They found 2061 patients with liver cancer on records spanning the five years between 1994 and 1999. They also randomly selected more than 6,000 other similarly aged patients from the database, who had not been diagnosed with cancer.

They checked the claims records for known risk factors for liver cancer, including hepatitis C and B viral infections, alcoholic liver disease, and haemochromatosis (poor absorption and storage of iron). They also checked the claims records for diabetes.

Compared with patients who did not have the disease, those with liver cancer were around twice as likely to be male and of non-white ethnicity. And the proportion of patients with diabetes was almost twice as high among those with liver cancer.

The authors suggest that diabetes might be an independent risk factor for liver cancer, rates of which have increased significantly in the US in recent years.