Commentscomments powered by Disqus
Fluctuations in cognition and alertness vary independently in dementia with Lewy bodies
Donald L. Bliwise PhD*, Michael K. Scullin PhD, Lynn Marie Trotti MD, MSc
Article first published online: 22 OCT 2013
Department of Neurology, Emory University School of Medicine, Atlanta, Georgia, USA
*Correspondence to: Dr. Donald L. Bliwise, Department of Neurology, Emory University School of Medicine, 1841 Clifton Road, Room 509, Atlanta, GA 30329, USA; firstname.lastname@example.org
Funding agencies: National Institutes of Health grant numbers R01 NS-050595 (to D.L.B.), F32 AG-041543 (to M.K.S)., KL2 RR-025009 (to L.M.T.), UL1 RR-025008 (Atlanta Clinical and Translational Science Institute).
Relevant conflicts of interest/financial disclosures: Nothing to report.
Fluctuations in mental status are 1 of the core diagnostic criteria for dementia with Lewy bodies (DLB) and are thought to reflect variability in daytime alertness. Previous attempts to study fluctuations have been limited to caregiver reports, observer rating scales, short segments of electroencephalography, or motor-dependent, reaction time tests. Concordance among such measures is often poor, and fluctuations remain difficult to quantify. We compared fluctuations in cognition and alertness in patients with DLB (n=13) and idiopathic Parkinson's disease (PD) (n=13), a condition associated with deficits in daytime alertness. We systematically and repeatedly collected cognitive and physiologic measures during a 48-hour inpatient protocol in a sound-attenuated sleep laboratory in a geriatric hospital. Cognitive fluctuations were analyzed using coefficients of variation (COVs) derived from performance on a bedside examination familiar to clinicians (digit span). Alertness fluctuations were assessed objectively using COVs from the polysomnographically-based Maintenance of Wakefulness Test. Despite predictably lower mean digit span performances, DLB patients demonstrated significantly greater cognitive fluctuations than PD patients (P < 0.001), even when groups were matched on general cognitive impairment. There were no group differences in alertness fluctuations, although DLB patients were less alert than PD patients not receiving dopaminergics. The prevailing assumption that fluctuations in cognition in DLB are reflected in fluctuations in daytime alertness was not supported by objective, physiological measurements. Fluctuating mental status in DLB patients can be detected with repeated administration of a simple bedside exam that can be adapted to a clinic setting. © 2013 International Parkinson and Movement Disorder Society.
Fluctuations are 1 of the core criteria for dementia with Lewy bodies (DLB) and are conceptualized as involving pronounced variation in alertness and cognition.[1, 2] Studies that have examined fluctuations in cognition in such patients typically have operationalized fluctuations by either caregiver/clinician–based questionnaires and/or ratings, [3-7] or more objectively by motor-dependent, reaction time tests. Such tests are often derived from proprietary pharmaceutical testing programs,[8-12] and they can be challenging to administer across a broad range of synucleinopathic patients, because of well-known motor deficits in both simple and choice reaction time characterizing Parkinsonian patients. Considerations such as these make it difficult to integrate objective measurements of fluctuating cognition into routine clinical practice.
In this study, we examined within-patient fluctuations in cognition in DLB patients over a period of 48 hours using a simple, verbally administered, bedside protocol (digit span) not affected by slow manual dexterity. Because the construct of fluctuating mental status implicitly and explicitly has invoked the notion that varying levels of alertness parallel those fluctuations in cognition,[5, 7, 11, 12, 14] we used the Maintenance of Wakefulness Test (MWT) as an objective, physiological measure of alertness. As a comparison group, we also examined such fluctuations in another patient group known to incur deficits in daytime alertness, idiopathic Parkinson's disease (PD).
Patients and Methods
DLB patients (n = 13) met consensually agreed upon criteria from the Third International Consensus conference for the disorder. A detailed presentation of central, core, suggestive, and supportive features for the patients is shown in Table 1. PD patients (n = 64) demonstrated at least 2 of the cardinal signs for PD diagnosis, including bradykinesia, rigidity, resting tremor, and postural instability.[15, 16] None of the PD patients were demented. Twelve had Mini-Mental State Exam (MMSE) scores of less than 28 (n = 6 with less than 27) (range, 23–27), and 4 of these had some mild memory complaints, but a formal diagnosis of mild cognitive impairment was not rendered in these patients. MMSE scores of less than 26 are occasionally encountered in PD patients without dementia. We examined separately the subgroup of PD patients with low MMSE scores in our case-control analysis.
Table 1. Clinical characteristics of DLB patients (n = 13)
|79 M||+||VH, P||RBD−, NS NA||G, D, S||Nonfocal atrophy||Preceding|
|73 M||+||FC, VH, P||RBD+, NS NA||U||Nonfocal atrophy||Concurrent|
|58 M||+||VH, P||RBD+, NS NA||D||Nonfocal atrophy||Preceding|
|69 M||+||VH, P||RBD−, NS+||G, D, DL||Nonfocal atrophy||Preceding|
|72 M||+||FC, VH, P||RBD−, NS NA||G, D, S||Nonfocal atrophy||Preceding|
|74 M||+||FC, VH, P||RBD−, NS NA||G, S, DL||Nonfocal atrophy, WMH||Concurrent|
|69 M||+||VH, P||RBD+, NS NA||D, U, S||WNL||Concurrent|
|79 M||+||P||RBD+, NS+||D||Nonfocal atrophy||Concurrent|
|71 M||+||VH, P||RBD+, NS−||D, U||Nonfocal atrophy||Preceding|
|71 M||+||VH, P||RBD−, NS NA||EEG||Nonfocal atrophy||Preceding|
|73 F||+||P||RBD−, NS−||D, U, G||Nonfocal atrophy||Concurrent|
|69 M||+||VH, P||RBD+, NS−||D, G, S||NA||Concurrent|
|59 M||+||FC, P||RBD+, NS NA||EEG||WMH||Preceding|
Values follow the criteria given in Ref. . Low dopamine transporter uptake was not included in the table as it was not used routinely in our center.
Demographic, clinical, and medication information for both the DLB and PD patients are presented in Table 2. There was no difference in benzodiazepine use. Relative to DLB patients, a larger proportion of PD patients used dopamine agonists, although there were no differences in the proportion using levodopa. l-Dopa dosage was higher in the PD patients. Usage of antipsychotics and cholinesterase inhibitors was higher in DLB patients. Fourteen PD patients used neither l-dopa nor dopamine agonists. A board-certified neurologist rated each patient's parkinsonism with the Unified Parkinson's Disease Rating Scale (UPDRS) motor subscale. Dose equivalence for those patients taking dopaminergic medications followed standard conversions. Approximately half (62%) of the DLB patients, but only 2 of the PD patients, received cholinesterase inhibitors. Only 1 patient (PD) used stimulant medication, and 14 PD patients were not taking any dopaminergics. All patients completed the Epworth Sleepiness Scale (ESS); for several of the DLB patients, caregivers assisted in the completion of this scale.
Table 2. Demographics and clinical data for Parkinson's disease (n = 64) and dementia with Lewy bodies (n = 13) patients
|Parkinson's disease||Dementia with Lewy bodies||P|
|Age, y||62.95 (9.70)||70.46 (6.24)||0.009|
|Gender, % male||66%||92%||0.06|
|ESS||10.25 (4.44)||10.88 (5.21)||0.65|
|MMSE||28.55 (1.73)||22.85 (4.62)||0.001|
|UPDRS||17.56 (8.91)||26.40 (11.38)||0.008|
|Disease duration, y||5.57 (3.95)||4.08 (3.54)||0.22|
|Levodopa dosage equivalent||460.22 (379.62)||233.35 (276.63)||0.045|
|Medication usage, n (% using mg)|
|Levodopa||42 (65.6)||7 (53.8)||0.42|
|Dopamine agonists||35 (54.7)||1 (7.7)||0.002|
|Benzodiazepines, GABAergic site-specific agonists||9 (14.1)||2 (15.4)||0.99|
|Cholinesterase inhibitors||2 (3.1)||8 (61.5)||0.0001|
|Antipsychotics||0 (0%)||6 (46.2)||0.0001|
|Antidepressants||22 (34.4)||8 (61.5)||0.067|
|MAO-B inhibitors||18 (28.1)||1 (7.7)||0.17|
Patients were recruited from the Movement Disorders Clinic at Emory University School of Medicine in Atlanta, Georgia, USA. Advertisements were placed in waiting rooms and in physicians' offices. Any patient interested in participating was contacted by the first author (D.B.), who provided details and arranged scheduling. Exclusion criteria included serious medical comorbidities (eg, history of stroke, myocardial infarction, active cancer, uncontrolled diabetes), history of central nervous system (CNS) trauma or infection, history of neurosurgical intervention (pallidotomy, subthalamic deep brain stimulation), or current or lifetime Axis I psychiatric diagnosis.
Data collection occurred between March 2008 and May 2012 in a sound-attenuated laboratory at Wesley Woods Geriatric Hospital in Atlanta, Georgia, USA. All patients provided written informed consent at the beginning of the study. The protocol conformed to the ethical standards of the Declaration of Helsinki for protection of human subjects and was approved by the Institutional Review Board at Emory University.
Bedside Verbally Administered Cognitive Testing
We employed digit span forward and backward as brief assessments of cognitive fluctuations. These were performed immediately prior to each MWT trial (Fig. 1) (maximum = 8 trials per patient). The digit span administration protocol was similar to that used in standard intelligence testing, except that the order of digits varied across trials. In the forward condition, the technician read a string of digits aloud (eg, 2–9-6-1) and the patient was asked to repeat these aloud in the order presented. If the patient was able to do this correctly, an additional digit was added to the sequence. Each number of digit strings was presented twice and the particular trial was discontinued when the patient made errors on 2 consecutive presentations. Following digit span forward testing, the patient was instructed to repeat a string of digits backward (eg, in the above case, 1–6-9-2). The backward recall always began with a 2-digit practice sequence to ensure that patients understood the procedure. All patients were able to do at least 2 digits backward when presented initially. The score for digit span was the highest number of digits successfully recalled prior to first error, calculated separately for forward, for backward, and for the sum of forward and backward. These scores were calculated separately for each of the 8 trials.
Physiologic Alertness Testing
Patients were given the MWT during which they were asked to recline in bed (head elevated) in a darkened room and were instructed to try to remain awake. Each trial of the test lasted 40 minutes and the trials were evenly spaced throughout the daytime hours for a total of 8 trials over the 48-hour period. For each MWT trial, we generated 2 measurements: the time (in minutes) to the first 30-second epoch of sleep (sleep latency) and the proportion of the 40-minute nap opportunity in which the patient was asleep (sleep efficiency). A full description of the procedure has been described. Using a similar procedure, a mean sleep latency of 32.6 minutes has been reported in normal subjects.
Clinical Fluctuations Scale
A subset of patient caregivers/spouses of the patients (n = 58) completed this previously published scale. Initially, the scale contained 19 items but the original validation suggested that 4 items that related to alertness were most sensitive to differentiating DLB. For the current analyses we examined these originally described 4 items, but selected 6 additional items (items 1, 2, and 13–16) that specifically asked caregivers about fluctuations in cognition. We analyzed the original 4 items (“alertness” subscale) and the 6 additional items (“cognitive” subscale) separately.
Inpatient (Laboratory) Procedures
During the 48-hour protocol (Fig. 1), patients maintained their typical schedule for meals, medications, and nocturnal sleep. They were not kept in time isolation. One patient was studied at a time, usually beginning on Thursday night and ending late Saturday afternoon. To ensure the absence of inadvertent sleep that would compromise the validity of our physiological measurements, patients underwent constant behavioral monitoring, both directly and with infrared video monitoring from an adjacent room, by 1 of 2 research technologists who monitored only that patient around the clock. Video monitoring continued during overnight polysomnography. Sleep was prohibited during the daytime, except during MWT trials. Patients were allowed anywhere in the 1400-square-foot laboratory during the 48-hour period but otherwise remained sequestered. They were not allowed to consume caffeinated beverages. Between testing procedures, patients watched television, browsed the Internet, used the telephone, and read magazines. The protocol included 2 consecutive nights of overnight polysomnography; some of these results have been reported[22, 24]; but, in brief, nocturnal polysomnography measures were unrelated to the measures of cognitive fluctuations. Some patients (PD, n = 9; DLB, n = 2) completed fewer than 8 testing trials (eg, leaving for social engagements or were only able to schedule a single day), but exclusion of these patients did not change the major results.
Comparisons between DLB and PD patients were made using chi-square tests for categorical variables (or Fisher's exact test, when expected cell frequencies were less than 5) and t tests for continuous measures. We used Levene's test correction for unequal variances when necessary, and as an additional check, we performed nonparametric Mann-Whitney U tests. For the digit span and MWT data, we characterized each patient by their mean level of performance averaged across all trials, but the main outcome measure was the assessment of intraindividual variability in performance. We accomplished the latter by calculating for each patient a coefficient of variation (COV), which divided each patient's standard deviation (SD) by their mean. This ratio expresses variance after accounting for differences in a patient's mean score. The COV allowed comparison of DLB and PD patients on measures for which their average level of performance may differ (eg, digit span). Additionally, we performed several case-control subanalyses by matching PD and DLB cases on general cognitive ability (MMSE) or age, and repeated the COV analysis. We also compared the 14 PD patients not receiving dopaminergics to the 13 DLB patients on COV measures. Spearman's rho was used to examine correlations between objective measures of fluctuations (digit span, MWT) and caregiver-assessments of fluctuations, as well as dopaminergic dose equivalents. In cases with missing data, analyses are reported based on existing data. Statistical significance was determined using an alpha level of 0.05.
Analyses of Mean Levels
Mean level comparisons between DLB and PD groups are included in Table 3. As would be expected, DLB patients performed more poorly relative to PD patients on both digits forward and backward, and, by definition, on the sum of forward + backward. There were no differences in cognition or alertness between DLB patients receiving or not receiving cholinesterase inhibitors. Mean levels of alertness derived from the MWT did not differentiate the 2 groups (Table 3); however, DLB patients fell asleep more rapidly than the 14 PD patients not taking dopaminergics (t = 2.13, P = 0.04). Mean levels of caregiver reported fluctuations on the alertness subscale did not significantly distinguish the groups, but caregiver reported fluctuations on the cognition subscale were significantly higher in the DLB patients relative to the PD patients (Table 3).
Table 3. Mean values and COVs for patients with Parkinson's disease and dementia with Lewy bodies
|Parkinson's disease||Dementia with Lewy bodies||P|
|Digits forward, mean (SD)||6.24 (0.85)||4.90 (0.97)||<0.001|
|Digits backward, mean (SD)||4.09 (1.11)||2.63 (0.99)||<0.001|
|Digits forward + backward, mean (SD)||10.29 (1.79)||7.53 (1.70)||<0.001|
|COV digits forward, mean (SD)||0.13 (0.05)||0.27 (0.20)||0.02|
|COV digits backward, mean (SD)||0.32 (0.19)||0.64 (0.43)||0.02|
|COV digits forward + backward, mean (SD)||0.15 (0.09)||0.25 (0.08)||<0.001|
|Maintenance of wakefulness test|
|Sleep latency (min), mean (SD)||18.39 (12.68)||18.42 (15.17)||0.99|
|Sleep efficiency (%), mean (SD)||28.11 (23.41)||27.29 (28.70)||0.92|
|COV sleep latency, mean (SD)||0.65 (0.37)||0.75 (0.62)||0.59|
|COV sleep efficiency, mean (SD)||0.97 (0.71)||0.83 (0.82)||0.53|
|Caregiver fluctuations scale|
|Alertness subscale, mean (SD)||0.74 (0.95)||1.33 (0.89)||0.06|
|Cognitive subscale, mean (SD)||1.61 (1.90)||4.25 (1.06)||<0.001|
Analyses of Objectively Measured Fluctuations
The primary goal of this study was to compare objectively measured cognitive fluctuations between groups. COVs of digit span forward, backward, and the sum of digits forward + backward were significantly higher (t tests using Levene's adjustment, when appropriate) in DLB patients relative to the PD patients (Table 3). The digit span COV group differences were significant with nonparametric Mann-Whitney U tests as well (forward: P = 0.003, backward: P = 0.017, combined: P < 0.001). A group difference also was seen when digit span performance was expressed more simply as the maximal intraindividual difference between the highest and lowest number of digits (ie, the range) recalled correctly by a given patient during the 48-hour protocol. So, for example, for digits backward, 10 of 13 DLB patients showed a best minus worse difference of 3 digits, whereas for the PD patients only 26 of 64 showed a difference of that magnitude (χ2 = 5.72, P = 0.02). This suggests that a relatively basic, easily computed metric for performance fluctuations may be of value.
Because the PD and DLB groups differed in age, we matched 13 PD patients to the 13 DLB patients on the basis of age and sex. Results were similar. Comparisons between the 14 PD patients not receiving any dopaminergics and the 13 DLB patients also showed parallel findings. Specific comparisons between those using versus not using all other medication classes in Table 2 indicated no significant effect for digit span COV measures to be higher or lower for those using medication within either patient group. Dopaminergic dose equivalents were unrelated to all digit span COV measures for PD and DLB patients taking this drug class. Taken together, these results indicate that medication usage did not account for the differences in cognitive fluctuations.
Although intraindividual differences in cognition existed within this sample, intraindividual variability in physiologically measured alertness (MWT) was similar between patient groups. Using correlational analyses, no associations existed between MWT-derived COV and digit-span derived COV within PD, DLB, or both groups combined (all P values >0.17), further indicating that fluctuations in alertness cannot be equated with fluctuations in cognition, at least within these patient groups. Within PD patients only, those not receiving dopaminergics showed higher MWT-defined COV in sleep efficiency than those receiving dopaminergics, suggesting that use of this medication class may “smooth” daytime alertness to some degree (mean [SD] MWT COV for patients [n = 50] using dopaminergics = 0.81 [0.60] vs mean [SD] MWT COV for patients [n = 14] not using dopaminergics = 1.56 [0.80]; t = 3.84, P < 0.001).
Because some of the PD patients showed mild cognitive impairments, we examined group differences in COVs comparing the 13 DLB patients to the 6 PD patients who had MMSE scores ≤26 (M = 24.3). Despite also being statistically similar in age, UPDRS, and gender (all P values > 0.10), the DLB patients still demonstrated significantly greater COVs for digit span combined (P = 0.002), digits forward (P = 0.04), and digits backward (P = 0.004). The DLB and PD subgroup did not differ on MWT COVs (sleep latency, P = 0.60, or sleep efficiency, P = 0.99). Results were similar when comparing the 13 DLB patients to the 12 PD patients who had MMSE scores ≤ 27.
Associations Between Objectively Measured Cognitive Fluctuations and Caregiver-Reported Fluctuations
Scores from the 4-item alertness subscale on the caregiver questionnaire were unrelated to all COV measures of digit span performance; however, scores on the 6-item cognitive subscale were related to the COV of the sum of digits forward + backward (rho = 0.33, P = 0.01). These data suggested only partial validation of caregiver's report of cognitive fluctuations with objective measurements made in the inpatient setting.
Our data demonstrated that fluctuations in cognition were indeed a distinguishing feature of DLB, but that those fluctuations are independent from alertness, at least when the latter is measured with the MWT. The simple digit span procedure, when administered repeatedly, proved quite sensitive to fluctuations. Although representing a core feature of the condition, operationalization of fluctuations in mental status in DLB patients has been inconsistent across research studies[25, 26] and probably remains even more so in clinical practice. Ratings made by others, including caregivers, or by clinicians,[7, 11] presumably provide data integrating the more long-term experiences" > of those closest to patients, but the caregiver scale used here was only modestly related to COVs of objective measures of cognitive fluctuations. Previous reaction time studies[9, 11, 12, 27] analyzing COVs on a second-to-second basis often have been used to define fluctuations; however, their duration of administration (typically 90 seconds) provides only the narrowest of temporal windows on a phenomenon that is probably far less circumscribed, especially when compared to caregivers' cumulative experiences of the identified patient's fluctuating cognition over much longer intervals of time. By way of example, time of day in such studies is often left unspecified, though 1 study noted that all testing occurred between 10:00 and 11:00 am, thus emphasizing the possibility that some variability across the remainder of the day might have been missed. Reaction time studies involving DLB patients have also typically,[11, 12] but not always,<0" > excluded patients with frank motor impairments who could not perform such tests, thus limiting their relevance for testing trial-to-trial cognitive fluctuations in patients with bradykinesia, rigidity, or tremor. For all these reasons, we believe that repeated administration of a simple, verbally administered bedside exam (ie, digit span), could hold considerable practical relevance. Although digit span performance may be less influenced by slowed manual reaction time, it may nonetheless reflect a more generalized bradyphrenia, reflecting impairment and slower processing of recent verbally presented material, a phenomenon recognized in Parkinsonian patients
Fluctuations in alertness form a cornerstone of current conceptualizations of variability in cognition in DLB.[1, 4, 14] Our data suggest that the 2 phenomena may be different. Using Likert scale ratings, Escandon et al. also suggested that cognitive fluctuations in dementia were independent from rated levels of alertness, but repetitive measures of cognition or alertness were not performed in that study. Prior attempts to measure fluctuations physiologically with spontaneous electroencephalographic (EEG) activity examined variability of delta band spectral power over a limited 1-hour window at unspecified times of day.[7, 11, 12, 29] These data indicated dampened levels of alertness in DLB patients at a single interval of time but did not speak to variability in those levels of alertness across the entirety of daytime hours as a correlate of fluctuating mental status. We did not perform spectral analyses of the EEG across the daytime hours in our study, which leaves open the possibility that some intrasubject variability might have been missed.
Because the differential diagnosis of DLB using clinical criteria continues to present challenges[2, 25, 30, 31] even for astute clinicians, our findings of a simple-to-administer measure of cognitive fluctuations may bear relevance upon important diagnostic issues of determination of intrapatient variability. The greatest relevance of the variability data may be the potential for the differentiation of DLB from other types of dementia, such as Alzheimer's disease (AD). Our protocol did not include AD patients, who might have been expected to show reduced variability in cognition relative to the DLB patients. Future work would be required to confirm the utility of this approach in other forms of dementia.
Our study has several weaknesses. In addition to the aforementioned lack of testing in AD patients, perhaps the group for whom presumed lack of variability in cognition might be the most important factor for differential diagnosis, our PD and DLB groups differed in several respects. The PD patients were considerably younger than the DLB patients and this might have influenced our findings, but the subanalyses of age-matched and sex-matched PD patients showed similar results both for COVs based on digit span and for MWT. Although usage of l-dopa was no more common in PD relative to DLB patients, the l-dopa dosage equivalent was higher in the PD patients. Interestingly, use of dopaminergics did not impact measures of fluctuations in cognition, but they may have served to “smooth” out fluctuations in alertness seen in untreated PD, perhaps, as we have shown, by making patients sleepier. DLB patients did not show this effect, but our small sample size (reflecting the intensive nature of the protocol) and relatively low power for within DLB analyses may have worked against detecting such medication effects. Similarly, although cholinesterase inhibitors are associated with improvement in reaction time and attention tasks in DLB, we could not demonstrate any corresponding effects on fluctuations here.
These data were collected under the unique conditions of an intensive, 48-hour inpatient protocol with careful and continuous behavioral monitoring. Mega et al. proposed that a 5-point fluctuation in total MMSE score over an interval of 6 months would constitute evidence of substantial fluctuation. Our data suggest that substantial variability was present over 2 consecutive days. Certainly some repetitive sampling of cognition could be performed in clinic setting on a single day, though individual clinics would need to determine how many repetitive administrations over how short a time interval would be (1) tolerable to patients and their families; (2) amenable to adequate staffing; and (3) yield data on intraindividual performance of comparable validity to what we have proposed here. It also is possible that technology transfer (eg, smartphone applications for repeated verbal administration of digit span) might even be used by caregivers at home, many of whom are deeply involved in acquiring an accurate diagnosis and appropriate treatment for the patients under their care.
M.K.S. was partially supported by an Emory University Cottrell Fellowship. We are appreciative of Anthony Wilson and Sophia Greer at Emory University who provided assistance in the collection and management of the data. D.L.B., M.K.S., and L.M.T. had full access to all the data and they take responsibility for the integrity of the data.
Research Project: A. Conception Bliwise; B. Organization Bliwise, Trotti; C. Execution Bliwise, Scullin, Trotti; Statistical Analysis: A. Design Bliwise, Scullin; B. Execution Scullin, Bliwise; C. Review and Critique Bliwise, Scullin, Trotti; Manuscript Preparation: A. Writing of the first draft Bliwise, Scullin;
B. Review and Critique Bliwise, Scullin, Trotti.
Dr. Bliwise is a consultant for Ferring Pharmaceuticals, Morehouse School of Medicine and the New England Research Institute; Dr. Scullin has no disclosures; and Dr. Trotti was previously a consultant for UCB Pharma.
- McKeith IG, Dickson DW, Lowe J, et al. Diagnosis and management of dementia with Lewy bodies: third report of the DLB Consortium. Neurology 2005;65:1863–1872.
- McKeith I, Mintzer J, Aarsland D, et al. Dementia with Lewy bodies. Lancet Neurol 2004;3:19–28.
- Bradshaw J, Saling M, Hopwood M, Anderson V, Brodtmann A. Fluctuating cognition in dementia with Lewy bodies and Alzheimer's disease is qualitatively distinct. J Neurol Neurosurg Psychiatry 2004;75:382–387.
- Escandon A, Al-Hammadi N, Galvin JE. Effect of cognitive fluctuation on neuropsychological performance in aging and dementia. Neurology 2010;74:210–217.
- Ferman TJ, Smith GE, Boeve BF, et al. DLB fluctuations: specific features that reliably differentiate DLB from AD and normal aging. Neurology 2004;62:181–187.
- Varanese S, Perfetti B, Monaco D, et al. Fluctuating cognition and different cognitive and behavioural profiles in Parkinson's disease with dementia: comparison of dementia with Lewy bodies and Alzheimer's disease. J Neurol 2010;257:1004–1011.
- Walker MP, Ballard CG, Ayre GA, et al. The clinician assessment of fluctuation and the one day fluctuation assessment scale. Br J Psychiatry 2000;177:252–256.
- Ballard C, O'Brien J, Gray A, et al. Attention and fluctuating attention in patients with dementia with Lewy bodies and Alzheimer's disease. Arch Neurol 2001;58:977–982.
- Ballard C, Walker M, O'Brien J, Rowan E, McKeith I. The characterization and impact of ‘fluctuating’ cognition in dementia with Lewy bodies and Alzheimer's disease. Int J Geriatr Psychiatry 2001;16:494–498.
- Ballard CG, Aarsland D, McKeith I, et al. Fluctuations in attention: PD dementia vs DLB with parkinsonism. Neurology 2002;59:1714–1720.
- Walker MP, Ayre GA, Cummings JL, et al. Quantifying fluctuating in dementia with Lewy bodies, Alzheimer's disease, and vascular dementia. Neurology 2000;54:1616–1625.
- Walker MP, Ayre GA, Perry EK, et al. Quantification and characterization of fluctuating cognition in dementia with Lewy bodies and Alzheimer's disease. Dement Geriatr Cogn Disord 2000;11:327–335.
- Pullman SL, Saunders-Pullman R. Assessing disability in movement disorders: quantitative techniques and rating scales. In: Watts RL, Standaert DG, Obeso JA, eds. Movement Disorders. 3rd ed. New York: McGraw-Hill Medical; 2012:17–32.
- Serrano C, Garcia-Borreguero D. Fluctuations in cognition and alertness in Parkinson's disease and dementia. Neurology 2004;63:S31–S34.
- Langston JW, Widner H, Goetz CG, et al. Core assessment program for intracerebral transplantations (CAPIT). Mov Disord 1992;7:2–13.
- Litvan I, Bhatia KP, Burn DJ, et al. SIC Task Force appraisal of clinical diagnostic criteria for parkinsonian disorders. Mov Disord 2003;18:467–486.
- Folstein MF, Folstein SE, McHugh PR. A practical method for grading the cognitive state of patients for the clinician. J Psychiatry Res 1975;12;189–198.
- Aarsland D, Bronnick K, Williams-Gray C, et al. Mild cognitive impairment in Parkinson disease. Neurology 2010;75:1062–1069.
- Hobson DE, Lang AE, Martin WRW, Razmy A, Rivest J, Fleming J. Excessive daytime sleepiness and sudden onset sleep in Parkinson's disease: a survey by the Canadian Movement Disorders Group. JAMA 2002;287:455–463.
- Johns MW. A new method for measuring daytime sleepiness: the Epworth Sleepiness Scale. Sleep 1991;14:540–545.
- Mitler MM, Gujavarty KS, Browman CP. Maintenance of wakefulness test: a polysomnographic technique for evaluation of treatment efficacy in patients with excessive somnolence. Electroencephalogr Clin Neurophysiol 1982;53:658–661.
- Bliwise DL, Trotti LM, Wilson AG, et al. Daytime alertness in Parkinson's disease: potentially dose-dependent, divergent effects by drug class. Mov Disord 2012;27:1118–1124.
- Doghramji K, Mitler MM, Sangal RB, et al. A normative study of the maintenance of wakefulness test (MWT). Electroencephalogr Clin Neurophysiol 1997;103:554–562.
- Scullin MK, Trotti, LM, Wilson AG, Greer SA, Bliwise DL. Nocturnal sleep enhances working memory training in Parkinson's disease but not Lewy body dementia. Brain 2012;135:2789–2797.
- Lee DR, Taylor J-P, Thomas AJ. Assessment of cognitive fluctuation in dementia: a systematic review of the literature. Int J Geriatr Psychiatry 2012;27:989–998.
- Imamura T, Hirono N, Hashimoto M, et al. Clinical diagnosis of dementia with Lewy bodies in a Japanese dementia registry. Dement Geriatr Cogn Disord 1999;10:210–216.
- Bradshaw JM, Saling M, Anderson V, Hopwood M, Brodtmann A. Higher cortical deficits influence attentional processing in dementia with Lewy bodies, relative to patients with dementia of the Alzheimer's type and controls. J Neurol Neurosurg Psychiatry 2006;77:1129–1135.
- Sagar HJ, Sullivan EV, Gabrieli JDE, Corkin S, Growdon JH. Temporal ordering and short-term memory deficits in Parkinson's Disease. Brain 1988;111:525–539.
- Walker MP, Ayre GA, Ashton CH, et al. A psychophysiological investigation of fluctuating consciousness in neurodegenerative dementias. Hum Psychopharmacol Clin Exp 1999;14:483–489.
- Mega MS, Masterman DL, Benson DF, et al. Dementia with Lewy bodies: reliability and validity of clinical and pathologic criteria. Neurology 1996;47:1403–1409.
- Nelson PT, Jicha GA, Kryscio RJ, et al. Low sensitivity in clinical diagnoses of dementia with Lewy bodies. J Neurol 2010;257:359–366.
- Rowan E, McKeith IG, Saxby BK, et al. Effects of donepezil on central processing speed and attentional measures in Parkinson's disease with dementia and dementia with Lewy bodies. Dement Geriatr Cogn Disord 2007;23:161–167.