Genetic analysis of the timing of human puberty

Puberty is an important developmental stage during which reproductive function is achieved. The timing of puberty is highly variable in the general population with high heritability (estimates range from 50-80%). However, the specific variants that influence pubertal timing in the general population remain unknown. Association studies represent one approach to identifying the genetic basis of the variability in pubertal timing. I employed three association-based approaches to identify specific variants that underlie the population variation in pubertal timing. First, I used a candidate gene approach to study common variation in ten genes related to hypogonadotropic hypogonadism (HH) and/or Kallmann syndrome (KS), disorders of absent puberty, concluding that variation in the known HH/KS genes is unlikely to play a substantial role in regulating pubertal timing in the general population. I also tested for association between estimated genetic ancestry and age at menarche and found evidence of association in several racial/ethnic groups. Second, I conducted a genome-wide association (GWA) study for age at menarche using pooled DNA samples. I was able to enrich for SNPs with allele frequency differences between early and late menarche groups. However, these SNPs did not replicate in external data sets. I also identified SNPs informative for Native Hawaiian ancestry. Third, I attempted replication of several of the most significant SNPs from GWA studies for age at menarche conducted by other groups and SNPs that have been definitively associated with obesity and height, phenotypes that may be related to puberty. My results for several SNPs were consistent with results from the GWA studies for age at menarche, but further replication samples are required to demonstrate definitive evidence of association. I observed no strong evidence of association with age at menarche at any of the height or obesity loci, although two SNPs in the obesity-associated gene FTO were associated with BMI in women with early but not late age at menarche. Taken together, these data suggest that individual variants that regulate pubertal timing are likely to have modest effects. Large sample sizes or other approaches will be required to identify genetic variants associated with pubertal timing in the general population.