Fig. 6

Proportion of edge frequency involved in feedback loops in the top 1000 high-symptom-level networks (red) and bottom 1000 low-symptom-level networks (green). Each point represents the average proportion of occurrences for a specific edge across all time points, with error bars indicating one standard deviation. Edges are ordered by their frequency in high-symptom-level networks. Larger solid dots highlight the 13 most frequently observed edges in each group. The differences between the two groups (high − low symptom networks) are visualized in the bottom plot with a gray dashed line. Positive values indicate edges more frequently observed in high-symptom networks, while negative values reflect edges more commonly present in low-symptom networks. Note that the y-axis is inverted. Asterisks indicate edges with descriptively higher frequency in one group, based on non-overlapping standard deviations. The bar plot in the top right corner displays the average proportion of the sad \(\rightleftarrows\) glt feedback loop in each network group across all time points, with error bars representing one standard deviation. This loop appears more frequently in the high-symptom networks (mean = 0.53, SD = 0.04) than in the low-symptom networks (mean = 0.15, SD = 0.01).