Search tips
Search criteria

Results 1-15 (15)

Clipboard (0)

Select a Filter Below

more »
Year of Publication
Document Types
1.  Strategies to Optimize the Impact of Nutritional Surveys and Epidemiological Studies12 
Advances in Nutrition  2013;4(5):545-547.
The development of nutrition and health guidelines and policies requires reliable scientific information. Unfortunately, theoretical considerations and empirical evidence indicate that a large percentage of science-based claims rely on studies that fail to replicate. The session “Strategies to Optimize the Impact of Nutrition Surveys and Epidemiological Studies” focused on the elements of design, interpretation, and communication of nutritional surveys and epidemiological studies to enhance and encourage the production of reliable, objective evidence for use in developing dietary guidance for the public. The speakers called for more transparency of research, raw data, consistent data-staging techniques, and improved data analysis. New approaches to collecting data are urgently needed to increase the credibility and utility of findings from nutrition epidemiological studies. Such studies are critical for furthering our knowledge and understanding of the effects of diet on health.
PMCID: PMC3771144  PMID: 24038252
2.  A mixed-effects model approach for estimating the distribution of usual intake of nutrients: The NCI method 
Statistics in medicine  2010;29(27):10.1002/sim.4063.
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.
PMCID: PMC3865776  PMID: 20862656
statistical distributions; diet surveys; nutrition assessment; mixed-effects model; nutrients; percentiles
3.  Income and race/ethnicity are associated with adherence to food-based dietary guidance among U.S. adults and children 
Income and race/ethnicity are associated with differences in dietary intakes that may contribute to health disparities among the United States population.
To examine alignment of intakes of food groups and energy from solid fats, added sugars, and alcohol with the 2005 Dietary Guidelines and MyPyramid, by family income and race/ethnicity.
Data from the National Health and Nutrition Examination Survey (NHANES), a cross-sectional nationally-representative survey, for 2001-2004.
Persons aged 2 years and older for whom reliable dietary intake data were available (n=16,338) were categorized by income (lowest, middle, and highest) and race/ethnicity (non-Hispanic White (NHW), non-Hispanic Black (NHB), and Mexican-American (MA)).
Statistical analyses performed
The National Cancer Institute (NCI) method was used to estimate the proportions of adults and children in each income and race/ethnic group whose usual intakes met the recommendations.
Higher income was associated with greater adherence to recommendations for most food groups; the proportions meeting minimum recommendations among adults in the highest income group were double that observed for the lowest income group for total vegetables, milk, and oils. Fewer differences by income were apparent among children. Among the race/ethnic groups, the proportions meeting recommendations were generally lowest among NHB. Marked differences were observed for milk—15% of NHB children met the minimum recommendations compared to 42% of NHW children and 35% of MA children; a similar pattern was evident for adults. One in five Mexican American adults met the dry beans and peas recommendations compared to approximately 2% of NHW and NHB. Most adults and children consumed excess energy from solid fats and added sugars irrespective of income and race/ethnicity.
The diets of some subpopulations, particularly individuals in lower-income households and NHB, are especially poor in relation to dietary recommendations, supporting the need for comprehensive strategies to enable healthier dietary intake patterns.
PMCID: PMC3775640  PMID: 22709767
dietary guidance; income; race/ethnicity; disparities; usual dietary intakes; 24-hour recall; NHANES
5.  Reactivity and its association with body mass index across days on food checklists 
Characterizing relationships between diet, body weight, and health is complicated by reporting errors in dietary intake data that are associated with body weight. The objectives of this study were to assess changes in reporting across days (reactivity) on food checklists and associations between reactivity and body mass index (BMI) using data from two cross-sectional studies: 1) the Recontacting Participants in the Observing Protein and Energy Nutrition study (n = 297), which was conducted in 2003–2004 and included a 7-day checklist and a 4-day food record (FR), and 2) the America’s Menu Daily Food Report Study (n=530), which was conducted in 1996 and included a 30-day checklist. Zero-inflated Poisson regression was used to assess effects of reporting day on frequency of consumption for the checklists and number of items reported for the FR. Interactions between day and BMI were tested using contrast statements. Frequency of reported consumption declined across days among males and females for total items and many of the eight food groups on the 7-day checklist; among females, the effect of reporting day differed by BMI category for the meat, fish, and poultry group. Smaller declines across days were observed for some of the 22 food groups on the 30-day checklist; no interactions with BMI were apparent. No reporting day effects were observed in the FR data. The results suggest inconsistent reactivity across days, possibly reflecting changes in reporting or consumption behavior. However, the effects are generally small and independent of body weight, suggesting that checklists are potentially useful for the study of body weight and diet.
PMCID: PMC3269781  PMID: 22308230
dietary assessment; food checklist; measurement error; reactivity; body mass index; obesity
7.  No effect of meat, meat cooking preferences, meat mutagens or heme iron on lung cancer risk in the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial 
Recent epidemiological studies have suggested that red and processed meat may increase the risk of lung cancer. Possible underlying mechanisms include mutagens produced during high temperature cooking or preservation, or formed endogenously from heme iron in meat. We used data from 99,579 participants of both screened and non-screened arms of the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO), aged 55–74 years, to investigate whether meat type, cooking method, doneness level, intake of specific meat mutagens 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (MeIQx), 2-amino-3,4,8-trimethylimidazo[4,5-f]quinoxaline] (DiMeIQx), 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP), and benzo(a)pyrene (B(a)P)] and heme iron are associated with lung cancer. Participants’ diet was assessed prospectively using a 124-item food frequency questionnaire and an additional meat-cooking module. Dietary data were used in conjunction with a database to estimate intake of MeIQx, DiMeIQx, PhIP, B(a)P and heme iron. After up to 8 years of follow-up, 782 incident lung cancer cases were ascertained. Lung cancer risk was not associated with the consumption of either red (men: HRQ5vs.Q1 = 1.11, 95% CI = 0.79–1.56, Ptrend = 0.42; women: HRQ5vs.Q1 = 1.30, 95% CI = 0.87–1.95, Ptrend = 0.65) or processed meat (men: HRQ5vs.Q1 = 1.12, 95% CI = 0.83–1.53, Ptrend = 0.22; women: HRQ5vs.Q1 = 0.98, 95% CI = 0.68–1.41, Ptrend = 0.32) in multivariable models. High temperature cooking methods, level of meat doneness, meat mutagens and heme iron had no effect on lung cancer risk. In this population, we found no association between meat type, cooking method, doneness level, or intake of specific meat mutagens or heme iron and lung cancer risk.
PMCID: PMC2970721  PMID: 20232386
Meat; diet; lung cancer; meat mutagens; heme iron; PLCO
8.  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.
PMCID: PMC3406506  PMID: 22848190
Bayesian approach; latent variables; measurement error; mixed effects models; nutritional epidemiology; zero-inflated data
9.  Physical Activity in US Youth: Impact of Race/Ethnicity, Age, Gender, & Weight Status 
To describe physical activity (PA) levels by race/ethnicity, age, gender, and weight status in a representative sample of U.S. youth.
Cross-sectional data from the 2003–4 and 2005–6 National Health and Nutrition Examination Survey (NHANES) were combined and analyzed. Youth ages 6 to 19 with at least 4 10-hour days of PA measured by accelerometry were included (N=3,106). Outcomes included mean counts per minute and minutes spent in moderate to vigorous PA (MVPA).
6 to 11 year olds spent more time (88 min/day) in MVPA than 12 to 15 (33 min/day) and 16–19 (26 min/day) year olds (p<.001 for both). Females spent fewer min/day in MVPA than males (p<.001). Overall, obese youth spent 16 fewer min/day in MVPA than normal weight youth. However, non-Hispanic White (NHW) males spent 3–4 fewer min/day in vigorous PA than Mexican American (MA) (p=.004) and non-Hispanic Black (NHB) (p<.001) males but had lower obesity rates; and obese 12 to 15 year old MA recorded similar minutes in MVPA per day as normal weight MA (p>.050). There was a significant 3-way age-BMI-race/ethnicity interaction for mean min/day in MVPA (p<.001). Adjustment for total energy intake did not qualitatively alter these results.
Females and older youth were the least active groups. Obese youth were generally less active, but this did not hold uniformly across race/ethnic groups. Cultural or biological factors could moderate the association between PA and obesity in youth.
PMCID: PMC3242154  PMID: 21084930
NHANES; Moderate To Vigorous; Accelerometer; BMI; Adolescent
The annals of applied statistics  2011;5(2B):1456-1487.
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.
PMCID: PMC3145332  PMID: 21804910
Bayesian methods; Dietary assessment; Latent variables; Measurement error; Mixed models; Nutritional epidemiology; Nutritional surveillance; Zero-Inflated Data
11.  Available Carbohydrates, Glycemic Load, and Pancreatic Cancer: Is There a Link? 
American Journal of Epidemiology  2010;171(11):1174-1182.
High-carbohydrate diets have been linked to pancreatic cancer risk in case-control studies, but prospective studies have shown mostly null results. The authors investigated the associations of glycemic load, glycemic index, and carbohydrate intake with pancreatic cancer risk in the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial. Dietary intake was assessed by using a self-administered questionnaire. Between 1998 and 2006 (median follow-up = 6.5 years), 266 incident, confirmed pancreatic cancers were identified among 109,175 participants. Hazards ratios and 95% confidence intervals were adjusted for sex, smoking, body mass index, and total energy. Overall, elevated risks for pancreatic cancer were observed in the 90th versus 10th percentile of glycemic load (hazards ratio (HR) = 1.45, 95% confidence interval (CI): 1.05, 2.00), available carbohydrate (HR = 1.47, 95% CI: 1.05, 2.06), and sucrose (HR = 1.37, 95% CI: 0.99, 1.89) intake. The positive association for available carbohydrate intake was observed during the first 4 years of follow-up (HR<2 years = 2.60, 95% CI: 1.34, 5.06; HR2–<4 years = 1.94, 95% CI: 1.06, 3.55) but not subsequently (HR = 0.86, 95% CI: 0.52, 1.44); the opposite pattern was observed for total fat and saturated fat intake. Rather than being causal, the short-term increase in pancreatic cancer risk associated with high available carbohydrate and low fat intake may be capturing dietary changes associated with subclinical disease.
PMCID: PMC2915491  PMID: 20452999
diet; dietary carbohydrates; dietary fats; glycemic index; pancreatic neoplasms; prospective studies
12.  Differences in fruit and vegetable intake among Hispanic subgroups in California - Results from the 2005 California Health Interview Survey 
To compare total fruit and vegetable intake in cup equivalents (FVC) and its individual components among Hispanic subgroups in California.
Data are from the adult portion of the 2005 California Health Interview Survey. Hispanic/Latino subjects (n = 7954) were grouped into six subcategories (Mexican, Central American, Caribbean, Spanish American, South American, > 1 group). FVC was estimated from frequency responses about seven food categories. T-test and chi-squares were estimated to assess differences in sociodemographic characteristics across Hispanic subgroups. Multivariate linear regressions using SUDAAN were conducted to obtain means of FVC and its components by Hispanic subgroups controlling for confounders.
Hispanic subgroups did not differ in their intake of total FVC (mean cups = 3.4 and 2.9 for men and women, respectively). Small but significant differences (p<0.01) were found across Hispanic subgroups in individual FVC components (green salad (women only), cooked dried beans and non-fried white potatoes) after adjusting for potential sociodemographic and acculturation confounders.
Hispanic FVC intake did not meet the national recommendation, although their reported intake is higher compared to other race/ethnicity groups. The public health message remains the same: to increase FVC. Examination of intake for subgroups of Hispanics may enhance the utility of dietary information for surveillance, program and message design, and intervention and evaluation.
PMCID: PMC2823482  PMID: 19857629
13.  Modeling Data with Excess Zeros and Measurement Error: Application to Evaluating Relationships between Episodically Consumed Foods and Health Outcomes 
Biometrics  2009;65(4):1003-1010.
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.
PMCID: PMC2881223  PMID: 19302405
Dietary measurement error; Dietary survey; Episodically consumed foods; Excess zero models; Food frequency questionnaire; Fish; Individual usual intake; Mercury; Nonlinear mixed models; Regression calibration; 24-hour recall
14.  Race/ethnic Variation in Serum Levels of IGF-I and IGFBP-3 in US Adults 
The IGF axis plays a significant role in normal growth and development and variation in IGFs is associated with health outcomes. Past studies report variation in IGF levels among race/ethnic groups known to differ in disease incidence. This paper reports on race/ethnic variation in serum levels of IGF-I and IGF-BP3 in a nationally representative and ethnically diverse sample of US adults.
Serum IGF-I and IGFBP-3 levels from the fasting subsample (n = 6061) of respondents to the US National Health and Nutrition Examination Survey III (NHANES III) were analyzed using an IGF-I ELISA (Diagnostic Systems Laboratory (DSL) 10–5600) and an IGFBP-3 IRMA (DSL 6600). The NHANES is a combined examination and interview survey of a nationally representative sample of US adults. Regression analyses were used to estimate cross-sectional associations between the IGF axis and demographic variables.
In unadjusted analyses, serum IGF-I levels were higher in males than in females, and IGFBP-3 levels were higher in females than in males. Both analytes were lower in older adults. Univariate analyses indicate that serum levels of IGF-I are lower in female Non-Hispanic Whites (NHW) (256 [4.9]) and Hispanics (249 [6.6]) than in Non-Hispanic Blacks (NHB) (281 [4.9]). However, in males, IGF levels in NHWs (287 [3.6]) and NHBs (284 [4.3]) are similar and levels in Mexican-Americans are only moderately reduced (265 [3.4]). Notably, NHB’s have the highest molar ratio of IGF-I:IGFBP-3 at all ages. After adjustment for age and BMI, gender and race/ethnicity differences persist.
These cross-sectional data support exploration of the IGF axis as an explanation for some race/ethnic differences in cancer incidence.
PMCID: PMC2702997  PMID: 18812263
Cancer; Insulin-Like Growth Factor; Race/Ethnicity; Age
15.  A new method for estimating the usual intake of episodically-consumed foods with application to their distribution 
We propose a new statistical method that uses information from two 24-hour recalls (24HRs) to estimate usual intake of episodically-consumed foods.
Statistical Analyses Performed
The method developed at the National Cancer Institute (NCI) accommodates the large number of non-consumption days that arise with foods by separating the probability of consumption from the consumption-day amount, using a two-part model. Covariates, such as sex, age, race, or information from a food frequency questionnaire (FFQ), may supplement the information from two or more 24HRs using correlated mixed model regression. The model allows for correlation between the probability of consuming a food on a single day and the consumption-day amount. Percentiles of the distribution of usual intake are computed from the estimated model parameters.
The Eating at America's Table Study (EATS) data are used to illustrate the method to estimate the distribution of usual intake for whole grains and dark green vegetables for men and women and the distribution of usual intakes of whole grains by educational level among men. A simulation study indicates that the NCI method leads to substantial improvement over existing methods for estimating the distribution of usual intake of foods.
The NCI method provides distinct advantages over previously proposed methods by accounting for the correlation between probability of consumption and amount consumed and by incorporating covariate information. Researchers interested in estimating the distribution of usual intakes of foods for a population or subpopulation are advised to work with a statistician and incorporate the NCI method in analyses.
PMCID: PMC2517157  PMID: 17000190
Usual intake; Episodically-consumed foods; statistical methods

Results 1-15 (15)