Comparative effectiveness of treatment of actinic keratosis with topical fluorouracil and imiquimod in the prevention of keratinocyte carcinoma: A cohort study.

BACKGROUND: The effectiveness of 5-fluorouracil compared with that of imiquimod for preventing keratinocyte carcinoma is unknown. OBJECTIVE: To compare the effectiveness of 5-fluorouracil and that of imiquimod in preventing keratinocyte carcinoma in a real-world practice setting. METHODS: We identified 5700 subjects who filled prescriptions for 5-fluorouracil or imiquimod for treatment of actinic keratosis in 2007. An intention-to-treat analysis controlling for potential confounding variables was used to calculate 2- and 5-year cumulative risk differences for subsequent keratinocyte carcinoma overall and in field-treated areas. RESULTS: 5-Fluorouracil was associated with a statistically significant decreased risk of any keratinocyte carcinoma compared with imiquimod (adjusted hazard ratio [aHR], 0.86; 95% confidence interval [CI], 0.76-0.97), but there were no significant differences in risk by tumor subtype (for squamous cell carcinoma: aHR, 0.89; 95% CI, 0.74-1.07; for basal cell carcinoma: aHR, 0.87; 95% CI, 0.74-1.03) or site-specific keratinocyte carcinoma (aHR, 0.96; 95% CI, 0.81-1.14). There were no significant differences in 2- or 5-year cumulative risk of keratinocyte carcinoma among those treated with 5-fluorouracil versus with imiquimod. LIMITATIONS: Generalizability to other practice settings may be limited. CONCLUSIONS: Whereas 5-fluorouracil was more effective in reducing keratinocyte carcinoma risk overall, we found no differences in the short- or long-term risk of subsequent site-specific keratinocyte carcinoma in a real-world practice setting.

A real-world, community-based cohort study comparing the effectiveness of topical fluorouracil versus topical imiquimod for the treatment of actinic keratosis.

BACKGROUND: The most widely used topical agents for the field-based treatment of multiple actinic keratoses (AKs) are 5-fluorouracil and imiquimod, but their comparative effectiveness has not been assessed in a real-world setting. OBJECTIVE: We compared the effectiveness of 5-fluorouracil and imiquimod in reducing risk for subsequent AKs in a large, integrated health care delivery system in northern California. METHODS: In this cohort study, we identified adult health plan members who had an AK diagnosed in 2007 and who subsequently filled a prescription for 5-fluorouracil or imiquimod (N = 5700). We followed subjects for subsequent AKs identified by the International Classification of Diseases codes and estimated the 2-year (short-term) and 5-year (long-term) differences in cumulative risk while controlling for potential confounding by pretreatment variables. RESULTS: 5-Fluorouracil reduced the short-term incidence of subsequent AKs (cumulative risk difference -4.54% [95% confidence interval, -7.91% to -1.17%]), but there was no statistically significant evidence of a long-term decreased risk (cumulative risk difference -1.43% [95% confidence interval, -3.43% to 0.05%]) compared with that with imiquimod. LIMITATIONS: This is a retrospective study with limited ascertainment of all relevant potential confounding variables. CONCLUSION: We found that 5-fluorouracil appeared to be significantly more effective than imiquimod in the short-term, but not long-term, prevention of subsequent AKs.

Risk of cancer in Asian Americans: a Kaiser Permanente cohort study.

PURPOSE: To supplement published cohort data about incident cancer in Asian Americans (Asians) including risk of specific Asian ethnic groups. METHODS: A cohort study in 124,193 persons (13,344 Asians) with baseline examination data in 1978-1985 used Cox proportional hazards models with seven covariates to estimate hazard ratios (HRs) and 95 % confidence intervals (CIs). RESULTS: Through 2012 cancer was diagnosed in 18,687 persons including 1,522 Asians. Compared to Whites, the HR (CIs) for any cancer in Asians was 0.8 (0.7-0.9, p < 0.001). Lower Asian risk was stronger for men (HR = 0.7, p < 0.001) than for women (HR = 0.9, p = 0.003). Lower Asian vs. White risks with p < 0.05 were found for cancers of the upper airway digestive area, hematologic malignancies, melanoma, and cancers of the prostate, bladder, and brain. Melanoma contributed substantially to lower Asian risk, especially in women. HRs for specific Asian groups versus Whites follow: Chinese = 0.9 (p < 0.001), Japanese = 0.9 (p = 0.01), Filipinos = 0.8 (p < 0.001), South Asians = 0.5 (p < 0.001), and Other Asians = 0.7 (p = 0.006). Both South Asian men and women had lower risk than Whites, and South Asians had lower risk than any other racial/ethnic group. CONCLUSIONS: Asians had lower cancer risk than Whites, due to lower risk of several cancer types. Each Asian ethnic group had lower risk than Whites with South Asians at the lowest risk.

Antidepressants and testicular cancer.

PURPOSE: Re-examine association of fluoxetine and paroxetine with risk of testicular cancer noted in drug screening, with 4 years more follow-up and expanded study of these and other antidepressant drugs. METHODS: In the Kaiser Permanente Medical Care Program in Northern California, 906 men with testicular cancer diagnosed August 1996-December 2010 were compared with 38,253 matched controls with race/ethnicity recorded regarding receipt of antidepressant drugs at least 2 years before diagnosis or control index date. Analyses emphasized duration of use and histological subgroups. RESULTS: With control for race/ethnicity and use of other antidepressant drugs, odds ratios (OR) and 95 % confidence intervals (CI) for associations with testicular cancer were as follows: fluoxetine 1.22 (0.88-1.71), paroxetine 1.19 (0.78-1.83), and 1.21 (0.92-1.58) for all serotonin reuptake inhibitors. There was no statistically significant association with risk of all testicular cancers or their histological subtypes for any individual drug or for tricyclics or all antidepressants combined except for citalopram with all testicular cancers 2.55 (1.43-4.52) and those of mixed histology 4.36 (1.50-12.68) and nefazodone with embryonal cancers 9.79 (1.85-51.81). These could readily be chance findings in the context of the many analyses that were performed. Duration of use was not associated with risk of the drugs and drug groups with sufficient numbers of exposed cases for analysis. CONCLUSIONS: We found little evidence to support a testicular carcinogenic effect of fluoxetine, paroxetine, or other antidepressant drugs, but a weakly positive association is not ruled out. The signals in prior screening may have been due to chance and/or uncontrolled confounding.

Moderate alcohol intake and cancer: the role of underreporting.

PURPOSE: There is compelling evidence that heavy alcohol drinking is related to increased risk of several cancer types, but the relationship of light-moderate drinking is less clear. We explored the role of inferred underreporting among light-moderate drinkers on the association between alcohol intake and cancer risk. METHODS: In a cohort of 127,176 persons, we studied risk of any cancer, a composite of five alcohol-associated cancer types, and female breast cancer. Alcohol intake was reported at baseline health examinations, and 14,880 persons were subsequently diagnosed with cancer. Cox proportional hazard models were controlled for seven covariates. Based on other computer-stored information about alcohol habits, we stratified subjects into 18.4 % (23,363) suspected of underreporting, 46.5 % (59,173) not suspected of underreporting, and 35.1 % (44,640) of unsure underreporting status. RESULTS: Persons reporting light-moderate drinking had increased cancer risk in this cohort. For example, the hazard ratios (95 % confidence intervals) for risk of any cancer were 1.10 (1.04-1.17) at <1 drink per day and 1.15 (1.08-1.23) at 1-2 drinks per day. Increased risk of cancer was concentrated in the stratum suspected of underreporting. For example, among persons reporting 1-2 drinks per day risk of any cancer was 1.33 (1.21-1.45) among those suspected of underreporting, 0.98 (0.87-1.09) among those not suspected, and 1.20 (1.10-1.31) among those of unsure status. These disparities were similar for the alcohol-related composite and for breast cancer. CONCLUSIONS: We conclude that the apparent increased risk of cancer among light-moderate drinkers may be substantially due to underreporting of intake.

Risk of cancer in patients exposed to gabapentin in two electronic medical record systems.

PURPOSE: High doses of gabapentin were associated with pancreatic acinar cell tumors in male Wistar rats, but there is little published epidemiological data regarding gabapentin and carcinogenicity. We explored the association between gabapentin and cancer in a US medical care program and followed up nominally significant associations in a UK primary care database. METHODS: In the US Kaiser Permanente Northern California (KPNC) health system, we performed nested case-control analyses of gabapentin and 55 cancer sites and all cancers combined using conditional logistic regression. Up to 10 controls were matched to each case on year of birth, sex, and year of cohort entry. No other covariates were included in models. Only dispensings for gabapentin 2 years or more before index date were considered. Nominally significant associations with an OR > 1.00 and p < 0.05 for three or more dispensings versus no dispensings were followed up by similar nested case-control analyses in the UK General Practice Research Database (GPRD), adjusting for potential indications for gabapentin and risk factors for the specific cancers. RESULTS: The following analyses had OR > 1.00 and p < 0.05 for three or more dispensings of gabapentin versus no dispensing (2-year lag) in KPNC and were also examined in the GPRD: all cancers, breast, lung and bronchus, urinary bladder, kidney/renal pelvis, stomach, anus/anal canal/anorectum, penis, and other nervous system. These cancers were not statistically significantly associated with gabapentin in the GPRD case-control studies (2-year lag). The GPRD and KPNC studies did not identify a statistically significant increased risk of pancreatic cancer with more than two prescriptions of gabapentin in the 2-year lagged analyses. CONCLUSIONS: The epidemiological data in a US cohort with up to 12 years of follow-up and a UK cohort with up to 15 years of follow-up do not support a carcinogenic effect of gabapentin use. However, the confidence intervals for some analyses were wide, and an important effect cannot be confidently excluded.

Association of tea consumption and cutaneous squamous cell carcinoma.

Laboratory and epidemiologic studies suggest a protective effect of tea consumption on risk of cutaneous squamous cell carcinoma (SCC). We designed a case-control study to examine the association between putative protective exposures, including tea consumption, and SCC risk using a large health maintenance organization population. Cases (n=415) were defined as Kaiser Permanente Northern California (KPNC) members with a pathology-verified SCC in 2004 and controls (n=415) were age-, gender-, and race-matched members with no previous history of skin cancer. Tea consumption and SCC risk factors were ascertained by questionnaire. Odds ratios (OR) and 95% confidence intervals (CI) were calculated using conditional logistic regression to estimate the association of SCC with regular use, as well as dose and duration of tea consumption. Risk factor adjusted models included education, smoking, hair and eye color, skin type, family history of skin cancer, and history of freckling, sunburns, sun exposure, and tanning bed use. Adjusted analyses showed no reduction in SCC risk with regular consumption of tea (OR=1.11, 95% CI: 0.81-1.54). Examining duration, dose, and combined duration and dose exposure variables did not alter findings. We found no evidence that tea consumption was associated with cutaneous SCC risk.

Supplement use and risk of cutaneous squamous cell carcinoma.

BACKGROUND: Laboratory and epidemiologic studies suggest that certain dietary supplements may alter risk of cutaneous squamous cell carcinoma (SCC). OBJECTIVE: We sought to examine the association between supplement use and SCC risk. METHODS: Cases (n = 415) were defined as Kaiser Permanente Northern California members with a pathology-verified SCC in 2004 and control subjects (n = 415) were age-, sex-, and race-matched members with no history of skin cancer. Supplement use and SCC risk factors were ascertained by questionnaire. Associations of SCC with use of multivitamins; vitamins A, C, D, and E; and grape seed extract were estimated as odds ratios and 95% confidence intervals using conditional logistic regression. Models were adjusted for SCC risk factors and other supplement use. RESULTS: Grape seed extract users had a significantly decreased risk of cutaneous SCC (adjusted odds ratio 0.26, confidence interval 0.08-0.89, P = .031). Multivitamin use was associated with a borderline significant reduction in SCC risk (adjusted odds ratio 0.71, confidence interval 0.51-1.00, P = .049). Use of vitamins A, C, D, and E was not associated with SCC risk. LIMITATIONS: The data may be prone to recall and selection bias because of the case-control design. No information was obtained on dose or duration of supplement use. CONCLUSIONS: Use of grape seed extract may be associated with a decreased risk of cutaneous SCC. The other supplements included in our study did not reveal clear associations with SCC risk.

Haloperidol and Prostate Cancer Prevention: More Epidemiologic Research Needed.

CONTEXT: The antipsychotic drug haloperidol has antiproliferative and growth-inhibiting properties on prostate cancer cell lines in vitro by binding the sigma 1 protein. Evidence is needed regarding a possible preventive association in men. OBJECTIVE: To examine whether our epidemiologic data support an inverse association of haloperidol use with risk of prostate cancer. DESIGN: These case-control analyses used conditional logistic regression to estimate relative risk by odds ratios (ORs) adjusting for race/ethnicity and aspects of medical care related to detection of prostate cancer. We tested 3 other commonly used antipsychotic drugs, risperidone, quetiapine, and olanzapine, for sigma 1 protein binding and inhibition of clonogenic growth of prostate cancer cells. Use of any of these by men was considered use of a comparator drug. MAIN OUTCOME MEASURES: 1) association of haloperidol with prostate cancer; 2) sigma 1 binding and clonogenic growth. RESULTS: Probably owing to small numbers of haloperidol recipients, evidence of a preventive association was inconsistent, depending on the definition of long-term use. If duration of use was greater than 1 year, the odds ratio (OR) was 0.38 (95% confidence interval (CI) = 0.14-1.01) for haloperidol and 0.80 (95% CI = 0.66-0.98) for the comparator drug; if the duration of use was greater than 2 years, the OR was 0.66 (95% CI = 0.24-1.76) for haloperidol and 0.84 (95% CI = 0.66-1.08) for the comparator drug. Unlike haloperidol, risperidone, quetiapine, and olanzapine did not bind sigma 1 or inhibit clonogenic growth. CONCLUSION: Given the laboratory evidence, our ambiguous epidemiologic findings should encourage more epidemiologic evaluation of haloperidol use and risk of prostate cancer. Finding a negative association could be a scientific advance in prostate cancer prevention but would not be sufficient basis for recommending the prescription of haloperidol for that purpose.

Epidemiologic evaluation of pharmaceuticals with limited evidence of carcinogenicity.

Thorough review by the International Agency for Research on Cancer (IARC) has resulted in classifying many substances, including pharmaceuticals, as probably or possibly carcinogenic to humans, based on experiments on animals or limited data on humans. We evaluated 9 such pharmaceuticals for evidence of carcinogenicity in patients receiving them in a large medical care program with automated pharmacy records and a cancer registry. Nested case-control analyses were performed in a cohort of 6.5 million subscribers with up to 12 years of follow-up, focusing on cancer sites suggested by previous evidence and other sites with odds ratio of at least 1.50, p < 0.01 and some evidence of dose-response. Unmeasured confounding was estimated in sensitivity analyses. We found some supportive evidence for carcinogenicity of griseofulvin, metronidazole and phenytoin and for the known carcinogen, cyclophosphamide, which was added for validation of our data and analyses. Findings for chloramphenicol, iron-dextran complex, phenoxybenzamine and phenobarbital were essentially non-contributory. Confounding by cigarette smoking and prior thyroid disease could account, respectively, for associations of oxazepam with lung cancer and propylthiouracil with thyroid cancer. Although not definitive, these findings should be considered in the evaluations of these pharmaceuticals.

Pharmaceuticals that cause mammary gland tumors in animals: findings in women.

Risk of breast cancer in women was assessed for eight pharmaceuticals that produce mammary tumors in experimental animals, using nested case-control analyses in two cohorts with prescription records in a comprehensive medical care program. The two cohorts were: (1) earlier cohort: 78,118 female members who received prescriptions in 1969-1973, of whom 2,467 developed breast cancer, and (2) later cohort: 3,289,408 female members who received prescriptions in 1994-2006 of whom 24,528 developed breast cancer. Longest follow-up was until June 30, 2006. Ten randomly selected concurrent control women were age-matched to almost every case. Relative risks were estimated by conditional logistic regression. Case ascertainment was lagged by 2 years, or unlagged and subdivided by number of prescriptions received. Some analyses were controlled for hormone use and sensitivity analyses were conducted to estimate the effects of uncontrolled confounding. In the later cohort furosemide, and metronidazole showed statistically significant but very small increases in relative risk (ranging from 1.07 to 1.13). Of these, only furosemide showed increased risk in the earlier cohort: 2-year lag relative risk 1.66 (95% confidence interval 1.23-2.24) or as low as 0.97, assuming uncontrolled positive confounding. Griseofulvin showed significant increases in the later cohort: relative risk for three or more prescriptions 1.48 (1.08-2.03) or as low as 1.23 assuming uncontrolled positive confounding and non-significant increases were noted in the earlier cohort. Our findings are limited by their inconsistency across the two cohorts and our inability to directly control for most established breast cancer risk factors. Although inconclusive, our findings suggest a need for more research on furosemide and griseofulvin.

Screening pharmaceuticals for possible carcinogenic effects: initial positive results for drugs not previously screened.

OBJECTIVE: To screen commonly used prescription drugs for possible carcinogenic effects. METHODS: In a large health care program we identified 105 commonly used drugs, not previously screened. Recipients were followed for up to 12½ years for incident cancer. Nested case-control analyses of 55 cancer sites and all combined included up to ten matched controls per case, with lag of at least 2 years between drug dispensing and cancer. Positive associations entailed a relative risk of 1.50, with p ≤ 0.01 and higher risk for three or more, than for one prescription. Evaluation included further analyses, searches of the literature, and clinical judgment. RESULTS: There were 101 associations of interest for 61 drugs. Sixty-six associations were judged to have involved substantial confounding. We found evidence that of the remaining 35, the following associations may not be due to chance: sulindac with gallbladder cancer and leukemia, hyoscyamine with nonHodgkin lymphoma, nortriptyline with esophageal and hepatic cancer, oxazepam with lung cancer, both fluoxetine and paroxetine with testicular cancer, hydrochlorothiazide with renal and lip cancer, and nifedipine with lip cancer. CONCLUSIONS: These preliminary findings suggest that further studies are indicated regarding sulindac, hyoscyamine, nortriptyline, oxazepam, fluoxetine, paroxetine, hydrochlorothiazide, and nifedipine.

Statin use and risk of basal cell carcinoma.

OBJECTIVE: We examined the association between statin use and basal cell carcinoma (BCC) risk. METHODS: We identified all members of a large integrated health care delivery system with a diagnosis of a histologically proven BCC in 1997. Subsequent BCCs were identified through 2006 from health plan electronic pathology records. Longitudinal exposure to statins and other lipid-lowering agents was determined from automated pharmacy records. We used extended Cox regression to examine the independent association between receipt of statin therapy (ever vs never, cumulative duration) and risk of subsequent BCC. To minimize confounding by indication, we conducted sensitivity analyses in the subset of individuals considered eligible for lipid-lowering therapy based on national guidelines. RESULTS: Among 12,123 members given a diagnosis of BCC who had no prior statin exposure, 6381 developed a subsequent BCC during follow-up. Neither "ever use of statins" (adjusted hazard ratio 1.02, 95% confidence interval: 0.92-1.12) or cumulative duration of statin (adjusted hazard ratio 1.02/year, 95% confidence interval: 0.99-1.11) was associated with subsequent BCC after adjustment for age, sex, and health care use. Risk estimates did not change appreciably when the analysis was limited to the subset of individuals who met eligibility criteria for initiating statin therapy. There was also no significant association between use of non-statin antilipemics and subsequent BCC (adjusted hazard ratio 1.10, 95% confidence interval: 0.76-1.58). LIMITATIONS: No information was available for BCC risk factors, such as sun sensitivity and sun exposure. CONCLUSIONS: Among a large cohort of individuals with BCC, statin therapy was not significantly associated with risk of subsequent BCC.

Gabapentin and Cancer Risk: Updated Findings from Kaiser Permanente Northern California.

CONTEXT: Epidemiologic analyses of gabapentin use and cancer risk in Kaiser Permanente Northern California were previously carried out in a collaborative study and independently evaluated in a UK database. OBJECTIVE: To update these epidemiologic analyses with 7.5 more years of follow-up. DESIGN: Case-control analyses using conditional logistic regression to estimate relative risk by odds ratios using the prior collaboration's criteria for identifying positive drug-cancer associations and our more stringent criteria requiring stronger association, lower p values, and evidence of dose response. New associations were reanalyzed with additional control for limited measures of smoking and alcohol use. MAIN OUTCOME MEASURES: Gabapentin-cancer associations. RESULTS: No previously found associations met our stringent criteria, but cancers of the mouth/pharynx, esophagus, liver, and vagina did. All odds ratios for 3 or more and 8 or more prescriptions were moderately reduced by control for smoking and alcohol. Substantial elevations of risk of mouth/pharynx, liver, and vaginal cancers were associated with only 1 prescription dispensed. Sensitivity analyses aimed at possible confounding and other biases did not change our conclusions but did reveal a markedly increased risk of vaginal cancer in gabapentin users with epilepsy compared with users without. CONCLUSION: The reduced magnitude of relative risk with control for smoking and alcohol use suggests confounding by known risk factors. Biologically implausible elevated risk from just 1 prescription suggests confounding by indication. Either or both of these concerns applies to each of the 4 cancer sites associated with gabapentin use. Updated analyses show little if any evidence for carcinogenic effects of gabapentin.

Helicobacter pylori infection and development of pancreatic cancer.

BACKGROUND: Infection with Helicobacter pylori is an established risk factor for gastric cancer. Results from two studies suggest that it may also be a risk factor for pancreatic cancer. METHODS: We conducted a nested case control study among 128,992 adult subscribers to the Kaiser Permanente Medical Care Program who had been enrolled in a multiphasic health checkup from 1964 to 1969. Serum collected during the checkup was maintained frozen, and subjects were followed for cancer. Cases consisted of 104 randomly selected subjects among 507 who developed pancreatic cancer in the cohort. Controls consisted of 262 pancreatic cancer-free subjects from a pool of 730 controls previously tested for studies conducted on this cohort. Controls were individually matched to cases on age, gender, race, site, and date of multiphasic health checkup. Control sera were compared with cases for antibodies to H. pylori and the CagA protein. The effects of smoking, alcohol consumption, obesity, and years of education were also investigated. RESULTS: Neither H. pylori [odds ratio (OR), 0.85; 95% confidence interval (95% CI), 0.49-1.48] nor its CagA protein (OR, 0.96; 95% CI, 0.48-1.92) was associated with subsequent development of pancreatic cancer. Smoking (OR, 2.09; 95% CI, 1.17-3.74) and greater number of years of education (OR, 2.13; 95% CI, 1.23-3.69) were risk factors for pancreatic cancer, whereas alcohol consumption and obesity were not. CONCLUSION: Our results suggest that H. pylori infection is not associated with development of pancreatic cancer.

The confounded relation of coffee drinking to coronary artery disease.

After decades of conflicting studies, the relation of coffee drinking to coronary artery disease (CAD) risk remains unresolved. Using Cox proportional-hazards models with 5 covariates, 127,212 subjects who supplied baseline data at voluntary health examinations from 1978 to 1985 were studied. Subsequently, 8,357 subjects were hospitalized for CAD. Coffee drinking was unrelated to CAD risk in 58,888 never smokers, but in ex-smokers and current baseline smokers, daily coffee intake was associated with higher CAD risk. This disparity was generally consistent in stratified subgroups. In conclusion, this relation of coffee consumption to increased CAD risk only in smokers could be explained by incomplete control for smoking, by other traits of smokers, or by an adverse biologic interaction of a coffee ingredient with smoking effect on CAD.

Coffee, cirrhosis, and transaminase enzymes.

BACKGROUND: A minority of persons at risk develop liver cirrhosis, but knowledge of risk modulators is sparse. Several reports suggest that coffee drinking is associated with lower cirrhosis risk. METHODS: We studied 125,580 multiethnic members of a comprehensive prepaid health care plan without known liver disease who supplied baseline data at voluntary health examinations from 1978 to 1985. Subsequently, through 2001, 330 of them were diagnosed with liver cirrhosis. Review of medical records confirmed the diagnosis of cirrhosis and ascertained probable etiology. The association of coffee drinking with cirrhosis was estimated by Cox proportional hazards models with 7 covariates. We also did a cross-sectional analysis of baseline aspartate aminotransferase and alanine aminotransferase levels, studied by logistic regression. RESULTS: In the cohort study, relative risks of alcoholic cirrhosis (199 subjects) for coffee drinking (vs none) were less than 1 cup per day, 0.7 (95% confidence interval [CI], 0.4-1.1); 1 to 3 cups, 0.6 (95% CI, 0.4-0.8; P<.001); and 4 or more cups, 0.2 (95% CI, 0.1-0.4; P<.001). For 131 subjects with nonalcoholic cirrhosis, relative risks were less than 1 cup, 1.2 (95% CI, 0.6-2.2); 1 to 3 cups, 1.3 (95% CI, 0.8-2.1); and 4 or more cups, 0.7 (95% CI, 0.4-1.3). These relative risks for coffee drinking were consistent in subsets. Tea drinking was unrelated to alcoholic or nonalcoholic cirrhosis. In the cross-sectional analyses, coffee drinking was related to lower prevalence of high aspartate aminotransferase and alanine aminotransferase levels; for example, the odds ratio of 4 or more cups per day (vs none) for a high aspartate aminotransferase level was 0.5 (95% CI, 0.4-0.6; P<.001) and for a high alanine aminotransferase level, 0.6 (95% CI, 0.6-0.7; P<.001), with stronger inverse relations in those who drink large quantities of alcohol. CONCLUSION: These data support the hypothesis that there is an ingredient in coffee that protects against cirrhosis, especially alcoholic cirrhosis.