Mentholated cigarettes and smoking habits in whites and blacks.

OBJECTIVE: To determine if cigarette mentholation is associated with the frequency of smoking and with quitting, and whether mentholation explains racial differences in these two smoking behaviours. DESIGN: Cross sectional analysis of case-control data on smoking and lung cancer. SUBJECTS: Limited to 19 545 current and former cigarette smokers. MAIN OUTCOME MEASURES: Smoking > 20 cigarettes per day (cpd) versus < or = 20 cpd, and continued smoking versus quit smoking. RESULTS: Among blacks, the prevalence odds ratio (POR) of heavy smoking (> or = 21 cpd) associated with mentholated cigarettes versus non-mentholated cigarettes was 0.7 (95% confidence interval (CI) 0.5 to 0.9) in current smokers and 0.6 (95% CI 0.4 to 0.9) in former smokers. Among whites, the corresponding POR were 0.9 (95% CI 0.8 to 1.0) and 0.9 (95% CI 0.8 to 1.0). Blacks were less likely to have been heavy smokers than whites, but the difference was unrelated to cigarette mentholation. The POR of continued smoking versus quitting, associated with mentholated cigarettes was 1.1 (95% CI 1.0 to 1.2) for both blacks and whites. CONCLUSION: Smoking > 20 cpd was independently associated with white race. Among blacks, smoking < or = 20 cpd was independently associated with mentholated cigarettes. The risk of quitting was not associated with cigarette menthol flavour.

Handheld cellular telephones and risk of acoustic neuroma.

The hypothesis that intracranial energy deposition from handheld cellular telephones causes acoustic neuroma was tested in an epidemiologic study of 90 patients and 86 control subjects. The relative risk was 0.9 (p = 0.07) and did not vary significantly by the frequency, duration, and lifetime hours of use. In patients who used cellular telephones, the tumor occurred more often on the contralateral than ipsilateral side of the head. Further efforts should focus on potentially longer induction periods.

Smoking and lung cancer risk in American and Japanese men: an international case-control study.

Rates of lung cancer in American men have greatly exceeded those in Japanese men for several decades despite the higher smoking prevalence in Japanese men. It is not known whether the relative risk of lung cancer associated with cigarette smoking is lower in Japanese men than American men and whether these risks vary by the amount and duration of smoking. To estimate smoking-specific relative risks for lung cancer in men, a multicentric case-control study was carried out in New York City, Washington, DC, and Nagoya, Japan from 1992 to 1998. A total of 371 cases and 373 age-matched controls were interviewed in United States hospitals and 410 cases and 252 hospital controls in Japanese hospitals; 411 Japanese age-matched healthy controls were also randomly selected from electoral rolls. The odds ratio (OR) for lung cancer in current United States smokers relative to nonsmokers was 40.4 [95% confidence interval (CI) = 21.8-79.6], which was >10 times higher than the OR of 3.5 for current smokers in Japanese relative to hospital controls (95% CI = 1.6-7.5) and six times higher than in Japanese relative to community controls (OR = 6.3; 95% CI = 3.7-10.9). There were no substantial differences in the mean number of years of smoking or average daily number of cigarettes smoked between United States and Japanese cases or between United States and Japanese controls, but American cases began smoking on average 2.5 years earlier than Japanese cases. The risk of lung cancer associated with cigarette smoking was substantially higher in United States than in Japanese males, consistent with population-based statistics on smoking prevalence and lung cancer incidence. Possible explanations for this difference in risk include a more toxic cigarette formulation of American manufactured cigarettes as evidenced by higher concentrations of tobacco-specific nitrosamines in both tobacco and mainstream smoke, the much wider use of activated charcoal in the filters of Japanese than in American cigarettes, as well as documented differences in genetic susceptibility and lifestyle factors other than smoking.

Cancer mortality in the United States and Germany.

PURPOSE: The USA and Germany are currently two of the world's leading industrial nations with comparable standards of living and considerable similarities in lifestyle. Fifty years ago, i.e., in the years following the Second World War, the living conditions in the two countries were completely different. If it is true that the major part of cancer occurrence is lifestyle-related, we should see corresponding discrepancies and assimilations on the level of cancer occurrence. METHODS: As an exercise in descriptive epidemiology, we compare the time trends in German and US cancer mortality in order to examine whether they parallel indeed the differences and changes in lifestyle factors of the two countries. RESULTS: Overall, we found the cancer mortality of the two countries converging to rather similar rates. However, in detail, the data indicate various inconsistencies between the patterns of lifestyle factors and cancer mortality in the two countries: similar lung cancer rates, despite rather different patterns of cigarette consumption, or decreasing rectal cancer mortality, despite increasing prevalence of risk factors, are examples. CONCLUSIONS: Promising changes with regard to relevant risk factors indicate that the recent decline of cancer mortality in both countries will continue. Nevertheless, vigorous action towards primary prevention in Germany and more effective screening programs in both countries appear recommendable.

Handheld cellular telephone use and risk of brain cancer.

CONTEXT: A relative paucity of data exist on the possible health effects of using cellular telephones. OBJECTIVE: To test the hypothesis that using handheld cellular telephones is related to the risk of primary brain cancer. DESIGN AND SETTING: Case-control study conducted in 5 US academic medical centers between 1994 and 1998 using a structured questionnaire. PATIENTS: A total of 469 men and women aged 18 to 80 years with primary brain cancer and 422 matched controls without brain cancer. MAIN OUTCOME MEASURE: Risk of brain cancer compared by use of handheld cellular telephones, in hours per month and years of use. RESULTS: The median monthly hours of use were 2.5 for cases and 2.2 for controls. Compared with patients who never used handheld cellular telephones, the multivariate odds ratio (OR) associated with regular past or current use was 0.85 (95% confidence interval [CI], 0.6-1.2). The OR for infrequent users (<0. 72 h/mo) was 1.0 (95% CI, 0.5-2.0) and for frequent users (>10.1 h/mo) was 0.7 (95% CI, 0.3-1.4). The mean duration of use was 2.8 years for cases and 2.7 years for controls; no association with brain cancer was observed according to duration of use (P =.54). In cases, cerebral tumors occurred more frequently on the same side of the head where cellular telephones had been used (26 vs 15 cases; P =.06), but in the cases with temporal lobe cancer a greater proportion of tumors occurred in the contralateral than ipsilateral side (9 vs 5 cases; P =.33). The OR was less than 1.0 for all histologic categories of brain cancer except for uncommon neuroepitheliomatous cancers (OR, 2.1; 95% CI, 0.9-4.7). CONCLUSIONS: Our data suggest that use of handheld cellular telephones is not associated with risk of brain cancer, but further studies are needed to account for longer induction periods, especially for slow-growing tumors with neuronal features.

Breast cancer risk in relation to adipose concentrations of organochlorine pesticides and polychlorinated biphenyls in Long Island, New York.

To assess a possible etiological role of organochlorine compounds in breast cancer development on Long Island, a high-risk region of New York State, concentrations of organochlorine pesticides and polychlorinated biphenyls (PCBs) were measured in the adipose tissue of 232 women with breast cancer and 323 hospital controls admitted to surgery for benign breast disease or non-breast-related conditions. Seven pesticide residues and 14 PCB congeners were assayed via a supercritical fluid extraction method followed by gas chromatography with electron capture detection. After adjustment for age and body mass index, which were strongly correlated with organochlorine levels, adipose concentrations of 1,1-dichloro-2,2-di(4-chlorophenyl)ethylene, total pesticides, and total polychlorinated biphenyls (PCBs) did not differ significantly between cases and controls. The relative abundance of individual pesticide species and PCB congeners was similar in cases and controls. Odds ratios adjusted for age, BMI, hospital, and race gave no evidence of a dose-response for 1,1-dichloro-2,2-di(4-chlorophenyl)ethylene, total pesticides, or total PCBs, whether stratified by estrogen receptor status or not. Breast cancer risk among Long Island residents was not elevated compared with residents of the adjacent New York City borough of Queens. We did not confirm a previously reported association between breast cancer risk and levels of PCB congener 118 (2,3',4,4',5-pentachlorobiphenyl), nor did we observe an association with the most abundant congener 153 (2,2',4,4',5,5'-hexachlorobiphenyl), a strong inducer of phase I enzymes that was reported recently to have estrogenic properties. Only PCB congener 183 (2,2',3,4,4',5',6-heptachlorobiphenyl), which is also an inducer, was significantly associated with risk, with an adjusted odds ratio of 2.0 (95% confidence interval, 1.2-3.4) in women with adipose levels >5.67 ng/g; the biological importance of this observation is unclear without confirmation in additional studies. Although neither the present nor other studies have provided convincing evidence of an association between body burden of 1,1,1-trichloro-2,2-bis(4-chlorophenyl)ethane and PCBs with cancer of the breast, these compounds are rated as "possible" and "probable" human carcinogens, respectively, by the International Agency for Research on Cancer. Investigations of associations with cancer at other sites should be carried out.

Comparison of GSTM polymorphisms and risk for oral cancer between African-Americans and Caucasians.

Two members of the mu class of glutathione S-transferase (GST) genes, GSTM1 and GSTM3, have polymorphic alleles which have been associated with altered levels of GST mu protein expression and may be linked to increased risk for several tobacco-related cancers. Oral cancer is a tobacco-related disease that affects African-American men at a significantly higher incidence than Caucasian men. To examine the potential role of GSTM polymorphisms in risk for oral cancer in African-Americans and Caucasians, the prevalences of the GSTM1 null and GSTM3 intron 6 polymorphisms were examined in 63 African-American and 101 Caucasian patients with histologically confirmed primary oral cancer, as well as in 133 African-American and 213 Caucasian matched control subjects. In African-Americans, the odds ratio for oral cancer associated with the GSTM1 (0/0) genotype was 3.1 [95% confidence interval (CI) = 1.1-8.5], with the association between the GSTM1 (0/0) genotype and oral cancer risk strongest in heavy smokers [i.e. > 24 pack-years; odds ratio (OR) = 5.4, 95% CI = 1.2-24]. Using the potentially most protective GSTM1 [+]/GSTM3 (B/B) genotype as the reference group, increased risk for oral cancer was observed in African-Americans with the GSTM1 [+]/GSTM3 [(A/A) + (A/B)] (OR = 2.2, 95% CI = 0.82-6.0), GSTM1 (0/0)/GSTM3 (B/B) (OR = 4.3, 95% CI = 1.1-16), and GSTM1 (0/0)/GSTM3 [(A/A) + (A/B)] (OR = 6.6, 95% CI = 1.2-38) genotypes (P < 0.01, trend test). No significant associations were observed between GSTM genotype and oral cancer risk in Caucasians. These results suggest that the GSTM1 null and GSTM3 intron 6 polymorphisms play an important role in risk for oral cancer among African-Americans and implicates the mu class of GSTs as important tobacco carcinogen detoxifying enzymes in this population.

Vegan diet-based lifestyle program rapidly lowers homocysteine levels.

BACKGROUND: Plasma homocysteine levels have been directly associated with cardiac disease risk. Current research raises concerns as to whether comprehensive lifestyle approaches including a plant-based diet may interact with other known modulators of homocysteine levels. METHODS: We report our observations of homocysteine levels in 40 self-selected subjects who participated in a vegan diet-based lifestyle program. Each subject attended a residential lifestyle change program at the Lifestyle Center of America in Sulphur, Oklahoma and had fasting plasma total homocysteine measured on enrollment and then after 1 week of lifestyle intervention. The intervention included a vegan diet, moderate physical exercise, stress management and spirituality enhancement sessions, group support, and exclusion of tobacco, alcohol, and caffeine. B vitamin supplements known to reduce blood homocysteine levels were not provided. RESULTS: Subjects' mean homocysteine levels fell 13%: from 8.66 micromol/L (SD 2.7 micromol/L) to 7.53 micromol/L (SD 2.12 micromol/L; P < 0.0001). Subgroup analysis showed that homocysteine decreased across a range of demographic and diagnostic categories. Conclusions. Our results suggest that broad-based lifestyle interventions favorably impact homocysteine levels. Furthermore, analysis of Lifestyle Center of America program components suggests that other factors in addition to B vitamin intake may be involved in the observed homocysteine lowering.

Relative abundance of organochlorine pesticides and polychlorinated biphenyls in adipose tissue and serum of women in Long Island, New York.

Some organochlorine pesticides (OCPs) and PCBs are under investigation as possible risk factors for breast cancer because of their estrogenic properties and widespread presence in the environment. It is important to know whether adipose tissue used by some investigators and serum assays used by others can provide comparable information on body burden. Concentrations of seven OCPs or their breakdown products as well as 14 PCB congeners were measured in the adipose tissue and serum of 293 women enrolled as controls in a case-control study of environmental factors for breast cancer in Long Island, New York, a high-risk region. Adipose OCP/PCB levels were measured using a supercritical fluid extraction method developed by the authors. 1,1-Dichloro-2,2-di(4-chlorophenyl)ethylene (p,p'-DDE) was detected in all adipose and serum samples; two chlordane derivatives, beta-hexachlorocyclohexane (a lindane isomer) and hexachlorobenzene, were detected in at least 92% of adipose samples. The di-ortho hexachlorinated PCB congeners 2,4,5,2',4',5'-hexachlorobiphenyl and 2,3,4,2',4',5'-hexachlorobiphenyl were detected in all adipose and over 98% of serum samples. 1,1-Dichloro-2,2-di(4-chlorophenyl)ethylene comprised 77% of total pesticide residues in adipose and 71% in serum. 2,4,5,2',4',5'-Hexachlorobiphenyl comprised 24% of adipose and 21% of serum PCBs. The relative concentration patterns of the 14 PCB congeners were similar to those reported in other human studies and were also typical of patterns reported in environmental samples from various biota, including mammals and birds, but differed substantially from patterns reported in occupationally exposed workers. All adipose-serum correlations for pesticides and most PCBs were statistically significant. Either serum or adipose OCP/PCB levels of a variety of environmental organochlorine compounds may serve as useful biomarkers of body burden.

A case/control study of risk factors for major salivary gland cancer.

BACKGROUND: Salivary gland cancer (SGC) is a rare disease with a largely unknown origin. Because cancer of the tongue and mouth floor is caused primarily by smoking and alcohol consumption, we investigated the role of tobacco, alcohol, and other possible risk factors in the development of SGC in a hospital-based study. METHODS: Interviews were obtained from 128 patients with newly diagnosed histologically confirmed SGC and from 114 age- and gender-matched controls by using a structured questionnaire. All patients were interviewed at bedside by a trained interviewer. RESULTS: No differences in levels of education were found between the two groups. Cigarette smoking and alcohol consumption did not independently or jointly increase the risk of SGC. Chewing tobacco and snuff use were also unrelated factors. The odds ratio for low body mass in men was 0.46 (p = 0.05). There was no relation with body mass in women. An examination of employment history and job-related exposures revealed no occupational risk factors. CONCLUSION: These findings show that smoking, alcohol consumption, and most occupational exposures are unrelated to SGC.

Lung cancer risk and workplace exposures in black men and women.

There are little data on workplace exposures and lung cancer risk in blacks. An ongoing case-control study of lung cancer that included 550 black men and women with lung cancer and 386 age-matched controls was examined by reported occupational exposures and job titles. In men, significant associations were observed with reported exposure to asbestos [odds ratio (OR), 1.8; 95% confidence intervals (CI) 1.03-3.1] and coal dust (OR, 2.8; 95% CI 1.1-7.0). Elevated but nonsignificant risks of 1.4 or more were detected for the following occupations: police/security guards, farmers/farm workers, laborers, and motor-vehicle drivers. In women, nonsignificant increased risks were found with reported exposure to paint (OR, 1.8) and gas fumes (OR, 4.9). Women employed as farmers/farm workers and building maintenance workers had elevated but nonsignificant risks.

Differences in the urinary metabolites of the tobacco-specific lung carcinogen 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone in black and white smokers.

Incidence and mortality rates for lung cancer in the United States are significantly greater in blacks than in whites. This disparity cannot be explained by differences in smoking behavior. We hypothesize that the observed racial differences in risk may be due to differences in the metabolic activation or detoxification of the tobacco-specific lung carcinogen 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK). To test this, different biomarkers of NNK exposure and metabolism, including the urinary metabolite 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL) and the presumed detoxification product [4-(methylnitrosamino)-1-(3-pyridyl)but-1-yl]-beta-O-D-glucosiduronic acid (NNAL-Gluc), were examined along with questionnaire data on lifestyle habits and diet in a metabolic epidemiological study of 34 black and 27 white healthy smokers. Results demonstrated that urinary NNAL-Gluc:NNAL ratios, a likely indicator of NNAL glucuronidation and detoxification, were significantly greater in whites than in blacks (P < 0.02). In addition, two phenotypes were apparent by probit analysis representing poor (ratio < 6) and extensive (ratio > or = 6) glucuronidation groups. The proportion of blacks falling into the former, potentially high-risk group was significantly greater than that of whites (P < 0.05). The absolute levels of urinary NNAL, NNAL-Gluc, and cotinine were also greater in blacks than in whites when adjusted for the number of cigarettes smoked. None of the observed racial differences could be explained by dissimilarities in exposure or other sociodemographic or dietary factors. Also, it is unlikely that the dissimilarities are due to racial differences in preference for mentholated cigarettes, because chronic administration of menthol to NNK-treated rats did not result in either increases in urinary total NNAL or decreases in NNAL-Gluc:NNAL ratios. Altogether, these results suggest that racial differences in NNAL glucuronidation, a putative detoxification pathway for NNK, may explain in part the observed differences in cancer risk.

CYP1A1 and GSTM1 polymorphisms and oral cancer risk.

The importance of both the CYP1A1 exon 7 (ile:val) and GSTM1 (0/0) polymorphisms in oral cancer susceptibility was assessed by examining polymorphic prevalences in 135 patients with oral cancer and 135 noncancer controls frequency-matched by age at diagnosis (+/- 5 years), race, sex, and institute of patient recruitment. The prevalence of the GSTM1 (0/0) genotype was approximately 51% in both cases and controls. The prevalence of the CYP1A1 (ile:val) polymorphism [including both the (ile/val) and (val/val) genotypes] was significantly higher in cases as compared to controls (17.6% versus 7.6%, respectively; crude odds ratio, 2.6; confidence interval, 1.2-5.7). No association was observed between polymorphic prevalence and levels of smoking or alcohol consumption in cases. These results suggest that the GSTM1 null genotype is not associated with oral cancer risk. These results also suggest that individuals with the CYP1A1 exon 7 ile:val polymorphism are at increased risk for oral cancer, and that this risk may not be influenced by differences in exposure to tobacco smoke.

Risk of squamous cell carcinoma and adenocarcinoma of the lung in relation to lifetime filter cigarette smoking.

BACKGROUND: Over the past few decades, the incidence of adenocarcinoma (AC) of the lung increased much more rapidly than that of squamous cell carcinoma (SCC) in men and women. During this time period, filter cigarettes with substantially reduced "tar" and nicotine yields in the smoke came to dominate the market. METHODS: The risk of SCC and AC in lifelong smokers of filter cigarettes relative to lifelong nonfilter cigarette smokers was assessed in a case-control study performed between 1977 and 1995 with 2292 lung carcinoma patients and 1343 hospital controls who were current smokers. RESULTS: Odds ratios (OR) for SCC in male and female subjects who had smoked filter cigarettes exclusively during their lives were slightly reduced relative to lifetime nonfilter cigarette smokers in men (OR = 0.8; 95% confidence interval [CI], 0.5-1.2), and significantly reduced in women (OR = 0.4; 95% CI, 0.2-0.8). No reduction in risk was observed for AC of the lung in men or women. CONCLUSIONS: Evidence that the increasing predominance of AC over SCC may be due in part to the reduced risk of SCC (but not AC) associated with lifelong filter cigarette smoking is strongest in women; for men, further studies that include larger numbers of lifetime filter smokers are needed to confirm this finding. A lack of protection against AC from low yield filter cigarettes may result from smokers' "compensating" with deeper and more frequent inhalation, thereby increasing delivery of carcinogens to the peripheral lung. The smoke of modern cigarettes also contains higher concentrations of nitrosamines that primarily produce AC.

Cigarette smoking and large cell carcinoma of the lung.

Large cell carcinoma is the fourth most common histological type of lung cancer in the United States. Cigarette smoking causes large cell lung cancer, but it is uncertain whether the effect varies with the amount and duration of smoking. This uncertainty stems from ambiguity in the histopathological classification of large cell cancer, especially before 1971, and the relatively infrequent occurrence of large cell cancer in epidemiological studies. The present case-control investigation demonstrates that the risk of large cell cancer increases with both the frequency and number of years of cigarette smoking. The odds ratio associated with smoking two or more packs/day was 37.0 (95% confidence interval, 16.4-83.2) in men and 72.9 (35.4-150.2) in women. It is concluded that cigarette smoking is the predominant cause of large cell lung cancer.

Impact of filter cigarette smoking on lung cancer histology.

BACKGROUND: The rates of lung adenocarcinoma cancer have risen more rapidly than the rates of lung squamous cell cancer over the past 2 decades. METHODS: A case-control study was carried out to assess the impact of long-term filter cigarette smoking on the risk of squamous cell carcinoma (SCC) and adenocarcinoma (AC) of the lung. RESULTS: Odds ratios for SCC among subjects who had smoked only filter cigarettes were reduced relative to lifetime nonfilter cigarette smokers by 30% for men and by 60% for women, but no risk reduction was observed for AC of the lung. CONCLUSION: The predominance of AC over SCC may be due in part to the fact that smokers of very low yield cigarettes tend to compensate for the lower nicotine levels by inhaling more deeply and frequently, leading to greater exposure of the peripheral lung to the carcinogens in tobacco smoke, and in part to the increased concentration of nitrosamines that preferentially produce AC in laboratory animals.

Breast cancer: weighing the evidence for a promoting role of dietary fat.

It has been hypothesized that a high-fat diet promotes the development of postmenopausal breast cancer. This contention is supported by data showing high international correlations between fat intake and breast cancer rates, modest positive associations with a high-fat diet in case-control studies, and animal model studies that have consistently demonstrated that dietary fat influences mammary cancer development at several stages in the carcinogenic process. A number of plausible biologic mechanisms have been suggested that may explain such promotional effects. In contrast, dietary fat intake is unrelated to the risk of breast cancer in cohort studies. The conflicting findings from cohort studies have created uncertainty regarding nutritional recommendations and breast cancer prevention. After reviewing key scientific findings that are relevant to this issue, the following conclusion is drawn: In the absence of data from dietary intervention trials, the weight of available evidence suggests that the type and amount of fat in the diet is related to postmenopausal breast cancer and that the inability to detect associations within populations (cohort studies) is because of measurement error and the relative homogeneity of diets measured. It is expected that the results from intervention trials will clarify this issue.

The epidemiology of left-handedness in a hospital population.

PURPOSE: We evaluated the association between left-handedness (LH) and age, education, cigarette smoking, alcohol consumption, and disease status in a case-control study of 8801 hospitalized patients with cancer and those with other conditions. METHODS: Subjects were interviewed in person using a structured questionnaire that contained detailed sections of lifestyle behaviors. RESULTS: The overall prevalences of LH were 7.6% among men and 6.5% among women. Among both sexes LH declined with increasing age (P < 0.05). After adjustment for age, the following associations were observed. Men had a higher risk of LH than women. The prevalence of LH was lower in ever-married subjects compared with never-married subjects (odds ratio [OR] for men, 0.7; 95% confidence intervals [CI], 0.5-0.9; for women, OR, 0.5; 95% CI, 0.3-0.9). Among men, the prevalence of LH was not associated with race, years of education, smoking status, or levels of alcohol consumption. The risk of LH was elevated in men diagnosed with fractures as compared with all other male patients (OR, 2.4; 95% CI, 1.3-6.7). Among women, LH was not associated with race, smoking, or hormonal and reproductive factors, but LH was more common among female high-school and college graduates and among self-reported alcoholics. The odds ratio of LH was significantly lower in women with breast cancer (OR, 0.3; 95% CI, 0.1-0.7). CONCLUSIONS: The increased risk of serious injuries in LH is not a result of higher alcohol use. Handedness might be an important factor in the safe use of industrial equipment.

Smoking and pancreatic cancer in men and women.

Most studies of smoking and pancreatic cancer have used male subjects or combined men and women together in statistical analyses. There is little information on the relative risk of smoking and pancreatic cancer in women. Because of the high case-fatality rate, many of these studies were also based on information gathered from proxy respondents, in which smoking habits may not be recalled with certainty. A hospital-based study of 484 male and female patients with pancreatic cancer and 954 control subjects was conducted based on direct interviews of incident cases. Compared to never smokers, the odds ratio (OR) for current cigarette smokers was 1.6 [95% confidence interval (CI), 1.1-2.4] for men and 2.3 (95% CI, 1.4-3.5) for women. In women, but not in men, there was a trend in the ORs with years of daily cigarette consumption (P < 0.01). Filter cigarettes offered no protective advantage compared to nonfilter cigarettes. Among men, the OR was 2.1 (95% CI, 1.2-3.8) for pipe/ cigar smokers and 3.6 (95% CI, 1.0-12.8) for tobacco chewers. Tobacco smoke causes pancreatic cancer when inhaled into the lungs. Tobacco juice may also cause pancreatic cancer when ingested or absorbed through the oral cavity. These data suggest that smoking is a cause of pancreatic cancer in women and that the risks for female smokers are comparable to male smokers. Nevertheless, the causes of most pancreatic cancers are unknown.