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.

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.

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.

Americans do not meet federal dietary recommendations

A longstanding goal of dietary surveillance has been to estimate the proportion of the population with intakes above or below a target, such as a recommended level of intake. However, until now, statistical methods for assessing the alignment of food intakes with recommendations have been lacking. The purposes of this study were to demonstrate the National Cancer Institute's method of estimating the distribution of usual intake of foods and determine the proportion of the U.S. population who does not meet federal dietary recommendations. Data were obtained from the 2001-2004 NHANES for 16,338 persons, aged 2 y and older. Quantities of foods reported on 24-h recalls were translated into amounts of various food groups using the MyPyramid Equivalents Database. Usual dietary intake distributions were modeled, accounting for sequence effect, weekend/weekday effect, sex, age, poverty income ratio, and race/ethnicity. The majority of the population did not meet recommendations for all of the nutrient-rich food groups, except total grains and meat and beans. Concomitantly, overconsumption of energy from solid fats, added sugars, and alcoholic beverages ("empty calories") was ubiquitous. Over 80% of persons age ≥ 71 y and over 90% of all other sex-age groups had intakes of empty calories that exceeded the discretionary calorie allowances. In conclusion, nearly the entire U.S. population consumes a diet that is not on par with recommendations. These findings add another piece to the rather disturbing picture that is emerging of a nation's diet in crisis.

A population's distribution of Healthy Eating Index-2005 component scores can be estimated when more than one 24-hour recall is available

The USDA's Healthy Eating Index-2005 (HEI-2005) is a tool to quantify the quality of diet consumed by individuals in the U.S. It comprises 12 components expressed as ratios of a food group or nutrient intake to energy intake. Components are scored on a scale from 0 to M, where M is 5, 10, or 20. Ideally, the HEI-2005 is calculated on the basis of the usual, or long-term average, dietary intake of an individual. In recent cycles of the NHANES, intake data have been collected via 24-h recalls for more than 1 d on most participants. We present here a statistical method to estimate a population's distribution of usual HEI-2005 component scores when >or=2 d of dietary information is available for a sample of individuals from the population. Distributions for the total population and for age-gender subgroups may be estimated. The method also yields an estimate of the population's mean total HEI-2005 score. Application of the method to NHANES data for 2001-2004 yielded estimated distributions for all 12 components; those of total vegetables (range 0-5), whole grains (range 0-5), and energy from solid fats, alcoholic beverages, and added sugars (range 0-20) are presented. The total population mean scores for these components were 3.21, 1.00, and 8.41, respectively. An estimated 30% of the total population had a score of <2.5 for total vegetables. This is the first time, to our knowledge, that estimated distributions of usual HEI-2005 component scores have been published.

Evaluating the food environment: application of the Healthy Eating Index-2005

BACKGROUND

The Healthy Eating Index-2005 (HEI-2005), a tool designed to evaluate concordance with the 2005 Dietary Guidelines, has been used to monitor the quality of foods consumed by Americans. Because the HEI-2005 is not tied to individual requirements and is scored on a per 1000 kcal basis, it can be used to assess the overall quality of any mix of foods.

PURPOSE

The goal of this paper is to examine whether the HEI-2005 can be applied to the food environment.

METHODS

Two examples were selected to examine the application of the HEI-2005 to the food environment: the dollar menu displayed at a fast-food restaurant (coded and linked to the MyPyramid Equivalents Database and the Food and Nutrient Database for Dietary Studies) to represent the community level and the 2005 U.S. Food Supply (measured with food availability data, loss-adjusted food availability data, nutrient availability data, and Salt Institute data) to represent the macro level.

RESULTS

The dollar menu and the 2005 U.S. Food Supply received 43.4 and 54.9 points, respectively (100 possible points). According to the HEI-2005, for the offerings at a local fast-food restaurant and the U.S. Food Supply to align with national dietary guidance, substantial shifts would be necessary: a concomitant addition of fruit, dark-green vegetables, orange vegetables, legumes, and nonfat milk; replacement of refined grains with whole grains; and reduction in foods and food products containing sodium, solid fats, and added sugars.

CONCLUSIONS

Because the HEI-2005 can be applied to both environmental- and individual-level data, it provides a useful metric for studies linking data across various levels of the socioecologic framework of dietary behavior. The present findings suggest that new dietary guidance could target not only individuals but also the architects of our food environment.

Healthfulness of the U.S. food supply: little improvement despite decades of dietary guidance

BACKGROUND

Every 5 years for the past several decades, the USDHHS and the U.S. Department of Agriculture have issued and updated the Dietary Guidelines for Americans, which form the basis of federal nutrition policy and have shown remarkable consistency across various editions among the major themes.

PURPOSE

This paper examines whether the U.S. food supply is sufficiently balanced to provide the recommended proportions of various foods and nutrients per the amount of energy, whether this balance has shifted over time, and which areas of the food supply may have changed more than others.

METHODS

The Healthy Eating Index-2005 (HEI-2005) was used to measure the dietary quality of the U.S. food supply, from 1970 to 2007. Sources of data were the USDA's food availability data, loss-adjusted food availability data, and nutrient availability data, and the U.S. Salt Institute's data on salt sold for human consumption.

RESULTS

Total HEI-2005 scores improved by about 10 points between 1970 and 2007, but they never achieved even 60 points on a scale from 0 to 100. Although meats and total grains were supplied generally in recommended proportions, total vegetables, total fruit, whole fruit, and milk were supplied in suboptimal proportions that changed very little over time. Saturated fat, sodium, and calories from solid fat, alcoholic beverages, and added sugars were supplied in varying degrees of unhealthy abundance over the years. Supplies of dark-green/orange vegetables and legumes and whole grains were entirely insufficient relative to recommendations, with virtually no change over time.

CONCLUSIONS

Deliberate efforts on the part of policymakers, the agriculture sector, and the food industry are necessary to provide a supply of foods consistent with nutrition recommendations and to make healthy choices available to all.

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.

Comparing 3 dietary pattern methods--cluster analysis, factor analysis, and index analysis--With colorectal cancer risk: The NIH-AARP Diet and Health Study

The authors compared dietary pattern methods-cluster analysis, factor analysis, and index analysis-with colorectal cancer risk in the National Institutes of Health (NIH)-AARP Diet and Health Study (n = 492,306). Data from a 124-item food frequency questionnaire (1995-1996) were used to identify 4 clusters for men (3 clusters for women), 3 factors, and 4 indexes. Comparisons were made with adjusted relative risks and 95% confidence intervals, distributions of individuals in clusters by quintile of factor and index scores, and health behavior characteristics. During 5 years of follow-up through 2000, 3,110 colorectal cancer cases were ascertained. In men, the vegetables and fruits cluster, the fruits and vegetables factor, the fat-reduced/diet foods factor, and all indexes were associated with reduced risk; the meat and potatoes factor was associated with increased risk. In women, reduced risk was found with the Healthy Eating Index-2005 and increased risk with the meat and potatoes factor. For men, beneficial health characteristics were seen with all fruit/vegetable patterns, diet foods patterns, and indexes, while poorer health characteristics were found with meat patterns. For women, findings were similar except that poorer health characteristics were seen with diet foods patterns. Similarities were found across methods, suggesting basic qualities of healthy diets. Nonetheless, findings vary because each method answers a different question.

The population distribution of ratios of usual intakes of dietary components that are consumed every day can be estimated from repeated 24-hour recalls

Estimating the population distribution of the usual intake of a nutrient relative to that of another nutrient requires determination of individual-level ratios. If intake data are available on a per-day basis, as with 24-h dietary recalls, those ratios can be determined in 1 of 2 ways: as the usual ratio of intakes or the ratio of usual intakes. Each of these ratios has its own meaning and determination; the ratio of usual intakes is conceptually consistent with determinations obtained from FFQ data. We present a method for estimating the ratio of usual intakes that uses bivariate modeling of the 2 nutrient intakes in question. Application of the method to the NHANES data for the years 2001-2004 yielded estimated distributions for percent of usual energy intake from total fat, percent of usual energy intake from saturated fat, and usual sodium intake per 1000 kcal (4184 kJ) of usual energy intake. Distributions for both the total population and for age-gender subgroups were estimated. Approximately 60% of adults (>19 y) had a usual total fat intake that was within the recommended range of 20-35% of total energy, but only approximately 34% had a usual saturated fat intake <10% of total energy. The results changed only minimally when the other definition of usual intake, the usual ratio of intakes, was adopted.

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.

Differences between food group reports of low-energy reporters and non-low-energy reporters on a food frequency questionnaire

BACKGROUND

Low-energy reporters (LERs) and non-LERs differ with respect to several characteristics, including self-reported intake of foods. Limited data exist regarding food intake difference between LERs and non-LERs identified using doubly labeled water (DLW).

OBJECTIVE

In the Observing Protein and Energy Nutrition Study (September 1999-March 2000), differences were examined between food group reports of LERs and non-LERs on a food frequency questionnaire (FFQ) (n=440).

DESIGN

LERs were identified using DLW. Responses of LERs (n=220) and non-LERs (n=220) for 43 food groups on the FFQ were examined in three ways: whether they reported consuming a food group (yes/no), how frequently they reported consuming it (times per day), and the reported portion size (small, medium, or large). Analyses were adjusted for total energy expenditure from DLW.

RESULTS

LERs, compared to non-LERs, were less likely to report consumption for one food group among women (soft drinks/regular). Among men, there was no difference between LERs and non-LERs with respect to reporting consumption of food groups. Reported mean daily frequency of consumption was lower among LERs compared with non-LERs for 23 food groups among women and 24 food groups among men (18 food groups were similar in men and women). In addition, reported mean portion sizes were smaller for LERs compared with non-LERs for six food groups among women and five food groups among men (three food groups were similar in men and women). Results varied minimally by sex and body mass index.

CONCLUSIONS

LERs, compared with non-LERs, were more likely to differ regarding their reported frequency of consumption of food groups than their reported consumption (yes/no) or portion size of food groups. Results did not vary greatly by sex or body mass index. It still remains unclear whether improvement in questionnaire design or additional tools or methods would lead to a decrease in differential reporting due to LER status on an FFQ.

Considerations for an obesity policy research agenda

The rise in obesity levels in the U.S. in the past several decades has been dramatic, with serious implications for public health and the economy. Experiences in tobacco control and other public health initiatives have shown that public policy may be a powerful tool to effect structural change to alter population-level behavior. In 2007, the National Cancer Institute convened a meeting to discuss priorities for a research agenda to inform obesity policy. Issues considered were how to define obesity policy research, key challenges and key partners in formulating and implementing an obesity policy research agenda, criteria by which to set research priorities, and specific research needs and questions. Themes that emerged were: (1) the embryonic nature of obesity policy research, (2) the need to study "natural experiments" resulting from policy-based efforts to address the obesity epidemic, (3) the importance of research focused beyond individual-level behavior change, (4) the need for economic research across several relevant policy areas, and (5) the overall urgency of taking action in the policy arena. Moving forward, timely evaluation of natural experiments is of especially high priority. A variety of policies intended to promote healthy weight in children and adults are being implemented in communities and at the state and national levels. Although some of these policies are supported by the findings of intervention research, additional research is needed to evaluate the implementation and quantify the impact of new policies designed to address obesity.

A population's mean Healthy Eating Index-2005 scores are best estimated by the score of the population ratio when one 24-hour recall is available

The USDA Healthy Eating Index-2005 (HEI-2005) is a tool to quantify and evaluate the quality of diet consumed by the U.S. population. It comprises 12 components, expressed as ratios of a food group or nutrient to energy intake. The components are scored on a scale from 0 to M, where M is 5, 10, or 20. Ideally, the HEI-2005 is calculated on the basis of the usual dietary intake of an individual. Intake data, collected via a 24-h recall, are often available for only 1 d for each individual. In this article, we examine how best to estimate a population's mean usual HEI-2005 component and total scores when 1 d of dietary information is available for a sample of individuals from the population. Three methods are considered: the mean of individual scores, the score of the mean of individual ratios, and the score of the ratio of total food group or nutrient intake to total energy intake, which we call the population ratio. We investigate via computer simulation which method is the least biased. The simulations are based on statistical modeling of the distributions of intakes reported by 738 women participating in the Eating at America's Table Study. The results show that overall, the score of the population ratio is the preferred method. We therefore recommend that the quality of the U.S. population's diet be assessed and monitored using this method.

Index-based dietary patterns and risk of colorectal cancer: the NIH-AARP Diet and Health Study

The authors compared how four indexes-the Healthy Eating Index-2005, Alternate Healthy Eating Index, Mediterranean Diet Score, and Recommended Food Score-are associated with colorectal cancer in the National Institutes of Health-AARP Diet and Health Study (n = 492,382). To calculate each score, they merged data from a 124-item food frequency questionnaire completed at study entry (1995-1996) with the MyPyramid Equivalents Database (version 1.0). Other variables included energy, nutrients, multivitamins, and alcohol. Models were stratified by sex and adjusted for age, ethnicity, education, body mass index, smoking, physical activity, and menopausal hormone therapy (in women). During 5 years of follow-up, 3,110 incident colorectal cancer cases were ascertained. Although the indexes differ in design, a similarly decreased risk of colorectal cancer was observed across all indexes for men when comparing the highest scores with the lowest: Healthy Eating Index-2005 (relative risk (RR) = 0.72, 95% confidence interval (CI): 0.62, 0.83); Alternate Healthy Eating Index (RR = 0.70, 95% CI: 0.61, 0.81); Mediterranean Diet Score (RR = 0.72, 95% CI: 0.63, 0.83); and Recommended Food Score (RR = 0.75, 95% CI: 0.65, 0.87). For women, a significantly decreased risk was found with the Healthy Eating Index-2005, although Alternate Healthy Eating Index results were similar. Index-based dietary patterns that are consistent with given dietary guidelines are associated with reduced risk.

A comparison of food-based recommendations and nutrient values of three food guides: USDA's MyPyramid, NHLBI's Dietary Approaches to Stop Hypertension Eating Plan, and Harvard's Healthy Eating Pyramid

The purpose of this research was to compare food-based recommendations and nutrient values of three food guides: the US Department of Agriculture's MyPyramid; the National Heart, Lung, and Blood Institute's Dietary Approaches to Stop Hypertension Eating Plan, and Harvard University's Healthy Eating Pyramid. Estimates of nutrient values associated with following each of the food guides at the 2,000-calorie level were made using a composite approach. This approach calculates population-weighted nutrient composites for each food group and subgroup, assuming average choices within food groups. Nutrient estimates were compared to the Dietary Reference Intakes and other goals and limits. Recommendations were similar regarding almost all food groups for both the type and amount of foods. Primary differences were seen in the types of vegetables and protein sources recommended and the amount of dairy products and total oil recommended. Overall nutrient values were also similar for most nutrients, except vitamin A, vitamin E, and calcium. These food guides were derived from different types of nutrition research, yet they share consistent messages: eat more fruits, vegetables, legumes, and whole grains; eat less added sugar and saturated fat; and emphasize plant oils.

Communicating healthy eating: lessons learned and future directions

Achieving and maintaining wide-scale positive dietary change is a complex and formidable endeavor, given the current food environment. Moreover, for positive change to occur, nutrition messages should be communicated in a scientifically precise, yet practical and motivating manner. This challenge was the impetus for the organization of a 2-day workshop hosted by the National Cancer Institute (NCI) and the Division of Nutrition Research Coordination (DNRC), both of the National Institutes of Health (NIH). The conference included communication, nutrition, and behavioral scientists, market researchers, media advocates, journalists, and public policy experts. Discussions regarding communication efforts and the best methods to craft, deliver, and evaluate the impact of nutrition messages illustrated both the challenges and the opportunities we face. During the discussions, important recommendations for nutrition communicators and interventionists emerged, based on existing knowledge from the communications field, lessons learned thus far, and noted gaps in our knowledge.

What can communication science tell us about promoting optimal dietary behavior?

Four of the 10 leading causes of death can be attributed to poor dietary behaviors. Nutrition professionals continue to struggle with the most effective ways to deliver nutrition messages that will result in changes in dietary behavior. On July 14-15, 2005, the National Cancer Institute and the Division of Nutrition Research Coordination, both of the National Institutes of Health, hosted a meeting to explore the state of the science concerning this issue. This paper provides an introduction to that meeting and the articles that resulted from it.

Most Americans eat much less than recommended amounts of fruits and vegetables

OBJECTIVE

To estimate the proportions of the population meeting recommendations for fruit and vegetable intake, we first estimated the usual intake distributions of total fruits and vegetables and then compared the results to the 5 A Day recommendation and to the recommendations for fruits and vegetables combined, found in the new US Department of Agriculture food guide, MyPyramid.

DESIGN/SUBJECTS

The primary dataset was created from one 24-hour recall from each of 8,070 respondents in the 1999-2000 National Health and Nutrition Examination Survey. Variances were estimated using one or two 24-hour recalls from 14,963 respondents in the 1994-1996 Continuing Survey of Food Intakes by Individuals.

STATISTICAL ANALYSIS

The statistical method developed at Iowa State University was used for estimating distributions of usual intake of dietary components that are consumed daily. It was modified to allow the adjustment of heterogeneous within-person variances using an external estimate of heterogeneity.

RESULTS

In 1999-2000, only 40% of Americans ate an average of five or more (1/2)-cup servings of fruits and vegetables per day. The proportions of sex-age groups meeting the new US Department of Agriculture recommendations ranged from 0.7% of boys aged 14 to 18 years, whose combined recommendation is 5 cups, to 48% of children aged 2 to 3 years, whose combined recommendation is 2 cups.

CONCLUSIONS

Americans need to consume more fruits and vegetables, especially dark green and orange vegetables and legumes. Nutritionists must help consumers realize that, for everyone older than age 3 years, the new recommendations for fruit and vegetable intakes are greater than the familiar five servings a day.

The Food Propensity Questionnaire: Concept, Development, and Validation for Use as a Covariate in a Model to Estimate Usual Food Intake

OBJECTIVE

Twenty-four-hour recalls capture rich information on food consumption, but suffer from inadequately measuring usual intakes of episodically consumed foods. We explore using food frequency questionnaire (FFQ) data as covariates in a statistical model to estimate individual usual intakes of episodically consumed foods and their distributions and describe the development of the Food Propensity Questionnaire, an FFQ introduced in the 2003-2004 National Health and Nutrition Examination Survey.

DESIGN

We analyzed data from 965 adult participants in the Eating at America's Table Study who completed four 24-hour recalls and an FFQ. We assessed whether or not increasing FFQ-reported frequency was associated with both number of 24-hour recall consumption days and amounts reported.

RESULTS

For 52 of 56 food groups (93%), and 218 of 230 individual foods (95%), there were significant monotonically increasing relationships between FFQ frequency and 24-hour recall probability of consumption. For 47 of 56 food groups (84%) and 55 of 230 (24%) individual foods, there were significant positive correlations between FFQ frequencies and consumption-day mean intake.

CONCLUSIONS

We found strong and consistent relationships between reported FFQ frequency of food and food-group consumption and probability of consumption on 24-hour recalls. This supports the premise that frequency data may offer important covariate information in supplementing multiple recalls for estimating usual intake of food groups.

Patterns of health behavior in U.S. adults

BACKGROUND

Associations between health-related behaviors are important for two reasons. First, disease prevention and health promotion depend on understanding both prevalence of health behaviors and associations among such behaviors. Second, behaviors may have synergistic effects on disease risk.

METHODS

We document patterns of adherence to recommendations concerning five behaviors (physical activity, tobacco use, alcohol consumption, fruit and vegetable consumption, and dietary fat intake) in U.S. adults (n = 15,425) using data from the Third National Health and Nutrition Examination Survey. Division of individuals into categories associated with adherence or nonadherence to lifestyle recommendations results in 32 patterns of adherence/nonadherence.

RESULTS

Proportions of U.S. adults with 21 of 32 behavior patterns characterized here deviated from proportions expected if health behaviors are independent of each other. The two extreme patterns, all adherence (5.9%) and all nonadherence (4.9%), were found in about double the proportion expected. Age, gender, race/ethnicity, education, and income were associated with a number of patterns, including the two extremes.

CONCLUSIONS

This analysis of behavior patterns highlights population subgroups of public health importance, provides a benchmark for studies of multivariate associations between health behaviors, and supports a multidimensional model of health behavior.

Evaluating the impact of population changes in diet, physical activity, and weight status on population risk for colon cancer (United States)

OBJECTIVE

To estimate the effects of observed population-level changes in risk factors on population risk and incidence of disease.

METHODS

Trends in a set of risk factors for colon cancer (vegetable intake, red meat intake, alcohol consumption, physical activity levels, and weight status) were modeled for the US adult population over the years 1975-1995 and combined with relative risk estimates from epidemiologic studies and a probability distribution for the induction period to estimate the percentage change in incidence rates from 1985 to 1995 due to the five risk factors. A sensitivity analysis was performed to account for imprecision related to estimates of trends in behavior and epidemiologic risk.

RESULTS

Increased vegetable intake and decreased intakes of red meat and alcohol reduced risk, while reduced physical activity and increased body mass index increased risk for colon cancer. When all five factors were considered together, change in the average population relative risk was small and the risk factors accounted for little of the recently observed decline in incidence.

CONCLUSIONS

Although these factors have the potential to greatly affect risk of colon cancer and incidence rates, little of that potential was realized since adverse trends neutralized what progress had been made in the areas of vegetables, red meat, and alcohol consumption.

Let the pyramid guide your food choices: capturing the total diet concept

This paper discusses how the guideline "Eat a variety of foods" became "Let the Pyramid guide your food choices," presents background information on the food guidance system upon which the Food Guide Pyramid is based and reviews methods that have been used to assess aspects of the total diet, i.e., the variety, moderation and proportionality, promoted by this guidance. The methods include measures of dietary variety, patterns based on Pyramid food group intakes and scoring methods comprised of multiple dietary components. Highlights of results from these methods include the following. Although approximately one third of the U.S. population eat at least some food from all Pyramid food groups, only approximately 1-3% eat the recommended number of servings from all food groups on a given day. Fruits are the most commonly omitted food group. Vegetables and meat are the groups most commonly met by adults, and dairy the most commonly met by youth. Intakes of specific types of vegetables (i.e., dark green, deep yellow) and of grains (i.e., whole grains) are well below that recommended; intakes of total fat and added sugars exceed current recommendations. Scoring methods show those diets of the majority of the population require improvement, and that diets improve with increases in education and income. This paper also discusses the limitations and strengths of these approaches, and concludes with suggestions to improve current food guidance and methods to assess the total diet.