Geoffroy · Scientific reports 2015 · candidate gene association study · n=269

Genetic association study of circadian genes with seasonal pattern in bipolar disorders.

Cited 60 times in the scientific literature.

Level 4 - case-series / case-control

Case-control candidate gene association study comparing bipolar disorder patients with and without a seasonal pattern

PubMed 25989161 · doi:10.1038/srep10232 · record verified 2026-08-30

What was done

The study evaluated 269 Caucasian patients with bipolar disorder, comparing those with and without a seasonal pattern, recruited from university-affiliated psychiatric departments in France. Researchers performed single-marker and gene-based genetic association analyses on 349 single nucleotide polymorphisms (SNPs) across 21 circadian genes and 3 melatonin pathway genes, adjusting for multiple testing with false discovery rate and Bonferroni methods.

What was found

Fourteen SNPs across six circadian genes (NPAS2, CRY2, ARNTL, ARNTL2, RORA, RORB) showed nominal associations with seasonal patterns. After false discovery rate correction, five SNPs in NPAS2 remained significant: rs6738097 (pc = 0.006), rs12622050 (pc = 0.006), rs2305159 (pc = 0.01), rs1542179 (pc = 0.01), and rs1562313 (pc = 0.02). Gene-based analysis identified significant associations for rs6738097 in NPAS2 (empirical p = 0.0003, surviving Bonferroni correction) and rs1554338 in CRY2 (empirical p = 0.005), with epistasis analysis indicating an additive effect between these two markers.

Why it matters

The findings identify NPAS2 and CRY2 variations as potential genetic correlates of seasonal patterns in bipolar disorder, highlighting circadian clock pathways as potential contributors to seasonal vulnerability.

Limits

The sample size was small (n = 269), limiting statistical power and increasing susceptibility to type I and II errors common in candidate-gene studies. The study included only Caucasian individuals from France, limiting generalizability to other ancestries and latitudes. Odds ratios, effect sizes, and independent replication cohorts were not reported in the abstract.

Cited by