In 2005 in the US, for men aged 35 or over, there were 142,205 deaths from seven cancers considered to be caused by smoking.

If these people had never smoked, Lee and Hamling estimated that the numbers would have reduced by 104,737.

If smokeless tobacco was introduced to a similar population of non smokers, this meta-analysis shows that any increase in risk would be negligible compared to the lives saved by reducing cigarette use.

Lee said: 'Our paper shows very clearly that, in marked contrast to smoking, smokeless tobacco use carries little or no risk of cancer.'

These results were published in the open access journal BMC Medicine.