CP: care partner
IADL: instrumental activities of daily living
MCI: mild cognitive impairment
MoCA: Montreal Cognitive Assessment
PD: Parkinson disease
pwPD: people with Parkinson disease
SCC: subjective cognitive complaint
WNL: within normal limits
Parkinson disease (PD) is characterized by motor and non-motor symptoms, including a broad spectrum of physical, sensory, cognitive, behavioral, emotional, and autonomic symptoms (Lee and Koh, 2015). Cognitive symptoms in people with Parkinson disease (pwPD) are common and are among the most disabling non-motor features of the disease. Cognitive symptoms are also strongly associated with an increased risk of nursing home placement (Aarsland et al, 2000), increased economic burden (Vossius et al, 2011), and reduced quality of life for both the patient and his or her care partner (CP) (Lawson et al, 2017). Among the cognitive impacts of PD are decreases in executive functioning, attention, and visuospatial and memory skills (Janvin et al, 2005; Watson and Leverenz, 2010).
Although impaired performance on psychometric measures is well documented in pwPD, the subjective experience of cognitive decline in pwPD is not fully understood (Aarsland et al, 2017). This lack of knowledge has important clinical and research implications.
The promulgation of proposed criteria for mild cognitive impairment (MCI) due to PD (Litvan et al, 2012) has brought attention to subjective cognitive complaints (SCCs) in pwPD. Although documentation of perceived cognitive decline (from the perspective of the patient, CP, or clinician) is basic to these criteria, there are key limitations. Notably, the type and degree of cognitive complaints that should be queried and the way in which symptoms are described are not specified.
Prior work suggests that pwPD commonly experience SCCs of reduced executive function, attention, and processing speed, as well as an increase in verbal fluency problems (Koster et al, 2015). Early reports of visuospatial and verbal memory dysfunction are associated with a higher risk of developing objective cognitive impairment (Mills et al, 2016). However, the relationship between SCCs and objective impairment appears variable (Copeland et al, 2016; Hong et al, 2018; Lehrner et al, 2014; Mills et al, 2016; Sitek et al, 2011).
In addition to variability in which cognitive symptoms are assessed, another assessment issue relates to how best to capture SCCs in pwPD; for example, how valid for pwPD are SCC inventories that were developed for other conditions? Two prior studies of SCCs in pwPD (Lehrner et al, 2014; Sitek et al, 2011) used tools designed to identify symptoms in people with amnestic disorders such as Alzheimer disease that focused on memory functioning (Dujardin et al, 2010). Although memory complaints are present in pwPD, executive or visual problems are usually more salient in this population.
Furthermore, on commonly used questionnaires developed specifically for pwPD, such as the Parkinson’s Disease Questionnaire–39 (Peto et al, 1998), cognitive domains are minimally addressed by queries about memory and concentration only, and they are therefore likely to omit items or symptoms that are particularly important or unique to pwPD. A further concern is that broader questionnaires tend to use cognitive concepts that are jargon heavy, which may not bring forth ecologically valid responses from patients or CPs (Mills et al, 2016). By excluding the patient- and CP-lived experience of cognitive decline, those impacted by PD may not be able to clearly communicate the patient’s cognitive state.
Also missing from some of these prior investigations (Hong et al, 2018; Lehrner et al, 2014; Mills et al, 2016) is the perspective of the CP, which may differ from the patient’s (Schiehser et al, 2013). This deficit is important because the CP may indicate the specific impact of cognitive failures on the patient’s daily functioning and the priorities for potential treatment of these difficulties. When CPs’ perspectives were considered, the studies were limited by focusing specifically on a particular segment of the PD population, for example, those with MCI (Copeland et al, 2016) or only memory symptoms (Sitek et al, 2011).
Even if the SCCs of pwPD can be assessed successfully, there are concerns that merely naming complaints does not provide meaningful information about a patient’s actual objective cognitive status. The SCCs of pwPD have been criticized as lacking sensitivity and specificity to cognitive deficits (Copeland et al, 2016; Dujardin et al, 2010), leading to the suggestion that pwPD have minimal insight into their cognitive decline. However, such studies were based on dichotomous impressions of cognitive status (ie, presence or absence of the complaint) and may not reflect the more subtle, multidimensional nature of the degree of impairment (Castro et al, 2016; Copeland et al, 2016; Dujardin et al, 2010; Hong et al, 2018; Mills et al, 2016).
Despite numerous potential problems and difficulties in capturing the subjective experience of cognitive decline in pwPD, the goal of modern patient-centered outcomes research is to identify and understand the outcomes that are important to the patient as well as to his or her CP. Understanding the patient’s experience of cognitive decline is therefore central to successful treatment. To this end, it is important to note that a patient’s perception of cognitive impairment has a strong effect on his or her quality of life (Schrag et al, 2000; Vlagsma et al, 2016). Over and above a patient’s objective cognitive deficits, his or her perception of a cognitive deficit needs to be understood and possibly treated if the goal is to improve the patient’s quality of life.
Because symptom complaints tend to guide clinical assessment and treatment, understanding the impact of SCCs is critical for improving patient-centered outcomes. Furthermore, how patients and their CPs perceive symptoms may influence their impression of treatment success, or lack thereof. We are not aware of any studies that have explored priorities for cognitive treatments for pwPD.
To address these concerns, we pursued four goals:
- Report how patients and their CPs describe, in their own terms, the cognitive impacts of PD.
- Explore the convergent validity of SCCs with measured cognition, reported instrumental activities of daily living (IADL), and CP distress.
- Describe how role (patient versus CP) and degree of cognitive impairment relate to SCCs.
- Report patients’ and CPs’ treatment priorities for cognitive treatment outcomes.
Study procedures and methods were reviewed and approved by the local institutional review board at Baylor Scott and White Health, with participants providing written informed consent before beginning the study activities. Dyads consisting of a person with PD and his or her CP were recruited to participate in either the focus group (22 dyads) or the questionnaire phase (50 dyads) of the study. Volunteers were recruited from our local movement disorders clinic, community-based PD support groups, and the Michael J. Fox trial finder. Volunteers were paid $15 for their participation.
Inclusion criteria for the pwPD were being older than 18 years, having a diagnosis of idiopathic PD by a neurologist, being fluent in English, having an available CP to complete study activities, and having an adequate understanding of the study to participate meaningfully and provide consent (as determined by study staff). Patients were excluded if they did not meet all inclusion criteria or had been diagnosed with a comorbid neurodegenerative condition. Inclusion criteria for the CPs were being older than 18 years, knowing the person with PD well, and providing assistance to the person with PD.
Because we have previously described the 22 dyads in the focus group phase of our study (Benge et al, 2018), we do not repeat the information on participants here. Participants in the focus group phase provided information for construction of the questionnaire used in the second phase. In that second phase, which we report here, 50 pwPD/CP dyads independently completed the questionnaire. The pwPD had an average age of 71.8 years (SD=10.6), 80% were men, and 70% had at least some college education (2% did not finish high school, 28% high school degree, 14% some college, 12% associate’s degree, 28% bachelor’s, 16% graduate degree). The average Hoehn and Yahr rating scale (1967) was 2.1 (SD=0.87). The CPs had an average age of 65.4 years (SD=13.4), 86% were women, 80% were spouses, 69% reported daily contact with the person with PD, and 69% had at least some college education. The sample (pwPD and CPs) was mainly white/Caucasian by self-report (96%).
Focus Group Phase
To summarize what we previously reported, the focus group phase of our study centered on pwPD’s and CPs’ priorities for outcomes from cognitive assessment and cognitive treatment (Benge et al, 2018). During that initial phase, 22 dyads (22 pwPD and 22 CPs interviewed separately) were invited to generate examples of cognitive symptoms or difficulties that they observed daily, and for which they hoped to see improvement if the person with PD underwent cognitive treatment. Research-trained Parkinson’s Advocates in Research, in conjunction with a licensed neuropsychologist (J.F.B.), grouped the participants’ responses into domains in order to create an SCC scale reflecting the pwPD’s and CPs’ experiences of cognitive decline and their priorities for potential treatment. Table 1 lists the 25 individual items (unevenly distributed among the five domains) that comprise this scale and are the basis of the questionnaire used in the second phase of our study.
Fifty new pwPD/CP dyads completed the SCC questionnaire separate from one another. For each of the 25 items on the questionnaire, each person with PD rated his or her difficulty (or, in the case of the CPs, their patient’s difficulty) with each particular SCC on a scale of 0–10, with 0 indicating no difficulty and 10 representing the worst difficulty imaginable. Finally, participants identified what they considered to be the top five symptoms on the list that would be important targets for cognitive treatment.
Objective Measures: Cognitive Characterization and Other Assessments
To assess the general cognitive functioning of the pwPD, we administered the Montreal Cognitive Assessment (MoCA; Nasreddine et al, 2005). In addition, the pwPD’s motor and disability ratings were provided by study staff, using the Hoehn and Yahr rating scale. Assessments were completed by the neuropsychologist and a trained research assistant.
Cognitive categorization of the pwPD was based on their total MoCA scores to screen for objective cognitive impairment, as well as their Hoehn and Yahr rating scales indicating impairment in ADL. The MoCA is a multidomain cognitive screening instrument that has been validated in pwPD (Brown et al, 2016; Dalrymple-Alford et al, 2010; Kandiah et al, 2014). A score of 26 or greater indicates cognitive functioning within normal limits (WNL); less than 26 indicates cognitive impairment. We have previously shown (Benge et al, 2017) that in pwPD, the MoCA consists of three subscales: (a) memory (ie, memory items), (b) executive function (ie, verbal fluency, clock drawing, cube copy, backward digit span, and serial sevens subtraction), and (c) verbal attention (ie, a cross-loading of serial sevens subtraction, forward digit span, and sentence repetition). In the present study, we evaluated the participants’ MoCA scores in terms of the total score and the three component subscales.
The CPs of the pwPD also completed scales. Determination of the pwPD’s functional ability (IADL) was based on scores from the Penn Parkinson’s Daily Activities Questionnaire—15 (Brennan et al, 2016). The Penn Parkinson’s Daily Activities Questionnaire—15 is a newly developed instrument that is designed to evaluate IADL difficulties in pwPD. It is sensitive and specific to the presence of PD-related dementia (Brennan et al, 2016). A score of 44 or greater indicates normal IADL function; a score of 43 or lower indicates functional impairment. In the present study, the pwPD were divided into three groups: (a) no cognitive impairment (WNL cognitive screen and no IADL impairment), (b) MCI (abnormal cognitive screen and no IADL impairment), and (c) suspected dementia (abnormal cognitive screen and reported IADL impairment). The CPs also completed the Zarit Burden Interview (Zarit et al, 1980) to measure caregiver distress.
On the basis of their MoCA scores and Penn Parkinson’s Daily Activities Questionnaire—15 ratings, nine pwPD were classified as WNL cognition, 20 as MCI, and 21 as suspected dementia.
Most statistical analyses were conducted in JASP version 86 (JASP, 2018), with secondary analyses conducted in Stata 13.0 (StataCorp, 2013). In keeping with our stated aims, we took the individual SCCs from the questionnaire and used Pearson product-moment correlations to explore the convergent validity of the SCCs with measured cognition, reported ADL, and reported CP distress. We also used coefficient alpha to assess interitem reliability for the thematically similar items (ie, the items in each of the five domains of SCCs) and the SCC instrument as a whole. Next, we explored how role (patient versus CP) and degree of cognitive impairment (WNL cognition, MCI, and dementia) relate to SCCs. To determine the influence of cognitive state, reporter’s role, and their interaction on the SCCs, we used repeated-measures analysis of variance with reporter (paired self and CP reports) serving as the repeated measure and cognitive grouping (WNL, MCI, and dementia) serving as the between-groups variable. Because of the uneven sample sizes in the cognitive impairment groups, we used Levene’s test to evaluate for violations of the assumption of homogeneity of variances for all analyses. Where appropriate, Friedman’s test, the nonparametric equivalent of repeated-measures analysis of variance, was used with Dunn’s post hoc tests. Finally, priorities for cognitive treatment outcomes were identified by adding up the items that were most commonly identified as the top five priorities for cognitive treatment in each of the pwPD and CP groups.
Cognitive Symptoms Identified in the Focus Group Phase
In conjunction with the Parkinson’s Advocates in Research, we grouped the 25 cognitive symptoms identified by the focus groups into five broad themes, or domains: memory, language and communication, attention and processing speed, executive functioning, and fluctuations/episodic confusion (Table 1). Table 2 lists descriptive data for the SCCs by domain and reporter.
Interitem Reliability and Convergent Validity of SCC Items and Subscales
Among pwPD’s self-report of SCCs, all subscale scores for the five domains and the total score reached a high level of reliability, without the removal of any items: memory (α=0.91), language and communication (α=0.88), attention and processing speed (α=0.80), executive functioning (α=0.94), fluctuations and episodic confusion (α=0.83), and total score (α= 0.97).
Among CPs’ report of SCCs, all subscales for the five domains and the total score also reached an excellent level of reliability, without the removal of any items: memory (α=0.90), language and communication (α=0.91), attention and processing speed (α=0.90), executive functioning (α=0.96), fluctuations and episodic confusion (α=0.82), and total score (α= 0.98).
Table 3 shows the correlations between pwPD- and CP-reported cognitive difficulties and scores from the MoCA, CP-reported IADLs, and CP burden. Correlations between SCCs and these measures reported by pwPD were statistically significant, though weak in magnitude. Correlations between SCCs reported by CPs and objective cognitive impairment as assessed by the MoCA were moderately statistically significant.
When we looked at the MoCA subscales, there was little domain specificity, with significant correlations between MoCA memory, MoCA executive, and MoCA attention items with all CP-reported subjective cognition scales. Interestingly, the pwPD-reported SCCs correlated more with attention-screening test measures from the MoCA than did the CP-reported SCCs.
SCCs as reported by both the pwPD and their CPs strongly negatively correlated with IADL functioning, and self-reported SCCs bore more modest associations with IADL functioning. The CPs reported an increase in caregiver stress as their pwPD’s cognitive difficulties increased.
Impact of Reporter and Cognitive Status on SCC Reporting
Figure 1 illustrates how reporter status (person with PD [self] versus CP) and cognitive status of the patient (normal cognition [WNL], MCI, dementia) were related to SCC reporting in total and within each SCC domain.
For total SCC, there were significant effects for reporter (F1,47=6.204, P=0.016, η²=0.087), cognitive category (F2,47=32.64, P<0.001, η²=0.581), and the interaction (F2,47=18.457, P<0.001, η²=0.251). These relationships and interaction effects are modeled in Figure 1A. Dementia was associated with a higher number of reported SCCs regardless of who was reporting the symptoms (patient or CP), with no significant difference between reports for patients with cognition WNL and those with MCI (t=0.804, P>0.05). For patients assessed as either WNL or MCI, the CPs reported fewer cognitive issues than did the pwPD; however, for the dementia group, the CPs reported significantly more SCCs than did the pwPD.
For memory complaints, there were significant effects for cognitive category (F2,47=21.67, P<0.001, η²=0.480) and the interaction (F2,47=6.693, P=0.003, η²=0.219) (Figure 1B). Memory complaints were significantly higher in pwPD with dementia than in those with cognition WNL or MCI (P<0.001 for all contrasts), and there was no difference between memory complaints in pwPD with cognition WNL and those with MCI. The main effect for reporter was not significant. For patients with cognition WNL or MCI, self-reported memory complaints were greater than those by CPs; however, for those with dementia, self-reports were lower than those by CPs.
Analysis of language and communication symptom variances (Figure 1C) revealed a violation of the assumption of equality of variance (Levene’s test; CP report [F2,47=8.5, P<0.001]). However, using nonparametric equivalents, we noted a similar pattern to other cognitive domains. The degree of pwPD self-reported language SCCs increased with cognitive disability (χ2=9.173, P=0.012), and post hoc Dunn’s comparisons showed a stepwise increase in language complaints, with a statistically significant increase in language problems from either WNL or MCI cognition to dementia. The same pattern held with CP-reported language problems (χ2=35.34, P<0.0001), with the same pattern noted on post hoc tests. Although Wilcoxon signed rank tests did not reveal an overall difference between pwPD self- and CP-reported language problems for the sample as a whole (Z=0.904, P=0.3661), there were statistically significant differences in the individual groups. Both pwPD in the WNL and MCI cognitive groups self-reported more language problems than did their CPs, but the opposite pattern was found in the dementia group (all P<0.01).
For attention and processing speed symptoms (Figure 1D), a main effect was observed for cognitive categorization (F2,47=15.37, P<0.001, η²=0.395), and the interaction was significant (F2,47=5.71, P=0.006, η²=0.186). Significantly more attention and processing speed difficulties were reported for pwPD with dementia than those with cognition WNL or MCI (P<0.001 for all contrasts); there was no difference between attention and processing speed complaints in pwPD with cognition WNL and MCI cognition. The main effect for reporter was not significant. For pwPD with cognition WNL and MCI, self-reported attention and processing speed SCCs were greater than those reported by CPs; however, for those with dementia, self-report was lower than that by CPs.
For the executive functioning symptoms (Figure 1E), the assumption of equality of variance was violated for CPs (F2,47=4.775, P= 0.013). We again decomposed the analyses using nonparametric equivalents. The pwPD self-reported executive dysfunction grew with increasing cognitive disability (χ2=9.413, P=0.0246), with post hoc Dunn’s comparisons showing that pwPD with dementia identified significantly more executive dysfunction than did those with either WNL or MCI (P<0.01). The same pattern held with CP-reported executive problems (χ2=31.38, P<0.0001), with the same post hoc test pattern. Wilcoxon signed rank tests did not reveal an overall difference between pwPD self- and CP-reported executive problems for the sample as a whole (Z=−1.18, P=0.2370). However, there were statistically significant differences in the individual groups: patients with MCI reported more executive dysfunction than did their CPs, and patients with dementia reported less executive dysfunction than did their CPs (all P<0.05).
For fluctuations in cognitive ability and episodic confusion (Figure 1F), there were significant effects for cognitive category (F2,47=16.42, P<0.001, η²=0.411) and the interaction (F2,47=13.479, P=0.003, η²=0.136). Cognitive fluctuation complaints were significantly higher in pwPD with dementia than in those with cognition WNL or MCI (P<0.01 for all contrasts); there were no differences between pwPD with cognition WNL and MCI. The main effect for reporter was not significant. For patients with cognition WNL or MCI, the self-reported fluctuation complaints were greater than those reported by CPs; for patients with dementia, self-report was less than CP report.
Cognitive Treatment Priorities
The five items most frequently identified by pwPD as priorities for cognitive treatment were as follows:
- Improving the ability to recall the name of a person or place (endorsed by 46% of the sample as a top five priority; Language item 5 in Table 1)
- General word-finding difficulties (36%; Language item 4 in Table 1)
- Misplacing items (34%; Memory item 4 in Table 1)
- Recalling recent events (32%; Memory item 2 in Table 1)
- Difficulties following instructions (28%; Language item 1 in Table 1)
In contrast, for the CP group, the five most frequently identified priorities for cognitive treatment were as follows:
- Improvement in decision-making (50%; Executive Functioning item 6 in Table 1)
- Staying focused/paying attention (38%; Attention and Processing speed item 1 in Table 1)
- Improving a reduced desire/interest in starting tasks (36%; Executive Functioning item 8 in Table 1)
- Fluctuations in thinking abilities (30%; Fluctuations item 3 in Table 1)
- Misplacing items (30%; Memory item 4 in Table 1)
In our study, pwPD and their CPs identified numerous cognitive symptoms that mimicked the broad cognitive impacts of the disease (Muslimović et al, 2009). Many of these symptoms are similar to SCCs that have been identified in other neurodegenerative conditions (Farias et al, 2017). However, pwPD and their CPs also described cognitive symptoms seldom included among SCC measures developed for other conditions, such as fluctuations in cognitive abilities, prospective memory lapses, and difficulties recalling names distinct from more general word-finding difficulties. These SCCs are not commonly evaluated in other cognitive conditions.
A recent study estimated that non-motor fluctuations occur in ∼40% of pwPD within the first 5 years from disease onset, and among the most frequently reported symptoms were fluctuations in cognition, including episodes of dulled thinking and concentration difficulty (Kim et al, 2018). Nearly one third of the CPs in our study identified improving such fluctuations in thinking abilities as a treatment priority. This finding highlights the need to take into account the specific experiences of pwPD and CPs in order to identify cognitive treatment targets. This approach involves going beyond using the more general questions developed for other conditions.
Among the pwPD in this study, 46% identified improving name- or word-finding difficulties as a treatment priority. Focus groups and our Parkinson’s Advocates in Research team distinguished between perceived difficulties in retrieving names and more general word-finding difficulties. A prior investigation, however, suggested that pwPD and their CPs may not identify language problems accurately (Copeland et al, 2016). It is notable that the objective measures frequently used to identify language problems in pwPD (ie, verbal abstractions or confrontation noun-naming tasks) may not measure the same cognitive constructs necessary for personal semantic access.
Overall, our findings suggest that considering the unique aspects of cognitive decline in pwPD will be fruitful for understanding these symptoms and measuring them in ways that align with the lived experiences of pwPD.
Relationship Between SCCs and Objectively Measured Cognition
Although prior work suggested limitations in the use of SCCs for identifying pwPD with cognitive impairment (Copeland et al, 2016), the current study found that SCCs were correlated with objectively measured cognition. This relationship was stronger for CP-reported cognitive difficulties than for pwPD self-report. Although our finding may appear to suggest reduced awareness of cognitive deficits in pwPD, subsequent group-level analyses demonstrated that the SCC rating was higher in pwPD than in CPs when the pwPD had cognition WNL or MCI. The pwPD with dementia still reported greater SCCs, but the issues appeared more noticeable and were rated as more severe by the CPs. Our findings suggest that clinicians and researchers must consider both the reporter and the overall stage of a patient’s cognitive decline when evaluating the potential utility of SCCs. Furthermore, future studies will be needed to identify symptom type and degree of difficulty that optimally distinguishes cognitive groups (ie, those SCCs that best distinguish between cognition WNL and MCI).
There was little cognitive domain specificity between the subjective reports of the pwPD and CPs and objective cognition, as indexed by correlations between SCCs and the three subscales of the MoCA (Table 3). Interestingly, pwPD-reported cognitive difficulties had significant correlation with MoCA attention items, whereas CP-reported cognitive difficulties did not. This suggests that pwPD notice attention dysfunction more readily than do their CPs. In contrast, CPs notice memory and executive dysfunction more than do the pwPD. We observed the expected convergent validity with reported cognitive complaints and other measures of caregiver distress and IADL functioning. Thus, our findings suggest that SCCs from both pwPD and their CPs can provide valid information on current cognitive functioning in line with other measures, although more work is needed to associate domains of cognitive complaints with instruments used in clinical practice.
Cognitive Treatment Priorities
The cognitive treatment priorities differed between pwPD and their CPs: pwPD identified improving word- and name-finding as a priority (36% and 46%, respectively), whereas 50% of CPs identified improving decision-making as a priority. These priorities highlight the need for a thorough understanding of these cognitive symptoms in particular. For example, there are unique aspects of language decline in pwPD that are virtually unexplored, such as difficulty with verb processing (Cardona et al, 2013) and proper noun anomia. Likewise, decision-making in daily life reflects a broad array of cognitive and situational factors and may require development of sensitive and specific measures for these difficulties in day-to-day life in view of possible limitations to the ecological validity of measures such as the Iowa Gambling Test (Bechara et al, 1994). More work is needed targeting ecologically valid measures of symptoms and patient-centered outcomes that are important to pwPD and CP.
Summary, Limitations, and Future Directions
In this study, both pwPD and their CPs identified a broad array of cognitive impacts of the disease, including some PD-specific symptoms such as fluctuations in cognitive ability, word-finding difficulties, and decision-making problems. SCCs were endorsed more frequently by pwPD than CPs in the absence of dementia, whereas symptom report was higher in CPs than pwPD in the presence of dementia. Symptom complaints, particularly those reported by CPs, correlated well with objective measures of cognition. The pwPD and the CPs identified SCCs and decision-making, respectively, as particularly relevant cognitive endpoints. Our findings highlight the need to further refine instrumentation to assess SCCs in an ecologically valid manner.
Our study had two important limitations. First, we used a relatively small, predominantly Caucasian sample of pwPD and CPs. Future studies will need to further investigate SCCs and treatment targets in broader and more diverse PD populations to determine if demographic factors impact SCCs more broadly. Second, the cognitive characterization of our sample of pwPD was limited to MoCA screening. Because depression is strongly associated with reports of cognitive symptoms (Alzahrani and Venneri, 2015), future studies should include affective measures to help tease apart the influence of a patient’s emotional state on his or her particular cognitive symptoms.
The authors thank Teresa Heemsbergen, Ronald A. Nelson, David O. Konesheck, Bonny Konesheck, Charles S. Aaron (our Parkinson’s Advocates in Research), and Melissa Ainslie (our research coordinator) for their contributions to this research.
Aarsland D, Creese B, Politis M, et al. 2017. Cognitive decline in Parkinson disease
. Nat Rev Neurol. 13:217–231.
Aarsland D, Larsen JP, Tandberg E, et al. 2000. Predictors of nursing home placement in Parkinson’s disease: a population‐based, prospective study. J Am Geriatr Soc. 48:938–942.
Alzahrani H, Venneri A. 2015. Cognitive and neuroanatomical correlates of neuropsychiatric symptoms in Parkinson’s disease: a systematic review. J Neurol Sci. 356:32–44.
Bechara A, Damasio AR, Damasio H, et al. 1994. Insensitivity to future consequences following damage to human prefrontal cortex. Cognition. 50:7–15.
Benge JF, Balsis S, Madeka T, et al. 2017. Factor structure of the Montreal Cognitive Assessment items in a sample with early Parkinson’s disease. Parkinsonism Relat Disord. 41:104–108.
Benge JF, Heemsbergen T, Nelson RA, et al. 2018. Toward patient-centered outcomes for cognitive evaluations: the perspective of those affected by Parkinson’s disease. The Clinical Neuropsychologist. 32:1303–1318.
Brennan L, Siderowf A, Rubright JD, et al. 2016. The Penn Parkinson’s Daily Activities Questionnaire—15: psychometric properties of a brief assessment of cognitive instrumental activities of daily living in Parkinson’s disease. Parkinsonism Relat Disord. 25:21–26.
Brown DS, Bernstein IH, McClintock SM, et al. 2016. Use of the Montreal Cognitive Assessment and Alzheimer’s Disease-8 as cognitive screening measures in Parkinson’s disease. Int J Geriatr Psychiatry. 31:264–272.
Cardona JF, Gershanik O, Gelormini-Lezama C, et al. 2013. Action-verb processing in Parkinson’s disease: new pathways for motor-language coupling. Brain Struct Funct. 218:1355–1373.
Castro PCF, Aquino CC, Felício AC, et al. 2016. Presence or absence of cognitive complaints in Parkinson’s disease: mood disorder or anosognosia? Arquivos de Neuro-psiquiatria. 74:439–444.
Copeland JN, Lieberman A, Oravivattanakul S, et al. 2016. Accuracy of patient and care partner identification of cognitive impairments in Parkinson’s disease–mild cognitive impairment. Mov Disord. 31:693–698.
Dalrymple-Alford JC, MacAskill MR, Nakas CT, et al. 2010. The MoCA: well-suited screen for cognitive impairment in Parkinson’s disease. Neurology. 75:1717–1725.
Dujardin K, Duhamel A, Delliaux M, et al. 2010. Cognitive complaints in Parkinson’s disease: its relationship with objective cognitive decline. J Neurol. 257:79–84.
Farias ST, Lau K, Harvey D, et al. 2017. Early functional limitations in cognitively normal older adults predict diagnostic conversion to mild cognitive impairment. J Am Geriatr Soc. 65:1152–1158.
Hoehn MM, Yahr MD. 1967. Parkinsonism: onsent, progression, and mortality. Neurology. 17:427–442.
Hong JY, Lee Y, Sunwoo MK, et al. 2018. Subjective cognitive complaints
and objective cognitive impairment in Parkinson’s disease. J Clin Neurol. 14:16–21.
Janvin CC, Larsen JP, Salmon DP, et al. 2005. Cognitive profiles of individual patients with Parkinson’s disease and dementia: comparison with dementia with lewy bodies and Alzheimer’s disease. Mov Disord. 21:337–342.
Kandiah N, Zhang A, Cenina AR, et al. 2014. Montreal Cognitive Assessment for the screening and prediction of cognitive decline in early Parkinson’s disease. Parkinsonism Relat Disord. 20:1145–1148.
Kim A, Kim HJ, Shin CW, et al. 2018. Emergence of non-motor fluctuations with reference to motor fluctuations in Parkinson’s disease. Parkinsonism Relat Disord. 54:79–83. doi:10.1016/j.parkreldis.2018.04.020
Koster DP, Higginson CI, MacDougall EE, et al. 2015. Subjective cognitive complaints
in Parkinson disease
without dementia: a preliminary study. Appl Neuropsychol Adult. 22:287–292.
Lawson RA, Yarnall AJ, Johnston F, et al. 2017. Cognitive impairment in Parkinson’s disease: impact on quality of life of carers. Int J Geriatr Psychiatry. 32:1362–1370.
Lee HM, Koh SB. 2015. Many faces of Parkinson’s disease: non-motor symptoms of Parkinson’s disease. J Mov Disord. 8:92–97.
Lehrner J, Moser D, Klug S, et al. 2014. Subjective memory complaints, depressive symptoms and cognition in Parkinson’s disease patients. Eur J Neurol. 21:1276–1284.
Litvan I, Goldman JG, Tröster AI, et al. 2012. Diagnostic criteria for mild cognitive impairment in Parkinson’s disease: Movement Disorder Society Task Force guidelines. Mov Disord. 27:349–356.
Mills KA, Mari Z, Pontone GM, et al. 2016. Cognitive impairment in Parkinson’s disease: association between patient-reported and clinically measured outcomes. Parkinsonism Relat Disord. 33:107–114.
Muslimović D, Post B, Speelman JD, et al. 2009. Cognitive decline in Parkinson’s disease: a prospective longitudinal study. J Int Neuropsychol Soc. 15:426–437.
Nasreddine ZS, Phillips NA, Bédirian V, et al. 2005. The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc. 53:695–699.
Peto V, Jenkinson C, Fitzpatrick R. 1998. PDQ-39: a review of the development, validation and application of a Parkinson’s disease quality of life questionnaire and its associated measures. J Neurol. 245 (suppl 1):S10–S14.
Schiehser DM, Liu L, Lessig SL. 2013. Predictors of discrepancies in Parkinson’s disease patient and caregiver ratings of apathy, disinhibition, and executive dysfunction before and after diagnosis. J Int Neuropsychol Soc. 19:295–304.
Schrag A, Jahanshahi M, Quinn N. 2000. What contributes to quality of life in patients with Parkinson’s disease? J Neurol Neurosurg Psychiatry. 69:308–312.
Sitek EJ, Sołtan W, Wieczorek D, et al. 2011. Self-awareness of memory function in Parkinson’s disease in relation to mood and symptom severity. Aging Ment Health. 15:150–156.
StataCorp. 2013. Stata Statistical Software: Release 13 [computer software]. College Station, Texas: StataCorp LP.
Vlagsma TT, Koerts J, Fasotti L. 2016. Parkinson’s patients’ executive profile and goals they set for improvement: why is cognitive rehabilitation not common practice? Neuropsychol Rehabilitation. 26:216–235.
Vossius C, Larsen JP, Janvin C, et al. 2011. The economic impact of cognitive impairment in Parkinson’s disease. Mov Disord. 26:1541–1544.
Watson GS, Leverenz JB. 2010. Profile of cognitive impairment in Parkinson’s disease. Brain Pathol. 20:640–645.
Zarit SH, Reever KE, Bach-Peterson J. 1980. Relatives of the impaired elderly: correlates of feelings of burden. Gerontologist. 20:649–655.