Primary brain tumors following traumatic brain injury ± a population-based cohort study in Sweden Catharina Nygren 1, *, Johanna Adami 2 , Weimin Ye 2 , Rino Bellocco 2 , Jean-Luc af Geijerstam 3 , JoÈrgen Borg 1 & Olof NyreÂn 2 1 Department of Rehabilitation Medicine, Karolinska Institute, Danderyd Hospital, SE-182 88, Stockholm, Sweden; Ph.: 46-8655 5484; Fax: 46-86226555; E-mail: catharina.nygren@mbox304.swipnet.se; 2 Department of Medical Epidemiology, Karolinska Institute, Stockholm, Sweden; 3 Swedish Council of Technology Assessment in Health Care, Stockholm, Sweden (*Author for correspondence) Received 25 August 2000; accepted in revised form 24 April 2001 Key words: brain tumor, epidemiology, risk factor, traumatic brain injury. Abstract Objectives: The aim of this study was to explore the association between traumatic brain injury and brain tumor development. Methods: A cohort of patients hospitalized for traumatic brain injury during 1965±1994 was compiled using the Swedish Inpatient Register. Complete follow-up through 1995 was attained through record linkage with the Swedish Cancer Register, the Cause of Death Register, and the Emigration Register. Standardized incidence ratios (SIRs), de®ned as the ratios of the observed to the expected numbers of brain tumors, were used as the measure of relative risk. The expected number of brain tumors was calculated by multiplying the observed person-time by age-, gender- and calendar year-speci®c incidence-rates derived from the general Swedish population. Results: The cohort included 311,006 patients contributing 3,225,317 person-years. A total of 281 cases of brain tumors were diagnosed during follow-up. No associations were found between traumatic brain injury and the risk of primary brain tumors, neither overall (SIR: 1.0; 95% con®dence interval (CI): 0.9±1.2), nor in analyses broken down by main groups of brain tumors. Strati®ed analyses according to age at entry into the cohort, year of follow- up, and severity of the brain injury all showed essentially the same null results. Conclusion: No association between traumatic head injury and primary brain tumors has been found. Introduction Brain tumors include mainly lethal cancers, particularly glioblastomas and anaplastic astrocytomas. Meningio- mas account for 20% of all brain tumors and are clas- si®ed as benign. The etiology of brain tumors remains largely unknown. Dierent medical conditions, occupa- tional exposures, and reproductive factors have tenta- tively been associated with subsequent development of brain cancer. However, the only established causal factor is high-dose ionizing radiation [1±4]. Although addressed in several epidemiologic studies [5±12] the relationship between head traumas and risk of brain tumor remains unclear. One proposed mechanism is genetic changes with activation of cellular oncogenes and inactivation of tumor suppression genes following tissue repair through cell division [13]. Another suggest- ed mechanism is a co-carcinogenic action through in¯ammation in conjunction with foreign material [14]. A recent cohort study [6] showed a weak association while the results of 11 case±control studies were incon- sistent [5, 7±12, 15±18]. Given that impaired mental functions are common, a case±control design with recalled exposure histories may be unsuitable for etio- logical studies in brain cancer patients. To explore the association between traumatic brain injury and brain tumor risk in greater detail, and with greater precision, we conducted a large population- based cohort study using high-quality computerized registers in Sweden. Cancer Causes and Control 12: 733±737, 2001. 733 Ó 2001 Kluwer Academic Publishers. Printed in the Netherlands.