Residential noise exposure and children's emotional and behavioural difficulties in France: a birth cohort study

Over 20% of Europe's population is chronically exposed to transport noise levels considered detrimental to health, yet evidence on its impact on children's mental health remains limited. We aimed to investigate the effect of different sources of residential noise exposure on children's emotional and behavioural difficulties.

In this birth cohort study, data from the nationally representative French ELFE cohort were analysed from birth to age 10 years (2011-22). We excluded children with missing data on the study's exposures from age 2 months to 5·5 years and on outcome variables at age 10·5 years, twins, children in foster care, and children with profound hearing impairment. Residential exposure to road traffic, aircraft, and railway noise was assessed at age 2 months, 3·5 years, and 5·5 years using GeoClimate and NoiseModelling open-source software and public geographical, traffic, and noise map data. Children's mental health was assessed using the parent-reported Strengths and Difficulties Questionnaire at age 10·5 years. We applied marginal structural models with inverse probability weighting to estimate the effect of time-varying noise exposure through childhood.

Of the 7914 children included, 4044 (51·1%) were male and 3870 (48·9%) were female. At age 2 months, 2480 (31·3%) children were exposed to low levels (<55 A-weighted decibels [dB(A)]) of road traffic noise exposure, 2772 (35·0%) were exposed to medium levels (55-<60 dB[A]), and 2662 (33·6%) were exposed to high levels (≥60 dB[A]). Persistent exposure to high levels of road traffic noise (≥60 dB[A] across childhood) was associated with higher levels of internalising symptoms at age 10·5 years (adjusted regression coefficient β=0·38, 95% CI 0·14 to 0·63; p=0·003), particularly emotional symptoms (β=0·20, 0·02 to 0·37; p=0·026) and peer relationship problems (β=0·19, 0·06 to 0·32; p=0·009). No significant effects were detected for high exposure to road traffic noise at age 3·5 years. At age 2 months, 161 (2·0%) children were exposed to railway noise of 50 dB(A) or higher. Exposure to railway noise (at least once to ≥50 dB[A] during childhood) was associated with lower levels of internalising symptoms (β=-0·61, 95% CI -1·10 to -0·12; p=0·008), although this association was no longer significant after correction for multiple testing. Exposure to aircraft noise was low, with 185 (2·3%) children exposed to 50 dB(A) or higher at age 2 months, and no significant associations were observed between exposure to aircraft noise and children's emotional and behavioural difficulties.

Although residual confounding cannot be entirely excluded, our results support the need for policies aimed at reducing residential road traffic noise. They require replication in other large-scale cohorts and further investigation of the underlying mechanisms is needed.

None.