Search

Background: The relationship between diet and pancreatic cancer remains unclear. In this study, we assessed the role of diet and nutrition as risk factors for pancreatic cancer, using data obtained from direct interviews only, rather than data from less reliable interviews with next of kin. We evaluated whether dietary factors could explain the higher incidence of pancreatic cancer experienced by black Americans compared with white Americans., Methods: We conducted a population-based case-control study of pancreatic cancer diagnosed in Atlanta (GA), Detroit (MI), and 10 New Jersey counties from August 1986 through April 1989. Reliable dietary histories were obtained for 436 patients and 2003 general-population control subjects aged 30-79 years., Results: Obesity was associated with a statistically significant 50%-60% increased risk of pancreatic cancer that was consistent by sex and race. Although the magnitude of risk associated with obesity was identical in blacks and whites, a higher percentage of blacks were obese than were whites (women: 38% versus 16%; men: 27% versus 22%). A statistically significant positive trend in risk was observed with increasing caloric intake, with subjects in the highest quartile of caloric intake experiencing a 70% higher risk than those in the lowest quartile. A statistically significant interaction between body mass index (weight in kg/height in m2 for men and weight in kg/height in m1.5 for women) and total caloric intake was observed that was consistent by sex and race. Subjects in the highest quartile of both body mass index and caloric intake had a statistically significant 180% higher risk than those in the lowest quartile., Conclusions: Obesity is a risk factor for pancreatic cancer and appears to contribute to the higher risk of this disease among blacks than among whites in the United States, particularly among women. Furthermore, the interaction between body mass index and caloric intake suggests the importance of energy balance in pancreatic carcinogenesis.

A cross-sectional study may be more feasible than a cohort or case-control study for examining the effect of a genetic mutation on cancer penetrance outside of cancer families. The kin-cohort design uses volunteer probands selected from a population with a relatively high frequency of the mutations of interest. By considering the cancer risk in first-degree relatives of mutation-positive and -negative probands as a weighted average of the risk in carriers and noncarriers, with weights calculated assuming a known mode of inheritance, one can infer the penetrance of the mutations. The estimates of penetrance by age 70 years for three specific mutations in the BRCA1 and BRCA2 genes common among Ashkenazi Jews for the first occurrence of breast or ovary cancer is 63%. The kin-cohort design can be a useful tool for quickly estimating penetrance from volunteers in a setting in which the mutation prevalence is relatively high.

Mutations in APC are classically associated with familial adenomatous polyposis (FAP), a highly penetrant autosomal dominant disorder characterized by multiple intestinal polyps and, without surgical intervention, the development of colorectal cancer (CRC). APC is a tumour-suppressor gene, and somatic loss occurs in tumours. The germline T-to-A transversion responsible for the APC I1307K allele converts the wild-type sequence to a homopolymer tract (A8) that is genetically unstable and prone to somatic mutation. The I1307K allele was found in 6.1% of unselected Ashkenazi Jews and higher proportions of Ashkenazim with family or personal histories of CRC (ref. 2). To evaluate the role of I1307K in cancer, we genotyped 5,081 Ashkenazi volunteers in a community survey. Risk of developing colorectal, breast and other cancers were compared between genotyped I1307K carriers and non-carriers and their first-degree relatives.

The association of diet, smoking/drinking and occupation with subsequent risk of fatal colorectal cancer was investigated in a cohort of 17,633 white males aged 35 and older, who completed a mail questionnaire in 1966. During the subsequent 20 years of follow-up, 120 colon cancer and 25 rectal cancer deaths were identified. Due to small numbers, no significant dose-response trends were observed in the study, but risk of colon cancer was elevated among heavy cigarette smokers (> or = 30/day; RR = 2.3, 95% CI 0.9-5.7), heavy beer drinkers (> or = 14 times/month; RR = 1.9, 95% CI 1.0-3.8) and white-collar workers (RR = 1.7, 95% CI 1.0-3.0) or crafts workers within service and trade industries (RR = 2.6, 95% CI 1.1-5.8). In addition, an increased risk was seen for those who consumed red meat more than twice a day (RR = 1.8, 95% CI 0.8-4.4). Risk patterns for cancers of the colon and rectum combined were similar to those reported for cancer of the colon, but the estimates were somewhat dampened. Our findings support previous reports that a high intake of red meat and a sedentary life-style may increase the risk of colon cancer.

Human papillomavirus (HPV), particularly HPV 16, is linked to the development of cervical cancer. However, the role of HPV 16 in a number of extra-cervical epithelial tumours is controversial. To assess exposure to HPV 16 in patients with different cancers, we conducted a large serosurvey of 905 patients with 21 types of tumours and measured IgG to HPV 16 virus-like particles (VLPs) using a well characterized enzyme linked immunosorbent assay (ELISA). Patients with cervical cancer were considered 'positive controls', as about half were expected to be specifically HPV 16-related. A non-cancer study group consisting of 48 patients with endocrine disorders (eg diabetes) was also tested. HPV 16 antibody prevalence was highest in patients with cancers of the cervix (52% +/- 7%), vulva (27% +/- 9%), vagina (27% +/- 13%) and penis (63% +/- 16%). Seroprevalence was much lower in the non-cancer group (4% +/- 3%) and all other cancer patients: uterus (9% +/- 4%); ovary (4% +/- 3%); testis (5% +/- 4%); prostate (6% +/- 4%); squamous carcinoma (6% +/- 3%) and adenocarcinoma (4% +/- 3%) of the lung; rectum (2% +/- 2%); pancreas (8% +/- 4%); colon (2% +/- 2%); stomach (0%); oesophagus (8% +/- 4%); buccal cavity (12% +/- 5%); breast (10% +/- 4%); non-melanomatous (9% +/- 6%) and melanomatous (6% +/- 3%) skin tumours; bladder (8% +/- 4%); and kidney (2% +/- 2%). The results confirm the strong relation of HPV with several lower anogenital tract tumours, but, at least in this population, fail to identify additional epithelial tumours associated with high seroprevalence of HPV 16 VLP antibodies.

Objectives: To investigate the reproducibility of wire codes to characterise residential power line configurations and to determine the extent to which wire codes provide a proxy measure of residential magnetic field strength in a case-control study of childhood leukaemia conducted in nine states within the United States., Methods: Misclassification of wire codes was assessed with independent measurements by two technicians for 187 residences. The association between categories of wire code and measured level of magnetic field was evaluated in 858 residences with both a wire code measurement and a 24 hour measurement of the magnetic field in the bedroom. The strength of the association between category of wire code and risk of leukaemia was examined in two regions with different average levels of magnetic field in homes with high categories of wire code., Results: The reproducibility of any of three different classifications of wire codes was excellent (kappa > or = 0.89). Mean and median magnetic fields, and the percentage of homes with high magnetic fields increased with increasing category for each of the wire code classification schemes. The size of the odds ratios for risk of leukaemia and high categories of wire code did not reflect the mean levels of the magnetic field in those categories in two study regions., Conclusion: Misclassification of categories of wire code is not a major source of bias in the study. Wire codes provide a proxy measure of exposure to residential magnetic fields. If magnetic fields were a risk factor for leukaemia, however, there would be some attenuation of risk estimates based on wire codes because of misclassification of exposure to magnetic fields at both extremes of the wire code range. The lack of an association between high categories of wire code and risk of leukaemia cannot be explained by a failure of the wire code classification schemes to estimate exposure to magnetic fields in the study area.

As part of a comprehensive study of residential magnetic field exposure in nine midwestern and mid-Atlantic states, we evaluated the use of appliances by 640 patients with acute lymphoblastic leukemia, 0-14 years of age, diagnosed between 1989 and 1993, and 640 matched control children. Mothers were interviewed regarding use of electrical appliances during their pregnancy with the subject and the child's postnatal use. The risk of acute lymphoblastic leukemia was elevated in children whose mothers reported use of an electric blanket or mattress pad during pregnancy [odds ratio (OR) = 1.59; 95% confidence interval (CI) = 1.11-2.29] but was reduced for use of sewing machines during pregnancy (OR = 0.76; 95% CI = 0.59-0.98). The risk of acute lymphoblastic leukemia was increased with children's use of electric blankets or mattress pads (OR = 2.75; 95% CI = 1.52-4.98) and three other electrical appliances (hair dryers, video machines in arcades, and video games connected to a television), but the patterns of risk for duration in years of use and frequency of use were inconsistent for most appliances used by children. Risks rose with increasing number of hours per day children spent watching television, but risks were similar regardless of the usual distance from the television. The inconsistency in the dose-response patterns for many appliances, reporting and selection bias, and the lack of an effect for measured 60 Hertz magnetic fields or wire codes in our companion study must be considered before ascribing these associations to exposures from magnetic fields.

Background: Several ecologic analyses have shown significant positive associations between mean indoor radon concentrations and risk of leukemia at all ages (acute myeloid leukemia and chronic lymphocytic leukemia) and for children (all leukemia, acute myeloid leukemia, and acute lymphoblastic leukemia [ALL]). As part of an age-matched, case-control study of childhood ALL in the United States, we investigated the association between the incidence of ALL in children under age 15 years and indoor radon exposure., Methods: Radon detectors were placed in current and previous homes of subjects where they resided for 6 months or longer. Children were included in analyses if radon measurements covered 70% or more of the 5-year period prior to diagnosis for case subjects (or from birth for case subjects under age 5 years) and the corresponding reference dates for control subjects. Radon levels could be estimated for 97% of the exposure period for the eligible 505 case subjects and 443 control subjects., Results: Mean radon concentration was lower for case subjects (65.4 becquerels per cubic meter [Bqm(-3)]) than for control subjects (79.1 Bqm(-3)). For categories less than 37, 37-73, 74-147, and 148 or more Bqm(-3) of radon exposure, relative risks based on matched case-control pairs were 1.00, 1.22, 0.82, and 1.02, respectively, and were similar to results from an unmatched analysis. There was no association between ALL and radon exposure within subgroups defined by categories of age, income, birth order, birth weight, sex, type of residence, magnetic field exposure, parental age at the subject's birth, parental occupation, or parental smoking habits., Conclusions: In contrast to prior ecologic studies, the results from this analytic study provide no evidence for an association between indoor radon exposure and childhood ALL.

Several earlier case-control studies reported inverse associations of cervical squamous intraepithelial lesions (SIL) with high dietary or biomarker levels of carotenoids, folate, and vitamins C and E. However, most studies did not measure the primary causal factor, cancer-associated genital human papillomaviruses (HPV), now detected by sensitive viral DNA tests. This nested case-control study assessed whether high dietary intakes of these nutrients, plus zinc and vitamin A, reduced SIL risk in cancer-associated HPV DNA-positive women. Using a 60-item food-frequency questionnaire, nutrient estimates were obtained for 33 incident cases with high-grade lesions, 121 with low-grade lesions, 97 with equivocal SIL, and 806 cytologically normal controls sampled from a large prospective cohort study. Baseline cervicovaginal lavages were tested for HPV DNA by the polymerase chain reaction. Among DNA-positive cases (n = 68) and controls (n = 69), age-adjusted odds ratios (ORs) of SIL in the highest vs. the lowest nutrient quartiles were 1.4 [95% confidence interval (CI) = 0.5-4.2] for vitamin A, 0.6 (CI = 0.2-2.0) for beta-carotene, 1.3 (CI = 0.4-3.6) for vitamin C, 1.0 (CI = 0.4-3.6) for vitamin E, 0.7 (CI = 0.3-2.1) for folate, and 0.8 (CI = 0.3-2.2) for zinc. ORs in HPV DNA-negative women approximated 1.0, with the exception of vitamin E (OR = 0.5, CI = 0.3-0.9). These results do not support a protective role for the above nutrients against low-grade or equivocal SIL, which constituted the majority of diagnoses in this study.

We conducted a methodologic study to validate a quantitative retrospective exposure assessment method used in a follow-up study of workers exposed to benzene. Assessment of exposure to benzene was carried out in 672 factories in 12 cities in China. Historical exposure data were collected for 3179 unique job titles. The basic unit for exposure assessment was a factory/work-unit/job-title combination over seven periods between 1949 and 1987. A total of 18,435 exposure estimates was developed, using all available historical information, including 8477 monitoring data. Overall, 38% of the estimates were based on benzene monitoring data. The highest time-weighted average exposures occurred in the rubber industry (30.7 ppm), particularly for rubber glue applicators (52.6 ppm). Because of its recognized link with benzene exposure, the association between a clinical diagnosis of benzene poisoning (hematotoxicity) and benzene exposure was evaluated (412 cases and 614,509 person-years) to validate the exposure-assessment method. Relative risks of benzene hematotoxicity increased very sharply with increasing estimated intensity of benzene exposure. Odds ratios were 2.2 (95% CI: 1.7-2.9), 4.7 (95% CI: 3.4-6.5), and 7.2 (95% CI: 5.3-9.8) for the intensity levels of less than 5 ppm, 5-19 ppm, 20-39 ppm, and 40 and more ppm, respectively. This sharp trend between benzene hematotoxicity and estimated exposure to benzene indicated that the exposure-estimation method used in this cancer epidemiology study is reliable.

The use and interpretation of energy-adjustment regression models in nutritional epidemiology has been vigorously debated recently. There has been little discussion, however, regarding the effect of dietary measurement error on the performance of such models. Contrary to conventional assumptions invoked in the standard treatment of the effect of measurement error in regression analysis, reporting errors in dietary studies are usually biased, correlated with true nutrient intakes and with each other, heteroscedastic, and nonnormally distributed. Methods developed in this paper allow for this more complex error structure and are therefore more appropriate for dietary data. For practical illustration, these methods are applied to data from the Women's Health Trial Vanguard Study. The results demonstrate considerable shrinkage in the magnitude of the estimated main exposure effect in energy-adjustment models due to attenuation of the true effect and contamination from the effect of an adjusting covariate. In most cases, this shrinkage causes a sharply reduced statistical power of the corresponding significance test in comparison with measurement without error. These results emphasize the need to understand the measurement error properties of dietary instruments through validation/calibration studies and, where possible, to correct for the impact of measurement error when applying energy-adjustment models.

Background: Benzene is a widely distributed environmental contaminant known to cause leukemia, particularly acute nonlymphocytic leukemia, and perhaps other hematologic neoplasms and disorders. Few epidemiologic studies, however, have been able to address relationships between the extent of benzene exposure and the level of risk., Purpose: A large cohort study was carried out in China to evaluate the risks of developing specific hematologic neoplasms and selected related disorders in relationship to quantitative estimates of occupational benzene exposure., Methods: A cohort of 74828 benzene-exposed and 35805 unexposed workers employed from 1972 through 1987 in 12 cities in China was identified and followed to determine the incidence of hematologic neoplasms and related disorders. Estimates of benzene exposure were derived from work histories and available historic benzene measurements. Existing pathologic material and supporting medical records were reviewed to establish diagnoses of disease. Relative risks (RRs) (i.e., ratios of incidence rates for specific hematologic neoplasms and related disorders in the benzene-exposed group to incidence rates in the unexposed group) were determined by use of Poisson regression analysis, with stratification by age and sex., Results: For workers historically exposed to benzene at average levels of less than 10 parts per million (ppm), the RR for all hematologic neoplasm combined was 2.2 (95% confidence interval [CI] = 1.1-4.2), and, for the combination of acute nonlymphocytic leukemia and related myelodysplastic syndromes, the RR was 3.2 (95% CI = 1.0-10.1). For individuals who were occupationally exposed to benzene at constant levels of 25 ppm or more, the RR for the combination of acute nonlymphocytic leukemia and related myelodysplastic syndromes was 7.1 (95% CI = 2.1-23.7). Workers with 10 or more years of benzene exposure had an RR of developing non-Hodgkin's lymphoma of 4.2 (95% CI = 1.1-15.9), and the development of this neoplasm was linked most strongly to exposure that had occurred at least 10 years before diagnosis (i.e., distant exposure) (P for trend = .005, two-sided). In contrast, the risk for the combination of acute nonlymphocytic leukemia and related myelodysplastic syndromes was significantly increased among those with more recent benzene exposure (P for trend = .003, two-sided), but it was not linked to distant exposure (P for trend = .51, two-sided)., Conclusions: The results of this study suggest that benzene exposure is associated with a spectrum of hematologic neoplasms and related disorders in humans. Risks for these conditions are elevated at average benzene-exposure levels of less than 10 ppm and show a tendency, although not a strong one, to rise with increasing levels of exposure. The temporal pattern of benzene exposure appears to be important in determining the risk of developing specific diseases.

Benzene is a ubiquitous occupational hematotoxin and leukemogen, but people vary in their response to this toxic agent. To evaluate the impact of interindividual variation in enzymes that activate (i.e., CYP2E1) and detoxify (i.e., NQO1) benzene and its metabolites, we carried out a case-control study in Shanghai, China, of occupational benzene poisoning (BP; i.e., hematotoxicity), which we show is itself strongly associated with subsequent development of acute nonlymphocytic leukemia and the related myelodysplastic syndromes (relative risk, 70.6; 95% confidence interval, 11.4-439.3). CYP2E1 and NQO1 genotypes were determined by PCR-RFLP, and CYP2E1 enzymatic activity was estimated by the fractional excretion of chlorzoxazone (fe(6-OH)) for 50 cases of BP and 50 controls. Subjects with both a rapid fe(6-OH). and two copies of the NQO1 609C-->T mutation had a 7.6-fold (95% confidence interval, 1.8-31.2) increased risk of BP compared to subjects with a slow fe(6-OH) who carried one or two wild-type NQO1 alleles. In contrast, the CYP2E1 PstI/RsaI polymorphism did not influence BP risk. This is the first report that provides evidence of human susceptibility to benzene-related disease. Further evaluation of susceptibility for hematotoxicity and hematological malignancy among workers with a history of occupational exposure to benzene is warranted.

Serological assays for measuring antibodies to human papillomavirus type 16 (HPV-16) virus-like particles (VLPs) have become important epidemiologic tools in recent years. However, the interlaboratory replicability of these assays has not been assessed. In this investigation, three laboratories tested a panel of specimens obtained from two different groups: 265 subjects in a vulvar cancer case-control study and 107 healthy volunteer blood donors. Each laboratory used an enzyme-linked immunosorbent assay (ELISA), but no attempt was made to standardize assay procedures among the three laboratories. The data showed good day-to-day intralaboratory replicability in laboratory 1 (correlation coefficient, > or = 0.88) and good intra-assay variability in laboratory 3 (correlation coefficient, > or = 0.93). Interlaboratory correlations, likewise, ranged between 0.61 and 0.80 in both case-control study subjects and healthy blood donors, indicating that ELISA optical density (OD) values between laboratories were linearly related regardless of the population. Kappa coefficients (kappa), based on each laboratory's categorical interpretation of its results (as positive or negative), showed good agreement (kappa, > 0.6) in case-control study subjects and moderate agreement (kappa, > or = 0.4) in blood donors, a population that had few strongly positive sera. When OD values near seropositive cutoffs were treated as indeterminates, there was little discordance between laboratories in either population. The data suggest that each laboratory measured the same humoral immune response and that their HPV-16 VLP ELISAs performed similarly (Pearson correlations). Interlaboratory differences, however, probably due to reagents and procedures, were considerably greater than intralaboratory day-to-day variability. Interlaboratory agreement in determining seropositivity (kappa) could be improved by sharing positive and negative serum controls and by treating marginal results as indeterminate. As part of continuing cooperation to improve interlaboratory agreement, we are preparing bulk serum control specimens to be shared and made available to interested researchers.

Background: Carriers of germ-line mutations in BRCA1 and BRCA2 from families at high risk for cancer have been estimated to have an 85 percent risk of breast cancer. Since the combined frequency of BRCA1 and BRCA2 mutations exceeds 2 percent among Ashkenazi Jews, we were able to estimate the risk of cancer in a large group of Jewish men and women from the Washington, D.C., area., Methods: We collected blood samples from 5318 Jewish subjects who had filled out epidemiologic questionnaires. Carriers of the 185delAG and 5382insC mutations in BRCA1 and the 6174delT mutation in BRCA2 were identified with assays based on the polymerase chain reaction. We estimated the risks of breast and other cancers by comparing the cancer histories of relatives of carriers of the mutations and noncarriers., Results: One hundred twenty carriers of a BRCA1 or BRCA2 mutation were identified. By the age of 70, the estimated risk of breast cancer among carriers was 56 percent (95 percent confidence interval, 40 to 73 percent); of ovarian cancer, 16 percent (95 percent confidence interval, 6 to 28 percent); and of prostate cancer, 16 percent (95 percent confidence interval, 4 to 30 percent). There were no significant differences in the risk of breast cancer between carriers of BRCA1 mutations and carriers of BRCA2 mutations, and the incidence of colon cancer among the relatives of carriers was not elevated., Conclusions: Over 2 percent of Ashkenazi Jews carry mutations in BRCA1 or BRCA2 that confer increased risks of breast, ovarian, and prostate cancer. The risks of breast cancer may be overestimated, but they fall well below previous estimates based on subjects from high-risk families.

This paper reports on the enrollment phase of a population-based natural history study of cervical neoplasia in Guanacaste, a rural province of Costa Rica with consistently high rates of invasive cervical cancer. The main goals of the study are to investigate the role of human papillomavirus (HPV) infection and its co-factors in the etiology of high-grade cervical neoplasia, and to evaluate new cervical cancer screening technologies. To begin, a random sample of censal segments was selected and enumeration of all resident women 18 years of age and over was conducted with the aid of outreach workers of the Costa Rican Ministry of Health. Of the 10738 women who were eligible to participate, 10049 (93.6%) were interviewed after giving written informed consent. After the interview on cervical cancer risk factors was administered, a pelvic examination was performed on those women who reported previous sexual activity. The pelvic examination included a vaginal pH determination and collection of cervical cells for cytologic diagnosis using three different techniques. Additional cervical cells were collected for determination of the presence and amount of DNA from 16 different types of HPV, and two photographic images of the cervix were taken and interpreted offsite by an expert colposcopist. Finally, blood samples were collected for immunologic and micronutrient assays. Women with any abnormal cytologic diagnosis or a positive Cervigram, as well as a sample of the whole group, were referred for colposcopy, and biopsies were taken when lesions were observed. The enrollment screening will serve as the basis for a prevalent case-control study, and the members of the cohort free from serious disease will be followed actively, at intervals of no more than a year, to study the natural history of HPV infection and the origins of high-grade squamous intraepithelial lesions (HSIL). Details of the field operation are outlined, with particular reference to the realization of this kind of study in developing countries. Descriptive data on the prevalence of disease and exposure to various risk factors are also presented.

Serological markers of squamous intraepithelial lesions (SILs), the precursors of cervical cancer, have not been studied extensively. To screen for antibody responses that might be associated with SILs, we measured IgG and/or IgA to nine antigens based on papillomaviruses, the infectious cause of SIL and cervical cancer, using an ELISA format. Cases were 59 women with low grade SIL (LSIL) and 38 with high grade SIL (HSIL). Controls were 50 women chosen to minimize the possibility that they ever had SILs [individuals who had no history of SIL and repeatedly tested negative for cervical human papillomavirus (HPV) DNA], frequency age-matched to cases. The data showed that five antibodies had strong positive associations with SILs and that one was inversely related to SILs. By studying these antibodies in pairs, furthermore, we found that case-control differences were enhanced. In particular, the combination of IgG to an epitope in the E6 protein of HPV 16 (E6:10) and IgA to HPV 16 virus-like particles (VLPs) was detected in 53% of LSILs and 65% of HSILs but only 9% of controls. These same responses were both negative in just 6% of LSILs and zero HSILs, compared to 59% of controls. Notably, E6:10 IgG and HPV 16 VLP IgA were not correlated with each other, and the other antibody responses positively associated with SILs could be broken into two groups: those correlated with E610 IgG and those correlated with HPV 16 VLP IgA. Overall, the data suggest that several papillomavirus antibodies may be strongly related to SILs, and that they can be divided into at least two independent groups of humoral immune reactions.

Background: Cigarette smoking has been shown to increase oxidative DNA damage in human sperm cells. Assessment of the role of cigarette smoking in the etiology of childhood cancer has focused primarily on the effect of maternal smoking. Similar studies in relation to paternal smoking, however, have been inconclusive. Few studies have evaluated the effect of paternal smoking in the preconception period, and most of these could not disentangle the effects of paternal from maternal smoking., Purpose: We investigated the relationship of paternal smoking, particularly in the preconception period, with childhood cancer among offspring of the nonsmoking mothers., Methods: We conducted a population-based, case-control study in Shanghai, People's Republic of China, where the prevalence of smoking is high among men but extremely low among women. The study included 642 childhood cancer case patients (<15 years of age) and their individually matched control subjects. Information concerning parental smoking, alcohol drinking, and other exposures of the index child was obtained by direct interview of both parents of the study subjects. Odds ratios (ORs), derived from conditional logistic regression models, were used to measure the association between paternal smoking and risk of childhood cancers., Results and Conclusions: Paternal preconception smoking was related to a significantly elevated risk of childhood cancers, particularly acute leukemia and lymphoma. The risks rose with increasing pack-years of paternal preconception smoking for acute lymphocytic leukemia (ALL) (P for trend = .01), lymphoma (P for trend = .07), and total cancer (P for trend = .006). Compared with children whose fathers had never smoked cigarettes, children whose fathers smoked more than five pack-years prior to their conception had adjusted ORs of 3.8 (95% confidence interval [CI] = 1.3-12.3) for ALL, 4.5 (95% CI = 1.2-16.8) for lymphoma, 2.7 (95% CI = 0.8-9.9) for brain tumors, and 1.7 (95% CI = 1.2-2.5) for all cancers combined. Statistically significant increased risks of cancer were restricted to children under the age of 5 years at diagnosis or those whose fathers had smoked during all of the 5 years prior to conception., Implications: Further studies are needed to confirm the association of paternal smoking with increased risk of cancer in offspring, to clarify the pattern of risks in relation to the timing of cigarette smoking, and to elucidate the biologic mechanism involved in predisposing the offspring to cancer. For example, it may be that paternal smoking induces prezygotic genetic damage that, in turn, acts as the predisposing factor.

Previous occupational cohort studies of benzene-exposed workers have for the most part used only death certificates to validate diagnoses of workers developing leukemia and other hematopoietic and lymphoproliferative malignancies and related disorders (HLD). In a follow-up study of 74,828 benzene-exposed workers and a comparison group of 35,805 nonexposed workers from 12 cities in China, we sought to characterized clinicopathologically and to confirm diagnoses of all cases of HLD. Using medical records, laboratory hematology results, and histopathology, U.S. and Chinese expert hematopathologists, blinded to exposure status, carried out a detailed review using standardized evaluation forms. Key among the findings were a notable diversity of malignant and nonneoplastic hematopoietic and lymphoproliferative disorders, documentation of excess myelodysplastic syndromes among benzene workers, and widespread dyspoiesis involving all hematopoietic cell lines. As sophisticated clinicopathologic characterization and corresponding classification schemes for HLD become increasingly widespread, it is recommended that future epidemiologic investigations of benzene workers incorporate similarly detailed morphologic evaluation. In extending follow-up of this cohort of young workers, we will continue to use all available clinical, laboratory hematology, and pathology data as well as cytogenetic and biochemical markers to characterized various HLD outcomes. These careful surveillance mechanisms should also provide additional insight into carcinogenic mechanisms of benzene and allow comparison of the molecular pathogenesis of HLD induced by benzene versus chemotherapy, radiation, or other exposure.

A large cohort of 74,828 benzene-exposed and 35,805 nonexposed workers employed between 1972 and 1987 in 12 cities in China was followed to determine mortality from all causes. Benzene-exposed study subjects were employed in a variety of occupations including coating applications, and rubber, chemical, and shoe production. Mortality was slightly increased among workers with greater cumulative exposure to benzene (ptrend < 0.05), but this excess was largely due to cancer deaths (ptrend < 0.01). Deaths due to lymphatic and hematopoietic malignancies (ptrend = 0.01) and lung cancer (ptrend = 0.01) increased with increasing cumulative exposure to benzene. Investigations continue to relate benzene exposure to specific lymphatic and hematopoietic malignancies and other causes of death.

We present a validation study of a quantitative retrospective exposure assessment method used in a follow-up study of workers exposed to benzene. Assessment of exposure to benzene was carried out in 672 factories in 12 cities in China. Historical exposure data were collected for 3179 unique job titles. The basic unit for exposure assessment was a factory/work unit/job title combination over seven periods between 1949 and 1987. A total of 18,435 exposure estimates was developed, using all available historical information, including 8477 monitoring data. Overall, 38% of the estimates were based on benzene monitoring data. The highest time-weighted average exposures were observed for the rubber industry (30.7 ppm) and for rubber glue applicators (52.6 ppm). Because of its recognized link with benzene exposure, the association between a clinical diagnosis of benzene poisoning and benzene exposure was evaluated to validate the assessment method that we used in the cohort study. Our confidence in the assessment method is supported by the observation of a strong positive trend between benzene poisoning and various measures, especially recent intensity of exposure to benzene.

An expanded cohort study of 74,828 benzene-exposed and 35,805 unexposed workers were followed during 1972 to 1987, based on a previous study in 12 cities in China. A small increase was observed in total cancer mortality among benzene-exposed compared with unexposed workers (relative risk [RR] = 1.2). Statistically significant excesses were noted for leukemia (RR = 2.3), malignant lymphoma (RR = 4.5), and lung cancer (RR = 1.4). When risks were evaluated by leukemia subtype, only acute myelogenous leukemia was significantly elevated (RR = 3.1), although nonsignificant excesses were also noted for chronic myelogenous leukemia (RR = 2.6) and acute lymphocytic leukemia (RR = 2.3). A significant excess was also found for aplastic anemia.

A large cohort study of 74,828 benzene-exposed and 35,805 unexposed workers employed between 1972 and 1987 in 12 cities in China were followed to determine mortality from all causes and the incidence of lymphohematopoietic malignancies and other hematologic disorders. Benzene-exposed study subjects were employed in a variety of occupations, including painting, printing, and the manufacture of footwear, paint, and other chemicals. All-cause mortality was similar in the benzene-exposed and unexposed comparison group. Statistically significant excess deaths were noted among benzene-exposed subjects for leukemia (RR = 2.3, 95% CP 1.1-5.0), malignant lymphoma (RR = 4.5, 95% CI: 1.3-28.4), and nonneoplastic diseases of the blood (RR = 95% CP 2.5-infinity), and a marginally significant excess was noted for lung cancer (RR = 1.4, 95% CI: 1.0-2.0). Risk was significantly elevated for the incidence of all lymphohematopoietic malignancies (RR = 2.6, 95% CI: 1.5-5.0), malignant lymphoma (RR = 3.5, 95% CI: 1.2-14.9), and leukemia (RR = 2.6, 95% CI.. 1.3-5.7). Among the leukemia subtypes, only acute myelogenous leukemia (AML) incidence was significantly elevated (RR = 3.1, 95% CI: 1.2-10.7), although nonsignificant excesses were also noted for chronic myelogenous leukemia (CML) (RR = 2.6, 95% CI: 0.7-16.9) and lymphocytic leukemias (RR = 2.8, 95% CI.. 0.5-54.5). Significant excesses were found for aplastic anemia (RR = infinity, 95% CI: 2.2-co) and myelodysplastic syndrome (RR = infinity, 95% CI: 1.7-infinity). Employment in benzene-associated occupations in China is associated with a wide spectrum of myelogenous and lymphocytic malignant diseases and related disorders. Investigations continue to assess the nature of these associations.

We examined the relation between area measurements of residential magnetic fields and personal dosimetry measurements among 64 control children age 2-14 years from the National Cancer Institute-Children's Cancer Group's nine-state case-control study of childhood leukemia. During a typical weekday, an activity diary was completed, and a 24-hour measurement was obtained in each child's bedroom. According to the activity diaries, children spent more than 40% of the 24 hours in their bedrooms, and 68% of their time at home. We found that at-home personal dosimetry levels were highly correlated with total personal dosimetry levels in children under 9 years (Spearman correlation coefficient, R = 0.94), whereas the correlation was lower in older children (R = 0.59). For all children combined, bedroom 24-hour measurements correlated well with at-home personal dosimetry levels (R = 0.76). The 24-hour bedroom measurement was a useful predictor of both at-home and total personal dosimetry measurements. Particularly for younger children, our data suggest that in-home area measurements predict both current residential and current total magnetic field exposures. This information will be valuable for assessing the validity of exposure assessment in previous and ongoing studies and for developing measurement protocols for future studies.

In a population-based case-control study including 449 directly interviewed cases and 707 controls, we assessed the risk of renal cell cancer associated with height, weight, body mass index (BMI), and frequency of weight changes. Odds ratios and 95% confidence intervals were estimated by using logistic regression models. Among women, risk increased with increasing usual BMI (P for trend < 0.001). A nearly 4-fold risk was found among the 10% of women with the highest usual BMI (odds ratio = 3.8; confidence interval = 1.7-8.4). Among men, no clear trend was observed with usual weight or BMI, although the highest risk (30-50%) generally was seen among those in the upper deciles of weight or BMI. There was no clear indication that excess BMI early or late in life disproportionately affected risks. Risk also was not related to patterns of weight fluctuations or use of diet pills. Our study supports previous observations linking renal cell cancer risk to increased BMI among women and suggests a weaker association in men. Given the increasing prevalence of obesity and the rising incidence of renal cell cancer in the United States, additional studies are needed to disentangle the effects of BMI from various correlates and to identify the mechanisms by which obesity affects risk.

In Shanghai, China, age-adjusted incidence rates for pancreatic cancer have increased steadily, beginning in the early 1970s. To examine the effects of diet on this cancer, a population-based case-control study was conduct. Cases (n = 451) were permanent residents of Shanghai, 30-74 years of age, newly diagnosed with pancreatic cancer between October 1, 1990 and June 30, 1993. Decreased cases (19%) were excluded from the study. Controls (n = 1552) were selected among Shanghai residents, frequency matched to cases by gender and age. Information on usual adult dietary intake was obtained by trained interviewers in person, using a food frequency questionnaire. Dietary associations were measured by odds ratios and 95% confidence intervals. Risks of pancreatic cancer were inversely associated with consumption of vegetables (P for trend among men = 0.03; among women = 0.15) and fruits (P among men = 0.02; among women = 0.08). Reductions in risk were related also to intake of dietary fiber and micronutrients abundant in plant sources, such as vitamins C and E and carotene. There was also an inverse association with egg consumption (P for trend among men = 0.08; among women = 0.001). No consistent positive associations were observed with intake of other food groups, including preserved animal foods, fresh red meat, organ meat, poultry, and staple foods. On the other hand, risks increased with frequency of consumption of preserved vegetables and foods that were deep fried, grilled, cured, or smoked, providing clues to the possible role of nitrosamines, polycyclic aromatic hydrocarbons, and heterocyclic aromatic amines. The inverse associations observed with intake of dietary fat and protein in our study were unexpected, although these findings were based on consumptions well below the average intake in Western countries, where most previous studies on pancreatic cancer were conducted. Our results suggest that dietary variations have contributed little to the rising trends of pancreatic cancer in Shanghai. However, given the improving food availability and changing dietary patterns in China, further study of dietary and nutritional risk factors for pancreatic cancer appears warranted.

To evaluate renal pelvis and ureter (RPU) cancer risk in relation to lifetime use of analgesics, a population-based case-control study was carried out in 3 areas of the United States. Among 502 cases and 496 controls diagnosed and interviewed during 1983-1986, no significant increases in risk were found for any of the non-prescription and prescription analgesics evaluated or among regular users of phenacetin, acetaminophen or aspirin. Neither cumulative lifetime ingestion nor duration of regular use of these 3 drugs, whether alone or in combination, was associated with significantly increased risk of RPU cancer, although a slight excess was observed among long-term users of acetaminophen. Risk was not increased among persons reporting highest cumulative dose and/or longest duration of phenacetin use. Although our study of RPU cancer is the largest to date, it was nonetheless limited by the small number of regular analgesic users and the relatively low response rates. Because of the relatively recent onset of widespread use of acetaminophen, its pharmacologic similarity to phenacetin, a known urothelial carcinogen, and the elevation in risk seen in long-term users, further surveillance of this analgesic is warranted.

Although recent data have suggested an association between renal cell cancer and the use of diuretics, it remains unclear whether these medications or hypertension is the important risk factor. In a population-based case-control study including 440 renal cell cancer cases, spouses of an additional 151 cases, and 691 controls, we assessed renal cell cancer risk associated with hypertension and use of diuretics and other antihypertensive medications. Risks increased with the use of diuretics or other drugs that lower blood pressure, especially among persons who reported no history of hypertension. After adjustment for hypertension, the use of diuretics alone was associated with a 40% excess risk (OR = 1.4; 95% CI = 0.8-2.2), while use of other antihypertensive drugs was linked to a 2-fold risk (OR = 2.0; 95% CI = 1.2-3.3). The excess risk was not restricted to any specific products, and no trend was observed with estimated lifetime consumption of any product. Furthermore, risk was not potentiated by the presence of both hypertension and the use of antihypertensive drugs. Among persons who did not use antihypertensive drugs, a history of hypertension was associated with a significant 40-50% excess risk of renal cell cancer. Excluding subjects with hypertension diagnosed within 5 years of cancer diagnosis or interview had only a small effect on risk. These findings suggest small effects on renal cell cancer risk associated with hypertensive disease and with the use of diuretics and other antihypertensive drugs, but it is difficult to disentangle the separate effects due to potential misclassification of highly correlated events.(ABSTRACT TRUNCATED AT 250 WORDS)

Most of the literature on the effect of nondifferential misclassification and errors in variables either addresses binary exposure variables or discusses continuous variables in the classical error model, where the error is assumed to be uncorrelated with the true value. In both of these situations, an imperfectly measured exposure always attenuates the relation, at least in the univariate setting. Furthermore, measuring a confounder with error independent of the exposure, even while measuring the exposure of interest perfectly, leads to partial control of the confounding. For many variables measured in epidemiology, particularly those based on self-report, however, errors are often correlated with the true value, and these rules may not apply. Epidemiologists need to be wary of deviations from the classical error model, since poor measurement might occasionally explain a positive finding even when the error does not differ by disease status.

A population-based cohort of 120 Danish men, discharged with a hospital diagnosis of primary hemochromatosis from 1977 to 1989, was followed up to 1989 for subsequent cancer risk. Nineteen subjects (including 6 with primary liver cancers) were excluded from the analysis, either because they died within the same month of hemochromatosis diagnosis or because they had cancer prior to diagnosis of hemochromatosis. Among the 101 remaining subjects, 4 primary liver cancers occurred one year or more after the diagnosis of hemochromatosis, far surpassing the expected number based on incidence rates from the Danish population (standardized incidence ratio 92.9, 95% confidence interval 25.0 to 237.9). The excess of liver cancer was associated with cirrhosis and included cholangiocarcinoma as well as hepatocellular carcinoma. Significantly elevated risks were also observed for non-hepatic cancers (13 cases; SIR 3.5, 95% CI 1.9 to 6.0), notably esophageal cancer (2 cases; SIR 42.9, 95% CI 4.8 to 154.9) and skin melanoma (2 cases; SIR 27.8, 95% CI 3.1 to 100.3). The results of this population-based study are in accordance with the hypothesis that patients with primary hemochromatosis have a substantial risk of primary liver cancer. Further studies of hemochromatosis may be useful in clarifying the relation of non-hepatic malignancies to body iron stores in the general population.

Benzene is recognized internationally as a leukemogen, but the available data to clarify dose-response relationships and examine risks of malignancies other than leukemia are sparse. A collaborative study was therefore carried out to expand on a previous retrospective cohort mortality study of Chinese benzene-exposed workers. Methods and resources used in the 16-year follow-up of 74,828 benzene-exposed and 35,805 unexposed workers employed for any length of time during 1972-1987 in 712 factories in 12 cities in China are described. Details are provided of the study organization, assessment of benzene exposures since 1949, characterization of factories and workers by exposure status, city, and sex, identification and confirmation of cancers and other deaths, and quality control procedures. The distinguishing features of the study are discussed in relation to earlier cohort studies, and study limitations as well as strengths are presented.

Background: Renal cell cancer, although still relatively uncommon, has been increasing in incidence in the United States and other countries around the world., Purpose: Since previous studies have suggested an association with high intake of meat, we sought to further examine the role of diet in renal cell cancer risk., Methods: Patients with histologically confirmed renal cell cancer that had been diagnosed between July 1, 1988, and December 31, 1990, were identified through the Minnesota Cancer Surveillance System, a statewide cancer registry. The patients eligible for inclusion in this study were white residents of Minnesota between 20 and 79 years of age. Control subjects were selected from the general population of Minnesota residents; subjects under age 65 were selected by use of a random-digit-dialing method and those 65 years or older were sampled from the Health Care Financing Administration files. Population-based control subjects were frequency-matched to cases by sex and 5-year age groups. A total of 690 patients and 707 control subjects were interviewed. Patients and control subjects were similar in distribution by sex, age, and educational level. Usual adult dietary intakes were assessed by questionnaire, and odds ratios were calculated by logistic regression analyses., Results: Significantly increased risks of renal cell cancer were observed with increasing consumption of several food groups, including red meat (P for trend = .05), high-protein foods (P = .01), and staple (grains, breads, and potatoes) foods (P = .009). When examined by macronutrient status, risks increased monotonically with the amount of protein intake, from 1.2 (95% confidence interval [CI] = 0.7-1.9) to 1.4 (95% CI = 0.8-2.5) and 1.9 (95% CI = 1.0-3.6) (P for trend = .03) in the second, third, and fourth quartiles of intake, respectively, after adjustment for age, sex, caloric intake, body mass index, and cigarette smoking. No significant or consistent associations were detected with the intake of other dietary nutrients or beverages., Conclusion: Although an independent effect of dietary protein has not been previously associated with renal cell cancer, high protein consumption has been related to development of other chronic renal conditions that may predispose an individual to this cancer., Implication: These findings should prompt further study of dietary protein and its potential contribution to the origins of renal cell cancer.

Gender differences in risk for leukemia and other selected and combined disease categories were examined by major occupational category for 74,828 benzene-exposed workers compared to 35,805 unexposed workers from 12 cities in China. No significant differences in the relative risks for total mortality and cancer mortality were found between female and male benzene-exposed workers, although risks tended to be somewhat higher among male than among female employees. Both female and male workers in several occupational categories had notably increased risks for all hematopoietic and lymphoproliferative (HLP) malignant and nonmalignant disorders combined and for total leukemia. Variation in risk for HLP disorders by occupational category was observed in both genders, with highest risks for male and female chemical manufacturing workers, female nonproduction employees, and male printers. However, the numbers of leukemia and other HLP malignancies in each category were small. The findings suggest that both female and male benzene-exposed workers in several occupational categories experience excess leukemia and other HLP disorders with relatively minor gender differences. Although this population is one of the largest cohorts of benzene-exposed workers studied to date, evaluation of the observed variation in risk for HLP neoplasms among the occupational groups for workers of each gender is limited by the small numbers of these relatively rare malignancies.

Case-control studies can often be made more efficient by using frequency matching, randomized recruitment, stratified sampling, or two-stage sampling. These designs share two common features: (1) some "first-stage" variables are ascertained for all study subjects, while complete variable ascertainment is carried out for only a selected subsample, and (2) the subsampling of subjects for "second-stage" variable ascertainment depends jointly on their disease status and their observed first-stage variables. Because first-stage variables alter the subsampling fractions, standard analyses require a multiplicative specification of any joint effects of a second- and a first-stage variable. We show that by making use of missing data methods, maximum likelihood estimates can be obtained for risk parameters of interest, even those characterizing interactions between first- and second-stage variables. Joint effects can thus be modelled flexibly, with allowance for both additive and multiplicative models. Preliminary data from a case-control study of lung cancer as related to age, sex, and smoking provide an example, leading to the suggestion that the combined effect of age and smoking is multiplicative.

In population-based case-control studies, an attempt is made to identify all incident cases diagnosed in a specified population during a fixed time interval. Assuming that this goal is met allows one to obtain measures of risk other than relative risks. In this paper, we describe three approaches to estimate exposure-specific incidence rates. Approach 1 relies on estimating crude incidence rates of the disease in strata defined, for instance, by age and geographic area, and combining them with relative risk estimates from the case-control data. In approaches 2 and 3, baseline incidence rates and relative risks are estimated jointly. Approach 2 is based on a pseudo-likelihood, while, in approach 3, the problem is regarded as a missing data problem and a full likelihood is maximized. We applied these three approaches to a study of bladder cancer. Our three sets of estimates of exposure-specific incidence rates were in close agreement, while there appeared to be greater precision with approaches 2 and 3.

We propose an alternative to a long questionnaire that may increase quality while reducing the cost and effort of participants and researchers. In the 'partial questionnaire design', information about the exposure of interest is obtained from all subjects, while zero, one, or more disjoint subsets of questions about possible confounders are asked to randomly selected subgroups. The proposed analyses exploit the fact that the uncollected data can be considered to be missing at random. We show that it is possible to obtain high efficiency for estimating the effect of exposure of interest, adjusted for confounding, while substantially shortening average questionnaire length.

Background: Patients with non-Hodgkin's lymphoma (NHL) are at increased risk for second cancers. Few studies, however, include long-term survivors, and none report risk for second cancer among NHL patients surviving 15 or more years., Purpose: Our aim was to examine the pattern of second cancers among long-term survivors of NHL., Methods: A cohort of 6171 patients diagnosed with NHL as a first primary cancer and who survived 2 or more years was identified within population-based tumor registries in Sweden, Ontario, and Iowa and within the affiliated tumor registry of The Netherlands Cancer Institute. Nearly 1000 NHL patients lived 15 or more years after diagnosis. Tumor registry files were searched for new invasive primary malignancies., Results: Second cancers were reported in 541 subjects (observed-to-expected ratio [O/E] = 1.37; 95% confidence interval = 1.26-1.49), with significant excesses seen for all solid tumors (O/E = 1.28), acute nonlymphocytic leukemia (O/E = 4.83), melanoma (O/E = 2.38), Hodgkin's disease (O/E = 12.0), and cancers of the lung (O/E = 1.36), brain (O/E = 2.33), kidney (O/E = 2.07), and bladder (O/E = 1.77). Among 15-year survivors, significantly increased risks persisted for all second cancers (O/E = 1.45), solid tumors (O/E = 1.37), bladder cancer (O/E = 3.24), and Hodgkin's disease (O/E = 25.0). The actuarial risk of developing a second cancer 3-20 years after diagnosis of NHL was 21%, compared with a population expected cumulative risk of 15%., Conclusions: Patients with NHL continue to be at significantly elevated risk of second primary cancer for up to two decades following diagnosis. The pattern of risk suggests the influence of treatment as well as factors associated with the underlying disease., Implications: Quantitative studies of second cancer following NHL are needed to clarify the role of antecedent therapy, shared risk factors, host susceptibility, and other etiologic and diagnostic influences. Despite the generally advanced age of patients with NHL, the persistently elevated risk of second cancers should alert clinicians to the importance of continued medical surveillance.

Risk factors for pancreatic cancer were evaluated in a cohort study of 17,633 White men in the United States who responded to a mailed questionnaire in 1966 and were followed-up through 1986 for mortality. Cigarette smoking and alcohol consumption were found to be important risk factors for pancreatic cancer. Risks increased significantly with number of cigarettes smoked, reaching fourfold for smokers of 25 or more cigarettes per day relative to nonsmokers. Alcohol intake also was related significantly to risk, with consumers of 10 or more drinks per month having three times the risk of nondrinkers, but dose-response trends among drinkers were not smooth. Coffee consumption was unrelated to risk. Dietary analyses revealed a rising rate of pancreatic cancer mortality with increasing consumption of meat after adjustment for other risk factors. Men in the highest quartile of meat intake had about three times the risk of those in the lowest quartile. No consistent association, however, was observed for consumption of fruits, vegetables, or grains. This study confirms cigarette smoking as an important risk factor for pancreatic cancer, and provides evidence that elevated intake of alcohol and meat may increase the risk of this fatal malignancy.

Background: Experimental studies have provided strong evidence that human papillomavirus (HPV) is the long-sought venereal cause of cervical neoplasia, but the epidemiologic evidence has been inconsistent., Purpose: Given improvements in HPV testing that have revealed a strong link between sexual activity history and cervical HPV infection, we conducted a large case-control study of HPV and cervical intraepithelial neoplasia (CIN) to evaluate whether sexual behavior and the other established risk factors for CIN influence risk primarily via HPV infection., Methods: We studied 500 women with CIN and 500 control subjects receiving cytologic screening at Kaiser Permanente, a large prepaid health plan, in Portland, Ore. The established epidemiologic risk factors for CIN were assessed by telephone interview. We performed HPV testing of cervicovaginal lavage specimens by gene amplification using polymerase chain reaction with a consensus primer to target the L1 gene region of HPV. Unconditional logistic regression analysis was used to estimate relative risk of CIN and to adjust the epidemiologic associations for HPV test results to demonstrate whether the associations were mediated by HPV., Results: The case subjects demonstrated the typical epidemiologic profile of CIN: They had more sex partners, more cigarette smoking, earlier ages at first sexual intercourse, and lower socioeconomic status. Statistical adjustment for HPV infection substantially reduced the size of each of these case-control differences. Seventy-six percent of cases could be attributed to HPV infection; the results of cytologic review suggested that the true percentage was even higher. Once HPV infection was taken into account, an association of parity with risk of CIN was observed in both HPV-negative and HPV-positive women., Conclusion: The data show that the great majority of all grades of CIN can be attributed to HPV infection, particularly with the cancer-associated types of HPV., Implications: In light of this conclusion, the investigation of the natural history of HPV has preventive as well as etiologic importance.

A case-control study of 107 colon cancer cases and 486 controls from an oncological clinic in Istanbul was conducted to examine the association between occupational physical activity and colon cancer in Turkey, where incidence of this disease is low. Only two of the four activity measures showed evidence of an increased colon cancer risk for sedentary jobs (time spent sitting OR = 1.5 and occupational energy expenditure OR = 1.6); neither was statistically significant. Subjects below age 55 showed higher risk associated with sedentary jobs than did the older age group, probably due to their adoption of a more western lifestyle, including dietary habits, less activity, and other factors that may interact to increase the risk of colon cancer.

In the Bethesda System, noninvasive squamous abnormalities are classified as atypical squamous cells of undetermined significance (ASQUS), low-grade squamous intraepithelial lesions, and high-grade squamous intraepithelial lesions. The Bethesda System eliminates two diagnostic distinctions that are made in the dysplasia/carcinoma in situ and cervical intraepithelial neoplasia (CIN) classifications, CIN1 vs koilocytotic atypia and CIN2 vs CIN3, and maintains three others, negative vs ASQUS, ASQUS vs koilocytotic atypia, and CIN1 vs CIN2. To determine whether the diagnostic distinctions preserved in the Bethesda System are made more consistently than those eliminated, we analyzed the interobserver reproducibility of two cytopathologists in classifying 257 smears. The findings indicate that the distinctions retained in the Bethesda System are more reproducible than those eliminated. Specifically, cases classified as koilocytotic atypia were distinguished from CIN1 no more reproducibly than predicted by chance, whereas CIN2 and CIN3 were distinguished as consistently as any other pair of diagnoses examined. In 13 cases in which there was interobserver discordance, one reviewer classified the smear as ASQUS and the other reviewer diagnosed CIN2 or CIN3. The findings in this study suggest that smears showing koilocytotic atypia and/or CIN1 may be reported as low-grade squamous intraepithelial lesions without further specification. In contrast, smears showing high-grade squamous intraepithelial lesions may be further classified as CIN2 or CIN3 in accordance with the Bethesda guidelines. Since the diagnosis of ASQUS is applied to smears showing a wide spectrum of changes, management of patients with the diagnosis of ASQUS will be facilitated by providing an explanatory note and/or recommendations when appropriate.

Among 17,633 U.S. white male insurance policy holders whose use of tobacco was characterized in a 1966 self-administered questionnaire, there were 49 deaths from non-Hodgkin's lymphoma (NHL) and 21 from multiple myeloma (MM) during a 20-year follow-up. Men who had ever smoked cigarettes had an elevated mortality from NHL (RR = 2.1; CI = 0.9-4.9), with risk almost four-fold greater among the heaviest smokers (RR = 3.8; CI = 1.4-10.1) compared with those who used no tobacco. In contrast, risk of MM was only slightly elevated among those who had ever smoked cigarettes (RR = 1.3; CI = 0.4-3.9) and without evidence of a dose-response trend. Since this is the first cohort study suggesting a link between cigarette smoking and NHL and findings from case-control studies have been inconsistent, additional clarification should be sought from larger incidence-based cohort investigations.

In 1966, a cohort of White males aged 35 or over, who were policy-holders with the Lutheran Brotherhood Insurance Society (United States), completed a mail questionnaire on tobacco use, diet, and demographic characteristics. During the 20 years of follow-up, 219 lung cancer deaths occurred. Besides the strong relationship with cigarette smoking, we observed an effect on lung cancer risk among current users of cigars or pipes who were nonsmokers of cigarettes (relative risk [RR] = 3.5, 95 percent confidence interval [CI] = 1.0-12.6) or who were past/occasional users of cigarettes (RR = 2.7, CI = 1.4-5.3). In addition, elevated risks (from 1.5 to 2.6) of lung cancer were found among craftsmen and laborers, with the highest risks among subjects who worked in the mining or manufacturing industry. No association between current (as of 1966) use of beer or hard liquor and lung cancer was observed, although past users were at elevated risk. An inverse association between lung cancer and intake of fruits was observed, and risks of lung cancer were lower among persons in the highest dietary intake quintiles of vitamins A and C. Except for oranges, however, none of the inverse associations with fruits or dietary nutrients had statistically significant trends. The findings from this cohort study add to the evidence of an adverse effect of cigar/pipe smoking and possibly protective effect of dietary factors on lung cancer risk.

The association of oral contraceptive use with liver cancer was examined in a study of 76 deaths from primary liver cancer, 22 deaths from cancer of the intrahepatic bile ducts, and 629 controls among women aged 25 to 49 years. The subjects in the study are from the 1986 National Mortality Followback Survey, which included a questionnaire sent or administered to the next-of-kin of almost 20,000 deceased individuals in the United States. Information on a number of lifestyle factors was collected, including questions on oral contraceptive use. Increased risks of primary liver cancer were found for ever-users (odds ratio [OR] = 1.6, 95 percent confidence interval [CI] = 0.9-2.6), and for long-term (greater than or equal to 10 years) users (OR = 2.0, CI = 0.8-4.8) of oral contraceptives. When the analysis was restricted to subjects whose spouse or parent was the respondent, more pronounced risks were seen for ever-users (OR = 2.7, CI = 1.4-5.3) and long-term users (OR = 4.8, CI = 1.7-14.0). No clear excess risk was found for cancer of the intrahepatic bile ducts. This study, the largest to date, adds to the number of investigations demonstrating an increased risk of primary liver cancer with use, particularly long-term use, of oral contraceptives.

In a 20-year follow-up (1966-86) of 17,633 White males who described tobacco use in a mailed questionnaire sent in 1966, there were 74 deaths from leukemia (including 30 myeloid, 30 lymphatic, and 14 other and unspecified leukemia). Among men who ever smoked cigarettes, increased risks were observed for lymphatic (relative risk [RR] = 2.7), and other and unspecified leukemia (RR = 1.5); risks rose with increasing number of cigarettes smoked, although the dose-response relationship was statistically significant only for total leukemia. Mortality from myeloid leukemia was not elevated, except among those smoking over a pack of cigarettes per day. Results from this cohort support a relationship between cigarette smoking and leukemia. Further studies are needed to elucidate subtype associations with cigarette smoking.

Hypothetical examples are presented that show that bias can result from misclassification of clinical outcomes defined as one of several events, even though the assessment of each component event was masked. In one example, the effect of the overall misclassification is to make a treatment that reduces the risk appear to increase it; in another example, misclassification causes the overall treatment effect to appear stronger than it actually is. These anomalies are due to the fact that the misclassification of the overall outcome can be differential, even though the misclassification of the individual components is nondifferential.

We develop approximate methods to compare the efficiencies and to compute the power of alternative potential designs for sampling from a cohort before beginning to collect exposure data. Our methods require only that the cohort be assembled, meaning that the numbers of individuals Nkj at risk at pairs of event times tk and tj greater than or equal to tk are available. To compute Nkj, one needs to know the entry, follow-up, censoring, and event history, but not the exposure, for each individual. Our methods apply to any "unbiased control sampling design," in which cases are compared to a random sample of noncases at risk at the time of an event. We apply our methods to approximate the efficiencies of the nested case-control design, the case-cohort design, and an augmented case-cohort design, compared to the full cohort design, in an assembled cohort of 17,633 members of an insurance cooperative who were followed for mortality from prostatic cancer. The assumptions underlying the approximation are that exposure is unrelated both to the hazard of an event and to the hazard for censoring. The approximations performed well in simulations when both assumptions held and when the exposure was moderately related to censoring.

A design is proposed for case-control studies in which selection of subjects for full variable ascertainment is based jointly on disease status and on easily obtained "screening" variables that may be related to the disease. Recruitment of subjects follows an independent Bernoulli sampling scheme, with recruitment probabilities set by the investigator in advance. In particular, the sampling can be set up to achieve, on average, frequency matching, provided prior estimates of the disease rates or odds ratios associated with screening variables such as age and sex are available. Alternatively--for example, when studying a rare exposure--one can enrich the sample with certain categories of subject. Following such a design, there are two valid approaches to logistic regression analysis, both of which allow for efficient estimation of effects associated with the screening variables that were allowed to bias the recruitment. The statistical properties of the estimators are compared, both for large samples, based on asymptotics, and for small samples, based on simulations.

A cohort of 17,633 white males age 35 and older responded to a mailed epidemiological questionnaire in 1966 and was followed until 1986 to determine the risk of cancer associated with diet, tobacco use, and other factors. During the 20-year follow-up, 149 fatal prostate cancer cases were identified. Relative risks for prostate cancer were significantly elevated among cigarette smokers (relative risk, 1.8; 95% confidence interval, 1.1-2.9) and users of smokeless tobacco (relative risk, 2.1; 95% confidence interval, 1.1-4.1). No significant associations were found with frequency of consumption of meats, dairy products, fruits, or vegetables. There were no overall significant associations between consumption of vitamin A from animal sources (retinol) and provitamin A from plant sources (carotene) and risk, but positive trends were seen for ages under 75, while inverse associations were found at older ages. Beverage consumption, including drinking coffee and alcohol, was unrelated to risk. Marital status, education, rural/urban status, and farming residence were also unrelated to the risk of fatal prostate cancer. The findings add to limited evidence that tobacco may be a risk factor for prostate cancer, but fail to provide clues to dietary or other risk factors.