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

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

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

Are the 2019 Canada's Food Guide Recommendations on Healthy Food Choices Consistent with the EAT-Lancet Reference Diet from Sustainable Food Systems?

BACKGROUND

The diet proposed by the EAT-Lancet Commission, which supports both health and environmental sustainability, provides an opportunity to assess the sustainability of food-based dietary guidelines.

OBJECTIVES

The primary objective was to assess the alignment of the 2019 Canada's Food Guide (CFG) with the EAT-Lancet diet. To do so, an index assessing adherence to the EAT-Lancet diet was developed and evaluated.

METHODS

Data from 1147 adults were used from the cross-sectional PRÉDicteurs Individuals, Sociaux et Environnementaux (PREDISE) study conducted between 2015 and 2017 in the province of Québec. The EAT-Lancet Dietary Index (EAT-I) was developed to evaluate adherence to the EAT-Lancet diet. Adherence to the 2019 CFG was assessed using the Healthy Eating Food Index (HEFI)-2019. Associations between the HEFI-2019 score and component scores and the EAT-I score were examined using linear regression models.

RESULTS

The mean EAT-I score (/80) in this population was 33.4 points [95% confidence interval (CI): 32.2, 34.6]. EAT-I scores were consistent with expected differences in diet quality between females and males (+6.9 points, 95% CI: 4.8, 9.0) and between adults aged 50-65 y and 18-34 y (+4.3 points, 95% CI: 1.6, 7.0). The mean HEFI-2019 (/80) score was 44.9 points (95% CI: 44.1, 45.7). The HEFI-2019 was strongly associated with the EAT-I (ß = 0.76, 95% CI: 0.72, 0.80). Among the 10 components of the HEFI-2019, components such as the whole-grain foods (ß =4.01, 95% CI: 3.49, 4.52), grain foods ratio (ß =3.65, 95% CI: 3.24, 4.07), plant-based protein foods (ß =2.41, 95% CI: 2.03, 2.78), and fatty acids ratio (ß =3.12, 95% CI: 2.72, 3.51) showed the strongest associations with the EAT-I.

CONCLUSIONS

These results suggest that recommendations in the 2019 CFG are largely coherent with the EAT-Lancet diet underscoring the complementarity and compatibility of the 2019 CFG for sustainability and health promotion purposes.

One Avocado per Day as Part of Usual Intake Improves Diet Quality: Exploratory Results from a Randomized Controlled Trial

Background

Few clinical trials have evaluated diet quality change as a predictor of intervention effectiveness.

Objectives

The aim of this study was to examine changes in the Healthy Eating Index (HEI)-2015 after a food-based intervention, and assess the associations between HEI-2015 change and intervention effects on cardiometabolic risk-related outcomes.

Methods

The Habitual Diet and Avocado Trial was a 26-wk, multicenter, randomized, controlled parallel-arm study. Participants were 1008 individuals aged ≥25 y with abdominal obesity (females ≥ 35 inches; males ≥ 40 inches). The avocado-supplemented diet group was provided 1 avocado per day, and the habitual diet group maintained their usual diet. Change in diet quality was assessed using the HEI-2015 from a single 24-h recall conducted at 4 time points. Mixed models were used for analysis.

Results

The avocado-supplemented diet group had a greater increase in the HEI-2015 (4.74 points; 95% CI: 2.93, 6.55) at 26 wk than the habitual diet group. Compared with the habitual diet group, the avocado-supplemented diet group had greater increases in the following HEI-2015 components from baseline: total vegetables (0.99 points; 95% CI: 0.77, 1.21), fatty acid ratio (2.25 points; 95% CI: 1.74, 2.77), sodium (1.03 points; 95% CI: 0.52, 1.55), refined grains (0.82 points; 95% CI: 0.32, 1.31), and added sugars (0.84 points; 95% CI: 0.49, 1.19). No differences in HEI-2015 improvements were observed by race, ethnicity, study site, body mass index, or age category. In the avocado-supplemented diet compared with the habitual diet group, the HEI-2015 increased in females (6.50 points; 95% CI: 4.39, 8.62) but not in males (0.02 points; 95% CI: -3.44, 3.48). Median HEI-2015 change was not associated with intervention-related changes in cardiometabolic disease risk factors.

Conclusions

Intake of 1 avocado per day for 26 wk in adults with abdominal obesity increased adherence to the Dietary Guidelines for Americans. Changes in diet quality did not predict changes in risk factors for cardiometabolic disease.This trial was registered at clinicaltrials.gov as NCT03528031 (https://clinicaltrials.gov/study/NCT03528031).

Effect of an mHealth weight loss intervention on Healthy Eating Index diet quality: the SMARTER randomised controlled trial

In the few weight loss studies assessing diet quality, improvements have been minimal and recommended calculation methods have not been used. This secondary analysis of a parallel group randomised trial (regsitered: https://clinicaltrials.gov/ct2/show/NCT03367936) assessed whether self-monitoring with feedback (SM + FB) v. self-monitoring alone (SM) improved diet quality. Adults with overweight/obesity (randomised: SM n 251, SM + FB n 251; analysed SM n 170, SM + FB n 186) self-monitored diet, physical activity and weight. Real-time, personalised feedback, delivered via a study-specific app up to three times daily, was based on reported energy, fat and added sugar intake. Healthy Eating Index 2015 (HEI-2015) scores were calculated from 24-hour recalls. Higher scores represent better diet quality. Data were collected August 2018 to March 2021 and analysed spring 2022. The sample was mostly female (78·9 %) and white (85·4 %). At baseline, HEI-2015 total scores and bootstrapped 95 % CI were similar by treatment group (SM + FB: 63·11 (60·41, 65·24); SM: 61·02 (58·72, 62·81)) with similar minimal improvement observed at 6 months (SM + FB: 65·42 (63·30, 67·20); SM: 63·19 (61·22, 64·97)) and 12 months (SM + FB: 63·94 (61·40, 66·29); SM: 63·56 (60·81, 65·42)). Among those who lost ≥ 5 % of baseline weight, HEI-2015 scores improved (baseline: 62·00 (58·94, 64·12); 6 months: 68·02 (65·41, 71·23); 12 months: 65·93 (63·40, 68·61)). There was no effect of the intervention on diet quality change. Clinically meaningful weight loss was related to diet quality improvement. Feedback may need to incorporate more targeted nutritional content.

Relationship between Adherence to the 2019 Canada's Food Guide Recommendations on Healthy Food Choices and Nutrient Intakes in Older Adults

BACKGROUND

Following Canada's food guide (CFG) recommendations should ensure adequate nutrient intakes. Older adults have increased needs for certain nutrients and nutrient density; the extent to which adherence to CFG recommendations can help reduce inadequate nutrient intakes is unknown.

OBJECTIVES

We aimed to assess the relationship between adherence to CFG recommendations on healthy food choices and intake of key nutrients in adults 65 y and older from the Canadian Community Health Survey 2015 - Nutrition.

METHODS

Secondary analysis of data from 4093 older adults of the Canadian Community Health Survey 2015 - Nutrition (mean age, 73.6 y, 54% females). Dietary intakes were measured using an interviewer-administered 24-h dietary recall, including 1 repeat in a subsample (42%). The National Cancer Institute multivariate method was used to estimate usual (i.e., long-term) dietary intakes. Adherence to CFG recommendations was measured using the Healthy Eating Food Index (HEFI)-2019 score. Simple linear and logistic regression models estimated the effect of increased HEFI-2019 score on usual nutrient intakes and the prevalence of inadequate nutrient intakes (i.e., below the estimated average requirements), respectively.

RESULTS

Compared with the prevalence of inadequate intakes at median HEFI-2019 score (46.4/80 points), a higher HEFI-2019 (+11 points) was associated with reductions in the prevalence of inadequate intakes of magnesium, vitamin B6 and protein [-19.8% (95% confidence interval (CI): -30.8, -8.9), -12.7% (95% CI: -22.5, -3.0), and -4.7% (95% CI: -9.4, -0.1), respectively]. In contrast, data for higher HEFI-2019 scores were compatible with increased prevalence of inadequate intakes of folate, vitamin D, and calcium [4.0% (95% CI: -8.4, 16.3), 2.6% (95% CI: 1.1, 4.0), and 2.3% (95% CI: -3.0, 7.5), respectively].

CONCLUSIONS

Based on dietary intakes of Canadian older adults in 2015, increasing the degree of adherence to CFG recommendations on healthy food choices may reduce nutrient intake inadequacy for most key nutrients except folate, vitamin D, and calcium.

Usual nutrient intakes and diet quality among United States older adults participating in the Supplemental Nutrition Assistance Program compared with income-eligible nonparticipants

BACKGROUND

The proportion of older adults with food insecurity at 8% has increased faster than that of the general United States population from 2001 to 2017. Many low-income food-insecure older adults rely on food-assistance programs, such as the Supplemental Nutrition Assistance Program (SNAP), for meeting energy and nutrient needs, whereas others are eligible but do not participate. Neither updated nutrient intake estimates nor potential differences in meeting the Dietary Reference Intakes from foods alone and with dietary supplements (DS) among low-income older adults using or eligible for SNAP are known.

OBJECTIVES

This study assessed and compared national estimates of usual nutrient adequacy and dietary quality of United States older adults using SNAP and income-eligible nonparticipants.

METHODS

Usual dietary intake was estimated among older adults (≥60 y; n = 2582) in the 2007-2016 NHANES cross-sectional national survey. Data on food-assistance participation and eligibility (poverty-income-ratio ≤130%), DS use, and ≥24-h dietary recalls were used. The NCI method (Markov Chain Monte Carlo approach) was applied to estimate mean usual nutrient intakes, proportion of inadequate nutrient intake, and dietary quality using the 2015 Healthy Eating Index.

RESULTS

Neither usual nutrient intake from dietary and total sources nor dietary quality differed between older adult SNAP participants and eligible nonparticipants. Low dietary quality and high percentage of inadequate intake for several nutrients were apparent among both groups, especially from food sources alone, including vitamins A (56%), C (55%), D (97%), E (99%), calcium (73%), and magnesium (74%), but rates were attenuated when DS were also considered (i.e., 36% reduced risk for vitamin D inadequacy).

CONCLUSIONS

Diet quality and usual nutrient intake among older adult SNAP participants and eligible nonparticipants were poor, but DS lowered the risk of nutrient inadequacy. Future policies and programs should focus on improving the intake of vitamins A, C, D, E, calcium, and magnesium and dietary quality for all older adults.

Mortality and Life Expectancy Lost in Canada Attributable to Dietary Patterns: Evidence From Canadian National Nutrition Survey Linked to Routinely Collected Health Administrative Databases

Using 5 diet quality indexes, we estimated the mortality and life expectancy lost, at the national level, attributable to poor dietary patterns, which had previously been largely unknown. We used the Canadian Community Health Survey 2004, linked to vital statistics (n = 16,212 adults; representing n = 22,898,880). After a median follow-up of 7.5 years, 1,722 deaths were recorded. Population attributable fractions were calculated to estimate the mortality burden of poor dietary patterns (Dietary Guidelines for Americans Adherence Index 2015, Dietary Approaches to Stop Hypertension, Healthy Eating Index, Alternative Healthy Eating Index, and Mediterranean Style Dietary Pattern Score). Better diet quality was associated with a 32%-51% and 21%-43% reduction in all-cause mortality among adults aged 45-80 years and ≥20 years, respectively. Projected life expectancy at 45 years was longer for Canadians adhering to a healthy dietary pattern (average of 5.2-8.0 years (men) and 1.6-4.1 (women)). At the population level, 26.5%-38.9% (men) and 8.9%-22.9% (women) of deaths were attributable to poor dietary patterns. Survival benefit was greater for individuals with higher scores on all diet indexes, even with relatively small intake differences. The large attributable burden was likely from assessing overall dietary patterns instead of a limited range of foods and nutrients.

Transitioning to sustainable dietary patterns: learnings from animal-based and plant-based dietary patterns in French Canadian adults

Introduction

Many dietary guidelines promote the substitution of animal proteins with plant-based proteins for health benefits but also to help transitioning toward more sustainable dietary patterns. The aim of this study was to examine the food and nutrient characteristics as well as the overall quality and costs of dietary patterns consistent with lower intakes of animal-based protein foods and with higher intakes of plant-based protein foods among French Canadian adults.

Methods

Dietary intake data, evaluated with 24 h recalls, from 1,147 French-speaking adults of the PRÉDicteurs Individuels, Sociaux et Environnementaux (PREDISE) study conducted between 2015 and 2017 in Québec were used. Usual dietary intakes and diet costs were estimated with the National Cancer Institute's multivariate method. Consumption of animal- and plant-based protein foods was classified into quarters (Q) and differences in food and nutrient intakes, Healthy Eating Food Index (HEFI)-2019 scores and diet costs across quarters were assessed using linear regression models adjusted for age and sex.

Results

Participants with lower intakes of animal-based protein foods (Q1 vs. Q4) had a higher HEFI-2019 total score (+4.0 pts, 95% CI, 0.9 to 7.1) and lower daily diet costs (-1.9 $CAD, 95% CI, -2.6 to -1.2). Participants with higher intakes of plant-based protein foods (Q4 vs. Q1) had a higher HEFI-2019 total score (+14.6 pts, 95% CI, 12.4 to 16.9) but no difference in daily diet costs (0.0$CAD, 95% CI, -0.7 to 0.7).

Discussion

In a perspective of diet sustainability, results from this study among French-speaking Canadian adults suggest that a shift toward a dietary pattern focused primarily on lower amounts of animal-based protein foods may be associated with a better diet quality at lower costs. On the other hand, transitioning to a dietary pattern focused primarily on higher amounts of plant-based protein foods may further improve the diet quality at no additional cost.

Healthy Eating Index Diet Quality in Randomized Weight Loss Trials: A Systematic Review

BACKGROUND

Weight loss interventions focus on dietary and physical activity changes to induce weight loss. Both through weight loss and independent of it, diet quality is important for reducing chronic disease risk. However, whether and how diet quality changes over the course of a behavioral intervention is unclear.

OBJECTIVE

To systematically review the evidence from randomized controlled trials on the effect of behavioral interventions on diet quality as defined by the Healthy Eating Index (HEI) among adults with overweight and obesity.

METHODS

PubMed, Ebscohost CINAHL, Embase, OVID APA PsycInfo, Scopus, and Web of Science were searched through May 2021. Inclusion criteria comprised randomized controlled trial design, a primary or secondary aim of weight loss, a sample of US adults with overweight or obesity, measurement using the HEI-2005, 2010, or 2015, and assessment of the time by treatment effect. Interventions must have included behavioral components and lasted at least 3 months. Risk of bias was assessed using the Cochrane Risk of Bias 2 tool. The systematic review protocol was published on Open Science Framework.

RESULTS

Of 3,707 citations retrieved, 18 studies met inclusion criteria. A wide array of behavioral interventions were assessed, including in-person and mobile health interventions as well as those prescribing intake of specific foods. Risk of bias in the included studies primarily arose from the measurement of the outcome variable. Sample sizes ranged from 34 to 413 participants. Nine studies used multiple dietary recalls, with few using the recommended method of Healthy Eating Index calculation. Changes in diet quality ranged from no improvement to a 20-point improvement. More often, improvement was in the 4- to 7-point range.

CONCLUSIONS

The evidence for the efficacy of behavioral weight loss interventions for improving diet quality among adults with overweight and obesity is limited. Modest improvements in HEI scores were observed in the reviewed studies.

Greater adherence to the 2019 Canada's Food Guide recommendations on healthy food choices reduces the risk of cardiovascular disease in adults: a prospective analysis of UK Biobank data

BACKGROUND

Canada's Food Guide (CFG) was profoundly revised in 2019, but the extent to which adherence to recommendations on healthy food choices reduces the risk of cardiovascular disease (CVD) is unknown.

OBJECTIVES

The aim of this study was to examine how greater adherence to the 2019 CFG's recommendations on healthy food choices influences the risk of incident CVD.

METHODS

Participants were a sample of adults without history of CVD, diabetes, or cancer from the UK Biobank prospective cohort study. Usual dietary intakes were estimated by modeling data from repeated Web-based 24-h dietary recalls using the National Cancer Institute multivariate method. Adherence to key CFG recommendations on healthy food choices was assessed using the Healthy Eating Food Index (HEFI)-2019, which has a maximum of 80 points. The CVD outcome was a composite of fatal and nonfatal myocardial infarction and ischemic stroke. Cox regression models adjusted via inverse probability weighting were used to estimate CVD risks. Counterfactual models were used to interpret risks of hypothetical changes in the HEFI-2019 score.

RESULTS

A total of 136,698 participants met the eligibility criteria (55% females; mean age: 57.2 y; range: 40-75 y). During the 11-y follow-up, there were 2843 cases of incident CVD. Compared with no change in the HEFI-2019 score, increasing the HEFI-2019 score of all participants to the 90th percentile of the score distribution (58.1 points) hypothetically reduced the risk of CVD by 24% (RR: 0.76; 95% CI: 0.58, 0.94; absolute risk difference: -0.58%).

CONCLUSIONS

These results suggest that greater adherence to the 2019 CFG recommendations on healthy food choices reduces the 11-y risk of CVD in middle-aged and older adults.

Adhering to Canada's Food Guide Recommendations on Healthy Food Choices Increases the Daily Diet Cost: Insights from the PREDISE Study

The aim of this study was to assess the association between daily diet costs and the Healthy Eating Food Index (HEFI)-2019, an index that reflects the alignment of dietary patterns to recommendations on healthy food choices in the 2019 Canada’s Food Guide (CFG). Dietary intake data from 24 h recalls, completed between 2015 and 2017, of 1147 French-speaking participants of the web-based multicenter cross-sectional PRÉDicteurs Individuels, Sociaux et Environnementaux (PREDISE) study in Quebec were used. Diet costs were calculated from dietary recall data using a Quebec-specific 2015–2016 Nielsen food price database. Usual dietary intakes and diet costs were estimated using the National Cancer Institute’s multivariate method. Linear regression models were used to evaluate associations between diet costs and HEFI-2019 scores. When standardized for energy intake, a higher HEFI-2019 score (75th vs. 25th percentiles) was associated with a 1.09 $CAD higher daily diet cost (95% CI, 0.73 to 1.45). This positive association was consistent among different sociodemographic subgroups based on sex, age, education, household income, and administrative region of residence. A higher daily diet cost was associated with a higher HEFI-2019 score for the Vegetables and fruits, Beverage, Grain foods ratio, Fatty acids ratio, Saturated fats, and Free sugars components, but with a lower score for the Sodium component. These results suggest that for a given amount of calories, a greater adherence to the 2019 CFG recommendations on healthy food choices is associated with an increased daily diet cost. This highlights the challenge of conciliating affordability and healthfulness when developing national dietary guidelines in the context of diet sustainability.

The relationship between moderate to vigorous physical activity and metabolic syndrome: a Bayesian measurement error approach

Metabolic Syndrome (MetS) is a serious condition that can be an early warning sign of heart disease and Type 2 diabetes. MetS is characterized by having elevated levels of blood pressure, cholesterol, waist circumference, and fasting glucose. There are many articles in the literature exploring the relationship between physical activity and MetS, but most do not consider the measurement error in the physical activity measurements nor the correlations among the MetS risk factors. Furthermore, previous work has generally treated MetS as binary, rather than directly modeling the risk factors on their measured, continuous space. Using data from the National Health and Nutrition Examination Survey (NHANES), we explore the relationship between minutes of moderate to vigorous physical activity (MVPA) and MetS risk factors. We construct a measurement error model for the accelerometry data, and then model its relationship between MetS risk factors with nonlinear seemingly unrelated regressions, incorporating dependence among MetS risk factors. The novel features of this model give the medical research community a new way to understand relationships between MVPA and MetS. The results of this approach present the field with a different modeling perspective than previously taken and suggest future avenues of scientific discovery.

A Guide to Dietary Pattern-Microbiome Data Integration

The human gut microbiome is linked to metabolic and cardiovascular disease risk. Dietary modulation of the human gut microbiome offers an attractive pathway to manipulate the microbiome to prevent microbiome-related disease. However, this promise has not been realized. The complex system of diet and microbiome interactions is poorly understood. Integrating observational human diet and microbiome data can help researchers and clinicians untangle the complex systems of interactions that predict how the microbiome will change in response to foods. The use of dietary patterns to assess diet-microbiome relations holds promise to identify interesting associations and result in findings that can directly translate into actionable dietary intake recommendations and eating plans. In this article, we first highlight the complexity inherent in both dietary and microbiome data and introduce the approaches generally used to explore diet and microbiome simultaneously in observational studies. Second, we review the food group and dietary pattern-microbiome literature focusing on dietary complexity-moving beyond nutrients. Our review identified a substantial and growing body of literature that explores links between the microbiome and dietary patterns. However, there was very little standardization of dietary collection and assessment methods across studies. The 54 studies identified in this review used ≥7 different methods to assess diet. Coupled with the variation in final dietary parameters calculated from dietary data (e.g., dietary indices, dietary patterns, food groups, etc.), few studies with shared methods and assessment techniques were available for comparison. Third, we highlight the similarities between dietary and microbiome data structures and present the possibility that multivariate and compositional methods, developed initially for microbiome data, could have utility when applied to dietary data. Finally, we summarize the current state of the art for diet-microbiome data integration and highlight ways dietary data could be paired with microbiome data in future studies to improve the detection of diet-microbiome signals.

Evaluation of the Healthy Eating Food Index (HEFI)-2019 measuring adherence to Canada's Food Guide 2019 recommendations on healthy food choices

The objective of this study was to evaluate the construct validity and reliability of the Healthy Eating Food Index-2019 (HEFI-2019), which was developed to measure adherence to Canada's Food Guide 2019 (CFG-2019) recommendations on healthy food choices. Dietary intake data from 24-hour dietary recalls in the 2015 Canadian Community Health Survey-Nutrition were used for that purpose. Multidimensionality was examined using principal component analysis. Mean scores were compared among subgroups of the population. The association between scores and energy intake was assessed using Pearson correlations. Cronbach's alpha was calculated to assess reliability. The estimated mean HEFI-2019 score (/80) was 43.1 (95% CI, 42.7 to 43.6) among Canadians aged 2 years and older. The first and 99th percentiles were 22.1 and 62.9 points. The mean HEFI-2019 score for smokers was 7.2 points lower than for non-smokers (95% CI, -8.5 to -5.9). The HEFI-2019 was weakly correlated with energy intake (r = -0.13; 95% CI, -0.20 to -0.06). The principal components analysis revealed at least 4 dimensions. Cronbach's alpha was 0.66 (95% CI, 0.63 to 0.69). Evidence of construct validity and internal consistency support the use of the HEFI-2019 to assess adherence to CFG-2019's recommendations on healthy food choices. Novelty: Examination of the HEFI-2019's psychometric properties is needed prior to implementation. Analyses support the construct validity and internal consistency of the HEFI-2019. Interpretation of the total HEFI-2019 score must be accompanied by its components' scores, considering it assesses multiple dimensions of food choices.

Daily food insecurity is associated with diet quality, but not energy intake, in winter and during COVID-19, among low-income adults

BACKGROUND

Food insecurity (FI) is a dynamic phenomenon. Experiences of daily FI may impact dietary outcomes differently within a given month, across seasons, and before or during the COVID-19 pandemic.

OBJECTIVES

The aims of this study were to investigate the association of short-term FI with dietary quality and energy 1) over six weeks in two seasonal months and 2) before and during the COVID-19 pandemic.

METHODS

Using an ecological momentary assessment framework on smartphones, this study tracked daily FI via the 6-item U.S. Adult Food Security Survey Module and dietary intake via food diaries in 29 low-income adults. A total of 324 person-days of data were collected during two three-week long waves in fall and winter months. Generalized Estimating Equation models were applied to estimate the daily FI-diet relationship, accounting for intrapersonal variation and covariates.

RESULTS

A one-unit increase in daily FI score was associated with a 7.10-point (95%CI:-11.04,-3.15) and 3.80-point (95%CI: -6.08,-1.53) decrease in the Healthy Eating Index-2015 (HEI-2015) score in winter and during COVID-19, respectively. In winter months, a greater daily FI score was associated with less consumption of total fruit (-0.17 cups, 95% CI: -0.32,-0.02), whole fruit (-0.18 cups, 95%CI: -0.30,-0.05), whole grains (-0.57 oz, 95%CI: -0.99,-0.16) and higher consumption of refined grains (1.05 oz, 95%CI: 0.52,1.59). During COVID-19, elevated daily FI scores were associated with less intake of whole grains (-0.49 oz, 95% CI: -0.88,-0.09), and higher intake of salt (0.34 g, 95%CI: 0.15,0.54). No association was observed in fall nor during the pre-COVID-19 months. No association was found between daily FI and energy intake in either season, pre-COVID 19, or during-COVID-19 months.

CONCLUSION

Daily FI is associated with compromised dietary quality in low-income adults in winter months and during the COVID-19 period. Future research should delve into the underlying factors of these observed relationships.

Today’s Mediterranean Diet in Greece: Findings from the National Health and Nutrition Survey—HYDRIA (2013–2014)

Background: This study aimed to investigate the food and macronutrient intake of the population in Greece and evaluate its adherence to the Greek traditional Mediterranean diet. Methods: Adults over 18 years old (n = 4011) were included from the 2013–2014 National Health and Nutrition survey—HYDRIA. Dietary intake was collected using two 24-h recall interviews and a nonquantitative food frequency questionnaire. Macronutrient intakes were calculated using an updated version of the Greek FCT. Results: Only 28.3% of the adult population had high adherence to the Greek traditional Mediterranean diet, with a higher percentage (39.7%) observed for participants over 65 years compared to those under 65 years (25.5%). Differences in adherence to the MD were observed among the four geographical regions in Greece. Younger adults had a higher intake of meat, cereals, alcoholic and nonalcoholic beverages, and sugar products than older individuals who consumed more vegetables, fruits, legumes, dairy, fish, and lipids (mainly from olive oil). Adults do not meet the international dietary recommendations for the intake of several foods and macronutrients. Conclusions: The adult Greek population, especially younger people, has headed away from the Greek traditional Mediterranean diet. These observations indicate potential detrimental consequences in terms of morbidity and mortality.

The Total Nutrient Index is a Useful Measure for Assessing Total Micronutrient Exposures Among US Adults

BACKGROUND

Most dietary indices reflect foods and beverages and do not include exposures from dietary supplements (DS) that provide substantial amounts of micronutrients. A nutrient-based approach that captures total intake inclusive of DS can strengthen exposure assessment.

OBJECTIVES

We examined the construct and criterion validity of the Total Nutrient Index (TNI) among US adults (≥19 years; nonpregnant or lactating).

METHODS

The TNI includes 8 underconsumed micronutrients identified by the Dietary Guidelines for Americans: calcium; magnesium; potassium; choline; and vitamins A, C, D, and E. The TNI is expressed as a percentage of the RDA or Adequate Intake to compute micronutrient component scores; the mean of the component scores yields the TNI score, ranging from 0-100. Data from exemplary menus and the 2003-2006 (≥19 years; n = 8861) and 2011-2014 NHANES (≥19 years; n = 9954) were employed. Exemplary menus were used to determine whether the TNI yielded high scores from dietary sources (women, 31-50 years; men ≥ 70 years). TNI scores were correlated with Healthy Eating Index (HEI) 2015 overall and component scores for dairy, fruits, and vegetables; TNI component scores for vitamins A, C, D, and E were correlated with respective biomarker data. TNI scores were compared between groups with known differences in nutrient intake based on the literature.

RESULTS

The TNI yielded high scores on exemplary menus (84.8-93.3/100) and was moderately correlated (r = 0.48) with the HEI-2015. Mean TNI scores were significantly different for DS users (83.5) compared with nonusers (67.1); nonsmokers (76.8) compared with smokers (70.3); and those living with food security (76.6) compared with food insecurity (69.1). Correlations of TNI vitamin component scores with available biomarkers ranged from 0.12 (α-tocopherol) to 0.36 (serum 25-hydroxyvitamin D), and were significantly higher than correlations obtained from the diet alone.

CONCLUSIONS

The evaluation of validity supports that the TNI is a useful construct to assess total micronutrient exposures of underconsumed micronutrients among US adults.

Bayesian Copula Density Deconvolution for Zero-Inflated Data in Nutritional Epidemiology

Estimating the marginal and joint densities of the long-term average intakes of different dietary components is an important problem in nutritional epidemiology. Since these variables cannot be directly measured, data are usually collected in the form of 24-hour recalls of the intakes, which show marked patterns of conditional heteroscedasticity. Significantly compounding the challenges, the recalls for episodically consumed dietary components also include exact zeros. The problem of estimating the density of the latent long-time intakes from their observed measurement error contaminated proxies is then a problem of deconvolution of densities with zero-inflated data. We propose a Bayesian semiparametric solution to the problem, building on a novel hierarchical latent variable framework that translates the problem to one involving continuous surrogates only. Crucial to accommodating important aspects of the problem, we then design a copula based approach to model the involved joint distributions, adopting different modeling strategies for the marginals of the different dietary components. We design efficient Markov chain Monte Carlo algorithms for posterior inference and illustrate the efficacy of the proposed method through simulation experiments. Applied to our motivating nutritional epidemiology problems, compared to other approaches, our method provides more realistic estimates of the consumption patterns of episodically consumed dietary components.

Shifts towards healthy diets in the US can reduce environmental impacts but would be unaffordable for poorer minorities

Environmental implications of food choice are the focus of increasingly extensive research, but less is known about the impacts of dietary patterns of different socio-economic groups of a country, and the trade-offs between nutritional quality and environmental impacts of diet within those groups. We evaluate the impacts of US household dietary patterns on greenhouse gas emissions, blue water footprint, land use and energy consumption across supply chains using an environmentally extended input-output analysis. We compare the nutritional quality of these dietary patterns using healthy eating index scores across individuals' income and other socio-economic characteristics. Individuals with higher income or education levels are more likely to adopt healthier diets but are also responsible for larger environmental impacts of diet primarily due to a higher consumption of dairy and livestock products, seafood and items with lower energy density but higher nutrient density. Our optimization shows that a healthy diet with lower environmental impacts is achievable within current food budgets for almost 95% of people, and results in average decreases of 2% in food-related greenhouse gas emissions, 24% in land use and 4% in energy consumption, but a 28% increase in blue water consumption. However, such dietary patterns are unaffordable for 38% of Black and Hispanic individuals in the lowest income and education groups. Policies that affect income and food prices making nutritious food more affordable would be needed to achieve better nutrition and improved environmental outcomes simultaneously, particularly for more vulnerable socio-economic groups.

Association of food insecurity with dietary intakes and nutritional biomarkers among US children, National Health and Nutrition Examination Survey (NHANES) 2011-2016

BACKGROUND

Food insecurity is associated with poorer nutrient intakes from food sources and lower dietary supplement use. However, its association with total usual nutrient intakes, inclusive of dietary supplements, and biomarkers of nutritional status among US children remains unknown.

OBJECTIVE

The objective was to assess total usual nutrient intakes, Healthy Eating Index-2015 (HEI-2015) scores, and nutritional biomarkers by food security status, sex, and age among US children.

METHODS

Cross-sectional data from 9147 children aged 1-18 y from the 2011-2016 NHANES were analyzed. Usual energy and total nutrient intakes and HEI-2015 scores were estimated using the National Cancer Institute method from 24-h dietary recalls.

RESULTS

Overall diet quality was poor, and intakes of sodium, added sugars, and saturated fat were higher than recommended limits, regardless of food security status. Food-insecure girls and boys were at higher risk of inadequate intakes for vitamin D and magnesium, and girls also had higher risk for inadequate calcium intakes compared with their food-secure counterparts, when total intakes were examined. Choline intakes of food-insecure children were less likely to meet the adequate intake than those of their food-secure peers. No differences by food security status were noted for folate, vitamin C, iron, zinc, potassium, and sodium intakes. Food-insecure adolescent girls aged 14-18 y were at higher risk of micronutrient inadequacies than any other subgroup, with 92.8% (SE: 3.6%) at risk of inadequate intakes for vitamin D. No differences in biomarkers for vitamin D, folate, iron, and zinc were observed by food security status. The prevalence of iron deficiency was 12.7% in food-secure and 12.0% in food-insecure adolescent girls.

CONCLUSIONS

Food insecurity was associated with compromised intake of some micronutrients, especially among adolescent girls. These results highlight a need for targeted interventions to improve children's overall diet quality, including the reduction of specific nutrient inadequacies, especially among food-insecure children. This study was registered at clinicaltrials.gov as NCT03400436.

A review of statistical methods for dietary pattern analysis

Background

Dietary pattern analysis is a promising approach to understanding the complex relationship between diet and health. While many statistical methods exist, the literature predominantly focuses on classical methods such as dietary quality scores, principal component analysis, factor analysis, clustering analysis, and reduced rank regression. There are some emerging methods that have rarely or never been reviewed or discussed adequately.

Methods

This paper presents a landscape review of the existing statistical methods used to derive dietary patterns, especially the finite mixture model, treelet transform, data mining, least absolute shrinkage and selection operator and compositional data analysis, in terms of their underlying concepts, advantages and disadvantages, and available software and packages for implementation.

Results

While all statistical methods for dietary pattern analysis have unique features and serve distinct purposes, emerging methods warrant more attention. However, future research is needed to evaluate these emerging methods’ performance in terms of reproducibility, validity, and ability to predict different outcomes.

Conclusion

Selection of the most appropriate method mainly depends on the research questions. As an evolving subject, there is always scope for deriving dietary patterns through new analytic methodologies.

Iodine nutrition: Disorders, monitoring and policies

Iodine is an essential mineral nutrient and an integral component of thyroid hormones. Iodine deficiency is typically associated with goiter, but can have more serious health implications. Adequate iodine status is important for normal brain development. Iodine deficiency in utero or in early life can cause severe neurological and cognitive impairment. Over the last three decades, global efforts have reduced the prevalence of iodine deficiency disorders (IDD) in many areas of the world with implementation of nutrition policies and programs such as "salt" iodization. However, in a number of areas iodine deficiency is still widespread. Iodine deficiency in remote regions with high poverty will be more difficult to eradicate. Efforts to eliminate IDD in affected areas and sustaining successful iodine programs will be a priority given the substantial public health and economic benefits. A key component will be periodic monitoring of population iodine status to ensure sufficient intakes and the absence of excessive intakes. Median urinary iodine concentration (UIC), a validated biomarker for assessing population iodine status, will facilitate monitoring. Research validating "usual" UIC for use in combination with the Estimated Average Requirement cut-point method will expand its utility and allow accurate determination of the prevalence of inadequate intakes in populations. Further research on the development of biomarkers for assessment of individual iodine status for routine patient care will be important.

Relationship of added sugars intakes with physiologic parameters in adults: an analysis of national health and nutrition examination survey 2001-2012

Introduction

Consumption of added sugars (AS) has been associated with increased risk for liver disease and risk factors for cardiovascular disease. The objective of this study was to further understand the relationship of AS intake with liver enzymes and risk factors for cardiovascular disease in adults (n = 29,687) participating in the National Health and Nutrition Examination Survey (NHANES) 2001–2012.

Methods

Individual usual intake (IUI) of AS was estimated using the Markov Chain Monte Carlo ratio method using two days of 24-hour dietary recalls gathered using standardized protocols. Subjects were separated into six consumption groups: 0 to <5%, 5 to <10%, 10 to <15%, 15 to <20%, 20 to <25% and ≥25% of energy as added sugars. Linear and group trends were determined using regression analyses for liver enzymes, cardiovascular risk factors, weight parameters, glucose, metabolic syndrome, and C-reactive protein. Logistic regression calculated odds ratios for these biomarkers above established risk levels (p < 0.01).

Results

There was an inverse group trend association with AS IUI for lower body mass index (beta = −0.12 kg/m² across AS intake groups); % overweight/obese or % obese also showed inverse group trend (−1.44, −0.77, % units across AS intake groups, respectively). Liver enzymes did not show a significant association with AS IUI. Mean plasma glucose levels (mg/dL) showed an inverse linear (beta = −0.13 mg/dL per AS intake); and group association with energy from AS IUI (beta = −0.71mg/d L across AS intake groups). There was no association of AS intake with the risk of elevated uric acid levels; however, the odds of reduced uric acid levels was 17% lower with increasing intake of energy from AS IUI in group trend analysis only (0.83; 0.72–0.95 [99^th CI]).

Conclusion

Although it is sensible to consume AS in moderation, results suggested that higher intakes of AS were not consistently associated with physiologic parameters. Since the data were cross-sectional, they cannot be used to assess cause and effect. Thus, additional studies are warranted to confirm these findings with more rigorous study designs.

Comparison of Various Methods to Determine Added Sugars Intake to Assess the Association of Added Sugars Intake and Micronutrient Adequacy

Different methods for determining the effect of added sugars intake among children and adults on meeting recommended nutrient intakes were compared using 24 h dietary recall data from the National Health and Nutrition Examination Survey (NHANES) 2011-2014. Four methods were used to determine deciles of added sugars intake (as the percentage of total calories): 1 day intake, 2 day average intake, and individual usual intake (UI) determined with the National Cancer Institute (NCI) and the multivariate Markov Chain Monte Carlo methods. Percentages of the population below the Estimated Average Requirement (EAR) for calcium and vitamin D/above the Adequate Intake (AI) for potassium and dietary fiber for each decile of added sugars intake were assessed with the NCI method. Using regression analyses, added sugars intake deciles (by any method) in children were inversely associated (p < 0.001) with percentages below the EAR/above the AI of vitamin D, calcium, potassium, and fiber. In adults, added sugars intake deciles were inversely associated with meeting recommendations for vitamin D, potassium, and fiber. There were no significant between-method differences for regression coefficients for any nutrients investigated. Overall, these methods showed a similar association of added sugars intake with nutrient inadequacy/adequacy; therefore, method preference may depend more on practical reasons.

Older adults with obesity have higher risks of some micronutrient inadequacies and lower overall dietary quality compared to peers with a healthy weight, National Health and Nutrition Examination Surveys (NHANES), 2011-2014

OBJECTIVE

To evaluate total usual intakes and biomarkers of micronutrients, overall dietary quality and related health characteristics of US older adults who were overweight or obese compared with a healthy weight.

DESIGN

Cross-sectional study.

SETTING

Two 24-h dietary recalls, nutritional biomarkers and objective and subjective health characteristic data were analysed from the National Health and Nutrition Examination Survey 2011-2014. We used the National Cancer Institute method to estimate distributions of total usual intakes from foods and dietary supplements for eleven micronutrients of potential concern and the Healthy Eating Index (HEI)-2015 score.

PARTICIPANTS

Older adults aged ≥60 years (n 2969) were categorised by sex and body weight status, using standard BMI categories. Underweight individuals (n 47) were excluded due to small sample size.

RESULTS

A greater percentage of obese older adults compared with their healthy-weight counterparts was at risk of inadequate Mg (both sexes), Ca, vitamin B6 and vitamin D (women only) intakes. The proportion of those with serum 25-hydroxyvitamin D < 40 nmol/l was higher in obese (12 %) than in healthy-weight older women (6 %). Mean overall HEI-2015 scores were 8·6 (men) and 7·1 (women) points lower in obese than in healthy-weight older adults. In addition, compared with healthy-weight counterparts, obese older adults were more likely to self-report fair/poor health, use ≥ 5 medications and have limitations in activities of daily living and cardio-metabolic risk factors; and obese older women were more likely to be food-insecure and have depression.

CONCLUSIONS

Our findings suggest that obesity may coexist with micronutrient inadequacy in older adults, especially among women.

Semiparametric Estimation of the Distribution of Episodically Consumed Foods Measured With Error

Dietary data collected from 24-hour dietary recalls are observed with significant measurement errors. In the nonparametric curve estimation literature, much of the effort has been devoted to designing methods that are consistent under contamination by noise, and which have been traditionally applied for analyzing those data. However, some foods such as alcohol or fruits are consumed only episodically, and may not be consumed during the day when the 24-hour recall is administered. These so-called excess zeros make existing nonparametric estimators break down, and new techniques need to be developed for such data. We develop two new consistent semiparametric estimators of the distribution of such episodically consumed food data, making parametric assumptions only on some less important parts of the model. We establish its theoretical properties and illustrate the good performance of our fully data-driven method in simulated and real data. Supplementary materials for this article are available online.

STRATOS guidance document on measurement error and misclassification of variables in observational epidemiology: Part 2-More complex methods of adjustment and advanced topics

We continue our review of issues related to measurement error and misclassification in epidemiology. We further describe methods of adjusting for biased estimation caused by measurement error in continuous covariates, covering likelihood methods, Bayesian methods, moment reconstruction, moment-adjusted imputation, and multiple imputation. We then describe which methods can also be used with misclassification of categorical covariates. Methods of adjusting estimation of distributions of continuous variables for measurement error are then reviewed. Illustrative examples are provided throughout these sections. We provide lists of available software for implementing these methods and also provide the code for implementing our examples in the Supporting Information. Next, we present several advanced topics, including data subject to both classical and Berkson error, modeling continuous exposures with measurement error, and categorical exposures with misclassification in the same model, variable selection when some of the variables are measured with error, adjusting analyses or design for error in an outcome variable, and categorizing continuous variables measured with error. Finally, we provide some advice for the often met situations where variables are known to be measured with substantial error, but there is only an external reference standard or partial (or no) information about the type or magnitude of the error.

Best (but oft-forgotten) practices: sample size and power calculation for a dietary intervention trial with episodically consumed foods

Dietary interventions often target foods that are underconsumed relative to dietary guidelines, such as vegetables, fruits, and whole grains. Because these foods are only consumed episodically for some participants, data from such a study often contains a disproportionally large number of zeros due to study participants who do not consume any of the target foods on the days that dietary intake is assessed, thus generating semicontinuous data. These zeros need to be properly accounted for when calculating sample sizes to ensure that the study is adequately powered to detect a meaningful intervention effect size. Nonetheless, this issue has not been well addressed in the literature. Instead, methods that are common for continuous outcomes are typically used to compute the sample sizes, resulting in a substantially under- or overpowered study. We propose proper approaches to calculating the sample size needed for dietary intervention studies that target episodically consumed foods. Sample size formulae are derived for detecting the mean difference in the amount of intake of an episodically consumed food between an intervention and a control group. Numerical studies are conducted to investigate the accuracy of the sample size formulae as compared with the ad hoc methods. The simulation results show that the proposed formulae are appropriate for estimating the sample sizes needed to achieve the desired power for the study. The proposed method for sample size is recommended for designing dietary intervention studies targeting episodically consumed foods.

Application of the Healthy Eating Index-2015 and the Nutrient-Rich Food Index 9.3 for assessing overall diet quality in the Japanese context: Different nutritional concerns from the US

Objectives

While it is widely perceived that the diet consumed by Japanese is healthy, empirical evidence supporting this notion is limited. In this cross-sectional study, we assessed the overall diet quality of Japanese using the Healthy Eating Index-2015 (HEI-2015) and Nutrient-Rich Food Index 9.3 (NRF9.3), and compared diet quality scores between Japanese and Americans.

Methods

We used 1-d dietary record data from 19,719 adults (aged ≥20 y) in the Japanese National Health and Nutrition Survey 2012 and the first 24-h dietary recall data from 4614 adults in the US NHANES 2011–2012.

Results

As expected, a higher total score of the HEI-2015 and NRF9.3 was associated with favorable patterns of overall diet in the Japanese population. The range of total score was wide enough for both HEI-2015 (5th percentile 37.2; 95th percentile 67.2) and NRF9.3 (5th percentile 257; 95th percentile 645). Both HEI-2015 and NRF9.3 distinguished known differences in diet quality between sex, age, and smoking status. The mean total scores of HEI-2015 and NRF9.3 were similar between Japanese (51.9 and 448, respectively) and US adults (52.8 and 435, respectively). However, component scores between the 2 populations were considerably different. For HEI-2015, Japanese had higher scores for whole fruits, total vegetables, green and beans, total protein foods, seafood and plant proteins, fatty acids, added sugars, and saturated fats, but lower scores for total fruits, whole grains, dairy, refined grains, and sodium. For NRF9.3, the intakes of vitamin C, vitamin D, potassium, added sugars, and saturated fats were more favorable in Japanese, while those of dietary fiber, vitamin A, calcium, iron, magnesium, and sodium were less favorable.

Conclusions

This study suggests the usefulness of HEI-2015 and NRF9.3 for assessing the diet quality of Japanese, as well as for highlighting different nutritional concerns between Japan and the US.

Daily Sampling Reveals Personalized Diet-Microbiome Associations in Humans

Diet is a key determinant of human gut microbiome variation. However, the fine-scale relationships between daily food choices and human gut microbiome composition remain unexplored. Here, we used multivariate methods to integrate 24-h food records and fecal shotgun metagenomes from 34 healthy human subjects collected daily over 17 days. Microbiome composition depended on multiple days of dietary history and was more strongly associated with food choices than with conventional nutrient profiles, and daily microbial responses to diet were highly personalized. Data from two subjects consuming only meal replacement beverages suggest that a monotonous diet does not induce microbiome stability in humans, and instead, overall dietary diversity associates with microbiome stability. Our work provides key methodological insights for future diet-microbiome studies and suggests that food-based interventions seeking to modulate the gut microbiota may need to be tailored to the individual microbiome. Trial Registration: ClinicalTrials.gov: NCT03610477.

Breakfast Consumption Is Positively Associated with Usual Nutrient Intakes among Food Pantry Clients Living in Rural Communities

BACKGROUND

Breakfast consumption has declined over the past 40 y and is inversely associated with obesity-related diet and health outcomes. The breakfast pattern of food pantry clients and its association with diet is unknown.

OBJECTIVE

The objective is to investigate the association of breakfast consumption with diet quality and usual nutrient intakes among food pantry clients (n = 472) living in rural communities.

METHODS

This was an observational study using cross-sectional analyses. English-speaking participants ≥18 y (or ≥19 y in Nebraska) were recruited from 24 food pantries in rural high-poverty counties in Indiana, Michigan, Missouri, Nebraska, Ohio, and South Dakota. Participants were surveyed at the pantry regarding characteristics and diet using 24-h recall. A second recall was self-completed or completed via assisted phone call within 2 wk of the pantry visit. Participants were classified as breakfast skippers when neither recall reported breakfast ≥230 kcal consumed between 04:00 and 10:00; breakfast consumers were all other participants. The Healthy Eating Index-2010 was modeled with breakfast pattern using multiple linear regression. Mean usual intake of 16 nutrients was estimated using the National Cancer Institute Method and compared across breakfast pattern groups. Usual nutrient intake was compared with the Estimated Average Requirement (EAR) or Adequate Intake (AI) to estimate the proportion of population not meeting the EAR or exceeding the AI.

RESULTS

A total of 56% of participants consumed breakfast. Compared with breakfast skippers, breakfast consumers had 10-59% significantly higher usual mean intakes of all nutrients (P ≤ 0.05), and had 12-21% lower prevalence of at-risk nutrient intakes except for vitamin D, vitamin E, and magnesium.

CONCLUSIONS

Adult food pantry clients living in rural communities experienced hardships in meeting dietary recommendations. Breakfast consumption was positively associated with usual nutrient intakes in this population. This trial was registered at clinicaltrials.gov as NCT03566095.

Evaluation of the Healthy Eating Index-2015

BACKGROUND

The Healthy Eating Index (HEI), a diet quality index that measures alignment with the Dietary Guidelines for Americans, was updated with the 2015-2020 Dietary Guidelines for Americans.

OBJECTIVE AND DESIGN

To evaluate the psychometric properties of the HEI-2015, eight questions were examined: five relevant to construct validity, two related to reliability, and one to assess criterion validity.

DATA SOURCES

Three data sources were used: exemplary menus (n=4), National Health and Nutrition Examination Survey 2011-2012 (N=7,935), and the National Institutes of Health-AARP (formally known as the American Association of Retired Persons) Diet and Health Study (N=422,928).

STATISTICAL ANALYSES

Exemplary menus: Scores were calculated using the population ratio method. National Health and Nutrition Examination Survey 2011-2012: Means and standard errors were estimated using the Markov Chain Monte Carlo approach. Analyses were stratified to compare groups (with t tests and analysis of variance). Principal components analysis examined the number of dimensions. Pearson correlations were estimated between components, energy, and Cronbach's coefficient alpha. National Institutes of Health-AARP Diet and Health Study: Adjusted Cox proportional hazards models were used to examine scores and mortality outcomes.

RESULTS

For construct validity, the HEI-2015 yielded high scores for exemplary menus as four menus received high scores (87.8 to 100). The mean score for National Health and Nutrition Examination Survey was 56.6, and the first to 99th percentile were 32.6 to 81.2, respectively, supporting sufficient variation. Among smokers, the mean score was significantly lower than among nonsmokers (53.3 and 59.7, respectively) (P<0.01), demonstrating differentiation between groups. The correlation between diet quality and diet quantity was low (all <0.25) supporting these elements being independent. The components demonstrated multidimensionality when examined with a scree plot (at least four dimensions). For reliability, most of the intercorrelations among the components were low to moderate (0.01 to 0.49) with a few exceptions, and the standardized Cronbach's alpha was .67. For criterion validity, the highest vs the lowest quintile of HEI-2015 scores were associated with a 13% to 23% decreased risk of all-cause, cancer, and cardiovascular disease mortality.

CONCLUSIONS

The results demonstrated evidence supportive of construct validity, reliability, and criterion validity. The HEI-2015 can be used to examine diet quality relative to the 2015-2020 Dietary Guidelines for Americans.

Applications of the Healthy Eating Index for Surveillance, Epidemiology, and Intervention Research: Considerations and Caveats

The Healthy Eating Index (HEI) is a measure of diet quality that can be used to examine alignment of dietary patterns with the Dietary Guidelines for Americans. The HEI is made up of multiple adequacy and moderation components, most of which are expressed relative to energy intake (ie, as densities) for the purpose of calculating scores. Due to these characteristics and the complexity of dietary intake data more broadly, calculating and using HEI scores can involve unique statistical considerations and, depending on the particular application, intensive computational methods. The objective of this article is to review potential applications of the HEI, including those relevant to surveillance, epidemiology, and intervention research, and to summarize available guidance for appropriate analysis and interpretation. Steps in calculating HEI scores are reviewed and statistical methods described. Consideration of salient issues in the calculation and interpretation of scores can help researchers avoid common pitfalls and reviewers ensure that articles reporting on the use of the HEI include sufficient details such that the work is comprehensible and replicable, with the overall goal of contributing to knowledge on dietary patterns and health among Americans.

Combining biomarker and food intake data: calibration equations for citrus intake

BACKGROUND

Measurement error associated with self-reported dietary intake is a well-documented issue. Combining biomarkers of food intake and dietary intake data is a high priority.

OBJECTIVES

The aim of this study was to develop calibration equations for food intake, illustrated with an application for citrus intake. Further, a simulation-based framework was developed to determine the portion of biomarker data needed for stable calibration equation estimation in large population studies.

METHODS

Calibration equations were developed using mean daily self-reported citrus intake (4-d semiweighed food diaries) and biomarker-derived intake (urinary proline betaine biomarker) data from participants (n = 565) as part of a cross-sectional study. Different functional specifications and biomarker transformations were tested to derive the optimal calibration equation specifications. The simulation study was developed using linear regression for the calibration equations. Stability in the calibration equation estimations was investigated for varying portions of biomarker and intake data "qualities."

RESULTS

With citrus intake, linear regression on nontransformed biomarker data resulted in the optimal calibration equation specifications and produced good-quality predicted intakes. The lowest mean squared error (14,354) corresponded to a linear regression model, defined with biomarker-derived estimates of intakes on the original scale. Using this model in a subpopulation without biomarker data resulted in an average mean ± SD citrus intake of 81 ± 66 g/d. The simulation study suggested that in large population studies, biomarker data on 20-30% of the subjects are required to guarantee stable estimation of calibration equations. This article is accompanied by a web application ("Bio-Intake"), which was developed to facilitate measurement error correction in self-reported mean daily citrus intake data.

CONCLUSIONS

Calibration equations proved to be a useful instrument to correct measurement error in self-reported food intake data. The simulation study demonstrated that the use of food intake biomarkers may be feasible and beneficial in the context of large population studies.

Early Experience Analyzing Dietary Intake Data from the Canadian Community Health Survey-Nutrition Using the National Cancer Institute (NCI) Method

Background: One of the underpinning elements to support evidence-based decision-making in food and nutrition is the usual dietary intake of a population. It represents the long-run average consumption of a particular dietary component (i.e., food or nutrient). Variations in individual eating habits are observed from day-to-day and between individuals. The National Cancer Institute (NCI) method uses statistical modeling to account for these variations in estimation of usual intakes. This method was originally developed for nutrition survey data in the United States. The main objective of this study was to apply the NCI method in the analysis of Canadian nutrition surveys. Methods: Data from two surveys, the 2004 and 2015 Canadian Community Health Survey—Nutrition were used to estimate usual dietary intake distributions from food sources using the NCI method. The effect of different statistical considerations such as choice of the model, covariates, stratification compared to pooling, and exclusion of outliers were assessed, along with the computational time to convergence. Results: A flowchart to aid in model selection was developed. Different covariates (e.g., age/sex groups, cycle, weekday/weekend of the recall) were used to adjust the estimates of usual intakes. Moreover, larger differences in the ratio of within to between variation for a stratified analysis or a pooled analysis resulted in noticeable differences, particularly in the tails of the distribution of usual intake estimates. Outliers were subsequently removed when the ratio was larger than 10. For an individual age/sex group, the NCI method took 1 h–5 h to obtain results depending on the dietary component. Conclusion: Early experience in using the NCI method with Canadian nutrition surveys data led to the development of a flowchart to facilitate the choice of the NCI model to use. The ability of the NCI method to include covariates permits comparisons between both 2004 and 2015. This study shows that the improper application of pooling and stratification as well as the outlier detection can lead to biased results. This early experience can provide guidance to other researchers and ensures consistency in the analysis of usual dietary intake in the Canadian context.

How Well Do U.S. Hispanics Adhere to the Dietary Guidelines for Americans? Results from the Hispanic Community Health Study/Study of Latinos

Purpose: To examine alignment between 2010 Dietary Guidelines for Americans (DGA) and dietary choices of individuals in the Hispanic Community Health Study/Study of Latinos (HCHS/SOL) between 2008 and 2011.

Methods: Data are from 15,633 adults 18–74 years from the population-based cohort in HCHS/SOL. The Healthy Eating Index (HEI) 2010 was used to measure diet quality. Means and standard errors (SEs) for the HEI total and each of the 12 component scores were calculated overall and by Hispanic/Latino heritage, sex, age group, and measures of acculturation. Linear regression was used to examine correlates of the HEI 2010 total score. All analyses accounted for complex survey design.

Results: The overall HEI mean of 63.8 (SE: 0.4) varied across groups from a high (healthier diet) of 71 (SE: 0.9) among Mexicans to a low of 56 (SE: 0.1) among Puerto Ricans. The proportion with a maximum score for the HEI components varied across heritage groups; >25% of adults adhered to recommendations for total proteins, and seafood and plant proteins, whole fruits, and greens and beans, with the exception of Cubans and Puerto Ricans, who had lower adherence scores for the latter two. The components with the lowest adherence were sodium (<2%) and fatty acids (overall 7.4%) among all heritage groups. Characteristics associated with better adherence included sociodemographic variables, Spanish language preference, weight status, medical conditions, and lifestyle behaviors.

Conclusions: Individuals with Mexican, Dominican, and Central American heritage had better overall dietary quality compared to other groups. However, all can improve their eating habits to align more with the DGAs by reducing sodium consumption and improving fatty acid ratios.

Robust Clustering with Subpopulation-specific Deviations

The National Birth Defects Prevention Study (NBDPS) is a case-control study of birth defects conducted across 10 U.S. states. Researchers are interested in characterizing the etiologic role of maternal diet, collected using a food frequency questionnaire. Because diet is multi-dimensional, dimension reduction methods such as cluster analysis are often used to summarize dietary patterns. In a large, heterogeneous population, traditional clustering methods, such as latent class analysis, used to estimate dietary patterns can produce a large number of clusters due to a variety of factors, including study size and regional diversity. These factors result in a loss of interpretability of patterns that may differ due to minor consumption changes. Based on adaptation of the local partition process, we propose a new method, Robust Profile Clustering, to handle these data complexities. Here, participants may be clustered at two levels: (1) globally, where women are assigned to an overall population-level cluster via an overfitted finite mixture model, and (2) locally, where regional variations in diet are accommodated via a beta-Bernoulli process dependent on subpopulation differences. We use our method to analyze the NBDPS data, deriving pre-pregnancy dietary patterns for women in the NBDPS while accounting for regional variability.

Combining biomarker and self-reported dietary intake data: A review of the state of the art and an exposition of concepts

Classical approaches to assessing dietary intake are associated with measurement error. In an effort to address inherent measurement error in dietary self-reported data there is increased interest in the use of dietary biomarkers as objective measures of intake. Furthermore, there is a growing consensus of the need to combine dietary biomarker data with self-reported data. A review of state of the art techniques employed when combining biomarker and self-reported data is conducted. Two predominant methods, the calibration method and the method of triads, emerge as relevant techniques used when combining biomarker and self-reported data to account for measurement errors in dietary intake assessment. Both methods crucially assume measurement error independence. To expose and understand the performance of these methods in a range of realistic settings, their underpinning statistical concepts are unified and delineated, and thorough simulation studies are conducted. Results show that violation of the methods' assumptions negatively impacts resulting inference but that this impact is mitigated when the variation of the biomarker around the true intake is small. Thus there is much scope for the further development of biomarkers and models in tandem to achieve the ultimate goal of accurately assessing dietary intake.

Correcting for measurement error in fractional polynomial models using Bayesian modelling and regression calibration, with an application to alcohol and mortality

Exposure measurement error can result in a biased estimate of the association between an exposure and outcome. When the exposure-outcome relationship is linear on the appropriate scale (e.g. linear, logistic) and the measurement error is classical, that is the result of random noise, the result is attenuation of the effect. When the relationship is non-linear, measurement error distorts the true shape of the association. Regression calibration is a commonly used method for correcting for measurement error, in which each individual's unknown true exposure in the outcome regression model is replaced by its expectation conditional on the error-prone measure and any fully measured covariates. Regression calibration is simple to execute when the exposure is untransformed in the linear predictor of the outcome regression model, but less straightforward when non-linear transformations of the exposure are used. We describe a method for applying regression calibration in models in which a non-linear association is modelled by transforming the exposure using a fractional polynomial model. It is shown that taking a Bayesian estimation approach is advantageous. By use of Markov chain Monte Carlo algorithms, one can sample from the distribution of the true exposure for each individual. Transformations of the sampled values can then be performed directly and used to find the expectation of the transformed exposure required for regression calibration. A simulation study shows that the proposed approach performs well. We apply the method to investigate the relationship between usual alcohol intake and subsequent all-cause mortality using an error model that adjusts for the episodic nature of alcohol consumption.

Usual diet quality among 8- to 12-year-old Brazilian children

Abstract: Nutritional surveys are important information sources for public policy in the food and nutrition field. They focus on assessing usual dietary patterns, because health outcomes result from the long-term intake. Here we aimed to evaluate diet quality adjusted for day-to-day variance among Brazilian children. Data were collected between March 2013 and August 2015. The sample included 8- to 12-year-old children (n = 1,357) from public schools from all administrative regions of a Brazilian city. One 24-h dietary recall (24HR) was collected for the whole sample and two 24HR for two non-consecutive days of the same week for a subsample. The Healthy Eating Index-2010 (HEI-2010) was adapted to Brazilian food habits and the Brazilian dietary guidelines were used to evaluate diet quality. Statistical analysis included a multipart, nonlinear mixed model with correlated random effects proposed by the U.S. National Cancer Institute to correct diet quality for day-to-day variance. The adapted HEI-2010 total score was 51.8. Children with poorer diet quality (< 10th percentile) scored less than 41.1, and children with higher diet quality (> 90th percentile) scored more than 62.4. The overall adequacy of adapted HEI-2010 components was low. Higher adequacy percentages were identified for total protein foods (94.9%), greens (62.3%), and seafood and plant proteins (52.2%). Seven components showed less than 10% of adequacy: refined grains, fatty acids, dairy, sodium, total vegetable, whole grains, and empty calories. This study identified the main inadequacies among children’s diet quality, which can guide promotion actions for healthy eating.

Clustering in General Measurement Error Models

This paper is dedicated to the memory of Peter G. Hall. It concerns a deceptively simple question: if one observes variables corrupted with measurement error of possibly very complex form, can one recreate asymptotically the clusters that would have been found had there been no measurement error? We show that the answer is yes, and that the solution is surprisingly simple and general. The method itself is to simulate, by computer, realizations with the same distribution as that of the true variables, and then to apply clustering to these realizations. Technically, we show that if one uses K-means clustering or any other risk minimizing clustering, and a multivariate deconvolution device with certain smoothness and convergence properties, then, in the limit, the cluster means based on our method converge to the same cluster means as if there is no measurement error. Along with the method and its technical justification, we analyze two important nutrition data sets, finding patterns that make sense nutritionally.

Epidemiologic analyses with error-prone exposures: review of current practice and recommendations

PURPOSE

Variables in observational studies are commonly subject to measurement error, but the impact of such errors is frequently ignored. As part of the STRengthening Analytical Thinking for Observational Studies Initiative, a task group on measurement error and misclassification seeks to describe the current practice for acknowledging and addressing measurement error.

METHODS

Task group on measurement error and misclassification conducted a literature survey of four types of research studies that are typically impacted by exposure measurement error: (1) dietary intake cohort studies, (2) dietary intake population surveys, (3) physical activity cohort studies, and (4) air pollution cohort studies.

RESULTS

The survey revealed that while researchers were generally aware that measurement error affected their studies, very few adjusted their analysis for the error. Most articles provided incomplete discussion of the potential effects of measurement error on their results. Regression calibration was the most widely used method of adjustment.

CONCLUSIONS

Methods to correct for measurement error are available but require additional data regarding the error structure. There is a great need to incorporate such data collection within study designs and improve the analytical approach. Increased efforts by investigators, editors, and reviewers are needed to improve presentation of research when data are subject to error.