Abstract
Abstract
Background
Subjective cognitive decline (SCD) and mild cognitive impairment (MCI) are high-risk factors for dementia. We developed a cognitive measurement tool for screening SCD and MCI in community-dwelling elderly individuals.
Objective
This study investigated the feasibility of using the Brief Elderly Cognitive Screening Inventory (BECSI) as a screening measure for MCI and SCD in community elderly.
Design
A cross-sectional validation study.
Participants
The study included 1642 community-dwelling older adults aged ≥60 years.
Outcome measures
The Cronbach’s α and split-half coefficients were calculated to test its reliability. The BECSI scores of the normal control group, SCD group and MCI group were compared. The internal consistency analysis, correlation analysis with the neuropsychiatric inventory (NPI) and core neuropsychological test (CNT) were conducted. The screening efficacy of BECSI was verified by receiver operating characteristic curve.
Results
BECSI was a self-report questionnaire. Its Cronbach’s α coefficient and split-half coefficient were respectively 0.923 and 0.888. The correlation coefficients between the total score and individual items ranged from 0.185 to 0.813, and were also significantly correlated with NPI and CNT. Statistically significant differences were observed among the three groups in the total scores. The areas under the curves for distinguishing SCD from normal cognitive and MCI from SCD are 0.835 and 0.889, respectively, with the optimal cut-off points of 12.5 and 16.5.
Conclusion
BECSI is quick and easy to administer, and can be used as a feasible and useful measure for screening SCD and MCI in community-dwelling older adults.
Keywords: PSYCHIATRY, Old age psychiatry, Psychometrics
STRENGTHS AND LIMITATIONS OF THIS STUDY.
Brief Elderly Cognitive Screening Inventory (BECSI) is a self-developed and localised cognitive screening tool for the elderly that is culturally appropriate.
The development of BECSI is based on the theoretical foundation established through a literature review, with a focus on the assessment of memory function, while also taking into account daily living abilities and psychiatric symptoms, balancing comprehensiveness and simplicity.
The preliminary application of BECSI in the elderly community in China has confirmed its good reliability and validity.
This study adopted stratified cluster random sampling, and the sample is representative in terms of demographic characteristics.
If the relationship between BECSI scores and imaging can be clarified, the results will be more reliable.
Introduction
With the development of social economy and the improvement of medical conditions, the elderly population has increased sharply, the problems related to ageing have become increasingly prominent and the incidence of dementia caused by various reasons has increased significantly. China has the largest elderly population in the world, and the prevalence of dementia is as high as 6.0% for individuals aged 60 years and older, of which 60% are Alzheimer’s disease (AD).1 The early stages of AD development include subjective cognitive decline (SCD) and mild cognitive impairment (MCI), during which aggressive intervention may impede disease progression, but once it progresses to the dementia stage, there is currently no cure.2 MCI and SCD are common among elderly people in Chinese communities, overall MCI prevalence was estimated to be 15·5%, representing 38·77 million people in China, but the proportion of those who seek medical help proactively is low, as they believe that memory loss is a natural part of ageing and does not require treatment. Currently, the rate of dementia diagnosis in China is only 26.9%, and most patients are already in the irreversible stage of AD dementia at the time of diagnosis.3 Hence, screening for SCD or MCI in the elderly population and early intervention is of great significance. Currently, the elderly population over 60 years old in China exceeds 260 million,4 while the availability of community neuropsychology professionals remains limited. Faced with such a large target population, there is an urgent need for a simple and reliable self-report cognitive screening tool to identify high-risk individuals for dementia, who can then be referred to professional staff for further objective cognitive assessment and clinical diagnosis.
At present, there are many methods for screening SCD and MCI, such as the Memory Assessment Questionnaire (MAC-Q), Everyday Cognition (E-Cog), Subjective Cognitive Decline Questionnaire (SCD-Q), Subjective Cognitive Decline Interview (SCD-I) and Dichotomy Question.5,9 However, some of these scales involve only one or fewer cognitive domains, and some need to be implemented by professionals. In the early stage of dementia, there may be subtle changes in multiple cognitive domains. Relying solely on memory assessment may not be sufficient to fully detect early-stage dementia. The MAC-Q mainly focuses on memory-related issues and provides limited assessment of other cognitive domains, such as executive function and language ability. The E-Cog covers multiple cognitive domains, including memory, language, visual-spatial ability, planning, organisation and attention; however, it lacks assessment of emotional and behavioural changes. The Dementia Screening Questionnaire for the Elderly can comprehensively assess cognitive function, daily living ability and psychopathological symptoms, but it takes a lot of time.10 The Montreal Cognitive Assessment (MoCA) has a good discriminative ability for MCI screening, but some items have cultural specificity, such as the ‘animal naming’ item. In addition, the educational level of rural elderly people in China is generally low, and the copy cube and draw clock items cannot be completed by elderly people without writing experience. Although the Beijing version of MoCA has been localised, its accuracy in discriminating MCI is lower than that of Western studies.11 Other commonly used cognitive screening tools include the Mini-Mental State Examination (MMSE) and the Clock Drawing Test (CDT).12 13 MMSE is significantly influenced by educational attainment. It has a ‘ceiling effect’ for elderly individuals with higher educational levels and is prone to false positives for those with lower education. CDT mainly reflects spatial structure ability and is also not suitable for illiterate and low-educated elderly people. Both of these tools can effectively distinguish AD from normal individuals, but their detection performance for MCI is limited.
To date, there is no consensus on the best screening tool for SCD. Due to variations in cultural and social backgrounds across different countries or regions, the outcomes of neuropsychological assessments are somewhat influenced, thereby limiting their applicability for replication and utilisation; thus, it is necessary to customise localised screening tools according to the national conditions. In the early stage, we prepared the Brief Elderly Cognitive Screening Inventory (BECSI), a self-reported questionnaire with simple content and easy operation, which has been proved to have good reliability and validity in community promotion and application.14 15 The purpose of this study was to retest BECSI in a large community sample to evaluate its efficacy and feasibility in screening SCD and MCI.
Methods
Study procedure and participant selection criteria
The investigation was performed from September 2023 to February 2024, and a stratified cluster random sampling method was used to select the participants. We randomly selected two administrative districts in Wuxi City. One represents the urban core area, where the economy is mainly driven by commerce and services, the proportion of elderly residents is relatively high and the general level of education among the population is relatively high. The other district is located in the suburbs, with an economy centred on manufacturing and tourism, a relatively younger population and a slightly lower average educational level among the elderly. These two districts are representative in terms of geography and demographics, reflecting the broader population characteristics of the study area. To obtain a sufficient sample size within the limited time and resources and ensure the feasibility and efficiency of the study, we randomly selected four streets within the two districts and recruited elderly people ≥aged 60 as research participants. Inclusion criteria: Han ethnicity, 60–85 years old, primary school or above education, and vision and hearing accessibility. Exclusion criteria: serious physical and mental illness, significant anxiety and depression, and had been diagnosed with dementia. A household survey was conducted based on the provided list of elderly individuals from the community health service centre. General demographic data and medical history were collected through interviews, followed by inviting eligible elderly participants to undergo neuropsychological assessments and health examinations at the community health service centre. Informed consent was obtained from all participants and their guardians. According to the sample size estimation formula, n = (Zα/2) ² × P (1 – P)/d², by referring to the literature,1,3 with probability P set at 0.2, allowable error d = 0.1 × P, and a significance level of 0.05, the sample size n = 1.96² × 0.2 × (1–0.2)/(0.1×0. 2)² = 1536. Considering a 10% dropout rate, the sample size was expanded to 1650 cases.
Patient and public involvement
Patients or the public were not involved in the design, conduct, reporting, or dissemination plans of our research.
Assessment and diagnosis procedure
Subjective cognitive function was evaluated using the BECSI, a self-report questionnaire designed to assess perceived changes in cognitive function over the past 6 months. The BECSI comprises 14 items that measure five functional domains: memory function (six items), verbal function (two items), temporal orientation (two items), execution efficiency (two items) and psychopathological behaviour (two items). Each item is assigned a score based on the following grading scale: 0=have none in the preceding 6 months; 1=rarely (1–2 times per month in the preceding 6 months); 2=sometimes (1–2 times per week in the preceding 6 months) and 3=always (daily occurrence in the preceding 6 months). Each item of the BECSI was scored on a scale ranging from 0 to 3, with the total score ranging from 0 to 42, with higher scores indicating more pronounced SCD symptoms. For the BECSI, see online supplemental appendix 1.
Objective cognitive assessment included (1) a battery of core neuropsychological tests (CNTs) contains four cognitive domains: (a) memory function: Picture-Symbol Matching (PSM); (b) visual space function: Block Design (BD); (c) language function: Word Completion (WC); and (d) executive function: Trail-Making Test (TMT).16 17 The raw scores obtained from the aforementioned four CNTs were converted into standardised scores, where higher scores indicate better cognitive functioning (>7 is considered normal, 5–7 indicates MCI and ≤4 suggests severe cognitive impairment). (2) Global cognitive: Quick Cognitive Screening Scale for the Elderly (QCSS-E) was used, including 12 subtests of immediate memory, object naming, visual spatial ability, language fluency, numerical breadth, abstraction ability, auditory imitation, visual imitation, command following capability, delayed memory recall capacity, simple calculation skills, as well as time and space orientation. The total score on this scale is 90, with higher scores indicating better cognitive functioning.18 (3) Clinical Dementia Rating Scale (CDR): evaluated six cognitive domains, including memory, orientation, judgement and problem solving, community affairs, home and hobbies, as well as personal care. The CDR is established by clinical scoring rules, where CDR=0 indicates no dementia and CDR 0.5, 1, 2 or 3 indicates questionable, mild, moderate or severe.19 (4) Other measures employed include the Hachinski Ischemic Score (HIS), which was used to differentiate between degenerative and vascular dementia while the activities of daily living (ADL) was used to evaluate social functioning.20 21
The assessment of psycho-behavioural symptoms included (1) the neuropsychiatric interview (NPI-Q), which evaluated 12 psychotic symptoms: delusions, hallucinations, agitation, dysphoria, anxiety, apathy, irritability, euphoria, disinhibition, aberrant motor behaviour, night-time behaviour disturbances, and appetite and eating abnormalities.22 Each abnormal behaviour was rated on a scale of 0–3 based on severity for a total score of 36. A higher score indicated more severe psychotic symptoms. The cut-off score for dementia was set at ≥10. (2) The Hamilton Anxiety Scale (HAMA) and the Hamilton Depression Scale (HAMD-17) were used for assessing negative affective states.23 A total score ≤7 on either HAMA or HAMD-17 was considered within the normal range, while a score >7 indicated a significant level of anxiety or depression.
The personnel involved in the screening work included psychiatrists, psychology postgraduate students and psychometricians. All researchers received unified training and passed the consistency test, achieving a high inter-rater reliability (>90%). Clinical interviews and cognitive assessments were conducted jointly by psychiatrists and psychology postgraduate students, while core neurological tests were carried out by psychometricians. The health check-up encompassed physical examination, neurological examination and laboratory tests. Based on the interview and examination results, the two chief physicians finally determined the diagnosis. The diagnostic criteria for SCD and MCI followed the 2014 International Working Group (SCD-I) research standards and Petersen diagnostic criteria, respectively.24,26 Finally, participants were divided into three groups: (1) normal cognitive (NC) group: individuals had normal scores in all cognitive domains (scores of PSM, Build Block (BB), WC, TMT>7, QCSS-E≥81 and CDR=0). (2) SCD group: individuals had cognitive complaints and minimal impairment in one domain (only one of PSM, BB, WC and TMT scored between 5 and 7, and the others were all >7, QCSS-E≥81 and CDR=0). (3) MCI group: individuals had cognitive complaints and mild impairment in cognitive domains (two or more of PSM, BB, WC and TMT scored between 5 and 7, others >7 and QCSS-E≥71, CDR=0.5). Participants in all three groups preserved ability to perform daily activities and social functions; HIS score<4, no serious physical disease and mental disorder.
Statistical analysis
Data were analysed using SPSS 24.0 software (IBM Corporation, Armonk, New York, USA). Continuous variables with a normal distribution were described as the mean±SD, while continuous variables with a skewed distribution were expressed as the median and quartile, and categorical variables were expressed as n (%). χ2 test or analysis of variance or K-W test was used to assess group differences, and p<0 .05 was considered to be statistically significant. GLM was used to analyse the influencing factors of BECSI score. Item analysis was carried out by using the item difficulty (the comparison between the average score of all subjects in an item and the full score value of the item), item differentiation (the comparison between the top 27% high-scoring group and the bottom 27% low-scoring group) and the response rate of different grades of each item. Cronbach’s α coefficient and Spearman–Brown coefficient method were used for reliability analysis. Validity was analysed by internal consistency analysis, correlation analysis with the NPI and CNT. Receiver operating characteristic (ROC) curves were used to evaluate the ability of BECSI to distinguish between NC-SCD and SCD-MCI, as well as the best cut-off score, sensitivity and specificity of BECSI.
Results
Participant characteristics
1650 elderly volunteers aged ≥60 were recruited, and a total of 1560 volunteers completed clinical interviews and various examinations. Based on the outcomes of the interview and examination, after excluding mild dementia, significant anxiety and depression, as well as serious physical and mental illness, a total of 1462 valid samples were obtained (see details in figure 1). These included 770 males and 692 females, with an average age of (71.59±6.28) years and an average education duration of (9.61±2.82) years. According to diagnostic criteria, 1462 participants were divided into the NC group (941 cases), the SCD group (224 cases) and the MCI group (297 cases). No statistical differences were observed on gender, marital status, smoking and body mass index (BMI) between the three groups. There were statistically significant differences in the average age, education level, alcohol drinking, hypertension, diabetes, HAMA, HAMD-17 and NPI-Q scores between the three groups (p<0.05). After controlling for age and education level, the analysis of covariance revealed statistically significant differences in QCSS-E and CNT scores among the three groups (p<0.001). The SCD group exhibited lower scores on PSM and TMT compared with the NC group, while scoring higher than the MCI group. There was no significant difference in BB, WC and QCSS-E scores between the SCD group and the NC group (p>0.05). The MCI group demonstrated lower QCSS-E and CNT scores compared with both the NC and SCD groups, as shown in table 1.
Figure 1. Sample flowchart. MCI, mild cognitive impairment; NC, normal cognitive; SCD, subjective cognitive decline.
Table 1. Comparison of demographic and clinical characteristics for NC, SCD and MCI.
| Index | NC (n=941) (%) |
SCD (n=224) (%) |
MCI (n=297) (%) |
χ2/F/H | P value | Pairwise comparison |
|---|---|---|---|---|---|---|
| Gender | 1.246 | 0.536 | ||||
| Male | 501 (53.24) | 121 (54.02) | 148 (49.83) | – | ||
| Female | 440 (46.76) | 103 (45.98) | 149 (50.17) | |||
| Age (years) | 71.13±6.26 | 72.40±6.38 | 73.85±5.50 | 23.088 | 0.001 | NC<SCD<MCI |
| Marriage status | 4.369 | 0.113 | ||||
| In marriage | 781 (83.00) | 186 (83.04) | 231 (77.78) | – | ||
| Single/others | 160 (17.00) | 38 (16.96) | 66 (22.22) | |||
| Education (years) | 9.93±2.93 | 9.33±2.96 | 8.79±2.07 | 20.458 | <0.001 | NC>SCD>MCI |
| Smoking | 4.567 | 0.102 | ||||
| Yes | 160 (17.00) | 28 (12.50) | 58 (19.53) | – | ||
| No | 781 (83.00) | 196 (87.50) | 239 (80.47) | |||
| Alcohol drinking | 30.545 | 0.000 | ||||
| Yes | 86 (9.14) | 16 (7.14) | 59 (19.87) | NC=SCD<MCI | ||
| No | 855 (90.86) | 208 (92.86) | 238 (80.13) | |||
| Hypertension | 7.180 | 0.028 | ||||
| Yes | 289 (30.71) | 72 (32.14) | 116 (39.06) | NC<MCI | ||
| No | 652 (69.29) | 152 (67.86) | 181 (60.94) | |||
| Diabetes | 13.539 | 0.001 | ||||
| Yes | 85 (9.03) | 33 (14.73) | 47 (15.82) | NC<SCD=MCI | ||
| No | 856 (90.97) | 191 (85.27) | 250 (84.18) | |||
| BMI (kg/m2) | 23.40±2.85 | 23.36±2.91 | 23.43±3.07 | 0.039 | 0.962 | – |
| HAMA (points) | 2.91±2.18 | 3.67±2.23 | 3.24±2.21 | 11.595 | <0.001 | NC<MCI<SCD |
| HAMD-17 (points) | 2.04±1.72 | 2.84±1.99 | 2.24±1.67 | 18.983 | <0.001 | NC=MCI<SCD |
| QCSS-E (points) | 81.48±5.68 | 80.94±5.94 | 74.75±2.74 | 188.577 | <0.001 | NC=SCD>MCI |
| NPI-Q (points) | 2.0 (1.0, 6.0) | 3.5 (2.0, 8.0) | 6.0 (2.0, 8.0) | 62.832 | <0.001 | NC<SCD=MCI |
| PSM (points) | 9.90±1.95 | 8.40±2.07 | 7.06±1.65 | 157.107 | <0.001 | NC>SCD>MCI |
| BB (points) | 9.91±1.87 | 9.88±3.43 | 8.63±2.06 | 27.901 | <0.001 | NC=SCD>MCI |
| WC (points) | 11.93±2.46 | 11.79±2.71 | 10.04±2.93 | 32.816 | <0.001 | NC=SCD>MCI |
| TMT (points) | 9.94±1.55 | 9.07±2.40 | 7.20±1.53 | 166.149 | <0.001 | NC>SCD>MCI |
Values are mean±SD or median (first quartile, third quartile).
BB, Build Block; BMI, body mass index; HAMA, Hamilton Anxiety Scale; HAMD-17, Hamilton Depression Scale; MCI, mild cognitive impairment; NC, normal cognitive; NPI-Q, Neuropsychiatric Interview; PSM, Picture-Symbol Matching; QCSS‐E, Quick Cognitive Screening Scale for the Elderly; SCD, subjective cognitive decline; TMT, Trail-Making Test; WC, Word Completion.
Comparisons of BECSI scores
The GLM analysis was conducted to identify impact factors by taking the total scores of BECSI as the dependent variable and taking age and educated level as independent variables. Results indicated that neither age nor education had a statistically significant effect on the total BECSI score (F age=3.203, p>0.05, Feducation=2.276, p>0.05). The BECSI scores of the three groups were compared, and the results revealed statistically significant differences (p<0.001) in both total and each domain scores (except verbal function) between the NC group and the SCD group, as well as between the SCD group and the MCI group, as shown in table 2.
Table 2. Comparison of total and subtest BECSI scores between NC, SCD and MCI.
| Index | NC (n=941) | SCD (n=224) | MCI (n=297) | F/H | P value | LSD |
|---|---|---|---|---|---|---|
| Memory function | 7.22±1.74 | 9.44±2.22 | 12.09±1.74 | 842.257 | <0.001 | NC<SCD<MCI |
| Temporal orientation | 0.44±0.51 | 0.70±0.47 | 0.95±0.77 | 98.235 | <0.001 | NC<SCD<MCI |
| Verbal function | 0.79±0.50 | 0.81±0.62 | 1.55±0.82 | 188.842 | <0.001 | NC=SCD<MCI |
| Execution efficiency | 0.71±0.64 | 1.39±0.76 | 2.39±1.25 | 481.245 | <0.001 | NC<SCD<MCI |
| Psychopathological behaviour | 1.0 (1.0, 1.0) | 1.0 (1.0, 2.0) | 2.0 (1.0, 3.0) | 223.242 | <0.001 | NC<SCD<MCI |
| BECSI | 10.37±2.44 | 13.95±2.79 | 19.27±3.39 | 1232.241 | <0.001 | NC<SCD<MCI |
BECSI, Brief Elderly Cognitive Screening Inventory; LSD, Least Significant Difference; MCI, mild cognitive impairment; NC, normal cognitive; SCD, subjective cognitive decline.
Item analysis
Project analysis revealed that the difficulty of each BECSI item was between 0.083 and 0.397. Except item 14, the differentiation of other items ranged from 0.349 to 0.770, and the average response rates of different grades (0–3 points) were 45.5%, 32.8%, 18.4% and 3.4% (online supplemental appendix 2). Item 14: “Are you feeling that you are not speaking as fluently as you used to?” The definition of this item may be somewhat ambiguous, resulting in low discrimination. We are considering revising this item 14 in future versions of BECSI as: “Are you feeling that your speaking speed is slower than before?”
Reliability and validity test
Cronbach’s α coefficient of BECSI was 0.923, and the Cronbach’s α coefficient of each item ranged from 0.913 to 0.927 (online supplemental appendix 3). Guttman Split-Half Coefficient for BECSI was 0.888. All items were significantly correlated with the total score, with the majority of items exhibiting a correlation coefficient exceeding 0.6. Except for items 13/14 and 5/14, the other items of BECSI were significantly correlated with each other (online supplemental appendix 4). The total score and domain scores (except for the Temporal Orientation—WC) of BECSI are significantly correlated with QCSS-E, NPI-Q, PSM, BD, WC and TMT, as presented in table 3.
Table 3. Correlation between BECSI and QCSS-E, NPI-Q and CNT scores.
| Index | QCSS-E | NPI-Q | PSM | BB | WC | TMT |
|---|---|---|---|---|---|---|
| Memory function | −0.428** | 0.218** | −0.400** | −0.201** | −0.239** | −0.340** |
| Temporal orientation | −0.160** | 0.070** | −0.175** | −0.077** | −0.005 | −0.138** |
| Verbal function | −0.230** | 0.175** | −0.181** | −0.148** | −0.178** | −0.165** |
| Execution efficiency | −0.222** | 0.135** | −0.221** | 0.142** | −0.094** | −0.140** |
| Psychopathological behaviour | −0.335** | 0.179** | −0.295** | −0.166** | −0.186** | −0.282** |
| BECSI | −0.448** | 0.242** | −0.416** | −0.226** | −0.241** | −0.350** |
**p<0.01.
BB, Build Block; BECSI, Brief Elderly Cognitive Screening Inventory; CNT, core neuropsychological test; NPI-Q, Neuropsychiatric Interview; PSM, Picture-Symbol Matching; QCSS‐E, Quick Cognitive Screening Scale for the Elderly; TMT, Trail-Making Test; WC, Word Completion.
ROC curve analysis
The results of the ROC curve analysis demonstrated that the BECSI total score and subtests of memory function, execution efficiency exhibited relatively strong abilities to distinguish between individuals with NC and SCD, with corresponding area under the curve (AUC) of 0.835, 0.785 and 0.733, respectively. Additionally, for discriminating SCD from MCI, the BECSI total score, memory function, verbal function and execution efficiency subtests showed a relatively strong discriminatory power, with AUCs of 0.889, 0.827, 0.738 and 0.723, respectively (refer to table 4 and figures2 3).
Table 4. ROC analysis of distinguishing SCD and MCI by BECSI total and subtest scores.
| Index | NC vs SCD | SCD vs MCI | ||||
|---|---|---|---|---|---|---|
| AUC | SE | 95% CI | AUC | SE | 95% CI | |
| BECSI | 0.835 | 0.015 | 0.806 to 0.863 | 0.889 | 0.014 | 0.862 to 0.916 |
| Memory function | 0.785 | 0.019 | 0.748 to 0.822 | 0.827 | 0.019 | 0.791 to 0.864 |
| Execution efficiency | 0.733 | 0.018 | 0.697 to 0.768 | 0.723 | 0.022 | 0.681 to 0.766 |
| Psychopathological behaviour | 0.600 | 0.022 | 0.557 to 0.644 | 0.652 | 0.025 | 0.603 to 0.701 |
| Temporal orientation | 0.632 | 0.020 | 0.593 to 0.672 | 0.581 | 0.025 | 0.533 to 0.630 |
| Verbal function | 0.503 | 0.023 | 0.458 to 0.549 | 0.738 | 0.022 | 0.696 to 0.781 |
AUC, area under the curve; BECSI, Brief Elderly Cognitive Screening Inventory; MCI, mild cognitive impairment; NC, normal cognitive; ROC, receiver operating characteristic curve; SCD, subjective cognitive decline.
Figure 2. Receiver operating characteristic curves of Brief Elderly Cognitive Screening Inventory (BECSI) in screening for normal cognitive versus subjective cognitive decline.
Figure 3. Receiver operating characteristic curves of Brief Elderly Cognitive Screening Inventory (BECSI) in screening for subjective cognitive decline vs mild cognitive impairment.
Using clinical diagnosis as the reference standard, we calculated the cut-off value, sensitivity, specificity and Youden’s index of the total BECSI score for differentiating between SCD and MCI (table 5). The results indicated that a total BECSI score of 12.5 achieved the best balance between sensitivity (0.656) and specificity (0.877) for distinguishing NC from SCD, with a Youden’s index of 0.533 and an accuracy rate of 83.4%. Similarly, a total BECSI score of 16.5 achieved optimal sensitivity (0.817) and specificity (0.791) for discriminating between SCD and MCI, with a Youden’s index of 0.608 and an accuracy rate of 81.2%.
Table 5. Cut-off points, sensitivity, specificity and Youden’s index of BECSI total score for distinguishing SCD and MCI.
| NC vs SCD (n=1165) | SCD vs MCI (n=521) | ||||||
|---|---|---|---|---|---|---|---|
| BECSI cut-off | Sensitivity | Specificity | Youden’s index | BECSI cut-off | Sensitivity | Specificity | Youden’s index |
| 8.50 | 0.991 | 0.238 | 0.229 | 12.50 | 0.344 | 1.000 | 0.344 |
| 9.50 | 0.982 | 0.348 | 0.330 | 13.50 | 0.469 | 0.997 | 0.465 |
| 10.50 | 0.897 | 0.537 | 0.434 | 14.50 | 0.594 | 0.923 | 0.516 |
| 11.50 | 0.804 | 0.654 | 0.457 | 15.50 | 0.714 | 0.865 | 0.580 |
| 12.50 | 0.656 | 0.877 | 0.533 | 16.50 | 0.817 | 0.791 | 0.608 |
| 13.50 | 0.531 | 0.918 | 0.449 | 17.50 | 0.893 | 0.684 | 0.576 |
| 14.50 | 0.406 | 0.927 | 0.333 | 18.50 | 0.938 | 0.512 | 0.449 |
| 15.50 | 0.286 | 0.969 | 0.255 | 19.50 | 0.978 | 0.404 | 0.382 |
BECSI, Brief Elderly Cognitive Screening Inventory; MCI, mild cognitive impairment; NC, normal cognitive; SCD, subjective cognitive decline.
Discussion
The National Institute on Aging-Alzheimer Association classifies SCD as a third-stage preclinical AD, where overall cognitive functioning falls within normal range but some complex neurocognitive processes may exhibit minimal damage without reaching the objective impairment degree of MCI.27,29 In order to distinguish NC, SCD and MCI, we used neuropsychological test scores 1 SD below the normal mean as evidence of objective impairment. MCI must have objective impairment in two or more cognitive domains, while SCD impairment in no more than one cognitive domain. BD and WC mainly measure working memory, reasoning ability, processing speed and verbal fluency, while the PSM and TMT reflect the learning strategy and executive function for establishing new associations. Our study showed that after adjusting for confounding variables like age and education, it was found that the PSM and TMT scores of the SCD group were situated between those of the NC and MCI groups, indicating that patients with SCD presented mild impairments in their learning ability and executive function, while patients with MCI had mild impairments in all complex neurocognitive functions, which is consistent with other studies on SCD.30 31
China is a country with a large ageing population, and identifying cognitive impairment in the elderly at an early stage is a challenging issue that society has to confront. From a health economics perspective, self-reported cognitive screening questionnaires can be employed to evaluate an elderly person’s memory or cognitive function. Subsequently, for those with suspected cognitive impairment, a more in-depth cognitive assessment and examination should be conducted by trained professionals. BECSI was developed in response to this need and is a brief (14-item) and easy-to-use (self-report) cognitive screening questionnaire that can be completed in just 5 min. Cognitive complaints constitute the core symptoms of SCD or MCI. Some patients might undergo changes in personality and mood, along with a decline in work efficiency.32 33 A meta-analysis indicates that approximately 27% of individuals with SCD develop MCI within 4 years of follow-up, while approximately 14% progress to dementia. The clinical symptoms of MCI mainly cover three key aspects: cognitive decline, slight impairment of complex instrumental daily functioning and non-cognitive neuropsychiatric symptoms. However, the commonly used clinical cognitive function screening tools, such as MMSE, MoCA, SCD-Q and E-Cog, only focus on assessing the degree of cognitive impairment, while ADL is only used to evaluate daily living ability.34 These tools do not involve the assessment of neuropsychiatric symptoms, and although the NPI is used to assess mental symptoms, it does not cover cognitive function. This single-domain assessment method may reduce the sensitivity of the screening. Moreover, MCI can be further classified into single-domain MCI and multidomain MCI. The assessment of memory, attention, executive function, language and visuospatial function is crucial for the diagnosis of MCI subtypes and the analysis of the causes. Given this, during the development of BECSI, emphasis was placed on the assessment of multidomain cognitive functions, while also taking into account daily living abilities and mental behaviours, in order to more comprehensively reflect the clinical characteristics of patients. From the test results, the difficulty of the BECSI items was 0.083–0.390, indicating that the probability of symptoms appearing in community elderly people was <0.4. The average response rates for the different levels (0–3 points) were 45.5%, 32.8%, 18.4% and 3.4%, which were very close to the prevalence rates of SCD or MCI in domestic and foreign studies, indicating that the questionnaire content could accurately reflect the cognitive impairment symptoms of the elderly in the community and also demonstrates the representativeness of the sample in this study.35,37 The discriminant validity of each item ranged from 0.349 to 0.770, indicating that the questionnaire possesses excellent discriminatory validity and can effectively differentiate MCI or SCD from the general population. The α coefficient of the total score of BECSI was 0.923, and all items showed significant correlations with the total score. Except for a few items, all items were significantly correlated with each other, reflecting good internal consistency of the questionnaire items, which were both related and relatively independent. They could measure different cognitive functions independently and jointly measure overall cognitive function.
From the test results of elderly individuals in the community, a comparison of questionnaire scores among the NC group, SCD group and MCI group revealed that both the total score of the questionnaire and differences between groups within each domain were statistically significant. Specifically, the score of the NC group was significantly lower than that of the SCD group, and the score of the SCD group was significantly lower than that of the MCI group. These findings indicate good empirical validity for BECSI. In addition, BECSI is not affected by education level or age, suggesting that the assessment could serve as an effective early warning indicator for detecting early cognitive impairment in elderly individuals with diverse educational backgrounds and across different age groups in communities.
Some scholars have used E-Cog and the Prospective Retrospective Memory Questionnaire to assess SCD, revealing a significant association between self-reported cognitive complaints among community-dwelling older adults and deficits in episodic memory and executive function. This means that the cognitive problems perceived by the elderly themselves may reflect their actual decline in cognitive function. Moreover, these complaints can serve as indicators of subtle changes in objective cognition.38,40 QCSS-E is a global cognitive rating scale which is highly correlated with the MMSE and ADAS.18 ADAS is a comprehensive assessment scale used to evaluate the cognitive and non-cognitive functions of patients with AD. The cognitive subscale (ADAS-Cog) specifically assesses patients' memory, language, attention, and other cognitive abilities, and is a commonly used tool in clinical research. Similarly, the application of NPI in Chinese community elderly population also has good reliability and validity.41 The core neuropsychological tests (CNTs) used in this study are derived from the CCAS and the Halstead-Reitan Neuropsychological Battery (H-R). The CCAS is a localized cognitive assessment tool specifically designed for the Chinese population, covering multiple cognitive domains and capable of comprehensively evaluating an individual's cognitive functions. The H-R, on the other hand, is an internationally recognized neuropsychological test battery, primarily used to assess memory, attention, language, and executive functions. It has been widely used in China and has established normative criteria appropriate to the cultural context of China.16 17 The total score and domain scores of BECSI were significantly correlated with QCSS-E, NPI and CNT, which indicated that BECSI possesses good criterion-related validity. The findings of this study suggest that self-reported cognitive impairment might be one of the early behavioural markers of AD, while BECSI can be employed as a feasible tool for clue investigation.
ROC analysis demonstrated that the BECSI total score and subtests of memory function and execution efficiency exhibited excellent discriminative abilities between NC and SCD, and the AUC was 0.835, 0.785 and 0.733, respectively. Furthermore, the BECSI total score and subtests of memory function, execution efficiency and verbal function displayed substantial discriminative potential in distinguishing SCD from MCI, with AUC of 0.889, 0.827, 0.723 and 0.738, respectively. The findings indicated that memory impairment was the most prevalent and prominent symptom of early cognitive impairment, followed by executive function deficits and language dysfunction, while spatiotemporal orientation disorder and psychopathological symptoms were relatively less common in the early stages of cognitive impairment. Additionally, these results also suggested that SCD and MCI may share common pathological characteristics. As the disease progresses to the MCI stage, there is an expansion in the extent of damage and an increase in pathological severity. In the present study, when the cut-off value was set to 12.5, Youden’s index for distinguishing SCD from NC was the maximum (0.533), with a featuring moderate sensitivity (65.6%) and excellent specificity (87.7%), and the accuracy was (83.4%). Comparatively, when the cut-off value was set to 16.5, Youden’s index for distinguishing MCI from SCD was the maximum (0.608), with excellent sensitivity (81.7%) and good specificity (79.1), and the accuracy was (81.2%). It has been reported that the sum of the best sensitivity and specificity for identifying NC and SCD and NC and MCI by E-Cog was 1.24 and 1.42, respectively.6 38 The combined maximum sensitivity and specificity of MES in identifying NC and SCD, as well as NC and MCI, were 1.35 and 1.62, respectively. Additionally, the combined maximum sensitivity and specificity of SCD-Q9 in identifying NC and MCI was 1.55.42 43 Therefore, it can be observed that BECSI demonstrates high sensitivity and specificity when screening for SCD and MCI among similar scales. In routine practice such as in primary healthcare settings and during the initial diagnostic stage, if the BECSI score ranges from 13 to 16, an objective cognitive assessment should be carried out. If cognitive function is found to be normal, it is recommended to proceed with APOEε4 genetic testing to evaluate the risk. This strategy is straightforward and practical, aligning well with the resource capabilities of primary hospitals and the needs of patients. If the BECSI score exceeds 17, it is advisable to conduct imaging studies, laboratory tests and a comprehensive neuropsychological evaluation to establish a diagnosis. In cases where MCI is confirmed, further classification is necessary to support the development of a personalised treatment plan.44
In conclusion, BECSI demonstrates excellent performance in identifying SCD and MCI among the elderly in the community. Moreover, BECSI is easy to administer and does not require a significant amount of time; thus, it is suitable for screening early cognitive dysfunction in the elderly population and has high feasibility. If the total score is combined with sensitive subtests, the accuracy of detecting subtle cognitive decline may be higher.
The participants in this study were all regular elderly people from the community, with a relatively large sample size and good representativeness, making the results reliable. In addition, we excluded elderly individuals with obvious anxiety, depression and mild AD in our study to ensure the reliability of self-reported information. AD is a continuous spectrum of disorders, and its pathological changes can quietly emerge during the asymptomatic stage. Biomarker detection plays a crucial role in improving the accuracy of early AD diagnosis. However, this study has certain limitations. First, this was a cross-sectional study and cannot verify the predictive value of the results for future cognitive decline. Second, this study lacks detection of biomarkers related to AD and thus fails to verify the direct connection between the questionnaire and the pathology of AD. Future research needs to conduct longitudinal follow-up studies and comprehensively analyse multidimensional information such as clinical manifestations, imaging examinations and biomarker detection to further verify the potential and application value of BECSI in future AD screening.
Supplementary material
Footnotes
Funding: This study was supported by scientific and technological achievements of Wuxi Municipal Health Commission and appropriate technology extension projects (T202023).
Prepublication history and additional supplemental material for this paper are available online. To view these files, please visit the journal online (https://doi.org/10.1136/bmjopen-2024-097390).
Provenance and peer review: Not commissioned; externally peer reviewed.
Patient consent for publication: Not applicable.
Ethics approval: This study received ethical approval from the Ethics Committee of Wuxi Mental Health Center (no.: WXMH-CIRB2021LLky001). All protocols were conducted at the Wuxi Community Health Service Center. Since all the research subjects were elderly people over 60 years old, some might have had potential cognitive impairments. According to the Declaration of Helsinki, all participants and their guardians signed the informed consent form before participating in the study.
Patient and public involvement: Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Data availability statement
Data are available upon reasonable request.
References
- 1.Jia L, Du Y, Chu L, et al. Prevalence, risk factors, and management of dementia and mild cognitive impairment in adults aged 60 years or older in China: a cross-sectional study. Lancet Public Health. 2020;5:e661–71. doi: 10.1016/S2468-2667(20)30185-7. [DOI] [PubMed] [Google Scholar]
- 2.Jack CR, Jr, Albert MS, Knopman DS, et al. Introduction to the recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement. 2011;7:257–62. doi: 10.1016/j.jalz.2011.03.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Zhang H, Zhou Y, Ma J, et al. Understanding help-seeking decisions in people with subjective cognitive decline: A systematic review of qualitative studies. Geriatr Nurs (Lond) 2021;42:1507–16. doi: 10.1016/j.gerinurse.2021.10.013. [DOI] [PubMed] [Google Scholar]
- 4.National Burea of Statistics Bulletin of the Seventh National Census (No.5) http://www.stats.gov.cn/tjsj/zxfb/202105/t20210510_1817181.html Available.
- 5.Vogel A, Salem LC, Andersen BB, et al. Differences in quantitative methods for measuring subjective cognitive decline - results from a prospective memory clinic study. Int Psychogeriatr. 2016;28:1513–20. doi: 10.1017/S1041610216000272. [DOI] [PubMed] [Google Scholar]
- 6.Tomaszewski Farias S, Mungas D, Harvey DJ, et al. The measurement of everyday cognition: development and validation of a short form of the Everyday Cognition scales. Alzheimers Dement. 2011;7:593–601. doi: 10.1016/j.jalz.2011.02.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Rami L, Mollica MA, García-Sanchez C, et al. The Subjective Cognitive Decline Questionnaire (SCD-Q): a validation study. J Alzheimers Dis. 2014;41:453–66. doi: 10.3233/JAD-132027. [DOI] [PubMed] [Google Scholar]
- 8.Miebach L, Wolfsgruber S, Polcher A, et al. Which features of subjective cognitive decline are related to amyloid pathology? Findings from the DELCODE study. Alzheimers Res Ther. 2019;11:66. doi: 10.1186/s13195-019-0515-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Jessen F, Wiese B, Bachmann C, et al. Prediction of dementia by subjective memory impairment: effects of severity and temporal association with cognitive impairment. Arch Gen Psychiatry. 2010;67:414–22. doi: 10.1001/archgenpsychiatry.2010.30. [DOI] [PubMed] [Google Scholar]
- 10.Cheng ZH, Chao DL, Li D, et al. Development and Reliability and Validity of Dementia Screening Inventory for Elderly. J Clin Psychol. 2008;16:66. doi: 10.3969/jissn.1005-3611.2008.02.001. [DOI] [Google Scholar]
- 11.Yu J, Li J, Huang X. The Beijing version of the Montreal Cognitive Assessment as a brief screening tool for mild cognitive impairment: a community-based study. BMC Psychiatry. 2012;12:156. doi: 10.1186/1471-244X-12-156. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Thomann PA, Toro P, Dos Santos V, et al. Clock drawing performance and brain morphology in mild cognitive impairment and Alzheimer’s disease. Brain Cogn . 2008;67:88–93. doi: 10.1016/j.bandc.2007.11.008. [DOI] [PubMed] [Google Scholar]
- 13.Mitchell AJ. A meta-analysis of the accuracy of the mini-mental state examination in the detection of dementia and mild cognitive impairment. J Psychiatr Res . 2009;43:411–31. doi: 10.1016/j.jpsychires.2008.04.014. [DOI] [PubMed] [Google Scholar]
- 14.Wu Y, Xu WW, Cheng ZH, et al. Brief elderly cognitive screening inventory: development, reliability and validity. Chin J Geront. 2016;36:1211–3. doi: 10.3969/jissn.1005-9202.2016.05.086. [DOI] [Google Scholar]
- 15.Zhou XQ, Lu J, Mo J, et al. Association of Subjective Memory Complaints with Objective Cognitive Impairments in the. ElderlyChin Mental Health J. 2019;27:520–3. doi: 10.16128/j.cnki.1005-3611.2019.03.019. [DOI] [Google Scholar]
- 16.Cheng ZH, Sun JR, Yang BX, et al. Reliability and Validity of Chinese Cognitive Ability Scale. Chin Mental Health J. 2007;21:103–6. doi: 10.3321/j.issn.1000-6729.2007.02.01317. [DOI] [Google Scholar]
- 17.Gong YX. The Chinese revision of Halstead-Reitan neuropsychological test battery for adults. Acta Psychologica Sinica. 1986;18:433–42. [Google Scholar]
- 18.Wu Y, Jiang Y, Tang L, et al. Performance analysis of Quick Cognitive Screening Scale in screening early Alzheimer disease from community population. Chin Mental Health J. 2020;34:106–11. [Google Scholar]
- 19.Morris JC. The Clinical Dementia Rating (CDR): Current version and scoring rules. Neurology (ECronicon) 1993;43:2412–4. doi: 10.1212/WNL.43.11.2412-a. [DOI] [PubMed] [Google Scholar]
- 20.Hachinski VC, Lassen NA, dementia MJM. A cause of mental deterioration in the elderly. Lancet. 1974;2:207–10. doi: 10.1016/s0140-6736(74)91496-2. [DOI] [PubMed] [Google Scholar]
- 21.Zhang F, Yang W. Interaction between activities of daily living and cognitive function on risk of depression. Front Public Health. 2024;12:1309401. doi: 10.3389/fpubh.2024.1309401. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Cummings JL. The Neuropsychiatric Inventory: assessing psychopathology in dementia patients. Neurology (ECronicon) 1997;48:S10–6. doi: 10.1212/wnl.48.5_suppl_6.10s. [DOI] [PubMed] [Google Scholar]
- 23.Zhang MY. Handbook of psychiatric rating scale. Changsha: Hunan Science and Technology Press; 2016. pp. 143–316. [Google Scholar]
- 24.Jessen F, Amariglio RE, van Boxtel M, et al. A conceptual framework for research on subjective cognitive decline in preclinical Alzheimer’s disease. Alzheimers Dement. 2014;10:844–52. doi: 10.1016/j.jalz.2014.01.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Petersen RC. Mild cognitive impairment as a diagnostic entity. J Intern Med. 2004;256:183–94. doi: 10.1111/j.1365-2796.2004.01388.x. [DOI] [PubMed] [Google Scholar]
- 26.Jessen F, Amariglio RE, Buckley RF, et al. The characterisation of subjective cognitive decline. Lancet Neurol. 2020;19:271–8. doi: 10.1016/S1474-4422(19)30368-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.McKhann GM, Knopman DS, Chertkow H, et al. The diagnosis of dementia due to Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement. 2011;7:263–9. doi: 10.1016/j.jalz.2011.03.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Sun Y, Yang FC, Lin CP, et al. Biochemical and neuroimaging studies in subjective cognitive decline: progress and perspectives. CNS Neurosci Ther. 2015;21:768–75. doi: 10.1111/cns.12395. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Saykin AJ, Wishart HA, Rabin LA, et al. Older adults with cognitive complaints show brain atrophy similar to that of amnestic MCI. Neurology (ECronicon) 2006;67:834–42. doi: 10.1212/01.wnl.0000234032.77541.a2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Rabin LA, Smart CM, Crane PK, et al. Subjective Cognitive Decline in Older Adults: An Overview of Self-Report Measures Used Across 19 International Research Studies. J Alzheimers Dis. 2015;48 Suppl 1:S63–86. doi: 10.3233/JAD-150154. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Wolfsgruber S, Kleineidam L, Guski J, et al. Minor neuropsychological deficits in patients with subjective cognitive decline. Neurology (ECronicon) 2020;95:e1134–43. doi: 10.1212/WNL.0000000000010142. [DOI] [PubMed] [Google Scholar]
- 32.Fastame MC. Are subjective cognitive complaints associated with executive functions and mental health of older adults? Cogn Process. 2022;23:503–12. doi: 10.1007/s10339-022-01089-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Munoz N, Goma IFM, Valero S, et al. Personality Factors and Subjective Cognitive Decline: The FACEHBI Cohort. Behav Neurol. 2020:5232184. doi: 10.1155/2020/523218435. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Jekel K, Damian M, Wattmo C, et al. Mild cognitive impairment and deficits in instrumental activities of daily living: a systematic review. Alz Res Therapy. 2015;7:17. doi: 10.1186/s13195-015-0099-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Xue J, Li J, Liang J, et al. The Prevalence of Mild Cognitive Impairment in China: A Systematic Review. Aging Dis. 2018;9:706–15. doi: 10.14336/AD.2017.0928. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Hao L, Wang X, Zhang L, et al. Prevalence, Risk Factors, and Complaints Screening Tool Exploration of Subjective Cognitive Decline in a Large Cohort of the Chinese Population. J Alzheimers Dis. 2017;60:371–88. doi: 10.3233/JAD-170347. [DOI] [PubMed] [Google Scholar]
- 37.Petersen RC, Lopez O, Armstrong MJ, et al. Practice guideline update summary: Mild cognitive impairment [RETIRED]: Report of the Guideline Development, Dissemination, and Implementation Subcommittee of the American Academy of Neurology. Neurology (ECronicon) 2018;90:126–35. doi: 10.1212/WNL.0000000000004826. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Song M, Lee SH, Kim SY, et al. Measurement of Subjective Cognitive Decline (SCD) Using Korean-Everyday Cognition (K-ECog) as a Screening Tool: a Feasibility Study. Dement Neurocogn Disord . 2021;20:80–8. doi: 10.12779/dnd.2021.20.4.80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Steinberg SI, Negash S, Sammel MD, et al. Subjective memory complaints, cognitive performance, and psychological factors in healthy older adults. Am J Alzheimers Dis Other Demen. 2013;28:776–83. doi: 10.1177/1533317513504817. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Crawford J, Smith G, Maylor E, et al. The Prospective and Retrospective Memory Questionnaire (PRMQ): Normative data and latent structure in a large non-clinical sample. Memory. 2003;11:261–75. doi: 10.1080/09658210244000027. [DOI] [PubMed] [Google Scholar]
- 41.Xie HG, Wang LN, Wang W, et al. Prevalence survey of the mental and behavioral disturbance in elderly people with normal cognitive function in community. CTTER. 2004;8:5223–35. doi: 10.3321/j.issn:1673-8225.2004.25.028. [DOI] [Google Scholar]
- 42.Pan FF, Huang L, Chen KL, et al. A comparative study on the validations of three cognitive screening tests in identifying subtle cognitive decline. BMC Neurol. 2020;20:78. doi: 10.1186/s12883-020-01657-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Hao L, Jia J, Xing Y, et al. An application study-subjective cognitive decline Questionnaire9 in detecting mild cognitive impairment (MCI) Aging Ment Health. 2022;26:2014–21. doi: 10.1080/13607863.2021.1980860. [DOI] [PubMed] [Google Scholar]
- 44.Caffò AO, Spano G, Tinella L, et al. The Prevalence of Amnestic and Non-Amnestic Mild Cognitive Impairment and Its Association with Different Lifestyle Factors in a South Italian Elderly Population. Int J Environ Res Public Health. 2022;19:3097. doi: 10.3390/ijerph19053097. [DOI] [PMC free article] [PubMed] [Google Scholar]