Characterizing energy intake misreporting and its effects on intake estimations, in the Portuguese adult population

Eating out of home in Portugal: characterisation and effects on dietary intake

This cross-sectional study aims to describe and compare energy, nutrient intake and food consumption according to eating location and by age groups using data from the National Food, Nutrition and Physical Activity Survey (IAN-AF 2015/2016). Dietary intake was estimated by two non-consecutive days of food diaries (children)/24-h recalls (other age groups), and four eating location categories were defined according to the proportion of meals consumed at out-of-home locations: Home (at least 80 % of meals at home), Other Homes, School or Work and Restaurants and Other Places. The majority of meals (69·1 %) were consumed at home. Meals were also often taken at school by children and adolescents and in restaurants and similar outlets by adults and elderly. Children and adolescents in the School or Work category ate more fruit, vegetables and pulses and cereals and starchy tubers, whereas adults in this category ate more red and processed meats, sugar-sweetened beverages and sweets. Compared with Home category, Restaurants and Other Places was associated with worse diet adequacies among children (β = -1·0; 95 % CI = -2·0, -0·04), adolescents: (β = -2·4; 95 % CI = -3·2, -1·5) and adults (β = -1·3; 95 % CI = -1·6, -1·0) reflecting higher intakes of energy, fat, trans-fatty acids and SFA, and Na. The elderly consumed more free sugars and fat when eating out of home in general. Overall, findings reflect important variation in nutrient profiles by eating location, with meals taken at school or work contributing to higher consumption of nutrient-dense foods and those taken in restaurants and other similar settings implying higher consumption of energy-dense foods.

Comparing self-reported energy intake using an online dietary tool with energy expenditure by an activity tracker

OBJECTIVE

The aim of this study was to compare self-reported total energy intake (TEI) collected using an online multiple-pass 24-h dietary recall tool (Intake24) with total energy expenditure (TEE) estimated from Fitbit Charge 2-improved algorithms in adults from the NoHoW trial (12-mo weight maintenance after free-living weight loss).

METHODS

Bland-Altman plots were used to assess the level of agreement between TEI and TEE at baseline and after 12 mo. The ratio of TEI to TEE was also calculated.

RESULTS

Data from 1323 participants (71% female) was included in the analysis (mean ± SD: age 45 ± 12 y, body mass index 29.7 ± 5.4 kg/m2, initial weight loss 11.5 ± 6.5 kg). The TEI was lower than TEE on average by 33%, with limits of agreement ranging from -91% to +25%. Men, younger individuals, those with higher body mass index, those with the greater weight loss before enrollment, and those who gained weight during the study underestimated to a greater extent.

CONCLUSIONS

These findings contribute to the ongoing research examining the validity of technology-based dietary assessment tools.

Beef Intake Is Associated with Higher Nutrient Intake and Nutrient Adequacy in U.S. Adolescents, NHANES 2001-2018

Nutrient adequacy among adolescents is of concern due to higher nutrient requirements for their developing bodies as well as the gap between the current nutrient intake and the recommendations. The objective of this study was to determine beef intake and assess the relationship between beef consumption and nutrient intake and nutrient adequacy in male and female adolescents, 14-18 years of age. Dietary recalls collected during the What We Eat in America (WWEIA) portion of the National Health and Nutrition Examination Survey (NHANES) cycles 2001-2018 were utilized to determine beef intake. Usual nutrient intakes were determined with the National Cancer Institute method in conjunction with day 1 and day 2 total nutrient files. Nutrient adequacy was assessed by calculating the percentage of the population below the estimated average requirement (EAR) or above the adequate intake (AI). The average beef intake of male and female adolescent beef consumers was 57.9 ± 2.4 and 46.8 ± 2.2 g with a 90th percentile of 82.3 ± 4.3 and 67.8 ± 3.5 g, respectively. Compared to non-consumers, beef consumers had a 10% or higher intake of calcium, iron, phosphorus, potassium, total choline, vitamin B12, and zinc. Over 50% of the adolescent population (regardless of beef consumption) had intakes below the EAR for calcium, magnesium, and vitamins A, C, D, and E. The percentage of the beef-consuming population below the EAR was lower for calcium, copper, folate, iron, phosphorus, zinc, and vitamins B12 and B6 as compared to non-consumers. Additionally, the portion of the population above the AI for sodium was higher in female beef consumers as compared to non-consumers. We estimate approximately 900,000 to 1,400,000, 400,000-700,000, 200,000-600,000, and 200,000-400,000 fewer adolescents to be below the EAR for zinc, phosphorus, vitamin B12, and iron, respectively if beef non-consumers were to consume beef. This study suggests beef can help increase the nutrient intake and nutrient adequacy in the diets of adolescents, helping to close important gaps for this nutritionally vulnerable population. While recommendations to reduce beef intake are widely prevalent, this could result in unintended nutritional consequences regarding under-consumed nutrients including those of public health concern important for adolescent health.

Methodological approaches for the assessment of bisphenol A exposure

Bisphenol A (BPA) is an endocrine disruptor used in food contact materials, by the application of polycarbonate plastics and epoxy resins. The main objective of this study is to compare the estimate of daily BPA exposure at 13 years of age and in the adult Portuguese population, using different methodological approaches, and assess the associations between this exposure and sociodemographic characteristics.

METHODOLOGY

Cross-sectional data of 13-years follow-up from a population-based birth cohort Generation XXI (GXXI) (n = 2804) and from the National Food, Nutrition and Physical Activity Survey (IAN-AF 2015-2016) (n = 3845, ≥18 years old) was used. Dietary information was collected through three food diaries for adolescents and two non-consecutive 24-hour-recalls for adults. To estimate the daily exposure to BPA, three methodological approaches were used. "Food groups attribution" merged the food consumption data with the concentration of BPA in food groups. "Regression tree model" and "random forest" combined food consumption information with urinary BPA, measured in a subsample of 24-hour urine (in adolescents n = 216, and in adults n = 82), both used to predict BPA exposure in the remaining sample. The fit-index of the methodologies was assessed through the root mean square error (RMSE), mean absolute error (MAE) and Spearman correlation coefficient (ρ). Associations between BPA exposure and sociodemographic variables were tested by linear regression models, adjusted for sex, age groups (in adults) and educational level. Tolerable Daily Intake (TDI) of 0.2 ng/kg body weight (bw), recently proposed by the European Food Safety Authority (EFSA), was used for the risk characterization of BPA exposure.

RESULTS

The "random forest" was found as the best methodology to estimate the daily BPA exposure (adolescents: RMSE = 0.989, MAE = 0.727, ρ = 0.168; adults: RMSE = 0.193, MAE = 0.147, ρ = 0.250). The median dietary BPA exposure, calculated by "food groups attribution", was 79.1 and 46.1 ng/kg bw/day for adolescents and adults, respectively, while "random forest" estimated a BPA exposure of 26.7 and 38.0 ng/kg bw/day. 99.9% of the Portuguese population presented a daily exposure above TDI. Male adolescents, females and higher educated adults, were those more exposed to BPA.

CONCLUSIONS

The estimated daily BPA exposure strongly depends on the methodological approach. Food groups attribution may overestimate the exposure while the random forest appears to be a better methodological approach to estimate BPA exposure. Nevertheless, for all methods, the Portuguese population presented an unsafe BPA exposure by largely exceeding the safe levels proposed by EFSA.

Validity and reproducibility of a food frequency questionnaire assessing food group intake in the PERSIAN Cohort Study

Purpose

A semi-quantitative food frequency questionnaire (FFQ) was developed for use in the Prospective Epidemiological Research Studies in IrAN (PERSIAN Cohort), investigating non-communicable disease risk factors. This study aimed to assess the validity and reproducibility of this FFQ, through food group intake.

Methods

Participants, recruited from seven PERSIAN cohort centers, completed the FFQ at the beginning of the study (FFQ1) and at the end (FFQ2), with a 12-month interval in between, during which two 24-h dietary recalls (24 h) were completed each month. Correlation coefficients of the median intake of food groups recorded by the FFQs were compared to those of the 24 h to assess validity, and the two FFQs were compared to assess reproducibility of findings.

Results

Overall, data from 978 participants were included in this validation analysis. Of the 26 food groups assessed, Tea, Sugars, Whole/Refined Grains, and Solid Fats/Oils, had the strongest correlations (0.6-0.79), while Red Meat, Chicken and Eggs showed moderate correlations (0.42-0.59). The weakest correlations observed belonged to Fresh fruit Juice and Other Meats (0.23-0.32). Reproducibility was assessed among those who completed both FFQ1 and FFQ2 (n = 848), revealing moderate to strong correlations in all food groups, ranging from 0.42 in Legumes to 0.72 in both Sugar and Sweetened Drinks.

Conclusion

The PERSIAN Cohort FFQ is appropriate to rank individuals based on food group intake.

Usual dietary intake, physical activity, weight loss, and body composition after five years of Roux-en-Y gastric bypass

OBJECTIVE

To estimate usual dietary intake (UDI), physical activity (PA), and their association with weight loss and body composition in patients who underwent Roux-en-Y gastric bypass (RYGB) after five years in the Federal District, Brazil.

METHODS

We assessed anthropometry and body composition using bioimpedance, and dietary intake and PA with three nonconsecutive 24-h recalls. PC-Side was used to estimate UDI. Dietary patterns (DPs) were identified through principal component analysis, and association between UDI and PA with percentage of total weight loss (%TWL) and fat-free mass (FFM) through multinomial logistic regression.

RESULTS

Sample (n = 124) presented mean (SD) age of 48.9 (9.4) years, median (IQR) of 9 years (7-10) post RYGB, current BMI = 32.3 kg/m² (28.8-35.7), %TWL = 24.7% (10.9), and FFM = 45.1 kg (41.1-51.9). Mean usual energy intake of 1556 kcal/d, with adequate protein intake, poor fiber intake, and excessive carbohydrate, total fat, and added sugar intake, compared to dietary guidelines. Calcium, vitamins C, D, and E presented the greatest inadequacy (15%, 24%, 32%, and 49% of individuals, respectively, reported usual intake below EAR); 83 participants were considered active/very active, according PA. DP with high energy, protein, total fat, saturated fat, and sodium intake, was negatively associated with %TWL (OR = 0.545, p = 0.037). Protein intake was positively associated with FFM (OR = 1.091, p = 0.004). PA was not associated with %TWL or FFM.

CONCLUSION

Participants demonstrated intake of carbohydrate, fat, fiber, added sugar not in accordance with guidelines. A DP rich in energy, protein, total fat, saturated fat, and sodium appears to decrease TWL. However, protein intake appears to increase FFM.

Lower Energy-Adjusted Nutrient Intakes Occur Among Food Energy Under-Reporters With Poor Mental Health

Background

Food energy under-reporting is differentially distributed among populations. Currently, little is known about how mental health state may affect energy-adjusted nutrient intakes among food energy under-reporters.

Methods

Stratified analysis of energy-adjusted nutrient intake by mental health (poor vs. good) and age/sex was conducted using data from Canadian Community Health Survey (CCHS) respondents (14–70 years; n = 8,233) who were deemed as under-reporters based on Goldberg's cutoffs.

Results

Most were experiencing good mental health (95.2%). Among those reporting poor mental health, significantly lower energy-adjusted nutrient intakes tended to be found for fiber, protein, vitamins A, B₂, B₃, B₆, B₉, B₁₂, C, and D, and calcium, potassium, and zinc (probability measures (p) < 0.05). For women (51–70 years), all micronutrient intakes, except iron, were significantly lower among those reporting poor mental health (p < 0.05). For men (31–50 years), B vitamin and most mineral intakes, except sodium, were significantly lower among those reporting poor mental health (p < 0.05). Among women (31–50 years) who reported poor mental health, higher energy-adjusted intakes were reported for vitamin B₉ and phosphorus (p < 0.05).

Conclusions

Among food energy under-reporters, poor mental health tends to lower the report of specific energy-adjusted nutrient intakes that include ones critical for mental health. Future research is needed to discern if these differences may be attributed to deviations in the accurate reports of food intakes, measurement errors, or mental health states.

Risk characterization of dietary acrylamide exposure and associated factors in the Portuguese population

Acrylamide exposure, mainly resulting from food cooking and processing, has been associated with a higher risk of health problems, due to genotoxic effects. This study aims to estimate acrylamide dietary exposure of the Portuguese population and its associated factors. Dietary data collected through 2 non-consecutive 24 hour recalls or food diaries from a representative sample of the Portuguese population from the National Food, Nutrition and Physical Activity Survey was used (n = 5811; 3-84 years). Occurrence data of acrylamide in food were obtained from EFSA. The margins of exposure (MOE) were calculated for peripheral neuropathy and neoplastic effects. The association between acrylamide and socio-demographic characteristics was estimated through linear regression models. For the total population, the estimated median daily dietary exposure per body weight to acrylamide was 0.38 µg/kg/day, ranging from 0.14 to 0.88 µg/kg/day for the 5^th and 95^th percentile, respectively. Children aged between 1-2 years had the highest acrylamide exposure (median 0.75 µg/kg/day, 95^th percentile 1.41 µg/kg/day). For the peripheral neuropathy and neoplastic effects, the median MOE estimated was 1140 and 451, respectively. Men compared to women had a higher acrylamide dietary exposure, as well as smokers compared to non-smokers. Elderly and less educated individuals were inversely associated with acrylamide exposure. 'Bread and rusks' (24.2%) were the main source of acrylamide, followed by 'coffee' (21.3%). The current dietary exposure to acrylamide in the Portuguese population is of concern mainly regarding neoplastic effects. Our results point to the need to reduce exposure to acrylamide, especially in men, young children, higher educated individuals and smokers.

Identification of psychological correlates of dietary misreporting under laboratory and free-living environments

Errors inherent in self-reported measures of energy intake (EI) are substantial and well documented, but correlates of misreporting remain unclear. Therefore, potential predictors of misreporting were examined. In Study One, fifty-nine individuals (BMI = 26·1 (sd 3·8) kg/m2, age = 42·7 (sd 13·6) years, females = 29) completed a 14-d stay in a residential feeding behaviour suite where eating behaviour was continuously monitored. In Study Two, 182 individuals (BMI = 25·7 (sd 3·9) kg/m2, age = 42·4 (sd 12·2) years, females = 96) completed two consecutive days in a residential feeding suite and five consecutive days at home. Misreporting was directly quantified by comparing covertly measured laboratory weighed intakes (LWI) with self-reported EI (weighed dietary record (WDR), 24-h recall, 7-d diet history, FFQ). Personal (age, sex and %body fat) and psychological traits (personality, social desirability, body image, intelligence quotient and eating behaviour) were used as predictors of misreporting. In Study One, those with lower psychoticism (P = 0·009), openness to experience (P = 0·006) and higher agreeableness (P = 0·038) reduced EI on days participants knew EI was being measured to a greater extent than on covert days. Isolated associations existed between personality traits (psychoticism and openness to experience), eating behaviour (emotional eating) and differences between the LWI and self-reported EI, but these were inconsistent between dietary assessment techniques and typically became non-significant after accounting for multiplicity of comparisons. In Study Two, sex was associated with differences between LWI and the WDR (P = 0·009), 24-h recall (P = 0·002) and diet history (P = 0·050) in the laboratory, but not home environment. Personal and psychological correlates of misreporting identified displayed no clear pattern across studies or dietary assessment techniques and had little utility in predicting misreporting.

Dietary Protein and Amino Acids in Vegetarian Diets-A Review

While animal products are rich in protein, the adequacy of dietary protein intake from vegetarian/vegan diets has long been controversial. In this review, we examine the protein and amino acid intakes from vegetarian diets followed by adults in western countries and gather information in terms of adequacy for protein and amino acids requirements, using indirect and direct data to estimate nutritional status. We point out that protein-rich foods, such as traditional legumes, nuts and seeds, are sufficient to achieve full protein adequacy in adults consuming vegetarian/vegan diets, while the question of any amino acid deficiency has been substantially overstated. Our review addresses the adequacy in changes to protein patterns in people newly transitioning to vegetarian diets. We also specifically address this in older adults, where the issues linked to the protein adequacy of vegetarian diets are more complex. This contrasts with the situation in children where there are no specific concerns regarding protein adequacy because of their very high energy requirements compared to those of protein. Given the growing shifts in recommendations from nutrition health professionals for people to transition to more plant-based, whole-food diets, additional scientific evidence-based communications confirming the protein adequacy of vegetarian and vegan diets is warranted.