Ires Ghielen

43 Replication factor analysis in neuropsychiatric PD patients Introduction Anxiety disorders are present in an estimated 31% of patients with Parkinson’s Disease (PD), with generalized anxiety disorder being the most common [1, 2]. Around 45% of PD patients suffer from clinically relevant anxiety [3]. Both from clinical experience and previous research, anxiety can be linked to fluctuations in available dopamine related to timing of PD medication and is often comorbid or secondary to depression, psychosis, and cognitive decline [3-5]. Anxiety is therefore one of the most prevalent neuropsychiatric features in PD patients and can be very debilitating. In addition, its presence is one of the most significant predictors of health-related quality of life in PD [6, 7], and greatly impacts caregiver burden [8]. Anxiety can worsen motor symptoms, such as tremor and freezing [9, 10] and can create a vicious cycle, in which motor symptoms trigger anxiety or distress, which can in turn exacerbate motor symptoms. These reciprocal interactions between motor and non-motor symptoms are not only seen in clinical practice but are supported by scientific research [11-13]. Despite its large impact on patient well- being and caregiver burden, anxiety in PD is still poorly understood and evidence on the treatment of anxiety in PD patients is limited [14, 15]. Since anxiety and other PD-related symptoms are so highly intertwined, diagnosing anxiety is complicated by its overlap with motor (e.g., tremor, rigidity, freezing) and autonomic symptoms (e.g., excessive daytime sweating) [12, 16]. Self-report questionnaires, such as the Beck Anxiety Inventory (BAI), can be used to quantify anxiety symptoms [17]. Due to the aforementioned reciprocity, the interpretation of the total score on such self-report questionnaires can be complicated. Evaluation of the specific symptom dimensions (or subscales of a questionnaire) can be a useful approach [18]. One statistical technique for extracting these subscales is principal components analysis (PCA). In a previous study in PD patients [19], a factor solution of the BAI was not found by using PCA, which might be explained by the great heterogeneity of the investigated study population. In a previously conducted PCA by our research group, based on a large cohort of patients that were referred to our outpatient clinic for movement disorders, we found that the BAI encompasses one affective and four somatic factors [18]. A significant association between the symptoms of anxiety and depression was found, and severity of motor symptoms showed significant associations with the somatic (and not affective) factors of the BAI. Another study investigated the dimensionality of the BAI by principal axis factoring, and found one distinct PD motor subscale [12]. 3