Does Age make a Difference in Testing Gene-Environment Interactions

• In Event: 2-038 - New Directions in Research on (Epi)Gene-Environment Interplay in the Development of Psychopathology

Integrative Statement

Gene-environment interactions (GxE) are assumed to play an important role in the development of an individual, in particular child and adolescents. From a behavioural perspective, the presence of developmental periods is clear and widely described in literature. In biology overall, a similar paradigm is applied, with puberty one of the best examples of age and development related changes to the body. In genetics, however, this topic has received less attention. Embryological research focuses on genes activated in a particular time window, but in childhood and adolescence this has not been explored from that perspective. Furthermore, ongoing changes in the environment and changes following from the developmental stage can also impact GxE results. The nature of the influence of age on gene-environment interactions remains unknown. The preliminary results presented, aim to increase the understanding of role of age and development on GxE. We hypothesised that effect sizes and significance for gene-environment interaction can differ by age for each SNP.
Cross-sectional analyses were performed on a subset (N=927) of the STRATEGIES sample (N=1,116). Individuals were selected based on caucasian descent and age. Three groups of 1 year span were formed with the 12.30 and 15.30yrs as lower and upper cut-off: 12.30yrs(N=313), 13.30yrs(N=369), and 14.30yrs(N=245). Based on our previous work that urged better understanding of the role of age in GxE for the gene SLC6A11, we selected this gene for further examination. Genotyping was performed using an Illumina iSelect Custom chip resulting in 5052 SNPs after quality control, from which we only selected the 65 SNPs of the gene of interest. Subscales corresponding to adolescents’ perceived support from the Leuven adolescent perceived parenting scale (LAPPS) and the Short Scale for Parenting Behavior (VSOG) were used as environmental factor of interest. Adolescent reported depressive symptoms (CES-D scale; Center for Epidemiologic Studies Depression) was the outcome in the analyses. GxE analyses were run as regression analyses per SNP, corrected for gender. Groups were not statistically different with regard to the outcome and sample size. A decline in perceived support by age was noted. ANOVA was used at the meta-level to distinguish characteristics of GxE beta and p-value distributions by age.
At the significance level applied for SNPs, corrected for 65 SNPs tested in three age groups (p < 2.56e-04), only the youngest group (age between 12.30 and 13.30) showed a significant interaction: rs377410 (estimate: -0.48, p=2.11e-04, F(305,6)=10.30). The number of SNPs below p<0.05 in this age group is higher (22 SNPs) than the other two age groups (13.30-14.30: 2 SNPs, 14.30-15.30: 0 SNPs). An ANOVA showed that p-value distribution was different over the three groups (ANOVA F(2,192)=11.26, p= 2.37e-05). This was also the case for the estimates (ANOVA F(2,192)=4.77, p=0.0095). Ranked p-values show that the first 9% are solely from the youngest group. The oldest group enters only at the 19th percentile. The results from this study suggest that age stratification can indeed be relevant in evaluations of gene-environment interplay. The impact of genes can change by age (sensitive periods).

Authors

• Evelien Van Assche, KU Leuven, University
  Presenting Author

• Tim Moons, Leuven University
  Non-Presenting Author

• Luc Goossens, KU Leuven
  Non-Presenting Author

• Stephan Claes, KU Leuven
  Non-Presenting Author