TABLE 4-2 Multiple Logistic Regression Analysis of Major Depression in Past 12 Months Among Females


Odds ratio (95% C.I.a)


Father sad for 2+ weeks during childhood

3.17 (1.80, 5.60)

< 0.0001

Closeness of mother: summary scoreb (higher values denote less closeness)

1.24 (1.07, 1.44)


Father social problems: summary scoreb (higher values denote more problems)

0.80 (0.65, 0.99)


Parental neglect: summary scoreb (higher scores denote more neglect)

1.21 (1.01, 1.47)


Ageb (years)

1.36 (1.03, 1.80)


Relationship status (overall p = 0.07)



Separated, widowed, divorced vs. Married, cohabitating

1.37 (0.72, 2.59)


Never married vs. married, cohabitating

1.77 (1.09, 2.90)


Any PTSD before age 18

3.41 (1.98, 5.88)

< 0.0001

NOTES: PTSD = Posttraumatic stress disorder; N = 1,035 (176 with major depression, 859 without); model p-value < 0.0001; generalized R2 = 0.12; C-statistic = 0.66.

aConfidence interval.

bFor a one standard deviation difference.

SOURCE: Tabulations based on the National Comorbidity Survey-Replication (see

history can be included in public use data sets to be further analyzed, e.g., identifying the perpetrator of the trauma or neglect.

Timing of Exposure

The timing of exposure has received quite a bit of attention as a potential moderator, addressing the question: Is there a sensitive period for exposure to maternal depression? Some theories suggest that the first year of life, approximately, may represent a sensitive period given both infants’ dependence on their caregivers and the centrality of responsive, sensitive caregiving for children’s development of secure attachment and other aspects of emotion regulation (Essex et al., 2001). Two prospective studies conducted in Great Britain came to somewhat different conclusions regarding the role of postpartum depression alone in predicting adverse outcomes for children, regardless of subsequent episodes. Hay and colleagues studied youth from low-income homes. In contrast to cognitive development, for which postnatal depression has been found to be associated with children’s later cognitive functioning (at ages 11 and 16) regardless of subsequent exposures to maternal depression, postnatal depression plus at least one subsequent episode of depression in mothers predicted children’s behavioral

The National Academies of Sciences, Engineering, and Medicine
500 Fifth St. N.W. | Washington, D.C. 20001

Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement