Should we be frightened of bracken? A review of the evidence.
OBJECTIVE: To assess the risk to human health of the plant bracken (Pteridium sp). DESIGN: An evaluation of studies of human and animal populations exposed to bracken, together with a review of expert reports and advice to the public. MAIN RESULTS: Bracken induced disease has been demonstrated in animals in both laboratory and field studies. Depending on the species, diseases in animals associated with the plant have included; cancers of the alimentary and urogenital tract, lung and breast; haematuria; retinal degeneration; and, thiamine deficiency. Potential exposure of human populations is through: food either directly (people in some parts of the world eat bracken as a traditional dish) or indirectly by consuming animals fed on bracken; milk; water; inhalation and ingestion of spores; and insect vectors. Four studies of human populations (two analytical and two observational) failed to assess adequately confounding factors and other sources of bias, so that conclusions about a risk to human health from bracken cannot firmly be drawn. Establishing exposure is also extremely difficult in populations (such as the United Kingdom) where direct consumption of bracken is rare. CONCLUSION: Bracken is a common plant worldwide. It is toxic to many animal species and to several organ systems. There is no tumour (or other disease) that is pathognomic of exposure in animals, though cancers of the alimentary and urogenital tract seem to be the most commonly associated. It is not possible to extrapolate from animal models to humans. Studies of human populations, do not establish a clear risk of bracken to human health, largely because of methodological problems. Testing the evidence against traditional criteria of causality only fulfils the criterion of biological plausibility. Despite this, current public information implies a serious risk to human health from bracken, and increasing media coverage of the subject is likely to lead to greater public concern. Further epidemiological studies are required.