Contributions of worry and social anxiety to paranoia in clinical and non‐clinical samples: Cross‐sectional and longitudinal results

Abstract Objectives Theoretical models implicate social anxiety and worry as potential drivers of paranoia; however, it is unknown which aspect of anxiety relates more strongly to paranoia, whether these relationships differ in clinical versus non‐clinical populations and how these variables relate to one another over time. Methods 143 schizophrenia spectrum disorder (SSD) and 89 non‐clinical comparison (NCC) participants completed assessments of worry (Dunn Worry Questionnaire), social anxiety (Liebowitz Social Anxiety Scale) and paranoia (Revised‐Green Paranoid Thoughts Scale). A subset (72 SSDs; 66 NCCs) completed longitudinal assessments over 6 months. Hierarchical regression was used to analyze the relative contribution of worry and social anxiety on paranoia severity within each group, for cross‐sectional and longitudinal analyses. Results Worry, social anxiety, and paranoia were significantly elevated in SSD and intercorrelated. In NCCs, social anxiety accounted for more variance in paranoia than worry (16% vs. 3% in R 2 ‐change), but in SSDs, worry accounted for more variance in paranoia than social anxiety (14% vs. 5% R 2 ‐change). Longitudinal associations followed a similar pattern. In NCCs, change in social anxiety explained more variance than change in worry (18.5% vs. 6%), and in SSDs, change in worry explained more variance than change in social anxiety (26.5% vs. 0.5%), with social anxiety not adding meaningful variance to paranoia severity, above and beyond worry. Discussion Social anxiety and worry are both highly relevant for paranoia. In sub‐clinical samples, paranoia may be more related to social anxiety, whereas worry may relate more to paranoia in clinical samples. These findings can inform intervention.