Network analysis of symptoms in a Parkinson patient using experience sampling data: An n=1 study

Background Around 50% of Parkinson's disease patients experience motor fluctuations after long-term treatment with levodopa. These fluctuations may be accompanied by mood fluctuations. Routine cross-sectional assessments cannot capture the extent of these motor and mood fluctuations and their possible associations. Experience sampling techniques that use frequently repeated measurements of symptoms over time are able to capture such fluctuations. Based on such data, longitudinal associations between symptoms can be studied using network analysis. Aim The purpose of this study is to identify longitudinal associations between motor symptoms and mood states in a patient with Parkinson's disease. Methods A 53-year-old man with Parkinson's disease and motor fluctuations collected experience sampling data during 34 consecutive days. A set of dependent variables included tremor, rigidity, balance problems, and "on/off" state, and the mood variables anxiety, cheerful, and "down." Independent variables were the same variables assessed at the preceding measurement. Regression coefficients were calculated and presented in a network graph. Results In this patient, anxiety and cheerfulness had a central position within the symptom network. Higher anxiety was prospectively associated with increased rigidity and tremor and with feeling "down." Cheerfulness was associated with less tremor. Balance problems were not influenced by cheerfulness nor anxiety, but increased balance problems were associated with reduced cheerfulness at the next assessment. Feeling "down" did not influence self-reported motor symptom severity at the next assessment. Conclusion This n = 1 study shows that network analysis of experience sampling data may reveal longitudinal associations of self-reported motor symptoms and mood states that may have relevance for treatment strategies. (c) 2018 International Parkinson and Movement Disorder Society