Magnesium in drinking water and cardio-vascular disease--an epidemiological dilemma.

The relation between drinking water and cardiovascular disease has been evaluated in a number of studies. Certain of these show strong relationships whereas others do not. This review analyses the methodological aspects on epidemiological studies in terms of dose-range, confounding factors and multiple exposures. It is concluded that there is good evidence for a relation between drinking water quality and cardio-vascular disease, that several methodological criteria for such studies need to be considered, that there is little evidence for magnesium as the single causative agent, and that the relevant exposure should be considered as a mixture of minerals.