Abstract
BACKGROUND: Neuropsychiatric fluctuations (NpsyF) are frequent and disabling in people with Parkinson's disease (PD). In OFF-medication, NpsyF entail minus neuropsychiatric symptoms (NPS) like anxiety, apathy, sadness, and fatigue. In ON-medication, NpsyF consist in plus NPS, such as high mood, hypomania, and hyperactivity. Accurate identification of these NpsyF is essential to optimize the overall PD management. Due to lack of punctual scales, the neuropsychiatric fluctuation scale (NFS) has been recently designed to assess NpsyF in real time. The NFS comprises 20 items with two subscores for plus and minus NPS, and a total score. OBJECTIVE: To evaluate the psychometric properties of the NFS in PD. METHODS: PD patients with motor fluctuations and healthy controls (HC) were assessed. In PD patients, the NFS was administrated in both the ON-and OFF-medication conditions, together with the movement disorders society-unified Parkinson disease rating scale parts I-IV. Depression (Beck depression scale II), apathy (Starkstein apathy scale) and non-motor fluctuations items of the Ardouin scale of behaviour in PD (ASBPD OFF and ON items) were also assessed. NFS internal structure was evaluated with principal component analysis consistency (PCA) in both medication conditions in PD patients and before emotional induction in HC. NFS internal consistency was assessed using Cronbach's alpha coefficient. NFS convergent and divergent validity was measured through correlations with BDI-II, Starktein, and ASBPD OFF and ON non motor items. Specificity was assessed comparing NFS global score between the HC and PD populations. Sensitivity was evaluated with t-student test comparing the ON-and the OFF-medication conditions for NFS global score and for minus and plus subscores. RESULTS: In total, 101 consecutive PD patients and 181 HC were included. In PD patients and HC, PCA highlighted one component that explained 32-35 and 42% of the variance, respectively. Internal consistency was good for both the NFS-plus (alpha =0.88) and NFS-minus items (alpha =0.8). The NFS showed a good specifity for PD (p < 0.0001) and a good sensitivity to the medication condition (p < 0.0001). CONCLUSION: The satisfactory properties of the NFS support its use to assess acute neuropsychiatric fluctuations in PD patients, adding to available tools.