Acta Neurol Scand 2014: 131: 219–224 DOI: 10.1111/ane.12310
© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd ACTA NEUROLOGICA SCANDINAVICA
Education-corrected CERAD identifies MCI and dementia in Parkinson’s disease Karrasch M, Laatu S, Ellfolk U, Marttila R, Martikainen K. Education-corrected CERAD identifies MCI and dementia in Parkinson’s disease. Acta Neurol Scand 2015: 131: 219–224. © 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd. Objectives – This study examined whether controlling for educational background in the CERAD cognitive screening battery would affect the likelihood of patients with Parkinson’s disease to fulfill criteria for mild cognitive impairment (PD-MCI) and dementia (PDD). Materials & Methods – One-hundred seventeen patients with PD were studied. Cognitive impairment was determined as two subtest scores falling below either the standard cutoff scores or education-corrected cutoff scores. The presence of dementia was determined by clinical interview or Clinical Dementia Rating. Patients were then classified as PD-MCI and PDD according to cognitive test performance and presence/ absence of dementia. Results – The number of cognitively impaired patients (PD-MCI or PDD) was significantly higher when educationcontrolled cutoff scores were used (62.5% vs 38%). Correspondingly, the number of false negatives (demented PD patients performing normally in CERAD) was significantly lower when educationcorrected cutoff scores were used (4% vs 10%). Conclusions – Controlling for education increases the sensitivity of the CERAD for PD-MCI and PDD.
Introduction
Cognitive impairment is one of the most frequent non-motor symptoms of Parkinson’s disease (PD) (1). Although cognitive symptoms are typically relatively mild in early disease stages, epidemiological studies have shown that between 49 and 75% of patients eventually progress to Parkinson’s disease dementia (PDD) (2, 3). Diagnostic criteria for PDD were published in 2007 (3). The diagnosis of PDD is assigned when a patient has impairment on more than one cognitive domain representing a decline from premorbid levels, and the impairment is severe enough to impair everyday life. Cognitive impairment was recommended to be assessed either with short cognitive screening tools (Level I testing) or with more extensive neuropsychological testing including measures of several cognitive domains (Level II testing). The term PD-MCI (Parkinson’s disease mild cognitive impairment) has been coined for the transition stage between normal or near-normal cognitive function and dementia in PD (4, 5).
M. Karrasch1, S. Laatu2, U. Ellfolk1, R. Marttila3, K. Martikainen4 1 Department of Psychology and Logopedics, Abo Akademi University, Turku, Finland; 2Outpatient Ward of Neuropsychiatry, Turku University Hospital, Turku, Finland; 3Division of Clinical Neurosciences, Turku University Hospital, Turku, Finland; 4The Finnish Parkinson Association, Turku, Finland
Key words: cognitive screening; CERAD; education; Parkinson’s disease; PDD; PD-MCI M. Karrasch, Department of Psychology and Logopedics, Abo Akademi University, FIN-20500 Turku, Finland Tel.: +358-2-2154405 Fax: +358-2-2154534 e-mail: [email protected] Accepted for publication August 27, 2014
Diagnostic criteria for PD-MCI, published in 2012 (5), also included recommendations for both Level I and II cognitive testing. Several different screening tests for mild cognitive impairment in PD have been tested and developed (for a review, see 6). Some tests have been adopted from research on Alzheimer’s disease (AD) (MoCA, MDRS), while others have been specifically developed for PD (SCOPA-Cog, PD-CRS, PANDA). In Finland, the CERAD is the most widely used cognitive screening test in clinical geriatric settings, although it was originally developed to identify cognitive changes in AD (7, 8). The CERAD includes tests measuring both posteriorcortical, such as visuospatial function and language function, as well as medial temporal, such as verbal learning and episodic memory consolidation, which are sensitive to cortical dementia of the AD-type. It also includes tests of frontostriatal, executive functions, including verbal fluency and clock drawing. In fact, we recently showed significant differences in several CERAD subtests between patients with PD who were functioning 219
Karrasch et al. normally in everyday life and those who did have cognitive problems affecting everyday functioning (9). There were significant effects in executive subtests, but the largest effects were observed in visuospatial function, learning and delayed recall, which supports the notions that as cognitive functioning deteriorates in PD, the domains affected include the frontostriatally mediated executive functions and free recall, as well as more posterior-cortical functions (e.g., visuospatial functions) (1, 10). Educational background did, however, significantly affect most of the CERAD subtests in the patients with PD. A significant effect of education on CERAD scores has also been shown in normal elderly people (11). The likelihood of a patient fulfilling criteria for cognitive impairment or dementia is most probably affected by how cutoff scores are determined. In most cases, a general cutoff score is used, irrespective of educational background. The aim of this study was to examine whether controlling for education on the CERAD would affect the likelihood that patients with PD would meet the criteria for PD-MCI (5) and PDD (3) in a sample of 117 patients with PD. Material & methods
Data from two studies were pooled to ensure the representativeness of the sample for PD in all disease stages. Of 117 patients with PD included in the analyses, 89 were recruited consecutively from a multidisciplinary inpatient rehabilitation course in a movement disorders rehabilitation center in Finland, and 28 were enrolled through a national survey in collaboration with the Finnish Parkinson Association and from two outpatient neurology clinics. The previously established diagnoses of idiopathic PD were confirmed by clinical neurologic examination on the basis of UK Brain Bank criteria (12). The Parkinson medication was expressed as mean levodopa-equivalent daily doses (LEDD) calculated as suggested by Tomlinson et al. (13). Cognitive testing, neurologic examination, UPDRS rating and Hoehn and Yahr rating were conducted while patients were receiving their regular PD medication. The CERAD was conducted by a trained nurse or a clinical neuropsychologist. The presence of dementia (14) was determined using the Clinical Dementia Rating (CDR, 15) or a thorough clinical interview and examination. For a subset of patients whose next of kin could be interviewed (n = 72), the Alzheimer’s Disease Cooperative Study–Activities of Daily Living Inventory (ADCS-ADL, 16) was conducted to measure ADL function, and the Neuro220
psychiatric Inventory (NPI, 17) to measure neuropsychiatric symptoms. All the 117 patients gave their written informed consent for participation in the studies. The studies were conducted in accordance with the Declaration of Helsinki and were approved by the Ethical Committee of the South-Western Finland Health Care District. The mean age of the patients was 64.5 years (SD 9.4). The average years of education was 11.6 (SD 3.5). Mean UPDRS-III score was 23.6 (SD 10.7, range 4–59), mean Hoehn & Yahr stage was 2.4 (SD 0.8, range 1–5), and mean disease duration was 6.6 years (SD 5.2, range 0.2–28). We classified the patients into those fulfilling the Level 1 criteria in two ways, both for PDMCI (5) and PDD (3). The first classification was based on established cutoff scores for the Finnish CERAD. There are cutoff scores for nine of the subtests, which are based on large normative Finnish data (7, 18, 19). The cutoff scores have been established to achieve optimal sensitivity/specificity trade-off for differentiating between normal aging and early/mild AD. The tests and cutoffs are as follows: (1) semantic fluency (generating animal names during 1 min,
© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd ACTA NEUROLOGICA SCANDINAVICA
Education-corrected CERAD identifies MCI and dementia in Parkinson’s disease Karrasch M, Laatu S, Ellfolk U, Marttila R, Martikainen K. Education-corrected CERAD identifies MCI and dementia in Parkinson’s disease. Acta Neurol Scand 2015: 131: 219–224. © 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd. Objectives – This study examined whether controlling for educational background in the CERAD cognitive screening battery would affect the likelihood of patients with Parkinson’s disease to fulfill criteria for mild cognitive impairment (PD-MCI) and dementia (PDD). Materials & Methods – One-hundred seventeen patients with PD were studied. Cognitive impairment was determined as two subtest scores falling below either the standard cutoff scores or education-corrected cutoff scores. The presence of dementia was determined by clinical interview or Clinical Dementia Rating. Patients were then classified as PD-MCI and PDD according to cognitive test performance and presence/ absence of dementia. Results – The number of cognitively impaired patients (PD-MCI or PDD) was significantly higher when educationcontrolled cutoff scores were used (62.5% vs 38%). Correspondingly, the number of false negatives (demented PD patients performing normally in CERAD) was significantly lower when educationcorrected cutoff scores were used (4% vs 10%). Conclusions – Controlling for education increases the sensitivity of the CERAD for PD-MCI and PDD.
Introduction
Cognitive impairment is one of the most frequent non-motor symptoms of Parkinson’s disease (PD) (1). Although cognitive symptoms are typically relatively mild in early disease stages, epidemiological studies have shown that between 49 and 75% of patients eventually progress to Parkinson’s disease dementia (PDD) (2, 3). Diagnostic criteria for PDD were published in 2007 (3). The diagnosis of PDD is assigned when a patient has impairment on more than one cognitive domain representing a decline from premorbid levels, and the impairment is severe enough to impair everyday life. Cognitive impairment was recommended to be assessed either with short cognitive screening tools (Level I testing) or with more extensive neuropsychological testing including measures of several cognitive domains (Level II testing). The term PD-MCI (Parkinson’s disease mild cognitive impairment) has been coined for the transition stage between normal or near-normal cognitive function and dementia in PD (4, 5).
M. Karrasch1, S. Laatu2, U. Ellfolk1, R. Marttila3, K. Martikainen4 1 Department of Psychology and Logopedics, Abo Akademi University, Turku, Finland; 2Outpatient Ward of Neuropsychiatry, Turku University Hospital, Turku, Finland; 3Division of Clinical Neurosciences, Turku University Hospital, Turku, Finland; 4The Finnish Parkinson Association, Turku, Finland
Key words: cognitive screening; CERAD; education; Parkinson’s disease; PDD; PD-MCI M. Karrasch, Department of Psychology and Logopedics, Abo Akademi University, FIN-20500 Turku, Finland Tel.: +358-2-2154405 Fax: +358-2-2154534 e-mail: [email protected] Accepted for publication August 27, 2014
Diagnostic criteria for PD-MCI, published in 2012 (5), also included recommendations for both Level I and II cognitive testing. Several different screening tests for mild cognitive impairment in PD have been tested and developed (for a review, see 6). Some tests have been adopted from research on Alzheimer’s disease (AD) (MoCA, MDRS), while others have been specifically developed for PD (SCOPA-Cog, PD-CRS, PANDA). In Finland, the CERAD is the most widely used cognitive screening test in clinical geriatric settings, although it was originally developed to identify cognitive changes in AD (7, 8). The CERAD includes tests measuring both posteriorcortical, such as visuospatial function and language function, as well as medial temporal, such as verbal learning and episodic memory consolidation, which are sensitive to cortical dementia of the AD-type. It also includes tests of frontostriatal, executive functions, including verbal fluency and clock drawing. In fact, we recently showed significant differences in several CERAD subtests between patients with PD who were functioning 219
Karrasch et al. normally in everyday life and those who did have cognitive problems affecting everyday functioning (9). There were significant effects in executive subtests, but the largest effects were observed in visuospatial function, learning and delayed recall, which supports the notions that as cognitive functioning deteriorates in PD, the domains affected include the frontostriatally mediated executive functions and free recall, as well as more posterior-cortical functions (e.g., visuospatial functions) (1, 10). Educational background did, however, significantly affect most of the CERAD subtests in the patients with PD. A significant effect of education on CERAD scores has also been shown in normal elderly people (11). The likelihood of a patient fulfilling criteria for cognitive impairment or dementia is most probably affected by how cutoff scores are determined. In most cases, a general cutoff score is used, irrespective of educational background. The aim of this study was to examine whether controlling for education on the CERAD would affect the likelihood that patients with PD would meet the criteria for PD-MCI (5) and PDD (3) in a sample of 117 patients with PD. Material & methods
Data from two studies were pooled to ensure the representativeness of the sample for PD in all disease stages. Of 117 patients with PD included in the analyses, 89 were recruited consecutively from a multidisciplinary inpatient rehabilitation course in a movement disorders rehabilitation center in Finland, and 28 were enrolled through a national survey in collaboration with the Finnish Parkinson Association and from two outpatient neurology clinics. The previously established diagnoses of idiopathic PD were confirmed by clinical neurologic examination on the basis of UK Brain Bank criteria (12). The Parkinson medication was expressed as mean levodopa-equivalent daily doses (LEDD) calculated as suggested by Tomlinson et al. (13). Cognitive testing, neurologic examination, UPDRS rating and Hoehn and Yahr rating were conducted while patients were receiving their regular PD medication. The CERAD was conducted by a trained nurse or a clinical neuropsychologist. The presence of dementia (14) was determined using the Clinical Dementia Rating (CDR, 15) or a thorough clinical interview and examination. For a subset of patients whose next of kin could be interviewed (n = 72), the Alzheimer’s Disease Cooperative Study–Activities of Daily Living Inventory (ADCS-ADL, 16) was conducted to measure ADL function, and the Neuro220
psychiatric Inventory (NPI, 17) to measure neuropsychiatric symptoms. All the 117 patients gave their written informed consent for participation in the studies. The studies were conducted in accordance with the Declaration of Helsinki and were approved by the Ethical Committee of the South-Western Finland Health Care District. The mean age of the patients was 64.5 years (SD 9.4). The average years of education was 11.6 (SD 3.5). Mean UPDRS-III score was 23.6 (SD 10.7, range 4–59), mean Hoehn & Yahr stage was 2.4 (SD 0.8, range 1–5), and mean disease duration was 6.6 years (SD 5.2, range 0.2–28). We classified the patients into those fulfilling the Level 1 criteria in two ways, both for PDMCI (5) and PDD (3). The first classification was based on established cutoff scores for the Finnish CERAD. There are cutoff scores for nine of the subtests, which are based on large normative Finnish data (7, 18, 19). The cutoff scores have been established to achieve optimal sensitivity/specificity trade-off for differentiating between normal aging and early/mild AD. The tests and cutoffs are as follows: (1) semantic fluency (generating animal names during 1 min,
