2002 International International Biennial Scientific Meeting

INTERACTIONS BETWEEN BIOLOGICAL AND PSYCHOSOCIAL RISK FACTORS IN POSTPARTUM DEPRESSION AND ANXIETY

Lori E Ross, Edward M Sellers, and Myroslava K Romach

Psychopharmacology and Dependence Research Unit, Sunnybrook and Women's College Health Sciences Centre and Institute of Medical Science, University of Toronto

l.ross@utoronto.ca

Objectives: This study aimed to develop and test an interactional model of women's mental health during the perinatal period using structural equation modeling (SEM).

Methods: Measures of psychosocial and demographic risk factors, measures of depression (Edinburgh Postnatal Depression Scale, Hamilton Rating Scale for Depression), and measures of anxiety (Brief Symptom Inventory) were administered to 150 women at 36-40 weeks gestation and 6 weeks postpartum. Blood samples were also collected at each visit and analyzed for progesterone and cortisol concentrations.

Results: In multiple regression analyses, demographic (e.g., relationship status, income), clinical (e.g., personal and family psychiatric history) and psychosocial factors (e.g., social support, life stress) were all significantly correlated with depression and anxiety scores, while hormonal variables (plasma progesterone and cortisol concentrations) were not. Using structural equation modeling (SEM) techniques, these variables were entered into a model of perinatal mood disturbance which proposed that "biological risks" (hormone levels, psychiatric history variables) influence mood and anxiety indirectly through interactions with "stressors" (e.g., social support, life stress). Results of model testing revealed that while hormonal variables had neither a direct nor an indirect role in models of depression, they were significant contributors to models of perinatal anxiety.

Conclusions: Although hormonal variables may not have direct effects on women's moods during pregnancy and postpartum, results of structural equation modeling suggest that they do play an important role in perinatal anxiety, possibly by mediating sensitivity to psychosocial stressors. These results highlight the importance of investigating biological variables in conjunction with psychosocial and demographic risk factors when examining complex mental health issues such as postpartum depression.