This Week In The Bmj

Vitamin K and leukaemia: one study finds no link, but others don’t completely exclude a risk

BMJ 1998; 316 doi: (Published 17 January 1998) Cite this as: BMJ 1998;316:a

In the early 1990s two reports by Golding et al suggested that the risk of childhood cancer, particularly leukaemia, was doubled in children who had received prophylactic intramuscular vitamin K at birth. None of the subsequent research found any significant association, but these studies used different methods and were hampered by lack of power. This week's issue includes four more studies: all provide further evidence against an effect of the vitamin K of the size suggested by Golding et al.

The first, by McKinney et al, failed to show any significant association between cancer and vitamin K (p 173). In a population based case-control study in Scotland designed to identify risk factors for childhood cancer they investigated vitamin K by using information in hospital records and found no significantly raised risks for leukaemia or any other type of cancer. Accounting for other factors that might explain an association, such as type of delivery or deprivation, did not alter their findings.

Two studies (case-control and ecological) by Passmore et al (p 178, 184) and another case-control study by Parker et al (p 189) do not, however, exclude a possible association with leukaemia. Passmore et al consider that this might be explained by the association they found between type of delivery and risk of leukaemia. Parker et al found no association with non-leukaemia malignant disease but were unable to rule out an association with acute lymphoblastic leukaemia, especially for that diagnosed in children aged 1-6 years. Passmore et al and Parker et al suggest that their results support a policy of giving intramuscular vitamin K only to babies at high risk of vitamin K deficiency bleeding, with oral vitamin K being given to the others.