Metabolic phenotypes of obesity: frequency, correlates and change over time in a cohort of postmenopausal women.

OBJECTIVE: The possibility that a subset of persons who are obese may be metabolically healthy-referred to as the 'metabolically healthy obese' (MHO) phenotype-has attracted attention recently. However, few studies have followed individuals with MHO or other obesity phenotypes over time to assess change in their metabolic profiles. The aim of the present study was to examine transitions over a 6-year period among different states defined simultaneously by body mass index (BMI) and the presence/absence of the metabolic syndrome (MetS). METHODS: We used repeated measurements available for a subcohort of participants enrolled in the Women's Health Initiative (N=3512) and followed for an average of 6 years to examine the frequency of different metabolic obesity phenotypes at baseline, the 6-year transition probabilities to other states and predictors of the risk of different transitions. Six phenotypes were defined by cross-tabulating BMI (18.5-<25.0, 25.0-<30.0, ⩾30.0 kg m-2) by MetS (yes, no). A continuous-time Markov model was used to estimate 6-year transition probabilities from one state to another. RESULTS: Over the 6 years of follow-up, one-third of women with the healthy obese phenotype transitioned to the metabolically unhealthy obese (MUO) phenotype. Overall, there was a marked tendency toward increased metabolic deterioration with increasing BMI and toward metabolic improvement with lower BMI. Among MHO women, the 6-year probability of becoming MUO was 34%, whereas among unhealthy normal-weight women, the probability of 'regressing' to the metabolically healthy normal-weight phenotype was 52%. CONCLUSIONS: The present study demonstrated substantial change in metabolic obesity phenotypes over a 6-year period. There was a marked tendency toward metabolic deterioration with greater BMI and toward metabolic improvement with lower BMI.

Longitudinal association of measures of adiposity with serum antioxidant concentrations in postmenopausal women.

BACKGROUND/OBJECTIVES: The relationship between obesity and circulating levels of antioxidants is poorly understood. Most studies that have examined the association of adiposity with blood or tissue concentrations of antioxidant micronutrients have been cross-sectional, and few have compared the associations for indices of overall obesity and central obesity. Our aim was to prospectively examine the longitudinal association of body mass index (BMI), waist circumference (WC), waist circumference-height ratio (WCHtR) and waist-hip ratio (WHR) with major serum antioxidants in a population of postmenopausal women. SUBJECTS/METHODS: We used a subsample of participants in the Women's Health Initiative aged 50-79 years at entry with available fasting blood samples and anthropometric measurements obtained at multiple time points over 12.8 years of follow-up (N=2672). Blood samples were used to measure α-carotene, ß-carotene, ß-cryptoxanthin, lutein+zeaxanthin, α-tocopherol, γ-tocopherol and retinol at baseline, and at years 1, 3 and 6. We used mixed-effects linear regression analyses to examine associations between anthropometric measures and serum antioxidants at baseline and over time, controlling for covariates. RESULTS: In longitudinal analyses, carotenoids, and particularly ß-carotene, were strongly and inversely associated with BMI, WC and WCHtR and less so with WHR. α-Tocopherol showed a strong positive association with WHR but not with other anthropometric measures, whereas γ-tocopherol was positively and strongly associated with BMI, WC, WCHtR and less so with WHR. Retinol was positively associated with WHR. The inverse association of several carotenoids with anthropometric measures was stronger in never and former smokers compared with current smokers and in women without the metabolic syndrome. The inverse association of carotenoids with obesity measures may reflect reduced micronutrient concentrations owing to inflammation associated with obesity. CONCLUSIONS: In the present study, the strongest observed associations between anthropometric variables and micronutrients were an inverse association of WC with serum ß-carotene and a positive association of WC with γ-tocopherol.

Association of anthropometric measures and hemostatic factors in postmenopausal women: a longitudinal study.

BACKGROUND AND AIMS: Obesity has been associated with increased levels of hemostatic factors. However, few studies have compared change in different anthropometric measures of adiposity in relation to change in levels of hemostatic factors. Our aim was to examine prospectively the association of change in body mass index (BMI), waist-hip ratio (WHR), waist circumference (WC), and waist circumference-height ratio (WHtR) with change in markers of hemostasis in a population of postmenopausal women. METHODS AND RESULTS: A subsample of women in the Women's Health Initiative (WHI) cohort had fasting blood samples and anthropometric measurements obtained at multiple time points over 12.8 years of follow-up. Of these, we studied the 2593 women who were not in the intervention arm of any WHI clinical trial. Their blood samples were used to measure plasma fibrinogen, factor VII antigen activity, and factor VII concentration at baseline, and at years 1, 3, and 6. We conducted mixed-effects linear regression analyses to examine the longitudinal association between change in anthropometric factors and change in hemostatic factors, adjusting for a wide range of potential confounding factors. In longitudinal analyses using repeated measures, change in BMI, WC, and WHtR were all positively associated with change in all 3 hemostatic factors. Change in anthropometric variables was most strongly associated with change in fibrinogen. CONCLUSIONS: Our results suggest that an increase in adiposity over time is robustly associated with increased levels of hemostatic factors. Registration number of clinical trial: NCT00000611.

US pediatric population-level associations of DXA-measured percentage of body fat with four BMI metrics with cutoffs.

OBJECTIVE: Four body mass index (BMI) metrics--BMI, BMI z-score, BMI percentile and BMI%--are commonly used as proxy measures for children's adiposity. We sought to determine a BMI metric that is most strongly associated with measured percentage of body fat (%BF) in the US pediatric population stratified by sex, age and race/ethnicity, and to determine cutoffs that maximize the association for each BMI metric. SUBJECTS, DESIGN AND METHODS: %BF was measured by dual-energy X-ray absorptiometry among N=6120 US boys and girls aged 8.0-17.9 years old from the National Health and Nutrition Examination Survey 1999-2004. We fit piecewise linear regression models with cutoffs to %BF data using each BMI metric as the predictor stratified by sex, race/ethnicity and age. The slopes were modeled differently before and after the cutoffs which were determined on the basis of grid searches. RESULTS: BMI z-score was in general most strongly associated with %BF for both boys and girls. The associations of the four BMI metrics were lowest for boys aged 12-13.9 years and girls aged 16-17.9 years, and strongest for Mexican-American boys and for non-Hispanic Black girls. Overall, the associations were stronger for girls than for boys. In boys, BMI had the lowest association with %BF (R(2)=0.39) for all ages combined. The fold changes in slopes before and after cutoffs were greatest in general for BMI percentiles regardless of age, sex and race/ethnicity. BMI z-score cutoffs were 0.4 for both boys and girls for all ages combined. Except for BMI, the slopes after the cutoffs were in general greater than those before. CONCLUSIONS: All BMI metrics were strongly associated with %BF when stratified by age and race/ethnicity except that BMI was the least associated with %BF in boys for all ages combined. Overall, BMI z-score was superior for evaluation of %BF, and its cutoff of 0.4 can also serve as a threshold for careful monitoring of weight status.

Optimal scaling of weight and waist circumference to height for maximal association with DXA-measured total body fat mass by sex, age and race/ethnicity.

BACKGROUND: Body mass index (BMI; weight (Wt)/height (Ht) (in kg m(-2)) and waist circumference (WC) are widely used as proxy anthropometric measures for total adiposity. Little is known about what scaling power of 'x' in both Wt(kg)/Ht(m)(x) and WC(m)/Ht(m)(x) is maximally associated with measured total body fat mass (TBFM). Establishing values for x would provide the information needed to create optimum anthropometric surrogate measures of adiposity. OBJECTIVE: To estimate the value of 'x' that renders Wt/Ht(x) and WC/Ht(x) maximally associated with DXA-measured TBFM. SUBJECTS: Participants of the NHANES 1999-2004 surveys, stratified by sex (men, women), race/ethnicity (non-Hispanic whites, non-Hispanic blacks, Mexican-Americans), and age(18-29, 30-49, 50-84 years). METHODS: We apply a grid search by increasing x from 0.0-3.0 by increments of 0.1 to the simple regression models, TBFM=b0+b1*(Wt/Ht(x)) and TBFM=b0+b1*(WC/Ht(x)) to obtain an estimate of x that results in the greatest R(2), taking into account complex survey design features and multiply imputed data. RESULTS: R(2)'s for BMI are 0.86 for men (N=6544) and 0.92 for women (N=6362). The optimal powers x for weight are 1.0 (R(2)=0.90) for men and 0.8 (R(2)=0.96) for women. The optimal power x for WC is 0, that is, no scaling of WC to height, for men (R(2)=0.90) or women (R(2)=0.82). The optimal powers for weight across nine combinations of race/ethnicity and age groups for each sex vary slightly (x=0.8-1.3) whereas the optimal scaling powers for WC are all 0 for both sexes except for non-Hispanic black men aged 18-29y (x=0.1). Although the weight-for-height indices with optimal powers are not independent of height, they yield more accurate TBFM estimates than BMI. CONCLUSION: In reference to TBFM, Wt/Ht and Wt/Ht(0.8) are the optimal weight-for-height indices for men and women, respectively, whereas WC alone, without Ht adjustment, is the optimal WC-for-height index for both sexes. Thus, BMI, an index independent of height, may be less useful when predicting TBFM.

A longitudinal study of serum insulin and glucose levels in relation to colorectal cancer risk among postmenopausal women.

BACKGROUND: It is unclear whether circulating insulin or glucose levels are associated with increased risk of colorectal cancer. Few prospective studies have examined this question, and only one study had repeated measurements. METHODS: We conducted a prospective study of colorectal cancer risk using the subsample of women in the Women's Health Initiative study whose fasting blood samples, collected at baseline and during follow-up, were analysed for insulin and glucose. Cox proportional hazards models were used to assess associations with colorectal cancer risk in both baseline and time-dependent covariates analyses. RESULTS: Among 4902 non-diabetic women with baseline fasting serum insulin and glucose values, 81 incident cases of colorectal cancer were identified over 12 years of follow-up. Baseline glucose levels were positively associated with colorectal cancer and colon cancer risk: multivariable-adjusted hazard ratio (HR) comparing the highest (≥99.5 mg dl(-1)) with the lowest tertile (<89.5 mg dl(-1)): 1.74, 95% confidence interval (CI) 0.97-3.15 and 2.25, 95% CI: 1.12-4.51, respectively. Serum insulin and homeostasis model assessment were not associated with risk. Analyses of repeated measurements supported the baseline results. CONCLUSION: These data suggest that elevated serum glucose levels may be a risk factor for colorectal cancer in postmenopausal women.

Repeated measurements of serum carotenoid, retinol and tocopherol levels in relation to colorectal cancer risk in the Women's Health Initiative.

BACKGROUND/OBJECTIVE: Previous cohort studies examining the association of serum antioxidant levels and risk of colorectal cancer have used a single (baseline) measurement only. In the present study, we assessed the association of serum levels of eight antioxidant nutrients in relation to risk of colorectal cancer, using repeated measurements. SUBJECTS/METHODS: Data on a subsample of women in the Women's Health Initiative with repeated measurements of serum retinol, α-carotene, ß-carotene, ß-cryptoxanthin, lutein+zeaxanthin, lycopene, α-tocopherol and γ-tocopherol during follow-up were included in the analysis. Cox proportional hazards models were used to estimate hazard ratios (HR) and 95% confidence intervals (CIs). RESULTS: Among 5477 women with baseline serum antioxidant values, 88 incident cases of colorectal cancer were identified over a median follow-up time of 12 years. Serum antioxidants measured at baseline generally showed no association with risk of colorectal cancer, although serum ß-carotene at baseline showed a non-significant inverse association with colon cancer alone. Furthermore, using the repeated measurements of ß-carotene, the average of all measurements was inversely associated with risk of both colorectal and colon cancer: HRs for highest vs lowest tertile 0.54, 95% CI 0.31-0.96, and 0.47, 95% CI 0.25-0.88, respectively. No associations were seen with other antioxidant nutrients in the repeated measure analyses. CONCLUSIONS: In this study, baseline levels of antioxidant nutrients were not associated with risk of colorectal or colon cancer; however, using repeated measures, a relatively high serum level of ß-carotene (average of all measurements) was inversely associated with risk of colon and colorectal cancer in postmenopausal women.

Biomarkers of body iron stores and risk of developing type 2 diabetes.

AIM: Iron may contribute to the pathogenesis of type 2 diabetes mellitus (DM) by inducing oxidative stress and interfering with insulin secretion. Elevated ferritin levels are associated with increased DM risk among healthy individuals. However, it is yet unknown if ferritin predicts DM incidence among high-risk individuals with impaired glucose tolerance (IGT). Furthermore, the association between soluble transferrin receptors (sTfR), a novel marker of iron status, and DM risk has not yet been prospectively investigated in these individuals. We conducted this study to evaluate the association between baseline levels of ferritin and sTfR and the risk of developing DM among overweight and obese individuals at high risk of DM. METHODS: This nested case-control study (280 cases and 280 matched controls) was conducted within the placebo arm of the Diabetes Prevention Program, is a clinical trial conducted among overweight/obese individuals with IGT. Ferritin and sTfR levels were measured by immunoturbidimetric assays. Incident DM was ascertained by annual 75-g oral glucose tolerance test and semi-annual fasting glucose. RESULTS: Compared with controls, cases had higher sTfR levels (3.50 +/- 0.07 vs. 3.30 +/- 0.06 mg/l; p = 0.03), but ferritin levels were not statistically different. The multivariable odds ratios (OR) and 95% confidence intervals (95% CI) for DM incidence comparing highest with the lowest quartiles of sTfR was 2.26 (1.37-4.01) (p-trend: 0.008). CONCLUSIONS: Modestly elevated sTfR levels are associated with increased DM risk among overweight and obese individuals with IGT. Future studies should evaluate factors determining sTfR levels and examine if interventions that lower body iron stores reduce DM incidence.

Dietary intake of selected B vitamins in relation to risk of major cancers in women.

Although folic acid has been investigated for its potential to inhibit carcinogenesis, few epidemiologic studies have assessed the effects of intake of thiamin, riboflavin, and niacin, which may reduce cancer risk by acting as cofactors in folate metabolism or by other mechanisms. Using data from a large cohort of Canadian women, we examined the association of dietary intake of these nutrients, as well as intake of folate, methionine, and alcohol, with cancers of the breast, endometrium, ovary, colorectum, and lung ascertained during an average of 16.4 years of follow-up. After exclusions, the following numbers of incident cases were available for analysis: breast, n=2491; endometrium, n=426; ovary, n=264; colorectum, n=617; and lung, n=358. Cox proportional hazard models were used to estimate risk of each cancer with individual nutrients and to explore possible effect modification by combinations of nutrients on cancer risk. Few significant associations of intake of individual B vitamins with the five cancers were observed. Alcohol consumption showed a modest positive association with breast cancer risk but not with risk of the other cancers. There was no evidence of effect modification among the nutrients. This large study provides little support for an association of dietary intake thiamin, riboflavin, niacin, folate, or methionine with five major cancers in women.

Dietary iron and haem iron intake and risk of endometrial cancer: a prospective cohort study.

We used data from a large cohort study of Canadian women to assess the association of meat intake and dietary intake of iron and haem iron with risk of endometrial cancer. Among 34,148 women with an intact uterus at baseline and followed for a mean of 16.4 years, we identified 426 incident endometrial cancer cases. Data from a food frequency questionnaire administered at baseline were used to calculate intake of all meats, red meat, total dietary iron, iron from meat, haem iron, and non-haem iron. Analyses were carried out using Cox proportional hazards models with adjustment for known risk factors and covariates. We found no association of intake of meat or any of the dietary iron-related variables with risk of endometrial cancer.

A cohort study of dietary iron and heme iron intake and risk of colorectal cancer in women.

In a cohort study of 49,654 Canadian women, we assessed the association of colorectal cancer with total iron and heme iron intake, excluding iron supplements. Among women aged 40-59 years, followed for an average of 16.4 years, we identified 617 incident colorectal cancer cases. Data from a food frequency questionnaire administered at baseline were used to calculate red meat intake and intake of total dietary iron, iron from meat, and heme iron. Analyses were carried out for all cases and for the proximal colon, distal colon, and rectum, using Cox proportional hazards models. We found no association of intake of iron, heme iron, or iron from meat with risk of colorectal cancer overall or with any of the subsites, nor was there effect modification by alcohol consumption or hormonal replacement therapy.

A cohort study of reproductive and hormonal factors and renal cell cancer risk in women.

We examined the association of reproductive and hormonal factors with renal cell cancer risk in a cohort study of 89 835 Canadian women. Compared with nulliparous women, parous women were at increased risk (hazard ratio (HR) 1.78, 95% confidence interval (CI) 1.02-3.09), and there was a significant gradient of risk with increasing levels of parity: relative to nulliparous women, women who had > or =5 pregnancies lasting 4 months or more had a 2.4-fold risk (HR=2.41, 95% CI=1.27-4.59, P for trend 0.01). Ever use of oral contraceptives was associated with a modest reduction in risk. No associations were observed for age at first live birth or use of hormone replacement therapy. The present study provides evidence that high parity may be associated with increased risk of renal cell cancer, and that oral contraceptive use may be associated with reduced risk.

Electromagnetic fields and breast cancer on Long Island: a case-control study.

The EMF and Breast Cancer on Long Island Study (EBCLIS) was a case-control study designed to evaluate the possible association between exposure to electromagnetic fields (EMFs) and breast cancer. Eligible women were participants in the population-based Long Island Breast Cancer Study Project, were under 75 years of age at enrollment, were residentially stable, and were identified between August 1, 1996, and June 20, 1997. Of those eligible, 576 cases and 585 controls participated in EBCLIS (87% and 83%, respectively). In-home data collection included various spot and 24-hour EMF measurements, ground-current magnetic field measurements, wire mapping of overhead power lines servicing the home, and an interview. Odds ratios and 95% confidence intervals were based on multivariate logistic regression analyses. All odds ratios were close to 1 and nonsignificant. For the highest quartile of 24-hour EMF measurements, the odds ratio was 0.97 (95% confidence interval (CI): 0.69, 1.37) in the bedroom and 1.09 (95% CI: 0.78, 1.51) in the most lived-in room. For the highest exposure category of ground-current measurements, the odds ratio was 1.13 (95% CI: 0.88, 1.44) in the bedroom and 1.08 (95% CI: 0.85, 1.38) in the most lived-in room. These and other EBCLIS results agree with other recent reports of no association between breast cancer and residential EMF exposures.

Urinary estrogen metabolites and breast cancer: a case-control study.

Preliminary studies suggest that the estrogen metabolite 16 alpha-hydroxyestrone is associated with breast cancer, whereas 2-hydroxyestrone is not. However, epidemiological studies evaluating this relationship and taking established risk factors for breast cancer into account are lacking. The purpose of this study was to examine the association of the ratio of the urinary estrogen metabolites (2-hydroxyestrone and 16 alpha-hydroxyestrone) and of the individual metabolites with breast cancer. A spot urine sample, a brief history, and clinical data were collected from breast cancer cases (n = 42) and from women coming to the hospital for a routine mammogram or attending a free breast cancer screening (n = 64). 2-Hydroxyestrone and 16 alpha-hydroxyestrone were measured by enzyme immunoassay, and the estrogen metabolite ratio (EMR; 2-hydroxyestrone:16 alpha-hydroxyestrone) was computed. Cases and controls were similar in terms of age (mean age of cases, 53.8 +/- 15.1 years, versus 54.2 +/- 10.4 years for controls; P = 0.9) and demographics. Mean EMR was not associated with breast cancer overall (1.67 +/- 0.80 versus 1.72 +/- 0.66; P = 0.7). However, in postmenopausal women, the mean EMR was significantly lower in cases compared to controls (1.41 +/- 0.73 versus 1.81 +/- 0.71; P = 0.05). The multivariate adjusted odds ratios for the intermediate and lowest tertiles of the EMR relative to the highest among postmenopausal women were 9.73 (95% confidence interval, 1.27-74.84) and 32.74 (95% confidence interval, 3.36-319.09), respectively. The test for trend was highly significant (P = 0.003). Analyses of the individual metabolites indicated that 16 alpha-hydroxyestrone was a strong risk factor. The EMR did not show any consistent associations with age, race/ethnicity, age at first birth, parity, body mass index, family history of breast cancer, smoking, or alcohol intake. These data suggest a strong, inverse association of the EMR and a strong positive association of 16 alpha-hydroxyestrone with breast cancer in postmenopausal women. Larger studies are needed to confirm these results and to assess the relationship of the EMR and of the individual metabolites with breast cancer, with attention to menopausal status and clinical factors and with adjustment for known breast cancer risk factors.

Dietary and lifestyle correlates of passive smoking in Hong Kong, Japan, Sweden, and the U.S.A.

From epidemiologic studies in several countries, passive smoking has been associated with increased risk for lung cancer, respiratory diseases, and coronary heart disease. Since the relative risks derived from those studies are weak, i.e. relative risk less than two, we investigated whether poorer diets and less healthy lifestyles might act as confounders and be correlated with having a smoking husband on a cross-cultural basis. Characteristics of never-smoked wives with or without smoking husbands were compared between 530 women from Hong Kong, 13,047 from Japan, 87 from Sweden, and 144 from the U.S. In all four sites, wives with smoking husbands generally ate less healthy diets. They had a tendency to eat more fried food but less fruit than wives with nonsmoking husbands. Other healthy traits, e.g. avoiding obesity, dietary cholesterol and alcohol, or taking vitamins and participating in preventive screening were also less prevalent among wives with smoking husbands. These patterns suggest that never-smoked wives with smoking husbands tend to share the same less healthy dietary traits characteristic of smokers, and to have dietary habits associated with increased risk for lung cancer and heart disease in their societies. These results emphasize the need to take into account the potential confounding effects of diet and lifestyle in studies evaluating the health effects of passive smoking, especially since it is known that the current prevalence rates of smoking among men is indirectly associated with social class and education in affluent urban societies.

Aspects of the epidemiology of lung cancer in smokers and nonsmokers in the United States.

While it is well-established that smoking is the predominant risk factor for lung cancer, it is clear that factors other than smoking and occupational exposure play a role in some lung cancers, and particularly adenocarcinoma. Data from a large, hospital-based case-control study are used to examine the association of smoking-related risk factors (amount smoked, filter status, mentholation, and differences in smoking habits between blacks and whites) and selected factors other than smoking (environmental tobacco smoke, previous primary cancer and radiotherapy, reproductive and endocrine factors, and body mass index) with lung cancer. Although smoking shows a dose-response relationship with all major lung cancer cell types, the strength of the relationship is weaker for adenocarcinoma, suggesting that other risk factors must play an important role for this cell type. In blacks and whites of both sexes, odds ratios for lung cancer increased with increasing cumulative tobacco tar intake and decreased with years since quitting smoking. Use of mentholated cigarettes was associated with no greater risk for lung cancer than that associated with the use of nonmentholated cigarettes. Exposure to environmental tobacco smoke generally showed little relation to lung cancer risk. In particular, exposure of nonsmoking wives to a husband's smoking showed no increase in risk. A history of a reproductive primary cancer and a history of radiotherapy were each associated with a fourfold increase in risk in female nonsmokers. An association of lean body mass with lung cancer was observed in current smokers, ex-smokers, and female never smokers. These results are discussed in the context of existing studies. In conclusion, variation in lung cancer rates between populations may be due to: (1) differences in effective exposure to tobacco smoke carcinogens; (2) differences in factors which modify the effect of tobacco smoke, including differences in host susceptibility and metabolism of carcinogens, or (3) differences in exposure to other independent risk factors for lung cancer.

Increased risk of Helicobacter pylori associated with birth in wartime and post-war Japan.

BACKGROUND: Helicobacter pylori infection is now widely recognized as a cause of stomach cancer. We assessed trends in H. pylori infection in Japan, a population with high rates of gastric malignancy. METHODS: Using an enzyme-linked immunosorbent assay (ELISA), we tested sera collected between 1980 and 1993 from Tokyo University Hospital patients for anti-H. pylori IgG. Patients ranged in age from 0 to 94 years. Helicobacter pylori prevalence was then assessed for age and/or birth cohort effects. RESULTS: Of 1207 sera, 470 (38.9%) were positive for H. pylori IgG. By univariate analysis, both older age and birth in an earlier decade were associated with an increased risk of infection. Age-specific prevalence of H. pylori by birth cohort suggested increases in infection during the decades from 1900 to 1959, and age-specific decreases since 1960. In multivariate analysis, H. pylori infection increased with age and was most prevalent among those born in the 1940s and 1950s. CONCLUSION: Relative to other birth cohorts, people born in the 1940s and 1950s have a higher prevalence of H. pylori. This increased prevalence of infection among those born in wartime Japan likely attests to the impact of compromised living conditions on acquisition of H. pylori, and may portend continued high rates of gastric cancer in forthcoming years.

Relation between exposure to environmental tobacco smoke and lung cancer in lifetime nonsmokers.

To assess the relation between exposure to environmental tobacco smoke throughout life and lung cancer in lifetime nonsmokers, the authors conducted in-person interviews with 41 male and 69 female never-smoking lung cancer cases and 117 male and 187 female never-smoking controls between 1983 and 1990 as part of a hospital-based case-control study of tobacco-related cancers. Cases had newly diagnosed, histologically confirmed primary carcinoma of the lung. Controls were matched to cases on age (+/- 5 years), sex, race, hospital, and year of interview. Subjects were asked about environmental tobacco smoke exposure in childhood, in adulthood at home, in different jobs, and in transportation and social situations. In addition to amount smoked by family members in the subject's presence, subjects were asked to rate the intensity of each exposure, and married subjects were asked whether their spouse smoked in the bedroom. Several independent indicators of exposure to smoking by spouses were strongly correlated, thereby increasing confidence in the classification of exposure status. The reproducibility of environmental tobacco smoke variables was good for qualitative measures (yes/no), in agreement with previous studies. There were few associations of exposure in specific settings with lung cancer. Males whose wives smoked had odds ratio of 1.60 (95% confidence interval (CI) 0.67-3.82) and females whose husbands smoked had an odds ratio of 1.08 (95% CI 0.60-1.94). While this study had limited sample size, the pattern of odds ratios shows little indication of an association of environmental tobacco smoke with lung cancer in nonsmokers.