A recent association study identified three single nucleotide polymorphisms (SNPs) (rs1538979, rs821577 and rs821633) in Disrupted in Schizophrenia 1 (DISC1) gene that, either independently or in combination, influence risk of schizophrenia and bipolar disorder.1 We investigated the association between these SNPs and the personality trait emotional stability/neuroticism (a potential endophenotype for mental illness), and anxiety and depression in two relatively healthy older Scottish cohorts (Lothian Birth Cohort of 1921 (LBC1921),2 n=360; and Lothian Birth Cohort of 1936 (LBC1936),3 n=1065).
Personality was assessed at a mean age of ∼81 years (LBC1921) or ∼70 years (LBC1936). Emotional stability was measured using the IPIP Big-Five 50-item inventory4 and, for LBC1936 only, neuroticism was assessed using the NEO five-factor inventory.5 Mood was assessed at a mean age of ∼79 years (LBC1921) or ∼70 years (LBC1936) using scores on the Hospital Anxiety and Depression scale (HADS).6 LBC1921 was genotyped on a Sequenom platform (Sequenom, San Diego, CA, USA)1 and LBC1936 using TaqMan (Applied Biosystems, Pleasonton, CA, USA) at the Wellcome Trust Clinical Research Facility, Edinburgh. Genotype frequencies did not differ significantly from the Hardy–Weinberg equilibrium (P>0.01) and were LBC1921 (LBC1936); rs1538979, C/C=262 (749); T/C=80 (230); T/T=10 (29), rs821577, G/G=61 (193); G/T=163 (492); T/T=118 (351) and rs821633, C/C=35 (94); C/T=128 (429); T/T=175 (483). Phenotypic outliers with z-scores greater than ±3 were removed before analyses. Descriptive statistics of the phenotypic variables are listed in the Supplementary Table. General linear modelling was performed in SPSS v13 (SPSSinc, Chicago, IL, USA) using carrier status (as defined in Hennah et al. (2008)1) for each SNP or combination of SNPs and sex as fixed factors. As previous studies have shown that variance in DISC1 has stronger effects in females than in males,1 the models were also run analysing males and females separately. LBC1921 and LBC1936 were first analysed separately and then, for HADS and emotional stability, combined analyses were performed with cohort added to the models as a fixed factor.
In LBC1921, female carriers of the rs821577 risk allele had significantly higher HADS anxiety scores than did non-carriers (P=0.044, η2=0.021). This association was replicated in LBC1936 (P=0.033, η2=0.009), and became more significant when data from the two cohorts were combined (P=0.004, η2=0.0012) (Table 1). In the combined analysis, male carriers of both the rs821577 and rs821633 risk alleles had significantly lower HADS anxiety scores (P=0.016, η2=0.009). Female carriers of the risk alleles had significantly higher HADS anxiety scores (P=0.021, η2=0.008). LBC1921 participants who were ∼9 years older than LBC1936 had significantly higher HADS depression scores (P<0.05 for all models). This is likely to be partly owing to them being in poorer health because of being older. In LBC1921, female carriers of both the rs821577 and rs821633 risk alleles had higher HADS depression scores than did non-carriers (P=0.032, η2=0.024). This was not replicated in LBC1936 (P=0.434), but was significant in a combined analysis of both cohorts (P=0.015, η2=0.009) (Table 1). In LBC1936, IPIP emotional stability scores and NEO neuroticism scores were highly inversely correlated (r=−0.76, n=939, P<0.001). In the combined cohort analysis, female carriers of the rs821577 risk allele had lower emotional stability scores than did non-carriers (P=0.024, η2=0.008). In LBC1936, female carriers of both the rs821577 and rs821633 risk alleles had significantly lower emotional stability (P=0.038, η2=0.010) and higher neuroticism (P=0.013, η2=0.014) scores than did non-carriers. In the combined cohort analysis, female carriers of these two risk alleles had significantly lower emotional stability scores (P=0.015, η2=0.010).
We have shown that rs821577, either independently or in the presence of the risk allele for SNP rs821633, is associated with increased anxiety and depression and with higher levels of neuroticism, in women, in two cohorts of older normal individuals. These risk alleles have previously been associated with schizophrenia, bipolar disorder, and social and physical anhedonia, mainly in females.1, 7 Reports suggest that heritability of neuroticism, anxiety and depression is higher in females than in males, and that the genes involved differ between men and women.8, 9, 10 We tested SNPs and models specifically chosen on the basis of earlier evidence, and hence do not believe that stringent multiple testing corrections are appropriate, but these results need to be replicated in other suitable cohorts before variation in DISC1 is fully accepted as contributing to normal variation in neuroticism and mood.
Conflict of interest
The authors declare no conflict of interest.
About this article
Journal of Affective Disorders (2014)