Postpartum psychosis and the association with sociodemographic and obstetric factors

Abstract

Objective: To examine the association between postpartum psychosis and the sociodemographic factors age, education, marital status and year of delivery (study 1). To examine the association between neighbourhood socioeconomic characteristics and postpartum psychosis (study 2). To examine the association between postpartum psychosis and obstetric factors (study 3). To examine the association between non-puerperal readmission and years of follow-up among women with postpartum psychosis and to examine the impact of age, type of psychosis, previous psychiatric hospitalisation and education on the risk of non-puerperal readmission (study 4) among these women.

Methods: Data sources were the Swedish population registers and medical registers. Postpartum psychosis was defined as hospital admission due to psychotic disorder within three months after delivery, which was the outcome variable in studies 1 3. Study 1 followed 502,767 first-time mothers between Jan. 1, 1986, and Dec. 31, 1997, for postpartum psychosis. Study 2 followed 485,199 first-time mothers between Jan. 1, 1986, and Sept. 30, 1998, for postpartum psychosis. Study 3 followed 1,133,368 first-time mothers between Jan. 1, 1975, and Dec. 31, 2003, for postpartum psychosis. Study 4 followed 1,340 women with postpartum psychosis between Jan. 1, 1975, and Dec. 31, 2004, for non-puerperal readmission due to psychiatric disorder. The explanatory variables included age (studies 1 4), education (studies 1, 2, 4) marital status (studies 1, 2), year of delivery (studies 1 3), neighbourhood income (study 2), previous psychiatric hospitalisation (studies 3, 4), years of follow-up (study 4), type of psychosis (study 4), calendar year (study 4), and 19 different obstetric factors (study 3). Cox regression models (studies 1 3) and Cox frailty regression models (study 4) were used in the statistical analysis.

Results: Old age in the mother, not living with the father of the child, living in a socioeconomically deprived neighbourhood, preterm birth and acute caesarean section were associated with increased risk of postpartum psychosis among first-time mothers after adjustment for possible confounders. However, the HRs for these factors were only moderately increased, which implies that the absolute increase in risk is small. In contrast, previous hospitalisation for psychiatric disorders was associated with a more than 100-fold increased risk of postpartum psychosis. The risk of non-puerperal readmission due to a psychiatric disorder remained high for many years after the postpartum psychosis, for all women in the study group. In addition, the risk of non-puerperal readmission due to a psychiatric disorder was higher among women with low educational level, previous psychiatric hospitalisation and schizophrenia.

Conclusion: Stress from adverse sociodemographic and obstetric conditions can possibly play a role in the development of postpartum psychosis, although these associations might be confounded by unknown previous psychiatric illness. Postpartum psychosis is often part of a severe and chronic psychiatric illness, with high non-puerperal readmission rates for many years after the postpartum psychosis. Finally, previous psychiatric disorder is a strong risk factor for postpartum psychosis. Therefore, it is very important to consider previous psychiatric disorder in antenatal and postnatal care.