Descriptive epidemiology of birth defects thought to arise by new mutation

The current study is the first to examine the association of a broad range of sociodemographic factors with conditions thought to arise most of the time by de novo mutation.

Data were taken from 1999 to 2009 from the Texas Birth Defects Registry (TBDR), a statewide active surveillance program. We used Poisson regression to generate crude and adjusted measures of association; the latter included models with all variables and with a parsimonious subset of variables.

There were 1694 cases with any of the phenotypes in the panel, 1100 cases in a subpanel with ≥90% of cases thought to arise de novo, 523 with chromosomal deletion disorders, and 243 with imprinting disorders. In the most parsimonious models, there was an increasing time trend in all groups except imprinting (p ≤ 0.01). Plurality (twins, triplets, etc.) was associated with greater risk of all groups except chromosomal deletions (p ≤ 0.01). Parental age showed strong trends with all groups; paternal age was most important for the total and imprinting groups (p ≤ 0.0001), and maternal age for the others (p ≤ 0.04). De novo mutation phenotypes were more prevalent among offspring of fathers who are non-Hispanic White compared with some other race/ethnic groups.

This study suggests that birth defects arising by new mutation may be more prevalent among offspring of older parents and in plural births. The increasing time pattern and race/ethnic pattern may be related to greater use of or access to genetic tests. This approach to mutation epidemiology seems feasible for birth defects registries to consider.