Quantifying altitude of human habitation in studies of human health using geographical name server data.

Almost all studies examining the effects of altitude on human health have estimated the geographical altitude of defined regions, yet the primary interest lies in where people live, not the land around them. Populations are not homogenously distributed across altitudes. We propose a straightforward and computationally simple method for estimating the average altitude of habitation within the regional units for which health statistics are typically reported (such as counties). The United States Board on Geographical Names database contains records for over 2.7 million places, which can be processed to select places that are associated with human habitation. These points can easily be averaged by region yielding a representative altitude of human habitation within city, county, state regions, or by longitude and latitude zones. We provide an example of using this approach in a study of human health, and compare it with three other previously used methods of estimating altitude for counties.