Strengths and limitations:
Our study is unique in its size and breadth, including over one million
individuals, stretching as far back as 1939. The size of the study
allowed us to calculate a precise estimate for the familial risk of PA.
While the UPDB allows for population-level results, we were able to
collect data at the individual level using birth and death certificates
as well as electronic medical records. This allowed for careful
inclusion of cases, stratification of separate pregnancies for each
individual, and the ability to address confounding environmental and
clinical risk factors. We used a novel alternative method for evaluating
heritability, the “ego-driven” risk analysis, which allowed for
adjustment of clinical confounders as well as evaluating familial risk
using the relatives as the study population rather than the individual.
One limitation to generalizability is that our cohort arises from a
population of individuals living in a single state in the US. While the
state’s current population resulted from a smaller founder population in
the 1800s, recent genetic investigations have revealed a similar
proportion of genetic heterogeneity to the rest of the US
population.19 The inclusion of all cases of PA in
research is often limited by under-reporting and reliance on the
subjective judgement of health care professionals. We attempted to
mitigate this limitation by including cases of “antepartum bleeding”
while excluding placenta previa in our case definition. While there
might be causes of antepartum bleeding independent of PA, like cervical
or vaginal bleeding, PA is the most common leading cause of vaginal
bleeding throughout pregnancy.1,7 Our results revealed
a trend towards the null with the inclusion of “antepartum bleeding”:
from 20% estimated heritability to 18%. While we attempted to adjust
for all known risk factors for PA in our heritability analyses, one
possible confounding factor includes shared environments in family
pedigrees, which could have led to an overestimation in our risk.