Urinary concentrations of phthalate/DINCH metabolites and body mass index among European children and adolescents in the HBM4EU Aligned Studies: A cross-sectional multi-country study

Abstract

Background: Phthalates are ubiquitous in the environment. Despite short half-lives, chronic exposure can lead to endocrine disruption. The safety of phthalate substitute DINCH is unclear.

Objective: To evaluate associations between urinary concentrations of phthalate/DINCH metabolites and body mass index (BMI) z-score among children and adolescents.

Method: We used Human Biomonitoring for Europe Aligned Studies data from 2876 children (12 studies, 6-12 years, 2014-2021) and 2499 adolescents (10 studies, 12-18 years, 2014-2021) with up to 14 phthalate/DINCH urinary metabolites. We used multilevel linear regression to assess associations between phthalate/DINCH concentrations and BMI z-scores, testing effect modification by sex. In a subset, Bayesian kernel machine regression (BKMR) and quantile-based g-computation assessed important predictors and mixture effects.

Results: In children, we found few associations in single pollutant models and no interactions by sex (p-interaction > 0.1). BKMR detected no relevant exposures (posterior inclusion probabilities, PIPs < 0.25), nor joint mixture effect. In adolescent single pollutant analysis, mono-ethyl phthalate (MEP) concentrations were associated with higher BMI z-score in males (β = 0.08, 95 % CI: 0.001,0.15, per interquartile range increase in ln-transformed concentrations, p-interaction = 0.06). Conversely, mono-isobutyl phthalate (MiBP) was associated with a lower BMI z-score in both sexes (β = -0.13, 95 % CI: -0.19, -0.07, p-interaction = 0.74), as was sum of di(2-ethylhexyl) phthalate (∑DEHP) metabolites in females only (β = -0.08, 95 % CI: -0.14, -0.02, p-interaction = 0.01). In BKMR, higher BMI z-scores were predicted by MEP (PIP=0.90) and MBzP (PIP=0.84) in males. Lower BMI z-scores were predicted by MiBP (PIP=0.999), OH-MIDP (PIP=0.88) and OH-MINCH (PIP=0.72) in both sexes, less robustly by DEHP (PIP=0.61) in females. In quantile g-computation, the overall mixture effect was null for males, and trended negative for females (β = -0.11, 95 % CI: -0.25, 0.03, per joint exposure quantile).

Conclusion: In this large Europe-wide study, we found age/sex-specific differences between phthalate metabolites and BMI z-score, stronger in adolescents. Longitudinal studies with repeated phthalate measurements are needed.