Linking a latent variable trait-state-occasion model of emotion regulation to cognitive control

ABSTRACT

Emotion dysregulation (ED) is a vulnerability factor for affective disorders that may originate from deficits in cognitive control (CC). Although measures of ED are often designed to assess trait-like tendencies, the extent to which such measures capture a time-varying (TV) or state-like construct versus a time-invariant (TI) or trait-like personality characteristic is unclear. The link between the TV and TI components of ED and CC is also unclear. In a 6-wave, 5-month longitudinal study, community participants (n = 1281) completed the Difficulties in Emotion Regulation Scale (DERS-16), a commonly used measure of ED and measures of CC. A latent variable (trait-state-occasion) model showed that the proportion of TI factor variance (.80) was greater than the TV factor variance (.19). Although TV factor stability was significant, the coefficients were small in magnitude. Furthermore, regression weights for the ED TI factor (average $\beta$ = −.62) were significant and larger than those for the TV factor (average $\beta$ = −.10) in predicting latent CC at each of the six-time points. These findings suggest that ED, as assessed by the DERS-16, is largely TI and this TI component is more strongly linked to CC than the TV component.

KEYWORDS:

• Emotion regulation
• state
• trait
• longitudinal
• cognitive control

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Given the complexity of each TI-TV model and sample size limitations, it was not feasible to predict all six waves of Cognitive Control in a single model. However, results were largely consistent across the single-wave models. Regarding demographic variables, participant age was significantly associated attrition with older participants completing more surveys than younger participants. Race was also significantly associated with attrition. See Supplementary Tables B and Table C for models with age and ethnicity as covariates.

Additional information

Funding

The authors reported there is no funding associated with the work featured in this article.