.5 0.950 (0.939–0.961) <0.001 0.79 0.89 0.91 9.49 0.13 0.90   Elementary school 21.5 0.937 (0.925–0.950) <0.001 0.74 0.89 0.85 6.02 0.13 0.85   Junior high school or above 24.5 0.944 (0.919–0.969) <0.001 0.79 0.90 0.89 8.31 0.12 0.89 AD8  All 1.5 0.905 (0.894–0.917) <0.001 0.72 0.88 0.84 5.65 0.15 0.85  Gender   Male 1.5 0.912 (0.893–0.931) <0.001 0.74 0.87 0.87 6.79 0.15 0.87   Female 1.5 0.897 (0.883–0.912) <0.001 0.70 0.88 0.82 4.88 0.15 0.83  Educational level   Illiterate 1.5 0.891 (0.875–0.907) <0.001 0.66 0.92 0.74 3.56 0.10 0.77   Elementary school 1.5 0.898 (0.878–0.919) <0.001 0.71 0.84 0.87 6.66 0.19 0.87   Junior high school or above 1.5 0.913 (0.879–0.947) <0.001 0.74 0.82 0.92 10.65 0.20 0.92 IADL  All 6.5 0.925 (0.915–0.935) <0.001 0.70 0.89 0.81 4.75 0.13 0.82  Gender   Male 6.5 0.926 (0.910–0.942) <0.001 0.71 0.93 0.78 4.22 0.09 0.79   Female 6.5 0.928 (0.915–0.940) <0.001 0.71 0.86 0.84 5.50 0.16 0.84  Educational level   Illiterate 6.5 0.913 (0.898–0.929) <0.001 0.64 0.89 0.75 3.60 0.14 0.77   Elementary school 6.5 0.916 (0.898–0.934) <0.001 0.69 0.87 0.82 4.86 0.16 0.82   Junior high school or above 6.5 0.947 (0.917–0.978) <0.001 0.82 0.94 0.88 7.60 0.06 0.88 AD8+IADL  All 1.5/6.5 0.947 (0.939–0.955) <0.001 0.74 0.81 0.93 11.74 0.21 0.92 AUC, area under the curve; LR+, positive likelihood ratio; LR−, negative likelihood ratio; accuracy, classification accuracy; MMSE, Mini-Mental State Examination; AD8, The Ascertain Dementia 8-item Informant Questionnaire; IADL, Lawton’s Instrumental Activities of Daily Living scale. aNumber of participants: Dementia group, n = 584; Non-Dementia group, n = 8,221 (participants with any missing data in the IADL scale were excluded). bA suggested cutoff was decided using the optimised Youden Index, if the differences of the Youden Index among multiple candidates were within 0.01. Figure 1. View largeDownload slide Receiver operating characteristic curve for discriminating between elderly adults with and without dementia. (A) IADL; (B) MMSE; (C) AD8; and (D) IADL+AD8. Figure 1. View largeDownload slide Receiver operating characteristic curve for discriminating between elderly adults with and without dementia. (A) IADL; (B) MMSE; (C) AD8; and (D) IADL+AD8. An MMSE cutoff score of 20/21 resulted in an AUC value of 0.931. An AD8 cutoff score of 1/2 resulted in an AUC value of 0.905 (Table 2). There were no significant differences in the diagnostic and classification accuracy between the IADL scale, the MMSE, and the AD8 (P > 0.05). When the IADL and AD8 were combined, the AUC improved to 0.947 (95% CI = 0.939–0.955), sensitivity decreased by 8% (81%), specificity increased by 12% (93%), utility rose to excellent [LR+ (11.74)], and classification accuracy significantly increased by 10% [χ2] (1) = 58.36, P < 0.0001. IADL was gender- and education-independent The cutoff scores of the IADL and AD8 scales did not differ by gender or educational subgroup. However, different genders and educational subgroups had different MMSE cutoff scores (Table 2). Discussion Our most important finding was that Lawton’s IADL scale discriminated between older adults with and without dementia well as did the MMSE and the AD8. We also found that the IADL scale discriminated between those with and without dementia independently of the MMSE and AD8, even after comorbidities and lifestyle risk factors had been adjusted for. Several longitudinal studies [25, 37–39] reported that IADL performance predicts incident dementia. Our findings support the notion that declines in functional performance can be an independent predictor for dementia; thus, we conclude that IADL disabilities can be used as a screening index. Using IADL as a dementia screening strategy has been overlooked in favour of focusing on cognitive assessments [4]. This study supports that the IADL is a valid tool for identifying older adults with dementia in communities with a relatively low-prevalence rate (about 8–10%). An IADL scale score of 6 or less detected 92.5% of dementia among community-dwelling older adults (sensitivity = 0.89, specificity = 0.81), which was comparable with other, similar studies (estimated summarised AUC = 0.88, weighted average sensitivity = 90.18%, weighted average specificity = 78.79%) [29]. However, the diagnostic accuracy reported by some studies [19, 20] done in settings with higher prevalence rates, such as memory clinics, was not satisfactory, particularly the specificity. Therefore, we hypothesise that IADL measures are more effective for identifying community-dwelling older adults with dementia. This study is the first to compare the three frequently used tools to screen for dementia in community-dwelling older adults: Lawton’s IADL scale, the MMSE, and the AD8. All three detected more than 90% of the cases. The major concern is that factors unrelated to dementia, e.g. physiological degeneration and comorbidities, can also lead to functional deterioration in the older adults. Thus, using the IADL scale to screen for dementia might lead to less satisfactory specificity and a relatively high false-positive rate [29]. In the current study, the related discriminant indices of the IADL scale were as good as those of the AD8 and the MMSE. This might be because each screening tool has its own shortcomings. For example, the MMSE yielded a high false-positive rate in the older adults and in participants with lower education levels [4, 10, 16, 40]. We also found that the MMSE yielded gender and educational effects, but that the IADL and AD8 did not. Therefore, we can reasonably assume that although the IADL as a dementia screening scale did not have an ideal specificity, its discriminant validity as a screening tool was as good as that of the AD8 and the MMSE, with the additional benefit of easier administration. Furthermore, the IADL scale is frequently used in routine health examinations to provide information about a person’s functional status, which can be used to assess the care services they need [38]. Thus, its use for dementia detection will not add an extra burden for care providers. Moreover, evidence that IADL disabilities indicate a high risk of dementia should increase public awareness of dementia. It is interesting to note that the cutoff scores of Lawton’s IADL scale were identical for both genders, and that similar diagnostic accuracy was reported. Traditionally, gender differences would be considered when assessing IADL disabilities. A few studies [24, 41, 42] found that the patterns of functional abilities vary for men and women. However, the current study and Juva [23] found no substantial gender differences in IADLs. Therefore, the IADL scale can be used for both genders with the same standard. Recent research [13, 43, 44] found that combining two or more tools for dementia screening might increase the specificity and lower the probability of mislabeling a typically aging older adult as someone with dementia. This might prevent the expensive and potentially stressful series of tests for dementia, and minimise the possible adverse psychological effects of screening. Our findings showed that the correlations between the MMSE, AD8, and IADL were only moderate, and that each could independently predict dementia, which implied that each screening tool might evaluate different traits of dementia, and therefore complement each other. We suggest combining the IADL and AD8 results to reduce the false-positive rate. Those who screen positive on both will need a comprehensive evaluation. The current study has several strengths. It analysed a large representative sample from a nationwide epidemiology study that used face-to-face interviews. All participants underwent a detailed assessment to ensure a reliable diagnosis of dementia based on NIA-AA core clinical criteria, a primary diagnostic guideline for dementia research. Additionally, this is the first study that extensively compares the three most frequently used screening assessments (Lawton’s IADL, the MMSE, and the AD8). Lastly, we propose valid screening cutoff scores that allow immediate clinical application in community settings. This study has some limitations. First, we did not assess other factors that influence IADL, e.g. hand function, depression, and environmental factors. Second, there might be disagreement between self-reports and informant-based IADL data [45], but the data collected using these two methods were not differentiated in the study. Third, the results can be affected by ‘incorporation bias’, whereby the IADL disabilities, MMSE, and subjective complaints of cognitive problems (i.e. the AD8) form a part of the reference standard (the NIA-AA criteria). This bias is common in dementia diagnostic studies, according to the reporting standards for studies of diagnostic test accuracy in dementia [46]. Therefore, screening accuracy is associated with a tendency to overestimate the value of the three measures. In addition, validation has not been established for another population. Because of the likelihood of overfitting the data in this study, diagnostic cutpoints in an exploratory analysis in other cohorts are likely to be less accurate. Finally, the results of the current study can be generalised to settings with a low-prevalence rate of dementia and for people with similar cultural backgrounds, such as Asian communities. Additional prospective (longitudinal) studies aimed at establishing the predictive validity of the IADL scale in discriminating people with and without dementia are warranted. Because advanced activities of daily living (AADLs) (e.g. interests in hobbies, driving and trip planning) are more complex than are IADLs, future studies might want to investigate whether AADLs are more sensitive for detecting early-stage dementia [47]. Conclusion IADL disabilities independently discriminate between older people with and without dementia even after controlling for cognitive function. This study supports using Lawton’s IADL scale with a modified scoring method as an alternative dementia screening tool for community-dwelling older adults. Using the IADL as an initial screening tool can decrease the bias associated with gender and with educational levels. The discriminant ability of the IADL for dementia in community-dwelling older adults is comparable to that of the MMSE and AD8, the conventional cognitive-based screening tools. We recommend combining the IADL and the AD8 to improve specificity. Key points The discriminant ability of Lawton’s Instrumental Activities of Daily Living (IADL) scale to identify persons with dementia is comparable to that of Mini-Mental State Examination (MMSE), and Ascertain Dementia 8-item Informant Questionnaire (AD8). The combination of two informant-based tools, the Instrumental Activities of Daily Living (IADL) and Ascertain Dementia 8-item Informant Questionnaire (AD8), reduces the false-positive rate and improves specificity. The Instrumental Activities of Daily Living (IADL) is feasible to detect cognitive impairment in a routine healthcare examination. Supplementary data Supplementary data mentioned in the text are available to subscribers in Age and Ageing online. Acknowledgements We thank the study participants and their families for their cooperation and support, the staff members of Taiwan Alzheimer Disease Association, and the interviewers for their valuable contributions. Funding This work was supported in part by the Taiwan Ministry of Health and Welfare (grant number DOH101-TD-M-113-100001), and in part by the Taiwan Alzheimer Disease Association. The sponsors were not involved in executing the study, analyzing or interpreting the data, or writing the manuscript. Conflicts of interest None declared. References The very long list of references supporting this review has meant that only the most important are listed here and are represented by bold type throughout the text. The full list of references is available in Supplementary data available in Age and Ageing online. 1 Pot AM , Petrea I . Improving dementia care worldwide: ideas and advice on developing and implementing a National Dementia Plan. London: Bupa/ADI. 2013. Available at <www.alz.co.uk/sites/default/files/pdfs/global-dementia-plan-report-ENGLISH.pdf> (accessed 14 Feb, 2014). 2 Alzheimer’s Association . 2016 Alzheimer’s disease facts and figures . Alzheimers Dement 2016 ; 12 : 459 – 509 . CrossRef Search ADS PubMed 3 Olazarán J , Reisberg B , Clare L et al. . Nonpharmacological therapies in Alzheimer’s disease: a systematic review of efficacy . Dement Geriatr Cogn Disord 2010 ; 30 : 161 – 78 . Google Scholar CrossRef Search ADS PubMed 4 Lin JS , O’Connor E , Rossom RC et al. . Screening for cognitive impairment in older adults: a systematic review for the U.S. Preventive Services Task Force . Ann Intern Med 2013 ; 159 : 601 – 12 . Google Scholar PubMed 8 Borson S , Frank L , Bayley PJ et al. . Improving dementia care: the role of screening and detection of cognitive impairment . Alzheimers Dement 2013 ; 9 : 151 – 9 . Google Scholar CrossRef Search ADS PubMed 12 Tsoi KK , Chan JY , Hirai HW et al. . Cognitive tests to detect dementia: a systematic review and meta-analysis . JAMA 2015 ; 175 : 1450 – 8 . 13 Menon R , Larner AJ . Use of cognitive screening instruments in primary care: the impact of national dementia directives (NICE/SCIE, National Dementia Strategy) . Fam Pract 2011 ; 28 : 272 – 6 . Google Scholar CrossRef Search ADS PubMed 14 Fowler NR , Boustani MA , Frame A et al. . Effect of patient perceptions on dementia screening in primary care . J Am Geriatr Soc 2012 ; 60 : 1037 – 43 . Google Scholar CrossRef Search ADS PubMed 15 Prince M , Acosta D , Ferri CP et al. . The association between common physical impairments and dementia in low and middle income countries, and, among people with dementia, their association with cognitive function and disability. A 10/66 Dementia Research Group population-based study . Int J Geriatr Psychiatry 2011 ; 26 : 511 – 9 . Google Scholar CrossRef Search ADS PubMed 17 Barberger‐Gateau P , Commenges D , Gagnon M et al. . Instrumental activities of daily living as a screening tool for cognitive impairment and dementia in elderly community dwellers . J Am Geriatr Soc 1992 ; 40 : 1129 – 34 . Google Scholar CrossRef Search ADS PubMed 18 Cromwell DA , Eagar K , Poulos RG . The performance of instrumental activities of daily living scale in screening for cognitive impairment in elderly community residents . J Clin Epidemiol 2003 ; 56 : 131 – 7 . Google Scholar CrossRef Search ADS PubMed 19 De Lepeleire J , Aertgeerts B , Umbach I et al. . The diagnostic value of IADL evaluation in the detection of dementia in general practice . Aging Ment Health 2004 ; 8 : 52 – 7 . Google Scholar CrossRef Search ADS PubMed 20 Hancock P , Larner AJ . The diagnosis of dementia: diagnostic accuracy of an instrument measuring activities of daily living in a clinic-based population . Dement Geriatr Cogn Disord 2007 ; 23 : 133 – 9 . Google Scholar CrossRef Search ADS PubMed 21 Iavarone A , Milan G , Vargas G et al. . Role of functional performance in diagnosis of dementia in elderly people with low educational level living in Southern Italy . Aging Clin Exp Res 2007 ; 19 : 104 – 9 . Google Scholar CrossRef Search ADS PubMed 22 Koskas P , Henry-Feugeas MC , Feugeas JP et al. . The Lawton Instrumental Activities Daily Living/Activities Daily Living scales: a sensitive test to Alzheimer disease in community-dwelling elderly people? J Geriatr Psychiatry Neurol 2014 ; 27 : 85 – 93 . Google Scholar CrossRef Search ADS PubMed 23 Juva K , MÄKELÄ M , Erkinjuntti T et al. . Functional assessment scales in detecting dementia . Age Ageing 1997 ; 26 : 393 – 400 . Google Scholar CrossRef Search ADS PubMed 24 Hesseberg K , Bentzen H , Ranhoff AH et al. . Disability in instrumental activities of daily living in elderly patients with mild cognitive impairment and Alzheimer’s disease . Dement Geriatr Cogn Disord 2013 ; 36 : 146 – 53 . Google Scholar CrossRef Search ADS PubMed 29 Castilla‐Rilo J , López‐Arrieta J , Bermejo‐Pareja F et al. . Instrumental activities of daily living in the screening of dementia in population studies: a systematic review and meta‐analysis . Int J Geriatr Psychiatry 2007 ; 22 : 829 – 36 . Google Scholar CrossRef Search ADS PubMed 30 Lawton MP , Brody EM . Assessment of older people: self-maintaining and instrumental activities of daily living . Gerontologist 1969 ; 9 : 179 – 86 . Google Scholar CrossRef Search ADS PubMed 32 Sun Y , Lee HJ , Yang SC et al. . A nationwide survey of mild cognitive impairment and dementia, including very mild dementia, in Taiwan . PloS One 2014 ; 9 : e100303 . Google Scholar CrossRef Search ADS PubMed 37 Péres K , Helmer C , Amieva H et al. . Natural history of decline in instrumental activities of daily living performance over the 10 years preceding the clinical diagnosis of dementia: a prospective population‐based study . J Am Geriatr Soc 2008 ; 56 : 37 – 44 . Google Scholar CrossRef Search ADS PubMed 38 Fauth EB , Schwartz S , Tschanz JT et al. . Baseline disability in Activities of Daily Living predicts dementia risk even after controlling for baseline global cognitive ability and depressive symptoms . Int J Geriatr Psychiatry 2013 ; 28 : 597 – 606 . Google Scholar CrossRef Search ADS PubMed 39 Luck T , Luppa M , Angermeyer M et al. . Impact of impairment in instrumental activities of daily living and mild cognitive impairment on time to incident dementia: results of the Leipzig Longitudinal Study of the Aged . Psychol Med 2011 ; 41 : 1087 – 97 . Google Scholar CrossRef Search ADS PubMed 40 Milan G , Iavarone A , Vargas NF et al. . Effects of demographic and environmental variables on cognitive performance in a rural community sample of elderly people living in Southern Italy . Aging Clin Exp Res 2004 ; 16 : 398 – 402 . Google Scholar CrossRef Search ADS PubMed 41 Lechowski L , de Stampa M , Denis B et al. . Patterns of loss of abilities in instrumental activities of daily living in Alzheimer’s disease: the REAL cohort study . Dement Geriatr Cogn Disord 2008 ; 25 : 46 – 53 . Google Scholar CrossRef Search ADS PubMed 42 Pérès K , Helmer C , Amieva H et al. . Gender differences in the prodromal signs of dementia: Memory complaint and IADL-restriction. A prospective population-based cohort . J Alzheimers Dis 2011 ; 27 : 39 – 47 . Google Scholar CrossRef Search ADS PubMed 43 Narasimhalu K , Lee J , Auchus AP et al. . Improving detection of dementia in Asian patients with low education: combining the Mini-Mental State Examination and the Informant Questionnaire on Cognitive Decline in the Elderly . Dement Geriatr Cogn Disord 2007 ; 25 : 17 – 22 . Google Scholar CrossRef Search ADS PubMed 44 Chan QL , Shaik MA , Xu J et al. . The combined utility of a brief functional measure and performance-based screening test for case finding of cognitive impairment in primary healthcare . J Am Med Dir Assoc 2016 ; 17 ( 372 ): e9 – 11 . 46 Noel-Storr AH , McCleery JM , Richard E et al. . Reporting standards for studies of diagnostic test accuracy in dementia: The STARDdem Initiative . Neurology 2014 ; 83 : 364 – 73 . Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the British Geriatrics Society. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Age and AgeingOxford University Press

Published: Mar 8, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off