Micronutrients and Mood Disorders

: Using Mendelian Randomization to Explore Potentially Causal Nutritional Factors in Major Depressive Disorder

Abstract

Background
Major Depressive Disorder (MDD) is a leading cause of disability, and yet aetiological understanding remains limited. Treatments for MDD are often inaccessible, ineffective or poorly tolerated, making a recurrent episodic course the reality for many. Nutritional interventions could offer more acceptable and accessible options, particularly if targeted at underlying mechanisms. The discipline of Nutritional Psychiatry has been increasing in scientific credibility, although it remains challenged by parallel complexities of nutrition, and of psychiatry. This thesis aims to investigate nutritional risk factors for MDD using Mendelian Randomization (MR).

Methods
Two-sample MR was used to investigate the effect of micronutrient exposures on MDD and recurrent depression (rMDD) using GWAS summary statistics from the Psychiatric Genomics Consortium (PGC) MDD working group, (MDD_807k n=807,553, n_cases=246,363; MDD430k n=430,775, n_cases=116,209; rMDD_UKBiobank n=80,933, n_cases=17,451) and from the FinnGen consortium (rMDD_FinnGen, n=246,431, n_cases =19,388.)

Results
Genetic instruments for many vitamins and minerals were limited in strength, number or biological plausibility. However, MR methods to increase power by accounting for correlation between instruments identified possible protective effects of serum iron (OR_rMDD 0.90 per SD_iron increase; 95% CI 0.85, 0.95; p=3.0E^-4,) 25-hydroxyvitamin D (OR_rMDD 0.81 per SD_25(OH)D; 95% CI 0.66, 0.99; p=0.04), and erythrocyte
copper (OR_rMDD 0.97; per SD_copper; 95% CI 0.95- 0.99; p=4.1E^-4,) particularly in rMDD. Paradoxically adverse effects of serum selenium (OR_MDD430k 1.03 per SD_selenium; 95% CI 1.02, 1.05; p=3E^-4) and serum magnesium (OR_MDD430k 1.08 per SD_Magnesium; 95% CI 1.00, 1.08; p=0.06; and OR_rMDD 1.21 per SD_Magnesium; 1.01, 1.44;
p=0.04,) were less consistent and potentially driven by pleiotropy.

In contrast, genetic instruments for omega-3 fatty acids taken from the UK Biobank (n=115,078) provided ample power to detect a protective effect for omega-3 fatty acids on all MDD outcomes. The strongest overall evidence was for long-chain eicosapentaenoic acid (EPA) (OR_MDD430k 0.92 per SD_EPA; 95% CI 0.88 to 0.96; p=2.0E^-4; OR_{rMDDUKBiobank} 0.91 per SD_EPA; 0.77, 1.08; p=0.27; and OR_rMDDFinnGen 0.87 per SD_EPA; 95% CI 0.80, 0.96; p= 0.003.) Biological plausibility was supported by the lead variant located on the FADS2 gene, with evidence for genetic colocalisation in this region for our primary MDD sample (MDD_430k PP4_EPA 97.1%.) However, colocalisation for MDD_807k and rMDD_FinnGen, suggested distinct underlying variants, and horizontal pleiotropy was hard to rule out, as over 500 downstream traits were identified by Phenome Wide Association study (PheWAS).

In two-step MR, EPA reduced pro-inflammatory cytokines IL6_UKBiobank (𝛽_EPA -0.08; 95% CI -0.13, -0.03; p=7.1E^-4) and IL18_UKBiobank (𝛽_EPA -0.18; 95% CI -0.24, -0.12; p=2.5E^-9), colocalising with IL18_UKBiobank only (PP.H4_EPA 92.8%.) However, the effect of IL18_UKBiobank and IL6_UKBiobank on MDD_430k and rMDD_FinnGen was weak, as was evidence for inflammatory mediation using multivariable MR (MVMR.) This may reflect limitations of genetic instruments and methods, or it may highlight alternative mechanisms underlying the protective effects.

Conclusion
MR offers the potential to enhance research design and efficiency in nutritional psychiatry, by informing intervention development from the ground up, enabling insight into underlying pathophysiological mechanisms, and facilitating development of interventions to maximise benefit. While specifically identifying of a role for omega-3 fatty acids in MDD, this doctoral research highlights the potential for MR to uncover further nutritional risk factors in psychiatry, as sample sizes and methods evolve. In
combination with other methodologies, MR may offer the opportunity to optimise resources to benefit population mental health.

Date of Award	9 Dec 2025
Original language	English
Awarding Institution	University of Bristol
Supervisor	Maria C Borges (Supervisor), Jonathan Evans (Supervisor) & Richard M Martin (Supervisor)