Continuing Survey Of Food Intakes By Individuals Research Articles

Disparities in Snacking Trends in US Adults over a 35 Year Period from 1977 to 2012.

Background: Previous studies have indicated that snacking is contributing to increased calories in the American diet, and that the contribution of snacks to energy intake has increased in recent decades. Objective: The objective was to examine trends in the energy intake deriving from snacks and food sources of snacks for US adults over 35 years from 1977 to 2012, and whether these trends differ across sociodemographic groups. Methods: Participants included 74,291 US adults aged ≥19 years who participated in the 1977–1978 Nationwide Food Consumption Survey (NFCS); the 1989–1991 Continuing Survey of Food Intake by Individuals (CSFII), the 1994–1996 CSFII, the 1997–1998 CSFII, National Health and Nutrition Examination Survey (NHANES) 2003–2004, NHANES 2005–2006, NHANES 2009–2010 and NHANES 2011–2012; all surveys are a stratified random sample of the total civilian, non-institutionalized US population. Overall patterns of snacking, trends in energy intake from snacking, trends in snacking food and beverage sources and energy intake from snacks across racial/ethnic, age, education and income groups were examined. Results: For all US adults there was a significant increase in per capita energy intake from snacks from 1977 to 2012 (p < 0.01). Non-Hispanic blacks had the highest snacking energy intake from sugar-sweetened beverages (SSBs) in each year (p < 0.01 for all), with a peak intake in 2003–2006 of 76 kcal per capita per day. Intake of salty snacks increased in all groups between 1977 and 2012, with non-Hispanic blacks having the highest intake in 2009–2012 (p < 0.01). No significant differences were found between income or education groups. Conclusions: Results show that snacking remains a significant component of the US diet and the foods consumed at these snacks (SSBs, desserts and sweets and salty snacks) are not the types of foods recommended by the US dietary guidelines. Our finding that non-Hispanic blacks are consuming a higher proportion of SSBs and salty snacks than other groups, and showing the largest increase in energy intake deriving from snacks overall over the past 35 years, is indeed concerning.

A Consistent Database of Nutrient Intakes 1978‐Present Can Improve Analyses of Consumption Trends

For decades, the Federal Government has conducted nationally representative surveys of Americans' food and nutrient intakes. Major surveys used for dietary trend analyses include the U.S. Department of Agriculture (USDA)'s 1977–78 Nationwide Food Consumption Survey (NFCS) and the 1989–91 and 1994–96 Continuing Survey of Food Intakes by Individuals (CSFII), as well as the National Center for Health Statistics (NCHS)' National Health and Nutrition Examination Survey (NHANES). These surveys are used by researchers and policymakers to assess long‐term trends, yet methodological changes such as changes in nutrient databases can have important influences on results. The dietary references (i.e., nutrient values per 100 grams) utilized in the 1977–78 NFCS and 1989–91 CSFII differed from the 1994‐present databases, and did not contain some key nutrients like sodium. We link current dietary reference values to the 1977–78 NFCS and 1989–91 CSFII databases to create a consistent database that includes all the nutrients that can be analyzed today. We present examples of how the use of a consistent database can affect trend data using 1977–78 NFCS, the 1989–91 CSFII, and the continuous waves of NHANES 2003‐present. We describe the database we created to conduct these analyses, which will be publicly available upon completion.Support or Funding InformationSupported by U.S.Department of Agriculture Economic Research Service Cooperative Research Agreement No. 58‐4000‐6‐0009.

Adaptive LASSO for varying-coefficient partially linear measurement error models

This paper extends the adaptive LASSO (ALASSO) for simultaneous parameter estimation and variable selection to a varying-coefficient partially linear model where some of the covariates are subject to measurement errors of an additive form. We draw comparisons with the SCAD, and prove that both the ALASSO and the SCAD attain the oracle property under this setup. We further develop an algorithm in the spirit of LARS for finding the solution path of the ALASSO in practical applications. Finite sample properties of the proposed methods are examined in a simulation study, and a real data example based on the U.S. Department of Agriculture's Continuing Survey of Food Intakes by Individuals (CSFII) is considered.

U.S. Trends in dietary variety and its association with BMI and micronutrient intakes

One proposed contributor to excess energy intake is dietary variety, particularly energy‐dense variety (McCrory et al 1999). Using national survey data, we examined the potential role of dietary variety in the obesity epidemic and relationships among variety and micronutrient intake. Usual dietary intake data from the Continuing Survey of Food Intakes by Individuals (CSFII) 1989–91 and 1994–96 were coded and six dietary variety scores were calculated as described previously (Roberts et al 2005). We included adults aged 21–60 y, and excluded individuals with implausible energy intake reports (Huang et al 2005) and BMI&lt;17 kg/m2. Results showed all types of dietary variety increased from 1989–91 to 1994–96 (p≤.025), except for food group variety which remained constant. In both surveys, energy‐dense variety was higher and energy‐weak variety was lower with higher BMI (p≤.05). Micronutrient dense variety was positively associated with mean micronutrient intake (%EAR) in both surveys (p≤0.001), whereas energy‐dense variety was only weakly associated with greater micronutrient intake in 1989–91 and not associated in 1994–96. Increasing energy‐dense variety over time may lead to weight gain with minimal contribution to micronutrient intake.Grant Funding Source : n/a

Associations of food group and nutrient intake, diet quality, and meal sizes between adults and children in the same household: a cross-sectional analysis of U.S. households

BackgroundOne might assume that individuals living in the same household have similar dietary intakes of food groups and nutrients. However, the manner in which an adult's dietary intake affects children's food consumption, diet quality (defined as meeting intake recommendations), and meal sizes is understudied to date. The objective of this study was to estimate these relationships between minor children and the female or male head of household.MethodsDietary intakes of one randomly selected child of each age group (2-5, 6-11, or 12-18 years old (n = 2,380)) and that of the female/male head of household ((HH), proxy for mother and father) using multiple 24-hour recalls from the Continuing Survey of Food Intake by Individuals (CSFII) 1994-1996 was coded to reflect food group and nutrient density (servings/grams per 1,000 kcal). Linear or logistic regression models were used to determine the association between intakes, whether individuals' diets trended toward meeting her/his intake recommendations, and whether individuals were in the highest quintile for food group densities at four distinct eating occasions (breakfast, brunch/lunch, supper/dinner, or other) in each subject group. Stata's survey commands were used to fit linear or logistic regression models and obtain adjusted regression coefficients or odds ratios.ResultsAssociations between food group/nutrient densities were significant but weak to moderate. Adults with diets that trended toward meeting their intake recommendations doubled the odds for children to have diets that trended toward meeting the recommendations; for many meals, adults consuming in the highest quintile for food group density predicted that children's intakes were also in the highest quintile.ConclusionsFemale and male adults living in the same household significantly affect children's food group and nutrient intakes, diet quality, and meal sizes. There is an urgent need for in-depth analysis to elucidate the underlying mechanisms, especially for studies involving both the female and male HH.

Trends and correlates in meat consumption patterns in the US adult population

Few studies have examined recent shifts in meat consumption (MC), differences among US population groups, and the influence of psychosocial-behavioural factors. Nationally representative data collected for US adults aged >or=18 years in the 1988-1994 and 1999-2004 National Health and Nutrition Examination Survey (NHANES) and the 1994-1996 Continuing Survey of Food Intakes by Individuals (CSFII) and Diet and Health Knowledge Survey (DHKS) were used. We found a U-shaped trend in MC, a decrease between 1988-1994 and 1994-1996, and an increase from 1994-1996 to 1999-2004. NHANES 1988-1994 and 1999-2004 indicate that MC did not change significantly, particularly for all meat, red meat, poultry and seafood. Between 1994-1996 and 1999-2004, average MC, including red meat, poultry, seafood and other meat products, increased in men. Women's total MC decreased, mainly due to decreased red meat and other meat products, except for increased seafood. Noticeable differences existed in the changes across population groups. Black men had the largest increase in consumption of total meat, poultry and seafood; Mexican American men had the smallest increase in poultry, seafood and other meat products. In 1999-2004, ethnic differences in MC became greater in women than among women in 1994-1996. Associations between MC and energy intake changed over time. Perceived benefit of dietary quality and food label use were associated with reduced red MC. Noticeable differences exist in the shifts in MC across population groups and surveys. MC increased in men but decreased in women in recent years.

Maternal Employment and Household Food Production: Implications for Nutrition and Obesity

This paper explores the mechanisms through which maternal employment affects the quality of household nutrition and the implications for obesity of family members. Using comparable samples from the American Time Use Survey (ATUS) and the Continuing Survey of Food Intake by Individuals (CSFII), I find that maternal employment is negatively associated with time spent shopping for and preparing food and positively associated with the share of household food expenditure spent on food prepared away from home. This substitution of money for time in food consumption can have detrimental effects on the nutrition of both adults and children in the family. In married-couple families, I find that the quality of food-intake falls with maternal employment in all subgroups studied, and that obesity sometimes, but not always, rises with maternal employment. In single-mother families, however, my results suggest a weaker association between maternal labor supply, quality of food intake, and obesity. Overall, my findings suggest that maternal employment can explain very little of the time trend increase in obesity.

Weekend and Weekday Eating Patterns in Puerto Rican Schoolchildren Participating or Not Participating in the School Lunch Program

Number of eating occasions and meal patterns can have marked implications on the nutritional quality of children's diets. Various factors such as financial resources and mother's employment status have been examined by the Continuing Survey of Food Intake by Individuals (CSFII)and have been reported for children. We have examined additional factors ‐ those of participation in the School Lunch Program and weekday vs weekend eating patterns. To accomplish this, we have performed 24 hr recall studies with children at 3 school levels: elementay (n=101), intermediate (n=115) and high school (n=105). We have further analyzed eating patterns on the basis of gender and of median body mass index (BMI). Results show that the mean number of eating occasions to be higher in participants (P) than in non‐participants (NP, to be higher in elementary school which lowers further in intermediate thru high school and to be higher in weekdays vs weekends. Number of occasions exceeded 4 per day for all groups. It would appear that participation in the Lunch Program is related to a higher number of eating occasions per day. This suggests positively that meal skipping is less however this might also raise the risk of overconsumption. Supported by the USDA‐CSREES‐NRI, award #2003‐35200‐13590.

Estimates of Water Ingestion for Women in Pregnant, Lactating, and Non-Pregnant and Non-Lactating Child-Bearing Age Groups Based on USDA's 1994–96, 1998 Continuing Survey of Food Intake by Individuals

ABSTRACT Women in the child-bearing age of 15 to 44 years and, in particular, pregnant and lactating women in this age cohort are considered a sensitive subpopulation when assessing risk from ingestion of water because water borne contaminants may pose a risk not only to the mother but to the fetus or infant. This article presents estimates of daily average per capita water ingestion for women of child-bearing age and in three subgroups: pregnant, lactating, and non-pregnant/non-lactating women. Estimates of means and upper percentiles of subgroup ingestion distributions were generated using participant responses and survey weights from the United States Department of Agriculture's (USDA) 1994–96 and 1998 Continuing Survey of Food Intake by Individuals (CSFII). The ingestion estimates are empirical and not based on an assumed parametric distribution of daily average amount of water ingestion. Water occurring naturally in foods or added by manufacturers to commercial products is not included in the estimates presented. These estimates of water ingestion by women of child-bearing age are compared to those attributed to Ershow and Cantor (1989) by Burmaster (1998). These estimates, based on data collected in 1978, were used by Burmaster to characterize the distribution of daily average per capita ingestion as lognormal. The lognormal estimates of total water ingestion are generally greater than the total water ingestion estimates based on the CSFII data. Possible explanations for the differences are discussed.

Whole grain intake: The Baltimore Longitudinal Study of Aging

Our objective was to identify major dietary sources of whole grains and to describe the construction of a database of whole grain content of foods. Dietary information was collected with 7-d food records from men and women in the Baltimore Longitudinal Study of Aging, mean age 62.1 ± 16.0 years, who participated in the dietary assessment portion of the study ( n = 1516), and estimates of whole grain intake were obtained from a newly developed database. The Pyramid Servings database and 1994–1996 Continuing Survey of Food Intakes by Individuals (CSFII) recipe ingredients database were then used to calculate both servings and gram weights of whole grain intakes. Mean intakes of whole grains, refined grains, and total grains, as well as frequency of intake for major whole grain food groups and whole grain content for each group, were calculated. Top contributors of whole grains were ready-to-eat breakfast cereals (made with whole grain as well as bran), hot breakfast cereals (made with whole grain), multi-grain bread, and whole wheat bread. While more research is needed to better understand the benefits of whole grains, the development of research tools, including databases to accurately assess whole grain intake, is a critical step in completing such research.

The mitigating influence of time preference on the relation between smoking and BMI scores

Prior studies have found a strong negative relation between smoking and body mass index (BMI). Smoking and unhealthy eating both imply a preference for utility in the present at the expense of future consumption. This analysis seeks to proxy this preference through a composite index of equally weighted intertemporal behaviors to isolate the impact of smoking upon BMI independent of time discounting. Ordinary least squares regression is used to analyze BMI score based on a number of key independent variables. A total of 2970 individuals who provided data for the United States Department of Agriculture's (USDA) Continuing Survey of Food Intakes by Individuals (CSFII) 1994-1996, as well as the Diet and Health Knowledge Survey (DHKS). Adding time preference to a multivariate model inflates the magnitude of the smoking effect, consistent with discounted utility theory. Results suggest that full effect of the non-intertemporal aspects of smoking (e.g., the substitution of cigarettes for food) on BMI scores may be underestimated in previous studies that fail to account for the mitigating influence of time preference.

Estimated daily average per capita water ingestion by child and adult age categories based on USDA's 1994–1996 and 1998 continuing survey of food intakes by individuals

Water ingestion estimates are important for the assessment of risk to human populations of exposure to water-borne pollutants. This paper reports mean and percentile estimates of the distributions of daily average per capita water ingestion for a number of age range groups. The age ranges, based on guidance from the US EPA's Risk Assessment Forum, are narrow for younger ages when development is rapid and wider for older ages when the rate of development decreases. Estimates are based on data from the United States Department of Agriculture's (USDA's) 1994-1996 and 1998 Continuing Survey of Food Intake by Individuals (CSFII). Water ingestion estimates include water ingested directly as a beverage and water added to foods and beverages during preparation at home or in local establishments. Water occurring naturally in foods or added by manufacturers to commercial products (beverage or food) is not included. Estimates are reported in milliliters (ml/person/day) and milliliters per kilogram of body weight (ml/kg/day). As a by-product of constructing estimates in terms of body weight of respondents, distributions of self-reported body weights based on the CSFII were estimated and are also reported here.

Development of the Iowa bone nutrient food frequency questionnaire based on data from the US Department of Agriculture Continuing Survey of the Food Intake by Individuals

An easily-administered food frequency questionnaire (FFQ)/dietary screener was developed for current (adult) and retrospective (adolescent) intakes of nutrients important for bone development and maintenance. This tool quantified serving sizes and nutrients from foods using gender- and age-specific techniques. Nutrients of interest were calcium, vitamin D, caffeine and alcohol, and 15 categories of foods were selected for inclusion based on frequency of intake and nutrient density. Calcium-contributing foods were selected from published dietary intake assessment tools. Foods contributing vitamin D, caffeine and alcohol were selected based on nutrient density and Midwest consumption practices. Serving sizes were quantified in standard serving units or as small, medium and large servings. Food items selected for the FFQ/dietary screener were matched to foods from the United States Department of Agriculture (USDA) Continuing Survey of Food Intakes by Individuals (CSFII). Calcium, caffeine and alcohol values were assigned using CSFII data files at median values per 100 g intake. CSFII midpoint tertile frequency of intake values for males and females 14–18 and 25–45 years old were used to establish serving weights for small, medium and large servings. CSFII data files provide an efficient way for estimating typical intakes, serving sizes and nutrient values for target groups. Age- and gender-derived data provided realistic estimates of nutrient intakes when using FFQ/dietary screener assessment method.

How do socio-economic status, perceived economic barriers and nutritional benefits affect quality of dietary intake among US adults?

Socio-economic factors may affect diet quality, perhaps differentially across gender and ethnicity. The mechanism of this association is still largely unknown. We examined the independent effects of socio-economic status (SES), perceived barrier of food price (PBFP) and perceived benefit of diet quality (PBDQ) on diet quality indicators and indices (DQI(j,k)), across gender and ethnicity. Additionally, we estimated the mediation proportion of the effect of SES on DQI(j,k) through PBFP and PBDQ. Data from two cross-sectional surveys, the Continuing Survey of Food Intakes by Individuals (CSFII) and Diet and Health Knowledge Survey (DHKS) 1994-96 were used. Our sample consisted of 4356 US adults aged 20-65 years. With principal components analysis, SES (an index) was measured using household income per capita and education, and PBDQ was measured using an 11-item scale. PBFP was defined as the ratio of importance of food price score relative to nutrition. DQI(j,k) were assessed by a set of indicators and two indices including the Healthy Eating Index. The associations between SES, PBFP, PBDQ and DQI(j,k) varied significantly across gender and ethnic groups. PBFP acted as a mediator in the association between SES and selected DQI(j) indicators, namely energy, fat intake, sodium and simple sugar consumption (mediation proportion >10%), but not PBDQ. SES, PBFP and PBDQ all affect dietary intake, and vary by ethnicity and gender. Positive effect of SES on DQI(j,k) may be mediated by PBFP but not PBDQ which is an independent protective factor. Nutrition education is important to promote healthy eating.

How do socio‐economic status (SES), perceived economic barriers and nutritional benefits affect quality of dietary intake among US adults?

We examined independent effects of SES, perceived barriers of food price (PBFP) and benefits of diet quality (PBDQ) on diet quality indicators and indices (DQI), across gender and ethnicity using nationally representative data from the Continuing Survey of Food Intakes by Individuals (CSFII) and Diet and Health Knowledge Survey (DHKS) 1994–96 for 4,356 individuals aged 20–65 y. Diet was assessed using two 24-hour recalls. By principal components method, SES was constructed with household income per capita and education; and PBDQ, using an 11-item scale. PBFP is the ratio of importance of food price score relative to nutrition. Several DQI indicators and indices were considered. PBDQ z-score improved diet quality (β= −31 for energy, −0.34 for fat (% of energy), and 0.37 for fiber p<0.05). SES z-score among Whites, unlike African-Americans, reduced fat intake by β= −0.71 (p<0.05) and saturated fat by β= −0.43 (p<0.05) with a significant SES×ethnicity interaction. SES positively associated with fiber intake and was significantly stronger among women (β=0.95 vs. 0.72, p<0.05). Structural equation models revealed significant mediation of SES effects on diet quality through PBFP but not through PBDQ (mediation proportion>10%) for energy, total fat, sodium and sugar intakes. In conclusion, the effect of SES on diet quality varied by ethnicity and gender, and there was mediation of the effect of SES on diet quality by PBFP.

Nutrient Deficiency Classification Bias Relating to the 24-hour Dietary Recall: A Case Study With Dietary and Red Blood Cell Folate

SAB3-PD-07 The 24-hour recall data from food consumption surveys, such as the U.S. Continuing Survey of Food Intake by Individuals (CSFII) or the NHANES, are frequently used to estimate dietary exposure, which is often compared with recommended daily intake (RDI) for micronutrient to classify “nutrient deficiency status” for various population subgroups. However, the 24-hour recall method is not ideal for estimating long-term average daily food intakes since there are considerable variations in day-to-day intake of food within individuals. Single-day intake data usually result in a distribution that is flatter and wider than the true distribution of usual intakes of individuals in the population. Thus, the prevalence of high or low intakes (undernutrition or overnutrition) is overestimated with single 24-hour recall. The U.K. National Diet and Nutrition Surveys collect multiple days of intake data and could provide a mean to quantify the effects of intra-individual variation on estimates of usual food intakes. Individual's dietary folate intakes for day 1 through day 7 are estimated and folate status is established based on different combination of intake days and number of days prior to blood sampling. Using the red blood cell folate data collected in the U.K. survey, individual's folate deficiency status is also defined. Red blood cell folate concentration is thought to be a better index of tissue folate storage than serum folate, ie, longer-term measure of folate status. Classification of folate deficiency status based on the dietary intake estimates is compared with classification based on RBC folates. Bias in the classification of folate status using the single 24-hour recall is discussed.

Prediction and Evaluation of Dietary on Arsenic Exposures Using the SHEDS Model

SAB3-PD-04 Arsenic (As) is one of the important environmental contaminants with many documented health hazards. Dietary intake is one of the most important exposure pathways for arsenic. Approximately 40,000 person-days of data from the USDA's 1994–1996 Continuing Survey of Food Intake by Individuals (CSFII) and arsenic residues from the FDA's Total Dietary Survey (TDS) were combined in the Environmental Agency's Stochastic Human Exposure and Dose Simulation (SHEDS) model to estimate daily average dietary As exposure. Results ranged from approximately 0.2 to 0.5 μg/kg/day for different age groups, with higher As exposures predicted for the younger age groups (ages 1–5 years), mainly due to their lower body weight. Because of their special dietary habits, Asian Americans were predicted to have the highest dietary As exposure among the other groups studied (1 μg/kg/day, more than 3 times that of whites). The dietary As exposure for people living in urban areas was estimated to be about 1.6 greater than those living in rural areas on average. The dietary As exposure for populations from the Northeastern United States was found to be higher than those from other regions. The model was used to identify the 50 most important contributing food items, which contribute averagely more than 50% As dietary exposure. Seafood and rice were found to be the 2 most important food categories contributing to dietary As exposure. Evaluation of the SHEDS dietary model with duplicate food samples for 102 subjects from the National Human Exposure Assessment Survey (NHEXAS) showed that model results compare well against observed results. Although this work was reviewed by Environmental Agency and approved for publication, it may not necessarily reflect official Agency policy.

The effect of improved food composition data on intake estimates in the United States of America

The effect of improved food composition data on nutrient intake estimates was determined by re-analyzing dietary intake data from the Continuing Survey of Food Intake by Individuals (CSFII) 1994–1996, 1998 with the multi-year version of the Food and Nutrient Database for Dietary Studies (FNDDS) 1.0, wherein only the data improvements such as those due to new analytical data replaced the older values. Mean differences between the old and revised estimates were determined. Improved data resulted in minor but statistically significant ( P = 0.001 ) differences in mean intake estimates for most nutrients. Nutrients or food components with greatest differences included vitamin C, riboflavin, magnesium, and caffeine. As a result of these changes, 4–7% more adults have inadequate intakes of vitamin C and magnesium. Caffeine intakes were lower by about 25%. Changes in the food composition values for fluid milk, tomatoes, coffee, and mixed dishes appear to have had the most impact. These artifactual changes require adjustments to the earlier intake estimates to improve comparability with more current intakes.

Methodology for Adding Glycemic Load Values to the National Cancer Institute Diet History Questionnaire Database

Background A growing interest exists in using glycemic index and glycemic load as potentially important exposures in investigations of risk for a variety of chronic diseases. Objective We added values for glycemic index and glycemic load to the nutrient database of a commonly used dietary assessment instrument, the Diet History Questionnaire (DHQ). Design The nutrient database for the DHQ is based on 4,200 individual foods reported by adults in the 1994-1996 US Department of Agriculture Continuing Survey of Food Intakes by Individuals (CSFII). This list was condensed into 225 nutritionally similar groupings of individual foods. Using published glycemic index values we assigned glycemic index values to each of the individual CSFII foods in these food groups. In cases where CSFII foods did not correspond tightly to foods with published glycemic index values, we used decision criteria to assign glycemic index values. We then calculated sex- and serving size-specific glycemic load for each of the 225 food groups using the weighted mean method. Quality assessments were made to help evaluate the success of this method for assigning glycemic load values. Results Seventy-one percent of the top carbohydrate-contributing food groups had in excess of 90% of the CSFII mentions linked directly to a published glycemic index value (ie, no imputation was required), and 100% of these food groups had at least 50% of total mentions linked directly. Conclusions Using this method, it is now possible to use DHQ responses to assess the associations between reported glycemic load and glycemic index and risk of many chronic diseases in epidemiologic studies.

Research Methods in Foodservice Industry

This paper discusses and presents a binary logit model used to identify factors affecting the decision to dine out. The purpose of this paper is to convey an understanding of the logit model and to facilitate use of the model for research related to the foodservice industry. Data used in the binary logit application come from the 1994-1996 Continuing Survey of Food Intakes by Individuals (CSFII) and the Diet and Health Knowledge Survey (DHKS). The results of the empirical study shows that income had a positive effect on the probability of dining out, and those who were employed full time, part time, or temporarily unemployedwere more likely to dine out than unemployed respondents. Household size and age of the respondents had a negative effect on the probability of dining out. Education and gender did not affect the probability of dining out, while race of the household manager had a significant effect on the probability. White household managers were more likely to dine out than non-white household managers. The effects of urbanization and region were not significant.