The authors investigated whether the epidemiologic association between neuroticism — a personality trait marked by frequent and intense negative emotions — and Alzheimer's disease (AD) reflects a direct causal relationship or is explained by intervening health and behavioral factors. Two main questions guided the study: (1) is the neuroticism–AD relationship causal when assessed with genetic instruments, and (2) to what extent do modifiable risk factors mediate any observed association?
To address causality, the study applied a two-sample Mendelian randomization (MR) framework using publicly available genome-wide association study (GWAS) summary statistics. Genetic instruments for neuroticism were taken from Nagel et al. (2018), and outcome data for AD were taken from Bellenguez et al. (2022). The MR approach leverages genetic variants associated with the exposure (neuroticism) as instrumental variables to estimate a genetically informed effect on the outcome (AD), reducing confounding and reverse causation compared with conventional observational analyses. The authors report that MR analysis suggested the neuroticism–AD relationship is unlikely to be causal.
Mediation analyses were conducted using a subset of UK Biobank participants aged 60 and older who had complete data on neuroticism and the candidate mediators. The analytic sample included 121,825 controls (mean age = 63.9 ± 2.81; 61,993 females, 50.9%) and 1,277 individuals with AD (mean age = 65.6 ± 2.71; 628 females, 49.2%). All participants in this subset had complete information on neuroticism and the potential mediating variables evaluated by the authors.
The authors evaluated multiple health and behavioral factors as potential mediators of the association between higher neuroticism and increased AD risk. Three mediators reached statistical significance in the mediation models:
Depression: β = 0.048, p = 3 × 10−4.
Hypertension: β = 0.005, p = 2 × 10−4.
Alcohol consumption: β = 0.001, p = 1 × 10−5.
These findings indicate that a portion of the association between elevated neuroticism and AD in the observational data can be statistically attributed to these modifiable factors. The magnitudes of the reported beta coefficients and their p-values are those provided by the authors; the preprint does not report additional quantitative decomposition (for example, proportion mediated) in the abstract.
Combining MR and mediation results, the authors interpret the evidence as inconsistent with a direct genetic causal effect of neuroticism on AD. Instead, the association observed in phenotypic data appears to be largely explained by co-occurring, potentially modifiable health and behavioral conditions, specifically depression, hypertension, and alcohol consumption. In short, neuroticism may increase AD risk indirectly by increasing the prevalence or severity of these mediating conditions rather than acting as an independent causal factor.
The study highlights modifiable targets that could, if causally implicated by further research, be prioritized for intervention to reduce AD risk among individuals with high neuroticism scores. Specifically, better detection and treatment of depression, management and control of hypertension, and addressing harmful patterns of alcohol consumption are identified as pathways that may mediate increased AD risk in people with higher neuroticism. The authors note that understanding these mediating pathways may inform targeted prevention and therapeutic strategies aimed at reducing AD risk.
This work is reported as a preprint and has not undergone peer review; the authors explicitly caution that it should not be used to guide clinical practice. Details on study approvals are provided: UK Biobank received ethics approval from the Northwest Multi-Centre Research Ethics Committee, participants provided written consent, and data access was under application #61530 with the Centre for Addiction and Mental Health as lead institution. The authors declare no competing interests. All data produced in the study are available upon reasonable request to the authors. The abstract does not report additional sensitivity analyses, instrument strength metrics, or full mediation model parameters; those details were not reported in the abstract and would require consultation of the full preprint.
Conclusion
In this combined genetic and observational analysis, a two-sample MR framework did not support a direct causal effect of neuroticism on AD. Mediation analysis in a large UK Biobank subset identified depression, hypertension, and alcohol consumption as significant mediators, suggesting the phenotypic link between neuroticism and AD may be driven mainly by these modifiable factors. Further peer-reviewed research is needed to confirm these findings and to quantify mediation effects more fully.