Summary In a case-control study of 280 out of 426 consecutive patients with a recent diagnosis of nonHodgkin lymphoma (NHL) and 1827 control subjects, 53 (19%) and 230 (13%) respectively had received blood transfusions 1 year or more before the interview. Using an age-and sex-stratified analysis the odds ratio (OR) for transfusion was 1.74 (95% CI 1.24-2.44). ORs were also determined for transfusions received in the intervals 1 -5, 6 -15, 16-25 and > , Carli et al., 1994, Cartwright et al., 1994, National Board of Health and Welfare, 1995. The causes of this increase are unknown, and in an attempt to identify risk factors for NHL a case-control study has been initiated in the Southern Health Care Region of Sweden.According to cohort studies of cancer morbidity in blood recipients, transfusions may be a risk factor for NHL. In a Swedish study (Blomberg et al., 1993) an excess of NHL was observed in previously transfused subjects, the standardised morbidity ratio being 4.09 (95% CI 1.65-8.43). Cerhan et al. (1993) obtained information about past transfusions from women aged 55-69 years who were resident in Iowa and followed up for 5 years. For women who had ever received a blood transfusion the relative risk of NHL was 2.20 (95% CI 1.35-3.58). Memon and Doll (1994) identified nearly 13 000 infants who received blood transfusions shortly after birth in England, Wales and Scotland and recorded the incidence of subsequent malignant tumours. The incidence of NHL at 15-49 years of age was about twice that expected, but the excess was not statistically significant.Our ongoing case-control study offered an opportunity to further evaluate the relation between previous blood transfusions and the development of NHL. We report here on the association between transfusions and NHL. The relation between the time lapse since transfusion and the risk of lymphoma has also been studied.
Subjects and methods