Subjective judgments of rhythmic complexity in Parkinson's disease: Higher baseline, preserved relative ability, and modulated by tempo

Abstract

Previous research has demonstrated that people with Parkinson's disease (PD) have difficulties with the perceptual discrimination of rhythms, relative to healthy controls. It is not however clear if this applies only to simpler rhythms (a so called "beat-based" deficit), or if it is a more generalized deficit that also applies to more complex rhythms. Further insight into how people with PD process and perceive rhythm can refine our understanding of the well known problems of temporal processing in the disease. In this study, we wanted to move beyond simple/complex-dichotomy in previous studies, and further investigate the effect of tempo on the perception of musical rhythms. To this end, we constructed ten musical rhythms with a varied degree of complexity across three different tempi. Nineteen people with PD and 19 healthy controls part-took in an internet based listening survey and rated 10 different musical rhythms for complexity and likeability. In what we believe is the first study to do so, we asked for the participants subjective ratings of individual rhythms and not their capacity to directly compare or discriminate between them. We found an overall between-group difference in complexity judgments that was modulated by tempo, but not level of complexity. People with PD rated all rhythms as more complex across tempi, with significant group differences in complexity ratings at 120 and 150bpm, but not at 90bpm. Our analysis found a uniform elevated baseline for complexity judgments in the PD-group, and a strong association between the two groups' rank-ordering the rhythms for complexity. This indicates a preserved ability to discriminate between relative levels of complexity. Finally, the two groups did not significantly differ in their subjective scoring of likeability, demonstrating a dissimilarity between judgment of complexity and judgment of likeability between the two groups. This indicates different cognitive operations for the two types of judgment, and we speculate that Parkinson's disease affects judgment of complexity but not judgment of likeability.

Fig 1. Between-groups differences for complexity and likeability ratings across all rhythms.

Two-sided t-tests for between-group differences for overall complexity and likeability ratings (1.a), between-group differences for overall complexity (1.b) and likeability ratings (1.c) for three different tempi. Error bars represent 1 standard error of the mean (SEM).

Fig 2. Between-group differences for complexity ratings.

Between-groups fdr q-values from one-sided t-test p-values for averaged complexity ratings for each of the 10 individual rhythms. Error bars indicate 95% confidence interval. Y-axis indicates ratings of complexity (0–10) while X-axis are the 10 rhythms, ordered by increasing complexity ratings in the control group.

Fig 3. Likeability vs complexity-scores.

Quadratic curve-fit for both groups shows inverted U-curves (Wundt-curves) for the relationship between likeability scores (y-axis) and complexity scores (x-axis).