Sleeping problems during pregnancy, parental perinatal depressive symptoms and children's emotional problems : findings from the child-sleep and finnbrain birth cohorts

Johanna T. Pietikäinen

Research output: ThesisDoctoral ThesisCollection of Articles


Background and Objective: Depression affects up to 10–15% of women during pregnancy and the first postnatal year. Insomnia during pregnancy is a risk factor for postnatal depression, but it is unclear which insomnia symptoms and at which stage of pregnancy we should screen for preventive purposes. While maternal perinatal depression and its consequences for children’s emotional development have received attention, the longitudinal pattern of both maternal and paternal depressive symptoms and the father’s contribution to the risk of emotional symptoms in the offspring is less clear. The aims of this doctoral study were to investigate a) how risk factors, including prenatal sleep, associate with postnatal depressive symptoms (PDS), b) how the accumulation of such risk factors increases the risk of PDS, c) the longitudinal pattern of maternal and paternal depressive symptoms from pregnancy until two years postnatally, and d) how persistent depressive symptoms associate with children’s emotional problems at the ages of 2 and 5 years Materials and methods: This thesis consists of four individual studies from two Finnish birth cohorts. Studies I, III and IV were based on the CHILD-SLEEP birth cohort (n = 1667 mothers, n = 1598 fathers in late pregnancy and n = 949 children at the age of 2 years) and Study II on the FinnBrain birth cohort (n = 3808 mothers in early pregnancy). In Studies I and II, logistic regression analysis was performed, with a higher level of depressive symptoms (CES-D ≥10 (CS) or EPDS ≥11 (FB)) as the dependent variable and various insomnia symptoms as independent variables. In addition, in Study II, odd ratios were calculated for various PDS risk factor combinations and heat maps were constructed to visualize the accumulation of the PDS risk. In Study III, latent trajectory analyses were performed to examine the longitudinal pattern of maternal and paternal depressive symptoms. In Study IV, associations of maternal and paternal depressive symptom trajectories with children’s emotional problems at the ages of 2 and 5 years were examined with a general linear model. Results: In Studies I and II, we found associations between several insomnia symptoms in late pregnancy and PDS after adjusting for background variables and prenatal depressive symptoms: poor general sleep quality, short sleep of ≤7 h and long sleep latency of >20 min. In addition, in the cumulative models of Study II, we found that long sleep latency (≥20 min) in early pregnancy, decreased functioning in middle pregnancy, and insufficient sleep time during late pregnancy associated with PDS. The accumulation of several risk factors such as a history of depression, anxiety and multiparity substantially increased the risk of PDS, and the best model comprising background variables as well as measurements from early, middle and late pregnancy was able to predict 21.2% of PDS (Nagelkerke 0.21). In Study III, three stable depressive symptom trajectories were found for both mothers and fathers: stable low (n = 1053, 63.1% mothers, n = 1201, 74.9% fathers), stable intermediate (n = 470, 28.1% mothers, n = 362, 22.6% fathers) and stable high (n = 147, 8.8% mothers, n = 41, 2.6% fathers). Depression in one parent also associated with increased depressive symptoms in the spouse (χ2 = 104.6, p <0.001). In Study IV, we constructed combined parental depressive symptom trajectories and found a group with a higher level of persistent maternal depressive symptoms or depressive symptoms in both parents to associate with an increased risk of children’s emotional problems, whereas paternal depressive symptoms did not increase this risk for children if the mother was non-depressive. A higher level of maternal depressive symptoms was associated with a higher level of children’s emotional problems in a dose-dependent manner at the ages of both 2 and 5 years, whereas no such pattern was found in relation to a higher level of paternal depressive symptoms. Importantly, persistent subclinical maternal depressive symptoms increased the risk of externalizing and internalizing problems among both 2- and 5-year-olds. Conclusions: Long sleep latency in early pregnancy and several insomnia symptoms in late pregnancy might be vulnerability markers for an increased risk of PDS and thus potential screening items in order to better detect women at increased risk of PDS. The accumulation of risk factors in PDS should be taken into account when deciding when preventive interventions are necessary, and this study is the first to use heat maps to visualize such an accumulation. Maternal depressive symptoms should preferably be detected during pregnancy and treated with low threshold counselling and psychotherapeutic interventions as the first-line treatment. The screening of both parents is recommendable if one of them presents with depressive symptoms during the perinatal period. Individual, group and couple therapeutic interventions should be developed in the public sector in order to respond to the growing need for non-pharmacological treatments for perinatal mental health disorders.
Original languageEnglish
  • Paavonen, Juulia, Supervisor
  • Paunio, Tiina, Supervisor
Place of PublicationHelsinki
Print ISBNs978-951-51-6784-2
Electronic ISBNs978-951-51-6785-9
Publication statusPublished - 2020
MoE publication typeG5 Doctoral dissertation (article)

Bibliographical note

M1 - 159 s. + liitteet

Fields of Science

  • 3124 Neurology and psychiatry
  • 3123 Gynaecology and paediatrics

Cite this