The impacT of associaTed large-fiber peripheral neuropaThy on healTh-relaTed qualiTy of life in parkinson’s disease – resulTs from a romanian cohorT

background. Recent studies described a higher prevalence of peripheral neuropathy (PN) in Parkinson’s disease that was linked to L-Dopa exposure. Peripheral neuropathies are known causes of a decreased health-related quality of life (HrQoL). Until now, no studies addressed the issue of how or if associated PN in PD affects HrQoL. Methods. In a cross-sectional, observational study, 73 non-demented PD patients, from which 36 with confirmed PN based on clinical (using the Toronto Clinical Neuropathy Scale-TCSS) and nerve conduction studies completed the Romanian version of PDQ-39. results. Significant differences between mean scores in Motor (49.86 (27.61) vs. 31.50 (26.24), p = 0.005), Activities of daily living (49.86 (27.61) vs. 31.75 (30.10), p = 0.003) and body discomfort (52.54 (29.54) vs. 23.64 (18.48), p = 0.002) domains of PDQ-39 in the PN-PD group versus non-PN group were observed. TCSS significantly correlated to motor, emotional well-being and body discomfort domains (r = 0.406 p< 0.001; r = 0.316 p = 0.007; r = 0.356 p = 0.002, respectively). The multivariate linear regression model showed that motor impairment and PN correlated to motor domain (beta = 0.601, p = 0.000; beta = 0.211 p = 0.041, respectively) and PN significantly correlated to body discomfort domain (beta = 0.314, p = 0.020) of PDQ-39. conclusions. The presence of associated PN in PD determines a further deterioration of HrQol in subjects with already a poorer HrQol.


IntroductIon
Parkinson disease (PD), a prevalent progressive neurodegenerative disorder, causes motor disability as well as non-motor impairment and determines a significant impact on the quality of life (QoL) (1)(2)(3).
Health-related quality of life (HrQoL) in PD is essential in everyday practice because it reflects how a patient perceives the burden of the disease on his daily life and may predict the adherence to treatment or the number of presentation in the clin-ic. HrQoL in PD, until recently, was mostly related to the motor impairment of the central nervous system, primarily to motor fluctuations (2,3,22,23), but in the past decades, several papers highlighted that non-motor symptoms in PD determined a worsening in QoL (1,(24)(25)(26). Psychiatric symptoms are, by far, the non-motor symptoms that have the most significant impact on QoL (26)(27)(28)(29)(30)(31), but a recent study showed that the cumulative number of non-motor symptoms in PD is the predictor of a worse Qol in PD (32). Pain in PD is a well known non-motor symptom that can affect QoL (33,34). There are few reports on how peripheral neuropathic pain and sensory neuropathic symptoms influence self-reported QoL in PD (35,36).
Chronic polyneuropathy, regardless of its aetiology, was found to affect all aspects of HrQoL when assessed by the Short Form-36 questionnaire and compared to healthy controls (37). Painful neuropathies were associated with poor HrQoL (38,39).
Several specific questionnaires, like Parkinson Disease Questionnaire -39 (PDQ-39) and it's Short 8 item form, or Parkinson's Disease Quality of life Questionnaire (PDQL) are recommended for self-assessment of HrQoL in PD (40). Parkinson's Disease Questionnaire 39 and PDQ-39 Summary Index are a validated tool that evaluates HrQol in 8 domains -mobility, activities of daily living, emotional well-being, stigma, social support, cognition, communication and body discomfort (41).
Whatsoever, so far, there are no reports which inquired about QoL in PD patients with associated confirmed PN based, both on clinical and nerve conduction studies criteria. There is a need for reports on how associated PN in PD affects HrQoL in those patients with an already diminished quality of life due to PD itself.

MAterIAls And Methods
This study aimed to evaluate if confirmed PN can affect QoL in PD patients and to what extent.
in an observational, cross-sectional study, we included patients with PD according to UK Parkinson's Disease Brain Bank (42). The subjects were admitted to the Neurology I Department, a general neurology department, of County Emergency Clinical Hospital in Cluj-Napoca from January 2017 to June 2019.
Patients with other known causes of PN, plexopathies, radiculopathies, dementia, or acute medical conditions like stroke, heart attack or infections were excluded.
All patients completed the Romanian version of the PDQ-39 questionnaire in "on" state. The scores for each domain and the PDQ-39 summary Index were calculated accordingly (43). The PDQ-39 scores range from 0 to 100, where the higher scores indicate a worse HrQoL.
relevant data about PD duration, l-Dopa daily dose and psychiatric comorbidities were taken from charts. Neurological examination was made using the UPDRS scale part II -activities of daily living, part III -motor and part IV. Disease severity was assessed on the Hoehn and Yahr scale. Toronto Clinical Neuropathy Scale (TCSS) was used for clinical evaluation of the peripheral nervous system (44). Cognitive functions were evaluated using the universally used Mini-Mental state examination(MMSE) scale (45).
Al patients underwent electroneurographic evaluation. The full protocol for nerve conduction studies, assessment of the biological parameters and confirmed PN diagnosis, in the cohort were fully detailed in a previous study (46).
All included subjects gave their informed consent. This study was carried under the Helsinki Declaration and received approbation from the local ethics committee.

stAtIstIcAl AnAlyses
Statistical analysis was performed using SPSS 20. Descriptive statistics for continuous variables (e.g. age) were presented as mean and standard deviation. Descriptive statistics for categorical variables (e.g. gender) were presented as counts and proportions, and for statistical comparisons of proportions between groups, the Chi-square test was used. Even if the subgroups compared in the first part of the Results section are rather small, the main analysis presented here is parametrical (Independent samples t-tests) since statistical literature proved the robustness of parametrical procedures against non-normality of distributions (47). Anyway, to prove this robustness, the non-parametric version of this analysis (the Mann-Whitney test) is also presented. To overpass the possibility of false-positive results due to multiple comparisons, we also analysed the effect size for each comparison (Cohen's d) and decided upon the practical relevance of each result. Correlations between variables were assessed using the Pearson correlation coefficient, followed by multiple linear regression models, in order to have a more comprehensive picture about the predictors of the PDQ-39 domains set. The analysis of scatterplots for the bivariate relationship between each predictor and the criterion variable suggested that a linear analysis was appropriate. For all inferential analysis, a two-sided p-value < 0.05 was considered statistically significant.

results hrQol in Pn versus non-Pn groups
Seventy-three PD patients (49.3% females, mean age 65.44 (9.87)) from the previous study were included (46). Two patients scored below 24 points on MMse but due to motor impairment and 10 patients received medication for anxiety or depression. Those patients were still included. The entire group was separated into 2 subgroups -PN and non-PN. The PN group included 36 subjects with associated confirmed PN, evaluated in the previous study. The non-PN group included 37 subjects without PN. The PN group differed significantly from the non-PN regarding disease duration (p < 0.001), age (p = 0.051), L-Dopa daily dose (p 0.023), MMSE score (p = 0.003 and UPDRS activities of daily living (ADL) score, motor and part iV scores(p = 0.003, p = 0.007, p= 0.007, respec-tively). The non-parametric analysis confirmed the parametric results. Effect sizes varied from moderate to large. Demographic and clinical characteristics of the 2 groups are summarized in Table 1. PN patients experienced a significant worse QoL compared to non-PNP in all domains, except stigma and social support (34.43 (30.79) vs. 25.34 (26.43), p = 0.180; 20.54 (26.38) vs. 14.30 (19.42), p = 0.253), respectively. Significant differences between mean scores in motor, aDl and body discomfort domains of PDQ 39 in the 2 groups were observed. The non-parametric analysis confirmed the parametric results. Effect sizes varied from small (the case of non-significant differences) to moderate and large (the case of significant differences) ( Table 2).

Multivariate linear regression analyses
A multivariate linear regression model with the PDQ-39 domains and PDQ39-SI set as dependent variables and age, TCSS, UPDRS Motor score (part III) and MMSE as predictors was run. After adjusting for those variables, PN -as assessed by TCSS, still significantly correlated to the body dis-

dIscussIon
Patients in the PN group were older, in more advanced stages, with longer PD duration and scored higher in the UPDRS scale, than those in the non-PN group. PN in this cohort was previously shown to be related to L-Dopa intake; possible causes and relations of PN to PD in this cohort were previously discussed and PN was linked to vitamin B12 and folate deficiencies and high daily doses of L-Dopa (46). Nevertheless, PN in PD seems to be more common in advanced stages of PD and linked to higher doses of L-Dopa and might represent a more severe phenotype of PD (49).
This study included patients with confirmed PN of no other causes, excluding patients with plexopathy or radiculopathy; thus, these results depict the impact of signs and symptoms of PN on HrQoL, rather than focusing on specific sensory symptoms or neuropathic pain that are already predictors for a poor HrQoL (35,36). To the best of our knowledge, this is the first study to evaluate HrQol of PD patients with associated large fiber neuropathy.  PDQ-39 scores and PDQ-39-SI were significantly higher in the PN group than in the non-PN group, demonstrating that subjects in the PN group experienced a worse HrQoL compared to non-PNP in all domains, except stigma and social support.
These results highlighted that the associated impairment of peripheral nervous system had a significant impact on HrQoL in PD patients in the motor, activities of daily living, emotional well-being and body discomfort domains of HrQoL, as well as overall.
Motor disability, as assessed by UPDRS motor score showed significant correlations to the motor domain and activities of daily living of PDQ-39, as expected because, until recently, HrQoL in PD related significantly to motor impairment (2,3). After adjusting for this score, the multivariate regression model showed that polyneuropathy contributes to the burden of motor disability in everyday life, and may determine a further deterioration of HrQoL. Taken together, motor impairment and polyneuropathy had a substantial impact on overall HrQoL in PD, as shown by the multivariate regression model.
Motor complications, as assessed by UPDRS score part IV, influenced overall HrQoL and subdomains like motor function, activities of daily living or communication. Similar results were already outlined in other studies (22,23).
The multivariate regression model showed that the body discomfort domain was only influenced by PN, suggesting that positive sensory symptoms, like paresthesia, are responsible for poorer HrQoL in this aspect.
stigma and social support domains were not influenced by motor impairment, cognitive functions as assessed by MMSE, or polyneuropathy when taken separately nor in the multivariate regression model. These results indicate that these specific domains-stigma and social support can be instead more related to an individual's personality features, education levels and family dynamics.
Caution should be taken when interpreting these results, because this study was cross-sectional, on small sample size and single centre-based. Because of the study design, these associations cannot define a causal relationship. Furthermore, we did not assess the impact of non-motor symptoms, nor did we adjust the results for the presence of anxiety or depression. Non-motor symptoms, especially neuropsychiatric ones, are known predictors for worse HrQoL in PD (26).
The modality in which a patient experiences the burden of a chronic disease like PD is individual and is multifactorial influenced, as these results also show. The association of PN in PD may determine a further deterioration of HrQoL in subjects with already a poorer HrQoL as these results also show.
So far, there are no clinical good practice recommendations for treating a PD patient with PN, nor how to evaluate PD-associated PN. We believe these issues should be addressed in further studies, and guidelines are needed to be available, for patients to benefit from high-quality medical care.

conclusIons
Large fiber neuropathy was found to be more common in the advanced stages of Parkinson's disease. The impact of an associated condition like peripheral neuropathy caused a further motor impairment, as perceived by the subjects, which had a negative influence on health-related quality of life in PD patients. Patients with large fiber neuropathy and PD experienced worse quality of life as PD pa- There is a need for large-scale, prospective studies to evaluate how exactly large fiber impairment progresses in PD and how it impacts HrQoL when adjusting for non-motor symptoms, or after treatment of PN.