Methodologic issues in linking aggregated environmental and health data

✍ Scribed by Markku Nurminen; Tuula Nurminen

• Publisher: John Wiley and Sons
• Year: 2000
• Tongue: English
• Weight: 108 KB
• Volume: 11
• Category: Article
• ISSN: 1180-4009
• DOI: 10.1002/(sici)1099-095x(200001/02)11:1<63::aid-env381>3.0.co;2-q

✦ Synopsis

Epidemiologic studies of environmental exposures and their impacts on disease risk are an important and increasingly applied approach in public health assessment. However, environmental epidemiology often uses data that have been collected as temporal-spatial and demographic statistics, and thus are only available for analysis at the level of aggregate information. The need to conduct aggregate-level studies springs primarily from the diculty of obtaining high-quality, individual-level data on environmental exposures and extraneous covariates. This paper discusses the special characteristics of aggregate data and explains why great care must be exercised when the links between environment and health are analyzed. Further, this paper recalls ways to select an appropriate data-analytic method and strategy for epidemiologic studies, and to infer whether exposure to an environmental risk factor leads to a speci®c health outcome. Applicable statistical methods include ecologic analysis, time series analysis, multilevel modeling, and quantitative risk assessment. Finally, this paper outlines some methodologic requirements for linking environment and health data.