Searching for antenatal predictors of postnatal depressive symptomatology: unexpected findings from a study of obsessive-compulsive personality traits

Abstract

This prospective study set out to explore the relationship between obsessive-compulsive personality traits assessed antenatally and postnatal depressive symptomatology to determine whether predictive indices could be improved. In so doing the study unexpectedly raised some important questions regarding response rates during pregnancy. The study attempted to recruit all 106 women between 26 and 36 weeks pregnant attending two South London surgeries. Sixty-three participants completed a postal questionnaire which assessed obsessive-compulsive personality traits as well as other vulnerability factors using the Leverton Questionnaire (LQ) and Crown-Crisp Experiential Index (CCEI). At 8-12 weeks postpartum 91 of the participants completed the Edinburgh Postnatal Depression Scale (EPDS) when attending the surgeries for their postnatal visit. Vulnerability assessed by the LQ plus CCEI was related to EPDS scores but predictive utility was low since more than half received a vulnerable designation when assessed this late in pregnancy. Despite being chosen as the best content match to personality factors observed clinically as creating vulnerability to postnatal depression, the obsessive-compulsive personality scale on the Schedule for Non-adaptive and Adaptive Personality added no value because it was found to have little internal reliability. Interestingly, a failure to respond to the antenatal questionnaire was associated with higher postnatal EPDS scores. This raises the possibility that non-response to questionnaires conveys as much information as a particular score in a returned questionnaire. Clinicians should consider following up non-responders, not just screen positive women. Researchers should not consider incomplete data sets as representative of the population.