Relative Meaningfulness and Impacts of Symptoms in People with Early-Stage Parkinson’s Disease

Setting, Sample

This study was conducted collaboratively with Critical Path Institute, US Food and Drug Administration, individuals with PD, and academic and industry partners. It was designed as follow up to the parent WATCH-PD study (NCT03681015), which was a 12-month multi-center observational study evaluating use of a smartwatch and smartphone applications to detect cognitive and motor progression over time in individuals with early, untreated PD [4]. This dual-purpose follow-up study was conducted for the purpose of 1) identifying meaningful symptoms and impacts (i.e., functional or psychosocial changes resulting from symptoms) of early untreated PD and 2) exploring the extent to which the digital measures in the parent study were perceived as relevant to monitoring meaningful aspects of disease by participants. This manuscript presents results from Aim 1. All individuals with early PD who completed their final visit of the WATCH-PD parent study [13] within 6 months were eligible to participate (N = 54). Main inclusion criteria for the parent study were: 1) diagnosed PD duration≤2 years; 2) Modified Hoehn & Yahr stage≤2; and 3) not taking any PD medications. Main exclusion criteria were: 1) confounding comorbidities; 2) history of PD related falls; and 3) Montreal Cognitive Assessment score (MoCA) < 24 [4, 14]. The enrollment diagram is shown in Fig. 1. Participants from the parent study were randomly selected for interviews, contacted via phone, and screened for interest and eligibility. Forty (50%) were enrolled to correspond with parent study gender distribution with inclusion of all participants from underrepresented groups (N = 4). Sample size was based on maximal ranges identified in prior qualitative descriptive studies [15]. Data saturation was assessed after completing all 40 interviews to confirm adequacy [16]. Saturation was defined as the point after which no new or additional symptoms/impacts were identified in succeeding interviews. IRB approval was obtained (IRB# 00006429) and participants provided digital informed consent.

OPEN IN VIEWER

Fig. 1

Enrollment diagram.

Data collection

An integrated mixed-methods approach was used [15, 17] consisting of a preliminary survey followed by 1:1 online interviews that used symptom mapping in conjunction with a semi-structured interview protocol, as described below. All data collection procedures were developed in collaboration with people with Parkinson’s disease and the study advisory group, and were pretested (JC, JH).

Survey

Participants first completed a brief online survey (Supplement A) using Redcap, which is a secure, web-based software platform designed to support survey data capture [18, 19]. The purpose of the survey was to gather demographic data and preliminary qualitative information on personally important symptoms, which was used as a starting point for the symptom mapping interview. For the survey, participants were asked to describe all symptoms of PD that they experienced and explain which were most bothersome using open response items.

Online interview

One-week later, online interviews were conducted via Zoom videoconferencing with participants at home (average 100 min). As shown in Supplement B, interview questions focused on understanding PD symptoms, experiences, characteristics and impacts of symptoms, and relative bothersomeness. Symptom mapping [20] was used to delineate all personal symptoms and impacts as shown in Fig. 2. Symptom mapping is a hybrid mixed-method technique that enables the collection of qualitative data in a quantifiable manner by active categorization of experiences inside of a structured framework [20]. Interviews were conducted by a qualitative researcher experienced in these methods (JM; white, female, PhD-prepared advanced practice nurse, unacquainted with participants). Interviews were audio-visually recorded with permission.

OPEN IN VIEWER

Fig. 2

Sample participant symptom map.

Symptom mapping [20]

During the interview, a detailed concept map of the individual’s PD symptom experience was developed by the interviewer, as observed and directed by the participant via Zoom screen sharing [21]. Individual symptom maps were created using Xmind^TM software, with map levels organized hierarchically top to bottom from “Most bothersome” to “No current issues,” with an additional category “Not [personally] relevant to early PD”. This categorization of bothersomeness was intended to reflect the extent to which a symptom disturbed or distressed the individual.

Step 1. Prior to the interview, the researcher reviewed the survey data and entered reported symptoms into a preliminary map, with each symptom represented as a single yellow node.

Step 2. At the start of the interview, the participant was oriented to the mapping process and shown the preliminary map via screensharing. They were then asked to list and describe all symptoms of PD and how these impacted their life (past or present), while the researcher entered this information into the mutually viewable map.

Step 3. Next, the researcher probed for common symptoms of PD identified in prior literature (difficulties with tremor, walking, balance, fine motor, speech, thinking, mood, daytime sleepiness, fatigue, depth perception) if not spontaneously mentioned by the participant. Symptoms not experienced were categorized under “No current issues", or “Not relevant to early PD” based on the participants’ perspective.

Step 4 - Bothersomeness. Once symptoms and impacts were fully described, supporting details were collapsed (i.e., hidden) leaving only primary yellow “symptom” nodes visible. The participant was then asked to rank symptoms according to bothersomeness (i.e., how distressing the symptom was from most to least).

Step 5 - Importance. Lastly, participants identified which symptoms were most important to them overall and explained any differences between what was important vs. bothersome. A final opportunity was provided to review and edit the map, and copies of maps were given to participants at the end of the interview, if desired.

Data analysis

Content coding [22] was performed on symptom maps with thematic analysis of verbatim transcripts [23]. Maps were coded for type, frequency, and bothersomeness of symptoms and impacts. As shown in Supplement C, each hierarchical level in the map was associated with a Patient Reported Symptom Score (PRSS; range 0–4), where scores 1–4 indicated the symptom was present and degree of bothersomeness, 0 indicated the symptom was not present but still viewed as important to the participant, and “.” Indicated that the symptom was not present and not personally relevant.

Coding was performed in cycles by two coders (JM, PY) and differences resolved by consensus. In Cycle 1, open coding (i.e., no a priori coding schema) was performed on maps to develop a comprehensive list of symptom types using spreadsheets. Cycle 2, maps were coded again quantify frequencies and bothersomeness of each symptom by participant. In Cycle 3, maps were coded to derive a comprehensive list of all impacts with details on symptoms that contributed to the impact. Cycle 4, maps were re-coded to quantify frequencies of impacts by contributing symptoms for each participant. In Cycle 5, the frequencies of symptoms and impacts were compared to the Staunton conceptual model of early PD [11], with attention to divergence or alignment with conceptual domains and domain items [22]. Lastly, in Cycle 6, inductive thematic analysis was conducted on narratives using Nvivo12, starting with open coding to identify recurrent ideas within interviews, followed by pattern coding to identify dominant themes across interviews regarding how participants experienced and perceived early PD symptoms [24, 25]. Descriptive statistics were computed for demographic survey items. Independent T-tests were conducted in SPSS 28 to assess for any differences in symptom frequencies between those taking versus not taking PD medication.

Rigor

Procedures to ensure rigor included: co-development and pretesting of study procedures with people with PD (surveys, interview guide, mapping procedures), observing of interviews for consistency (RS), triangulated data collection approaches (preliminary survey, followed by mapping and cognitive interviewing), member-checking during interviews, peer-debriefing on thematic findings at weekly scheduled meetings, use of multiple coders, participant identifiers to show representativeness of quotes, and a formal audit trail [26]. Symptom maps were returned to participants who reviewed and confirmed validity of their personal data.

Sample and interview characteristics

Of 54 eligible participants from the parent study, one declined to participate, 5 could not be reached, and 8 were not solicited for interviews due to having achieved targeted sample size, as delineated in Fig. 1. Demographic data comparative to parent study demographics are displayed in Table 1. Participants were mostly white, male, and not taking PD medication at the time of interview. Data saturation for symptoms, impacts and themes was achieved by the 17th of 40 interviews, after which no new findings emerged.

Symptoms and impacts

Symptoms frequencies in early PD (All bothersome symptoms; PRSS 1-4)

Motor and non-motor symptoms of early PD are displayed by frequency of bothersomeness in Fig. 3. There were no significant differences in symptoms between those taking PD medications (N = 16/40; 40%) and those not taking PD medications (p > 0.05; range 0.083— 0.986). For all, tremor was the most commonly reported motor symptom (95%), followed by fine motor difficulties (87%), and slow movements (80%). Over half of people reported gait changes, stiffness/rigidity, and quiet voice. For example:

OPEN IN VIEWER

Fig. 3

Patient reported motor and non-motor symptoms of early Parkinson’s disease as shown in symptom maps (N = 40). PRSS 1 = Patient Reported Symptom Score – Likert scale rating of bothersomeness ranging from most bothersome to not bothersome. PRSS 4 = Symptoms that are present and most bothersome; PRSS 3 = Symptoms that are present and somewhat bothersome symptoms; PRSS 2 = Symptoms that are present and less bothersome; PRSS 1 = Symptoms that are present but not bothersome; Not present, Symptom not experienced. PRSS Scores are absolute. Figure includes both primary and contributing symptoms for each PRSS level. Percentage (%) represents the total percent of participants who experience the symptoms (encompassing 1–4). Tremor subcategories included: Hand tremor (85%) Leg/foot tremor (42.5%; Face/neck/Jaw tremor 12.5%). The following symptoms were reported by < 10% sample and are not represented in the graphs: dry mouth, diminished sensation, temperature dysregulation, sexual dysfunction, tearing of eyes, loss of appetite, double vision, right/left confusion, hemi-spatial neglect. Dyskinesias were not reported.

P14 : I move in slow motion like I’m a sloth. It’s just so frustrating ...  I can’t [ever] hurry. [It’s] like I’m turning into a stone.

The most common non-motor symptoms were nocturia (65%), feeling tired or fatigued (62.5%), difficulty concentrating (62.5%), and insomnia (60%), with more than half of participants reporting slow thinking, difficulty remembering, and anxiety.

“Most bothersome” symptoms (PRSS 4)

Participants identified an average of 10 “most bothersome” symptoms (range 0–28). Again, difficulties with tremor (63%), fine motor (48%), and slow movements (40%) were most reported. However, when evaluating symptoms categorically by area of impact (e.g., mobility & balance, speech, mood, cognitive changes) rather than individually, symptoms affecting mobility and balance (gait, posture, balance, slowness, stiffness/rigidity) were most bothersome to 57% (Supplement D). Other most bothersome categories included changes to speech (40%), disturbed sleep (30%), altered thinking (27%), and altered mood (22%).

“Most important” symptoms of early PD

Participants identified an average of 2 symptoms that they believed were “most important.” These were tremor (27%), fine motor (25%), slow movements (12.5%), and word finding difficulties (10%). As shown in Supplement E, when evaluated categorically rather than as individual symptoms, 32.5% identified issues of mobility as “most important” and 12.5% identified mood changes and cognitive difficulties.

Impacts of early PD symptoms

Figure 4 displays the frequencies of different impacts experienced in early PD, along with the symptom believed to have contributed to the impact. These were evaluated with reference to Staunton conceptual model, which included activities of daily living (ADL), physical, social, and emotional & psychological functioning, and fine motor domains [11]. In our analysis, “fine motor skills” (i.e., computer/smartphone use, handwriting) was subsumed under ADL based on descriptive patterns that indicated these impacts typically occurred within the context of ADLs, as described below.

OPEN IN VIEWER

Fig. 4

Impacts of disease in people with early Parkinson’s disease. Table comparative to the Patient-Centered Conceptual Model of symptoms and impacts in early PD from Staunton et al. (2022) [11]. Items with * were modified from or not present in the original model. Impacts affecting < 10% of total sample (N = 4/40) are not reported. Green shaded boxes visually represent number of participants reporting a particular impact as present. Red shaded boxes visually represent number of participants reporting a specific symptom as contributing to or causing the listed impact. Relative bothersomeness of impacts was not quantified and occurred across multiple symptom levels. Impacts were counted only once per participant.

ADL

The most discussed ADL impacts were altered sleep patterns and increased difficulty performing one’s job (70%). Writing (70%) and using a computer mouse/keyboard (68%) contributed largely to job-related difficulties, with close to half referencing cognitive changes (42.5%), including slower thinking, increased difficulty following sequences, and multitasking.

Difficulty using a computer appeared to have greatest impact on job function (e.g., sending emails, working in spread sheets), whereas handwriting affected both job and social interactions (ability to write notes/letters, holiday cards, or sign documents). Difficulties with the computer were most commonly due to slow fine motor movements (missing keys, double striking, holding down keys too long, difficulty manipulating the mouse, click/drag function). Other common ADLs impacts affecting more than half of participants included eating and drinking (choking, trouble using utensils due to tremor/fine motor) and increased difficulty driving (depth perception, anxiety, reaction times).

Physical functioning

Changes in walking, which affected exercise, were most common (75%; 63%, respectively). Less than one-third of participants experienced difficulties in other physical functioning areas, such as falling, getting up from sitting, standing, or climbing stairs. Most were physically active, exercised regularly, and had few physical limitations that were apparent to others.

Social

Many indicated PD affected their ability to communicate and interact as usual (78%), including energy to maintain social interactions (fatigue), ability to express ideas rapidly within a group context (cognitive), speak clearly (articulation) and be easily heard (quiet voice), or formulate written communications. PD also impacted personal relationships with others. This was most often reported with respect to immediate family, but also affected friends and co-workers. Impacts on relationships were due to increased dependence on others, tremor, cognitive issues affecting interactions, and mood changes.

Emotional/psychological

Nearly three-quarters of participants (73%) reported altered self-concept— i.e., viewing oneself as less capable, less competent, and less healthy, translating to decreased sense of well-being. Many also described preoccupation with PD (60%) and hypervigilance towards monitoring for symptoms and impacts, or fear of the future and inevitable disease progression (53%).

Approximately half (55%) reported embarrassment, most commonly with respect to tremor (47.5%) or other socially apparent symptoms (e.g., gait disturbances, flat affect) which sometimes led to feeling stigmatized (38%).

Global

In all, over 90% of participants reported a sense that PD limited what they could do, with many indicating doing things in general required substantially more effort than prior to having PD. A large number (40%) also reported increased personal discomfort ranging from pain to being unable to get comfortable and relax due to stiffness, cramping, or tremor.

Themes

Three key themes were identified with regards to what made symptoms meaningful from the perspective of people with early PD. Meaningful symptoms were those that were 1) personally important— whether actively present or not, with symptoms becoming more meaningful as 2) bothersomeness of an experienced symptom increased and 3) the symptom became more limiting. These themes are described below and supporting data are presented in Supplement F.

Theme 1. Symptoms can be important even when not present or present but not important

Personally important symptoms fell into two categories: 1) symptoms currently experienced that impacted physical and psychosocial functioning, and 2) symptoms not currently experienced that had potential to impact future physical and psychosocial functioning. For example, many people felt speaking and cognitive symptoms were very important (i.e., “staying me” [P28]) and they actively monitored for onset of cognitive and speech difficulties or engaged in activities to strengthen speech and cognition, even though currently without symptoms in that area. As one woman explained:

Theme 2. Symptoms can be bothersome even when not limiting or limiting but not bothersome.

Similar to theme 2, symptoms did not have to be limiting in order to be perceived as bothersome. For example, tremor was often perceived as bothersome, even though it did not actually limit ability to do things. For example:

The more limiting a symptom was, the more bothersome it was generally perceived to be, as seen here:

However, in some instances, symptoms caused limitations that were not viewed as bothersome. One individual who experienced substantially slower walking speed explained it this way:

Theme 3. Most bothersome symptoms are those which are limiting and have the greatest current negative impact on an individual’s sense of well-being and usual activities.

When discussing personally meaningful symptoms, individuals prioritized aspects of bothersomeness based upon a hierarchy of needs similar to that described by Maslow [27]: 1) physiologic needs (e.g., eating, breathing, sleep, pain/injury prevention, and ability to perform ADLs that meet these basic needs), 2) safety & security needs (symptoms affecting security, including employment or fear of the future, anxiety), 3) love and belonging needs (symptoms adversely impacting ability to communicate with others or interpersonal relationships), and 4) self-esteem needs (symptoms causing social embarrassment or affecting sense of oneself as a healthy competent person). When higher risk symptoms were not present, lower risk symptoms were prioritized instead.

For example, choking (Maslow level 1) was a very bothersome symptom, as were other symptoms that threatened personal safety, as shown in this comment:

Similarly, tremor was less bothersome when lacking direct impact and more bothersome when it caused pain or discomfort (Maslow level 1), interfered with work (Maslow level 2), or was apparent in social situations, resulting in a sense of stigma and embarrassment (Maslow levels 3-4). When individuals were able to mitigate the impact of the symptom or find workarounds (i.e., protecting hierarchy of needs), the symptom was viewed as less bothersome. For instance, loss of fine motor skills and slower movements were more bothersome when they affected the individual’s job and less bothersome after retirement when extensive computer work was not required. Similarly, when symptoms improved with medication use and became less limiting, they were viewed as less bothersome but still important.

Limitations and future directions

This study was conducted with people who were recruited from a prior study investigating the use of smartphones/smartwatches for PD symptom monitoring. Participants were predominantly higher socio-economic status, higher health/technologicallyliterate, white individuals, with qualitative data collected at a single time-point. While inclusion/exclusion criteria were compatible, the sampling approach affects generalizability and findings reported here may not reflect experiences of individuals from underrepresented groups or those with lower technological and health literacy. Sample size was also small and no significant differences in symptoms/impacts were seen between those taking and not taking PD medications. Lastly, in addition to a need for replication with larger and more demographically diverse samples, longitudinal data will be needed to support understanding of how meaningful symptoms change over time. Our data tentatively suggested that as life contexts change (e.g., transitions from working to retirement, change in living situations, or duration of disease) symptoms that are “most” bothersome may also change, which has been suggested elsewhere [5]. Thus, reevaluation of meaningful symptoms over time with the goal of developing population-based models to predict trends in PD symptom progression are warranted to guide long-term therapeutic objectives.

Conclusion

The findings and approaches described in this study can support rigorous, systematic identification and grading of meaningful symptoms and impacts of early PD, which is critical to selection of valid patient-centered endpoints for therapeutic trials. We believe the conceptual model and categorical classification of meaningfulness proposed here will be broadly relevant. Future work is needed to determine the extent to which this classification system can support interpretable evaluation of different outcomes assessments relevant to patients.