Emotion differentiation and depressive symptoms: a systematic review and meta-analysis

This study aimed to systematically examine the relationship between emotion differentiation (ED) and depressive symptoms. Cochrane Library, Web of Science, Embase, EBSCO, PubMed and Science Direct databases were searched from the time of build to March 25, 2023 for studies that: (i) were in English, (ii) peer-reviewed, (iii) reported correlations between PED/NED and depressive symptoms, and (iv) NED, PED, and NA were derived from emotional scores using ecological momentary assessment, daily diaries, and standardized stimuli ratings. Exclusions included duplicates, non-empirical research, studies not focused on PED/NED or depressive symptoms, conference abstracts, review articles, and studies lacking correlations between PED/NED and depressive symptoms. Review Manager 5.4 and Stata 17.0 software were used to perform meta-analysis, subgroup analysis, sensitivity analysis, meta-regression analysis and publication bias detection of the extracted data. A total of 16 samples with 2515 participants were included. Some evaluated studies found a negative correlation between negative emotion differentiation (NED) and depressive symptoms, while others reported positive association. No significant correlation was observed between positive emotion differentiation (PED) and depressive symptoms. Meta-analysis revealed a significantly negative association between NED and depressive symptoms (r = -0.14, 95% CI= [-0.21, -0.06], p < 0.01). However, PED was not significantly related to depressive symptoms (r = -0.02, 95% CI = [-0.07, 0.03], p = 0.41). Critically, the correlation between NED and depressive symptoms was moderated by mean negative affect (NA) (t = 3.50, p< 0.01, 95% CI = [0.297, 1.377]), whereas diagnostic status, assessment of NED, age and proportion of female did not emerge as moderators. This study confirmed the negative association between NED and depressive symptoms, in particular when experiencing more intensive negative emotions, whereas the null association between PED and depressive symptoms. In addition, the diagnostic status, assessment of NED, age and proportion of female did not affect the link. Our findings have important clinical implications for facilitating the comprehending, prevention and treatment of depression. We discuss limitations as well as suggestions for future research.