The associations between paternal postpartum depressive symptoms and testosterone and cortisol levels in hair over the first two years postpartum

Abstract

Background

After the birth of a child, also fathers may develop postpartum depression. Altered steroid hormone concentrations are discussed as a possible underlying mechanism, as these have been associated with depressive symptoms in previous studies outside the postpartum period. While higher paternal testosterone levels have been found to protect against paternal postpartum depressive symptoms (PPDS), an association between higher cortisol levels and PPDS has been seen in postpartum mothers, with no comparable studies available on fathers.

Objective

This study aimed to investigate cross-sectional and longitudinal associations between testosterone and cortisol levels in hair and PPDS over a period of 2 years postpartum.

Methods

Data from N = 226 fathers, who took part in the endocrine sub-study DREAM_HAIR of the longitudinal prospective cohort study DREAM, were used. PPDS were assessed 8 weeks, 14 months, and 24 months after birth using the Edinburgh Postnatal Depression Scale. At the same time, fathers provided 2 cm scalp-near hair samples in which testosterone (HairT) and cortisol (HairF) levels were determined. Cross-sectional and longitudinal associations between HairT, HairF and paternal PPDS were investigated.

Results

Correlation analyses showed a negative cross-sectional association between HairF levels and paternal PPDS 14 months after birth. A random intercept cross-lagged panel model revealed prospective relationships between paternal PPDS 8 weeks postpartum and HairF 14 months postpartum, and additionally between 14 months and 2 years postpartum in an exploratory model with similarly good model fit. No further significant associations of HairF with paternal PPDS emerged, and none of the analyses with HairT became significant. The overall pattern of results was confirmed when controlling for the influence of batch and storage time on HairT and HairF levels.

Conclusion

No consistent relationships between HairT or HairF and paternal PPDS emerged in this relatively healthy cohort. In HairF analyses with significant results, lower HairF was associated with more severe PPDS. Longitudinal results imply that altered cortisol secretion may rather follow than precede changes in paternal PPDS. Further research on hormonal changes in PPDS in fathers should consider possible covariates and examine fathers with higher depressive burden, which may help to identify fathers at risk and inform future preventive and interventive approaches.