Linking Daily-Life Interpersonal Stressors and Health Problems Via Affective Reactivity in Borderline Personality and Depressive Disorders

Abstract

Objective: Borderline personality disorder (BPD) is associated with unstable interpersonal relationships, affective instability, and physical health problems. In individuals with BPD, intense affective reactions to interpersonal stressors may contribute to the increased prevalence of health problems.

Methods: BPD (n = 81) and depressed participants (DD; n = 50) completed six daily ambulatory assessment prompts for 28 days. At each prompt, participants reported interpersonal stressors (disagreements, rejections, feeling let down), negative affect, and health problems in four domains (gastrointestinal, respiratory, aches, depressive symptoms). In multilevel moderated mediation models, we examined the indirect effects of interpersonal stressors on health problems via negative affect, by group.

Results: Interpersonal stressors were positively associated with negative affect in both groups (β values > 0.12, p values < .001), but more so for participants with BPD (βDay = 0.05, p < .001). Negative affect was positively associated with health problems across all domains (βMoment/Day values > 0.01, p values < .046), but associations were larger at the day level for respiratory symptoms in BPD (β = 0.02, p = .025) and for depressive symptoms in DD (β = 0.04, p < .001). Negative affect mediated the association of interpersonal stressors and health problems in both groups, with larger effects for the DD group for depressive problems (β = 0.02, p = .092) and for the BPD group for the other three domains (β values > 0.02, p values < .090).

Conclusions: Interpersonal stressors may contribute to increased physical health problems via an inability to regulate affective responses to such events. This pathway may be stronger in several health domains for those with BPD and may contribute to an elevated risk of morbidity and mortality in this disorder, suggesting a target for intervention to reduce these risks.

Figure 1.

Proposed mediation model including group (borderline versus depressed) as a moderating variable.

Figure 2.

Standardized indirect effect estimates of the association between IS and HP via NA. Within domain, borderline (BPD) and depressed (DD) effects are compared at A) momentary, B) day, and C) person levels. Error bars represent 95% bootstrapped confidence intervals.