[Incidence and mortality of central nervous system tumors in France: trends over the period 1978-2000 and influence of registration practices on results]. / Incidence et mortalité des tumeurs du système nerveux central en France: évolution de 1978 à 2000 et influence des pratiques d'enregistrement sur les résultats.

BACKGROUND: In France, cancer incidence figures are produced by cancer registries covering only 13.5% to 16% of the whole population of the country. Thus, to produce national figures, estimates have to be computed. Registration disparities between registries concerning tumors of the Central Nervous System (CNS) could have biased these estimates. METHODS: National estimates are based on modelling of the incidence/mortality ratio. The most recent estimations for year 2000 were calculated by the French Cancer Registry Network (FRANCIM) and the department of biostatistics of Lyon University Hospital. Since benign tumors are not recorded in some cancer registries, a new estimate of the incidence of CNS tumors was produced by estimating the number of benign tumors in these registries. RESULTS: In 2000 in France, the number of estimated cases of CNS tumors was 2697 in men and 2602 in women, with incidence rates (World standard) of 7.4 and 6.4 per 100,000 respectively. The incidence increased between 1978 and 2000, on an average by 2.25% per year in men and 3.01% per year in women. However, these estimates do not provide a correct picture of CNS incidence. First of all, pathological diagnoses are not performed in 3.5%-27.5% of the patients with CNS tumors registered in French registries. Second, figures for benign tumors (mainly meningiomas) were provided by only two of nine cancer registries. If benign tumors had been registered by all cancer registries, computed incidence would have increased by 12% for men and 26% for women. CONCLUSION: Incidence of CNS tumors is increasing in France, as in many other countries. To improve comparability with other countries, French cancer registries should also collect data on benign tumors. The discrepancies observed between registries in the proportion of patients without information on histology show differences in diagnostic practices and should be the starting point for a survey on this topic.

[Ovarian cancer: incidence and mortality in France]. / Les cancers de l'ovaire: incidence et mortalité en France.

OBJECTIVES: To describe the ovarian cancer incidence and mortality trends in France from 1980 to 2000. MATERIALS AND METHODS: Incidence data were obtained from 9 French departments covered by cancer registries which systematically record all cancers, of which those of the ovary. Mortality data has been provided by INSERM bureau in charge of the analysis of death certificates. RESULTS: With 4500 incident cases and 3500 deaths for the year 2000 in France, ovarian cancer still shows a poor prognosis. Age adjusted incidence and mortality rates are almost stable over the study period at an annual rate of 9 by 100000 for incidence and 5.5 by 100000 for mortality. However the risk of developing this cancer and the risk of dying from it, steadily decreased for the cohort born in 1930 to 1950, and the beginning of this reduction corresponds exactly to the beginning of oral contraception. We present comparisons between departments and analyze the border-line tumors in parallel with invasive cancers. CONCLUSION: The cancer of the ovary represents only a small part of female cancers (3.8%) in France and its incidence is stable over time, but its prognosis is very bad.

Medical and non-medical determinants of prostate cancer management: a population-based study.

Non-medical factors influencing treatment choices in prostate cancer are not well understood. We carried out a population-based study to obtain information on the management of prostate cancer patients. Our study population consisted of 1000 men diagnosed during 1995 from five French cancer registries. We looked at the main treatments performed in the year following diagnosis. Multivariate analysis was used to describe the determinants of the various treatment choices, simultaneously taking into account medical and non-medical factors. The probability of treatment by radical prostatectomy (RP) was 3 times higher in the Tarn area, whereas in the Calvados area the probability of treatment by radiotherapy was almost 6 times higher. The private sector favoured radical prostatectomy and hormonal therapy. In France, as in other developed countries, the initial treatment of prostate cancer varies greatly according to non-medical factors. This type of investigation, if carried out regularly, would make it possible to evaluate changes in practice patterns.

Cancer incidence and mortality in France over the period 1978-2000.

BACKGROUND: Monitoring cancer incidence and mortality time trends is essential for cancer research and health-care planning. French cancer registries do not cover the entire population and do not provide a representative sample of the national population. Our study aimed at estimating national cancer incidence and mortality trends over the longest period available. METHODS: Incidence and mortality data were collected over the period 1978-1997. Twenty-seven cancer sites were selected and age, sex and site specific incidence and mortality rates were estimated for each year from 1978 up to 2000. Observed incidence and mortality data in the population covered by cancer registries were modelled using age-cohort methods. An estimation of the incidence/mortality ratio was obtained from these models and applied to the mortality rates predicted from an age-cohort model for the entire French population. The person-years of observation were calculated cohort-wise from census data provided by the national institute of statistics RESULTS: Cancer incidence increased by 63% throughout the study period, from 170,000 new cases in 1980 to 278,000 in 2000. This evolution was due to demographic changes but also to an increase in the risk of cancer which was estimated to more than 35% during the same period. In men, this change is largely explain by the increase of prostate cancer incidence. Among women, the increase was dominated by the continuing increase in breast cancer incidence. Large increases were also seen for non-Hodgkin lymphoma, melanoma, and thyroid cancer in both genders and for lung cancer in women. Cancer mortality increased by 20% from 125,000 deaths in 1980 to 150,000 in 2000. This increase is less than that predicted from changes in demographic factors and corresponds in fact to a decrease in the risk of death estimated to about 8%, slightly greater for women than for men. This decrease is associated with a decreasing incidence for stomach cancers for both sexes, alcohol-related cancer for men and cervical cancer for women. Colo-rectal cancer decreasing mortality contributes to this improvement despite an incidence increase. CONCLUSION: Between 1980 and 2000, the study showed a large change in the cancer burden both quantitatively and qualitatively. Decrease in exposure, earlier diagnosis and therapeutic improvement explained part of this change, but overall the distribution of cancer cases shifted toward a distribution including less aggressive cancers. A striking divergence between incidence and mortality trends is observed for a great number of cancers. Prostate cancer shares with breast cancer the same pattern of a severe increasing incidence and a stable mortality. This points to important changes in medical practice and needs further analysis. The trend of lung cancer mortality among women should be emphasised since the situation will inevitably worsen in the coming years. It is already the third cause of cancer death among women.

[Lip, oral cavity and pharynx cancers in France: incidence, mortality and trends (period 1975-1995)]. / Les cancers de la lèvre, de la cavité buccale et du pharynx en France: incidence, mortalité et tendance (période 1975-1995).

With 10,882 estimated new cases in 1995 in France, lip, oral cavity and pharynx tumours rank 4th, representing 8.1% of all cancers in men. They are less frequent in women, with a sex ratio of 7. Based on the French cancer registries data which cover 13% of the metropolitan territory in 2000, both incidence and mortality increased until early 1980s to decrease thereafter. The main hypothesis proposed to explain the French leadership world-wide for these tumours deals with alcohol and tobacco consumption. Important differences observed between several areas within Europe, for some subsites, in connection with age or sex, are pointing toward the need of new studies about environment and/or genetics. Until now, comparisons between countries were made at the level of lip, oral cavity and pharynx category as a whole or by large subgroups. In this work we attempt to establish more accurate statistics, in order to comply with the situation of this cancer in France. Present results should encourage the scientific community to conduct site specific epidemiological studies.

Contacts with animals and humans as risk factors for adult brain tumours. An international case-control study.

While numerous studies have addressed the possible role of farming and related exposures as risk factors for brain tumours in adults, few of them have examined the potential effect of exposure to farm animals or pets. In an international multicentre case-control study, we investigated whether residence on a farm, contact with animals, or working in occupations with a high degree of potential contact with animals or humans were associated with brain tumours. Using a common questionnaire, 1177 cases of glioma, 330 with meningioma and 2478 controls from eight centres were interviewed about the exposures and, in particular, about their contacts with nine species of animals: dairy cattle, beef cattle, pigs, horses, sheep, goats, poultry, dogs and cats. Living or working on a farm was not a risk factor, for either glioma or meningioma. Except in some centres, there was no relationship between having contacts with farm animals or pets and the risk of brain tumour, for either type of tumour or either sex. In relation to seven industrial groups involving frequent human and/or animal contacts, no association was apparent for either glioma or meningioma. In relation to 25 occupational groups with potential frequent contact with humans and/or animals, for glioma there was a reduced risk for biological technicians (Odds Ratio (OR)/=0, P=0.01), and general farm workers (OR=0.66, 95% Confidence Interval (CI): 0.5-0.9). For meningioma, there was an increased risk for cooks (OR=2.0; CI: 1.2-3.4). With some exceptions, these results indicate no association between either the type of brain tumour and contacts with animals, or with occupations that include a high level of contact with animals or a high level of contact with humans.

[Cancer of the prostate in France: results of the survey CCAFU-FRANCIM]. / Le cancer de la prostate en France: résultats de l'enquête CCAFU-FRANCIM.

OBJECTIVE: The management, diagnosis and treatment of prostate cancer (PC) in the general population are poorly defined in France. The objective of this survey was to analyse the diagnostic and therapeutic modalities of prostate cancer in 1995, on the basis of a population derived from 4 French cancer registries. MATERIAL AND METHOD: A sample of 803 PCs diagnosed in 1995 were selected at random from the cases listed in the 4 registries (Bas-Rhin, Calvados, Isère and Tarn). Analysis by questionnaire concerned the modalities of diagnosis, clinical tumour stage and the treatment performed. Clinical stage (TNM 1992) was submitted to centralized coding. Logistic regression was used to quantify the various practices, taking into account the patient's age, PSA level and clinical stage. The probability of receiving each treatment modality was studied by using the same clinical determinants. RESULTS: The mean age of the patients was 71.6 years (range: 46-94). The clinical stage was T1 or T2 in 60% of cases, T3 or T4 in 14% of cases and N+ or M+ in 17% of cases. PSA (median: 18.2 ng/ml) was assayed in 92.4% of cases. The diagnosis was established by biopsy in 63% of cases and by endoscopic resection (TURP) in 32% (5% unknown). The main treatments were: radical prostatectomy (RP): 21.9%, radiotherapy: 19.4%, endocrine therapy: 33%, isolated TURP: 16.3%, conservative management: 6% and unspecified treatment: 5.6% of cases. 31% of cases received combinations of various treatment modalities. RP was performed more frequently in patients over the age of 60 years, for T2 tumours (OR: 3.3) and for 4 < PSA < 20 ng/ml. Radiotherapy tended to be reserved for older patients with T3-T4 tumours and 20 < PSA < 50 ng/ml. The frequency of endocrine therapy increased with age and PSA (> 50 ng/ml). The frequency of TURP and surveillance also increased with age, decreased with high PSA and essentially concerned T1 tumours. CONCLUSION: PC was diagnosed relatively early in France in 1995 with clinically localized tumours in 60% of cases. In this survey, 94% of patients received treatment during the year following diagnosis, with 40% of curative treatments and 31% of combined treatments.

[Variations of the practice of radical prostatectomy in France]. / Variations dans la pratique de la prostatectomie radicale en France.

OBJECTIVE: A description of radical prostatectomy (PR) practice in the general population in France has never been reported. The objective of this study was to analyse RP practice in France, and its determinants and geographical variations based on the CCAFU-Francim survey on prostate cancer (PC) diagnosed in 1995. MATERIAL AND METHOD: 175 PR were performed on the sample of 798 patients selected at random from the cases of PC identified in 1995 by four cancer registries (Bas-Rhin, Calvados, Isère and Tarn). Analysis was based on tumour characteristics, diagnostic methods and histopathological results. Multivariate analysis by logistic regression taking into account age, PSA and clinical stage studied variations between departments and types of urology practice (private or public). RESULTS: The mean age of the patients at the time of the diagnosis was 65.3 years (46-76). The median PSA was 18.3 ng/ml (1-184). The diagnosis was made by biopsies (91%) which were systematized in 74% of cases or by transurethral resection (7%). The clinical stage was T1 (22.3%), T2 (64%), T3 (8.6%), N+ (0.611) and unknown (4.5%). The pathological stage was pT2N0 (46.3%), pT3N0 (40%), pT4N0 (1.7%), pTxN0 (8.6%) and unknown (3.4%). Adjuvant treatment (radiotherapy: 13.7%, endocrine therapy: 13.7% or both: 31%) was performed in 54 patients (31%). Multivariate analysis showed that the adjusted probability to be treated by RP was 3 times higher in one department compared to others and 2.6 times higher in the private sector. CONCLUSION: This study of RP practice in the general population shows a concordance with the 1995 recommendations, but it also shows practice variations according to the region and the type of practice, reflecting different schools of thought and medical education within the same country.

A population-based study on management of prostate cancer in four regions of France.

In France, as in other European countries, management of prostate cancer in the population is rather variable. The objective was to analyse diagnosis and treatment modalities of prostate cancer in patients from French cancer registries.A sample of 803 patients with prostate cancer diagnosed in 1995 was drawn at random from cases recorded in four cancer registries in the geographical regions of Bas-Rhin, Calvados, Isère and Tarn. Diagnosis, clinical staging and treatment were analysed by questionnaire. Multivariate analysis by logistic regression was used to describe medical determinants (age, PSA, clinical staging) of the various treatment choices.The mean age of the patients was 71.6 y (range 46-94 y). Clinical staging showed 60% T(1-2) tumours, 15% T(3-4) and 17% N+or metastases. PSA rate (median 18.2 ng/ml) was assessed in 92.4% of patients. Prostate cancer was diagnosed by prostate biopsy in 63% and by transurethral resection of the prostate (TURP) in 32% of patients. The main treatments were radical prostatectomy 22%, radiotherapy 19.4%, hormonal therapy 33%, TURP alone 17.7% and expectant management 6%. The method of treatment was unknown in 5.7% of cases. Adjuvant radiotherapy or hormonal therapy had been used in 31% of cases. Logistic regression analysis showed that radical prostatectomy was most often performed in patients aged <60 y, three times more frequently for T(2) tumour and for PSA between 4 and 20 ng/ml. Radiotherapy was the most frequently applied treatment for patients aged between 65 and 75 y, especially in T(3) tumours.Approximately 75% of the patients in the study underwent a specific treatment for prostate cancer with a curative intention in 40%. This study provides a baseline to clinicians and public health authorities on the management of prostate cancer in France with recent data. This survey will be useful to compare future descriptive analysis and to provide data regarding changing clinical practice.Prostate Cancer and Prostatic Diseases (2001) 4, 154-160.

A population-based study on radical prostatectomy in France.

There has as yet been no descriptive study of the practice of radical prostatectomy in the general population in France. The objective of this work was to investigate the use of radical prostatectomy (RP) in France and its determinants and geographic variations. A total of 175 radical prostatectomies was identified in a random sample of 798 cases of prostate cancer recorded in 1995 by four cancer registries, in the departments of Bas-Rhin, Calvados, Isère and Tarn. Tumour characteristics, diagnostic procedures and histopathological results were analysed. Multivariate analysis by logistic regression was used to take into account age, prostate specific antigen (PSA) levels and clinical stage in order to study variations between geographical departments and sectors of activity (private or public). The mean age of the patients at the time of diagnosis was 65.3 y (range 46-76). Median PSA level was 18.2 ng/ml (range 1-184). Diagnosis was made by randomised biopsies in 73.8% of cases or by transurethral resection (6.9%). Clinical stage was classified T(1) (22.3%), T(2) (64%), T(3) (8.6%), N(+) (0.6%) and unknown (4.5%). The histopathological result was pT(2)N0 in 46.3% of cases, pT(3)N0 in 40%, pT(4)N0 in 1.7%, pT(x)N(+) in 8.6% and unknown in 3.4%. Adjuvant therapy (radiotherapy 13.7%, hormonal treatment 13.7% or both 3%) was administered in 54 patients (31%). Logistic regression showed that the probability of undergoing RP was three times higher in one department than in the other departments, and was 2.6 times as high in the private sector. This study on the practice of RP is the first performed in the general population in France. It shows that practice differs according to geographical region and sector of activity, indicating that schools of thought and medical culture vary within the same country.Prostate Cnacer and Prostatic diseases (2001) 4, 118-123

National cancer prevalence estimation in France.

In France, as in several other European countries, prevalence has to be estimated from the modelling of 2 of the 3 basic epidemiological measures of incidence, mortality, and survival. Since, in these countries, follow-up of cancer patients is only made in a few registries, we explored the feasibility of estimating prevalence in the absence of follow-up data. The method, which used only incidence and mortality, was validated on Danish data and applied to France. For this latter country, the estimation procedure is based on the recorded mortality data and an estimate of incidence for the entire country. It is applied to selected sites of cancer, which account for 80% of the estimated incidence. In 1992, the prevalence of patients who had such a diagnosis amounts to 538,000 women and 424, 000 men. The most frequent cancer sites are head and neck, breast, and large bowel. Most of the cancer sites present an increase in prevalence proportion between 1987 and 1992. The larger increases concern breast and prostate cancer.

[Patterns of health care delivery and breast cancer in the department of Isère in 1955]. / Filières de soins et cancer du sein dans le département de l'Isère en 1995.

BACKGROUND: Caring for cancer patients is expensive, warranting verification that health care organization works in a satisfactory way. A first step of this evaluation deals with the description of the pathway followed in the health care system by the patient. METHODS: 671 breast cancer cases were diagnosed in Isère in 1995. According to the place where each treatment (surgery, chemotherapy, radiotherapy) was performed, we described pathways for the patient, either entirely private, public or mixed. Characteristics of the patient (age, place of residence), of the disease (extent of disease, way of discovery) and of the physician (general practitioner, specialist) might have influenced the choice of this pathway. We described and tested the distribution of these characteristics within the 3 groups using univariate analysis. Relative risk of being affected to the private pathway compared to the public one was computed, after adjusting for age, type of physician, extent of disease, way of discovery and sanitary area, using a multivariate analysis (logistic regression). RESULTS: In the department of Isère, the private pathway cared for 55% of breast cancers, the public one 23% and the mixed one 19%. There was no preferential recruitment according to age, physician type, presence of metastasis or of the rural or urban residence. In sanitary area number 5, characterized by an important attraction of the patients by the nearby department of Rhône, 41% of the patients were cared for the private pathway, compared to 63% in sanitary area 4, where most patients were treated in the main town of Isère: Grenoble. After early breast cancer detection with mammography instead of breast cancer screening, probability of being cared for in the private pathway was 2-fold higher (OR = 2) than in the public one. CONCLUSION: In Isère department, early breast cancer detection with mammography is in favor of the private pathway. This is not true for physician type, neither for characteristics of the patient or extent of the disease. Finally, the distance to next department of oncology or radiotherapy plays a major role.

Cancer registry data based estimation of regional cancer incidence: application to breast and colorectal cancer in French administrative regions.

STUDY OBJECTIVE: In many countries, cancer registries cover only a small part of the national population. Cancer incidence for the rest of the country has therefore to be estimated. This can be done from mortality data using the relation between incidence and mortality observed in the cancer registry areas. Such an approach was used to study geographical variation and trend of colorectal and breast cancer incidence in France where 10% of the national population is covered by cancer registries. DESIGN: This study applies the incidence/mortality ratios of cancer registry areas to regional mortality data to obtain an estimation of cancer incidence at a given point in time. Age and period effects are included in the statistical models. MAIN RESULTS: The incidence estimations are given for 21 administrative regions and three time points (1985, 1990, 1995). The European standardised incidence rates for breast cancer ranged from 86.8 to 128.8. For colorectal cancer, these rates ranged from 48.2 to 79.6 for men, and from 32.5 to 48.8 for women. Breast cancer incidence has increased considerably between 1985 and 1995 with a higher increase in the north than in the south of France. The incidence of colorectal cancer has also increased, albeit to a lesser extent. CONCLUSION: The incidence estimation method proposed leads to regional incidence rates that are useful for planning health care services on a regional basis and may also be used to study regional differences in incidence. This method is useful when only partial incidence data are available.

Analysis and classification of interval cancers in a French breast cancer screening programme (département of Isère).

The objective of this study is to analyse detection rates and stage of diagnosis of interval cancers in the mass screening mammography programme of Isère (France), launched in 1990. Interval cancers are defined as breast cancers diagnosed within 30 months after a negative screening assessment, for women attending the programme between November 1990 and December 1994. Stages of diagnosis of these cancers are compared with those of screened cancers and to those of cancers diagnosed outside the programme. The rates of invasive interval cancers are 17.7% of the expected incidence rate during the first year, 60.0% during the second year and 58.8% after the second year. Sensitivity of the programme (one test every 30 months) is 74%; sensitivity at one year is 82%. Results are better for women aged 60-69 years than for younger women (50-59 years). Diagnosis is made at an early stage with 8% of in situ cases, and with 40% of very small tumours (sizes < or = 10 mm). Those stages are very close to the ones for screened cases. Interval cancer rates are low during the first year. Higher rates for the second year and early stages of diagnosis could be explained by self-referred screening practice in our area.

Role of medical history in brain tumour development. Results from the international adult brain tumour study.

In an international population-based case-control study carried out in 8 centres in 6 countries, we investigated the role of specific medical conditions in the aetiology of brain tumours in adults. Recruited were 1,178 glioma and 331 meningioma cases and 2,493 age- and gender-matched population controls. Only medical conditions occurring at least 2 years before brain tumour diagnosis were considered. Relative risks (RRs) and 95% confidence intervals (CIs) were estimated using a conditional logistic regression model. Heterogeneity between centres was tested. No association between meningioma and previous medical conditions was observed. For glioma, there was an increased risk associated with epilepsy (RR = 6.55, 95% CI 3.40-12.63), but this was considerably weaker for epilepsy of more than 20 years duration. The risk remained elevated after adjustment for use of anti-epileptic drugs. There was a statistically significant inverse association between glioma and all allergic diseases combined (RR = 0.59, 95% CI 0.49-0.71); this was also observed for specific allergic conditions, namely, asthma and eczema. Subjects who reported a history of infectious diseases (e.g., colds, flu) showed a 30% reduction in risk (RR = 0.72, 95% CI 0.61-0.85). The decreased risks for glioma in subjects reporting a history of allergic conditions or infectious diseases may indicate an influence of immunological factors on the development of glioma. The association between glioma and epilepsy has to be interpreted cautiously and needs further investigation.

[Survival and prognostic factors of germ cell tumors of the testis in the departments of Bas-Rhin, Isère, and Somme]. / Survie et facteurs pronostiques du cancer des cellules germinales du testicule dans les départements du Bas-Rhin, de l'isère et de la Somme.

The authors report a study of the prognosis of germ cell tumours of the testis in populations of the Bas-Rhin, Isère and Some departments. This study was performed on French Eurocare data (European testicular cancer register study). 247 patients diagnosed between 1987 and 1993 were studied. Estèv's method and model were used to estimate the relative survival and for prognosis analysis with reference to the French age-matched male population. 84.6% of patients were between the ages of 20 and 49 years. The 5-years relative survival was 90.1%; 95 CI% [85.5-94.8]. The 5-year relative survival was lower (63.9%) in advanced forms (stage III). Univariate analysis in 223 patients demonstrated the following significant prognostic factors: stage, metastatic sites, treatment, retroperitoneal lymph node dissection and testicular vascular invasion. Multivariate analysis identified: stage, age, CNS metastases, retroperitoneal lymph node dissection and treatment. This study confirms the generally good prognosis of this cancer. As reported in the literature, factors negatively influencing relative survival are those related to advanced disease. Age, a controversial factor in the literature, was correlated with relative survival in this study.

[Reporting of resected colonic carcinomas. Assessment of practices in 8 French counties in 1995]. / Examen anatomopathologique des cancers du côlon. Etat des pratiques dans huit départements français en 1995.

OBJECTIVES: The objective of this work was to study the reporting of resected colonic carcinomas and to compare them with Consensus Guidelines published in January 1998. METHODS: The study included 535 colonic carcinomas collected in 8 French registries in 1995, and chosen by drawing lots. For each report, the presence of the information requested by the Guidelines was looked for. Three synthetic variables were built: 2 scores and one qualitative at 2 classes. The influence of patients, tumour and health care system's characteristics was analysed on the 3 dependent variables. RESULTS: Some important variations were observed for information's percents in reporting. In monovariate analysis, scores were significantly influenced by cancer's sub-location, area of patient's residence, surgical center, type of laboratories and pathologist case volume. In multivariate analysis, significant heterogeneity in practices remained between geographical areas. Types of laboratories and pathologist case volume affected differently dependent variables. CONCLUSION: This study shows the necessity to assess the practices before consensus because of the impact of pathological forms in therapeutic decisions and variations observed.

An international case-control study of adult glioma and meningioma: the role of head trauma.

BACKGROUND: Increased brain tumour risk after head trauma suggested by case reports and clinical series has been previously studied epidemiologically with mixed results. An international multicentre case-control study investigated the role of head trauma from injury or sports participation in adult brain tumour risk. METHODS: In all, 1178 glioma and 330 meningioma cases were individually or frequency matched to 2236 controls. Only exposures that occurred at least 5 years before diagnosis and head injuries that received medical attention were considered. RESULTS: Risk for ever having experienced a head injury was highest for male meningiomas (odds ratio [OR] = 1.5, 95% confidence interval [CI] : 0.9-2.6) but was lower for 'serious' injuries, i.e. those causing loss of consciousness, loss of memory or hospitalization (OR = 1.2, 95% CI: 0.6-2.3). Among male meningiomas, latency of 15 to 24 years significantly increased risk (OR = 5.4, 95% CI: 1.7-16.6), and risk was elevated among those who participated in sports most correlated with head injury (OR = 1.9, 95% CI: 0.7-5.3). Odds ratios were lower for male gliomas (OR = 1.2, 95% CI : 0.9-1.5 for any injury; OR = 1.1, 95% CI: 0.7-1.6 for serious injuries) and in females in general. CONCLUSIONS: Evidence for elevated brain tumour risk after head trauma was strongest for meningiomas in men. Findings related to sports should be interpreted cautiously due to cultural variability in our data and our lack of complete data on physical exercise in general which appeared to be protective.