25 Medical Geography

Monday, 26 August 2013
2:10 PM - 5:30 PM
Washington III

ARE BIRTH DEFECT RATES HIGHER FOR RESIDENTS ON MOUNTAIN-TOP MINING (MTM) COUNTIES THAN THOSE OF NON-MINING COUNTIES: THE WEST VIRGINIA EXPERIENCE

Paper number: 25-7
Program: Sessions
Time: 4:30 PM – 4:50 PM

Authors

  • Ji Li Presenting Author
    Johns Hopkins University, Baltimore, MD
  • Shayhan A. Robbins
    Consultants In Epidemiology & Occupational Health LLC, Washington, DC
  • Steven H. Lamm
    Consultants In Epidemiology & Occupational Health LLC, Washington, DC

Abstract

Introduction: Ahern et al. (2011) reported that women resident in mountain-top mining (MTM) counties were significantly more likely to have children with birth defects (PRR = 1.26) than were women resident in non-mining counties, based on birth certificate recordations for Central Appalachia states (Kentucky, Tennessee, Virginia and West Virginia). Our re-analysis found that this comparison was actually between mountain-top mining counties of Kentucky and West Virginia and non-mining counties of Tennessee and Virginia. We proposed that more proper analysis could be obtained by limiting the comparison to counties within an individual state. West Virginia provided the most balanced distribution of counties.

Materials and Methods: We obtained West Virginia live birth certificate files (1990-2009) and indexed them by county of maternal residence and by hospital of birth. The Ahern classification of counties as non-mining, non-mountaintop mining, and mountaintop mining was used. Primary analysis was for any birth defect. Secondary analyses were for specific birth defects (N = 15) and non-specific birth defects (N=7) (e.g. “Other musculoskeletal”).

Results: The recorded birth defect rates in toto were significantly greater among births for women resident in mountain-top mining counties than for those resident in non-mining counties. Six hospitals which accounted for 84% of the births to residents of mountaintop mining counties had sufficient births (N>1000) from both residents of mountaintop mining counties and of non-mining counties such that hospital-specific, county mining status-specific rates could be calculated. Analysis of variance (ANOVA) revealed that hospital of birth was a significant determinant of birth defect rates but that mining status of county of resident (i.e., mountaintop mining vs. non-mining) was not. Graphic analysis confirmed these results.

Conclusion: The reported association of higher rates of birth defects for residents on mountaintop mining counties than for non-mining counties was found for West Virginia to be a consequence of confounding by hospital of birth rather than a consequence of county of residence.