Association studies between chromosomal regions 1q21.3, 5q21.3, 14q21.2 and 17q21.31 and numbers of children in Poland

Abstract

Abstract Number of children is an important human trait: studies have indicated associations with single-nucleotide polymorphisms (SNPs). Aim: evaluate further evidence for four associations using a large sample of Polish subjects. POPULOUS genetic data provided from anonymous, healthy, unrelated, Polish volunteers of both sexes (N=5760). SNPs (n=173): (a) 69 from the chromosome 17 H1/H2 inversion; (b) six from 1q21.3, 5q21.3 and 14q21.2; and (c) 98 random negative controls. Zero-inflated negative-binomial regression (z.i.) was performed (0-3 children/individual (NCI) as non-events; adjustors: year of birth, sex). Significance level p=0.05 with Bonferroni correction. With both sexes combined, no positive results were obtained from negative-control SNPs or (b). Statistically-significant differences (combined or male data) were obtained from highly-linked inversion SNPs: representative rs12373123 gave combined mean NCI: TT: 2.31 (n=1418); CT: 2.35 (n=554); CC: 2.44 (n=43) (genotype p=0.01; TTvs.CC p=0.004; CTvs.CC p=0.009). Recessive modeling indicated H2-homozygotes had 0.118 more children than H1-homozygotes+heterozygotes (z.i.-count estimates+standard errors: CT, -0.508+0.194; TT, -0.557+0.191). The non-over-dispersed count model detected no interactions. Conclusions: association between the H1/H2 inversion and numbers of children (previously reported in Iceland) has been confirmed, albeit using a different statistical model. One limitation is the small amount of data, despite initially ~6000 subjects. Causal studies require further investigation.