RESUMEN
BACKGROUND: The relationship between disturbed sleep and stress is well-documented. Sleep disorders and stress are highly prevalent during the perinatal period, and both are known to contribute to a number of adverse maternal and foetal outcomes. Arginine vasopressin (AVP) is a hormone and a neuropeptide that is involved in stress response, social bonding and circadian regulation of the sleep-wake cycle. Whether the AVP system is involved in regulation of stress response and sleep quality in the context of the perinatal mental health is currently unknown. The objective of the present study was to assess the relationship between levels of cumulative and ongoing psychosocial risk, levels of disordered sleep and AVP methylation in a community sample of pregnant and postpartum women. METHODS: A sample of 316 participants completed a battery of questionnaires during the second trimester of pregnancy (PN2, 12-14 weeks gestation), third trimester (PN3, 32-34 weeks gestation), and at 7-9 weeks postpartum (PP). Disordered sleep was measured using the Sleep Symptom Checklist at PN2, PN3 and PP; cumulative psychosocial risk was assessed with the Antenatal Risk Questionnaire (ANRQ) at PN2; salivary DNA was collected at the follow-up (FU, 2.9 years postpartum); and % methylation were calculated for AVP and for two of the three AVP receptor genes (AVPR1a and AVPR1b). Women were separated into high (HighPR) and low (LowPR) psychosocial risk groups, based on their scores on the ANRQ. RESULTS: Women in the HighPR group had significantly worse sleep disturbances during PN2 (p < .001) and PN3 (p < .001), but not at PP (p = .146) than women in the LowPR group. In HighPR participants only, methylation of AVP at intron 1 negatively correlated with sleep disturbances at PN2 (rs=-.390, p = .001), PN3 (rs=-.384, p = .002) and at PP (rs= -.269, p = .032). There was no association between sleep disturbances and AVPR1a or AVPR1b methylation, or between sleep disturbances and any of the AVP methylation for the LowPR group. Lastly, cumulative psychosocial stress was a moderator for the relationship between AVP intron 1 methylation and disordered sleep at PN2 (p < .001, adjusted R2 = .105), PN2 (p < .001, adjusted R2 = .088) and PP (p = .003, adjusted R2 = .064). CONCLUSIONS: Our results suggest that cumulative psychosocial stress exacerbates sleep disorders in pregnant women, and that salivary DNA methylation patterns of the AVP gene may be seen as a marker of biological predisposition to stress and sleep reactivity during the perinatal period. Further research is needed to establish causal links between AVP methylation, sleep and stress.