Attachment impacts cortisol awakening response in chronically depressed individuals.

Background: Early life experiences shape individual attachment, creating a template for regulating emotions in interpersonal situations, likely to persist across the lifespan. Research has shown that individual attachment creates vulnerability for depression, and also impacts the Hypothalamic-Pituitary-Adrenal (HPA) axis. Still, the relationship between attachment and the HPA axis in depressed individuals is unclear. Cortisol awakening response (CAR) has been recently investigated as a possibly useful physiological marker related to attachment insecurity and depression risk. However, research exploring the relationship between the CAR and attachment in individuals with chronic depression in either the presence or the absence of comorbid anxiety is lacking. The purpose of the current study was to fill this gap, by comparing the CAR in individuals with chronic depression with/without comorbid anxieties and controls. In addition, we also wanted to explore the relationship between attachment and the CAR in this group and to explore their predictive role for later depression severity.

Methods: Individuals experiencing a current depressive episode at least six months in length (cMDD; n = 63) and healthy controls (HC; n = 57) were enrolled in the study (total n = 120). Participants completed a structured clinical diagnostic interview (SCID-I) as well as measures of depression severity (Beck Depression Inventory-II (BDI-II) and Hamilton Rating Scale for Depression) and attachment dimensions (Experiences in Close Relationships scale; ECR) at baseline. In addition, participants provided salivary samples at four time points (i.e. 0 (S1), 30, 45 and 60 min) following awakening on two consecutive days. S1 cortisol, the area under the curve with respect to ground (AUCg) and increase (AUCi) were calculated based on the average values across both days. The HC and cMDD groups were compared on all measures. The CAR for individuals with cMDD alone (n = 14) and individuals with cMDD with two or more comorbid anxiety disorders (cMDD ≥ 2Anx; n = 30) were also compared. A subset of participants (n = 59) agreed to return for follow up one year later. Participants returning for follow up repeated the BDI-II and ECR. No salivary samples were collected at follow-up.

Results: The cMDD group had significantly lower S1 cortisol and AUCg compared to the HC group (both p ≤ 0.02). cMDD and cMDD ≥ 2Anx groups did not differ in their CAR. Regression analyses revealed that depression severity and the attachment interaction term was associated with lower S1 and AUCg cortisol (p < 0.01). Greater attachment avoidance was positively associated with S1 cortisol (p = 0.02), while mean awakening time on sample days was negatively associated with S1 cortisol. We also found a significant interaction between the attachment dimensions such that at low levels of attachment anxiety, attachment avoidance had a positive relationship with S1 cortisol and AUCg. The opposite relationship existed when attachment anxiety was high. Higher baseline BDI-II score and higher baseline attachment anxiety were predictive of higher scores on the BDI-II one-year later (both p < 0.05).

Conclusions: The current findings bring evidence that depression severity is associated with blunting of the CAR irrespective of the comorbid status with anxiety disorders. In addition, attachment avoidance may protect against the CAR blunting in individuals with low attachment anxiety. However, individuals with high attachment anxiety and avoidance might have additional CAR blunting. Attachment anxiety might be a good predictor of future depression severity.