- Significant inverse genetic correlation between zinc blood levels and treatment-resistant depression (rg = -0.20, p = 4.52 × 10^-3); non-significant trend for TRS.
- Copper blood concentration correlated positively with appetite and weight gain neurovegetative symptomatology (rg = 0.11, p = 0.0026).
- Mendelian randomisation and polygenic score analyses showed no causal or predictive effects, suggesting limited power, residual confounding, or alternative shared architectures.
Am J Med Genet B Neuropsychiatr Genet. 2026 Aug 19. doi: 10.1002/ajmg.b.70030. Online ahead of print.
ABSTRACT
Trace elements are pivotal to key biological processes, with possible effects on psychopathology. We investigated the hypothesis of shared genetic factors between trace elements levels, treatment resistance and related traits. We used genome-wide summary statistics for trace elements blood concentration, treatment-resistant depression (TRD), treatment-resistant schizophrenia (TRS), and related traits, including anhedonia, neurovegetative symptoms, suicidal ideation, and cognition. Cross-trait genetic correlations (rg) were estimated with High-Definition Likelihood with Bonferroni correction (α = 0.0031). Significant rg were tested by two-sample Mendelian randomization (MR) and polygenic scores (PGSs) in UK Biobank. We found an inverse rg between zinc and TRD (rg = -0.20, SE = 0.07, p = 4.52 × 10-3), and a consistent non-significant trend for TRS (p = 0.06). Copper correlated positively with appetite/weight gain (rg = 0.11, SE = 0.036, p = 0.0026). MR provided no evidence of causal effects of zinc levels on TRD or copper levels on appetite/weight gain, possibly reflecting weak instruments. The PGS of zinc was not associated with TRD, nor the PGS of copper with appetite/weight gain, in UK Biobank (n = 15,927 and n = 35,849, respectively). Our results suggested genetic correlations involving zinc and TRD, copper and neurovegetative symptomatology, without support for a causal model or predictive effect of the respective PGSs. This may suggest limited power, residual confounding, or alternative shared genetic architectures.
PMID:42619169 | DOI:10.1002/ajmg.b.70030
Share Evidence Blueprint
Save to Google Notes

Search Google Scholar
Save as PDF
⭐ My Revision List

