Search

Background In the United States, lung cancer death rates have been declining for decades, primarily as a result of pronounced decreases in cigarette smoking. It is unclear, however, whether there have been similar declines in mortality rates of lung cancer unrelated to smoking. We estimated trends in US lung cancer death rates attributable and not attributable to smoking from 1991 to 2018. Methods The study included 30- to 79-year-olds in the National Health Interview Survey who were linked to the National Death Index, 1991-2014. Adjusted hazard ratios for smoking status and lung cancer death were estimated, and age-specific population attributable fractions were calculated. Annual population attributable fractions were multiplied by annual US national lung cancer mortality, partitioning rates into smoking-attributable and smoking-unrelated lung cancer deaths. All statistical tests were 2-sided. Results During 1991-2018, the proportion of never smokers increased among both men (35.1%-54.6%) and women (54.0%-65.4%). Compared with those who had ever smoked, those who had never smoked had 86% lower risk (hazard ratio = 0.14; 95% confidence interval [CI] = 0.12 to 0.16) of lung cancer death. The fraction of lung cancer deaths attributable to smoking decreased from 81.4% (95% CI = 78.9 to 81.4) to 74.7% (95% CI = 78.1 to 71.4). Smoking-attributable lung cancer death rates declined 2.7% per year (95% CI = ‒2.9% to ‒2.5%) and smoking-unrelated lung cancer death rates declined 1.8% per year (95% CI = ‒2.0% to ‒1.5%); these declines have accelerated in recent years. Conclusions An increasing proportion of lung cancer deaths are unrelated to smoking based on declines in smoking prevalence. Smoking-unrelated lung cancer death rates have declined, however, perhaps because of decreases in secondhand smoke and air pollution exposure as well as treatment improvements. [ABSTRACT FROM AUTHOR]

Background Industrial facilities are not located uniformly across communities in the United States, but how the burden of exposure to carcinogenic air emissions may vary across population characteristics is unclear. We evaluated differences in carcinogenic industrial pollution among major sociodemographic groups in the United States and Puerto Rico. Methods We evaluated cross-sectional associations of population characteristics including race and ethnicity, educational attainment, and poverty at the census tract level with point-source industrial emissions of 21 known human carcinogens using regulatory data from the US Environmental Protection Agency. Odds ratios and 95% confidence intervals comparing the highest emissions (tertile or quintile) to the referent group (zero emissions [ie, nonexposed]) for all sociodemographic characteristics were estimated using multinomial, population density–adjusted logistic regression models. Results In 2018, approximately 7.4 million people lived in census tracts with nearly 12 million pounds of carcinogenic air releases. The odds of tracts having the greatest burden of benzene, 1,3-butadiene, ethylene oxide, formaldehyde, trichloroethylene, and nickel emissions compared with nonexposed were 10%-20% higher for African American populations, whereas White populations were up to 18% less likely to live in tracts with the highest emissions. Among Hispanic and Latino populations, odds were 16%-21% higher for benzene, 1,3-butadiene, and ethylene oxide. Populations experiencing poverty or with less than high school education were associated with up to 51% higher burden, irrespective of race and ethnicity. Conclusions Carcinogenic industrial emissions disproportionately impact African American and Hispanic and Latino populations and people with limited education or experiencing poverty thus representing a source of pollution that may contribute to observed cancer disparities. [ABSTRACT FROM AUTHOR]

Representative risk estimation is fundamental to clinical decision-making. However, risks are often estimated from non-representative epidemiologic studies, which usually under-represent minorities. Model-based methods use population registries to improve external validity of risk estimation but assume hazard ratios are generalisable from samples to the target finite population. 'Pseudoweighting' methods improve representativeness of studies by using an external probability-based survey as the reference, but the resulting estimators can be biased due to propensity model misspecification and inefficient due to highly variable pseudoweights or small sample sizes of minorities in the cohort and/or survey. We propose a two-step pseudoweighting procedure that post-stratifies the event rates among age/race/sex strata in the pseudoweighted cohort to the population rates, to produce efficient and robust pure risk estimation (i.e. a cause-specific absolute risk in the absence of competing events). For developing an all-cause mortality risk model representative for the USA, our findings suggest that hazard ratios for minorities are not generalisable, and that surveys can have inadequate numbers of events for minorities. Post-stratification on event rates is crucial for obtaining reliable risk estimation for minority subgroups. [ABSTRACT FROM AUTHOR]

Public health policymakers must make crucial decisions rapidly during a pandemic. In such situations, accurate measurements from health surveys are essential. As a consequence of limited time and resource constraints, it may be infeasible to implement a probability-based sample that yields high response rates. An alternative approach is to select a quota sample from a large pool of volunteers, with the quota sample selection based on the census distributions of available—often demographic—variables, also known as quota variables. In practice, however, census data may only contain a subset of the required predictor variables. Thus, the realized quota sample can be adjusted by propensity score pseudoweighting using a "reference" probability-based survey that contains more predictor variables. Motivated by the SARS-CoV-2 serosurvey (a quota sample conducted in 2020 by the National Institutes of Health), we identify the condition under which the quota variables can be ignored in constructing the propensity model but still produce nearly unbiased estimation of population means. We conduct limited simulations to evaluate the bias and variance reduction properties of alternative weighting strategies for quota sample estimates under three propensity models that account for varying sets of predictors and degrees of correlation among the predictor sets and then apply our findings to the empirical data. [ABSTRACT FROM AUTHOR]

Joinpoint regression can model trends in time-specific estimates from aggregated data. These methods have been developed mainly for nonsurvey data such as cancer registry data assuming that the time-specific estimates are uncorrelated from time point to time point. This independence assumption can be violated for trends in time-specific estimates from complex survey samples due to using the same primary sampling units across time and, therefore, the full variance–covariance matrix of the time-specific estimates should be incorporated into the regression model fitting. This article extends these joinpoint methods for analyzing complex survey data within the National Cancer Institute's Joinpoint software and empirically compares the extended method to existing methods for analyses of time trends in three surveys. [ABSTRACT FROM AUTHOR]

We study the efficiency of covariate‐specific estimates of pure risk (one minus the survival function) when some covariates are only available for case‐control samples nested in a cohort. We focus on the semiparametric additive hazards model in which the hazard function equals a baseline hazard plus a linear combination of covariates with either time‐varying or time‐invariant coefficients. A published approach uses the design‐based inclusion probabilities to reweight the nested case‐control data. We obtain more efficient estimates of pure risks by calibrating the design weights to data available in the entire cohort, for both time‐varying and time‐invariant covariate coefficients. We develop explicit variance formulas for the weight‐calibrated estimates based on influence functions. Simulations show the improvement in precision by using weight calibration and confirm the consistency of variance estimators and the validity of inference based on asymptotic normality. Examples are provided using data from the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial Study (PLCO). [ABSTRACT FROM AUTHOR]

Cohort studies provide information on relative hazards and pure risks of disease. For rare outcomes, large cohorts are needed to have sufficient numbers of events, making it costly to obtain covariate information on all cohort members. We focus on nested case‐control designs that are used to estimate relative hazard in the Cox regression model. In 1997, Langholz and Borgan showed that pure risk can also be estimated from nested case‐control data. However, these approaches do not take advantage of some covariates that may be available on all cohort members. Researchers have used weight calibration to increase the efficiency of relative hazard estimates from case‐cohort studies and nested cased‐control studies. Our objective is to extend weight calibration approaches to nested case‐control designs to improve precision of estimates of relative hazards and pure risks. We show that calibrating sample weights additionally against follow‐up times multiplied by relative hazards during the risk projection period improves estimates of pure risk. Efficiency improvements for relative hazards for variables that are available on the entire cohort also contribute to improved efficiency for pure risks. We develop explicit variance formulas for the weight‐calibrated estimates. Simulations show how much precision is improved by calibration and confirm the validity of inference based on asymptotic normality. Examples are provided using data from the American Association of Retired Persons Diet and Health Cohort Study. [ABSTRACT FROM AUTHOR]

Motivation Nonparametric multivariate analysis has been widely used to identify variables associated with a dissimilarity matrix and to quantify their contribution. For very large studies (⁠ n ≥ 5000 ⁠) and many explanatory variables, existing software packages (e.g. adonis and adonis2 in vegan) are computationally intensive when conducting sequential multivariate analysis with permutations or bootstrapping. Moreover, for subjects from a complex sampling design, we need to adjust for sampling weights to derive an unbiased estimate. Results We implemented an R function fast.adonis to overcome these computational challenges in large-scale studies. fast.adonis generates results consistent with adonis/adonis2 but much faster. For complex sampling studies, fast.adonis integrates sampling weights algebraically to mimic the source population; thus, analysis can be completed very fast without requiring a large amount of memory. Availability and implementation fast.adonis is implemented using R and is publicly available at https://github.com/jennylsl/fast.adonis. Supplementary information Supplementary data are available at Bioinformatics Advances online. [ABSTRACT FROM AUTHOR]

Recent epidemiologic studies have examined the association of fish consumption with upper gastrointestinal cancer risk, but the associations with n-3 and n-6 polyunsaturated fatty acid (PUFA) subtypes remain unclear. Using the National Institutes of Health–AARP Diet and Health Study (United States, 1995–2011), we prospectively investigated the associations of PUFA subtypes, ratios, and fish with the incidence of head and neck cancer (HNC; n  = 2,453), esophageal adenocarcinoma (EA; n  = 855), esophageal squamous cell carcinoma (n  = 267), and gastric cancer (cardia: n  = 603; noncardia: n  = 631) among 468,952 participants (median follow-up, 15.5 years). A food frequency questionnaire assessed diet. Multivariable-adjusted hazard ratios were estimated using Cox proportional hazards regression. A Benjamini-Hochberg (BH) procedure was used for false-discovery control. Long-chain n-3 PUFAs were associated with a 20% decreased HNC and EA risk (for HNC, quintile5 vs. 1 hazard ratio = 0.81, 95% confidence interval: 0.71, 0.92, and BH-adjusted P trend = 0.001; and for EA, quintile5 vs. 1 hazard ratio = 0.79, 95% confidence interval: 0.64, 0.98, and BH-adjusted P trend = 0.1). Similar associations were observed for nonfried fish but only for high intake. Further, the ratio of long-chain n-3:n-6 was associated with a decreased HNC and EA risk. No consistent associations were observed for gastric cancer. Our results indicate that dietary long-chain n-3 PUFA and nonfried fish intake are associated with lower HNC and EA risk. [ABSTRACT FROM AUTHOR]

Summary: For various reasons, cohort studies generally forgo probability sampling required to obtain population representative samples. However, such cohorts lack population representativeness, which invalidates estimates of population prevalences for novel health factors that are only available in cohorts. To improve external validity of estimates from cohorts, we propose a kernel weighting (KW) approach that uses survey data as a reference to create pseudoweights for cohorts. A jackknife variance is proposed for the KW estimates. In simulations, the KW method outperformed two existing propensity‐score‐based weighting methods in mean‐squared error while maintaining confidence interval coverage. We applied all methods to estimating US population mortality and prevalences of various diseases from the non‐representative US National Institutes of Health–American Association of Retired Persons cohort, using the sample from the US‐representative National Health Interview Survey as the reference. Assuming that the survey estimates are correct, the KW approach yielded generally less biased estimates compared with the existing propensity‐score‐based weighting methods. [ABSTRACT FROM AUTHOR]

We extended the mortality follow-up of a cohort of 25,460 workers employed at 8 acrylonitrile (AN)-producing facilities in the United States by 21 years. Using 8,124 deaths and 1,023,922 person-years of follow-up, we evaluated the relationship between occupational AN exposure and death. Standardized mortality ratios (SMRs) based on deaths through December 31, 2011, were calculated. Work histories and monitoring data were used to develop quantitative estimates of AN exposure. Hazard ratios were estimated by Cox proportional hazards regression. All-cause mortality and death from total cancer were less than expected compared with the US population. We observed an excess of death due to mesothelioma (SMR = 2.24, 95% confidence interval (CI): 1.39, 3.42); no other SMRs were elevated overall. Cox regression analyses revealed an elevated risk of lung and bronchial cancer (n = 808 deaths; for >12.1 ppm-year vs. unexposed, hazard ratio (HR) = 1.43, 95% CI: 1.13, 1.81; P for trend = 0.05), lagged 10 years, that was robust in sensitivity analyses adjusted for smoking and co-exposures including asbestos. Death resulting from bladder cancer (for >2.56 ppm vs. unexposed, lagged 10-year HR = 2.96, 95% CI: 1.38, 6.34; P for trend = 0.02) and pneumonitis (for >3.12 ppm-year vs. unexposed, HR = 4.73, 95% CI: 1.42, 15.76; P for trend = 0.007) was also associated with AN exposure. We provide additional evidence of an association between AN exposure and lung cancer, as well as possible increased risk for death due to bladder cancer and pneumonitis. [ABSTRACT FROM AUTHOR]

Cryptorchidism and hypospadias are the most common congenital anomalies of the genitourinary tract in males, but their etiology remains unclear. Placental insufficiency has been suggested to be linked to both conditions. Placental weight is a commonly used proxy measure for placental insufficiency; thus, we examined placental weight and other placental characteristics in relation to cryptorchidism and hypospadias in the Collaborative Perinatal Project, a US mother-child cohort study. Pregnant women were recruited between 1959 and 1965. The analysis contrasted boys with cryptorchidism (n = 413) and boys with hypospadias (n = 145) with boys without cryptorchidism (n = 23,799) and boys without hypospadias (n = 22,326). Odds ratios and 95% confidence intervals were calculated using unconditional logistic regression. In categorical analyses in which the middle tertile was the referent, cryptorchidism was inversely associated with placental weight (odds ratio = 0.66, 95% confidence interval: 0.46, 0.95) among white boys and positively associated with the lowest tertile of placental weight among black boys (odds ratio = 1.70, 95% confidence interval: 1.11, 2.59). We conclude that lower placental weight may be related to risk of cryptorchidism. Further investigation of placental functioning may offer insights into the etiology of cryptorchidism. [ABSTRACT FROM AUTHOR]

Misclassification of body mass index (BMI) categories arising from self-reported weight and height can bias hazard ratios in studies of BMI and mortality. We examined the effects on hazard ratios of such misclassification using national US survey data for 1976 through 2010 that had both measured and self-reported weight and height along with mortality follow-up for 48,763 adults and a subset of 17,405 healthy never-smokers. BMI was categorized as <22.5 (low), 22.5-24.9 (referent), 25.0-29.9 (overweight), 30.0-34.9 (class I obesity), and ≥35.0 (class II-III obesity). Misreporting at higher BMI categories tended to bias hazard ratios upwards for those categories, but that effect was augmented, counterbalanced, or even reversed by misreporting in other BMI categories, in particular those that affected the reference category. For example, among healthy male never-smokers, misclassifications affecting the overweight and the reference categories changed the hazard ratio for overweight from 0.85 with measured data to 1.24 with self-reported data. Both the magnitude and direction of bias varied according to the underlying hazard ratios in measured data, showing that findings on bias from one study should not be extrapolated to a study with different underlying hazard ratios. Because of misclassification effects, self-reported weight and height cannot reliably indicate the lowest-risk BMI category. [ABSTRACT FROM AUTHOR]

Background: Evidence suggests that obesity in adulthood is associated with increased risk of "clinically significant" prostate cancer. However, studies of body mass index (BMI) across the adult life course and prostate cancer risks remain limited.Methods: In a prospective cohort of 69 873 men in the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial, we examined associations of prediagnostic BMI across the adult life course with risk of incident prostate cancer and fatal prostate cancer (prostate cancer-specific mortality). At 13 years of follow-up, we identified 7822 incident prostate cancer cases, of which 3078 were aggressive and 255 fatal. BMI trajectories were determined using latent-class trajectory modeling. Cox proportional hazards regression was used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs).Results: BMI at age 20 years, 50 years, and baseline questionnaire (mean age = 63 years) were associated with increased risks of fatal prostate cancer (HRs = 1.27-1.32 per five-unit increase). In five BMI trajectories identified, fatal prostate cancer risk was increased in men who had a normal BMI (HR = 1.95, 95% CI = 1.21 to 3.12) or who were overweight (HR = 2.65, 95% CI = 1.35 to 5.18) at age 20 years and developed obesity by baseline compared with men who maintained a normal BMI. Aggressive and nonaggressive prostate cancer were not associated with BMI, and modest inverse associations were seen for total prostate cancer.Conclusions: Our results suggest that BMI trajectories during adulthood that result in obesity lead to an elevated risk of fatal prostate cancer. [ABSTRACT FROM AUTHOR]

Background: Coffee drinkers had a higher risk of lung cancer in some previous studies, but as heavy coffee drinkers tend to also be cigarette smokers, such findings could be confounded. Therefore, we examined this association in the nearly half a million participants of the US NIH-AARP Diet and Health Study.Methods: Typical coffee intake and smoking history were queried at baseline. During 4 155 256 person-years of follow-up, more than 9000 incident lung cancer cases occurred. We used Cox proportional hazards regression to estimate hazard ratios (HRs)and 95% confidence intervals for coffee intake and subsequent incidence of lung cancer. We also comprehensively adjusted for tobacco smoking and examined associations by detailed strata of tobacco use.Results: Coffee drinkers were far more likely to smoke than non-drinkers. Although coffee drinking was associated with lung cancer in age- and sex- adjusted models (HR for ≥ 6 cups/day compared with none: 4.56, 4.08-5.10), this association was substantially attenuated after adjusting for smoking (HR: 1.27, 1.14-1.42). Similar findings were observed for each different histological type of lung cancer, and for participants drinking predominantly caffeinated or decaffeinated coffee. Little evidence for an association was observed in our stratified analyses, either within never smokers or in most categories of tobacco use.Conclusions: Coffee drinking was positively associated with lung cancer in our study, although the association was substantially attenuated after adjustment for tobacco smoking. As our adjustment for lifetime tobacco use was imperfect, it is likely that the remaining association is due to residual confounding by smoking, although other explanations are possible. [ABSTRACT FROM AUTHOR]

We used male pattern baldness as a proxy for long-term androgen exposure and investigated the association of dermatologist-assessed hair loss with prostate cancer--specific mortality in the first National Health and Nutrition Examination Survey Epidemiologic Follow-up Study. From the baseline survey (1971-1974), we included 4,316 men who were 25-74 years of age and had no prior cancer diagnosis. We estimated hazard ratios and used Cox proportional hazards regressions with age as the time metric and baseline hazard stratified by baseline age. A hybrid framework was used to account for stratification and clustering of the sample design, with adjustment for the variables used to calculate sample weights. During follow-up (median, 21 years), 3,284 deaths occurred; prostate cancer was the underlying cause of 107. In multivariable models, compared with no balding, any baldness was associated with a 56% higher risk of fatal prostate cancer (hazard ratio = 1.56; 95% confidence interval: 1.02, 2.37), and moderate balding specifically was associated with an 83% higher risk (hazard ratio = 1.83; 95% confidence interval: 1.15, 2.92). Conversely, patterned hair loss was not statistically significantly associated with all-cause mortality. Our analysis suggests that patterned hair loss is associated with a higher risk of fatal prostate cancer and supports the hypothesis of overlapping pathophysiological mechanisms. [ABSTRACT FROM AUTHOR]

Concerns about high caffeine intake and coffee as a vehicle for added fat and sugar have raised questions about the net impact of coffee on health. Although inverse associations have been observed for overall mortality, data for cause-specific mortality are sparse. Additionally, few studies have considered exclusively decaffeinated coffee intake or use of coffee additives. Coffee intake was assessed at baseline by self-report in the Prostate, Lung, Colo-rectal and Ovarian Cancer Screening Trial. Hazard ratios were estimated using Cox proportional hazards models. Among 90,317 US adults without cancer at study baseline (1998-2001) or history of cardiovascular disease at study enrollment (1993-2001), 8,718 deaths occurred during 805,644 person-years of follow-up from 1998 through 2009. Following adjustment for smoking and other potential confounders, coffee drinkers, as compared with nondrinkers, had lower hazard ratios for overall mortality (<1 cup/day: hazard ratio (HR) = 0.99 (95% confidence interval (CI): 0.92, 1.07); 1 cup/day: HR = 0.94 (95% CI: 0.87, 1.02); 2-3 cups/day: HR = 0.82 (95% CI: 0.77, 0.88); 4-5 cups/ day: HR = 0.79 (95% CI: 0.72, 0.86); ≥6 cups/day: HR = 0.84 (95% CI: 0.75, 0.95)). Similar findings were observed for decaffeinated coffee and coffee additives. Inverse associations were observed for deaths from heart disease, chronic respiratory diseases, diabetes, pneumonia and influenza, and intentional self-harm, but not cancer. Coffee may reduce mortality risk by favorably affecting inflammation, lung function, insulin sensitivity, and depression. [ABSTRACT FROM AUTHOR]

Background: Statins (3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors) are widely prescribed to reduce cholesterol levels. Studies have suggested that statins are associated with reduced risk for liver cancer, but much of the evidence is from regions of the world with high liver cancer incidence rates. The current study examined the statins-liver cancer relationship in a low-rate region and examined the effects of preexisting liver disease and diabetes on that association. Methods: A nested case-control study was conducted within the United Kingdom's Clinical Practice Research Datalink (CPRD). Persons diagnosed with primary liver cancer between 1988 and 2011 were matched to controls at a four-to-one ratio. Matches stratified on liver disease and on diabetes were also completed. Odds ratios (ORs) and 95% confidence intervals (CIs) for associations of statins with liver cancer were estimated using conditional logistic regression. Results: In total, 1195 persons with primary liver cancer were matched to 4640 control patients. Statin use was associated with a statistically significantly reduced risk for liver cancer (ORadj = 0.55, 95% CI = 0.45 to 0.69), especially among current users (ORadj = 0.53, 95% CI = 0.42 to 0.66). The reduced risk was statistically significant in the presence (ORadj = 0.32, 95% CI = 0.17 to 0.57) and absence of liver disease (ORadj = 0.65, 95% CI = 0.52 to 0.81) and in the presence (ORadj = 0.30, 95% CI = 0.21 to 0.42) and absence of diabetes (ORadj = 0.66, 95% CI = 0.51 to 0.85). Conclusions: In the current study in a low-rate area, statin use was associated with a statistically significantly reduced risk for liver cancer overall. Risk was particularly reduced among persons with liver disease and persons with diabetes, suggesting that statin use may be especially beneficial in persons at elevated risk for liver cancer. [ABSTRACT FROM AUTHOR]

Epidemiologic cross-sectional, case-cohort, or case–control studies often select augmentation samples to supplement an existing (baseline) sample, primarily for the two reasons: (1) to increase the sample sizes from certain subdomains of interest that were not originally considered in the design of the baseline study and (2) to obtain samples from an extension of the target population. To address these two objectives, two-stage stratified sample designs are considered, where the stratification based on the expanded population at the second stage is not nested in the first stage strata. The sample weighting and Taylor linearization variance estimation for the two-stage stratified sample designs, involving re-stratification and population expansion, are provided for estimating population totals and logistic regression coefficients. Results from limited simulation studies and a logistic regression analysis of a study of human papillomavirus serology are provided. [ABSTRACT FROM PUBLISHER]

We investigated the association of demographic and behavioral factors with oral human papillomavirus (HPV) load for 18 high-risk types among 211 individuals with prevalent high-risk HPV within the National Health and Nutrition Examination Survey 2009-2010. Factors independently associated with HPV load above the median included older age (odds ratio, 1.04 per year increase [95% confidence interval, 1.01-1.07]; P = .004) and intensity of current smoking (P for trend <.001). A marginally greater percentage of men than women had an HPV load above the median (55.7% vs 32.8%; P = .069), and HPV load increased marginally with increasing alcohol use (P for trend = .062). In conclusion, older age and current smoking are associated with a high oral load of high-risk HPV types among individuals with a prevalent infection. [ABSTRACT FROM AUTHOR]

The World Health Organization (Geneva, Switzerland) and the National Heart, Lung, and Blood Institute (Bethesda, Maryland) have developed standard categories of body mass index (BMI) (calculated as weight (kg)/height (m)2) of less than 18.5 (underweight), 18.5–24.9 (normal weight), 25.0–29.9 (overweight), and 30.0 or more (obesity). Nevertheless, studies of BMI and the risk of death sometimes use nonstandard BMI categories that vary across studies. In a meta-analysis of 8 large studies that used nonstandard BMI categories and were published between 1999 and 2014 and included 5.8 million participants, hazard ratios tended to be small throughout the range of overweight and normal weight. Risks were similar between subjects of high-normal weight (BMI of approximately 23.0–24.9) and those of low overweight (BMI of approximately 25.0–27.4). In an example using national survey data, minor variations in the reference category affected hazard ratios. For example, choosing high-normal weight (BMI of 23.0–24.9) instead of standard normal weight (BMI of 18.5–24.9) as the reference category produced higher nonsignificant hazard ratios (1.05 vs. 0.97 for men and 1.06 vs. 1.02 for women) for the standard overweight category (BMI of 25.0–29.9). Use of the standard BMI groupings avoids problems of ad hoc and post hoc category selection and facilitates between-study comparisons. The ways in which BMI data are categorized and reported may shape inferences about the degree of risk for various BMI categories. [ABSTRACT FROM PUBLISHER]

Dietary iron intake and variation in iron homeostasis genes may affect colorectal neoplasia risk. This study examines the interaction of dietary iron intake and genetic variation in iron homeostatic genes on the outcomes of colorectal adenoma and colorectal cancer.Dietary iron intake and variation in iron homeostasis genes may affect colorectal neoplasia risk. We conducted two nested case–control studies within the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial: one of advanced colorectal adenoma (1205 cases; 1387 controls) and one of colorectal cancer (370 cases; 401 controls). Iron intake was estimated with a food frequency questionnaire and genotyping was performed for 21 genes. Unconditional logistic regression was used to estimate odds ratio (OR) and 95% confidence intervals (95% CIs) for colorectal neoplasia risk within quartiles of intake. Several single nucleotide polymorphisms (SNPs) modified the association between iron intake and the risk of adenoma or cancer. Dietary iron was positively associated with colorectal adenoma among three SNPs of HEPHL1, including carriers of the AA genotype at rs7946162 (ORQ4–Q1 = 2.22, 95% CI 1.15–4.27, Ptrend = 0.03; Pinteraction = 0.10), the TT genotype at rs2460063 (ORQ4–Q1 = 2.39, 95% CI 1.26–4.54, Ptrend = 0.02; Pinteraction = 0.04) and the GG genotype at rs7127348 (ORQ4–Q1 = 2.40, 95% CI 1.23–4.67, Ptrend = 0.02; Pinteraction = 0.09). Heme iron was positively associated with colorectal cancer among those with GG genotypes for ACO1 rs10970985 (ORQ4–Q1 = 2.45, 95% CI 3.40–8.06, Ptrend = 0.004; Pinteraction = 0.05). However, none of the associations were statistically significant after adjustment for multiple comparisons. Future studies should target the specific genes and SNPs for which the association was significant prior to multiple comparison correction. [ABSTRACT FROM AUTHOR]

In environmental epidemiology, measurements of exposure biomarkers often fall below the assay's limit of detection. Existing methods for handling this problem, including deletion, substitution, parametric regression, and multiple imputation, can perform poorly if the proportion of “nondetects” is high or parametric models are misspecified. We propose an approach that treats the measured analyte as the modeled outcome, implying a role reversal when the analyte is a putative cause of a health outcome. Following a scale reversal as well, our approach uses Cox regression to model the analyte, with confounder adjustment. The method makes full use of quantifiable analyte measures, while appropriately treating nondetects as censored. Under the proportional hazards assumption, the hazard ratio for a binary health outcome is interpretable as an adjusted odds ratio: the odds for the outcome at any particular analyte concentration divided by the odds given a lower concentration. Our approach is broadly applicable to cohort studies, case-control studies (frequency matched or not), and cross-sectional studies conducted to identify determinants of exposure. We illustrate the method with cross-sectional survey data to assess sex as a determinant of 2,3,7,8-tetrachlorodibenzo-p-dioxin concentration and with prospective cohort data to assess the association between 2,4,4′-trichlorobiphenyl exposure and psychomotor development. [ABSTRACT FROM PUBLISHER]

Background We investigated risk factors for inflammatory breast cancer (IBC), a rare, aggressive, and poorly understood breast cancer that is characterized by diffuse breast skin erythema and edema. Methods We included 617 IBC case subjects in a nested case-control study from the Breast Cancer Surveillance Consortium database (1994-2009). We also included 1151 noninflammatory, locally advanced, invasive breast cancers with chest wall/breast skin involvement (LABC), 7600 noninflammatory invasive case subjects without chest wall/ breast skin involvement (BC), and 93 654 control subjects matched to case subjects on age and year at diagnosis and mammography registry. We present estimates of rate ratios (RRs) and 95% confidence intervals (CI) from conditional logistic regression analyses for each case group vs control subjects based on multiply imputed datasets. Results First-degree family history of breast cancer and high mammographic breast density increased risk of IBC, LABC, and BC. High body mass index (BMI) increased IBC risk irrespective of menopausal status and estrogen receptor (ER) expression; rate ratios for BMI 30 and greater vs BMI less than 25 were 3.90 (95% CI = 1.50 to 10.14) in premenopausal women and 3.70 (95% CI = 1.98 to 6.94) in peri/postmenopausal women not currently using hormones. BMI 30 and greater slightly increased risk of ER-positive BC (RR = 1.40; 95% CI = 1.11 to 1.76). Statistically significant reductions in risk of ER-negative IBC with older age at first birth and of ER-positive IBC with higher education were not seen for LABC and BC of the same ER status. Conclusions Different associations with BMI, age at first birth, and education between IBC and/or LABC and BC suggest a distinct etiology for IBC. [ABSTRACT FROM AUTHOR]

BACKGROUND Debate exists about the relationship between hypovitaminosis D and cardiovascular (CVD) risk. METHODS This study investigated baseline (n=2,271) 25-hydroxyvitamin D (25(OH)D) and baseline and 4 year (n=1,957) CVD risk in a cohort of Finnish middle-aged male smokers. RESULTS The prevalences of measured hypertension, high pulse rate, diabetes, and coronary heart disease were 63%, 16%, 5%, and 10% at baseline and were 64%, 20%, 6%, and 16% at 4 years after baseline. The mean 25(OH)D was 41±18 nmol/L. At baseline, systolic blood pressure (β = −0.048; P = 0.02), and pulse rate (β = −0.043; P = 0.04) were both associated with lower 25(OH)D levels but not coronary heart disease or diabetes prevalence. On remeasuring CVD risk 4 years after baseline, the only significant association with baseline 25(OH)D levels was high pulse rate (β = −0.077; P = 0.001). In addition, a higher 25(OH)D level at baseline was associated with a change in pulse rate (β = −0.055; P = 0.01). These trends for hypertension (baseline) and high pulse rate (baseline and 4 years after baseline) were also seen on adjusted categorical analysis (Ptrend < 0.05). CONCLUSIONS Vitamin D deficiency at baseline was associated with hypertension in Finnish male smokers, but not after 4 years. These results are consistent with recent findings in other large cohort studies with measured blood pressure. Change in pulse rate over time continued to be significantly associated with lower 25(OH)D baseline levels; this new finding should be investigated further. [ABSTRACT FROM AUTHOR]

Circulating 25-hydroxyvitamin D (25(OH)D), a marker for vitamin D status, is associated with bone health and possibly cancers and other diseases; yet, the determinants of 25(OH)D status, particularly ultraviolet radiation (UVR) exposure, are poorly understood. Determinants of 25(OH)D were analyzed in a subcohort of 1,500 participants of the US Radiologic Technologists (USRT) Study that included whites (n = 842), blacks (n = 646), and people of other races/ethnicities (n = 12). Participants were recruited monthly (2008–2009) across age, sex, race, and ambient UVR level groups. Questionnaires addressing UVR and other exposures were generally completed within 9 days of blood collection. The relation between potential determinants and 25(OH)D levels was examined through regression analysis in a random two-thirds sample and validated in the remaining one third. In the regression model for the full study population, age, race, body mass index, some seasons, hours outdoors being physically active, and vitamin D supplement use were associated with 25(OH)D levels. In whites, generally, the same factors were explanatory. In blacks, only age and vitamin D supplement use predicted 25(OH)D concentrations. In the full population, determinants accounted for 25% of circulating 25(OH)D variability, with similar correlations for subgroups. Despite detailed data on UVR and other factors near the time of blood collection, the ability to explain 25(OH)D was modest. [ABSTRACT FROM PUBLISHER]

Background Nonsteroidal anti-inflammatory drugs (NSAIDs) have been shown to reduce chronic inflammation and risk of many cancers, but their effect on risk of hepatocellular carcinoma (HCC) and death due to chronic liver disease (CLD) has not been investigated. Methods We analyzed prospective data on 300504 men and women aged 50 to 71 years in the National Institutes of Health–AARP Diet and Health Study cohort and linked self-reported aspirin and nonaspirin NSAID use with registry-confirmed diagnoses of HCC (n=250) and death due to CLD (n=428, excluding HCC). We calculated hazard rate ratios (RRs) and their two-sided 95% confidence intervals (CIs) using Cox proportional hazard regression models with adjustment for age, sex, race/ethnicity, cigarette smoking, alcohol consumption, diabetes, and body mass index. All tests of statistical significance were two-sided. Results Aspirin users had statistically significant reduced risks of incidence of HCC (RR = 0.59; 95% CI = 0.45 to 0.77) and mortality due to CLD (RR = 0.55; 95% CI = 0.45 to 0.67) compared to those who did not use aspirin. In contrast, users of nonaspirin NSAIDs had a reduced risk of mortality due to CLD (RR = 0.74; 95% CI= 0.61 to 0.90) but did not have lower risk of incidence of HCC (RR = 1.08; 95% CI = 0.84 to 1.39) compared to those who did not use nonaspirin NSAIDs. The risk estimates did not vary in statistical significance by frequency (monthly, weekly, daily) of aspirin use, but the reduced risk of mortality due to CLD was statistically significant only among monthly users of nonaspirin NSAIDs compared to non-users. Conclusions Aspirin use was associated with reduced risk of developing HCC and of death due to CLD whereas nonaspirin NSAID use was only associated with reduced risk of death due to CLD. [ABSTRACT FROM PUBLISHER]

Advances in human genetics have led to epidemiological investigations not only of the effects of genes alone but also of gene–environment (G–E) interaction. A widely accepted design strategy in the study of how G–E relate to disease risks is the population-based case–control study (PBCCS). For simple random samples, semiparametric methods for testing G–E have been developed by Chatterjee and Carroll in 2005. The use of complex sampling in PBCCS that involve differential probabilities of sample selection of cases and controls and possibly cluster sampling is becoming more common. Two complexities, weighting for selection probabilities and intracluster correlation of observations, are induced by the complex sampling. We develop pseudo-semiparametric maximum likelihood estimators (pseudo-SPMLE) that apply to PBCCS with complex sampling. We study the finite sample performance of the pseudo-SPMLE using simulations and illustrate the pseudo-SPMLE with a US case–control study of kidney cancer. [ABSTRACT FROM AUTHOR]

Meat mutagens, including heterocyclic amines (HCAs), polycyclic aromatic hydrocarbons (PAHs) and N-nitroso compounds (NOCs), may be involved in colorectal carcinogenesis depending on their activation or detoxification by phase I and II xenobiotic metabolizing enzymes (XME). Using unconditional logistic regression to estimate odds ratios (OR) and 95% confidence intervals (CI), we examined the intake of five meat mutagens and >300 single nucleotide polymorphisms (SNPs) in 18 XME genes in relation to advanced colorectal adenoma (1205 cases and 1387 controls) and colorectal cancer (370 cases and 401 controls) within the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial. Dietary intake of meat mutagens was assessed using a food frequency questionnaire with a detailed meat-cooking module. An interaction was observed between 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (MeIQx) intake and the NAT1 polymorphism rs6586714 in the adenoma study (Pinteraction = 0.001). Among individuals carrying a GG genotype, high MeIQx intake was associated with a 43% increased risk of adenoma (95% CI 1.11–1.85, Ptrend = 0.07), whereas the reverse was observed among carriers of the A variant (OR = 0.50, 95% CI 0.30–0.84, Ptrend = 0.01). In addition, we observed some suggestive (P < 0.05) modifying effects for SNPs in other XME genes (UGT1A, CYP2E1, EPHX1, AHR and GSTM3), but these were not significant after adjustment for multiple testing. This large and comprehensive study of XME genes, meat mutagens and the risk of colorectal tumours found that a NAT1 polymorphism modified the association between MeIQx intake and colorectal adenoma risk. [ABSTRACT FROM AUTHOR]

Background: Previously, we have shown that increasing adult height is associated with increased risk of testicular germ-cell tumor (TGCT). Recently, a number of single nucleotide polymorphisms (SNPs) have been found to be related to height. We examined whether these SNPs were associated with TGCT and whether they explained the relationship between height and TGCT.Methods: We genotyped 15 height-related SNPs in the US Servicemen's Testicular Tumor Environmental and Endocrine Determinants (STEED) case-control study. DNA was extracted from buccal cell samples and Taqman assays were used to type the selected SNPs. We used logistic regression models to estimate odds ratios (ORs) and 95% confidence intervals (95%CIs).Results: There were 561 cases and 676 controls for analysis. Two SNPs were found to be associated with risk of TGCT, rs6060373 (CC vs TT, OR = 1.51, 95% CI: 1.06-2.15) and rs143384 (CC vs TT, OR = 1.53, 95% CI: 1.09-2.15). rs6060373 is an intronic polymorphism of ubiquinol-cytochrome c reductase complex chaperone (UQCC), and rs143384 is a 5'UTR polymorphism of growth differentiation factor 5 (GDF5). No individual SNP attenuated the association between height and TGCT. Adjustment for all SNPs previously associated with adult height reduced the associations between adult height and TGCT by ~8.5%, although the P-value indicated only weak evidence that this difference was important (P = 0.26).Conclusions: This novel analysis provides tentative evidence that SNPs which are associated with adult height may also share an association with risk of TGCT. [ABSTRACT FROM AUTHOR]

Background Knowledge of family cancer history is essential for estimating an individual’s cancer risk and making clinical recommendations regarding screening and referral to a specialty cancer genetics clinic. However, it is not clear if reported family cancer history is sufficiently accurate for this purpose. Methods In the population-based 2001 Connecticut Family Health Study, 1019 participants reported on 20 578 first-degree relatives (FDR) and second-degree relatives (SDR). Of those, 2605 relatives were sampled for confirmation of cancer reports on breast, colorectal, prostate, and lung cancer. Confirmation sources included state cancer registries, Medicare databases, the National Death Index, death certificates, and health-care facility records. Sensitivity, specificity, positive predictive value, and negative predictive value were calculated for reports on lung, colorectal, breast, and prostate cancer and after stratification by sex, age, education, and degree of relatedness and used to estimate report accuracy. Pairwise t tests were used to evaluate differences between the two strata in each stratified analysis. All statistical tests were two-sided. Results Overall, sensitivity and positive predictive value were low to moderate and varied by cancer type: 60.2% and 40.0%, respectively, for lung cancer reports, 27.3% and 53.5% for colorectal cancer reports, 61.1% and 61.3% for breast cancer reports, and 32.0% and 53.4% for prostate cancer reports. Specificity and negative predictive value were more than 95% for all four cancer types. Cancer history reports on FDR were more accurate than reports on SDR, with reports on FDR having statistically significantly higher sensitivity for prostate cancer than reports on SDR (58.9% vs 21.5%, P = .002) and higher positive predictive value for lung (78.1% vs 31.7%, P < .001), colorectal (85.8% vs 43.5%, P = .004), and breast cancer (79.9% vs 53.6%, P = .02). Conclusions General population reports on family history for the four major adult cancers were not highly accurate. Efforts to improve accuracy are needed in primary care and other health-care settings in which family history is collected to ensure appropriate risk assessment and clinical care recommendations. [ABSTRACT FROM PUBLISHER]

Complex sample designs, involving stratified and/or multistage sampling with sample weighting, along with frequency matching, are used to select controls or cases for case-control studies. Examples that motivated this paper are the Kaposi sarcoma case-control study that was conducted in Sicily and the US kidney cancer case-control study. Survey design-based approaches can be inefficient for the analysis of case-control studies with frequency matching. We propose a weighting method that post-stratifies control sample weights to the estimated population distribution of the matching variables among cases. This weighting maintains the efficiency of frequency matching. The method proposed is evaluated by using simulation studies and is applied to the two case-control studies. [ABSTRACT FROM AUTHOR]

In studies of weight and mortality, the construct of reverse causation has come to be used to imply that the exposure-outcome relation is biased by weight loss due to preexisting illness. Observed weight-mortality associations are sometimes thought to result from this bias. Evidence for the occurrence of such bias is weak and inconsistent, suggesting that either the analytical methods used have been inadequate or else illness-related weight loss is not an important source of bias. Deleting participants has been the most frequent approach to control possible bias. As implemented, this can lead to deletion of almost 90% of all deaths in a sample and to deletion of more overweight and obese participants than participants with normal or below normal weight. Because it has not been demonstrated that the procedures used to adjust for reverse causation increase validity or have large or systematic effects on relative risks, it is premature to consider reverse causation as an important cause of bias. Further research would be useful to elucidate the potential effects and importance of reverse causation or illness-related weight loss as a source of bias in the observed associations between weight and mortality in cohort studies. [ABSTRACT FROM PUBLISHER]

Identifying regions with the highest and lowest mortality rates and producing the corresponding color-coded maps help epidemiologists identify promising areas for analytic etiological studies. Based on a two-stage Poisson–Gamma model with covariates, we use information on known risk factors, such as smoking prevalence, to adjust mortality rates and reveal residual variation in relative risks that may reflect previously masked etiological associations. In addition to covariate adjustment, we study rankings based on standardized mortality ratios (SMRs), empirical Bayes (EB) estimates, and a posterior percentile ranking (PPR) method and indicate circumstances that warrant the more complex procedures in order to obtain a high probability of correctly classifying the regions with the upper and lower of relative risks for , and 0.2. We also give analytic approximations to the probabilities of correctly classifying regions in the upper of relative risks for these three ranking methods. Using data on mortality from heart disease, we found that adjustment for smoking prevalence has an important impact on which regions are classified as high and low risk. With such a common disease, all three ranking methods performed comparably. However, for diseases with smaller event counts, such as cancers, and wide variation in event counts among regions, EB and PPR methods outperform ranking based on SMRs. [ABSTRACT FROM AUTHOR]

Genetic data collected from surveys such as the Third National Health and Nutrition Examination Survey (NHANES III) enable researchers to investigate the association between wide varieties of health factors and genetic variation for the US population. Tests for trend in disease with increasing number of alleles have been developed for simple random samples. However, surveys such as the NHANES III have complex sample designs involving multistage cluster sampling and sample weighting. These types of sample designs can affect Type I error and power properties of statistical tests based on simple random samples. In order to address these issues, we have derived tests of trend based on Wald and quasi-score statistics, with and without assuming a genetic model, that account for the complex sampling design. The finite-sample properties of the proposed test procedures are evaluated via Monte Carlo simulation studies. We make recommendations about the choice of the test statistic depending on whether or not the underlying genetic model is known. Proposed test statistics are applied to NHANES III data to test for associations between the locus ADRB2 (rs1042713) and obesity, between VDR (rs2239185) and high blood lead level, and between TGFB1 (rs1982073) and asthma. [ABSTRACT FROM PUBLISHER]

For studies on population genetics, the use of representative random samples of the target population can avoid ascertainment bias. Genetic variation data from over a hundred genes were collected in a U.S. nationally representative sample in the Third National Health and Nutrition Examination Survey (NHANES III). Surveys such as the NHANES have complex stratified multistage cluster sample designs with sample weighting that can inflate variances and alter the expectations of test statistics. Thus, classical statistical tests of Hardy–Weinberg equilibrium (HWE) and homogeneity of HW disequilibrium (HHWD) for simple random samples are not suitable for data from complex samples. We propose using Wald tests for HWE and generalized score tests for HHWD that have been modified for complex samples. Monte Carlo simulation studies are used to investigate the finite sample properties of the proposed tests. Rao–Scott corrections applied to the tests were found to improve their type I error properties. Our methods are applied to the NHANES III genetic data for three loci involved in metabolizing lead in the body. [ABSTRACT FROM AUTHOR]

The authors examined associations between meat consumption (type, cooking method, and related mutagens), heme iron, nitrite/nitrate, and prostate cancer in a cohort of 175,343 US men aged 50–71 years. During 9 years of follow-up (1995–2003), they ascertained 10,313 prostate cancer cases (1,102 advanced) and 419 fatal cases. Hazard ratios comparing the fifth intake quintile with the first revealed elevated risks associated with red and processed meat for total (red meat: hazard ratio (HR) = 1.12, 95% confidence interval (CI): 1.04, 1.21; processed meat: HR = 1.07, 95% CI: 1.00, 1.14) and advanced (red meat: HR = 1.31, 95% CI: 1.05, 1.65; processed meat: HR = 1.32, 95% CI: 1.08, 1.61) prostate cancer. Heme iron, barbecued/grilled meat, and benzo[a]pyrene were all positively associated with total (HR = 1.09 (95% CI: 1.02, 1.17), HR = 1.11 (95% CI: 1.03, 1.19), and HR = 1.09 (95% CI: 1.00, 1.18), respectively) and advanced (HR = 1.28 (95% CI: 1.03, 1.58), HR = 1.36 (95% CI: 1.10, 1.69), and HR = 1.28 (95% CI: 1.00, 1.65), respectively) disease. Nitrite (HR = 1.24, 95% CI: 1.02, 1.51) and nitrate (HR = 1.31, 95% CI: 1.07, 1.61) intakes were associated with advanced prostate cancer. There were no clear associations for fatal prostate cancer. Red and processed meat may be positively associated with prostate cancer via mechanisms involving heme iron, nitrite/nitrate, grilling/barbecuing, and benzo[a]pyrene. [ABSTRACT FROM PUBLISHER]

A review of the article "Using the Peters-Belson Method in Equal Employment Opportunity Personnel Evaluations," by M.D. Sinclair and Q. Pan, which appeared on this issue of the journal is presented.

Background Exposure to endocrine-disrupting chemicals, such as persistent organochlorine pesticides, has been suggested to increase the risk of testicular germ cell tumors (TGCTs). Methods To study the relationship of POP exposure to TGCT risk, prediagnostic serum samples from 754 case subjects and 928 control subjects enrolled in the Servicemen's Testicular Tumor Environmental and Endocrine Determinants Study were analyzed for cis-nonachlor, trans-nonachlor, oxychlordane, total chlordanes, β-hexachlorocyclohexane, mirex, p,p′-dichlorodiphenyldichloroethylene (p,p′-DDE), and p,p′-dichlorodiphenyltrichloroethane. Adjusted odds ratios (ORs) and their associated 95% confidence intervals (CIs) for the risk of TGCT overall and for the histological subgroups, seminoma and nonseminoma, were estimated using multivariable logistic regression. All statistical tests were two-sided. Results TGCT risk was statistically significantly associated with higher plasma levels of p,p′-DDE (for highest quartile [Q4] vs lowest quartile [Q1], OR = 1.71, 95% Cl = 1.23 to 2.38, Ptrend = .0002) and of two chlordane components, cis-nonachlor (Q4 vs Q1, OR = 1.56, 95% Cl = 1.11 to 2.18, Ptrend = .009) and trans-nonachlor (Q4 vs Q1, OR = 1.46, 95% Cl = 1.07 to 2.00, Ptrend = .026). Seminoma risk was statistically significantly associated with p,p′-DDE (Q4 vs Q1, OR = 1.91, 95% Cl = 1.22 to 2.99, Ptrend = .0008), cis-nonachlor (Q4 vs Q1, OR = 1.93, 95% Cl = 1.27 to 2.93, Ptrend = .0045), trans-nonachlor (Q4 vs Q1, OR = 1.72, 95% Cl = 1.11 to 2.67, Ptrend = .033), and a chlordane metabolite, oxychlordane (Q4 vs Q1, OR = 1.64, 95% Cl = 1.04 to 2.60, Ptrend = .048), whereas nonseminoma risk showed a statistically significant association with p,p′-DDE only (Q4 vs Q1, OR = 1.63, 95% Cl = 1.10 to 2.42, Ptrend = .0044). Conclusions Increased exposure to p,p′-DDE may be associated with the risk of both seminomatous and nonseminomatous TGCTs, whereas exposure to chlordane compounds and metabolites may be associated with the risk of seminoma. Because evidence suggests that TGCT is initiated in very early life, it is possible that exposure to these persistent organic pesticides during fetal life or via breast feeding may increase the risk of TGCT in young men. [ABSTRACT FROM AUTHOR]

Background Vitamin D has been hypothesized to reduce cancer mortality through its effects on incidence and/or survival. Epidemiologic studies of the association of 25-hydroxyvitamin D [25(OH)D] and the risk of cancer, however, have been largely limited to incident cancers at a few sites. Methods A total of 16818 participants in the Third National Health and Nutrition Examination Survey who were 17 years or older at enrollment were followed from 1988-1994 through 2000. Levels of serum 25(OH)D were measured at baseline by radioimmunoassay. Cox proportional hazards regression models were used to examine the relationship between serum 25(OH)D levels and total cancer mortality (in the entire population or according to racelethnicity, sex, age, and retinol status) and mortality from specific cancers. Because serum was collected in the south in cooler months and the north in warmer months, we examined associations by collection season. All statistical tests were two-sided. Results We identified 536 cancer deaths in 146578 person-years. Total cancer mortality was unrelated to baseline vitamin D status in the entire population, men, women, non-Hispanic whites, non-Hispanic blacks, Mexican Americans, and in persons younger than 70 or 70 years or older. We found no interaction between vitamin D and season or vitamin D and serum retinol. Colorectal cancer mortality was inversely related to serum 25(OH)D level, with levels 80 nmol/L or higher associated with a 72% risk reduction (95% confidence interval = 32% to 89%) compared with lower than 50 nmol/L, Ptrend = .02. Conclusions Our results do not support an association between 25(OH)D and total cancer mortality, although there was an inverse relationship between 25(OH)D levels and colorectal cancer mortality. [ABSTRACT FROM AUTHOR]

Background. Human herpesvirus 8 (HHV-8) causes Kaposi sarcoma. In the United States, transmission routes for HHV-8 are uncertain. Methods. The National Health and Nutrition Examination Survey III sampled individuals from the US general population (1988-1994). We used enzyme immunoassays (EIAs) to measure HHV-8 antibodies (K8.1 and open reading frame [ORF] 73 antigens) in 13,894 surveyed adults. HHV-8 seroprevalence was examined according to sexual history and viral coinfection markers. Results. Overall, seroprevalence was low when a highly specific cutoff was used (K8.1, 1.6%; ORF73, 1.5%) but was higher when a less-specific cutoff was used (K8.1, 7.1%; ORF73, 7.4%). When the more-specific approach was used, K8.1 seroprevalence was similar in men and women. Men who have sex with men (MSM) had a higher K8.1 seroprevalence (8.2%). Among other men, K8.1 seroprevalence was marginally associated with duration of heterosexual activity (P = .1) and was positively associated with the lifetime number of sex partners (P = .04) and with coinfections with hepatitis B virus (6.1% vs. 1.2% without coinfection; P < .001) and herpes simplex virus 2 (2.7% vs. 1.0%; P = .1). Among women, K8.1 seroprevalence was not significantly related to duration of sexual activity, the lifetime number of sex partners, or viral coinfections. The ORF73 EIA revealed similar but less clear-cut patterns. Conclusions. Among men, HHV-8 transmission may occur through sexual activity, particularly sex with other men. No evidence was observed for heterosexual transmission to women. [ABSTRACT FROM AUTHOR]

Adjusted attributable risk (AR) is the proportion of diseased individuals in a population that is due to an exposure. We consider estimates of adjusted AR based on odds ratios from logistic regression to adjust for confounding. Influence function methods used in survey sampling are applied to obtain simple and easily programmable expressions for estimating the variance of . These variance estimators can be applied to data from case–control, cross-sectional, and cohort studies with or without frequency or individual matching and for sample designs with subject samples that range from simple random samples to (sample) weighted multistage stratified cluster samples like those used in national household surveys. The variance estimation of is illustrated with: (i) a weighted stratified multistage clustered cross-sectional study of childhood asthma from the Third National Health and Examination Survey (NHANES III), and (ii) a frequency-matched case–control study of melanoma skin cancer. [ABSTRACT FROM AUTHOR]

Previously reported estimates of deaths attributable to obesity in the United States have been based on a method that only partially adjusts for confounding and does not allow for effect modification. In this study, the authors investigated the possible magnitude and direction of bias in estimating deaths attributable to obesity when such a method is used. Hypothetical examples are based on 1991 US population data and published relative risks. Incomplete adjustment for confounding of the obesity-mortality relation by age and sex led to a 17% overestimation of deaths due to obesity. Additional bias resulted from slight differences between the derivation cohort and the target population. For example, a difference of three percentage points in the proportion of people 80 years of age or older led to a 42% overestimation of deaths due to obesity. In addition, these estimates appear to be sensitive to minor differences in relative risks between a derivation cohort and the target population. A difference of 0.20 in relative risks almost doubled the number of deaths (97% overestimation). Estimates of deaths attributable to obesity can be biased if confounding and effect modification are not properly taken into account or if the relative risks are not estimated accurately. [ABSTRACT FROM AUTHOR]

Background: The National Cancer Institute funded an 8-year, nonrandomized demonstration project for tobacco prevention and control, the American Stop Smoking Intervention Study (ASSIST). To evaluate ASSIST, we compared changes in adult smoking prevalence, per capita cigarette consumption, and tobacco control policies between the 17 ASSIST states and the 33 non-ASSIST states and the District of Columbia. Methods: The strength of tobacco control index was used to measure state-level program elements directed at tobacco control, and the initial outcomes index (IOI) was used to measure states' tobacco control policy outcomes. Prevalence data were obtained from the Tobacco Use Supplement to the Current Population Survey, and consumption data were obtained from the Tobacco Institute's bimonthly sales figures for cigarette packs moved from wholesale warehouses. Two-stage regression and mixed-effects linear modeling were used to analyze the various outcomes. Statistical analyses for testing individual regression coefficients were one-sided. Results: ASSIST states had a greater decrease in adult smoking prevalence than non-ASSIST states, with an adjusted difference of -0.63% (P = .049). Per capita cigarette consumption was not statistically significantly different between ASSIST and non-ASSIST states. However, an increase in the IOI of a state from the 25[sup th] to the 75[sup th] percentile was associated with a reduction in per capita cigarette consumption by 0.57 packs per person per month. State IOI was also inversely, albeit not statistically significantly, associated with smoking prevalence (regression coefficient = -0.11; P = .06). Conclusions: The reduction in adult smoking prevalence associated with ASSIST could have translated into approximately 278 700 fewer smokers nationwide if all states had implemented ASSIST. Investment in building state-level tobacco control capacity and promoting changes in tobacco control policies are effective strategies for reducing tobacco use. [ABSTRACT FROM AUTHOR]

Background: The National Cancer Institute funded an 8-year, nonrandomized demonstration project for tobacco prevention and control, the American Stop Smoking Intervention Study (ASSIST). To evaluate ASSIST, we compared changes in adult smoking prevalence, per capita cigarette consumption, and tobacco control policies between the 17 ASSIST states and the 33 non-ASSIST states and the District of Columbia.Methods: The strength of tobacco control index was used to measure state-level program elements directed at tobacco control, and the initial outcomes index (IOI) was used to measure states' tobacco control policy outcomes. Prevalence data were obtained from the Tobacco Use Supplement to the Current Population Survey, and consumption data were obtained from the Tobacco Institute's bimonthly sales figures for cigarette packs moved from wholesale warehouses. Two-stage regression and mixed-effects linear modeling were used to analyze the various outcomes. Statistical analyses for testing individual regression coefficients were one-sided.Results: ASSIST states had a greater decrease in adult smoking prevalence than non-ASSIST states, with an adjusted difference of -0.63% (P =.049). Per capita cigarette consumption was not statistically significantly different between ASSIST and non-ASSIST states. However, an increase in the IOI of a state from the 25th to the 75th percentile was associated with a reduction in per capita cigarette consumption by 0.57 packs per person per month. State IOI was also inversely, albeit not statistically significantly, associated with smoking prevalence (regression coefficient = -0.11;P =.06).Conclusions: The reduction in adult smoking prevalence associated with ASSIST could have translated into approximately 278 700 fewer smokers nationwide if all states had implemented ASSIST. Investment in building state-level tobacco control capacity and promoting changes in tobacco control policies are effective strategies for reducing tobacco use. [ABSTRACT FROM AUTHOR]