Abstract
Objective: To determine the risk, predictors, and prognosis of independence loss and institutionalization in patients with early Parkinson disease (PD).
Methods: We conducted a prospective population-based 5-year longitudinal study following 189 patients with incident PD and 174 controls matched for age, sex, and comorbidity. Health care status was assessed repeatedly with standardized interviews.
Results: More newly diagnosed patients with PD (15.9%) than controls (5.7%) were dependent in activities of daily living at baseline (relative risk [RR] 2.8, p = 0.004). During follow-up, 40.9% of the initially independent patients lost their independence vs 9.1% of controls (RR 4.5, p < 0.001). Higher age, shorter symptom duration, increasing motor severity, and presence of mild cognitive impairment at PD diagnosis independently predicted independence loss during follow-up. Dependency was irreversible in most (>95%) patients. Long-term care facility placement was needed in 8.8% of patients vs 0.6% of controls (RR 15.4, p = 0.001). More patients with PD admitted to long-term care facilities were fallers (RR 4.8, p < 0.001), had hallucinations (RR 4.4, p = 0.001), or had dementia (RR 4.2, p < 0.001) than home-dwelling patients. Once admitted to a long-term care facility, the age-adjusted RR for death during the study period was 5.5 (p = 0.002) vs patients never admitted and 25.1 (p < 0.001) vs controls.
Conclusions: Patients with early PD face a substantially greater risk of independence loss and institutionalization than well-matched controls. Independence loss is irreversible in most patients and represents a sinister prognostic factor in early PD. These findings have implications for patient management and health care planning.
GLOSSARY
- CCI=
- Charlson Comorbidity Index;
- CI=
- confidence interval;
- H&Y=
- Hoehn & Yahr;
- HR=
- hazard ratio;
- LED=
- levodopa equivalent dose;
- MADRS=
- Montgomery-Åsberg Depression Rating Scale;
- MCI=
- mild cognitive impairment;
- MMSE=
- Mini-Mental State Examination;
- PD=
- Parkinson disease;
- RR=
- relative risk;
- UPDRS=
- Unified Parkinson’s Disease Rating Scale
Functional independence is crucial for people's well-being and quality of life.1 Therefore, independence loss is increasingly recognized as an important event in neurodegenerative disorders such as Parkinson disease (PD).2 The level of dependence may range from mild functional impairment possible to overcome with help from relatives to more severe disability requiring assistance from professionals at home or in institutions. Hence, independence loss may not only result in major distress to patients and their relatives but also become a substantial burden to the health care system. This highlights the need for valid information on the occurrence, risk factors, and prognosis of dependency in PD.
Loss of independence has been studied previously in PD, but mainly in cross-sectional cohorts and patients with advanced PD recruited at hospitals or specialist clinics.3,–,9 A recent systematic review3 identified few longitudinal studies in this field, especially in early PD, and concluded that their methodologic quality in general was low. Notably, most studies gathered information on dependency from disability scales, which are not sensitive to detect independence loss,10 rather than directly by standardized interviews. Furthermore, we are unaware of population-representative studies, or studies that included a control group, which would be important to estimate the true frequency and relative risk of dependency in PD compared to the general population.
To close this knowledge gap, we prospectively assessed the frequency, risk factors, and prognosis of independence loss and institutionalization over 5 years in a large, population-based, and clinically well-characterized incident PD cohort and well-matched normal controls.
METHODS
Study design and participants.
All patients with PD participate in the Norwegian ParkWest project, a prospective, population-based, multicenter, longitudinal cohort study investigating the incidence, neurobiology, and prognosis of PD. The study design and multiple recruitment strategies to establish a population-based incident PD cohort have been described previously.11 After recruitment from the general population, a study neurologist with experience in movement disorders performed a screening visit. Patients with suspected incident PD underwent an extensive examination program at baseline, while still being drug-naive. Following baseline examinations, a study neurologist initiated and adjusted antiparkinsonian treatment according to best clinical judgement. We recorded type and dosage of all medications at each study visit, and calculated levodopa equivalent doses (LED) in line with current recommendations.12 Study neurologists also conducted standardized clinical follow-up visits every 6 months, with a more comprehensive examination program performed after 1, 3, and 5 years of follow-up. Of 212 suspected patients with PD initially included, 5 were not drug-naive at baseline and a further 18 were rediagnosed during follow-up. Thus, 189 patients were eligible for this study. All met widely acknowledged diagnostic criteria for PD13,14 at latest follow-up.
For comparison, we recruited 175 age- and sex-matched controls in the same geographical area from multiple sources, including friends and spouses of patients and members of public organizations for the elderly. Controls were investigated at baseline and after 1, 3, and 5 years of follow-up with standardized examinations identical to those performed in patients, except for PD-specific procedures. One control developed incident PD during follow-up and was excluded, leaving 174 controls eligible for this study.
Participants unable or not willing to attend the hospital were offered examinations at their residence to minimize attrition bias.
Standard protocol approvals, registrations, and consents.
The Regional Committee for Medical and Health Research Ethics, Western Norway, approved the study. We obtained signed written consent from all participants.
Assessments.
Baseline risk factors.
In both patients and controls, a study neurologist conducted medical and neurologic examinations and semi-structured interviews to obtain information on medical and drug history, as well as occupational status. We scored comorbid conditions according to the Charlson Comorbidity Index (CCI),15 a widely used, valid, and reliable16 measure of comorbidity. We examined global cognitive function using the Mini-Mental State Examination (MMSE)17 and depressive symptoms by the Montgomery-Åsberg Depression Rating Scale (MADRS).18 In patients with PD, additional assessments included the Unified Parkinson’s Disease Rating Scale (UPDRS)19 motor section (part III) to evaluate severity of parkinsonism and the modified Hoehn & Yahr (H&Y) scale20 to rate disease stage. Furthermore, evaluation of mild cognitive impairment in PD (PD-MCI) was performed according to established criteria21 by neurologists experienced in both movement disorders and dementias, as described previously.22
Milestones.
Clinicopathologic studies suggest a temporal clustering of events comprising falls, visual hallucinations, dementia, and residential care.23 These milestones have been proposed to designate the transition into advanced stages of PD, and were also assessed among the patients in our study. We defined falls according to the UPDRS item 13 (falling), with a score ≥1 defining presence of falls. We measured psychotic symptoms using UPDRS item 2 (thought disorder) with a score ≥2 defining psychosis (visual hallucinations or delusions), identical to previous work.24 Dementia evaluation was conducted by 2 study neurologists experienced in both movement disorders and dementias, according to established criteria,25 based on all available information, including assessment of behavioral features, neuropsychological tests, and standardized interviews with patients and caregivers, as described previously.26
Loss of independence.
We investigated dependency status at each study visit using standardized interviews addressing living location (at home or in intermittent or continuous facility care) and, if living at home, any source (e.g., friends, relatives, cleaning or food delivery personnel, community nurses) and type (e.g., administration of medication, dressing, personal hygiene, eating, house chores, general supervision) of help received. We defined loss of independence as receiving any help in activities of daily living, regardless of source, location, and type. Home care included help from any source, including community nursing services, in the patient's home, whereas facility care was defined as intermittent or long-term placement in a care facility. Some patients received assistance from multiple sources. In these instances, the assistance level reflecting the highest degree of disability at each visit was recorded for further analysis in order to avoid overestimating cumulative figures.
Statistical analysis.
We used the IBM (Armonk, NY, USA) Statistical Package for Social Sciences version 22.0 for statistical analysis. We used t tests or Mann-Whitney tests to compare between-group differences in continuous data and Pearson χ2 tests or Fisher exact tests for comparison of dichotomous data, and calculated relative risks (RR) with 95% confidence intervals (CI). We applied Kaplan-Meier survival analyses to calculate the cumulative incidence of independence loss, home care, and facility care. We used Cox logistic regression analyses to evaluate the effect of baseline demographic, clinical, and neuropsychiatric factors on loss of independence, considering the following independent variables: age, sex, motor symptom duration, CCI score, UPDRS motor score, MADRS score, MMSE score, and presence or absence of PD-MCI. We considered 2-tailed p values <0.05 significant.
RESULTS
Baseline characteristics.
Baseline characteristics are given in table 1. Of the 189 patients with PD included, 30 (15.9%) were not independent at baseline compared to 10 of 174 (5.7%) controls (RR 2.8 [95% CI 1.4–5.5], p = 0.004). Patients with PD who were dependent at baseline were older (74.3 vs 66.5 years, p < 0.001) and had more severe parkinsonism (UPDRS motor score 33.6 vs 21.4, p < 0.001), more advanced disease stage (H&Y stage 2.4 vs 1.8, p < 0.001), lower global cognitive function (MMSE score 26.2 vs 28.0, p = 0.021; proportion with PD-MCI 50.0% vs 17.0%, p < 0.001), and a higher degree of comorbidity (CCI 0.5 vs 0.25, p = 0.006) compared to the 159 patients who were independent at baseline. Among the 159 patients and 164 controls who were independent at baseline, there were, as expected, minor differences in MMSE and MADRS scores, but not age, sex, comorbidity, or occupational status.
Baseline characteristics of patients with Parkinson disease (PD) and controls
Study flow.
The flow of participants during follow-up is provided in figure 1. Of the 159 patients who were independent at baseline, 156 (98.1%) completed 1 year, 154 (96.9%) 3 years, and 141 (88.7%) 5 years of follow-up. Corresponding figures for controls were 155 (94.5%) at 1 year, 145 (88.4%) at 3 years, and 137 (83.5%) at 5 years. The main reason for loss to follow-up of patients with PD was death (n = 13/18), whereas withdrawal of consent was the most frequent reason among controls (n = 24/27). Controls who withdrew consent did not differ from those who remained in the study in terms of age, sex, cognitive function, depressive symptoms, or comorbidity. Follow-up yielded 1,701 observations in patients, 611 in controls, and 2,312 overall.
Patients were assessed at 6-month intervals; flowchart is simplified for readability. PD = Parkinson disease.
Treatment.
Patients with PD received antiparkinson treatment according to their study neurologist's best clinical judgement. LED increased progressively during the study, starting at none at baseline, increasing to a mean (SD) of 248.2 (165.0) mg after 1 year, 419.8 (227.6) mg after 3 years, and 586.3 (331.7) mg after 5 years of follow-up. No patients received advanced treatments, i.e., apomorphine, intrajejunal levodopa, or subthalamic nucleus deep brain stimulation during the study.
Loss of independence in PD.
Among patients who were independent at baseline, we observed a steady increase in dependency during follow-up (table 2, figure 2). The cumulative incidence of independence loss reached 6.3% (n = 10) 1 year, 25.8% (n = 41) 3 years, and 40.9% (n = 65) 5 years after diagnosis. The corresponding numbers for facility care admission were 0% at 1 year, 3.1% (n = 5) at 3 years, and 12.6% (n = 20) at 5 years of follow-up. Hence, 30.8% of those who lost independence in any activities of daily living were later admitted to facility care during follow-up, after a median of 1.99 (interquartile range 0.43–2.91) years following independence loss.
Point prevalence and cumulative incidence rates of independence loss among patients with Parkinson disease (PD) and controls during follow-up
Proportion of patients with Parkinson disease (blue) and controls (green) remaining independent through the study.
Relative risk of independence loss.
Compared to controls, the RR for overall incident independence loss in patients with PD was 2.1 (0.7–5.9, p = 0.18) after 1 year, 4.2 (2.2–8.2, p < 0.001) after 3 years, and 4.5 (2.7–7.5, p < 0.001) after 5 years of follow-up (table 2). The corresponding figures for needing home care were 2.1 (0.7–5.9, p = 0.19) after 1 year, 4.1 (2.1–8.0, p < 0.001) after 3 years, and 4.2 (2.5–7.1, p < 0.001) after 5 years. There was no difference in facility care admissions after 1 or 3 years of follow-up. After 5 years, however, 12.6% of patients vs 0.6% of controls had been admitted to an intermittent or long-term care facility, representing an RR of 20.6 (2.8–151.9, p = 0.003). When only considering long-term facility care admissions, the RR after 5 years of follow-up was 15.4 (2.1–115.8, p = 0.001).
Risk factor analysis.
In Cox proportional hazard logistic regression models, higher age (hazard ratio [HR] 1.07 [1.03–1.11] per year, p = 0.001), higher motor severity (HR 1.04 [1.01–1.07] per UPDRS motor score unit, p = 0.013), shorter motor symptom duration (HR 0.82 [0.69–0.98] per year, p = 0.031), and presence of PD-MCI (HR 1.81 [1.01–3.26], p = 0.048) were independent baseline predictors of independence loss during follow-up, whereas sex, comorbidity score, and depressive symptoms were not. For long-term facility care admissions, living alone was an important associated factor, resulting in an RR of 3.1 (1.1–8.4, p = 0.03); 8 of 14 patients were living alone when admitted (median 3.8 years after baseline), whereas 37 of the 135 nonadmitted patients lived alone at this time point.
Milestones.
At the time of long-term facility care admission, 7 of the 14 admitted patients had dementia, compared to only 16 of the 135 nonadmitted patients at this time point, yielding an RR of 4.2 (2.1–8.5, p < 0.001). Psychotic symptoms (5 of 14 vs 11 of 135, RR 4.4 [1.8–10.8], p = 0.001) and falls (10 of 14 vs 20 of 135, RR 4.8 [2.9–8.1], p < 0.001) were also more common in patients in need of long-term care facility placement. Twelve of the 14 patients (86%) admitted to facility care had one or more of these additional milestone conditions23 (dementia, hallucinations, or falls) at the time of admission, and all 14 patients contracted at least one additional milestone during follow-up.
Prognosis after independence loss.
Of the 65 patients who once lost independence, only 3 (4.6%) regained independence, whereas 53 (81.6%) remained dependent and 9 (13.8%) died during follow-up. The age-adjusted RR for death during the 5-year follow-up in patients who had lost independence was 1.9 (0.5–8.4, p = 0.38) compared to patients who remained independent, and 5.1 (0.96–27.1, p = 0.06) compared to the control group.
Of the 14 patients who were admitted to a long-term care facility, 10 (71.4%) remained in facility care and 4 (28.6%) died (after a median of 193 days from admission), whereas none regained independence. Mortality among patients under long-term facility care was increased, with age-adjusted RRs of 5.5 (1.9–16.0), p = 0.002, compared to patients living at home and 25.1 (5.1–124.3, p < 0.001) compared to controls (figure 3).
Survival of patients with Parkinson disease admitted to long-term care facilities (blue) vs home-dwelling patients (black) and controls (green).
DISCUSSION
This prospective population-based 5-year longitudinal study demonstrates a more than 4-fold increased risk of independence loss and 15-fold increased risk of long-term care facility placement in early PD compared to well-matched controls. In absolute numbers, more than 15% of patients were functionally dependent at diagnosis and 40% of the remainders lost their independence during follow-up. Once independence was lost, the chances of returning to an independent life were very low, and the risk of care facility admission during the 5-year study period was high, exceeding 30%. Prognosis after long-term care facility admission was grim, with one-third dying before study end and the rest remaining under facility care. These findings are important for patient management and health care planning.
Loss of independence is an important life event and a significant determinant of quality of life1 and caregiver burden.27 Despite this, there is a considerable uncertainty about the frequency, predictors, and prognosis of independence loss in PD. A recent systematic review of longitudinal studies in this field estimated that 10%–25% of surviving patients are dependent 5 years after diagnosis.3 Our findings suggest that these figures are substantial underestimates of the true rates of independence loss in PD. There are several potential explanations for this discrepancy, such as selection, attrition, and survival bias in previous cohorts, which we sought to minimize in our study using a population-based design, recruitment of an incident PD cohort, and frequent assessments during prospective follow-up. In addition, we assessed dependency status thoroughly using standardized clinical interviews rather than indirectly by disability scales, which are not sensitive to detect independence loss.10 Overall, we found that more than half of patients lose their independence within 5 years of diagnosis, which is at least twice as high as previously estimated. In addition, the RR of incident long-term care facility admission compared to well-matched controls exceeded 15 in our cohort. This is more than sevenfold higher than reported among patients with more advanced PD.28
Considering the high risk and dire consequences of independence loss in PD, early identification of those most likely to become dependent would be important. We found that, in addition to advanced age and shorter disease duration, higher motor disability and presence of PD-MCI at diagnosis predicted independence loss during follow-up. This suggests that early functional deficits in activities of daily living are caused by both motor and nonmotor symptoms in PD. It is well-known that there is an interrelation between motor and cognitive impairment in PD that becomes more prominent with disease progression. This is particularly true for motor symptoms that are poorly responsive to dopaminergic treatment, such as postural instability and gait problems, which are known to increase the risk of the disease milestones23 falls,29 psychosis,30 and dementia.31 These milestones have previously been identified as important risk factors for institutionalization in advanced PD,6,9 including studies analyzing decisions to institutionalize an affected family member.32,33 Our study extends these observations to be true also in early PD, as almost 90% of patients in need of long-term facility care had developed one of these milestones at time of admission, with a substantially higher prevalence compared to home-dwelling patients. These prospective data also support and extend evidence from clinicopathologic studies23 indicating a close relationship between these events as the disease enters a more advanced phase.
An important observation in our study was that loss of independence was irreversible in most patients. Indeed, fewer than 5% of those who became dependent subsequently regained independence during follow-up. This reversal rate is much lower than, for example, that reported for PD-MCI,22 which is increasingly recognized as an important prognostic marker in PD. In addition, we observed a high risk of institutionalization among those who lost independence. Combined, these findings (persistence and progression) strongly suggest that loss of independence is an early and highly predictive compound clinical marker of an unfavorable disease course in PD resulting from both motor and cognitive deterioration. This highlights the potential of independence loss as a useful outcome measure in studies of disease progression as well as clinical trials.
Our study differs from most previous studies in several important ways. First, it was an incident cohort study, and we aimed to include all newly diagnosed patients with PD in a specified geographical area during a predetermined time period. Second, while the study was community-based, patients were treated and followed by movement disorders neurologists, who provided expert confirmation of the diagnosis using widely accepted diagnostic criteria.13,14 Third, patients were monitored closely from time of diagnosis, and few were lost to follow-up for other reasons than death. Fourth, we recruited a control group, allowing us to provide risk estimates relative to nondiseased individuals matched for age, sex, and comorbidity. Finally, we prospectively assessed dependency status using clinical interviews rather than indirectly by disability scales. The less frequent follow-up visits in controls compared to patients with PD, although a potential limitation of this study, did not appear to be a sizable source of bias, as point prevalence rates and cumulative incidence estimates were comparable throughout follow-up. Therefore, our results are as valid and representative for the general PD population as possible.
Our finding that more than half of patients became dependent within 5 years of diagnosis may be considered to challenge the current opinion that PD has a benign early phase.34 This view is further substantiated by our observation that even minor functional independence loss appears to be persistent, with a high risk of progression to institutionalization within few years. Together, these observations may justify the addition of independence loss to the series of the more established clinical milestones in PD. We suggest that future research should focus on identification of biomarkers for independence loss, as dissection of the biological substrates of dependency would be important to reveal the mechanisms responsible for the accelerated neurodegenerative process that probably underlies the earlier entry into the more advanced disease stages.
AUTHOR CONTRIBUTIONS
Anders Bjornestad: study concept and design, analysis and interpretation of data, writing of the first draft. Ole-Bjørn Tysnes: study concept and design, acquisition of data, critical revision of the manuscript for important intellectual content, study supervision. Jan Petter Larsen: study concept and design, critical revision of the manuscript for important intellectual content, study supervision. Guido Alves: study concept and design, acquisition of data, analysis and interpretation of data, critical revision of the manuscript for important intellectual content.
STUDY FUNDING
This study was supported by the Research Council of Norway (grant 177966), the Western Norway Regional Health Authority (grants 911218 and 911949), and the Norwegian Parkinson Disease Association. The sponsors were not involved in the study design, in the collection, analysis, and interpretation of data, in the writing of the report, or in the decision to submit the article for publication.
DISCLOSURE
A. Bjornestad reports research support from the Western Norway Regional Health Authority (grant 911949). O. Tysnes and J. Larsen report no disclosures relevant to the manuscript. G. Alves reports personal fees from AbbVie and grants from the Norwegian Parkinson Disease Association outside the submitted work. Go to Neurology.org for full disclosures.
ACKNOWLEDGMENT
The authors thank the patients for their participation and the members of the Norwegian ParkWest study group and all other personnel involved in this study for their contributions.
Footnotes
Go to Neurology.org for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the article.
- Received April 29, 2016.
- Accepted in final form June 24, 2016.
- © 2016 American Academy of Neurology
REFERENCES
Disputes & Debates: Rapid online correspondence
NOTE: All authors' disclosures must be entered and current in our database before comments can be posted. Enter and update disclosures at http://submit.neurology.org. Exception: replies to comments concerning an article you originally authored do not require updated disclosures.
- Stay timely. Submit only on articles published within the last 8 weeks.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- 200 words maximum.
- 5 references maximum. Reference 1 must be the article on which you are commenting.
- 5 authors maximum. Exception: replies can include all original authors of the article.
- Submitted comments are subject to editing and editor review prior to posting.