Ires Ghielen

51 Replication factor analysis in neuropsychiatric PD patients subscales were exact replications, the cardiopulmonary and unsteadiness subscales showed a slightly different composition to the previously found subscales hypotension, hyperventilation, and trembling . The BAI total score and the BAI affective subscale score were significantly associated with the BDI score. This high correlation between the BAI and the BDI was also reported in the previous study [18]. This suggests a high co-occurrence of anxiety and depressive symptoms [5], as is also supported by Zhu and colleagues [4], who found that up to 70% of anxious PD patients also suffer from depression. The BAI total score and BAI affective subscale score were also significantly associated with the MoCA score. However, in the final adjusted model, this association was no longer significant due to confounding by the BDI score, explained by the high correlation between the BAI and BDI (see table S1 for Pearson’s correlations between BAI, BDI, UPDRS-III, MoCA, and age). In contrast to the previous study [18], no significant associations between somatic subscales and motor symptoms were found. The current patient sample showed more neuropsychiatric symptoms, but had a similar degree of motor dysfunction (see supplementary table S3 for the independent samples t-tests results of the comparison of the sample characteristics). The slightly different distribution of BAI-items over the somatic subscales might explain the absence of the associations with motor symptoms in the current sample. Based on the BAI total score, 77.4% of the current patient sample showed clinically relevant anxiety. However, only 52.9% of patients received a formal DSM-IV or -5 anxiety diagnosis from the assessing psychiatrist. This could be due to symptom overlap and high co-occurrence with depression, in which case a depressive disorder might be more fitting with the clinical presentation in some cases. In addition, anxiety symptoms can occur in the context of other psychiatric disorders, such as psychosis, dementia, and dopamine dysregulation syndrome. Besides neuropsychiatric comorbidity, anxiety in PD often has an atypical presentation, leading to the symptoms not fitting into the disorders listed in DSM-IV and -5 [27]. Another possible explanation is that the severity of anxiety may be overestimated by the BAI [16], since motor and autonomic symptoms of PD could have inflated the scores on this self-report instrument. Nevertheless, it must also be kept in mind that the same motor and autonomic symptoms can mask anxiety symptoms during clinical evaluation. The disentanglement of anxiety from motor symptoms in PD is both scientifically and clinically challenging [12, 16]. In our previous study [18], we found that the affective subscale was the only factor not associated with autonomic dysfunction, 3