Ecological Momentary Assessment of Voice & Psychological Factors: Group & Individual Mechanisms

Cross-sectional associations between voice and psychological factors are known, but changes over time offer opportunities to refine our understanding of their interactions and consider customized treatment options. Study objectives were to measure relationships between voice and psychological factors using ecological momentary assessment and applying (1) group-level time series analysis and (2) group and (3) individual causal modeling to identify key psychological factors relevant for voice outcomes.

Adults (N = 32) with primary muscle tension dysphonia completed multiple assessments daily for 10 days. Measures included items from the Voice Handicap Index-10, voice-adapted perceived present control scale, items from NIH PROMIS and the NIH Toolkit to assess distress, and the Positive and Negative Affect Scale. Group-level time series analysis was conducted using dynamic structural equation modeling; causal analysis utilized the Greedy Fast Causal Inference algorithm.

In group-level time series analyses, neither perceived control nor distress predicted subsequent timepoint voice handicap scores. In group-level causal modeling, anxiety was causal for voice handicap, but perceived control was not. Individual-level analyses identified various causal factors for voice handicap including perceived control and negative affect, and to a lesser extent, serenity, anxiety, somatic arousal, and stress.

Group-level analyses may obscure important heterogeneity that is identifiable using individual-level causal analyses. For example, perceived control was not identified as predictive or causal for voice handicap at the group level; but was a salient causal factor for voice handicap in some individuals. Causal modeling using intensive longitudinal datasets offers a potential avenue for individualized treatment approaches.