Dementia related to Parkinson’s disease (PD) is associated with increased mortality,Reference Levy, Tang and Louis 1 reduced quality of life,Reference Playfer 2 and increased caregiver distressReference Aarsland, Larsen, Karlsen, Lim and Tandberg 3 and is a major risk factor for nursing home placement.Reference Aarsland, Larsen, Tandberg and Laake 4 A patient with PD is five to six times more likely to have dementia than an age-matched control without PD.Reference Hobson and Meara 5 Increasing age is the most important risk factor for PD dementia (PDD), and it is more likely to occur in the postural instability and gait disturbance motor phenotype than in the tremor dominant phenotype of PD.Reference Burn, Rowan, Allan, Molloy, O'Brien and McKeith 6 Among non-motor symptoms (NMS), psychosis including visual hallucinations has been associated with increased risk of PDD.Reference Apaydin, Ahlskog, Parisi, Boeve and Dickson 7 Cognitive dysfunction is related to neurocirculatory abnormalities, especially orthostatic hypotension and supine hypertension.Reference Kim, Oh, Lee, Kim, Yang and Goldstein 8 However, it is uncertain whether various NMS are risk factors for dementia in PD.
Recently, we suggested that deficits of stereopsis are common in drug-naïve PD patients.Reference Kim, Park, Kim and Koh 9 Our previous voxel-based morphometry study suggested that abnormal stereopsis implicated cortical visual dysfunction as part of the NMS in PD.Reference Koh, Suh, Kim and Kim 10 The involvement of extra-striate visual cortices is also associated with visual hallucinations, one of the high risk factors for dementia in PD.Reference Meppelink, de Jong, Renken, Leenders, Cornelissen and van Laar 11 These findings suggest that abnormal stereopsis, which is related to the visual association area, may be a predicting factor for dementia in PD.
In this present study, we studied the decline of Mini-Mental Status Examination (MMSE) scores over two years according to the presence of NMS including stereopsis. To define the possible associations between NMS and the risk of developing dementia, we investigated the risk factors for dementia conversion within a two-year observation period in PD.
Methods
Standard protocol approvals, registrations, and patient consent
The study was approved by the Korea University Guro Hospital Institutional Review Board. All patients provided written informed consent.
Subjects
We recruited consecutively drug-naïve PD patients, meeting the United Kingdom Parkinson’s Disease Society’s criteria for idiopathic PD, from the Parkinson’s Disease Center at Korea University Guro Hospital from November 2008 to August 2010.Reference Hughes, Daniel, Kilford and Lees 12 To ascertain dopaminergic drug treatment response, we reviewed medical records of all included patients for two years. Exclusion criteria were as follows: (1) history of head trauma, stroke, exposure to anti-dopaminergic drugs, or central nervous system (CNS) infection; (2) abnormal thyroid function testing; (3) structural lesions or hydrocephalus on brain magnetic resonance imaging (MRI); (4) abnormal cognitive function (MMSE ≤24) and (4) red-flag signs suggesting Parkinson plus syndrome. We classified PD patients into either a dementia converting group or a dementia non-converting group.
Clinical assessment
Patients underwent a neurologic examination and a clinical assessment. Symptom history, level of education, comorbid diseases and medication history were assessed. The Hoehn and Yahr (H&Y) stage was determined,Reference Hoehn and Yahr 13 and the degree of disease severity was quantified by the Unified Parkinson’s Disease Rating Scale (UPDRS). 14 The Non-motor Symptoms Scale for Parkinson’s disease (NMSS) was used to evaluate the presence and severity of NMS. Depression was rated with the Korean version of the Montgomery-Asberg Depression Scale (K-MADS)Reference Ahn, Lee and Yi 15 or the Geriatric Depression Scale-15.Reference Yesavage, Brink and Rose 16 We assessed olfactory function using structural screening questionnaires including odor detection, discriminatuion and identification. Rapid eyeball movement (REM) sleep behavior disorders (RBD) was screened with RBD screening questionnaires.Reference Stiasny-Kolster, Mayer, Schafer, Moller, Heinzel-Gutenbrunner and Oertel 17 The two-step test for stereopsis was performed. The first step involved tests for visual acuity and strabismus. Patients with strabismus, nystagmus, ocular motility disturbance, or poor visual acuity in either eye (<20/40 Snellen fraction) were excluded from the study. The second step involved the measurement of stereopsis. Stereopsis was assessed using Titmus stereotest plates (Stereo Optic Co., Inc., IL, USA). Normal stereopsis was defined as an arc <60 seconds in the Titmus fly test. Previous studies of adult stereoacuity reported that a reduction of the Titmus fly test score was observed in five of 60 normal subjects.Reference Garnham and Sloper 18 All tests were conducted at a distance of 40 cm, under an illumination of 200 lux. The motor phenotype was determined according to the UPDRS motor score, using the method prescribed by Jankovic et al.Reference Jankovic, McDermott and Carter 19
The general cognitive status of each subject was evaluated by means of the MMSE and the Clinical Dementia Rating (CDR) scale at baseline and two years later.Reference Kang, Na and Hahn 20 For analytic purposes, the total daily L-dopa equivalent dose (total LEDD, mg/day) after two years was calculated based on theoretical equivalence from the literature.Reference Herzog, Volkmann and Krack 21 Dementia was diagnosed according to the Movement Disorders Society criteria of PD dementia, defined as impairment of ≥ two cognitive domains using neuropsychological tests associated with significant functional impairment from cognitive decline.Reference Dubois, Burn and Goetz 22 Detailed cognition was evaluated using the Seoul Neuropsychological Screening Battery,Reference Ahn, Chin and Park 23 which consists of an attention test (forward digit span, backward digit span and letter cancellation), a language and related function test (spontaneous speech, comprehension, repetition, naming measured by the Korean version of the Boston Naming test, reading, writing, finger naming, right-left orientation, body part identification, calculation and praxis), a visuospatial function test (drawing an interlocking pentagon and the Rey complex figure test), a verbal memory test (three word registration, three word recall, and the Seoul Verbal Learning Test which included immediate recall, delayed recall, and recognition), a non-verbal memory test (immediate recall, delayed recall and recognition of a Rey complex figure) and a frontal executive function test (motor impersistence, contrasting program, go-no-go test, fist-edge-palm, alternating hand movement, alternating square and triangle test, Luria loop, Controlled Oral Word Association Test: animal, supermarket and letter, Korean-color Word Stroop Test: word reading and color reading).
Statistical analysis
The data were expressed as mean±standard deviation (SD) or as n (%). Simple bivariate analyses were performed to identify potential predictors of cognitive decline. Non-categorical clinical variables including age were dichotomized at the median. Between-group comparisons were made using the Student’s t-test or one-way analysis of variance (ANOVA). In the second stage of analysis, variables significantly associated with cognitive decline in the bivariate analyses (P value of ≤0.05) were entered into a logistic regression analysis, using dementia conversion as the dependent variable. A backward stepwise method was used to remove non-significant predictors. All analyses were performed using Statistical Package for the Social Sciences (SPSS) 12.0 for Windows.
RESULTS
Baseline and clinical characteristics
A total of 80 patients were enrolled in the study. The mean (SD) interval of follow-up assessments from initial assessment was 2.2 (0.4) years. The demographic and clinical characteristics are shown in Table 1. Dementia-converting group had significant low MMSE scores, older age and higher UPDRS motor scores at baseline.
Demographic and baseline clinical data in dementia-converter and non-converter
Data=mean±SD (median), *Significant difference
MMSE decline over two years
The mean MMSE score (mean±SD) at initial evaluation was 27.9±3.6, and the mean decline of the MMSE score was 2.7±2.6 in 2 years. The mean decline of the MMSE score over 2 years in the dementia converting group (6.6±1.7) was significantly higher than that of non-converting group (1.4±1.2) (p<0.001). Among the non-motor symptoms, vivid dreaming (p=0.021), RBD (p=0.003), constipation (p=0.043), hyposmia (p=0.002), abnormal stereopsis (p=0.001) and depression (p=0.002) were associated with a more rapid rate of cognitive decline. Old age (≥62 years, median age, p<0.001), female gender (p=0.048), non-smoking (p=0.007), postural instability/ gait disturbance (PIGD) phenotype (p<0.001) and severe motor dysfunction (UPDRS III score >12, p<0.001) were also associated with a more rapid rate of cognitive decline. Table 2 summarizes the bivariate analyses of clinical variables versus rate of cognitive decline over two years.
Bivariate comparisons of demographic and clinical variables versus rate of cognitive decline over two years (change in MMSE), using the Student’s t-test or ANOVA
RBD=REM sleep behavior disorders; LED=levodopa equivalent dose; HY=Hoehn and Yahr; PIGD=postural instability and gait disturbance; UPDRS=Unified Parkinson’s disease rating scale; NMSS=Non-motor symptoms scale; CI=confidence interval; B=regression coefficient; S.E=standard error; Wald.=Wald statistic; df=degree of freedom; Sig.=significance; Exp(B)=expected regression coefficient
*Significant difference
Risk factors of dementia conversion
Logistic regression analysis demonstrated that vivid dreaming, RBD, hyposmia, depression and abnormal stereopsis were significant independent predictors of dementia risk within two years. The total score of NMSS was not an independent predictor of dementia risk. Old age (age ≥62 years), PIGD phenotype and motor disability (UPDRS III score >12) were also significant independent predictors of dementia risk over two years. Table 3 summarizes the logistic regression results of the factors predicting dementia converting.
Logistic regression results to predict dementia conversion
PIGD=postural instability and gait disturbance; RBD=REM sleep behavior disorders; UPDRS=Unified Parkinson’s disease rating scale; NMSS=Non-motor symptoms scale; CI=confidence interval; B=regression coefficient; S.E=standard error; Wald.=Wald statistic; df=degree of freedom; Sig.=significance; Exp(B)=expected regression.
*Significant difference
DISCUSSION
Dementia associated with PD is a common complication, affecting 80% of PD patients during the course of their illness.Reference Hely, Reid, Adena, Halliday and Morris 24 In our present study, 22.5% (18/80) of enrolled patients met the criteria of PD dementia at 24 months. A PD cohort study in Cambridge reported that the overall proportion of patients with dementia at 3.5 years is 10%.Reference Williams-Gray, Foltynie, Brayne, Robbins and Barker 25 A systematic review of prevalence studies of dementia in PD determined the prevalence to be 31.5%.Reference Aarsland, Zaccai and Brayne 26 Differences in study design, age and disease stage of subjects may account for these variations. The conversion rate to dementia of our study is relatively high. Our study was not a community based study but an hospital based study. Difference in study design may be considered as an explanation of high conversion rate.
Vivid dreaming, RBD, hyposmia, abnormal stereopsis and depression were significant NMS PD dementia predictors at 24 months in this study. These NMS are also associated with a more rapid rate of cognitive decline. Several cross-sectional and prospective studies show that RBD is associated with increased risk of dementia.Reference Vendette, Gagnon and Decary 27 , Reference Postuma, Bertrand and Montplaisir 28 A recent three-year longitudinal study demonstrated that severe olfactory dysfunction is a prodromal symptom of PD dementia.Reference Baba, Kikuchi and Hirayama 29 Although the diagnoses of RBD and hyposmia were not based on polysomnogram or the objective smell test, our questionnaire-based study also showed that RBD and hyposmia were the most significant risk factors for PD dementia. Recently, we reported that deficits of stereopsis are common in drug-naïve PD patients;Reference Kim, Park, Kim and Koh 9 we also suggested that the neural correlate of stereopsis may lie in the non-dominant extra-striate visual cortex with voxel-based morphometric study and abnormal stereopsis implicated in cortical visual dysfunction as part of the NMS in PD.Reference Koh, Suh, Kim and Kim 10 Visuospatial and perceptual deficits are more frequently observed in PD dementia patients experiencing pathological changes in the visual association area.Reference Mosimann, Mather, Wesnes, O'Brien, Burn and McKeith 30 These findings suggest that abnormal stereopsis, which is related to the visual association area, may be a predicting factor for dementia in PD. In this study, abnormal stereopsis was associated with increased risk of dementia. Although visual hallucinations are a well-known risk factor of PD dementia, we did not find a significant association between visual hallucinations and dementia conversion in our present study. Failure to detect a significant association is likely due to the subject characteristics in our study. The subjects of our study were drug-naïve patients, and the numbers of patients with visual hallucinations was small.
The most important risk factor for PD dementia among the demographic factors assessed is increasing age. In our present study, old age (≥62 years), PIGD phenotype and PD motor dysfunction, reflected by UPDRS III scores, are also PD dementia predictors.
This study of 80 cognitively normal de novo PD patients, observed prospectively for two years, resulted in an average annual decline of 1.3 points of the MMSE score. The average annual decline of the MMSE score in the dementia converting group was 3.28 points, which is significantly higher than that of the non-converting group (0.72 points). These results are similar to previous studies. Presence of depression, vivid dreaming, RBD, hyposmia, abnormal stereopsis, non-smoking and the PIGD phenotype were associated with a significantly more rapid decline of MMSE.
It must be acknowledged that there were several limitations to the current study. One limitation includes the dependence on subjective scale of NMS. The presence of olfactory dysfunction and RBD are dependent on subjective or caregiver recall. The second limitation is a short follow-up duration. A 2-year follow-up period is not adequate to assess advanced PD patients. Performing an extended follow-up study is therefore necessary.
With increased awareness of the non-motor risk factors for the development of dementia in PD, clinicians may be better able to actively monitor cognitive dysfunction in their PD patients who present with the particular risk factors detailed in this study.
