Relational Uncertainty and Interference from a Partner as Predictors of Demand/Withdraw in Couples with Depressive Symptoms

Abstract

This study examines associations between mechanisms of relational turbulence and demand/withdraw in couples with depressive symptoms. One hundred twenty-six couples in which one or both partners were professionally diagnosed with depression completed an online questionnaire. Actor effects appeared for relational uncertainty and interference from a partner. Partner effects were evident for self uncertainty and interference from a partner. The findings illustrate connections between qualities of the relationship and communication behaviors and illuminate interdependence in partners’ experiences. The results have pragmatic value for helping couples with depressive symptoms reduce negative communication patterns.

Keywords:

• Demand/Withdraw
• Depression
• Dyadic Data
• Relational Turbulence
• Romantic Relationships

Notes

1. Data collection procedures were approved by the university’s Institutional Review Board. Data from this study are also reported in Knobloch et al. (2016) and Sharabi et al. (2016).

2. Some participants reported more than one co-morbid condition, and others did not report any. Accordingly, these percentages represent the proportion of the sample that reported each condition.

3. We also examined the couple-level CES-D scores. Sixty-nine couples (54.7%) included one partner with a depression score at or above the clinical cut off, and 29 couples (23.0%) included two partners who met or exceeded the clinical cut off. In 28 couples (22.2%), neither partner had a depression score at or above the clinical cut off. We elected to keep these 28 couples in the sample because (a) they did meet the requirement of having a diagnosis, (b) current levels of depressive symptomology were likely decreased due to treatment efforts, and (c) we used the continuous measure of depressive symptoms as a covariate in the substantive tests, allowing for a range of depressive symptomology to be represented in analyses.

4. We also analyzed a different subset of the full dataset, in which we included the same-sex couples, but eliminated couples in which both partners reported a depression diagnosis. Then, we distinguished dyads by depression status (diagnosed, not diagnosed) instead of by sex. In this round of tests, the data revealed a similar pattern of associations, and no significant interactions between depression diagnosis status and the predictors. Thus, we elected to report the results with the larger sample distinguished by sex. Results of other tests are available from the first author.