Early childhood serum and urine metabolite signatures of internalizing, externalizing, and ADHD traits and their trajectories to adolescence: the Rhea mother–child cohort
INTRODUCTION: Early-life emotional and behavioral difficulties increase vulnerability to later psychopathology, but it remains unclear whether a biological footprint is detectable in early childhood. OBJECTIVES: We aimed to identify serum- and urine-based metabolite signatures of internalizing, externalizing, and attention-deficit/hyperactivity disorder (ADHD) traits in 6-year-old children and examine their associations with traits through adolescence. METHODS: We analyzed 196 participants of the Rhea cohort (Greece), with serum and urine metabolomics at age 6 and internalizing, externalizing, and ADHD traits assessed at ages 6, 11, and 15. Trait- and biofluid-specific metabolite signatures were identified using elastic net regression with cross-validation. Associations with subsequent traits at repeated time points were evaluated using confounder-adjusted linear mixed-effects models. RESULTS: Metabolite signatures comprising amino acids, acylcarnitines, phosphatidylcholines, and gut microbiota-related metabolites were identified for internalizing, externalizing, and ADHD traits. Serum metabolite scores were associated with increased internalizing (β = 0.40; 95% CI 0.14-0.68) and externalizing traits (β = 0.83; 95% CI 0.56-1.1). Urine metabolite scores were associated with internalizing (β = 0.42; 95% CI 0.16-0.68) and externalizing traits (β = 0.69; 95% CI 0.43-0.95), while the association with ADHD traits was nominal (β = 0.25; 95% CI: 0.006-0.502; nominal p = 0.045; BH-FDR p = 0.054). Associations were stable over time for internalizing and ADHD but weakened with age for externalizing traits. CONCLUSIONS: This proof-of-concept study shows that complementary serum and urine metabolomic profiles in early childhood are associated with subsequent internalizing and externalizing trait trajectories, while a nominal longitudinal association was observed between the urine metabolite profile and ADHD traits. These findings warrant evaluation in larger independent studies.
Authors
- Chrysi Mouatsou
- Konstantinos K. Tsilidis (ORCID: https://orcid.org/0000-0002-8452-8472)
- Anastasia Chrysovalantou Chatziioannou (ORCID: https://orcid.org/0000-0002-1973-7542)
- Manolis Kogevinas (ORCID: https://orcid.org/0000-0002-9605-0461)
- Katerina Koutra (ORCID: https://orcid.org/0000-0002-7478-8844)
- Hector C. Keun (ORCID: https://orcid.org/0000-0001-7358-8851)
- Christopher Papandreou (ORCID: https://orcid.org/0000-0002-6803-507X)
- Katerina Margetaki (ORCID: https://orcid.org/0000-0001-9955-1985)
- Marianna Karachaliou (ORCID: https://orcid.org/0000-0002-3014-0747)
- Martine Vrijheid (ORCID: https://orcid.org/0000-0002-7090-1758)
- Léa Maître (ORCID: https://orcid.org/0000-0003-3682-7117)
- Mariza Kampouri (ORCID: https://orcid.org/0000-0002-3359-5691)
- Ελένη Μπάρμπα
- Ariadni Konopidou
Institutions
- University of Crete (GR)
- Universitat Pompeu Fabra (ES)
- University of Ioannina (GR)
- Aristotle University of Thessaloniki (GR)
- Erasmus MC (NL)
- Hellenic Mediterranean University (GR)
- Institut de Recerca Biomèdica Catalunya Sud (ES)
- Barcelona Institute for Global Health (ES)
- Centro de Investigación Biomédica en Red de Epidemiología y Salud Pública (ES)
- Centre international de recherche sur le cancer (FR)
- Imperial College London (GB)
Publication Details
- Journal
- Metabolomics
- Published
- 2026-10-30
- DOI
- https://doi.org/10.1007/s11306-026-02550-1
- Primary Topic
- Metabolomics and Mass Spectrometry Studies
- Type
- article
- Field-Weighted Citation Impact
- 0.00