Notes on Perinatal Life Stress and Infant Temperament and Child Development (Thiel et al., 2021)

Background and rationale

• The perinatal period is a critical window for infant development and is highly dependent on parental well-being. Genetic factors, hormonal maternal stress responses, and postpartum experiences shape early development. There is a well-documented link between maternal psychopathology during the perinatal period and adverse child outcomes. Accumulating evidence indicates that in utero exposure to maternal stress has persistent effects on the infant, including links to preterm birth, poorer infant health, and later behavioral, cognitive, and interpersonal challenges.
• In particular, maternal distress during pregnancy may increase fearfulness and difficult temperament in infants and toddlers. Ongoing adversity in early temperament and emotion regulation can reinforce later behavioral problems, creating long-term adverse effects. Positive factors such as mother–infant bonding and social support can buffer these risks, but the potential detrimental effects of perinatal stress warrant further study.
• The developmental model of fetal programming posits that in utero exposure to maternal stress can trigger long-term fetal developmental responses and enduring changes in neuroendocrine regulation, possibly via placental mechanisms and epigenetic modifications. Stress hormones (e.g., cortisol) and placenta enzyme regulation (11β-HSD2) may modulate fetal exposure, potentially influencing brain development (amygdala, prefrontal and temporal cortices) and later social-emotional functioning. Epigenetic changes (e.g., DNA methylation) may mediate these effects.
• The present study investigates whether maternal perinatal life stress—assessed as the individual emotional impact of stressful life events—predicts infant temperament at 8 weeks and child development at 2 years, beyond demographic, pregnancy, childbirth-related, and postpartum factors.

Conceptualization of stress

• Stress is conceptualized not only as events that occur but as the person’s emotional response to those events. The study emphasizes the perinatal life stress burden as a function of perceived emotional impact of life events, captured by a 12-month retrospective assessment at 8 weeks postpartum.
• Stress is viewed through a person–environment transaction framework, balancing environmental demands with coping resources. The perinatal life stress measure integrates both exposure to events and the subjective distress they caused, aligning with the idea that outcomes depend on individual appraisal and physiological stress responses, not only on the occurrence of events.

Methods

• Design and sample
  • Data come from the Akershus Birth Cohort, a prospective Norwegian study following women from pregnancy through early childhood. Recruitment occurred November 2008–April 2010 at Akershus University Hospital.
  • Eligible and Norwegian-speaking women: 80% consented (n = 3{,}752) and completed baseline questionnaires.
  • Assessments occurred at four time points: during pregnancy week 17 (T1), week 32 (T2), 8 weeks postpartum (T3), and 2 years postpartum (T4).
  • Attrition analyses used logistic regression with predictors including maternal age, education, depressive, anxiety, and PTSD symptoms at week 17 and birth data to predict dropout within 2 years. Findings indicated older age and higher education reduced dropout, while postpartum depressive symptoms slightly increased dropout; anxiety and PTSD alone were not predictive of dropout.
  • Ethical approvals were obtained from regional committees in Norway; the study followed the Declaration of Helsinki. All procedures complied with ethical standards.
• Measures and constructs
  • Perinatal life stress (eight weeks postpartum assessment): 10 life events drawn from established scales. For each event, mothers reported whether it occurred in the past 12 months and rated emotional distress on a 3-point scale: 1 = emotionally not so difficult, 2 = difficult, 3 = very difficult. The total score is the sum across items. Stress level categories: 0 = low; 1–5 = moderate; >5 = high. The score reflects the subjective distress associated with life events, not just occurrence.
  • Difficult infant temperament (8 weeks): adapted version of the fussy/difficult subscale of the Infant Characteristics Questionnaire (ICQ). 10 items rated 1 (completely disagree) to 7 (completely agree); higher scores indicate more difficult temperament. Internal consistency in this sample: α = 0.83.
  • Child development at 2 years: Ages & Stages Questionnaire (ASQ-3) for gross motor, fine motor, and communication; ASQ: Social–Emotional (ASQ:SE) for social-emotional development.
  • ASQ-3 domains: each domain has 6 items, scored 10 = yes, 5 = sometimes, 0 = not yet; domain sums range 0–60. Norwegian version validated; internal consistency: gross motor α = 0.61, fine motor α = 0.42, communication α = 0.80.
  • ASQ:SE: 26 items, responses 0 = most of the time, 5 = sometimes, 10 = never/rarely; additional parent concerns add points. Total score 0–390, with higher scores indicating potential difficulties. Norwegian ASQ:SE validation; internal consistency α = 0.52.
  • Maternal mental health and related constructs
  • Postpartum depression (and depression over time): Edinburgh Postnatal Depression Scale (EPDS), 10 items, 0–3 per item, total 0–30. Internal consistency: α = 0.83 (gestational week 32), α = 0.85 (8 weeks postpartum), α = 0.87 (2 years postpartum).
  • Anxiety: Hopkins Symptom Checklist A (SCL-A), 10 items, 1–4 scale; total 10–40. Internal consistency: α = 0.78 (gestational week 32), α = 0.80 (8 weeks), α = 0.84 (2 years).
  • Prenatal attachment: 9-item short version of the Prenatal Attachment Inventory; response 1–4; total 9–36. Internal consistency: α = 0.82.
  • Social support: Oslo Social Support Scale (3 items), scores 3–14; higher scores indicate stronger support. Internal consistency: α = 0.55.
  • Prenatal depression and anxiety in the correlations table use the EPDS and SCL-A measures described above.
  • Demographics and obstetric variables (hospital birth records): education, employment, marital status, smoking and alcohol during pregnancy, maternal age, parity, prematurity, emergency cesarean, male child sex, obstetric complications, severe infant health complications (NICU admission), and postnatal health issues at 2 years.
  • Additional postnatal factors: social support at 8 weeks, continued breastfeeding for 12 months, and postnatal depression/anxiety at 8 weeks and 2 years.
• Statistical analysis
  • Analyses conducted in IBM SPSS Statistics, v25. Missing data handled with multiple imputation.
  • Preliminary analyses compared those with and without perinatal life stress using t-tests, chi-square tests, or Fisher exact tests as appropriate.
  • Bivariate relationships explored with Pearson correlations among perinatal life stress, ICQ difficult temperament (T3), ASQ-3 and ASQ:SE outcomes (T4), and demographic/childbirth/postpartum variables.
  • Hierarchical multiple regression to test whether perinatal life stress predicts:
  • Difficult temperament at 8 weeks (T3) while controlling for pregnancy variables, childbirth variables, and postchildbirth factors.
  • Fine motor and communication development at 2 years (T4) above and beyond prechildbirth, childbirth, and postchildbirth factors.
  • Social-emotional development at 2 years (T4) above and beyond the same covariates.
  • Regression steps were: Step 1 = prechildbirth variables (pregnancy-related), Step 2 = childbirth-related variables, Step 3 = postchildbirth variables, Step 4 = perinatal life stress.
  • Outcome measures treated as continuous in the regression models; R^2 and ΔR^2 reported to indicate explained variance and incremental contribution of perinatal life stress.
  • Correlation and regression results interpreted with standard significance levels: *p < 0.05; **p < 0.01; ***p < 0.001.

Participants and descriptive statistics

• Age range: 19–46 years; mean age roughly 31 years (SD ≈ 5).
• Marital/partnership status: ~95% married or living with partner; ~91% employed at least part-time; ~62% with 12+ years of education.
• Health and obstetric factors: 7% of infants admitted to NICU; 32% reported obstetric complications; 33% reported child health issues at 2 years; 52% delivered male infants; 12% had emergency cesarean; term delivery ≈ 94% (i.e., within 21 days of estimated due date or ≥ 258 days gestation).
• Perinatal life stress: ~68% reported at least moderate levels of life stress (n ≈ 2{,}532); comparison between those with moderate stress and those with low stress showed means and group differences in age, education, partnership status, employment, smoking, alcohol use, and breastfeeding duration.
  • Those with moderate perinatal life stress were younger on average (mean 30.7 vs 31.6 years; SDs ~4.9 and 4.2), p < 0.001, and less likely to have 12 years of education (57% vs 72%), p < 0.001.
  • Higher prevalence of unmarried/separated status, unemployment, smoking during pregnancy, and alcohol use during pregnancy; lower rates of continued breastfeeding for 12 months.

Key associations between perinatal life stress and outcomes

• Associations with outcomes (Table 1):
  • Perinatal life stress was associated with difficult infant temperament and with finer domains of child development (fine motor, communication, and social-emotional development), but not with gross motor development.
  • Correlations among study outcomes showed: difficult temperament (T3) negatively associated with gross motor, fine motor, and communication domains at 2 years (r ≈ -0.08 to -0.09 for gross/fine/communication; p < 0.05 to p < 0.001) and positively associated with social-emotional development at 2 years (r ≈ 0.22; p < 0.001).
  • Social-emotional development (ASQ:SE) correlated with other outcomes in a pattern consistent with the above.
• Hierarchical regression results (Table 2):
  • Difficult infant temperament at 8 weeks (T3): After entering prechildbirth, childbirth-related, and postchildbirth covariates, maternal perinatal life stress contributed an additional statistically significant amount of variance (
    ΔR^2 ≈ 0.01), on top of a model that explained about 12% of the variance in difficult temperament. The overall model explained roughly R^2 = 0.12. Other significant predictors included primiparity, male infant sex, obstetric complications, lower social support, and maternal postpartum depression; maternal age, prenatal depression, and anxiety at 8 weeks did not significantly contribute.
  • Fine motor and communication development at 2 years: Perinatal life stress did not contribute significantly to either fine motor or communication development beyond covariates.
  • Social-emotional development at 2 years: Perinatal life stress added a modest increment to explained variance (ΔR^2 ≈ 0.01), with the full model explaining about R^2 = 0.22. Significant covariates included lower maternal education, prenatal depression, low social support, difficult infant temperament, no continued breastfeeding for 12 months, postpartum depression and anxiety at 2 years, and lower fine motor and communication development; other factors (e.g., maternal age, marital status, employment, smoking, prematurity, child sex, early depression/anxiety) were not significant contributors.

Mechanisms and interpretation

• Prenatal stress may influence fetal neurodevelopment via the maternal–fetal HPA axis, placental regulation of cortisol exposure (e.g., down-regulation of placental 11β-HSD2), and subsequent effects on brain regions implicated in emotion regulation (amygdala, prefrontal and temporal cortices).
• In utero exposure to maternal stress may be linked to stable epigenetic changes (e.g., DNA methylation) that mediate the relationship between maternal stress and child outcomes.
• The findings support the fetal programming perspective by showing that subclinical, nonpathological levels of perinatal distress are associated with longer-term neurodevelopmental and social-emotional consequences, independent of postpartum psychopathology.
• The study emphasizes the relevance of assessing the subjective emotional impact of life events, not merely their occurrence, as offspring may be affected through the mother’s biophysiological stress response, which depends on individual perception of distress.

Clinical implications and recommendations

• The results underscore the need for person-centered perinatal care and enhanced support for mothers facing difficult life events during the perinatal period.
• Clinicians should maintain awareness of the potential in utero origins of later child social-emotional development and consider early interventions to support maternal mental health and coping with stress.
• Interventions could include screening for perinatal life stress, providing targeted psychosocial support during pregnancy and postpartum, and facilitating resources to bolster mother–infant bonding and social support.
• The authors advocate for routine prenatal care to integrate considerations of in utero development and maternal psychological health to facilitate timely intervention and potentially mitigate long-term adverse outcomes.
• Future research should examine the timing of perinatal life stress exposure (distinct gestational windows) and investigate the neurobiological and environmental mechanisms that mediate the relationship between maternal perinatal life stress and child outcomes.

Strengths, limitations, and robustness of findings

• Strengths:
  • Large, prospective cohort with data from pregnancy to 2 years postnatal.
  • Rich set of covariates spanning pregnancy, delivery, and postpartum periods.
  • Use of validated instruments (ICQ, ASQ-3, ASQ:SE, EPDS, SCL-A, Prenatal Attachment Inventory, Oslo Social Support Scale).
  • Focus on perinatal life stress as subjective distress to life events, rather than mere exposure, aligning with stress theory and potential biophysiological pathways.
• Limitations:
  • All measures rely on maternal self-report, which may introduce reporting bias and common-method variance.
  • Emotional stress data were collected retrospectively at 8 weeks postpartum; precise timing during gestation could not be pinpointed.
  • The study did not include objective emotional stress data from hospital records; timing during pregnancy remains a question for future work.
  • Some measures (e.g., ASQ:SE and some subscales) had modest internal consistency, especially in Norwegian translations, which may attenuate observed associations.
  • The sample was relatively homogeneous (Norwegian-speaking, predominantly White), potentially limiting generalizability to more diverse populations.
  • Attrition, although analyzed, may still bias estimates; those with severe prenatal distress might be underrepresented at follow-up.
  • Despite statistical significance, effect sizes were modest; nonetheless, the findings have clinical relevance given potential cumulative and long-term implications.

Conclusion

• The study provides prospective evidence that maternal perinatal life stress, defined by the subjective emotional impact of life events, contributes to difficult infant temperament at 8 weeks and to challenges in social-emotional development at 2 years, even after accounting for demographic, pregnancy-related, childbirth-related, and postpartum factors.
• These findings highlight the burden of life events perceived as distressing during the perinatal period and suggest potential long-term neurobiological and social-emotional consequences for the child. They support calls for person-centered perinatal care and enhanced support for mothers facing difficult life events to mitigate adverse developmental trajectories.
• Future research should investigate the timing of in utero exposure, underlying neurobiological mechanisms (e.g., HPA axis regulation, placental function, epigenetic modifications), and environmental contributors to better understand and intervene in the perinatal transmission of stress effects.

Glossary of key measures and terms

• Perinatal life stress: composite 10-item life event measure (past 12 months) with a 3-point distress rating per event; total score categorized as 0 (low), 1–5 (moderate), >5 (high).
• ICQ (Infant Characteristics Questionnaire) – Difficult temperament subscale: 10 items, 1–7 scale; higher scores indicate more difficult temperament; internal consistency in this study: α = 0.83.
• ASQ-3 (Ages & Stages Questionnaire, 3rd edition): domains of gross motor, fine motor, and communication; each domain sum ranges 0–60; higher is better development.
• ASQ:SE (Social–Emotional): total score 0–390; higher scores indicate potential social-emotional difficulties.
• EPDS: Edinburgh Postnatal Depression Scale; 0–30; higher scores indicate more depressive symptoms.
• SCL-A: Hopkins Symptom Checklist – Anxiety; 10 items; 10–40; higher scores indicate greater anxiety.
• Prenatal Attachment Inventory: 9-item short form; 9–36; higher scores indicate stronger prenatal attachment.
• Oslo Social Support Scale: 3 items; 3–14; higher scores indicate stronger social support.
• Regression metrics: R^2 represents explained variance; ΔR^2 indicates the incremental variance explained by a variable or block of variables.

Statistical details and representative numbers (selected highlights)

• Sample sizes and timing: baseline n = 3{,}752; 2-year follow-up subsample reported as n ≈ 3{,}572 in the results narrative.
• Perinatal life stress categories: 0 = low; 1–5 = moderate; >5 = high; measured at 8 weeks postpartum.
• Outcome correlations (approximate, from Table 1):
  • Difficult temperament at 8 weeks negatively related to later gross motor (r ≈ -0.08*), fine motor (r ≈ -0.09), and communication (r ≈ -0.06) at 2 years; positively related to social-emotional development at 2 years (r ≈ 0.22).
  • ASQ-3 and ASQ:SE domains show interrelations consistent with broader development patterns.
• Regression findings (Table 2 summaries):
  • Difficult temperament at 8 weeks: prechildbirth, childbirth-related, and postchildbirth factors explained a significant portion of variance; perinatal life stress added †≈ 0.01 of explained variance, total explained variance around R^2 ext{ (difficult temperament)} \approx 0.12.
  • Fine motor and communication development at 2 years: no significant contribution from perinatal life stress after accounting for covariates.
  • Social-emotional development at 2 years: perinatal life stress added ≈ 0.01 to explained variance; total model explained R^2 ext{ (social-emotional development)} \approx 0.22.
• Overall interpretation: modest effect sizes, but clinically meaningful long-term associations between maternal perinatal life stress and child development when following children to 2 years, supporting interventions during the perinatal period.

References to supplementary material

• Supplementary digital content (Supplemental Digital Content 1) contains flow chart details about attrition and the flow of participants through the study stages.