By Nathan
For years, researchers have known that severe conflict between parents can ripple through a household in ways that harm children. But the mechanism has been harder to pin down. A new study from Shanxi University offers the strongest evidence yet that in Chinese families, the damage travels through a specific psychological pathway: parental conflict breeds emotional insecurity in children, which undermines their sense of well-being, and that eroded well-being in turn disrupts their sleep.
What they found
The research team, led by Hongxia Hou, recruited 900 upper-grade primary school children from three public schools in northern China. The children, averaging 10.58 years of age, completed validated questionnaires measuring their perceptions of interparental conflict, emotional insecurity, subjective well-being, and sleep quality. This quantitative core was then enriched by semi-structured interviews with 23 of the children, giving the researchers a window into how children themselves make sense of family tension.
The headline result was clear: children who reported higher levels of interparental conflict also reported poorer sleep quality. The bivariate correlation was moderate and statistically robust (r=0.39, p<0.001), meaning that roughly 15 percent of the variation in children’s sleep quality could be accounted for by their exposure to parental conflict alone.
But the real contribution of the study lies in what happens when you introduce the intermediary variables. The researchers tested a serial mediation model with two intervening factors: emotional insecurity (the child’s subjective feeling of threat, instability, or worry arising from parental discord) and subjective well-being (a broad measure of how positively a child experiences their daily life, including happiness, life satisfaction, and positive affect).
The results supported the model. The direct path from interparental conflict to poorer sleep quality shrank substantially once emotional insecurity and subjective well-being were included as mediators, dropping to a beta weight of 0.11 (p=0.006). That may still be statistically significant, but the practical interpretation is that most of the original effect was indirect. Specifically, the indirect path through emotional insecurity alone accounted for a small but meaningful 0.014 of the total effect. The path through subjective well-being alone accounted for 0.019. And the full serial pathway (conflict leading to emotional insecurity, insecurity leading to lower well-being, and lower well-being leading to poorer sleep) contributed 0.004.
Taken together, these indirect routes indicate that interparental conflict damages sleep not primarily through some direct, brute-force mechanism, but by destabilizing a child’s inner emotional world and, through that instability, diminishing their overall experience of well-being. Sleep, in this model, is the downstream casualty of a cascade that begins with a threatened sense of safety.
The interview data added texture to the numbers. In the Chinese family context, children did not simply report feeling “upset” about conflict in the abstract. They interpreted parental arguments through culturally specific lenses: worries about the breakdown of family harmony, concerns about the quality of the parent-child bond, and a felt sense of relational responsibility. Some children described feeling that they needed to intervene or mediate, or that conflict placed them in a loyalty bind. This emotional labor, the study suggests, compounds the insecurity generated by the conflict itself, creating an ongoing cognitive and emotional burden that erodes well-being and, eventually, the ability to sleep restfully.
Why it matters
Poor sleep in childhood is not a trivial complaint. It is linked to impaired cognitive function, emotional dysregulation, behavioral problems, and long-term physical health consequences including obesity and cardiovascular risk. If parent-child dynamics are upstream drivers of childhood sleep problems, then intervening on the family environment may be more effective than targeting sleep directly with hygiene advice or melatonin supplements.
The study’s mediation model offers a precise target for intervention. If emotional insecurity is the first step in the cascade, then helping children feel safe and stable in the face of parental conflict may interrupt the chain before well-being erodes. This might involve explicit reassurance from parents that the conflict is not the child’s fault or responsibility, maintaining consistent routines and warmth even during periods of marital tension, and giving children age-appropriate strategies to distance themselves from adult arguments rather than feeling compelled to mediate.
At the same time, the finding that subjective well-being independently mediates the path suggests that building positive emotional resources in children is protective irrespective of the conflict level. Schools and families that invest in children’s happiness, social connection, and sense of mastery may be indirectly improving their sleep even when home tensions cannot be fully resolved.
The Chinese context is especially important here. In many Chinese families, the cultural value placed on family harmony can make open conflict feel particularly threatening to children, because the discord is not merely interpersonal but also a violation of an expected social order. The children in Hou’s study described exactly this: they felt that conflict signaled something wrong not just between parents but with the family unit itself. Interventions developed for Western populations may therefore need cultural adaptation to be effective in Chinese settings.
Limits
The study design is cross-sectional, which means the researchers measured all variables at a single time point. This is a fundamental constraint: while the proposed serial mediation model is theoretically grounded and statistically plausible, the data cannot prove temporal causation. It remains possible that children who sleep poorly for other reasons perceive their parents’ relationship more negatively, or that a third unmeasured variable (such as family socioeconomic stress or parental mental health) drives both conflict and poor sleep.
The sample draws from three schools in one region of northern China, limiting generalizability to Chinese children from other provinces, rural areas, or different socioeconomic backgrounds. The reliance on child self-report for all measures is both a strength (since sleep quality and emotional experience are inherently subjective) and a limitation (shared method variance may inflate the associations). The authors did not include objective sleep measurement such as actigraphy or parent-reported sleep problems that could corroborate the children’s accounts.
Finally, the mediating effects, while statistically significant, were small in absolute magnitude. The serial mediation pathway, for instance, contributed only 0.004 to the total effect. This suggests that emotional insecurity and subjective well-being are part of the picture but far from the whole story, and the practical clinical significance of such small indirect paths warrants cautious interpretation.
Bottom line
Interparental conflict does not simply disturb a child’s sleep through noise or disruption. It works through the child’s inner world: first generating emotional insecurity, then eroding subjective well-being, and only then harming sleep. For Chinese families, where conflict carries culturally specific meanings around family harmony and relational responsibility, parents and practitioners should attend to children’s emotional safety and daily happiness as modifiable pathways to better sleep. The study provides a well-reasoned model that, if confirmed in longitudinal research, could guide family-based sleep interventions in Chinese and other collectivist cultural contexts.
Source
Hou H, Guan Y, Luo T. The relationship between interparental conflict and sleep quality among Chinese primary school children: a serial mediation model of emotional insecurity and subjective well-being. BMC Psychology. 2026. PMID: 42469886. DOI: 10.1186/s40359-026-05152-7.