Observational analytic studies in multiple sclerosis: controlling bias through study design and conduct. The...

Rising multiple sclerosis incidence over the last 50 years and geographic patterns of occurrence suggest an environmental role in the causation of this multifactorial disease. Design options for epidemiological studies of environmental causes of multiple sclerosis are limited by the low incidence of the disease, possible diagnostic delay and budgetary constraints. We describe scientific and methodological issues considered in the development of the Australian Multicentre Study of Environment and Immune Function (the Ausimmune Study), which seeks, in particular, to better understand the causes of the well-known MS positive latitudinal gradient. A multicentre, case-control design down the eastern seaboard of Australia allows the recruitment of sufficient cases for adequate study power and provides data on environmental exposures that vary by latitude. Cases are persons with an incident first demyelinating event (rather than prevalent multiple sclerosis), sourced from a population base using a two tier notification system. Controls, matched on sex, age (within two years) and region of residence, are recruited from the general population. Biases common in case-control studies, eg, prevalence-incidence bias, admission-rate bias, non-respondent bias, observer bias and recall bias, as well as confounding have been carefully considered in the study design and conduct of the Ausimmune Study.