Table 4.
Statistics From the Analyses With Social Cognition and Functional Capacity Domains as Mediators Between Neurocognition (MCCB) and Functional Outcome
| X | Y | M | N | a/a’/a’’, estimates (CI) | b/b’/b’’, estimates (CI) | c, estimates (CI) | c’, estimates (CI) | Indirect effect estimate (CI) |
|---|---|---|---|---|---|---|---|---|
| Model 4 | ||||||||
| MCCB Composite score | SLOF Int | FEIT | 469 | 0.253 (−0.199, −0.306) | 0.075 (−0.005,0.156) | 0.101 (0.057,0.144) | −0.025 (−0.32,0.82) | † |
| MCCB Composite score | SLOF Int | TASIT1 | 469 | −0.201 (−0.172, −0.230) | 0.158 (0.007,0.309) | 0.031 (0.003,0.062) * | ||
| MCCB Composite score | SLOF Int | UPSA-B | 469 | 1.09 (0.96, 1.23) | .022 (−0.008,0.053) | † | ||
| Model 5 | ||||||||
| MCCB Composite score | SLOF ELS | FEIT | 467 | 0.253 (−0.199, −0.306) | 0.039 (−0.053,0.131) | 0.101 (0.057,0.144) | −0.025 (−0.32,0.82) | † |
| MCCB Composite score | SLOF ELS | TASIT1 | 467 | −0.201 (−0.172, −0.230) | 0.266 (0.094,0.439) | 0.053 (0.012,0.098) * | ||
| MCCB Composite score | SLOF ELS | UPSA-B | 467 | 1.09 (0.96, 1.23) | 0.158 (0.123,0.193) | 0.172 (0.126,0.221) * | ||
| Model 6 | ||||||||
| MCCB Composite score | SLOF WS | FEIT | 468 | 0.253 (−0.199, −0.306) | 0.124 (0.054,0.194) | 0.199 (0.159,0.239) | 0.072 (0.023,0.121) | 0.031 (0.012,0.053) * |
| MCCB Composite score | SLOF WS | TASIT1 | 468 | −0.201 (−0.172, −0.230) | 0.164 (0.034,0.295) | 0.033 (0.007,0.061) * | ||
| MCCB Composite score | SLOF WS | UPSA-B | 468 | 1.09 (0.96, 1.23) | 0.058 (0.031,0.084) | 0.063 (0.031,0.094) * | ||
Note: MCCB, MATRICS Consensus Cognitive Battery; SLOF, Specific Level of Functioning Scale; Int, interpersonal relationships; ELS, everyday life skills; WS, work skills; FEIT, Facial Emotion Identification Test; TASIT, The Awareness of Social Inference Test; UPSA-B, University of California San Diego (UCSD) Performance-based Skills Assessment; MATRICS, Measurement and Treatment Research to Improve Cognition in Schizophrenia.
In italics, P ≤ .05.
*Prerequisites to test for mediation are fulfilled, and the 95% CI of the indirect effect does not include zero and is therefore significant.
†Prerequisites to test for mediation are not fulfilled.
The results showed that (1) the direct effect of neurocognition on SLOF interpersonal relationships became nonsignificant, suggesting complete mediation through social emotion recognition (assessed with TASIT1); (2) the direct effect of neurocognition on SLOF everyday life skills became nonsignificant, suggesting complete mediation through social emotion recognition (assessed with TASIT1) and functional capacity; (3) the direct effect of neurocognition on SLOF work skills was reduced but remained significant, suggesting partial mediation through social emotion recognition (assessed with both FEIT and TASIT1) and functional capacity.