Citation

Patten SB. Soc. Psychiatry Psychiatr. Epidemiol. 2013; 48(6): 927-933.

Affiliation

Department of Community Health Sciences, University of Calgary, 3280 Hospital Drive NW, Calgary, AB, T2N 4Z6, Canada, patten@ucalgary.ca.

Copyright

(Copyright © 2013, Holtzbrinck Springer Nature Publishing Group)

DOI

10.1007/s00127-012-0603-9

PMID

23114784

Abstract

PURPOSE: Using a representative longitudinal cohort consisting of more than 8,000 community residents, this study sought to evaluate patterns of interaction between childhood adversity and adult stressors in relation to MDE. The goal was to interpret the interactions using epidemiologic theory. METHODS: A Canadian longitudinal study called the National Population Health Survey (NPHS) was the data source. This NPHS began in 1994 and the cohort has subsequently been interviewed every 2 years. Childhood adversities were assessed retrospectively and adult stressors and MDE were evaluated during follow-up. Interactions between various potential MDE risk factors were assessed on an additive scale using linear regression and on a multiplicative scale using logistic regression. RESULTS: Hypothesized interactions between negative childhood experiences and more recent stressors were apparent in statistical models adopting an additive (linear regression), but not multiplicative (logisitic), perspective. According to the component-cause model of etiology, this pattern suggests shared causal mechanisms. There was no general tendency for such interactions to occur with other risk factors. CONCLUSIONS: Biological mechanisms responsible for early life calibration of stress response systems may generate persistent sensitization to stressors, thereby increasing the risk of MDE following exposure to stressful events later in life. Reliance on multiplicative models such as logistic regression and log-binomial regression in psychiatric epidemiological studies may render etiologically important interactions more difficult to identify.

Language: en