PSYREFLECT
RESEARCHNovember 9, 20262 min read

A mother's one or two leftover earthquake symptoms already register in her toddler's behaviour

Key Findings
  • A Japanese birth cohort of 21,656 mother–child pairs looked at women who were not pregnant during the 2011 Great East Japan Earthquake and gave birth in the following years.
  • One year after delivery, 11,297 mothers with exposure data reported no earthquake-related traumatic stress symptoms, 1,489 reported one or two, and 332 reported three or four.
  • Compared with no symptoms, one or two symptoms were associated with higher odds of the child exceeding the CBCL cutoff: at age 2, OR 1.34 for internalizing, 1.29 for externalizing and 1.38 for total problems; at age 3, OR 1.38, 1.21 and 1.37.
  • Three or four symptoms showed stronger associations, for example OR 1.87 (95% CI 1.29–2.71) for internalizing problems at age 2; a test for trend was significant for 4 of the 6 outcomes.

The children in this cohort were born years after the March 2011 earthquake and tsunami, on average more than five, to women who were not pregnant when it struck. A team from Tohoku University asked whether the mothers' remaining disaster-related symptoms still showed in their children's behaviour at two and three.

Four questions a year after birth

The data come from the TMM BirThree birth cohort: pregnant women recruited at 50 obstetric clinics and hospitals in disaster-affected areas of north-eastern Japan between July 2013 and March 2017. Parents with depression or PTSD before the earthquake were excluded, leaving 21,656 pairs. One year postpartum, mothers answered four yes/no questions about the past week: unwanted memories or dreams of the disaster, distress at reminders, physical reactions, deliberate avoidance. Mothers then rated the child on the CBCL/1½–5, and the clinical cutoff served as the outcome.

The models adjusted for current maternal distress (K6), social isolation, insomnia, housing damage, income and other factors. With one or two symptoms, the odds of crossing the cutoff were higher for every scale at both ages, from OR 1.21 to 1.38. With three or four, they rose further for most outcomes. In mothers with severe current distress (K6 ≥13, n = 1,156) the link did not reach significance; in the rest it was present.

Where the threshold sits

My reading is that the group to watch here is the 1,489 women, not the 332. One intrusive memory, one deliberate detour around reminders – a mother like that passes a routine intake as well, and no one refers her anywhere. Her child, though, crossed the CBCL cutoff for total problems more often: 8.80% against 5.61% at age 2.

For perinatal and early-childhood work, this argues for asking a parent about past disasters and losses whatever the distance in years, and for recording a single persistent symptom as information about the child's environment, not as a pass.

Mothers who reported only one or two lingering earthquake symptoms, a level that rarely leads to referral, had children who crossed the clinical cutoff for behaviour problems more often at ages two and three.

Limitations

The study is observational. The four symptom items were developed for this cohort and have not been externally validated, and child behaviour was rated by the mothers themselves.

Source
Clinical and Experimental Pediatrics
Association between maternal traumatic stress related to Great East Japan Earthquake and behavioral problems in offspring
2026-10-22·View original ↗
Tags
  • Japan
  • intergenerational trauma
  • disaster mental health
  • maternal traumatic stress
  • early childhood behaviour
Related
Industry
Japan's methylphenidate registry slowed prescribing for under-20s far more than for adults
Journal of Pharmaceutical Health Care and SciencesRead →
Industry
California screened three million Medi-Cal members for childhood adversity, and parents were mostly left out
ACEs Aware (California Department of Health Care Services / Office of the California Surgeon General)Read →
Research
Rwandans conceived through genocidal rape show the largest shift in newer epigenetic clocks
Communications MedicineRead →
PsyReflect · Free · Mon & Thu
Get analyses like this every Monday and Thursday.
Only what matters for practice. Curated by a clinical psychologist. 5 minutes instead of 4 hours of monitoring.
← Previous
The 12-item ProQOL holds together better than the 30-item original
Next →
Japan's methylphenidate registry slowed prescribing for under-20s far more than for adults