Association of depression and self-reported health status by birthplace and citizenship status: Results from the 2010-2018 National Health Interview Survey.

BACKGROUND: Self-reported health (SRH) is an important indicator of mental health outcomes. More information, however, is needed on whether this association varies by birthplace (defined as US-born or non-US-born) and citizenship status (i.e., non-US-born citizen, non-US citizen, and US-born citizen). METHODS: We examined the associations between SRH and depression among non-US-born US citizens, non-US citizens, and US-born citizens aged 18 years and older using weighted cross-sectional data from the 2010-2018 National Health Interview Survey (n = 139,884). Logistic regression models were used to assess the association between depression and SRH by citizenship status, adjusting for covariates. RESULTS: US-born citizens reported the highest prevalence of depression (40.3 %), and non-US-born citizens reported the highest prevalence of poor/fair SRH (14.5 %). Individuals with fair/poor SRH had a significantly increased likelihood of depression relative to those with good/very good/excellent for non-US-born US citizens (Adjusted Odds Ratio [AOR] = 2.39, 95 % Confidence Interval [95 % CI] = 2.02-2.84), non-US citizens (AOR = 2.77, 95 % CI = 2.28-3.36), and US-born citizens (AOR = 2.31, CI = 2.18-2.44). LIMITATIONS: The study is cross-sectional, reducing the strength of determining causal relationships. Also, there is a possible response bias due to the self-reported nature of the data. CONCLUSIONS: Our study indicates that fair/poor SRH is significantly associated with an increased likelihood of depression regardless of an individual citizenship status. Additionally, immigrants with fair/poor SRH had higher increased odds of depression. Therefore, mental healthcare interventions tailored for immigrants can reduce mental health problems and disparities among immigrants.

Measurement error models with zero inflation and multiple sources of zeros, with applications to hard zeros.

We consider measurement error models for two variables observed repeatedly and subject to measurement error. One variable is continuous, while the other variable is a mixture of continuous and zero measurements. This second variable has two sources of zeros. The first source is episodic zeros, wherein some of the measurements for an individual may be zero and others positive. The second source is hard zeros, i.e., some individuals will always report zero. An example is the consumption of alcohol from alcoholic beverages: some individuals consume alcoholic beverages episodically, while others never consume alcoholic beverages. However, with a small number of repeat measurements from individuals, it is not possible to determine those who are episodic zeros and those who are hard zeros. We develop a new measurement error model for this problem, and use Bayesian methods to fit it. Simulations and data analyses are used to illustrate our methods. Extensions to parametric models and survival analysis are discussed briefly.

Parent-Child Nativity, Race, Ethnicity, and Common Mental Health Conditions Among United States Children and Adolescents.

OBJECTIVE: To examine associations between race, ethnicity, and parent-child nativity, and common mental health conditions among U.S. children and adolescents. METHODS: Data were from 2016 to 2019 National Survey of Children's Health, a US population-based, serial cross-sectional survey, and restricted to children who had access to health care. We used weighted multivariable logistic regression to examine the associations between race and ethnicity (Asian, Black, Hispanic, White, Other-race); mental health outcomes (depression, anxiety, and behavior/conduct problems) stratified by household generation; and between household generation and outcomes stratified by race and ethnicity, adjusting for demographics (age, sex, family income to poverty ratio, parental education), and an adverse childhood experience (ACE) score. RESULTS: When stratifying by household generation, racial and ethnic minority children generally had similar to lower odds of outcomes compared with White children, with the exception of higher odds of behavior/conduct problems among third + -generation Black children. When stratifying by race and ethnicity, third + generation children had increased odds of depression compared to their first-generation counterparts. Third + generation, racial and ethnic minority children had increased odds of anxiety and behavior/conduct problems compared with their first-generation counterparts. The associations generally remained significant after adjusting for the ACE score. CONCLUSIONS: Lower odds of common mental health conditions in racial and ethnic minority children could be due to factors such as differential reporting, and higher estimates, including those in third + generation children, could be due to factors including discrimination; systemic racism; and other factors that vary by generation and need further investigation to advance health equity.

Data-driven choice of a model selection method in joinpoint regression.

Selecting the number of change points in segmented line regression is an important problem in trend analysis, and there have been various approaches proposed in the literature. We first study the empirical properties of several model selection procedures and propose a new method based on two Schwarz type criteria, a classical Bayes Information Criterion (BIC) and the one with a harsher penalty than BIC (BIC3). The proposed rule is designed to use the former when effect sizes are small and the latter when the effect sizes are large and employs the partial R2 to determine the weight between BIC and BIC3. The proposed method is computationally much more efficient than the permutation test procedure that has been the default method of Joinpoint software developed for cancer trend analysis, and its satisfactory performance is observed in our simulation study. Simulations indicate that the proposed method performs well in keeping the probability of correct selection at least as large as that of BIC3, whose performance is comparable to that of the permutation test procedure, and improves BIC3 when it performs worse than BIC. The proposed method is applied to the U.S. prostate cancer incidence and mortality rates.

Parent-Child Nativity, Race, Ethnicity, and Adverse Childhood Experiences Among U.S. Children.

We examined the relationship between parent-child nativity, race/ethnicity, and adverse childhood experiences (ACEs) among US children. We found that a high proportion of Black/Hispanic children experienced ACEs, with variation by generation for Hispanic but not Black children, and a lower but increasing proportion of White children experienced ACEs by generation.

Characterizing Trends in Cancer Patients' Survival Using the JPSurv Software.

BACKGROUND: Improvements in cancer survival are usually assessed by comparing survival in grouped years of diagnosis. To enhance analyses of survival trends, we present the joinpoint survival model webtool (JPSurv) that analyzes survival data by single year of diagnosis and estimates changes in survival trends and year-over-year trend measures. METHODS: We apply JPSurv to relative survival data for individuals diagnosed with female breast cancer, melanoma cancer, non-Hodgkin lymphoma (NHL), and chronic myeloid leukemia (CML) between 1975 and 2015 in the Surveillance, Epidemiology, and End Results Program. We estimate the number and location of joinpoints and the trend measures and provide interpretation. RESULTS: In general, relative survival has substantially improved at least since the mid-1990s for all cancer sites. The largest improvements in 5-year relative survival were observed for distant-stage melanoma after 2009, which increased by almost 3 survival percentage points for each subsequent year of diagnosis, followed by CML in 1995-2010, and NHL in 1995-2003. The modeling also showed that for patients diagnosed with CML after 1995 (compared with before), there was a greater decrease in the probability of dying of the disease in the 4th and 5th years after diagnosis compared with the initial years since diagnosis. CONCLUSIONS: The greatest increases in trends for distant melanoma, NHL, and CML coincided with the introduction of novel treatments, demonstrating the value of JPSurv for estimating and interpreting cancer survival trends. IMPACT: The JPSurv webtool provides a suite of estimates for analyzing trends in cancer survival that complement traditional descriptive survival analyses.

Early childhood caries and intake of 100 percent fruit juice: Data from NHANES, 1999-2004.

BACKGROUND: The results of several studies conducted in the United States show no association between intake of 100 percent fruit juice and early childhood caries (ECC). The authors examined this association according to poverty and race/ethnicity among U.S. preschool children. METHODS: The authors analyzed data from the 1999-2004 National Health and Nutrition Examination Survey (NHANES) for 2,290 children aged 2 through 5 years. They used logistic models for caries (yes or no) to assess the association between caries and intake of 100 percent fruit juice, defined as consumption (yes or no), ounces (categories) consumed in the previous 24 hours or usual intake (by means of a statistical method from the National Cancer Institute). RESULTS: The association between caries and consumption of 100 percent fruit juice (yes or no) was not statistically significant in an unadjusted logistic model (odds ratio [OR], 0.76; 95 percent confidence interval [CI], 0.57-1.01), and it remained nonsignificant after covariate adjustment (OR, 0.89; 95 percent CI, 0.63-1.24). Similarly, models in which we evaluated categorical consumption of 100 percent juice (that is, 0 oz; > 0 and ≤ 6 oz; and > 6 oz), unadjusted and adjusted by covariates, did not indicate an association with ECC. CONCLUSIONS: Our study findings are consistent with those of other studies that show consumption of 100 percent fruit juice is not associated with ECC.

Non-hodgkin lymphoma risk and insecticide, fungicide and fumigant use in the agricultural health study.

Farming and pesticide use have previously been linked to non-Hodgkin lymphoma (NHL), chronic lymphocytic leukemia (CLL) and multiple myeloma (MM). We evaluated agricultural use of specific insecticides, fungicides, and fumigants and risk of NHL and NHL-subtypes (including CLL and MM) in a U.S.-based prospective cohort of farmers and commercial pesticide applicators. A total of 523 cases occurred among 54,306 pesticide applicators from enrollment (1993-97) through December 31, 2011 in Iowa, and December 31, 2010 in North Carolina. Information on pesticide use, other agricultural exposures and other factors was obtained from questionnaires at enrollment and at follow-up approximately five years later (1999-2005). Information from questionnaires, monitoring, and the literature were used to create lifetime-days and intensity-weighted lifetime days of pesticide use, taking into account exposure-modifying factors. Poisson and polytomous models were used to calculate relative risks (RR) and 95% confidence intervals (CI) to evaluate associations between 26 pesticides and NHL and five NHL-subtypes, while adjusting for potential confounding factors. For total NHL, statistically significant positive exposure-response trends were seen with lindane and DDT. Terbufos was associated with total NHL in ever/never comparisons only. In subtype analyses, terbufos and DDT were associated with small cell lymphoma/chronic lymphocytic leukemia/marginal cell lymphoma, lindane and diazinon with follicular lymphoma, and permethrin with MM. However, tests of homogeneity did not show significant differences in exposure-response among NHL-subtypes for any pesticide. Because 26 pesticides were evaluated for their association with NHL and its subtypes, some chance finding could have occurred. Our results showed pesticides from different chemical and functional classes were associated with an excess risk of NHL and NHL subtypes, but not all members of any single class of pesticides were associated with an elevated risk of NHL or NHL subtypes. These findings are among the first to suggest links between DDT, lindane, permethrin, diazinon and terbufos with NHL subtypes.

When performance management works: a study of the National Breast and Cervical Cancer Early Detection Program.

BACKGROUND: Little empirical evidence exists about the effectiveness of performance management systems in government. This study assessed the effectiveness of the performance management system of the National Breast and Cervical Cancer Early Detection Program (NBCCEDP) and explored why it works. METHODS: Generalized estimating equation models were used to assess change in program performance after the implementation of a performance management system. In addition, qualitative case study data including observations, interviews, and document review were analyzed using inductive methods. RESULTS: Five of the 7 indicators tested had statistically significant increases in performance postimplementation. Case study results suggest that the system is characterized by high-quality data, measures viewed by grantees as meaningful and fair, and institutionalized data use. CONCLUSIONS: Several factors help to explain the system's effectiveness including characteristics of the NBCCEDP program (eg, service delivery program), qualities of the indicators (eg, process level), financial investment in the system, and a culture of data use.

The Healthy Eating Index-2010 is a valid and reliable measure of diet quality according to the 2010 Dietary Guidelines for Americans.

The Healthy Eating Index (HEI), a measure of diet quality, was updated to reflect the 2010 Dietary Guidelines for Americans and the accompanying USDA Food Patterns. To assess the validity and reliability of the HEI-2010, exemplary menus were scored and 2 24-h dietary recalls from individuals aged ≥2 y from the 2003-2004 NHANES were used to estimate multivariate usual intake distributions and assess whether the HEI-2010 1) has a distribution wide enough to detect meaningful differences in diet quality among individuals, 2) distinguishes between groups with known differences in diet quality by using t tests, 3) measures diet quality independently of energy intake by using Pearson correlation coefficients, 4) has >1 underlying dimension by using principal components analysis (PCA), and 5) is internally consistent by calculating Cronbach's coefficient α. HEI-2010 scores were at or near the maximum levels for the exemplary menus. The distribution of scores among the population was wide (5th percentile = 31.7; 95th percentile = 70.4). As predicted, men's diet quality (mean HEI-2010 total score = 49.8) was poorer than women's (52.7), younger adults' diet quality (45.4) was poorer than older adults' (56.1), and smokers' diet quality (45.7) was poorer than nonsmokers' (53.3) (P < 0.01). Low correlations with energy were observed for HEI-2010 total and component scores (|r| ≤ 0.21). Cronbach's coefficient α was 0.68, supporting the reliability of the HEI-2010 total score as an indicator of overall diet quality. Nonetheless, PCA indicated multiple underlying dimensions, highlighting the fact that the component scores are equally as important as the total. A comparable reevaluation of the HEI-2005 yielded similar results. This study supports the validity and the reliability of both versions of the HEI.

Systems strategies to support cancer screening in U.S. primary care practice.

BACKGROUND: Although systems strategies are effective in improving health care delivery, little is known about their use for cancer screening in U.S. primary care practice. METHODS: We assessed primary care physicians' (N = 2,475) use of systems strategies for breast, cervical, and colorectal cancer (CRC) screening in a national survey conducted in 2007. Systems strategies included patient and physician screening reminders, performance reports of screening rates, electronic medical records, implementation of in-practice guidelines, and use of nurse practitioners/physician assistants. We evaluated use of both patient and physician screening reminders with other strategies in separate models by screening type, adjusted for the effects of physician and practice characteristics with multivariate logistic regression. RESULTS: Fewer than 10% of physicians used a comprehensive set of systems strategies to support cancer screening; use was greater for mammography and Pap testing than for CRC screening. In adjusted analyses, performance reports of cancer screening rates, medical record type, and in-practice guidelines were associated with use of both patient and physician screening reminders for mammography, Pap testing, and CRC screening (P < 0.05). CONCLUSION: Despite evidence supporting use of systems strategies in primary care, few physicians report using a comprehensive set of strategies to support cancer screening. IMPACT: Current health policy initiatives underscore the importance of increased implementation of systems strategies in primary care to improve the use and quality of cancer screening in the United States.

A NEW MULTIVARIATE MEASUREMENT ERROR MODEL WITH ZERO-INFLATED DIETARY DATA, AND ITS APPLICATION TO DIETARY ASSESSMENT.

In the United States the preferred method of obtaining dietary intake data is the 24-hour dietary recall, yet the measure of most interest is usual or long-term average daily intake, which is impossible to measure. Thus, usual dietary intake is assessed with considerable measurement error. Also, diet represents numerous foods, nutrients and other components, each of which have distinctive attributes. Sometimes, it is useful to examine intake of these components separately, but increasingly nutritionists are interested in exploring them collectively to capture overall dietary patterns. Consumption of these components varies widely: some are consumed daily by almost everyone on every day, while others are episodically consumed so that 24-hour recall data are zero-inflated. In addition, they are often correlated with each other. Finally, it is often preferable to analyze the amount of a dietary component relative to the amount of energy (calories) in a diet because dietary recommendations often vary with energy level. The quest to understand overall dietary patterns of usual intake has to this point reached a standstill. There are no statistical methods or models available to model such complex multivariate data with its measurement error and zero inflation. This paper proposes the first such model, and it proposes the first workable solution to fit such a model. After describing the model, we use survey-weighted MCMC computations to fit the model, with uncertainty estimation coming from balanced repeated replication.The methodology is illustrated through an application to estimating the population distribution of the Healthy Eating Index-2005 (HEI-2005), a multi-component dietary quality index involving ratios of interrelated dietary components to energy, among children aged 2-8 in the United States. We pose a number of interesting questions about the HEI-2005 and provide answers that were not previously within the realm of possibility, and we indicate ways that our approach can be used to answer other questions of importance to nutritional science and public health.

Are physicians' recommendations for colorectal cancer screening guideline-consistent?

BACKGROUND: Many older adults in the U.S. do not receive appropriate colorectal cancer (CRC) screening. Although primary care physicians' recommendations to their patients are central to the screening process, little information is available about their recommendations in relation to guidelines for the menu of CRC screening modalities, including fecal occult blood testing (FOBT), flexible sigmoidoscopy (FS), colonoscopy, and double contrast barium enema (DCBE). The objective of this study was to explore potentially modifiable physician and practice factors associated with guideline-consistent recommendations for the menu of CRC screening modalities. METHODS: We examined data from a nationally representative sample of 1266 physicians in the U.S. surveyed in 2007. The survey included questions about physician and practice characteristics, perceptions about screening, and recommendations for age of initiation and screening interval for FOBT, FS, colonoscopy and DCBE in average risk adults. Physicians' screening recommendations were classified as guideline consistent for all, some, or none of the CRC screening modalities recommended. Analyses used descriptive statistics and polytomous logit regression models. RESULTS: Few (19.1%; 95% CI:16.9%, 21.5%) physicians made guideline-consistent recommendations across all CRC screening modalities that they recommended. In multivariate analysis, younger physician age, board certification, north central geographic region, single specialty or multi-specialty practice type, fewer patients per week, higher number of recommended modalities, use of electronic medical records, greater influence of patient preferences for screening, and published clinical evidence were associated with guideline-consistent screening recommendations (p < 0.05). CONCLUSIONS: Physicians' CRC screening recommendations reflect both overuse and underuse, and few made guideline-consistent CRC screening recommendations across all modalities they recommended. Interventions that focus on potentially modifiable physician and practice factors that influence overuse and underuse and address the menu of recommended screening modalities will be important for improving screening practice.

Fitting a bivariate measurement error model for episodically consumed dietary components.

There has been great public health interest in estimating usual, i.e., long-term average, intake of episodically consumed dietary components that are not consumed daily by everyone, e.g., fish, red meat and whole grains. Short-term measurements of episodically consumed dietary components have zero-inflated skewed distributions. So-called two-part models have been developed for such data in order to correct for measurement error due to within-person variation and to estimate the distribution of usual intake of the dietary component in the univariate case. However, there is arguably much greater public health interest in the usual intake of an episodically consumed dietary component adjusted for energy (caloric) intake, e.g., ounces of whole grains per 1000 kilo-calories, which reflects usual dietary composition and adjusts for different total amounts of caloric intake. Because of this public health interest, it is important to have models to fit such data, and it is important that the model-fitting methods can be applied to all episodically consumed dietary components.We have recently developed a nonlinear mixed effects model (Kipnis, et al., 2010), and have fit it by maximum likelihood using nonlinear mixed effects programs and methodology (the SAS NLMIXED procedure). Maximum likelihood fitting of such a nonlinear mixed model is generally slow because of 3-dimensional adaptive Gaussian quadrature, and there are times when the programs either fail to converge or converge to models with a singular covariance matrix. For these reasons, we develop a Monte-Carlo (MCMC) computation of fitting this model, which allows for both frequentist and Bayesian inference. There are technical challenges to developing this solution because one of the covariance matrices in the model is patterned. Our main application is to the National Institutes of Health (NIH)-AARP Diet and Health Study, where we illustrate our methods for modeling the energy-adjusted usual intake of fish and whole grains. We demonstrate numerically that our methods lead to increased speed of computation, converge to reasonable solutions, and have the flexibility to be used in either a frequentist or a Bayesian manner.

A mixed-effects model approach for estimating the distribution of usual intake of nutrients: the NCI method.

It is of interest to estimate the distribution of usual nutrient intake for a population from repeat 24-h dietary recall assessments. A mixed effects model and quantile estimation procedure, developed at the National Cancer Institute (NCI), may be used for this purpose. The model incorporates a Box-Cox parameter and covariates to estimate usual daily intake of nutrients; model parameters are estimated via quasi-Newton optimization of a likelihood approximated by the adaptive Gaussian quadrature. The parameter estimates are used in a Monte Carlo approach to generate empirical quantiles; standard errors are estimated by bootstrap. The NCI method is illustrated and compared with current estimation methods, including the individual mean and the semi-parametric method developed at the Iowa State University (ISU), using data from a random sample and computer simulations. Both the NCI and ISU methods for nutrients are superior to the distribution of individual means. For simple (no covariate) models, quantile estimates are similar between the NCI and ISU methods. The bootstrap approach used by the NCI method to estimate standard errors of quantiles appears preferable to Taylor linearization. One major advantage of the NCI method is its ability to provide estimates for subpopulations through the incorporation of covariates into the model. The NCI method may be used for estimating the distribution of usual nutrient intake for populations and subpopulations as part of a unified framework of estimation of usual intake of dietary constituents.

Performance of common genetic variants in breast-cancer risk models.

BACKGROUND: Genomewide association studies have identified multiple genetic variants associated with breast cancer. The extent to which these variants add to existing risk-assessment models is unknown. METHODS: We used information on traditional risk factors and 10 common genetic variants associated with breast cancer in 5590 case subjects and 5998 control subjects, 50 to 79 years of age, from four U.S. cohort studies and one case-control study from Poland to fit models of the absolute risk of breast cancer. With the use of receiver-operating-characteristic curve analysis, we calculated the area under the curve (AUC) as a measure of discrimination. By definition, random classification of case and control subjects provides an AUC of 50%; perfect classification provides an AUC of 100%. We calculated the fraction of case subjects in quintiles of estimated absolute risk after the addition of genetic variants to the traditional risk model. RESULTS: The AUC for a risk model with age, study and entry year, and four traditional risk factors was 58.0%; with the addition of 10 genetic variants, the AUC was 61.8%. About half the case subjects (47.2%) were in the same quintile of risk as in a model without genetic variants; 32.5% were in a higher quintile, and 20.4% were in a lower quintile. CONCLUSIONS: The inclusion of newly discovered genetic factors modestly improved the performance of risk models for breast cancer. The level of predicted breast-cancer risk among most women changed little after the addition of currently available genetic information.

Influence of loop electrosurgical excision procedure on subsequent acquisition of new human papillomavirus infections.

BACKGROUND: The impact of loop electrosurgical excision procedure (LEEP) treatment for cervical precancerous lesions on subsequent acquisition of new human papillomavirus (HPV) infections is not well described. METHODS: The acquisition of new HPV infections was compared in HPV-positive women who underwent colposcopy and were treated by LEEP (n = 195) and those who were untreated (n = 1625) at entry into a 2-year study. Cumulative incidence rate ratios (IRRs) for treated versus untreated women at 6- and 24-months of follow-up, with 95% confidence intervals (CIs), were calculated for infection by individual HPV genotypes, any HPV genotypes, any carcinogenic HPV genotypes, any noncarcinogenic HPV genotypes, and phylogenetic groups of HPV genotypes. RESULTS: Treated women were 29% less likely than untreated women to have carcinogenic HPV genotypes detected at 6-month follow-up visits (IRR, 0.71; 95% CI, 0.50-1.00) and were 18% less likely to have these genotypes detected at 24-month follow-up visits (IRR, 0.82; 95% CI, 0.68-1.01). Treated women were 56% less likely to have HPV genotypes of the alpha9 phylogenetic species (which includes HPV-16) detected at 6-month follow-up visits (IRR, 0.44; 95% CI, 0.23-0.85) and were 40% less likely to have these genotype detected at 24-month follow-up visits (IRR, 0.60; 95% CI, 0.42-0.85). CONCLUSION: LEEP may reduce the acquisition of certain carcinogenic HPV genotypes related to HPV-16.

Modeling data with excess zeros and measurement error: application to evaluating relationships between episodically consumed foods and health outcomes.

Dietary assessment of episodically consumed foods gives rise to nonnegative data that have excess zeros and measurement error. Tooze et al. (2006, Journal of the American Dietetic Association 106, 1575-1587) describe a general statistical approach (National Cancer Institute method) for modeling such food intakes reported on two or more 24-hour recalls (24HRs) and demonstrate its use to estimate the distribution of the food's usual intake in the general population. In this article, we propose an extension of this method to predict individual usual intake of such foods and to evaluate the relationships of usual intakes with health outcomes. Following the regression calibration approach for measurement error correction, individual usual intake is generally predicted as the conditional mean intake given 24HR-reported intake and other covariates in the health model. One feature of the proposed method is that additional covariates potentially related to usual intake may be used to increase the precision of estimates of usual intake and of diet-health outcome associations. Applying the method to data from the Eating at America's Table Study, we quantify the increased precision obtained from including reported frequency of intake on a food frequency questionnaire (FFQ) as a covariate in the calibration model. We then demonstrate the method in evaluating the linear relationship between log blood mercury levels and fish intake in women by using data from the National Health and Nutrition Examination Survey, and show increased precision when including the FFQ information. Finally, we present simulation results evaluating the performance of the proposed method in this context.

Prospective study of serum retinol, beta-carotene, beta-cryptoxanthin, and lutein/zeaxanthin and esophageal and gastric cancers in China.

OBJECTIVE: This study examined the relationship between pretrial serum concentrations of retinol, beta-carotene, beta-cryptoxanthin, and lutein/zeaxanthin and the subsequent risk of developing esophageal squamous cell carcinoma and gastric cardia or non-cardia adenocarcinoma in subjects selected from a randomized nutritional intervention trial in Linxian, China, a region with epidemic rates of esophageal and gastric cardia cancer. METHODS: We used a stratified case-cohort design to select cohort members for inclusion in this study. In all we measured serum concentrations of the above vitamins in 590 esophageal, 395 gastric cardia, and 87 gastric non-cardia case subjects as well as in 1053 control subjects. Relative risks (RRs) were estimated using Cox proportional hazards models. RESULTS: Median values in our cohort were low for serum retinol (33.6 microg/dl), beta-carotene (4.3 microg/dl), and beta-cryptoxanthin (3.5 microg/dl), but were high for lutein/zeaxanthin (40.0 microg/dl). Gastric cardia cancer incidence fell 10% for each quartile increase in serum retinol (RR = 0.90, 95% CI = 0.83-0.99). For esophageal cancer, an inverse association with retinol levels was found only in male non-smokers (RR = 0.79 per quartile increase, 95% CI = 0.63-0.99). For gastric non-cardia cancer, an inverse association was limited to subjects 50 years old or younger (RR = 0.58 per quartile, 95% CI = 0.31-0.96). For beta-cryptoxanthin there was a borderline significant protective association for gastric non-cardia cancer (RR = 0.88 per quartile, 95% CI = 0.76-1.0). In contrast, we found the incidence of gastric non-cardia cancer increased (RR = 1.2 per quartile, 95% CI = 1.0-1.3) with increasing concentration of serum lutein/zeaxanthin. CONCLUSIONS: In this population, we found that low retinol and high lutein/zeaxanthin concentrations increased the risks of gastric cardia and gastric non-cardia cancer respectively. We found that there were no strong associations between any of the other analytes and any of the cancer sites.

Diet and sex hormones in girls: findings from a randomized controlled clinical trial.

BACKGROUND: Results of several studies have suggested that diet during adolescence may influence the risk of breast cancer in adulthood. We evaluated whether an intervention to lower fat intake among adolescent girls altered their serum concentrations of sex hormones that, in adults, are related to breast cancer development. METHODS: We conducted an ancillary hormone study among 286 of the 301 girls who participated between 1988 and 1997 in the Dietary Intervention Study in Children, in which healthy, prepubertal, 8- to 10-year-olds with elevated low-density lipoprotein cholesterol were randomly assigned to usual care or to a behavioral intervention that promoted a low-fat diet. Median time on the intervention was 7 years. Blood samples collected before randomization and at the year 1, year 3, year 5, and last visits were assayed to determine the girls' serum levels of sex hormones. All P values are two-sided. RESULTS: At the year 5 visit, girls in the intervention group had 29.8% (95% confidence interval [CI] = 5.4% to 47.9%; P =.02) lower estradiol, 30.2% (95% CI = 7.0% to 47.7%; P =.02) lower non-sex hormone binding globulin-bound estradiol, 20.7% (95% CI = 4.7% to 34.0%; P =.02) lower estrone, and 28.7% (95% CI = 5.1% to 46.5%; P =.02) lower estrone sulfate levels during the follicular phase of the menstrual cycle and 27.2% (95% CI = 5.7% to 53.1%; P =.01) higher testosterone levels during the luteal phase of the menstrual cycle than did girls in the usual care group. At the last visit, the luteal phase progesterone level was 52.9% (95% CI = 20.0% to 72.3%) lower for girls in the intervention group than for girls in the usual care group (P =.007). CONCLUSION: Modest reductions in fat intake during puberty are associated with changes in sex hormone concentrations that are consistent with alterations in the function of the hypothalamic-pituitary-ovarian axis. Whether these changes influence breast cancer risk is currently unknown.