Prenatal maternal stress is associated with lower cortisol and cortisone levels in the first morning urine of 45-month-old children

Abstract Prenatal stress (PS) has been related to altered hypothalamic-pituitary-adrenal (HPA) axis activity later in life. So far, studies in children assessing HPA axis functioning have focused on salivary cortisol, reflecting daytime activity. The present work is part of a prospective study and aims to extend knowledge about the association between PS and HPA axis regulation in children. To do so, we investigated cortisol, cortisone, and the ratio cortisone/(cortisone + cortisol) in the first morning urine of 45-month-old children in relation to several measures of maternal stress during pregnancy. Urinary cortisol and cortisone were measured by online turbulent flow chromatography coupled with high performance liquid chromatography-tandem mass spectrometry. PS was defined as: perceived stress for aim 1 (Perceived Stress Scale; n = 280); presence of self-reported (n = 371) and expert-rated psychopathology for aim 2 (Mini International Neuropsychiatric Interview; n = 281); continuous measures of anxiety and depression for exploratory aim 3 (State-Trait Anxiety Inventory and Edinburgh Postnatal Depression Scale; n = 280). Aim 1: Perceived maternal PS showed negative associations with cortisol and cortisone levels. Aim 2: The presence of expert-rated maternal psychopathology was associated with reduced morning cortisone. Aim 3: Continuous measures of anxiety and depression showed negative associations with cortisol and cortisone levels. After correcting for multiple testing, perceived maternal PS (aim 1) and prenatal level of anxiety (aim 3) were significant predictors of children’s urinary cortisol and cortisone in the morning (and, in the case of cortisone, also prenatal level of depression). The ratio cortisone/(cortisone + cortisol) as a global marker for the balance between the enzymes metabolizing cortisol to cortisone and vice versa (11β-hydroxysteroid dehydrogenases type 1 and 2; 11β-HSD1 and 2) was not associated with any measure of maternal PS (aims 1–3). The present study provides insight into possible programming effects of PS on nocturnal HPA axis activity and a proxy of 11β-HSD in a large sample. The results suggest that the nocturnal rate of cortisol production is lower in children exposed to PS, but do not support the hypothesis of divergent 11β-HSD activity.