Percephral and Motor SkiIh, 1991, 73, 1137-1138. O Perceptual and Motor Skills 1991

PREDICTING POSTPARTUM DEPRESSIVE SYMPTOMS: A STRUCTURAL MODELLING ANALYSIS ',* LESLEY A. GRAFF, DENNIS G. DYCK,JOHN R. SCHALLOW University of Manitoba Summary.-In this prospective, longitudinal study, 42 first-time mothers, recruited through prenatal classes, were assessed on variables predictive of postpartum depressive symptoms. Using a causal modelling structural analysis, it was found that prenatal depression, partners' support, and infant-related factors all contributed to the development of depression following childbirth.

The postpartum period has long been associated with adjustment difficulties for a number of women: depressive symptoms have been reported in up to 20 to 30% of new mothers (e.g., Atkinson & Rickel, 1984). Two etiological factors, couples' support and infants' behavior, have consistently been observed as influencing postpartum adjustment, albeit typically in independent studies. The current study allowed a more direct comparison of the influence of these two factors, as well as prepartum expectations of infants' behavior on the development of postpartum depressive symptoms. In this preliminary study, a structural model of postpartum depression (PPD) in first-time mothers was tested using LISREL VI. The model evaluated the direct contributions of prepartum depression and infants' actual behavior, as well as the direct and indirect influence of partners' support and expectations of infants' difficulty on postpartum depression. Forty-two primiparous couples were recruited from various prenatal classes offered over a 5-mo. period. They completed a set of questionnaires about eight weeks before their babies were due and again two months postpartum. Measures on the questionnaires included the Center for Epidemiological Studies-Depression scale, the Dyadic Adjustment Scale, and Expectations/Acrual Infant Behavior Scales (adapted from the Neonatal Perception Inventory). The correlation matrix of the measured variables was entered into the LISREL VI program, which then compared path coefficients to test the significance of relations among the latent and manifest variables, and over-all goodness-of-fit of the model described above. It was found that the proposed model adequately fitted the observed data, based on goodness-of-fit indices (GFI = .898; RMSR = ,093) and a nonsignificant value of chi squared (X,,2 (n = 42) = 23.70, p = .25). Results indicated that it was the discrepancy between expectations and infants' actual behavior that helped explain those 'Address correspondence to Lesley Graff, Department of Psychology, University of Manitoba, yinnipeg, Manitoba, Canada R3T 2N2. This study was supported by a Medical Research Council Studentship to L. A. Graff.

1138

L. A. GRAFF, ET AL.

variables' contribution to maternal postpartum depression. If the woman was expecting that her baby would be little trouble, and the infant was quite fussy, she would be more likely to become depressed. Depression during pregnancy was the strongest prenatal predictor of depression following childbirth. However, support from the partner, both prenatally and postnatally also influenced postpartum depression. The less affection and cohesion perceived by the woman in the couple's relationship, the more likely she would be depressed postnatally. The findings permit an integrat~onof research in the area by providing a clearer understanding of the complex interrelationships among the factors contributing to postpartum depression. Based on these results, a more comprehensive study testing a diathesis-stress model is currently underway, using a larger sample (n = 200) and a longer measurement period (six months between pre- and postnatal data collection). Additional measures include scales specific to the postpartum context, assessment of infants' temperament and infants' crying behavior, and assessment of maternal cognitive vulnerability to depression. REFERENCE ATKINSON,A,, & RICKEL,A. (1984) Postpartum depression in primiparous parents. lournu1 of Abnormal Psychology, 93, 115-119.

Accepted December 28, 1991.

Predicting postpartum depressive symptoms: a structural modelling analysis.

In this prospective, longitudinal study, 42 first-time mothers, recruited through prenatal classes, were assessed on variables predictive of postpartu...
67KB Sizes 0 Downloads 0 Views