Progressive cognitive impairment is a prominent non-motor symptom of Parkinson’s disease (PD) (Aarsland et al., 2021), and ranges from mild cognitive impairment (PD-MCI) (Litvan et al., 2012) to dementia (PDD) (Emre et al., 2007). Both diagnoses require evidence of difficulties on cognitive testing and reports of cognitive decline from the patient and/or care-partners. A PDD diagnosis additionally requires that cognitive impairments are sufficiently severe to interfere with the independent execution of instrumental activities of daily living (iADLs). Clinicians and researchers face several challenges in evaluating cognitively-mediated functional independence in PD, which can compromise diagnostic accuracy, and consequently, treatment planning and outcome measurements. First, the effect of cognitive difficulties must be distinguished from that of cardinal motor symptoms in PD (Cheon et al., 2015). Second, sampled activities should encompass a range of common iADLs to account for individual variability in roles and responsibilities (Kulisevsky et al., 2025). Third, severity of iADL difficulties should be captured as some changes in functioning may occur in PD-MCI, even prior to the onset of PDD (Kulisevsky et al., 2013).

Rating scales address some of these challenges by capturing the degree of difficulty across common iADLs. Although most iADL scales, which were primarily developed for use in Alzheimer’s disease, do not typically differentiate between the contributions of motor and cognitive symptoms to iADL impairments, some methods address this issue. One strategy has been to develop PD-specific iADL questionnaires, selecting activities with strong cognitive requirements (Brennan et al., 2016), and instructing responders to solely consider cognitive difficulties when rating functioning (Kulisevsky et al., 2013). Another approach has been to adapt existing iADL questionnaires either by eliminating items (Almeida et al., 2017) or modifying instructions (Cheon et al., 2015), although such modifications may alter the psychometric properties of these instruments. Alternatively, one group has modified scoring methods without changing test administration. Specifically, Becker et al. (2020) administered the standardized Functional Activities Questionnaire (FAQ) (Pfeffer et al., 1982), and categorized each item as motor, cognitive, or both. Their categories were based on each item’s degree of relationship with independent measures of motor symptoms [i.e., Unified Parkinson’s Disease Rating Scale (UPDRS-III] and global cognition [i.e., Montreal Cognitive Assessment (MoCA)]. The authors then derived motor and cognitive subscores by summing ratings for the respective item types and computed a quotient (FAQ_Q = cognitive subscore ÷ motor subscore). Patients classified as having cognitive iADL impairment (i.e., FAQ_Q > 1) showed greater impairments in attention/working memory and language, and were more likely to convert to PDD in the years that followed, highlighting the validity and possible prognostic value of this method (Becker et al., 2022a).

While innovative, this modified scoring approach relies on an indirect attribution of the underlying cause of the functional difficulty. Thus, it is unclear whether this corresponds to the respondent’s impression of the degree to which each type of symptom limits the patient’s ability to perform iADLs based on their lived experience. Second, given that the FAQ_Q was validated in patients in the early stages of PD, it is unclear whether the suggested threshold to identify the presence of cognitively-driven iADL impairment remains valid in patients with longer disease duration. Notably, the requirement that the cognitive subscore be greater than the motor subscore (FAQ_Q > 1) may not be met in patients with marked motor symptoms or fluctuations, such as those being considered for Deep Brain Stimulation (DBS). Accurate cognitive diagnosis is paramount in this group as PDD is considered a contraindication for DBS (Lang et al., 2006). Furthermore, although there is moderate agreement between patients’ and care-partners’ ratings on iADL questionnaires, including the FAQ (Becker et al., 2022b; Cholerton et al., 2020; Deck et al., 2019), the FAQ_Q was developed mainly from ratings by care-partners and did not expressly consider those of patients unless a care-partner’s ratings could not be obtained. It is common practice in PD to rely on care-partners’ reports, as patients’ insight may be compromised with increasing disease severity and cognitive impairment (Becker et al., 2022b; Deck et al., 2019). However, Cholerton et al. (2020) found that patient ratings are more sensitive than care-partner ratings in earlier stages of cognitive impairment (i.e., distinguishing normal cognition from PD-MCI). Lastly, several factors can influence care-partner ratings, including caregiver burden, quality or type of relationship to patient, education level, and cultural differences (Hackett et al., 2020), and knowledgeable care-partners may not always be available to complete rating forms. These findings highlight the importance of also examining patients’ self-report.

The present study examines both patients’ and care-partners’ evaluations of the source of iADL difficulties endorsed in pre-DBS patients with disabling motor symptoms and fluctuations. To this aim, the standardized FAQ was extended to include an explicit rating of the degree to which motor and cognitive symptoms contribute to iADL difficulties. This extension was administered after the standardized FAQ was completed to preserve the administration procedure and psychometric properties of the original form. We then explore whether ratings on the extended FAQ match the categorization used in the modified scoring method (Becker et al., 2020) in that items categorized as motor have higher reported limitation ascribed to motor symptoms than to cognitive symptoms, and conversely, items categorized as cognitive have higher reported limitation ascribed to cognitive symptoms than to motor symptoms. Thirdly, we examined the categorization of standardized FAQ items based on their degree of relatedness with independent measures of motor symptoms (UPDRS-III) and global cognition [Dementia Rating Scale-2 (DRS-2) (Jurica et al., 2004)]. This followed a similar approach to that used in previous research (Becker et al., 2020) in order to determine whether the same classification is achieved in PD patients with longer disease duration and more severe motor symptoms. Lastly, we derive a measure of cognitive iADL impairment based on patients’ and care-partners’ ratings on our extended FAQ. Specifically, we calculate the Reported Cognitive Burden (RCB) which measures the proportion of iADL difficulties ascribed to cognitive symptoms [i.e., cognitive ratings ÷ (cognitive + motor ratings)]. To validate our novel measure, we then examine the degree to which the RCB and the FAQ_Q relate to current cognitive status (cognitively normal vs. PD-MCI) and performance on neuropsychological tests across five cognitive domains.

We examined the validity of two approaches aimed at isolating the contribution of cognitive symptoms to iADL difficulties from that of motor symptoms based on responses on the FAQ (standard and extended versions) in PD patients with disabling motor symptoms and fluctuations. This distinction is a major challenge for clinicians and is critical for accurate cognitive diagnosis, prognosis, and intervention planning. We first applied a previously validated scoring method of the standard FAQ (Becker et al., 2020), which yields a quotient (FAQ_Q) reflecting the ratio of difficulties on FAQ items associated with objective cognitive measures to those on FAQ items related to motor scores. We then used a novel extension of the FAQ that explicitly asks patients and care-partners to rate the extent to which each endorsed iADL difficulty is due to motor and cognitive causes, respectively, and then converted these ratings into the Reported Cognitive Burden (RCB) score. In our pre-DBS cohort, motor symptoms were most often identified as the greater source of limitation by both patients and care-partners, even for items previously classified as “cognitive” (Becker et al., 2020). The FAQ_Q did not differ significantly between patients with PD-MCI and those with intact cognition (PD-CN) in our sample, and its correlations with objective cognitive measures were weak and did not reach statistical threshold. Thus, our results did not replicate previous findings in a sample with more severe motor disability. In contrast, our new RCB score, derived directly from attributions based on lived-experience, was higher in PD-MCI than PD-CN, and related to performance in attention and executive domains, regardless of whether ratings were those of patients or of care-partners. Thus, the RCB offers a sensitive means of detecting mild functional changes related to cognition in the context of disabling motor symptoms, including changes that precede the overt loss of independence seen in PD dementia (PDD). As such, it may provide support in the diagnosis of cognitive dysfunction in PD. In the sections that follow, we discuss how these findings align with and extend prior literature on questionnaire-based iADL measurement in PD, alternative approaches to iADL assessment, strengths and limitations, and future directions.

Determining the impact of cognitive decline on one’s ability to function independently is essential when making a dementia diagnosis but there is increasing recognition that subtle changes in the ability to carry out cognitively demanding iADLs may be present even at earlier stages of PD (Kulisevsky et al., 2013). Therefore, there is a pressing need to assess functional impairment on a continuum, to capture even minor changes that may occur early in the disease course, and to sample various relevant activities rather than relying on a patient’s ability to perform any single activity (Kulisevsky et al., 2025). Using standardized questionnaires allows researchers and clinicians to address these needs in a time- and resource-efficient way. However, it is important to recognize that most existing questionnaires were designed for different patient populations (typically Alzheimer’s disease). As such, they either do not ask about the source of disability (physical, cognitive, emotional, behavioral) or focus on iADLs impacted by memory and disorientation but not necessarily those impacted by executive dysfunction which are most common in PD (Marras et al., 2014). As a result, there has been increased effort toward adapting existing questionnaires to better suit PD populations (Almeida et al., 2017; Cheon et al., 2015) or developing PD-specific questionnaires such as the Penn Daily Activities Questionnaire (PDAQ) (Brennan et al., 2016) and the Parkinson’s Disease Cognitive Functional Rating Scale (PD-CFRS) (Kulisevsky et al., 2013).

An alternative approach is to modify only the scoring of an established questionnaire as done by Becker et al. (2020). This has the advantage of preserving the psychometric properties of the original standardized questionnaire as there is no modification of the items or of the test administration, and this is less resource-consuming than developing and validating a new PD-specific measure. An additional strength is the potential prognostic value of this method. Indeed, in a longitudinal study (Becker et al., 2022a), the presence of cognitive iADL impairment (FAQ_Q > 1) combined with PD-MCI at baseline predicted conversion to PDD more strongly than PD-MCI alone (HR = 12.01 vs. HR = 5.34). Although innovative, the FAQ scoring method treats individual iADLs as binary (either cognitive or motor), but our results indicate that it is common for iADL performance to be impacted by more than one source of difficulty, as patients and care-partners rated both motor and cognitive symptoms as contributing to all iADLs on the extended FAQ. This is further supported by Benge and Balsis (2016), who showed that care-partners often identified multiple sources of difficulty when asked about the impact of physical, cognitive, and emotional/behavioral problems on iADLs. Thus, approaches that classify items exclusively as “motor” or “cognitive” may be too coarse.

Moreover, because our sample was derived from patients undergoing evaluation for DBS candidacy—a treatment indicated for individuals with motor fluctuations, pronounced OFF periods, levodopa-induced dyskinesias, or tremor inadequately controlled with pharmacotherapy—it is anticipated that motor disability would strongly contribute to iADL difficulties. Accordingly, the proposed FAQ_Q > 1 cutoff may be less applicable in this cohort, as cognitive symptoms are unlikely to overshadow the motor contribution to iADLs in this population. This is further illustrated by the fact that 15.5% of our sample had a FAQ_Q > 1 based on care-partners’ ratings versus 23.9% of the sample in Becker et al. (2022a) despite having higher rates of PD-MCI (58% vs. 29%), longer disease duration (10.4 vs. 4.1 years) and higher intake of dopaminergic medications (1,365 mg vs. 511 mg per day). Our RCB may provide a useful alternative, or addition to the FAQ_Q, for addressing these challenges as it is derived from subjective ratings and represents cognitive contribution as a proportion of the overall difficulty rather than as a relative ratio between symptom types. Importantly, only the RCB and not the FAQ_Q differentiated PD-MCI from PD-NC in our pre-DBS population and was related to performance on tasks assessing executive functioning and attention, two domains that are particularly vulnerable in PD (Aarsland et al., 2021). These results contrast with prior ones using the FAQ_Q in non-surgical PD patients (Becker et al., 2020).

Although questionnaires are efficient and practical tools that can be readily utilized in routine clinical care, they inherently rely on self-report which may be influenced by respondents’ mood or level of insight, as well as by caregiver burden (Carlisle et al., 2023), potentially leading to either overestimation or underestimation of a patient’s functional abilities (Shulman et al., 2006; Pirogovsky et al., 2014). To provide a more objective assessment of how cognition impacts daily functioning, the Movement Disorder Task Force previously recommended using the Pill Questionnaire which asks patients to verbally describe their medication schedule (Dubois et al., 2007). If the patient can successfully identify the medications and their timing, cognition is considered to have no impact on daily living. However, patients’ declarative memory of their medication regimen may not necessarily match their daily medication adherence. Moreover, medication regimens can differ greatly across patients in terms of complexity and the frequency with which changes are made by their physician, making the difficulty of this task inconsistent. Also, while medication management is a particularly salient and repetitive part of disease management, this questionnaire might not reflect changes in other functional domains. Given these caveats, it is perhaps not surprising that this measure failed to diagnose around half of the probable PDD patients compared to neurologists’ assessments in one study (Lee et al., 2014). Notably, the patients diagnosed only by neurologists had lower motor symptoms and higher cognitive functioning, suggesting the Pill Questionnaires may lack sensitivity and only capture individuals on the more severe end of the PDD spectrum.

Performance-based tasks are standardized measures where patients are asked to perform one or more iADLs in a laboratory setting. They present an attractive alternative to questionnaires because of their standardized assessment and scoring, increased ecological validity, minimal reliance on patient or care-partner insight, and focus on real-life skills that may be targets for intervention (Pirogovsky et al., 2014). On these measures, PD patients perform more poorly on tasks with high cognitive demands such as “mock” medication (Shulman et al., 2006; Manning et al., 2012; Foster, 2014; Pirogovsky et al., 2014; Sumida et al., 2021; Schmitter-Edgecombe et al., 2022) and financial management (Martin et al., 2013; Pirogovsky et al., 2014; Ariesen et al., 2024). Poor performance on these tasks is often related to a variety of cognitive abilities (Foster, 2014; Holden et al., 2018; Fellows and Schmitter-Edgecombe, 2019; Sumida et al., 2021) including attention and executive functioning (Manning et al., 2012; Glonnegger et al., 2016; Sulzer et al., 2020; Schmitter-Edgecombe et al., 2022). However, correlations with cognitive abilities are not always replicated (Pirogovsky et al., 2014; Ariesen et al., 2024), and performance can also be related to the severity of motor symptoms (Manning et al., 2012; Foster, 2014; Glonnegger et al., 2016; Holden et al., 2018; Fellows and Schmitter-Edgecombe, 2019) and advanced age (Manning et al., 2012; Foster, 2014). Therefore, although performance-based measures are enticing, they are also imperfect indicators of cognitive iADL functioning. It is important to consider that patients are asked to perform these assessments outside of their home environment where they have established routines, and may be able to use compensatory strategies, assistive devices, and environmental cues to improve their functioning (Shulman et al., 2006; Pirogovsky et al., 2014; Schmitter-Edgecombe et al., 2022). Tasks performed in a laboratory are likely less emotionally salient than those performed at home with real consequences for functioning and independence (e.g., sorting beans into a dosette versus taking one’s own medications as prescribed which improve motor symptoms). Performance-based measures may also underestimate everyday ability as individuals may perform more poorly when they are aware that they are being observed and evaluated. Finally, logistical challenges such as lack of dedicated space to set up a faux environment and the time-consuming nature of many of these tasks limit their utility in clinical settings.

Our study has several strengths such as including a sample of consecutive and diverse PD patients (see Statucka and Cohn, 2019) with advanced disease who underwent comprehensive neuropsychological assessments with questionnaires completed by both patients and care-partners. Our recent work (Statucka et al., 2023) with an overlapping sample also highlighted the importance of considering patients’ and care-partners’ lived-experience, which we have once again done with our extension of the FAQ. While PD is an extreme case of physical disability, our method could also be applied to other populations who have both physical and cognitive limitations such as cerebrovascular disease (Babulal et al., 2015; Blomgren et al., 2019) or elderly populations with sensory impairments (e.g., vision and hearing loss) or orthopedic issues (Liu et al., 2016; Shin and Kim, 2022; Zhou et al., 2025).

In terms of limitations, while the use of consecutively collected clinical data within a universal health care system may be less prone to recruitment biases than prospective research studies, it results in a sample that is not representative of a general PD population. As our patients are considering DBS surgery for treatment of their severe motor symptoms and fluctuations, the degree to which motor and cognitive symptoms contribute to their iADL difficulties likely differ from other subsets of PD patients. Due to the large impact of motor symptoms on daily living, pre-DBS patients might underestimate the effect of cognitive symptoms or have less opportunity to engage in certain activities, making it difficult to estimate the degree to which other sources contribute. It is also possible that patients and care-partners tailor their responses in a way that they hope may be favorable to their DBS candidacy determination (e.g., perhaps minimizing cognitive difficulties and/or amplifying the negative impact of motor symptoms to functioning). Moreover, in our center, patients with clear dementia are excluded early in DBS evaluation process, prior to the completion of our comprehensive neuropsychological assessments where we collect FAQ data. Thus, we are unable to evaluate the diagnostic precision of the RCB or FAQ_Q in terms of PDD.

Another limitation is that we did not account for other variables such as mood in our analyses, though there is evidence that depressive symptoms are a strong predictor of iADLs in both the general PD and pre-DBS patient populations (Choi et al., 2019; Bezdicek et al., 2022). Although our results regarding the relationship of the novel RCB with cognitive status remained even after excluding individuals reporting a moderate-severe level of depressive symptoms on questionnaires (Supplementary Table 1), such scales do not provide a formal diagnosis of depression. Additionally, depression questionnaires are not specific to mood as they include items also related to PD symptoms such as subjective cognitive complaints, sleep disturbances, and autonomic symptoms. Thus, future research should examine how formally diagnosed mood disorders relate to iADL difficulties in PD. Finally, this study was cross-sectional and thus, we could not determine the prognostic value of having a higher RCB or FAQ_Q, nor investigate how pre-surgical motor and cognitive functions contribute to iADLs changes post-DBS. Thus, we aim to examine this approach in post-surgical patients who are likely to demonstrate improved motor symptoms and daily function, but perhaps cognitive decline following DBS. Future longitudinal studies should also aim at validating this method in non-surgical PD cohorts across cognitive diagnoses (i.e., intact cognition, PD-MCI, PDD) and disease stages, to examine to potential utility of this tool in the care of the broader PD population.

Our study once again emphasizes the importance of considering patients’ and care-partners’ lived-experiences to estimate the degree to which cognition impacts patients’ functional independence. The fact that similar results were obtained with patients’ and care-partners’ ratings is reassuring as it demonstrates that patients have sufficient insight into their own difficulties, and that their impressions should be considered especially when a knowledgeable care-partner is not available. Potential future extensions and applications of our findings are threefold. First, beyond patients with PD, our extended FAQ, and the RCB in particular, could be useful with other medically complex populations experiencing both physical and cognitive limitations (e.g., sensory or physical disability). Second, clarifying the impact of various symptoms on daily functioning is not only important for accurate cognitive diagnosis, but more broadly may inform treatment planning and engagement. For example, if a patient attributes their functional difficulties solely to physical symptoms, they may be unwilling to engage in cognitive remediation or treatment aimed at mood. Understanding patients’ attributions of their functional difficulties may lead to more personalized intervention plans with better outcomes for the individual. Finally, having a better understanding of the degree to which cognitive versus physical symptoms interfere with iADLs may also improve clinicians’ ability to identify individuals at elevated risk of developing dementia.