Ires Ghielen

52 Chapter 3 as measured with the Scales for Outcomes in Parkinson’s disease – Autonomic dysfunctions (SCOPA-AUT), and motor dysfunction as measured with the UPDRS- III. This is of clinical importance since it could indicate that certain items of the BAI should be weighted more heavily when screening for clinically relevant anxiety in PD. Therefore, in a post-hoc analysis, we investigated whether the score on the affective subscale of the BAI alone was better in predicting an anxiety disorder diagnosis given by a psychiatrist compared to the total score of the BAI. We calculated the area under the Receiver Operating Characteristic (ROC) curve of both the BAI total score and the affective subscale of the BAI in relation to an anxiety disorder diagnosis (yes or no). The area under the ROC curve of the BAI total score was 0.77 (sd = 0.04, p < 0.001). The affective subscale of the BAI showed an area under the ROC curve of 0.75 (sd = 0.05, p < 0.001). We can thus conclude that the BAI total score and the BAI affective subscale score have similar power in predicting an anxiety disorder diagnosis given by a psychiatrist in this PD patient sample. Using only the score on the affective subscale (7 items) in the prediction of an anxiety disorder diagnosis saves time and might be more practical compared to using the total BAI (21 items). In addition, this subscale might be considered as containing psychological and non- episodic anxiety items, which suspends the discussion about how to interpret the somatic BAI-items. To evaluate the affective subscale further, comparison with other screening options for anxiety in PD is useful. The Movement Disorder Association currently does not recommend one specific anxiety screening instrument (latest published research from 2008 [28]), but does recommend the non-motor rating scale (the MDS-NMS) in which four questions about anxiety are included [29]. Two of those questions represent two items of the affective subscale of the BAI, the other two ask about panic attacks and social anxiety. Another screening tool for anxiety, the Parkinson Anxiety Scale (PAS), is a self-report questionnaire that includes items specifically for non-episodic anxiety and avoidance behavior [30]. The PAS excludes almost all somatic symptoms of anxiety, except for panic related symptoms (e.g. shortness of breath and heart palpitations). All non-episodic anxiety items of the PAS are comparable to the items that clustered together in the affective subscale. This study has some limitations. The BAI focusses mostly on episodic anxiety, i.e. symptoms of panic disorder, while non-episodic anxiety, like in generalized anxiety disorder, is also common in PD. Unfortunately, autonomic dysfunction was not measured in the current patient sample, restricting the investigation of its associations with the (subscales of the) BAI. In terms of psychiatric diagnoses, this patient sample was not assessed with a structured clinical interview, to