Do adverse childhood experiences link to pre-pregnancy stress across generations?

Two generations, two childhoods, one pre-pregnancy window

The study asked whether adversity in a mother's childhood shows up as stress in her daughter's adult life just before pregnancy, and whether the daughter's own childhood adversity explains that link. It drew on 2,235 biologically linked mother-offspring pairs from the Grandmothers, Mothers, and Their Children's Health (GMATCH) study, surveyed in a 2023 follow-up; seven pairs verified as biologically unrelated were excluded from the 2,242 initially enrolled.

Both generations' ACEs were measured retrospectively with the Childhood Trauma Screener, which asks about experiences before age 18 across five subtypes: emotional abuse, physical abuse, sexual abuse, emotional neglect, and physical neglect. Any "yes" counted as ACE exposure, and scores ran from 0 to 5. The outcome belonged only to the younger generation: five yes/no items about the 12 months before pregnancy covering intense conflict with a spouse, serious conflict with family or friends, physical/psychological/sexual abuse, death of a relative or friend, and significant financial problems. Any reported stressor counted as pre-pregnancy stressful life events (pre-PSLEs).

That ordering matters for reading every number below. Maternal ACEs are the exposure, offspring ACEs the proposed intermediate, and offspring pre-PSLEs the outcome, sequenced by when each was experienced rather than by when it was reported.

ACEs were reported by 18.17% of mothers and 14.27% of offspring

Among the 2,235 pairs, 406 mothers (18.17%) and 319 offspring (14.27%) reported at least one ACE. Physical neglect was the most prevalent subtype among mothers, emotional neglect among offspring. Pre-PSLEs were considerably more common than ACEs: 659 offspring (29.49%) reported at least one, with financial stress and the death of a friend or relative tied as the most frequent single stressors at 12.21% each.

The authors note that these ACE prevalences are lower than in several comparison studies, including all-female studies in Iceland and Uganda and German estimates using the same screener. They offer three explanations: cultural norms in China that value enduring adversity, stigma and concern about family reputation, and differences in item counts and operational definitions across studies. These are the authors' interpretations of a prevalence gap, not tested findings, and they matter because underreporting would tend to weaken rather than inflate the observed associations.

Both generations' ACEs raised the odds, with a dose-response

In multivariable logistic regression adjusted for sociodemographic factors in both generations plus the offspring's pre-pregnancy anxiety or depression, maternal ACEs were associated with 1.79-fold higher odds of offspring pre-PSLEs (95% CI 1.41–2.27). Offspring ACEs showed a larger association, at 2.23 (95% CI 1.73–2.89). These are odds ratios for reporting any pre-PSLE, not counts of stressors or severity scores.

Both exposures followed a dose-response gradient. For maternal ACEs, one ACE gave an OR of 1.44 (1.07–1.92) and two or more gave 2.55 (1.79–3.64); for offspring ACEs the corresponding figures were 1.57 (1.13–2.18) and 3.58 (2.47–5.24), with trend tests below 0.001 in both cases. The paper reports that patterns held across all ACE subtypes, though the subtype-specific estimates vary widely in precision — maternal sexual abuse, for instance, rests on only 14 exposed mothers (OR 3.75, 95% CI 1.24–11.84).

Maternal ACEs tracked partner conflict, family conflict and money

Breaking the outcome into individual stressors shows where the cross-generation signal sits. Maternal ACEs were associated with higher odds of conflict with partners (OR 1.82, 95% CI 1.23–2.65), conflict with family members (OR 1.62, 1.13–2.28), and financial stress (OR 1.64, 1.19–2.25). Offspring ACEs, by contrast, were associated with all assessed stressors.

This is a comparison within one paper, not a claim that maternal ACEs spare some domains. The maternal associations are estimated separately per stressor, and the paper does not report a formal test of whether the maternal effect differs across stressors. What the pattern does support is that the cross-generation association is not carried by a single item such as financial hardship alone.

Offspring ACEs carried 79.64% of the maternal association

The mediation model is the paper's central mechanistic claim. Using structural equation modeling with acceptable fit (χ²/df = 2.05, RMSEA = 0.022, CFI = 0.993), the standardized indirect effect of maternal ACEs on offspring pre-PSLEs through offspring ACEs was 0.266 (95% CI 0.186–0.354), against a total effect of 0.334 (0.221–0.431). Dividing the indirect by the total gives the reported 79.64% mediated.

Read carefully, that figure describes the share of the total association accounted for by the mediator path in this model, not a share of cases or a probability that a given daughter will experience stress. The remaining roughly 20% is a direct path the model still estimates. The practical implication the authors draw is that the intergenerational link is largely reproduced through the next generation's own childhood exposure — which is why they argue preconception care should not focus only on stress during pregnancy.

Retrospective reports and one region limit the inference

All ACE and pre-PSLE data were self-reported retrospectively, which the authors say may introduce recall bias and lead to underestimation. The pre-PSLE instrument was a brief, study-specific questionnaire without formal validation and covered only five stressors, so the outcome is a coarse any/none indicator rather than a validated severity measure.

The authors also flag statistical power: although the cohort is described as modest in size, the number of women reporting pre-PSLEs was relatively small, which may produce imprecise or chance estimates — a concern that applies most acutely to the sparse subtype cells. Finally, all participants came from Huai'an, Jiangsu Province, with its own sociodemographic profile, limiting generalization to other regions or populations. Subgroup and sensitivity analyses (excluding 35 dyads with missing values and 9 offspring diagnosed with anxiety or depression before pregnancy) are reported as supporting robustness, but they do not remove these measurement constraints.

Screening should target the daughter's own childhood adversity

The most decision-relevant result is the comparison between the two exposures: offspring ACEs carried the stronger association with pre-PSLEs (OR 2.23 versus 1.79) and absorbed most of the maternal association. If a preconception or perinatal setting were to use brief childhood-adversity assessment to flag women at higher risk of substantial psychosocial stress before pregnancy — the use the authors propose — the daughter's own ACE history is the more proximate marker in this cohort, while maternal history adds context about how that risk arose.

The dose-response gradients sharpen this: the offspring estimate rises from 1.57 with one ACE to 3.58 with two or more, so the number of adversities, not merely their presence, separates higher- from lower-odds groups. The paper's own framing follows the same priority — early identification and intervention for ACEs within pre-pregnancy health management, paired with family- and community-based prevention of childhood adversity rather than attention to pregnancy stress alone.

Sources used in this answer