BACKGROUND: Few family interaction processes are more detrimental to children's health than family conflicts. Conflictual relationships in childhood predict a host of adverse health outcomes across the life course. The current study examines associations between household conflicts and the health of children aged 6-12 years and explores to which extent this may vary by socioeconomic status (SES) and multimorbidity (MM) in the household.

METHODS: Cross-sectional study using questionnaire data gathered between 2016 and 2020 as part of the Lolland-Falster Health Study (LOFUS) combined with routine register data on health care use and socio-demography from the Danish nationwide administrative databases. The study sample consisted of 1065 children 6-12 years old, who answered LOFUS4 or LOFUS11, from 777 households for which at least one adult answered LOFUS18. Main outcome was children's health complaints, defined as headache, abdominal pain, back pain, and sleep difficulties. Covariates included MM, SES, and conflicts, all three measured at household level. Multivariable logistic regression models were used.

RESULTS: Conflicts were negatively associated with children's health. This was most pronounced for general conflicts in the household, with increased complaints of abdominal pain, back pain, and sleep difficulties. The associations varied when we stratified the households according to MM and SES. Significant associations were found within households without MM for abdominal pain, and within households with MM and low SES, and without MM and with high SES for sleep difficulties. While the higher level of abdominal pain for the above indicated households were found for both internal and external conflicts in the household, the higher level in sleep difficulties was mostly driven by internal conflicts.

CONCLUSION: Children reporting frequent health complaints have a higher future health care use compared with children without such complaints. Our results indicate that growing up in a household with a high conflict level might be a predisposing factor.