Validating 24-Hour Dietary Recalls: A Comprehensive Guide for Research and Clinical Trials

Abstract

This article provides a comprehensive framework for the validation of 24-hour dietary recalls (24HR), a cornerstone method for assessing dietary intake in nutritional epidemiology, clinical trials, and public health monitoring. Aimed at researchers, scientists, and drug development professionals, we explore the foundational principles of 24HR methodology, including the standardized Automated Multiple-Pass Method (AMPM). The article details practical application and training protocols for diverse populations, addresses pervasive challenges like measurement error and under-reporting, and systematically reviews modern validation techniques—from comparator methods and biomarker use to the evaluation of emerging web-based and automated tools. The synthesis of these core intents offers a robust guide for ensuring data quality, enhancing study validity, and informing evidence-based health interventions.

The Science of Dietary Recall: Core Principles and Standardized Protocols

The 24-hour dietary recall (24HR) is a foundational, retrospective method for assessing dietary intake, designed to capture detailed information about all foods and beverages consumed by an individual over the previous 24-hour period [1] [2]. As a short-term assessment instrument, its primary purpose is to obtain a precise snapshot of daily intake, which can then be aggregated across a population to describe dietary patterns or, with repeated administration, to estimate an individual's habitual diet [3] [1]. Its utility spans national nutrition surveillance, such as the What We Eat in America component of the National Health and Nutrition Examination Survey (NHANES), epidemiological research investigating diet-disease relationships, and clinical trials evaluating nutritional interventions [1] [4]. The flexibility of the method, which can be administered by a trained interviewer or through automated self-administered systems, makes it a versatile tool for researchers and public health professionals [5].

Framed within the broader context of validation techniques research, this article details the standard protocols of the 24HR, its application across diverse settings, and the critical importance of validation in ensuring data accuracy. The ongoing development and refinement of the method, particularly through technology-assisted tools and statistical modeling, continue to enhance its validity and feasibility for large-scale studies [6] [7] [4].

Methodological Principles and Standard Protocols

Core Definition and Procedure

A 24HR is a structured interview or assessment that guides a respondent to recall and report all food and beverage consumption from the previous day, typically from midnight to midnight [1] [2]. A single recall typically requires 20 to 60 minutes to complete [1]. The method's open-ended nature is a key feature, allowing for detailed probing to capture comprehensive information that a respondent may not initially report [1].

The multiple-pass approach has been established as a gold-standard protocol to minimize memory lapse and enhance completeness [3] [8]. This method uses a staged interview process tailored to human cognition, comprising several distinct "passes" over the previous day's intake. While the exact number of passes can vary, a common and comprehensive framework is the five-pass method [3]:

• Pass 1: Quick List. The respondent provides a rapid, uninterrupted list of all foods and beverages consumed.
• Pass 2: Forgotten Foods. The interviewer uses specific probes to elicit commonly forgotten items (e.g., condiments, snacks, sugary drinks, alcohol).
• Pass 3: Time and Occasion. The respondent assigns a time and eating occasion (e.g., breakfast, lunch, snack) to each reported item.
• Pass 4: Detail Cycle. A detailed review of each food and beverage occurs, capturing description, portion size, cooking method, brand names (if applicable), and additions such as fats or sweeteners.
• Pass 5: Final Probe. A last opportunity is provided for the respondent to recall any additional items [3].

To improve the accuracy of portion size estimation, which is a known source of error, a variety of visual aids are employed. These include food photographs, household measures, three-dimensional food models, or two-dimensional grids [3] [1]. The data collected can be linked to food composition databases to estimate nutrient intake and to food pattern equivalent databases to assess adherence to dietary guidelines [1].

Modes of Administration

The traditional administration of the 24HR involves a trained interviewer conducting the assessment in person or by telephone [3] [2]. However, technological advances have led to the development and widespread adoption of automated, self-administered 24HR systems, which offer significant advantages in standardization, cost, and reduced researcher burden [5] [4].

Prominent examples of such tools include:

• ASA24 (Automated Self-Administered 24-Hour Dietary Assessment Tool): A free, web-based tool developed by the National Cancer Institute (NCI) that adapts the USDA's Automated Multiple-Pass Method (AMPM) for self-administration [5].
• Intake24: A web-based system developed in the UK, which has undergone multiple cycles of user testing [4].
• myfood24: An automated online dietary assessment tool that supports both self- and interviewer-administered recalls and includes features like a recipe builder and pop-up reminders for forgotten foods [9] [2].

Studies have shown that these web-based, self-administered tools can achieve levels of measurement error comparable to interviewer-administered methods, suggesting that the additional costs associated with interviewers may not always translate to improved accuracy [4].

Utility in Research and Data Applications

The 24HR is a highly flexible tool whose utility is dictated by the research question and design. The data generated can be applied to multiple analytical outcomes, as summarized in the table below.

Table 1: Dietary Dimensions Assessable by Multiple 24-Hour Dietary Recalls

Dietary Dimension	Possible to Assess?	Key Considerations
Energy and nutrient intake	Yes	Requires linkage to a nutrient composition database [1].
Intake of specific nutrients or foods	Yes	More accurate for commonly consumed items [3].
Infrequently consumed foods	Maybe	Requires a large number of recall days to capture [3].
Dietary pattern	Yes	Can identify patterns through statistical analysis of food group intake [3] [1].
Habitual diet of an individual	Yes	Requires multiple, non-consecutive recalls to account for day-to-day variation [3] [1].
Within-individual comparison	Yes*	Possible only when repeated measures are collected over time [3].
Meal composition & frequency	Yes	Captures context and timing of eating occasions [3] [1].
Eating environment	Yes	Can capture where food was obtained and consumed [1].

A critical distinction in the application of the 24HR is whether the goal is to assess the diet of a population or of an individual.

• For population-level assessment: A single 24HR per person from a large, representative sample is sufficient to estimate the group's mean usual intake of nutrients or foods [3] [1]. This application is fundamental to national nutrition surveillance.
• For individual-level assessment: A single day of intake is not representative of a person's habitual diet due to day-to-day variation. Therefore, estimating an individual's usual intake requires two or more non-consecutive 24HRs to account for this variability [3] [7] [1]. The number of recalls needed depends on the nutrient of interest and the desired precision; for example, it may take fewer days to estimate usual energy intake and many more days for a rarely consumed nutrient [3] [7].

To address the practical infeasibility of administering a large number of recalls, statistical methods have been developed to estimate usual intake distributions from a limited number of short-term measurements. The NCI method is one prominent example that is widely used because it corrects for measurement error and allows for the inclusion of covariates [7].

Table 2: Key Characteristics of 24-Hour Dietary Recall Administration

Characteristic	Single Recall	Multiple Recalls
Primary Use	Population mean intake	Individual usual intake & distribution
Number of Participants	Up to ~5000 [3]	Up to ~1000 [3]
Participant Burden	Low [3]	Very High [3]
Researcher Burden & Cost	Medium [3]	High [3]
Risk of Reactivity Bias	No [3]	No [3]
Major Type of Measurement Error	Random [1]	Random [1]

Experimental Protocols for Validation

Validating the 24HR against objective measures is a cornerstone of robust nutrition research. The following are detailed protocols for key validation experiments.

Protocol 1: Validation Against Observed Intake in a Controlled Feeding Study

This protocol evaluates the accuracy of a 24HR method by comparing reported intake to known, observed intake in a controlled setting [4].

Workflow:

Detailed Methodology:

• Participant Recruitment: Recruit a sample of healthy adults (e.g., n=150) representative of the target population [4].
• Controlled Feeding: Participants attend a research center on separate days to consume standardized breakfast, lunch, and dinner. The types and exact weights (using digital scales) of all foods and beverages served are recorded as "observed intake" [4].
• 24HR Administration: On the day following each feeding day, participants complete a 24HR using the method under evaluation (e.g., ASA24, Intake24). The sequence of methods should be randomized to control for order effects [4].
• Data Analysis:
  • Calculate the omission rate (percentage of consumed foods not reported) and intrusion rate (percentage of reported foods not consumed) [4].
  • Compare estimated versus observed values for energy, nutrients, and food groups using statistical tests like paired t-tests or linear mixed models [4].
  • Calculate reporting bias as the ratio of reported intake to observed intake [4].

Protocol 2: Validation Against Biomarkers

This protocol assesses the validity of the 24HR by comparing reported intake with objective biological markers, which are not reliant on self-report and are considered recovery biomarkers of true intake [9].

Workflow:

Detailed Methodology:

• Participant Recruitment and Dietary Assessment: Recruit eligible participants and instruct them to complete one or more 24HRs, sometimes as part of a multi-day weighed food record (WFR) using a web-based tool like myfood24 [9].
• Biomarker Collection: On the final day of the dietary recording period, collect biological samples from participants:
  • 24-hour urine collection to measure urinary nitrogen (a biomarker for protein intake) and urinary potassium [9].
  • Fasting blood samples to measure serum folate (a biomarker for folate intake) [9].
  • Indirect calorimetry can be used to measure resting energy expenditure, which helps identify misreporters of energy intake using the Goldberg cut-off [9].
• Data Analysis: Analyze the correlation between reported nutrient intakes and their corresponding biomarker levels using Spearman's rank correlation (ρ). For example, a strong correlation (e.g., ρ = 0.62 for folate) supports the validity of the tool for ranking individuals by intake [9].

The Scientist's Toolkit: Key Research Reagents and Materials

The following table details essential tools and materials required for implementing and validating the 24-hour dietary recall method.

Table 3: Essential Research Reagents and Solutions for 24HR Implementation

Item	Function / Application	Examples / Specifications
Automated 24HR Software	Self-administered data collection; standardizes the recall process and automates coding.	ASA24 [5], Intake24 [4], myfood24 [9] [2]
Food Composition Database (FCDB)	Converts reported food consumption into estimated nutrient intakes.	UK CoFID [6], USDA Food and Nutrient Database [1], Chinese Food Composition Tables [7]
Portion Size Estimation Aids	Helps respondents conceptualize and report the volume or weight of consumed foods.	Food photograph atlases [6] [4], household measures (cups, spoons) [3], food models [3], 2D grids [3]
Dietary Intake Biomarkers	Provides an objective, non-self-report reference for validating reported intakes of specific nutrients.	Urinary Nitrogen (for protein) [9], Urinary Potassium [9], Serum Folate [9], Doubly Labeled Water (for energy) [4]
Statistical Modeling Software	Applies specialized methods to estimate usual intake distributions from short-term recall data.	Software implementing the NCI Method [7], Multiple Source Method (MSM) [7]
Decamethrin-d5	Decamethrin-d5, CAS:1217633-23-2, MF:C22H19Br2NO3, MW:510.237	Chemical Reagent
Carbamazepine 10,11 epoxide-d2	Carbamazepine 10,11-Epoxide-d2 (Major)|RUO	Carbamazepine 10,11-Epoxide-d2 (Major) is For Research Use Only. It is a deuterated internal standard for accurate quantification of the active CBZ metabolite in pharmacokinetic and TDM studies.

The 24-hour dietary recall remains an indispensable tool in nutritional research and public health surveillance due to its detailed, quantitative, and flexible nature. Its core purpose—to capture a precise snapshot of daily food and beverage intake—enables a wide range of applications, from monitoring population diet quality to serving as a reference instrument in validation studies. The evolution of the method, driven by advancements in automated technology and statistical modeling, continues to address its inherent limitations, such as day-to-day variation and participant burden. However, the reliability of data generated by any 24HR method, traditional or novel, is contingent upon rigorous validation against objective standards like observed intake or recovery biomarkers. As dietary assessment moves further into the digital age, the principles of validation and standardization detailed in this article will remain paramount for ensuring that the 24HR continues to yield accurate and meaningful data for researchers, clinicians, and policymakers.

The USDA Automated Multiple-Pass Method (AMPM) represents the cornerstone of modern 24-hour dietary recall administration, serving as the foundational methodology for major national surveys including What We Eat in America, the dietary interview component of the National Health and Nutrition Examination Survey (NHANES) [10]. As a research-based, computerized method, AMPM employs a structured five-step multiple-pass approach specifically engineered to enhance complete and accurate food recall while simultaneously reducing respondent burden [10]. Its development marked a significant advancement in dietary assessment technology, establishing a new gold standard for interviewer-administered recalls conducted either in person or by telephone.

Within the context of dietary recall validation research, AMPM provides the critical benchmark against which newer assessment methods are measured. Understanding its architecture, operational mechanisms, and validation framework is essential for researchers designing studies to evaluate emerging dietary assessment technologies, including web-based platforms, mobile applications, and image-assisted recall systems.

AMPM Methodology: Architectural Framework

Core Operational Framework

The AMPM system operates as an interviewer-administered tool that guides respondents through a meticulously structured recall process for the previous 24-hour period [10]. Its computerized infrastructure incorporates extensive automation capabilities, including unique questions and response options specific to each food, dynamic routing based on previous responses, and automated edit checks during data entry [11]. The system utilizes comprehensive food lookup tables that reflect the contemporary food market and allows interviewers to add, change, or delete foods at any point during the interview process [11].

A critical companion tool to the AMPM system is the Food Model Booklet, which provides visual aids for estimating portion sizes accurately [11]. Furthermore, the method leverages the Food and Nutrient Database for Dietary Studies (FNDDS) for nutrient calculation, ensuring standardized nutritional analysis across all collected dietary data [11].

The Five-Pass Sequential Recall System

The AMPM's efficacy stems from its structured five-pass approach, each stage serving a distinct psychological and cognitive purpose in the retrieval process.

Table 1: The Five-Pass Sequence of the USDA AMPM

Step	Purpose	Cognitive Strategy
Quick List	Collect an unstructured list of all foods and beverages consumed the previous day.	Free recall without interference; respondent-driven narrative.
Forgotten Foods	Probe for frequently forgotten categories (e.g., sweets, beverages, snacks).	Cue-based retrieval; category prompting.
Time & Occasion	Collect time and eating occasion for each food.	Associative memory with temporal context and daily events.
Detail Cycle	For each food, collect detailed description, amount, and additions.	Deep processing and elaboration; portion size estimation.
Final Probe	Final opportunity to recall any additional items.	Comprehensive review and closure of recall process.

This multi-pass structure is deliberately designed to counteract the limitations of human memory through associative techniques (linking foods to daily events), systematic probing for commonly omitted items, and repetition with minimal burden on the respondent [11]. The respondent-driven approach allows the initial recall to be self-defined, establishing a cognitive foundation upon which subsequent passes build with increasingly specific probes for detail.

Validation Techniques and Research Applications

Reference Standards in Validation Studies

Validating 24-hour dietary recalls against objective measures requires sophisticated research designs. The following reference standards represent the current best practices for establishing criterion validity.

Table 2: Reference Standards for Dietary Recall Validation Studies

Validation Method	Description	Key Metrics	Applications in Research
Doubly Labeled Water (DLW)	Gold standard for measuring total energy expenditure through isotope elimination.	rEI:mEE ratio (reported Energy Intake to measured Energy Expenditure) [12].	Identifying under-/over-reporting; establishing energy intake plausibility [12].
Energy Balance Principle	Novel method calculating measured Energy Intake (mEI) as mEE + changes in energy stores (ΔES).	rEI:mEI ratio [12].	Direct comparison against reported intake; accounts for energy balance status [12].
Multiple 24-hour Recalls	Repeated administrations (typically 2-3 non-consecutive days including weekend) as reference.	Correlation coefficients (Spearman), ICCs, Bland-Altman analysis [13].	Assessing reliability and relative validity of intakes for nutrients/food groups [14].
Biomarkers	Objective measures in serum/urine (e.g., nitrogen, electrolytes, folate).	Triad method (correlation between recall, biomarker, and reference method) [15].	Validating intake of specific nutrients independent of reporting error [15].

Quantitative Validation Data

Recent studies provide robust quantitative evidence supporting the validation frameworks used with AMPM and similar methodologies.

Table 3: Validation Metrics from Recent Dietary Assessment Studies

Study & Comparison	Population/Focus	Key Correlation Coefficients	Reliability/Agreement Metrics
FFQ vs. 3-day 24HR [14]	Adults in Fujian, China (n=152)	Food groups: 0.41-0.72Nutrients: 0.40-0.70	>78% same/adjacent tertile classification; acceptable Bland-Altman agreement.
Foodbook24 Expansion [6]	Brazilian, Irish, Polish adults in Ireland	44% of food groups: r=0.70-0.9958% of nutrients: r=0.70-0.99	Food omissions varied by nationality (13-24%).
PERSIAN FFQ Validation [15]	Iranian adults (n=978), 24 recalls over 12 months	Energy: 0.57-0.63Protein: 0.56-0.62	Validity coefficients for biomarkers >0.4; high reproducibility for 19/30 nutrients.
Misreporting Analysis [12]	Adults with overweight/obesity (n=39)	N/A	50% under-reporting; novel energy balance method identified more over-reporting.

Experimental Protocols for Dietary Recall Validation

Protocol: Validation Against Doubly Labeled Water

Purpose: To identify under-reported, over-reported, and plausible self-reported energy intake (rEI) collected via 24-hour dietary recalls [12].

Materials: Doubly labeled water (¹⁸O and ²H), isotope ratio mass spectrometer, calibrated anthropometric scale, stadiometer, quantitative magnetic resonance (QMR) system or other validated body composition method, 24-hour dietary recall administration system (e.g., AMPM).

Procedure:

• Baseline Assessment: Collect baseline body weight, height, and body composition measurements (Day 1).
• DLW Administration: Administer pre-measured DLW dose orally (1.68 g/kg body water of ¹⁸O water and 0.12 g/kg body water of ²H water).
• Urine Collection: Collect urine samples pre-dose, 3-4 hours post-dose, and twice 12 days post-ingestion using a two-point protocol [12].
• Dietary Recall: Administer multiple 24-hour dietary recalls (3-6 non-consecutive days recommended) during the assessment period using the AMPM protocol.
• Follow-up Assessment: Repeat body weight and body composition measurements (Day 13).
• Analysis: Calculate measured energy expenditure (mEE) from DLW data using established equations [12]. Calculate mEI using the energy balance principle: mEI = mEE + ΔES (changes in energy stores). Classify recalls as under-reported, over-reported, or plausible based on rEI:mEE or rEI:mEI ratios and established cut-offs (e.g., ±1SD).

Protocol: Reliability and Validity Testing Against Multiple Recalls

Purpose: To evaluate the reliability and validity of a dietary assessment tool using repeated 24-hour dietary recalls as a reference method [14].

Materials: Dietary assessment tool to be validated (e.g., FFQ, web-based recall), 24-hour dietary recall system (e.g., AMPM), standardized food portion visual aids, nutrient analysis database.

Procedure:

• Participant Recruitment: Recruit participants representative of the target population (minimum n=100-200 recommended) [15].
• Initial Assessment: Administer the dietary tool to be validated (e.g., FFQ) at baseline.
• Multiple 24-hour Recalls: Administer multiple 24-hour dietary recalls (typically 2-3 non-consecutive days, including at least one weekend day) over the subsequent period [13].
• Reliability Assessment: Readminister the dietary tool after a suitable interval (e.g., 1 month) to assess test-retest reliability [14].
• Statistical Analysis:
  • Calculate Spearman correlation coefficients between the test tool and recalls for nutrients and food groups.
  • Compute intraclass correlation coefficients (ICCs) for reliability.
  • Perform Bland-Altman analysis to assess agreement between methods.
  • Use weighted Kappa statistics to evaluate tertile classification agreement.

Visualization of Methodological Frameworks

AMPM Five-Pass Methodology Workflow

Dietary Recall Validation Framework

The Scientist's Toolkit: Essential Research Reagents & Materials

Table 4: Essential Materials for Dietary Recall Validation Research

Item	Function/Application	Specifications/Standards
AMPM Interview System	Administration of standardized 24-hour dietary recalls.	USDA Automated Multiple-Pass Method protocol; computer-assisted interface [10] [11].
Food Model Booklet	Visual aids for portion size estimation during recalls.	Standardized images/representations of common foods and serving utensils [11].
Doubly Labeled Water Kit	Gold standard measurement of total energy expenditure.	¹⁸O water (10.8 APE) and ²H water (99.8 APE); urine collection materials [12].
Food Composition Database	Conversion of food intake to nutrient data.	Food and Nutrient Database for Dietary Studies (FNDDS) or country-specific equivalent [11].
Body Composition Analyzer	Assessment of changes in energy stores (ΔES).	Quantitative Magnetic Resonance (QMR) system or DEXA; standardized protocols [12].
Biological Sample Collection Kits	Biomarker validation (serum, urine, etc.).	24-hour urine collection containers; serum separation tubes; standardized processing protocols [15].
Statistical Analysis Software	Data processing and validation metrics calculation.	Capable of Spearman correlations, ICCs, Bland-Altman analysis, Kappa statistics [14] [12].
SDMA-d6	SDMA-d6, MF:C8H18N4O2, MW:208.29 g/mol	Chemical Reagent
(S)-(-)-Felodipine-d5	(S)-(-)-Felodipine-d5|Labelled Enantiomer Standard	(S)-(-)-Felodipine-d5 is a deuterated, vascular-selective calcium channel blocker enantiomer. For Research Use Only. Not for human consumption.

The USDA AMPM methodology represents a sophisticated psychological and technical approach to overcoming the inherent challenges of dietary recall. Its structured multi-pass system, grounded in cognitive research, provides a robust foundation for collecting dietary data in diverse research contexts. The validation techniques outlined—ranging from the gold standard DLW method to biomarker comparisons and multiple recall approaches—provide researchers with a comprehensive toolkit for evaluating and refining dietary assessment methodologies.

As dietary assessment evolves toward increasingly automated, web-based, and image-assisted platforms, the principles embedded in AMPM continue to inform next-generation tool development. The rigorous validation framework surrounding AMPM establishes the benchmark against which emerging technologies must be measured, ensuring the continued advancement of dietary assessment science in epidemiological research, clinical trials, and public health monitoring.

The 24-hour dietary recall (24HR) method generates a complex chain of data outputs, beginning with raw consumption reports and culminating in sophisticated dietary indices such as Food Pattern Equivalents (FPE). This transformation process enables researchers to quantify dietary intake against national guidelines and examine diet-disease relationships. In validation studies, understanding these outputs is crucial for selecting appropriate comparison metrics and interpreting results accurately [16] [17].

The National Health and Nutrition Examination Survey (NHANES), which collects dietary data through the "What We Eat in America" (WWEIA) component, exemplifies this structured approach. The analytical framework converts individual food reports into multiple data types suitable for different research questions, from simple nutrient totals to complex pattern analyses [16] [17].

Quantitative Data Output Tables

Primary Data Files from 24-Hour Dietary Recalls

Table 1: Core data files generated from 24-hour dietary recall analysis

Data File Type	Record Structure	Key Variables Contained	Primary Use in Analysis
Individual Foods File [16]	Multiple records per person (one per food item)	Food codes (e.g., USDA food code), eating occasion, time, food source, gram amount, nutrient values per food	Disaggregating intake by food item, meal pattern analysis, source of foods
Total Nutrient Intakes File [16]	One record per person	Total daily energy, macronutrients, micronutrients, dietary components	Assessing total daily nutrient adequacy, comparing to Dietary Reference Intakes (DRIs)
Food Pattern Equivalents File [17]	One record per person per day	Amounts consumed from each food group/subgroup (e.g., cup-equivalents of fruits, ounce-equivalents of protein foods)	Evaluating adherence to dietary guidelines (e.g., USDA Food Patterns)

Food Pattern Equivalents and Nutrient Outputs

Table 2: Key output metrics for food groups, subgroups, and select nutrients

Dietary Component Category	Specific Output Metrics	Unit of Measurement
Food Groups & Subgroups [17]	Total fruits; total vegetables; total grains; whole grains; refined grains; total protein foods; dairy and fortified soy alternatives; oils	Cup-equivalents, Ounce-equivalents, Gram-equivalents
Nutrients & Dietary Components [17]	Energy (kcal); dietary fiber; added sugars; calcium; iron; potassium; sodium; saturated fat; vitamin D; folate	Grams, Milligrams, Micrograms
Food Category Sources [17]	Contribution of specific food categories (e.g., cured meats, cheese, flavored milk) to total intake of food groups or nutrients	Absolute amount (e.g., grams) and percentage contribution

Experimental Protocols for Data Generation and Validation

Core Protocol: Generating Food Pattern Equivalents from Recall Data

The conversion of 24HR data into Food Pattern Equivalents (FPE) is a multi-stage process essential for comparing population intakes against dietary recommendations [17].

• Data Ingestion and Food Coding: Individual foods and beverages reported in the 24HR are first matched to standardized food codes within a reference database such as the USDA Food and Nutrient Database for Dietary Studies (FNDDS) [17].
• Food Disaggregation: Mixed dishes and prepared foods are broken down (disaggregated) into their constituent ingredients. For example, a "cheese pizza" report is decomposed into amounts of crust, cheese, and tomato sauce [3].
• Equivalents Calculation: Each ingredient is mapped to its corresponding USDA Food Pattern component. The weight of each ingredient is then converted into the appropriate unit of equivalence:
  • Cup-equivalents: Used for fruits, vegetables, and dairy. For example, 1 cup of milk = 1 cup-equivalent dairy, while 1.5 ounces of natural cheese = 1 cup-equivalent dairy.
  • Ounce-equivalents: Used for grains and protein foods. For example, 1 slice of bread = 1 ounce-equivalent grains, while ¼ cup of cooked beans = 1 ounce-equivalent protein foods [17].
• Aggregation: The equivalents for all components are summed for each individual to produce a daily total for each food group and subgroup. These data are stored in a FPE file, which facilitates analysis of diet quality and adherence to patterns like the Healthy U.S.-Style Eating Pattern [17].

Validation Protocol: Reference Method Comparison

Validation studies assess the accuracy of a novel dietary assessment tool by comparing its outputs against established reference methods. The following protocol outlines a comprehensive validation design.

• Study Population Recruitment: Recruit a target sample of participants. For example, a validation study for a new Experience Sampling-based Dietary Assessment Method (ESDAM) targeted 115 healthy volunteers to ensure adequate statistical power [18].
• Reference Data Collection:
  • Biomarkers: Collect objective biological measures to validate energy and nutrient intake.
    • Doubly Labeled Water (DLW): The gold standard for measuring total energy expenditure, used as a reference for reported energy intake [18].
    • Urinary Nitrogen: Used to estimate protein intake [18].
    • Serum Carotenoids and Erythrocyte Membrane Fatty Acids: Serve as biomarkers for fruit/vegetable and fatty acid intake, respectively [18].
  • Repeat 24-Hour Dietary Recalls: Administer multiple interviewer-led 24HRs using the Automated Multiple-Pass Method (AMPM) as a self-reported reference standard [18] [3].
• Test Method Data Collection: Concurrently, administer the dietary assessment method under validation (e.g., ESDAM, a web-based tool like Foodbook24) over a designated period, typically several days to two weeks [18] [6].
• Data Analysis for Validation:
  • Correlation Analysis: Calculate Spearman rank correlations between nutrient and food group intakes derived from the test method and the reference methods [18] [6].
  • Assessment of Agreement: Use Bland-Altman plots to visualize the limits of agreement between the two methods for key nutrients like energy and protein [18].
  • Method of Triads: Employ this statistical technique to quantify the measurement error of the test method, the 24HRs, and the biomarkers in relation to the unknown "true" dietary intake [18].

Workflow Visualization

Diagram 1: Data transformation and validation workflow for 24-hour dietary recalls, showing the progression from raw data to analytical outputs and validation pathways. DLW: Doubly Labeled Water; N: Nitrogen; FPED: Food Pattern Equivalents Database.

The Scientist's Toolkit: Research Reagent Solutions

Table 3: Essential databases, tools, and biological reagents for dietary recall validation research

Tool or Reagent	Type	Function in Research
USDA FNDDS [17]	Database	Provides energy and nutrient values for ~7,000 foods and beverages reported in WWEIA, NHANES; essential for converting food intake data into nutrient intake.
USDA FPED [17]	Database	Converts foods and beverages from FNDDS into 37 USDA Food Patterns components (e.g., cup-eq of fruits, oz-eq of grains); critical for assessing diet quality.
Doubly Labeled Water (DLW) [18]	Biomarker	The gold-standard objective measure of total energy expenditure; used to validate the accuracy of self-reported energy intake in validation studies.
Urinary Nitrogen [18]	Biomarker	An objective measure used to estimate and validate protein intake at the group level.
INTAKE24 / Oxford WebQ [3]	Software	Automated, web-based 24-hour recall systems that standardize data collection, reduce interviewer burden, and streamline nutrient analysis.
WWEIA Food Categories [17]	Classification System	A scheme of 167 mutually exclusive food categories used to analyze food sources and consumption patterns from NHANES dietary data.
Serum Carotenoids [18]	Biomarker	Objective biochemical measures that serve as a validation reference for reported fruit and vegetable intake.
2-Hydroxy(~13~C_6_)benzoic acid	2-Hydroxy(~13~C_6_)benzoic acid, CAS:1189678-81-6, MF:C7H6O3, MW:144.077 g/mol	Chemical Reagent
1H-indole-2-carboxylic acid	1H-Indole-2-Carboxylic Acid

The 24-hour dietary recall (24HR) is a cornerstone method for assessing dietary intake in nutritional epidemiology, clinical research, and public health monitoring [19] [1]. Its application spans from national surveys to intervention studies, owing to its quantitative nature, relatively low participant burden, and ability to capture detailed information about the total diet without altering immediate eating patterns [1] [3]. However, like all self-reported dietary assessment instruments, the 24HR is subject to inherent measurement errors that can compromise data validity if not properly acknowledged and mitigated [20].

Two of the most critical sources of these errors are day-to-day variation (random error) and reactivity (a potential systematic error). Day-to-day variation refers to the natural fluctuations in an individual's food intake from one day to the next, which obscures the measurement of their "usual" or habitual intake [20] [21]. Reactivity, conversely, occurs when the process of measurement itself influences the behavior being measured; though for unannounced 24HRs, this effect is considered minimal compared to methods like food records [1]. For researchers, particularly those in clinical and pharmaceutical development contexts where precise metrics are paramount, understanding these limitations is essential for designing robust studies, interpreting results accurately, and developing effective dietary interventions. This document outlines the nature of these limitations and provides standardized protocols for their management within a research validation framework.

Understanding Day-to-Day Variation

Definition and Impact on Data Quality

Day-to-day variation, or within-person variation, is a form of random error that arises from the fact that an individual's diet is not identical every day [20]. This variation is a function of two components:

• True Variation: Actual differences in food choices and portion sizes from day to day.
• Random Measurement Error: Inaccuracies introduced during the recall process, such as forgetting a food item or misestimating a portion size [20].

The primary impact of this variation is to lower the precision of the dietary intake data [20]. A single 24HR per participant provides a "snapshot" of intake that may not be representative of their long-term habitual consumption. This is especially problematic for nutrients and foods that are not consumed daily, such as vitamin A or liver [19]. When analyzing data, this error increases within-person variance, which flattens and widens the observed intake distribution for a group. This, in turn, reduces statistical power in association studies and leads to inaccurate estimates of the proportion of the population above or below a dietary threshold [20] [21].

Quantitative Data on Variation

The following table summarizes key aspects of day-to-day variation and its implications for study design.

Table 1: Impact of Day-to-Day Variation and Mitigation Strategies in 24-Hour Dietary Recalls

Aspect	Description	Implication for Research
Nature of Error	Random error that reduces precision [20].	Increases within-person variance, weakens observed diet-health relationships.
Effect on Distribution	Flattens and widens the intake distribution of a group [21].	Leads to biased estimates of the proportion of a population below or above a dietary cut-off point.
Variability by Nutrient	High for episodically consumed nutrients (e.g., Vitamin A, Vitamin C); lower for macronutrients [19].	The number of recall days needed to estimate usual intake varies by nutrient of interest.
Primary Mitigation	Collecting multiple non-consecutive 24HRs per person [20] [21] [1].	Enables use of statistical methods (e.g., NCI method) to remove within-person variation and estimate usual intake.
Subsampling Approach	Repeats collected on a random subset (≥30-40 individuals) of the population [20].	A cost-effective strategy to estimate the within-to-between person variance ratio for a population.

Experimental Protocol: Accounting for Day-to-Day Variation

Objective: To estimate the usual habitual intake distribution of a population for a specific nutrient (e.g., protein).

Materials:

• Trained interviewers or a validated automated self-administered system (e.g., ASA24, myfood24).
• Standardized protocol (e.g., Automated Multiple-Pass Method).
• Visual aids for portion size estimation (e.g., food photographs, models).
• Nutrient composition database.
• Statistical software capable of modeling usual intake (e.g., R, SAS with NCI macros).

Procedure:

• Study Design and Sampling:
  • Determine the number of recall days based on the nutrient of interest and study objectives. For many nutrients, 2 non-consecutive days is the minimum to apply statistical adjustment, though more may be needed for highly variable nutrients [21] [1] [3].
  • Administer recalls on random, non-consecutive days to capture different days of the week (including weekends) and, if feasible, different seasons to account for seasonal variation [20] [3].
  • For large cohort studies, consider a subsampling design where a representative subset of the population (e.g., 30-40 individuals per stratum) provides repeated recalls to estimate the population's variance ratio [20].

• Data Collection:
  • Collect 24HR data using a standardized multiple-pass method to minimize random measurement error [20]. The following workflow diagram illustrates a robust 24HR administration process suitable for research.

• Data Analysis:
  • Process the data by linking reported foods to a nutrient composition database to derive daily intake values for each participant and day [21].
  • Use statistical methods to estimate usual intake distributions. The National Cancer Institute (NCI) method is a widely accepted approach that models intake and episodicity to remove the effect of within-person variation [21] [1].
  • Account for nuisance effects like day of the week, sequence of recall, and interview mode in the statistical model [20] [21].

Understanding Reactivity

Definition and Context in 24HR

Reactivity is a systematic error that occurs when participants alter their normal dietary behavior because they are aware of being studied [1]. This is a well-known issue with prospective methods like food records, where participants may choose to eat simpler meals or consume "healthier" foods to make recording easier or due to social desirability bias [19].

A key advantage of the 24-hour dietary recall is that it is a retrospective method. When recalls are unannounced—meaning the participant does not know in advance which day they will be asked to recall—the potential for reactivity is significantly reduced because the diet has already been consumed [1] [3]. Therefore, for the standard 24HR, reactivity is generally not considered a major source of bias. However, in the context of validation techniques, it is crucial to distinguish the 24HR from other methods and to understand when reactivity might become a concern, such as in intensive longitudinal studies where participants complete frequent recalls over time.

Protocol for Minimizing Reactivity and Other Biases

Objective: To collect self-reported dietary data with minimal influence on actual eating behavior.

Procedure:

• Use Unannounced Recalls: Do not inform participants of the specific day they will be recalled in advance. This is the most effective strategy to prevent reactivity [3].
• Blind Participants to Study Hypotheses: If possible, keep participants unaware of the specific dietary hypotheses being tested to reduce the motivation to report in a socially desirable manner.
• Standardize Interviewer Behavior: Train interviewers to use a neutral, non-judgmental tone to avoid subtly cueing participants towards "desirable" responses [1].
• In Intervention Studies, Account for Differential Bias: In randomized controlled trials, the intervention group may become more aware of their diet and report it differently than the control group. Using an objective biomarker (e.g., doubly labeled water for energy) as a primary or calibration measure can help correct for this potential differential bias [21].

The Scientist's Toolkit: Research Reagent Solutions

Table 2: Essential Materials and Tools for 24HR Validation and Administration

Item	Function in Research	Example(s)
Automated Multiple-Pass Method (AMPM)	A structured interview protocol designed to enhance memory retrieval and reduce omissions, thereby minimizing random error [20] [3].	USDA AMPM; GloboDiet.
Self-Administered 24HR Systems	Reduces interviewer burden and cost, standardizes the questioning process, and allows for large-scale data collection [19] [22].	ASA24 (NIH), myfood24 [22] [23], Oxford WebQ.
Portion Size Estimation Aids	Visual tools to improve the accuracy of reported food amounts, reducing measurement error [1] [3].	Food photographs, 2D grids, household measure guides, food models.
Recovery Biomarkers	Objective, non-self-reported measures used to validate and calibrate self-reported intake data, correcting for systematic error like under-reporting [19] [21] [24].	Doubly Labeled Water (energy), Urinary Nitrogen (protein), Urinary Potassium (K), Urinary Sodium (Na).
Food Composition Database	Converts reported food consumption into estimated nutrient intakes. The choice of database directly impacts results [20] [21].	USDA FoodData Central, German BLS, UK Composition of Foods.
Usual Intake Modeling Software	Statistical packages that adjust intake distributions for within-person variation to estimate habitual intake [21].	NCI Method Macros (SAS/R), Multiple Source Method (MSM).
Abacavir-d4	Abacavir-d4, MF:C14H18N6O, MW:290.36 g/mol	Chemical Reagent
Nor Acetildenafil-d8	Nor-acetildenafil-d8|Isotopic Labeled Analog	Nor-acetildenafil-d8 is a deuterated internal standard for precise quantification of sildenafil analogs in research. For Research Use Only. Not for human or veterinary use.

The inherent limitations of day-to-day variation and the potential for reactivity are critical considerations in the validation and application of 24-hour dietary recalls. Day-to-day variation is a pervasive random error that must be addressed through study design (multiple recalls) and sophisticated statistical modeling to derive valid estimates of usual intake for a group. While reactivity is less of a concern for unannounced 24HRs compared to other dietary methods, vigilance is required to prevent other systematic biases like social desirability from influencing reports.

For researchers in drug development and high-stakes clinical research, adhering to the protocols outlined herein—employing multiple recalls, using recovery biomarkers for validation, standardizing data collection with automated tools, and applying appropriate statistical corrections—is essential. These practices transform the 24HR from a simple snapshot into a powerful, validated tool capable of generating reliable data on dietary exposure, which is fundamental to understanding the role of nutrition in health and disease.

From Theory to Practice: Implementing Recalls in Diverse Research Settings

Within the framework of 24-hour dietary recall validation research, selecting the appropriate data collection methodology is paramount to data integrity. The choice between interviewer-mediated and self-administered recalls represents a critical trade-off between data quality, participant burden, and logistical feasibility [25] [26]. Self-administered tools, primarily web-based systems like ASA24 (Automated Self-Administered 24-hour dietary assessment tool), R24W, and myfood24, offer a standardized, cost-effective approach for large-scale studies [5] [27] [22]. Conversely, interviewer-administered recalls, often employing the Automated Multiple-Pass Method (AMPM), provide a guided interaction that can enhance data completeness and accuracy, particularly for complex diets or specific populations [25] [28]. This application note synthesizes current evidence to delineate the advantages, limitations, and relative validity of each method, providing structured protocols and data-driven guidance for researchers and professionals in nutritional epidemiology and clinical trial design.

Comparative Analysis: Key Metrics and Performance Data

The decision between recall methods hinges on specific study objectives, population characteristics, and resource constraints. The quantitative data below summarizes key performance differences identified in validation studies.

Table 1: Key Differences Between Self-Administered and Interviewer-Administered 24-Hour Dietary Recalls

Metric	Self-Administered Recalls	Interviewer-Administered Recalls
Reported Energy & Nutrient Intake	Generally lower reported intakes for energy, fat, saturated fat, and sugar [25]. In some studies, higher reported intakes for certain nutrients (e.g., saturated fat) [27].	Higher reported energy and nutrient intakes; more food items reported (e.g., 25% more items) [25].
Participant Burden & Preference	70% preference reported in some adult studies for its convenience [29]. Adolescents often prefer interviewer-administered methods [26].	Preferred by adolescents [26] and individuals with technological or literacy barriers [25].
Data Completeness	Fewer food items reported per recall [25]. Prone to omission errors if users struggle with interface [25].	More complete food lists and detailed descriptions facilitated by interviewer probing [28].
Population Reach & Bias	Risk of exclusion and sampling bias: lower completion among older adults, non-white individuals, and those with lower education levels [25].	Higher inclusion rates across diverse demographics, including those with low literacy or limited tech access [25] [28].
Resource Allocation	Lower operational cost and staff time after development; automated coding [5] [29].	High cost and staff time for trained interviewers and data coding [26] [29].
Validation Against Biomarkers	Good validity for protein, potassium, sodium (e.g., R24W, myfood24) [30] [22]. Underreporting of energy vs. biomarkers is common [25].	Considered a robust standard; good agreement with urinary recovery biomarkers [22].

Table 2: Relative Validity of Selected Self-Administered 24-Hour Recall Tools

Tool Name	Population Validated In	Key Validation Findings	Citation
ASA24	Adults (Field Trial, n=~1,000)	87% of nutrients/food groups equivalent to interviewer-administered AMPM at 20% bound.	[29]
R24W	French-Canadian Adolescents (n=111)	Significant correlations for most nutrients; higher mean energy intake (8.8%) vs. interviewer recall.	[27]
myfood24-Germany	Adults (n=97)	Good agreement with weighed food records and urinary biomarkers for protein and potassium.	[22]
ASA24-Kids-2014	Adolescents (Pilot, n=20)	No significant decay in reporting quality over 6 weeks vs. interviewer recall, but technical difficulties and preference for interviewer.	[26]

Experimental Protocols for Method Comparison and Validation

To ensure the reliability of dietary data, rigorous validation of the chosen assessment method is essential. The following protocols outline standardized approaches for comparing self-administered tools against established benchmarks and for evaluating their feasibility in specific populations.

Protocol 1: Validation Against Interviewer-Administered Recall

Objective: To assess the relative validity of a web-based self-administered 24-hour dietary recall tool against a traditional interviewer-administered recall.

• Step 1: Study Design and Participant Recruitment

  • Employ a randomized crossover design where each participant completes both methods, with the order randomized to control for sequence effects [27] [29].
  • Recruit a sample size of approximately 100-150 participants to ensure adequate statistical power, ensuring diversity in age, sex, and socioeconomic status to evaluate bias [25] [29].

• Step 2: Data Collection

  • Self-Administered Recall: On the assigned day, participants receive an email link to complete the web-based recall (e.g., ASA24, R24W). The tool should guide them through a multi-pass process: a quick list, forgotten foods probe, detail cycle for portion sizes (using images), and a final review [5] [27].
  • Interviewer-Administered Recall: Conduct unannounced telephone interviews using the AMPM protocol. Trained interviewers use a structured script to guide participants through the five passes: quick list, forgotten foods, time and occasion, detail cycle, and final probe [26] [27]. Mail participants portion size aids (e.g., measuring cups, food model booklets) prior to the study [29].

• Step 3: Data Processing and Analysis

  • Export nutrient and food group data from both methods using standardized databases (e.g., FNDDS, CNF) [27] [29].
  • Perform statistical analysis using paired t-tests to compare mean intakes of energy and nutrients. Calculate Pearson or deattenuated correlation coefficients between methods. Use cross-classification analysis to determine the proportion of participants classified into the same or opposite quartiles [30] [27] [29].
  • Predefine equivalence bounds (e.g., ±20%) to determine clinical or practical significance of differences [29].

Protocol 2: Usability and Feasibility Testing in Challenging Populations

Objective: To identify barriers to self-completion and determine the need for interviewer support in populations such as the elderly, low-literacy, or low-income groups.

• Step 1: Participant Screening and Recruitment

  • Recruit participants from the target population (e.g., cancer survivors, older adults, low-income families) [25]. Collect baseline data on age, ethnicity, education level, internet use, and email proficiency [25].

• Step 2: Structured Data Collection with Integrated Support

  • Invite all participants to self-complete a 24-hour recall online. Offer standardized technical assistance and reminders [25].
  • Define a clear protocol for identifying "inability to complete": multiple failed attempts, self-reported overwhelming difficulty, or lack of an email address [25].
  • Offer an interviewer-administered recall as an alternative to those unable to self-complete. Document the specific reasons for failure (e.g., "unable to find foods," "lacked technological skills") [25].

• Step 3: Quantitative and Qualitative Analysis

  • Use multivariate regression analysis to identify demographic and socioeconomic factors (e.g., age, education, ethnicity) significantly associated with the need for an interviewer [25].
  • Compare completion rates, number of food items reported, and nutrient intakes between self-completers and those requiring an interviewer [25].
  • Analyze qualitative feedback from participants on usability challenges and preferences to inform tool optimization and study protocol design [25] [26].

The Scientist's Toolkit: Essential Research Reagent Solutions

The following table details key tools and methodologies central to advancing research in 24-hour dietary recall validation.

Table 3: Key Reagents and Tools for Dietary Recall Validation Research

Tool or Reagent	Function in Research	Application Notes
ASA24 (Automated Self-Administered 24-Hour Recall)	A free, web-based tool for collecting multiple, automatically coded 24-hour diet recalls and food records.	Most appropriate for respondents with a ~5th-grade reading level and comfort with computers; used in over 1,000 scientific publications [5].
AMPM (Automated Multiple-Pass Method)	The interviewer-administered gold standard protocol used in What We Eat in America, NHANES.	Structured into 5 passes to enhance memory and reduce omission; requires trained interviewers and is resource-intensive [26] [29].
Urinary Recovery Biomarkers (Nitrogen, Potassium)	Objective, non-self-report measures used to validate reported intakes of protein and potassium.	Protein intake is estimated from urinary nitrogen assuming 80% recovery; serves as an objective benchmark to assess validity of self-report tools [30] [22].
myfood24	A customizable, web-based 24-hour dietary recall tool adaptable for different countries and databases.	The German validation demonstrated good agreement with weighed food records and biomarkers, highlighting the importance of local database adaptation [22].
Portion Size Estimation Aids	Visual aids (photographs, food models, clay, cups) used to improve the accuracy of reported food amounts.	Mailed to participants for telephone interviews [29]; integrated as digital images in web-based tools [27].
Carteolol-d9hydrochloride	Carteolol-d9hydrochloride, MF:C16H25ClN2O3, MW:337.89 g/mol	Chemical Reagent
Niflumic Acid-d5	Niflumic Acid-d5, MF:C13H9F3N2O2, MW:285.24 g/mol	Chemical Reagent

Decision Framework and Visual Workflow

The choice between recall methods is not one-size-fits-all but should be guided by a structured assessment of study needs and population characteristics. The following workflow provides a logical pathway for this decision-making process.

Figure 1: Dietary Recall Method Selection Workflow. This diagram outlines a logical decision pathway for researchers choosing between interviewer-mediated and self-administered 24-hour dietary recalls, based on study priorities and participant characteristics.

The integration of self-administered web-based tools represents a significant advancement for large-scale dietary monitoring, offering scalability and cost-efficiency with reasonable validity for many research purposes [29] [22]. However, the persistence of digital exclusion necessitates a nuanced approach [25]. Reliance solely on self-administered methods risks introducing significant bias by systematically excluding older, less educated, and non-white populations [25]. Therefore, a one-size-fits-all approach is not recommended. For studies aiming for generalizable results across a diverse population, a mixed-mode design that offers an interviewer-administered alternative is not just a concession but a critical strategy to ensure equitable participation and data integrity [25]. The choice of method should be a deliberate decision, aligned with study goals and informed by a clear understanding of the target population's capabilities and constraints.

The accurate assessment of dietary intake is a cornerstone of nutritional epidemiology, essential for understanding diet-disease relationships and evaluating public health interventions. Traditional interviewer-administered 24-hour dietary recalls, while considered a gold standard, are resource-intensive, requiring trained personnel and imposing significant logistical burdens [31]. The emergence of automated, self-administered 24-hour dietary recall tools represents a paradigm shift, offering the potential to collect high-quality dietary data with greater efficiency, scalability, and reduced cost [32] [33]. This document provides Application Notes and Protocols for three prominent automated tools—ASA24, R24W, and FOODCONS—framed within the context of validation techniques for a broader thesis. It is structured to equip researchers, scientists, and drug development professionals with the practical knowledge to select, implement, and validate these tools in their studies.

Automated dietary assessment tools adapt the traditional multiple-pass methodology for digital platforms, guiding respondents through the recall process without direct interviewer assistance. The following table summarizes the core characteristics of ASA24, R24W, and FOODCONS.

Table 1: Technical Specifications and Availability of Automated 24-Hour Dietary Recall Tools

Feature	ASA24	R24W	FOODCONS
Primary Developer	National Cancer Institute (NCI), USA [5]	Université Laval, Canada [31]	Council for Agricultural Research and Economics, Italy [31]
Cost	Free for researchers [5] [34]	Information not specified in search results	Information not specified in search results
Access	Web-based, mobile-enabled [5] [34]	Web-based [31]	Web-based, on stand-alone computer via virtual machine [31]
Languages	English, Spanish (US Version); English, French (Canadian Version) [34]	French (Canadian version) [33]	Italian [31]
Data Collection Modes	24-hour recalls & food records [5]	24-hour recalls [33]	24-hour recalls & food diaries [31]
Underlying Methodology	USDA's Automated Multiple-Pass Method (AMPM) [5]	Multiple-Pass Method [33]	Multiple-Pass Method per EU Menu guidelines [31]
Food Composition Database	Food and Nutrient Database for Dietary Studies (FNDDS) [34]	Information not specified in search results	Italian food composition database [31]
Key Validation Study	Subar et al. (2020), IDATA Study [35]	Laramée et al. (2022) [35]	FOODCONS Italian Pilot Case Study (2025) [31]

Application Notes: Tool Selection and Implementation

ASA24 (Automated Self-Administered 24-Hour Dietary Assessment Tool)

ASA24 is a freely available, web-based tool developed by the National Cancer Institute (NCI). It enables the automated collection of both 24-hour recalls and food records [5]. As of 2025, it has been used in over 1,000 peer-reviewed publications and collects an average of 673 studies per month, underscoring its widespread adoption [5]. Its key advantage is the robust linkage to well-established American nutrient databases (FNDDS) and food group equivalents (FPED) [34]. The tool is mobile-enabled and has been validated in diverse populations, including adults, children (with parent proxy-reporting), and low-income groups [5] [35] [36].

R24W

R24W is a Canadian web-based 24-hour recall tool. A key study by Laramée et al. (2022) compared its usability against ASA24-Canada-2018 among French-speaking adults in Québec, providing critical validation data for its application in francophone populations [35]. While the searched results provide less detailed technical specification compared to ASA24, its inclusion in international reviews of dietary assessment tools confirms its relevance and utility in the field [33].

FOODCONS

FOODCONS is a software suite developed for Italian nutritional studies. Its 1.0 version supports both interviewer-led and self-administered 24-hour recalls, and it has been used in multiple Italian national consumption surveys [31]. A recent 2025 pilot study demonstrated strong agreement between self-administered and interviewer-led 24-hour recalls using the FOODCONS platform for data entry in both modes, validating its use for autonomous data collection [31]. A notable feature is its design for use on a stand-alone computer via a virtual machine, which can facilitate deployment in settings with limited internet connectivity [31].

Experimental Protocols for Tool Validation

Validation is critical to ensure that automated tools measure dietary intake accurately and with minimal systematic error. The following protocols detail methodologies from key validation studies.

Protocol 1: Comparison with Interviewer-Administered Recalls

This protocol validates a self-administered tool against the traditional benchmark.

• Citation: FOODCONS Italian Pilot Case Study (2025) [31]
• Objective: To compare food group, energy, and nutrient intakes derived from self-administered and interviewer-led 24-hour recalls using the same software (FOODCONS 1.0).
• Population: 39 Italian adults aged 18-64 years, excluding individuals with professional nutritional backgrounds.
• Design: A randomized crossover design. Participants were randomized into two groups (A and B). On the first study day, Group A completed a self-administered recall followed by an interviewer-led recall three hours later. Group B completed the tests in the reverse order. The process was repeated on a second, non-consecutive day, with the order swapped for each group.
• Data Collection: All recalls were completed using the FOODCONS 1.0 software, which implements a multiple-pass method. The two study days included at least one weekend day.
• Statistical Analysis: Paired t-tests or Wilcoxon signed-rank tests to compare mean intakes of energy and nutrients. Bland-Altman analysis to assess agreement between the two methods. Correlation coefficients (Pearson or Spearman) to evaluate concordance for food groups.

Protocol 2: Validation in a Specific Population (Parent Proxy-Reporting)

This protocol assesses the feasibility and validity of using a tool via a proxy reporter.

• Citation: Sharpe et al. (2021) [35] [36]
• Objective: To evaluate the feasibility of ASA24 for parent proxy-reporting of children's dietary intake and to compare intake estimates with national surveillance data.
• Population: Parents of children aged 4-15 years from the TARGet Kids! cohort in Canada.
• Design: A feasibility study. Parents were invited via email to complete an ASA24-Canada-2016 recall for their child. A subsample (~25%) was asked to complete a second recall about two weeks later to assess usual intake.
• Data Collection: Parents completed the recalls independently online. To mitigate social desirability bias, recalls were semi-unannounced. A $5 incentive was introduced partway through the study to improve response rates.
• Key Feasibility Metrics: Response rate, completion rate, median number of foods reported, median time to complete the recall.
• Validation Analysis: Descriptive statistics for nutrient intakes (energy, macronutrients, fiber, sodium, sugars) were compared with similar data from the 2015 Canadian Community Health Survey (CCHS), which uses interviewer-administered recalls.

Visualization of Workflows and Relationships

The following diagrams, generated using Graphviz DOT language, illustrate the experimental protocol and tool selection logic.

Dietary Recall Validation Protocol

Tool Selection Logic

The Scientist's Toolkit: Essential Research Reagents and Materials

The following table details key resources required for implementing and validating automated 24-hour dietary recall tools in a research setting.

Table 2: Essential Research Reagents and Solutions for Dietary Recall Validation

Item	Function/Description	Example/Note
Automated Recall Tool	The core software platform for self-administered data collection.	ASA24, R24W, or FOODCONS. Selection depends on target population and language [5] [33] [31].
Food Composition Database	Converts reported food consumption into nutrient intake data.	ASA24 links to FNDDS (USA) [34]; FOODCONS uses a national Italian database [31].
Pilot Testing Cohort	A small, representative sample for feasibility testing prior to main study.	Used to assess completion time, usability, and identify technical issues [5] [36].
Validation Standard	A benchmark against which the automated tool's accuracy is measured.	Can be interviewer-led recalls [31], feeding studies [35], or biomarkers like doubly labeled water [35].
Participant Incentives	Financial or other compensations to improve response and completion rates.	A $5 e-gift card was effective in increasing ASA24 completion rates in a feasibility study [36].
Usability Questionnaire	A survey to collect qualitative feedback on user experience and difficulties.	Critical for optimizing protocols and interpreting data quality, especially in novel populations [35].
Meclofenamic acid-d4	Meclofenamic acid-d4, CAS:1185072-18-7, MF:C14H11Cl2NO2, MW:300.2 g/mol	Chemical Reagent
Diethyltoluamide-d10	Diethyltoluamide-d10, CAS:1215576-01-4, MF:C12H17NO, MW:201.33 g/mol	Chemical Reagent

Within nutritional epidemiology and clinical research, the 24-hour dietary recall (24HR) serves as a cornerstone method for assessing individual and population-level dietary intake. Its utility spans from monitoring nutritional status in national surveys to serving as a key endpoint in intervention trials for chronic disease prevention [1] [3]. However, the accuracy of 24HR data is highly contingent upon the standardized administration of the method by interviewers and peer educators. A lack of fidelity in protocol application introduces measurement error, which can obscure true diet-disease relationships and compromise the validity of research findings [37]. This article details application notes and experimental protocols for training staff to achieve high-fidelity 24HR administration, directly supporting the rigorous validation techniques required for robust research.

Performance Quantification of 24-Hour Dietary Recalls

Understanding the potential sources of error and the expected performance of 24HR methods is fundamental to designing effective training and quality control protocols. The following tables summarize key quantitative findings on validity and reliability from recent studies.

Table 1: Criterion Validity of Energy Intake from Various 24HR Tools Against Doubly Labeled Water

24HR Tool / Study	Population	Mean Under-Reporting of Energy Intake	Correlation with TEE	Citation
Intake24 (First Recall)	98 UK adults (40-65 yrs)	25% (95% LoA: -73% to +68%)	0.31	[38]
Intake24 (Mean of 2 Recalls)	98 UK adults (40-65 yrs)	22% (95% LoA: -61% to +41%)	0.47	[38]
Automated Multiple-Pass Method (AMPM)	524 American adults (30-69 yrs)	11% (underestimation)	Information Missing	[39]
2 x 24-h Recall (EFSA Method)	120 Danish adults (18-60 yrs)	No significant under-reporting	Information Missing	[39]
7-day Food Diary	120 Danish adults (18-60 yrs)	Significant under-reporting (34% under-reporters)	Information Missing	[39]

Table 2: Reliability of Nutrient Intake Estimates from Repeated 24HR Administrations

Nutrient	Single Recall (ICC)	Two Recalls (ICC)	Citation
Energy	0.35	0.52	[38]
Fat	Information Missing	0.37	[38]
Iron	0.31	Information Missing	[38]
Non-Milk Extrinsic Sugars	0.43	0.63	[38]

Table 3: Food Reporting Accuracy in a Weighed Intake Study with Older Adults

Accuracy Metric	Overall Result	Difference by Sex (Women vs. Men)	Citation
Food Item Match Rate	71.4%	Significantly higher in women	[40]
Exact Match Rate	38.0%	Information Missing	[40]
Portion Size Estimation	Systematic overestimation	No significant difference by sex	[40]

Experimental Protocols for 24HR Validation

For researchers aiming to validate a 24HR tool or assess the effectiveness of a training program, the following protocols provide a methodological framework.

Protocol 1: Criterion Validation Against Doubly Labeled Water

This protocol uses the doubly labeled water (DLW) technique, the gold standard for measuring total energy expenditure (TEE) in free-living individuals, to validate reported energy intake (EI) [37] [38].

1. Objective: To assess the criterion validity of self-reported Energy Intake from a 24HR tool by comparing it against objectively measured TEE.

2. Materials and Reagents:

• Doubly Labeled Water: A mixture of stable isotopes (²Hâ‚‚O and H₂¹⁸O).
• Urine Collection Kit: Pre-labeled sterile sample bottles, cold packs, and a recording sheet for date and time.
• 24HR Tool: The automated or interviewer-administered system to be validated.
• Isotope Ratio Mass Spectrometer: For analyzing isotope enrichment in urine samples.

3. Experimental Workflow:

• Recruitment: Recruit a sample of ~100 participants, stratified by sex, age, and BMI to ensure population representation [38].
• Baseline Clinic Visit (Day 0):
  • Obtain informed consent.
  • Collect a baseline urine sample.
  • Administer a body weight-specific dose of DLW.
  • Provide training on the 24HR tool and the urine collection protocol.
• Free-Living Period (Days 1-9):
  • Participants collect one daily urine sample (excluding first void) for 9-10 days, recording the time and date.
  • Participants complete the 24HR tool at least twice on non-consecutive days during this period.
• Second Clinic Visit (Day 10):
  • Collect final urine sample and all stored samples.
  • Review and collect 24HR data.
• Data Analysis:
  • Calculate TEE from the isotope elimination rates using established equations [38].
  • Calculate the ratio of reported EI to TEE (EI/TEE). A ratio of 1 indicates perfect reporting.
  • Use Bland-Altman analysis to determine the mean bias (accuracy) and 95% limits of agreement (precision) between EI and TEE [38].
  • Calculate Pearson's correlation coefficient to assess the ability of the tool to rank individuals by their energy intake.

Protocol 2: Validation Against Weighed Food Records

This protocol uses weighed food records as a reference method, suitable for validating intake of specific foods, nutrients, and portion sizes [40].

1. Objective: To determine the accuracy of a 24HR in reporting food items and portion sizes compared to weighed food records in a controlled setting.

2. Materials:

• Digital Food Scales (precision ±1 g).
• Standardized Protocol Recording Forms for weighed records.
• Trained Dietitians to conduct both the weighing and the 24HR interviews.

3. Experimental Workflow:

• Participant Preparation: Recruit free-living participants. In a feeding study design, provide all meals and weigh each food item served.
• Weighed Intake Recording:
  • Participants consume meals in a designated area.
  • Trained staff weigh and record any leftovers to determine the exact net weight consumed for each food item.
• 24HR Administration:
  • Within 24 hours of the meal consumption, a trained interviewer, blinded to the weighed record, administers the 24HR.
  • The interview should follow a structured method like the Automated Multiple-Pass Method (AMPM) [1] [39].
• Data Analysis:
  • Match Rate: Calculate the percentage of actually consumed food items that were correctly reported in the 24HR [40].
  • Portion Size Accuracy: For matched items, calculate the absolute and relative difference between the weight consumed and the portion size reported in the 24HR.
  • Nutrient Intake Comparison: Convert both the weighed data and 24HR data to nutrient intakes and compare mean differences and correlations for key nutrients of interest.

Standardized 24HR Implementation Workflow

The diagram below outlines a standardized workflow for staff and peer educators to ensure fidelity during 24HR administration.

The Scientist's Toolkit: Essential Research Reagents & Materials

Table 4: Key Resources for 24HR Implementation and Validation

Item Name	Function/Application	Specifications & Examples
Automated 24HR Software	Standardizes the interview process, automates data entry and nutrient calculation.	Examples: USDA's AMPM, Intake24 [38], ASA24 [1], EPIC-Soft [41]. Feature: Integrated food composition databases and portion size image libraries.
Portion Size Estimation Aids	Helps participants conceptualize and report the volume of food consumed.	Types: Food photographs [38] [3], household measures (cups, spoons), 2-dimensional grids, food models [1].
Food Composition Database (FCDB)	Converts reported foods and beverages into estimated nutrient intakes.	Requirement: Must be culturally appropriate and regularly updated. Examples: USDA FCDB, local/regional databases (e.g., Chilean database for SER-24H) [42].
Food Patterns Equivalents Database (FPED)	Translates reported foods into guidance-based food groups (e.g., cups of fruit, tsp of added sugars).	Application: Essential for assessing adherence to dietary guidelines [1].
Doubly Labeled Water (DLW)	The gold-standard biomarker for validating total energy intake assessment in free-living individuals.	Application: Used in criterion validation studies [37] [38]. Consideration: High cost and technical complexity.
Recovery Biomarkers (e.g., Urinary Nitrogen)	Objective measures to validate intake of specific nutrients (e.g., protein via urinary nitrogen) [37].	Application: Provides an unbiased measure for validating specific nutrient intakes.
Pindolol-d7	Pindolol-d7, CAS:1185031-19-9, MF:C14H20N2O2, MW:255.36 g/mol	Chemical Reagent
Myrcene-d6	Myrcene-d6, CAS:75351-99-4, MF:C10H16, MW:142.27 g/mol	Chemical Reagent

The fidelity of 24-hour dietary recall administration is not a peripheral concern but a central component of data integrity in nutritional research. By implementing rigorous, standardized training protocols for staff and peer educators—underpinned by structured workflows, quantitative performance monitoring, and criterion-validation techniques—researchers can significantly mitigate measurement error. The protocols and tools detailed herein provide a roadmap for achieving this standardization, thereby enhancing the reliability and validity of dietary data used in etiological research, clinical trials, and national public health monitoring.

Within the framework of 24-hour dietary recall validation research, a one-size-fits-all approach is untenable. Accurate dietary intake assessment in specialized populations—specifically low-income, low-literacy, and pediatric groups—requires meticulously adapted methodologies that address unique socioeconomic, cognitive, and developmental constraints. These populations present distinct challenges, including limited technological access, cultural and linguistic barriers, and cognitive immaturity, which, if unaddressed, introduce significant measurement error and bias into nutrient intake estimates [43] [20]. This document provides detailed application notes and experimental protocols for validating and adapting 24-hour recall techniques to ensure data quality, reliability, and validity in these critical demographic segments. The guidance synthesizes recent empirical evidence to support researchers in generating robust, population-specific dietary data essential for informing clinical research, public health policy, and nutritional interventions.

Population-Specific Challenges and Adaptation Strategies

2.1 Low-Income and Low-Literacy Populations In low-income and low-literacy contexts, standard dietary assessment methods encounter barriers related to infrastructure, education, and cultural norms. Key challenges include limited or unreliable network connectivity for web-based tools, low literacy levels that preclude self-administered questionnaires, and a lack of familiarity with complex research protocols [43]. Furthermore, food insecurity can lead to high day-to-day variability in dietary intake, complicating the estimation of usual consumption [20].

Adaptation strategies must prioritize accessibility and cognitive simplicity. The Voice-Image Solution for Individual Dietary Assessment (VISIDA) system, validated in a Cambodian population, demonstrates a successful approach by leveraging voice recordings and images as primary data capture methods, thereby bypassing the need for literacy [43]. This method can function on smartphones without a continuous network connection and uses intuitive, visual data input. Additionally, careful protocol design must account for seasonal fluctuations in food availability and ensure proportionate representation of all days of the week to mitigate "nuisance effects" on intake data [20].

2.2 Pediatric Populations The primary challenge in pediatric dietary assessment, particularly for children under 12 years, is their ongoing cognitive development, which affects memory, attention span, and understanding of portion size estimation concepts [44] [45]. The age at which children can reliably self-report intake without parental assistance is not clearly defined and can vary significantly [44].

Adaptations should be developmentally tailored. For school-aged children (∼8 years and above), the 24-hour recall assisted by parental food records has been validated as a reliable method for group-level analysis [45]. For older children and adolescents (10-14 years), research from Burkina Faso indicates that with adequate training and probing, self-reported recalls can yield acceptable equivalence to observed intake, though some underestimation persists [44]. Utilizing interactive, age-appropriate tools on mobile devices can improve engagement and accuracy. Portion size estimation, a common source of error, can be improved by providing physical aids like standard bowls and plates or using validated photographic atlases [44].

Experimental Protocols for Validation and Data Collection

Validating adapted 24-hour recall methods against a reference standard is paramount to establishing their reliability in a target population. The following protocols outline structured approaches for this purpose.

3.1 Protocol for Validating a Novel Tool in a Low-Literacy Setting This protocol is based on the validation study of the VISIDA system in Cambodia [43].

• Objective: To determine the relative validity, test-retest reliability, and acceptability of a voice- and image-based dietary assessment system among low-literacy women and their children.
• Population: Mothers (≥18 years) and one of their children (≤5 years) from diverse socio-geographic settings (e.g., rural, semi-rural, urban).
• Study Design: A free-living, observational study with data collection over approximately four weeks.
• Ethical Considerations: Obtain ethical approval from relevant national and institutional review boards. Provide all study information and consent forms in the native language and secure recorded verbal consent to accommodate low literacy [43].
• Data Collection Workflow:
  • Demographic & Baseline Data: Collect household and participant demographic information.
  • Period 1 (Week 1): Dietary intake is collected using the novel tool (e.g., VISIDA) for 3 non-consecutive days, including one weekend day.
  • Period 2 (Weeks 2-3): Conduct three interviewer-administered multiple-pass 24-hour recalls on non-consecutive days. This serves as the reference method.
  • Period 3 (Week 4): Repeat the data collection using the novel tool for another 3 non-consecutive days to assess test-retest reliability.
  • Feedback Survey: Administer an acceptability survey to participants regarding the ease of use and burden of the novel tool.
• Statistical Analysis:
  • Use a linear mixed model to examine differences in estimated nutrient intakes between the three recording periods.
  • Assess test-retest reliability by comparing nutrient intakes from Period 1 and Period 3.
  • Analyze acceptability survey data using descriptive statistics.

3.2 Protocol for Validating 24-Hour Recall in Pediatric Populations This protocol is adapted from a validation study with adolescents in Burkina Faso [44].

• Objective: To validate the 24-hour recall method against the gold standard of observed weighed records in a pediatric population.
• Population: Children or adolescents stratified by age groups (e.g., 10-11 years and 12-14 years), recruited from schools and communities.
• Reference Method - Observed Weighed Record (OWR):
  • Trained research assistants accompany the child from the first to the last meal of the day.
  • All foods and beverages consumed are weighed before and after consumption using digital scales accurate to 1 gram.
  • Ingredients for recipe preparations are weighed when possible, or detailed descriptions are recorded.
• Test Method - 24-Hour Recall (24HR):
  • On the day following the OWR, a different trained research assistant conducts a multiple-pass 24-hour recall interview with the child.
  • Utilize portion size estimation aids (e.g., standard plates, bowls, playdough for solid foods, volumetric measures for liquids) that were provided to the household beforehand.
  • The child is asked to describe the main ingredients of any mixed dishes.
• Data Analysis:
  • Convert food intakes from both methods to nutrient intakes using a standardized food composition database.
  • Test for equivalence by comparing the ratios (24HR/OWR) for energy and key nutrients. Pre-specified equivalence margins (e.g., within ±10%, 15%, or 20%) should be established.
  • Analyze the prevalence of food omissions and misreporting by comparing the food lists from both methods.

Table 1: Key Considerations for Pediatric Validation Studies

Aspect	Consideration for Protocol	Rationale
Age Groups	Stratify by narrow age bands (e.g., 8-9, 10-11, 12-14 years).	Cognitive abilities and memory recall capacity develop rapidly during childhood [44].
Portion Aids	Provide physical aids (plates, bowls) and use age-appropriate aids like playdough.	Children have difficulty estimating portion sizes abstractly; physical aids improve accuracy [44].
Interviewer	Use interviewers trained in child-specific communication techniques.	Reduces intimidation and improves a child's ability to recall and communicate dietary information.
Parental Role	Define and standardize the role of parents (e.g., no assistance, memory prompt only).	Parental presence can influence a child's responses; a standardized protocol ensures consistency [44].

The Scientist's Toolkit: Research Reagent Solutions

Successful implementation of adapted dietary assessment requires specific tools and resources. The following table details essential "research reagents" for this field.

Table 2: Key Research Reagents and Resources for Dietary Recall Validation

Item Name	Function/Application	Specifications & Examples
Automated Self-Administered 24-Hour (ASA24)	A free, web-based system for automating multiple-pass 24-hour recalls and food records.	Utilizes the USDA's Automated Multiple-Pass Method (AMPM). Appropriate for respondents with a ~5th-grade reading level. Canadian and Australian versions available [5].
VISIDA (Voice-Image Solution for Individual Dietary Assessment)	A system using voice recordings and images as the primary data capture method.	Circumvents literacy barriers. Suitable for low-literacy populations; can function offline on smartphones [43].
Multiple-Pass 24-Hour Recall Protocol	A structured interview method to minimize forgotten foods and improve portion size estimation.	Consists of multiple passes (e.g., quick list, detail, review). Can be interviewer-administered or automated [20] [5].
Validated Portion Size Photographs	A visual aid for self-estimation of consumed food amounts.	Photographs of weighed servings of common foods. Reduces error compared to verbal descriptions alone [44].
Standardized Food Composition Database	Converts reported food consumption into nutrient intake data.	Must be context-specific, including local foods and recipes. Critical for accurate nutrient analysis in LMICs [20] [46].
Doubly Labeled Water (DLW)	The gold standard for measuring total energy expenditure.	Used as a reference method to validate the accuracy of energy intake reporting in 24-hour recalls and detect systematic under-reporting [20].
Hydroflumethiazide-13CD2	Hydroflumethiazide-13CD2 Stable Isotope	Hydroflumethiazide-13CD2 is a stable isotope-labeled internal standard for research on diuretic mechanisms and pharmacokinetics. For Research Use Only. Not for human use.
Flucytosine-13C,15N2	Flucytosine-13C,15N2, MF:C4H4FN3O, MW:132.07 g/mol	Chemical Reagent

Workflow and Data Analysis Diagrams

The following diagram illustrates the logical workflow for designing and executing a dietary recall validation study for a specific population, integrating the key phases of preparation, data collection, and analysis.

Validation Study Design Workflow

For the critical data analysis phase, the process of converting raw dietary data into validated usual intake estimates involves specific statistical procedures, as shown below.

Dietary Data Analysis Pathway

Rigorous validation of 24-hour dietary recalls for low-income, low-literacy, and pediatric groups is not merely a methodological refinement but a fundamental requirement for generating meaningful nutritional data. The strategies and protocols outlined herein—embracing technology-assisted voice and image capture, developmentally appropriate portion estimation aids, and robust statistical modeling of usual intake—provide a roadmap for enhancing accuracy and reducing measurement bias. By systematically addressing the unique barriers these populations face, researchers can significantly improve the quality of dietary assessment. This, in turn, strengthens the evidence base for public health initiatives, clinical guidelines, and policy decisions aimed at improving nutritional status and health outcomes in these vulnerable and often under-represented groups.

Mitigating Error and Bias: Strategies for Enhanced Data Quality

The 24-hour dietary recall (24HR) is a foundational tool in nutritional epidemiology, designed to capture detailed information about all foods and beverages consumed by an individual over the previous 24-hour period [1]. As a self-reported instrument, it is inherently susceptible to various forms of measurement error that can significantly impact data quality and subsequent research findings [47]. Understanding, identifying, and categorizing these errors is therefore crucial for researchers conducting dietary validation studies, particularly those focused on 24HR validation techniques. Measurement error in dietary assessment refers to the difference between the true dietary intake and the reported intake, arising from multiple sources throughout the data collection process [47]. These errors can be broadly classified as either random or systematic, each with distinct implications for data analysis and interpretation. In the context of a broader thesis on 24-hour dietary recall validation, this categorization provides the necessary framework for developing effective mitigation strategies and statistical corrections.

Theoretical Framework: Classification of Measurement Errors

Random Measurement Error

Random errors are unpredictable fluctuations in reported intake that do not follow a consistent pattern. In 24HRs, these primarily arise from day-to-day variation in an individual's diet and random misestimation of portion sizes [1] [47]. A key characteristic of random error is that it is non-differential, meaning it is unrelated to the true exposure level or the outcome of interest [48]. The primary consequence of random error is to attenuate (weaken) observed associations between dietary exposures and health outcomes, reducing statistical power and driving effect estimates toward the null [49] [48]. For instance, a single 24HR is unable to account for day-to-day variation in diet, and this inherent variability is a major source of random error when trying to estimate usual intake [1].

Systematic Measurement Error

Systematic error, or bias, refers to consistent inaccuracies that push measurements in a specific direction. Unlike random error, systematic error can be differential, where the reporting error is related to the outcome or another characteristic of the participant [48]. Major types of systematic error in 24HRs include:

• Recall Bias: The omission of foods, beverages, or specific details (e.g., additions like sauces or dressings) from the memory of consumption [47]. Commonly forgotten items include condiments, vegetables in mixed dishes, and ingredients in multicomponent foods [47].
• Social Desirability Bias: The systematic under-reporting of foods perceived as unhealthy (e.g., snacks, sweets) and over-reporting of foods perceived as healthy (e.g., fruits, vegetables) [9] [47].
• Systematic Portion Size Misestimation: A consistent tendency to over- or under-estimate the volumes of foods consumed, often influenced by the body weight and body image perceptions of the respondent [4].

The direction of bias introduced by systematic error is often unpredictable and can lead to either an over- or under-estimation of true intake, potentially creating spurious associations or masking real ones [48].

Table 1: Categorization and Impact of Measurement Errors in 24-Hour Dietary Recalls

Error Category	Sub-Type	Common Sources in 24HR	Primary Impact on Data
Random Error	Day-to-day variation	Natural fluctuation in daily food intake	Attenuation of associations toward null; reduced statistical power [1] [48]
	Random misestimation	Inaccurate portion size estimation due to rounding or guessing	Increased within-person variance [47]
Systematic Error	Recall Bias	Omission of foods (e.g., condiments, ingredients in mixed dishes) [47]	Under-estimation of intake for commonly forgotten items
	Social Desirability Bias	Under-reporting of "unhealthy" foods and over-reporting of "healthy" foods [9] [47]	Systematic distortion of reported food and nutrient intakes
	Portion Size Bias	Consistent over- or under-estimation of volumes	Fixed directional bias in energy and nutrient estimates

Quantitative Assessment of Measurement Error

Validation studies against objective biomarkers provide the most robust quantitative evidence of measurement error in 24HRs. The following table synthesizes key findings from recent studies that compare 24HR-derived estimates with biomarker measurements.

Table 2: Quantitative Evidence of Measurement Error from Biomarker Validation Studies

Reference & Tool	Nutrient/Food	Comparison Method	Key Metric & Result	Implied Error Type
myfood24 (Danish Adults) [9]	Protein	Urinary urea	Correlation (ρ) = 0.45	Random (Moderate correlation)
	Potassium	Urinary potassium	Correlation (ρ) = 0.42	Random (Moderate correlation)
	Folate	Serum folate	Correlation (ρ) = 0.62	Random (Stronger correlation)
NHANES 2014 (24HDR) [49]	Sodium	24-h Urinary Excretion	Mean Bias = -452 mg; Correlation = 0.27	Systematic (Bias) & Random (Low correlation)
	Potassium	24-h Urinary Excretion	Mean Bias = -315 mg; Correlation = 0.35	Systematic (Bias) & Random (Low correlation)
myfood24 (UK Study) [50]	Energy	Total Energy Expenditure	Attenuation Factor = ~0.2-0.3	Primarily Random (Attenuation)
	Protein, Potassium	Urinary Biomarkers	Partial Correlation = ~0.3-0.4	Primarily Random (Low correlation)

The data in Table 2 demonstrates that 24HRs consistently exhibit a combination of systematic bias (evidenced by mean differences) and substantial random error (evidenced by attenuation factors and modest correlation coefficients). The strong correlation for folate in the Danish study suggests that 24HRs may be more valid for ranking individuals by intake of certain nutrients, which is often sufficient for epidemiological studies, rather than for assessing absolute intake [9].

Experimental Protocols for Error Quantification

A robust validation study design is essential to isolate and quantify random and systematic errors. The following protocols, derived from the cited literature, provide a framework for such research.

Protocol 1: Biomarker-Based Validation Study

This protocol leverages objective biomarkers as the reference measure to quantify total measurement error [9] [50].

A. Objective: To assess the validity of a 24HR tool by comparing estimated intakes of specific nutrients with their corresponding recovery biomarkers.

B. Subjects and Study Design:

• Recruitment: Recruit a sample of metabolically stable adults. Participants should be weight-stable and willing to maintain their habitual diet and physical activity levels [9] [50].
• Study Design: A repeated cross-sectional study is recommended. Participants complete multiple 24HRs (e.g., at baseline and 4 weeks later) to also allow for reproducibility analysis [9].

C. Key Measurements and Procedures:

• Dietary Assessment: Participants complete the 24HR tool under investigation (e.g., myfood24) as a 7-day weighed food record or as multiple 24-hour recalls [9].
• Biological Sample Collection: On the final day of dietary recording, collect the following samples:
  • 24-hour Urine Collection: For analysis of urinary urea (protein biomarker), potassium, and sodium [9] [49] [50].
  • Fasting Blood Sample: For analysis of serum biomarkers such as folate [9] [50].
• Energy Expenditure Measurement: Use indirect calorimetry to measure Resting Energy Expenditure (REE) or doubly labeled water (DLW) for Total Energy Expenditure (TEE) to evaluate energy intake misreporting [9] [50].
• Data Analysis:
  • Calculate mean bias (reported intake minus biomarker value) to quantify systematic error [49].
  • Compute Spearman's rank correlation coefficients (ρ) between reported intake and biomarker values to assess the tool's ability to rank individuals, which is influenced by random error [9].
  • Apply the Goldberg cut-off to identify and classify participants as acceptable or misreporters of energy intake [9].

Protocol 2: Controlled Feeding Study with Item-by-Item Analysis

This protocol allows for a detailed analysis of food-specific errors, including omissions and intrusions, by comparing reported intake to a known, provided diet [51] [4].

A. Objective: To evaluate the accuracy of a 24HR in reporting specific food items and portion sizes under controlled conditions.

B. Subjects and Study Design:

• Recruitment: Recruit participants for a controlled feeding study. The sample can focus on specific groups of interest, such as older adults or different ethnicities [51].
• Study Design: Participants are provided with all meals and snacks for a set period (e.g., 5 days) in a controlled environment. Research staff monitor and record actual consumption [51].

C. Key Measurements and Procedures:

• Diet Provision: Prepare and serve all meals. Weigh all ingredients to the nearest 0.1g during preparation and pre-portioned meals to the nearest 1g before serving [51].
• Unobtrusive Documentation: Staff should monitor mealtimes to ensure compliance and document actual consumption without informing participants that a recall will be performed, to mimic a free-living setting and avoid primed memory [51] [4].
• 24HR Administration: On a randomly selected day, conduct an interviewer-administered 24HR using the Multiple-Pass Method (MPM). Use food models and measuring guides to aid portion size estimation [51].
• Data Analysis:
  • Food Item Accuracy: Classify each reported item as a match (correctly reported), exclusion (consumed but not reported), or intrusion (reported but not consumed) [51].
  • Portion Size Accuracy: Categorize portion sizes as corresponding (≤10% error), overreported, or underreported [51].
  • Nutrient Analysis: Calculate nutrient intakes from both the provided diet and the 24HR data and compare using paired t-tests [51].

Visualization of Error Pathways and Mitigation Strategies

The following diagram illustrates the pathways through which measurement errors are introduced during a 24HR assessment and highlights key mitigation strategies supported by the experimental protocols.

The Researcher's Toolkit: Essential Reagents and Materials

Table 3: Key Research Reagents and Materials for 24HR Validation Studies

Item	Specification / Example	Primary Function in Validation
Biomarker Assays	Urinary nitrogen (urea), potassium, sodium; Serum folate; Doubly Labeled Water (DLW)	Serve as objective, non-self-reported reference measures for specific nutrient intakes and total energy expenditure [9] [50].
Indirect Calorimeter	Device to measure Resting Energy Expenditure (REE) via oxygen consumption and carbon dioxide production.	Provides an objective measure of energy metabolism to identify misreporters of energy intake [9].
Standardized Food Composition Database	e.g., UK Composition of Foods, USDA FCDB, Branded Food Databases.	Converts reported food consumption into nutrient intake data; accuracy is critical to minimize database-related errors [6] [50].
Portion Size Estimation Aids	Food model booklets, image atlases, standardized photographs, geometric shapes (circles, grids), rulers [51] [4].	Helps participants conceptualize and report the volume of food consumed, reducing random and systematic errors in portion size estimation.
Web-Based 24HR Tool	e.g., myfood24, ASA24, Intake24, Foodbook24 [9] [6] [50].	Automated, self-administered tool that standardizes the recall process, reduces interviewer bias, and can incorporate portion images and multiple passes.
Structured Interview Protocol	Automated Multiple-Pass Method (AMPM) prompt sheets [50] [51].	Guides interviewers to probe for commonly forgotten foods (e.g., condiments, sauces) in a standardized way, mitigating recall bias.

The rigorous categorization of measurement errors into random and systematic types is not merely an academic exercise but a practical necessity for advancing the science of dietary assessment. As detailed in this application note, each error type has distinct origins and consequences, necessitating tailored validation approaches. The quantitative data and experimental protocols provided establish that while measurement error is an inherent challenge in 24HRs, its impact can be quantified through biomarker studies and controlled feeding designs, and mitigated through methodological choices such as the Multiple-Pass Method, portion size aids, repeated administrations, and statistical calibration. For researchers embarking on 24-hour dietary recall validation, a comprehensive strategy that acknowledges, measures, and adjusts for both random and systematic error is fundamental to producing reliable and meaningful data that can robustly inform public health policy and nutritional epidemiology.

Accurate dietary intake data is fundamental to nutritional epidemiology, clinical practice, and public health policy. However, self-reported dietary data, particularly from 24-hour recalls and food frequency questionnaires, is consistently compromised by systematic misreporting, with under-reporting of energy intake being the most prevalent and significant error [52]. This bias fundamentally undermines the validity of diet-disease association studies and the effectiveness of nutritional interventions.

The doubly labeled water (DLW) method has emerged as the gold standard for validating self-reported energy intake in free-living individuals. By measuring total carbon dioxide production through the differential elimination rates of stable isotopes of hydrogen (deuterium) and oxygen (O18), DLW provides an objective measure of total energy expenditure (TEE) with an analytical error of approximately 7% [53] [54]. In weight-stable individuals, TEE should equal energy intake, creating a robust criterion for identifying misreporting in dietary recalls.

This article synthesizes recent advances in using DLW methodology to quantify, understand, and address the pervasive challenge of dietary under-reporting, providing researchers with practical protocols and analytical frameworks to enhance the validity of dietary assessment in both research and clinical contexts.

Quantitative Insights into Under-Reporting

Prevalence and Magnitude of Under-Reporting

Recent studies utilizing DLW validation have consistently demonstrated alarmingly high rates of energy intake under-reporting across diverse populations, with specific subgroups exhibiting particularly pronounced effects.

Table 1: Prevalence of Energy Intake Under-Reporting Across Populations

Population	Under-Reporting Prevalence	Key Factors	Citation
Older Adults with Overweight/Obesity	50%	Higher BMI, older age	[52]
General Population (NDNS/NHANES)	27.4% (average misreporting)	BMI, sex, age	[53]
Adults (Men)	60%	Sex, physical activity level	[55]
Adults (Women)	56%	Sex, physical activity level	[55]
Older Korean Adults	~29% (portion size overestimation)	Cultural diet composition	[40]

The systematic nature of under-reporting introduces substantial bias in nutritional studies. Research has demonstrated that while self-reported energy intake (rEI) showed no significant relationship with weight (ß = 13.1, p = 0.06) or BMI (ß = 41.8, p = 0.11) when all reports were included, statistically significant relationships emerged when implausible reports were excluded using DLW-based methods [52]. This pattern confirms that under-reporting is not random but is correlated with anthropometric measures, potentially obscuring true associations between diet and health outcomes.

Predictive Equations for Energy Expenditure

While DLW remains the criterion method for TEE measurement, its high cost and technical demands limit widespread application. Consequently, researchers have developed predictive equations derived from large DLW datasets to estimate energy requirements and identify misreporting.

Table 2: Predictive Equations for Energy Expenditure Based on DLW Data

Equation Source	Sample Size	Input Variables	Variance Explained (R²)	Application Notes
Speakman et al. (2025)	6,497 individuals	Body weight, height, age, sex, ethnicity, elevation	69.8%	Uses natural log of body weight; includes interaction terms [53]
NASEM (2023)	International DLW dataset	Age, sex, weight, height, physical activity level	Not specified	Part of Dietary Reference Intakes update [56]
Porter et al. (2024)	1,657 older adults	Age-specific parameters for >65 years	Not specified	Integrated dataset of 39 DLW studies [56]

The equation developed by Speakman et al. represents a significant advancement in the field, incorporating not only basic anthropometric and demographic variables but also accounting for elevation above sea level and ethnicity, with specific coefficients for different ethnic groups [53]. The full equation takes the form:

ln(TEE) = -0.2172 + 0.4167 × ln(BW) + 0.006565 × Height - 0.02054 × Age + 0.0003308 × Age² - 0.000001852 × Age³ + 0.09126 × ln(Elevation) - 0.04092 × Sex + ethnicity terms + interaction terms

This equation enables researchers to calculate expected TEE and establish 95% predictive limits for identifying potentially misreported dietary recalls in large-scale studies where direct DLW measurement is impractical [53].

Experimental Protocols for DLW Validation

Core DLW Methodology for TEE Assessment

The DLW method involves precise protocols for isotope administration, sample collection, and analytical procedures to ensure accurate TEE measurement.

Protocol Title: Doubly Labeled Water Administration and Total Energy Expenditure Measurement

Principle: The method is based on the differential elimination of stable isotopes of hydrogen (deuterium, ²H) and oxygen (¹⁸O) from body water. Deuterium is eliminated as water, while oxygen is eliminated as both water and carbon dioxide. The difference in elimination rates therefore reflects carbon dioxide production [54].

Materials Required:

• Doubly labeled water (deuterium oxide and oxygen-18 water)
• Calibrated analytical balance for dose preparation
• Sterile urine collection containers
• Isotope ratio mass spectrometer
• Freezer facilities for sample storage (-20°C)

Step-by-Step Procedure:

• Pre-Dose Baseline Sample Collection:

  • Collect baseline urine sample before isotope administration
  • Participants should avoid food and water for 1-2 hours before baseline collection
  • Record exact time of collection [54]

• DLW Dose Preparation and Administration:

  • Prepare dose based on body weight or estimated total body water
  • Typical doses: ¹⁸O at 150-174 mg/kg and ²H at 70-80 mg/kg body weight
  • For obese participants, consider dosing based on estimated total body water to reduce costs [54]
  • Administer orally under supervision; ensure complete consumption

• Post-Dose Sample Collection:

  • Collect urine samples at 3-4 hours post-dose for isotope equilibration assessment
  • Continue daily urine collection for 12-14 days at approximately the same time each day
  • Participants should record exact collection time for each sample
  • Store samples in refrigerator during collection period; freeze at -20°C after completion [54]

• Isotopic Analysis:

  • Analyze urine samples using isotope ratio mass spectrometry
  • Determine elimination rates of both isotopes from the disappearance curves
  • Calculate pool sizes for hydrogen and oxygen [54]

• Calculation of Energy Expenditure:

  • Compute carbon dioxide production rate from differential isotope elimination
  • Convert to total energy expenditure using Weir equation: TEE = (1.106 × rCO₂ + 3.941 × rO₂) × 4.1868 kJ/day [54]
  • Assume respiratory quotient of 0.86 if no dietary data available, or derive from food diaries

Quality Control Considerations:

• Participants should maintain normal activity patterns during measurement period
• Record any deviations from protocol (missed samples, illness, unusual activity)
• Analyze samples in batches with appropriate standards and controls
• The two-point method (baseline and end point) can be used as a more cost-effective alternative to multiple sampling points [54]

Identification of Misreporting Using DLW-Based Methods

Protocol Title: Validation of Self-Reported Energy Intake Against DLW-Measured TEE

Principle: In weight-stable individuals, energy intake should equal total energy expenditure. Significant discrepancies between self-reported energy intake (rEI) and TEE measured by DLW (TEE_DLW) indicate misreporting [52].

Materials Required:

• DLW-derived TEE measurements (from Protocol 3.1)
• Self-reported energy intake data from 24-hour recalls or food records
• Anthropometric measurements (weight, height)
• Body composition data (if available)

Step-by-Step Procedure:

• Data Collection:

  • Obtain multiple (preferably 3-6 non-consecutive) 24-hour dietary recalls within the DLW measurement period
  • Measure body weight at beginning and end of DLW period to confirm weight stability
  • Calculate within-subject coefficient of variation for rEI to account for day-to-day variation [52]

• Calculation of Misreporting Indicators:

  • Compute rEI:TEE_DLW ratio for each participant
  • Calculate individual percentage bias: (rEI - TEE_DLW)/TEE_DLW × 100%
  • Group-level cutoffs: Classify as under-reported (< -1SD), plausible (±1SD), or over-reported (> +1SD) based on distribution of rEI:TEE ratios [52]

• Application of Predictive Equations (when direct DLW not available):

  • Use published equations (e.g., Speakman et al. [53]) to predict expected TEE
  • Calculate 95% prediction intervals for expected TEE
  • Identify potentially misreported records as those where rEI falls outside prediction intervals

• Novel Energy Balance Method (as alternative to TEE comparison):

  • Calculate measured energy intake (mEI) as: mEI = TEE_DLW + ΔES
  • Where ΔES (change in energy stores) is derived from body composition changes: ΔES = (ΔFM × 9.75) + (ΔFFM × 1.13) [52]
  • Compute rEI:mEI ratio as potentially more accurate comparison than rEI:TEE

Interpretation Guidelines:

• Under-reporting: rEI:TEE_DLW < 0.85 or below lower 95% prediction limit
• Plausible reporting: rEI:TEE_DLW between 0.85-1.15 or within 95% prediction limits
• Over-reporting: rEI:TEE_DLW > 1.15 or above upper 95% prediction limit
• Consider individual variability (±2.5% day-to-day variation in TEE, ±23% in rEI) when interpreting results [52]

The Researcher's Toolkit

Essential Research Reagents and Materials

Table 3: Key Reagents and Materials for DLW Studies

Item	Specifications	Function/Application	Technical Notes
Deuterium Oxide (²H₂O)	99.8% atomic percent excess (APE)	Stable isotope tracer for hydrogen	Dose: 0.12 g per kg body weight [52]
Oxygen-18 Water (H₂¹⁸O)	10.8% APE	Stable isotope tracer for oxygen	Dose: 1.68 g per kg body weight [52]
Isotope Ratio Mass Spectrometer	High-precision instrument	Measures isotopic enrichment in biological samples	Requires specialized facility and expertise [54]
Urine Collection Containers	Sterile, leak-proof	Sample collection over 14-day period	Label with participant ID, date, time of collection [54]
Automated Dietary Assessment Tool	e.g., ASA24, Intake24	Standardized collection of 24-hour recalls	Reduces administrative burden; enables automatic coding [5] [33]

Analytical Framework for Data Interpretation

The following diagram illustrates the decision process for identifying and addressing dietary misreporting in research studies:

The integration of doubly labeled water methodology into dietary assessment validation represents a paradigm shift in nutritional epidemiology. By providing an objective biomarker of energy expenditure, DLW has revealed the alarming extent and systematic nature of dietary misreporting, which disproportionately affects specific subgroups including individuals with higher BMI, older adults, and women.

The protocols and analytical frameworks presented herein provide researchers with practical tools to quantify and address this bias, either through direct DLW measurement or through predictive equations derived from large DLW datasets. As the field advances, the incorporation of these validation techniques will be essential for producing more reliable diet-disease association data and developing effective, evidence-based nutritional interventions.

Future directions should focus on refining predictive equations for diverse populations, reducing the costs and technical barriers of DLW methodology, and developing standardized protocols for identifying and correcting for misreporting in large-scale epidemiological studies.

Accurate dietary assessment is fundamental for advancing research in nutritional epidemiology, clinical nutrition, and public health policy development. The 24-hour dietary recall (24HR) stands as a widely employed methodology for capturing detailed dietary intake data. However, its accuracy is inherently challenged by several factors, including recall bias, portion size estimation errors, and day-to-day variability in food consumption [57] [58]. This application note delineates a refined protocol for 24HR administration, focusing on two critical optimization strategies: the implementation of advanced portion size estimation aids and the strategic use of multiple, non-consecutive recall days. Grounded in contemporary validation research, this protocol aims to enhance data reliability for researchers and professionals engaged in diet-health relationship studies.

Key Concepts and Quantitative Evidence

The optimization of 24HR protocols is supported by empirical evidence quantifying the impact of portion size aids and the number of recall days on data quality. The tables below summarize key findings from recent investigations.

Table 1: Impact of Protocol Design on Dietary Data Quality

Optimization Strategy	Key Evidence	Data Source & Context
Comprehensive Food Lists & Portion Images	86.5% (302/349) of consumed foods were available in the expanded food list; strong correlations (r=0.70-0.99) for 15/26 nutrients vs. interviewer-led recall [6].	Foodbook24 expansion for Brazilian, Irish, and Polish adults in Ireland [6].
Multiple Non-Consecutive Days (including weekends)	3-4 days of data required for reliable estimation (r > 0.8) of most micronutrients and food groups; significant day-of-week effects for energy, carbs, and alcohol [13].	"Food & You" digital cohort (n=958) in Switzerland, using the MyFoodRepo app [13].
Short Recall Intervals	Repeated 2-hour and 4-hour recalls reduce memory-related bias and improve accuracy compared to traditional 24-hour recalls [58].	Evaluation of the Traqq app among Dutch adolescents (n=102) [58].

Table 2: Minimum Days Required for Reliable Estimation of Dietary Components

Dietary Component	Minimum Days for Reliability (r > 0.8)	Notes
Water, Coffee, Total Food Quantity	1-2 days	Achieves high reliability quickly [13].
Macronutrients (Carbohydrates, Protein, Fat)	2-3 days	Good reliability with a short assessment period [13].
Micronutrients, Meat, Vegetables	3-4 days	Requires more days due to higher variability in consumption [13].

Experimental Protocols & Workflows

Protocol 1: Validation of Expanded Food Lists and Portion Size Aids

This protocol is designed to enhance the cultural and culinary relevance of dietary assessment tools, thereby improving portion size estimation.

1. Objective: To expand and validate a web-based 24HR tool with a comprehensive, culturally-specific food list and standardized portion size images. 2. Materials:

• Software Platform: A web-based 24HR tool (e.g., Foodbook24, ASA24) [6] [5].
• Food Composition Data: National nutrient databases (e.g., UK CoFID, Brazilian, Polish databases) and local sources for branded items [6].
• Portion Size Image Bank: High-quality, standardized photographs of foods and beverages in multiple serving sizes. 3. Methodology:
• Food List Expansion:
  • Review national food consumption surveys and relevant literature for the target population to identify frequently consumed foods [6].
  • Add new food items (e.g., 546 foods were added for Brazilian and Polish populations) and translate the interface into relevant languages [6].
  • Link new foods to nutrient composition data, prioritizing official databases and using food labels for branded products as a secondary source [6].
• Portion Size Estimation:
  • Derive medium portion sizes from the mean reported intake in national surveys.
  • Calculate smaller and larger portions using standard deviations (SDs) from the mean. For infrequently consumed foods with large SDs, use a fixed percentage (e.g., ±25%) [6].
  • Develop a corresponding image for each portion size option using a standardized photographic procedure.
• Validation:
  • Acceptability Study: Participants provide a visual record of their diet. Calculate the percentage of consumed foods available in the new list (e.g., 86.5% coverage) [6].
  • Comparison Study: Conduct a 24HR using the updated tool and an interviewer-led recall on the same day, repeated after two weeks. Analyze data using Spearman rank correlations and Mann-Whitney U tests to compare food group and nutrient intakes [6].

Protocol 2: Determining Minimum Days for Reliable Usual Intake

This protocol establishes the number of non-consecutive recall days needed to estimate usual intake for different nutrients and food groups.

1. Objective: To determine the minimum number of non-consecutive days required to obtain a reliable estimate of usual dietary intake at the population level. 2. Materials:

• Digital Dietary Assessment Tool: A mobile or web-based application for real-time tracking (e.g., MyFoodRepo, Traqq) [13] [58].
• Statistical Software: Capable of linear mixed models (LMM) and intraclass correlation coefficient (ICC) analysis (e.g., R, Python with statsmodels). 3. Methodology:
• Data Collection:
  • Recruit a large cohort (e.g., n > 900) and collect dietary data for an extended period (e.g., 2-4 weeks) [13].
  • Collect demographic and anthropometric data (age, sex, BMI) as covariates.
• Data Analysis:
  • Day-of-Week Analysis: Use LMM to test for significant variations in intake (energy, nutrients, food groups) across different days of the week, with participant as a random effect [13].
  • Minimum Days Estimation:
    • Coefficient of Variation (CV) Method: Calculate within- and between-subject variability to estimate the number of days needed to reach a reliability coefficient of r > 0.8 [13].
    • Intraclass Correlation (ICC) Analysis: Compute ICCs for all possible combinations of days to identify the optimal number and sequence of days (e.g., including weekdays and weekends) [13].

Logical Workflow for Protocol Optimization

The following diagram illustrates the decision-making workflow for integrating these optimization strategies into a 24HR study design.

The Researcher's Toolkit: Essential Reagents & Materials

Table 3: Key Research Reagent Solutions for 24HR Validation Studies

Item	Function & Application	Exemplars & Notes
Web-Based 24HR Platforms	Self-administered or interviewer-led dietary recall; automates data coding and nutrient analysis.	ASA24 [5], Intake24 [33], MyFood24 [33], Foodbook24 [6].
Food Composition Databases	Provides nutrient values for reported foods; essential for calculating nutrient intake.	USDA FNDDS [17], UK CoFID [6], local national databases (e.g., Brazilian, Polish) [6].
Standardized Portion Size Image Sets	Visual aids to improve accuracy of portion size estimation during recall.	Custom sets developed from national survey data [6]; images should represent multiple serving sizes.
Biomarker Assays	Objective validation method to assess reporting accuracy for specific nutrients.	Urinary nitrogen (for protein), urinary sodium, serum folate, serum fatty acids [15].
Statistical Analysis Packages	For processing complex dietary data and performing validity statistics (correlation, ICC, LMM).	R, Python (statsmodels), SAS, STATA.

Accurate dietary assessment using 24-hour dietary recalls (24HR) is fundamental to nutrition research, informing public health policy, clinical practice, and understanding diet-disease relationships. However, the validity of 24HR data is often compromised by contextual biases stemming from temporal and cultural factors. Day-of-the-week effects, seasonal variations, and cultural-linguistic barriers introduce measurement errors that can distort estimates of usual intake, leading to flawed conclusions and ineffective interventions [20] [59]. This document, framed within a broader thesis on 24-hour dietary recall validation techniques, outlines the sources of these biases and provides detailed application notes and protocols for mitigating their impact, ensuring data accurately reflect true consumption patterns.

The table below summarizes key evidence on the nature and impact of contextual biases, informing the need for specific mitigation strategies.

Table 1: Evidence Base for Contextual Biases in 24-Hour Dietary Recalls

Bias Type	Key Evidence	Impact on Dietary Data
Day-of-the-Week	Food consumption differs significantly between weekdays and weekends [20].	Single-day recalls can misrepresent habitual intake if day-of-week effects are not accounted for.
Seasonal	Dietary intakes vary by season, particularly for fresh fruits and vegetables [59].	Surveys conducted in one season provide a biased estimate of annual usual intake.
Cultural/Linguistic	Standard food lists often lack culturally-specific items, leading to under-reporting. Brazilian participants using a non-adapted tool omitted 24% of foods vs. 13% in an Irish cohort [6].	Systematically underestimates intake in diverse populations and hinders cross-cultural comparisons.
Recall Interval	Shorter retention intervals (time between eating and recall) significantly increase the number of foods reported, especially for evening meals (5.2 vs. 4.2 foods) [60].	Longer retention intervals increase memory-related under-reporting.

Experimental Protocols for Bias Mitigation

Protocol: Designing a 24HR Survey to Control for Temporal Biases

Application Note: This protocol is designed to integrate mitigation strategies for day-of-the-week and seasonal biases directly into the study design phase, reducing the need for complex statistical corrections later.

Detailed Methodology:

• Determine Recall Days:

  • For population-level estimates: Collect multiple non-consecutive 24HRs per participant. Evidence from Chinese adults shows that two non-consecutive days (NC2) can be equivalent to three consecutive days (C3) for estimating many dietary intakes, improving cost-efficiency [61].
  • For feasibility: If only one recall is possible per participant, ensure the sample is stratified and scheduled so that recall days are proportionally distributed across all seven days of the week [20].

• Schedule Data Collection:

  • Day-of-the-Week: Administer recalls on randomly selected days, ensuring all days of the week (including weekends) are equally represented in the overall sample [20].
  • Seasonal Variation: Conduct surveys over a longer period (e.g., 12 months) or conduct multiple shorter surveys randomly throughout the year to capture seasonal variation in food availability and consumption [20] [59].
  • Avoid Festive Periods: Exclude data collection on major feast days or holidays, as dietary practices during these times are often atypical and not representative of usual intake [20].

• Implement Quality Control:

  • Use a standardized 24HR protocol across all participants and time points, such as the Automated Multiple-Pass Method (AMPM), to minimize random error and enhance comparability [20] [27].

Protocol: Cultural and Linguistic Adaptation of a 24HR Tool

Application Note: This protocol provides a framework for adapting self-administered web-based 24HR tools for use with specific cultural or ethnic groups, ensuring their diets are accurately captured.

Detailed Methodology:

• Expand the Food List:

  • Identify Culturally-Relevant Foods: Review national food consumption surveys, scientific literature, and dietary guidelines from the target population's country of origin [6].
  • Add New Foods: Incorporate identified frequently consumed foods and beverages into the tool's database. For example, an expansion of the Foodbook24 tool for Brazilian and Polish populations in Ireland added 546 new food items [6].
  • Link to Nutrient Database: Assign appropriate nutrient composition data, prioritizing the host country's database (e.g., UK CoFID). For unique items, use nutrient composition databases from the country of origin [6].

• Translate and Localize the Interface:

  • Professional Translation: Translate the entire user interface, including food names, meal occasions, and instructions, into the target language(s) [6].
  • Cultural Appropriateness: Ensure that portion size images, meal names, and recipe examples are relevant and recognizable to the target culture.

• Validate the Adapted Tool:

  • Conduct an Acceptability Study: Have participants from the target group list their habitual diet and check the availability of these items in the adapted food list. A well-adapted list should represent >85% of commonly consumed foods [6].
  • Perform a Comparison Study: Compare dietary intakes (food groups and nutrients) from the adapted self-administered tool against a traditional method like an interviewer-administered 24HR. Strong correlations (e.g., r > 0.70) for most nutrients and food groups indicate successful adaptation [6] [27].

Protocol: Validating 24HR Data Using a Reference Measure

Application Note: This protocol is used to detect and correct for systematic errors like energy under-reporting, which can be exacerbated by contextual biases. It is particularly critical for validation studies within a thesis.

Detailed Methodology:

• Select an Appropriate Reference Measure:

  • Recovery Biomarkers: These are the gold standard for validating specific nutrients.
    • Energy: Doubly Labeled Water (DLW) to measure energy expenditure [20] [19].
    • Protein: Urinary nitrogen [20].
    • Sodium & Potassium: 24-hour urinary excretion [20] [19].
  • Weighed Food Records: Conducted on the same day as the 24HR to provide a direct comparison of reported versus actual intake, though they are also subject to error [20] [40].

• Execute the Validation Study:

  • Participant Recruitment: Recruit a sub-sample representative of the main study population.
  • Simultaneous Data Collection: Administer the 24HR and collect the reference measure (e.g., urine collection for 24 hours) over the same time period.
  • Blinding: Where possible, keep participants and staff blinded to the results of the reference measure during the 24HR administration to prevent bias.

• Analyze and Interpret Data:

  • Statistical Comparison: Use paired t-tests, correlation analyses, and Bland-Altman plots to assess the agreement between the 24HR and the reference measure [40] [27] [23].
  • Calculate Reporting Accuracy: Determine the degree of under- or over-reporting, for example, by calculating the ratio of 24HR energy intake to DLW-measured energy expenditure.

Visualization of Workflows

The following diagram illustrates the integrated workflow for designing a 24HR survey that addresses multiple contextual biases, as outlined in Section 3.1.

Diagram 1: 24HR survey design workflow for temporal bias mitigation.

The following diagram maps the systematic process for the cultural and linguistic adaptation of a 24HR tool, as described in Section 3.2.

Diagram 2: Workflow for cultural and linguistic adaptation of a 24HR tool.

The Scientist's Toolkit: Key Research Reagents & Materials

Table 2: Essential Materials and Tools for 24HR Validation and Bias Mitigation

Tool/Reagent	Function & Application Note
Standardized 24HR Protocol (e.g., AMPM)	A structured interview protocol using multiple passes (quick list, forgotten foods, detail cycle, final review) to enhance memory retrieval and reduce food omission [20] [27].
Doubly Labeled Water (DLW)	The gold-standard recovery biomarker for validating total energy intake by measuring metabolic carbon dioxide production and thus energy expenditure [20] [19].
24-Hour Urine Collection	A recovery biomarker for validating sodium, potassium, and protein (via nitrogen) intake. Requires careful participant instruction and compliance monitoring [20] [19].
Web-Based 24HR Tool (e.g., ASA24, Intake24)	Self-administered, automated systems that reduce interviewer burden and cost. They can be programmed for random, non-consecutive day recalls and adapted for cultural use [19] [6] [60].
Validated Portion Size Image Atlas	A library of food photographs depicting multiple serving sizes. Critical for allowing respondents to self-estimate portion sizes accurately in web-based or interviewer-led recalls [6] [27] [60].
Pictorial Recall Aids	Physical or digital aids (e.g., booklets of food pictures) given to participants post-consumption to aid subsequent recall. Proven to help identify omitted items, particularly beverages, snacks, and fruits [62].
National Food Composition Database	The foundational database linking consumed foods to their nutrient content. Must be updated and expanded to include culturally-specific foods for accurate nutrient intake calculation [6].
Statistical Modeling Software (e.,g., R, SAS with NCI Macros)	Software capable of implementing the National Cancer Institute (NCI) method or other statistical models to adjust for within-person variation and estimate usual intake distributions from short-term recalls [61].

Establishing Validity: A Framework for Comparing and Validating Recall Methods

Within nutritional epidemiology, the validation of dietary assessment tools is paramount for generating reliable data capable of informing public health policy and understanding diet-disease relationships. The 24-hour dietary recall (24HR) is widely used in national surveys and research to capture detailed short-term intake [3] [1]. However, to establish the validity of any dietary assessment method, it must be compared against a reference method that provides a superior approximation of true intake. This application note details the two primary categories of reference methods used in 24HR validation: weighed food records (WFR) and objective biomarkers, providing a structured comparison of their application, protocols, and quantitative performance.

Comparative Analysis of Reference Methods

The following tables summarize the core characteristics and validation outcomes of using Weighed Food Records and Biomarkers as reference methods against 24HR.

Table 1: Key Characteristics of Weighed Food Records and Biomarkers as Reference Methods for 24HR Validation

Feature	Weighed Food Records (WFR)	Objective Biomarkers
Primary Principle	Detailed, prospective recording of all foods and beverages consumed, with weights [63].	Measurement of biological compounds (in urine, blood) correlated with nutrient intake [15] [50].
Nature of Measurement	Self-reported, but with reduced reliance on memory due to concurrent recording.	Objective, independent of self-report.
Key Measured Outcomes	Energy and nutrient intake (e.g., protein, potassium, sodium) [63].	Energy expenditure (TEE via DLW); Nutrient-specific intake (e.g., urinary nitrogen for protein, urinary potassium for potassium) [39] [24] [50].
Major Source of Error	Participant burden and reactivity, leading to potential changes in diet or misreporting [3].	Biological variability in metabolism and excretion; Cost and complexity of analysis [39].
Ideal Application	Validation of energy and a wide range of nutrient intakes where detailed food composition data is available.	Gold-standard validation for specific, recoverable nutrients and total energy intake.

Table 2: Summary of Selected 24HR Validation Study Outcomes Against Reference Methods

Study Reference	24HR Tool / Method	Reference Method	Key Quantitative Findings (24HR vs. Reference)
myfood24-Germany, 2021 [63]	Web-based 24HR	Weighed Dietary Record (WDR) & Urinary Biomarkers	vs. WDR: Significant correlations for energy and all nutrients (r=0.45–0.87). No significant difference in mean energy intake. vs. Biomarkers: Protein intake underestimated by 10% vs. urinary nitrogen; no significant difference in potassium intake.
Danish Validation, 2023 [39]	2 x 24HR (Interviewer)	Doubly Labelled Water (TEEDLW)	Energy Intake: Mean reported EI (11.5 MJ/d) was the same as TEEDLW. Proportion of under-reporters was 4%.
US Biomarker Study, 2023 [24]	6 x ASA24 (Self-Administered)	Recovery Biomarkers (Urine, DLW)	Absolute Intakes: Systematically lower than biomarkers. Energy underestimated by 15-17% vs. DLW. Outperformed FFQs for absolute intakes.
PERSIAN Cohort, 2025 [15]	Interviewer FFQ (Validated by 24HR)	Serum & Urinary Biomarkers	Validity Coefficients: For urinary protein and sodium, and serum folate/fatty acids, coefficients were acceptably above 0.4.
Controlled Feeding, 2024 [64]	Four Technology-Assisted 24HRs	True Intake (Weighed Food)	Mean Energy Difference: Ranged from +1.3% (mFR-TA) to +15.0% (IA-24HR) of true intake. ASA24 and Intake24 showed reasonable validity.

Detailed Experimental Protocols

Protocol A: Validation Against Weighed Food Records

This protocol is designed to compare the 24HR against the detailed, prospective WFR for a comprehensive assessment of energy and nutrient intake validity [63].

Workflow Overview:

Step-by-Step Procedure:

• Participant Recruitment and Screening:

  • Recruit a sufficient sample size (e.g., ~100 participants) to detect clinically meaningful differences in nutrient intake [63].
  • Screen for eligibility: inclusion of adults fluent in the relevant language, with internet access, who are weight-stable and not on a weight-loss diet [63] [50].

• Study Visit 1: Training and Baseline Measurements:

  • Obtain written informed consent.
  • Conduct anthropometric measurements (weight, height).
  • Provide standardized, detailed oral and written instructions on completing the WFR [63].
  • Supply participants with materials: digital food scales, paper-based WDR forms, and instructions for describing foods in detail (e.g., brand names, recipes) [63].

• Intervention Phase: Concurrent Data Collection:

  • Weighed Food Record (WFR): Participants prospectively weigh and record all foods and beverages consumed, including leftovers, over a specified period (e.g., 3 consecutive days) [63].
  • 24-Hour Dietary Recall (24HR): On the final day of the WFR period (or for the same specific day), participants complete the 24HR tool under investigation. This can be done at a study center with an assistant or self-administered [63]. The 24HR should capture intake for the exact 24-hour period covered by the WFR.

• Study Visit 2: Data Submission:

  • Participants return their completed WFR forms.
  • Screen WFRs for inconsistencies and clarify entries with participants if necessary [63].

• Data Processing and Analysis:

  • Coding: Manually code the WFRs by trained coders using a standardized food composition database [63] [50].
  • Analysis: Calculate energy and nutrient intakes from both the WFR and the 24HR. Perform statistical comparison using correlation analyses (e.g., Pearson's), paired t-tests/Wilcoxon tests for mean differences, and Bland-Altman analysis to assess limits of agreement [63].

Protocol B: Validation Against Recovery Biomarkers

This protocol uses objective biomarkers to validate nutrient intake and energy expenditure, providing an unbiased assessment of the 24HR's accuracy [24] [50].

Workflow Overview:

Step-by-Step Procedure:

• Participant Recruitment and Screening:

  • Recruit metabolically stable adults (e.g., weight-stable, no chronic diseases affecting metabolism) [39] [50].
  • The required sample size depends on the biomarker and expected correlation but is often in the range of 50-120 participants [39] [50].

• Clinic Visit: Biomarker Administration and Baseline:

  • Obtain written informed consent.
  • Collect baseline urine and/or blood samples.
  • For Total Energy Expenditure (TEE): Administer a dose of Doubly Labelled Water (DLW) and instruct participants on subsequent urine collection over 1-2 weeks [39].
  • For nutrient-specific biomarkers (e.g., 24-hour urinary nitrogen/potassium), provide participants with collection containers and detailed instructions for a complete 24-hour urine collection [63] [50].

• At-Home Data Collection Period:

  • Biomarker Sampling:
    • DLW: Participants collect urine samples at home according to the study protocol (e.g., daily spot urine or full 24-hour collections over ~10 days) [39].
    • 24-hour Urine: Participants collect all urine for a 24-hour period, discarding the first void and collecting all subsequent voids, including the first void the next morning [63].
  • 24-Hour Dietary Recall (24HR): During the biomarker collection period (e.g., on the day of 24-hour urine collection), participants complete one or more 24HRs [63] [50]. The order of 24HR methods can be randomized to counterbalance learning effects [50].

• Final Clinic Visit and Sample Return:

  • Participants return urine samples and complete final questionnaires.
  • Verify the completeness of 24-hour urine collections through protocols (e.g., collection time, total volume) [63].

• Biomarker Analysis and Data Comparison:

  • Laboratory Analysis:
    • Urinary Nitrogen: Analyze using the Dumas method to estimate protein intake (intake = urinary N * 6.25 / 0.8) [63].
    • Urinary Potassium: Analyze using atomic absorption spectroscopy to estimate potassium intake (intake = urinary K / 0.8) [63].
    • DLW: Analyze isotope enrichments by isotope ratio mass spectrometry to calculate TEE [39].
  • Statistical Comparison: Compare reported energy intake from 24HR to TEE from DLW. Compare reported nutrient intakes to biomarker-estimated intakes using correlation coefficients, attenuation factors, and tests for mean differences [24] [50].

The Scientist's Toolkit: Research Reagent Solutions

Table 3: Essential Materials and Tools for 24HR Validation Studies

Item / Solution	Function / Application	Examples & Specifications
Doubly Labelled Water (DLW)	Gold-standard for measuring Total Energy Expenditure (TEE) in free-living individuals to validate reported energy intake [39].	^2Hâ‚‚^18O isotopes; analyzed via Isotope Ratio Mass Spectrometry.
24-Hour Urine Collection Kit	Collection of total urinary output over 24h for analysis of recovery biomarkers (e.g., nitrogen, potassium, sodium) [63] [50].	Includes large container (2-3L), portable cooler, written instructions, and protocol form to record start/stop times.
Automated 24HR Systems	Self-administered, standardized 24HR tools that reduce interviewer burden and cost, facilitating repeated measures [1] [6] [50].	ASA24 (US), myfood24 (UK/Germany), Intake24 (UK), Foodbook24 (Ireland).
Food Composition Databases	Convert reported food consumption into estimated nutrient intakes. Critical for both the test 24HR and the WFR method [63] [6] [50].	Country-specific databases (e.g., UK's CoFID, German BLS, USDA Food and Nutrient Database).
Portion Size Estimation Aids	Improve the accuracy of food amount reporting in 24HR and WFR.	Food photographs at multiple sizes [15] [6], household measures, digital food scales [63], food models [15] [1].
Standardized Protocols & SOPs	Ensure consistency in data collection, coding, and analysis, especially in multi-center studies or when using multiple coders [15] [50].	Interviewer training manuals [15], coder guidance flowcharts [50], biomarker collection instructions [63].

The choice between using a Weighed Food Record or objective Biomarkers as a reference method for validating 24-hour dietary recalls is not a matter of selecting a universally superior option, but rather of aligning the method with the specific research objectives, resources, and nutrients of interest. Weighed records offer a practical, comprehensive assessment across a wide range of nutrients, whereas biomarkers provide an unbiased, objective gold-standard for specific, recoverable nutrients and energy. The most robust validation studies strategically employ a combination of both methods to triangulate evidence and fully characterize the measurement error properties of the 24HR tool under investigation.

The integration of web-based tools in nutritional research, particularly for 24-hour dietary recalls (24HR), represents a significant advancement in data collection methodology. These tools offer the potential for scalable, cost-effective, and accurate dietary assessment, which is crucial for epidemiological studies, clinical trials, and public health monitoring. However, their adoption must be underpinned by rigorous validation to ensure data quality and reliability. This document provides a structured framework for assessing the relative validity and usability of web-based dietary assessment tools, with specific application to 24HR validation techniques. The protocols outlined herein are designed to equip researchers with standardized methodologies for evaluating these tools against established benchmarks and user-centric criteria, thereby ensuring the scientific integrity of collected dietary data.

Core Concepts: Validity and Usability

In the context of web-based dietary tools, validity refers to the accuracy with which an instrument measures the true dietary intake of an individual. It is not an inherent property of the tool but rather of the interpretation of the data it generates for a specific purpose and population [65]. Validity evidence is built from multiple sources, including content, response processes, and relationships to other variables [65].

Usability quantifies the user experience, measuring how effectively, efficiently, and satisfactorily a target user can interact with the web-based tool to complete their tasks [66] [67]. For a 24HR tool, high usability is critical for minimizing user error and maximizing long-term adherence.

Table 1: Key Dimensions of Validity and Usability

Dimension	Description	Common Metrics/Evidence
Relative Validity	The degree to which a tool's intake estimates correlate with those from a reference method or biomarker [68] [39].	Mean differences, correlation coefficients, cross-classification agreement.
Content Validity	Evidence that the tool adequately covers and represents the dietary components of interest [65].	Expert review, alignment with national dietary databases.
Usability: Effectiveness	The ability of users to complete tasks accurately and completely [67].	Task completion rate, error rate.
Usability: Efficiency	The resources expended by users to complete tasks [66].	Time on task, time per screen.
Usability: Satisfaction	Users' subjective perceptions and comfort while using the tool [67].	System Usability Scale (SUS), Net Promoter Score (NPS).

Quantitative Metrics for Assessment

A comprehensive assessment requires the collection of both quantitative and qualitative data. The following tables summarize key metrics for validity and usability.

Table 2: Core Metrics for Relative Validity Assessment

Metric	Definition	Application in Dietary Recall	Interpretation
Mean Difference (Bias)	The average difference between the test tool and the reference method.	(Energy from Web-24HR) - (Energy from Weighed Food Record).	A value close to 0 indicates minimal systematic bias.
Correlation Coefficient (Pearson/Spearman)	Measures the strength and direction of a linear relationship between two methods.	Correlation between fruit/vegetable intake from a tool and serum carotenoid levels [68].	Values closer to +1 or -1 indicate stronger agreement.
Percentage Match	The proportion of food items reported in the test tool that were actually consumed (vs. a reference) [40] [69].	Number of foods correctly recalled / Total number of foods consumed.	A higher percentage indicates better recall accuracy. Example: 71.4% match rate in older adults [69].
Intrusion Rate	The proportion of food items reported that were not consumed [40].	Number of foods incorrectly recalled / Total number of foods reported.	A lower rate indicates better accuracy and less false reporting.
Portion Size Ratio	The ratio of reported portion size to true portion size [40] [69].	Reported weight / Weighed weight.	A ratio of 1 indicates perfect accuracy. Example: A ratio of 1.34 indicates 34% overestimation [69].

Table 3: Core Metrics for Usability Assessment

Metric	Category	Calculation Formula	Interpretation
Completion Rate	Effectiveness	(Number of Completed Tasks / Total Assigned Tasks) x 100 [67]	A higher rate indicates the tool's interface is intuitive and navigable.
Misclick Rate	Effectiveness	(Number of Clicks on Incorrect Elements / Total Clicks) x 100	Highlights areas of UI confusion or poor information architecture.
Time on Task	Efficiency	Average time (seconds/minutes) taken by users to complete a specific task [66].	Compared against a benchmark; lower times can indicate greater efficiency.
System Usability Scale (SUS)	Satisfaction	Score from a 10-item questionnaire with a 5-point scale [70].	Scores are out of 100; above 68 are considered above average.
Net Promporter Score (NPS)	Satisfaction	% of Promoters - % of Detractors [67].	Measures user loyalty and likelihood to recommend the tool.

Experimental Protocols

Protocol 1: Assessing Relative Validity Against a Reference Method

This protocol validates a web-based 24HR tool by comparing it to a controlled feeding study, considered a high-quality reference.

Aim: To determine the relative validity of a web-based 24HR tool in estimating energy, nutrient, and food intake against weighed food records in a controlled setting. Materials:

• Weighing scales (precision ±1g)
• Web-based 24HR tool (e.g., ASA24 [5])
• Standardized food protocols
• Data management system (e.g., REDCap)

Procedure:

• Participant Recruitment: Recruit a sample representative of the target population (e.g., n=100+, considering age, sex, BMI). Obtain ethical approval and informed consent [40].
• Controlled Feeding: Provide participants with all meals for one or more days in a lab setting or as pre-weighed take-home packs. Discreetly weigh all food items before and after consumption to establish "true" intake [40] [69].
• Dietary Recall: Within 24 hours of the feeding period, administer the web-based 24HR tool. Participants should complete the recall independently.
• Data Processing: Extract nutrient and food intake data from the web-based tool. Process weighed food record data using the same nutrient database to ensure comparability.
• Statistical Analysis:
  • Calculate mean differences (test tool - reference) for energy and key nutrients. Use paired t-tests to assess significance.
  • Compute correlation coefficients (Pearson for normally distributed data, Spearman otherwise) to assess the strength of the relationship.
  • Assess agreement using Bland-Altman plots.
  • Calculate food-level accuracy metrics: Percentage Match, Intrusion Rate, and Portion Size Ratio [40] [69].

Validity Assessment Workflow

Protocol 2: Quantitative Usability Testing

This protocol employs an unmoderated, remote testing approach to collect quantitative usability data from target users.

Aim: To identify usability barriers and quantify the user experience of a web-based 24HR tool. Materials:

• Usability testing platform (e.g., Maze, UserTesting [71])
• Prototype or live version of the web-based 24HR tool
• Pre-defined test scenarios and tasks (e.g., "Report everything you ate yesterday for dinner")
• Post-test satisfaction questionnaire (e.g., SUS)

Procedure:

• Participant Recruitment: Recruit 20+ participants from the target population. For usability metrics, a sample size larger than 5 is required to achieve statistical confidence [66].
• Task Design: Create 3-5 core, realistic tasks that cover the primary functions of the 24HR tool (e.g., adding a food, specifying a portion size, modifying an entry).
• Unmoderated Testing: Deploy the test using a usability platform. Participants complete the tasks remotely without a moderator present.
• Data Collection: The platform automatically collects metrics including:
  • Completion Rate: For each task.
  • Time on Task: Average time to complete each task.
  • Misclick Rate: Clicks on non-interactive or incorrect elements.
  • Path Analysis: Deviation from an optimal path to completion.
• Post-Test Survey: Administer the System Usability Scale (SUS) to measure subjective satisfaction.
• Data Analysis:
  • Calculate descriptive statistics (mean, SD) for all metrics.
  • Benchmark completion rates against a target (e.g., 90%).
  • Analyze time on task to identify complex or inefficient workflows.
  • Use misclick maps to pinpoint specific UI elements causing confusion [67].

The Scientist's Toolkit: Research Reagent Solutions

Table 4: Essential Tools and Resources for Validation Research

Tool / Resource	Function in Research	Example Use Case	Key Features
ASA24 (Automated Self-Administered 24hr Recall)	A free, web-based system for collecting automatically coded 24-hour dietary recalls [5].	Serves as the "test tool" in a relative validity study against interviewer-administered recalls.	Based on USDA's AMPM; supports multiple recalls; available in multiple languages.
USDA AMPM (Automated Multiple-Pass Method)	A validated, 5-step interview methodology designed to enhance recall accuracy [39].	Used as the "gold standard" interviewer-administered protocol to validate a new web-based tool.	Structured passes (quick list, detail cycle, final review) to probe memory.
Doubly Labeled Water (DLW)	The gold standard biomarker for measuring total energy expenditure in free-living individuals [39].	Used to validate the accuracy of energy intake reporting in a dietary assessment tool.	High accuracy; used to identify under-/over-reporters.
Maze / Lookback	Usability testing platforms that facilitate remote, unmoderated (Maze) or moderated (Lookback) user testing [71].	Used to run Protocol 2, collecting completion rates, misclick data, and screen recordings.	Prototype integration, participant panels, automated reporting.
System Usability Scale (SUS)	A standardized 10-item questionnaire for measuring subjective usability [70].	Administered after usability testing to gather quantitative satisfaction data.	Quick to administer; provides a single, reliable score from 0-100.
NDSR / GloboDiet	Nutrient database and dietary analysis systems for processing food intake data.	Used to analyze and derive nutrients from both the test and reference dietary assessment methods.	Comprehensive food composition databases; standardized calculation protocols.

Within nutritional epidemiology, the validation of dietary assessment methods is a critical step to ensure that the data collected accurately reflects true intake. This is particularly vital for 24-hour dietary recalls (24-HDRs), a widely used method in both research and clinical settings. The process involves comparing the tool or method under investigation (the test method) against a reference method, which can be another, more detailed dietary assessment tool or objective biomarkers. This application note details the core statistical approaches—correlation coefficients, Bland-Altman plots, and cross-classification—employed to conduct these validations, providing researchers with a standardized framework for evaluating 24-HDR methodologies within broader thesis research on validation techniques.

Core Statistical Methodologies

The validation of dietary assessment methods relies on a suite of statistical tools to evaluate different aspects of agreement between the test and reference methods. The following table summarizes the purpose and application of the three primary approaches discussed in this protocol.

Table 1: Core Statistical Approaches for Dietary Assessment Validation

Statistical Approach	Primary Purpose	Key Interpretation Metrics
Correlation Coefficients	To measure the strength and direction of the linear relationship between two methods [72] [27].	Spearman's coefficient (r); Values closer to 1.0 indicate stronger agreement [27] [23].
Bland-Altman Plots	To visually assess the agreement between two methods by plotting differences against means, identifying systematic bias and proportional error [72] [73].	Mean difference (bias) and Limits of Agreement (LoA = mean difference ± 1.96 SD) [73].
Cross-Classification	To determine how well both methods classify individuals into the same intake categories (e.g., quartiles or tertiles) [27] [74].	Proportion classified into same/adjacent category; <5% grossly misclassified is a common benchmark [74].

Correlation Coefficients

Purpose and Rationale: Correlation coefficients quantify the strength and direction of the linear relationship between measurements obtained from the test method (e.g., a new web-based 24-HDR) and the reference method. A high correlation indicates that as intake values from one method increase, so do the values from the other method. Spearman's rank correlation coefficient is often preferred in dietary validation studies as it does not assume a normal distribution of data and is less sensitive to outliers [72] [27].

Protocol Application: In a recent validation of a web-based 24-HDR in Canadian adolescents, Spearman correlations were calculated for energy and 25 nutrients. The correlations were significant for most nutrients, ranging from 0.24 for thiamin to 0.52 for carbohydrates, demonstrating a moderate positive relationship between the web-based and interviewer-administered recalls [27]. Similarly, a validation study of a Food Frequency Questionnaire (FFQ) in China reported Spearman correlations for food groups and nutrients against a 3-day 24-HDR, with values ranging from 0.40 to 0.72, indicating moderate-to-good validity [74].

Bland-Altman Analysis

Purpose and Rationale: While correlation indicates a relationship, it does not measure agreement. Bland-Altman analysis is used to assess the degree to which two methods agree by plotting the difference between the two measurements against their mean for each subject [73]. This visual and analytical method helps identify any systematic bias (e.g., one method consistently reporting higher or lower values) and checks if the variability is consistent across the range of measurements.

Protocol Application: In the validation of a triple-pass 24-HDR against weighed food records in Ugandan children, Bland-Altman analysis showed the recall only marginally underestimated energy intake with a mean difference of 149 kJ (2.8%). The limits of agreement (LOA) were -1618 to 1321 kJ, indicating the range within which 95% of the differences between the two methods fell [73]. Another study using Bland-Altman plots for a web-based tool in adults with type 1 diabetes found no clear patterns of bias for energy and macronutrients, though the LOA were relatively wide, a common finding in dietary assessment [23].

Cross-Classification Analysis

Purpose and Rationale: This analysis evaluates the ability of the test method to correctly categorize individuals into groups of dietary intake (e.g., quartiles or tertiles) compared to the reference method. Good agreement is demonstrated when a high percentage of participants are classified into the same or adjacent category. This is crucial in nutritional epidemiology, where researchers are often interested in classifying subjects by level of intake rather than exact values.

Protocol Application: In the validation of the FFQ in China, researchers used tertile classification. They found that over 78% of participants were classified into the same or adjacent tertile for most food groups and nutrients, with less than 15% being grossly misclassified (e.g., lowest tertile by one method and highest by the other), supporting the questionnaire's validity [74]. In the Canadian adolescent study, cross-classification showed that 36.6% of participants were classified into the same quartile and 39.6% into the adjacent quartile using both the web-based and interview-based 24-HDR, with only 5.7% being misclassified [27].

Integrated Experimental Protocol for 24-HDR Validation

This section outlines a detailed protocol for validating a 24-hour dietary recall tool, integrating the statistical methodologies described above. The workflow for this validation process is summarized in the following diagram.

Study Design and Participant Recruitment

Objective: To recruit a representative sample of the target population for the 24-HDR tool. Procedure:

• Define Eligibility Criteria: Specify inclusion and exclusion criteria. For example, in a validation study for adolescents, participants may be recruited from schools and must be within a specific age range (e.g., 12-17 years) [27]. For adult populations, criteria often include stable body weight and not following a medically prescribed diet [18].
• Sample Size Calculation: Perform an a priori sample size calculation. A common recommendation is to recruit at least 100 participants to detect meaningful correlation coefficients (e.g., r ≥ 0.30) with 80% power and a 5% alpha error [18] [74]. Account for an expected dropout rate of 10-15%.
• Ethical Approval: Obtain approval from the relevant Institutional Review Board or Ethics Committee. Acquire written informed consent from all participants [73] [27].

Data Collection

Objective: To collect dietary intake data using both the test and reference methods. Procedure:

• Administer the Test Method: Have participants complete the 24-HDR tool under validation. If it is a web-based or app-based tool (e.g., R24W, Nutrition Data), provide clear instructions and training [27] [23]. Data should be collected over multiple non-consecutive days, including both weekdays and a weekend day, to account for day-to-day variation and estimate habitual intake. The number of recalls can vary (e.g., three recalls over a month [27]).
• Administer the Reference Method: The choice of reference method is critical.
  • Compared to Another Dietary Method: An interviewer-administered 24-HDR using the USDA Automated Multiple-Pass Method (AMPM) is a robust reference [27]. These should be conducted on different days than the test method recalls, ideally by trained dietitians blinded to the results of the test method.
  • Compared to Objective Biomarkers: For a more objective validation, compare estimated intake against nutritional biomarkers. For example:
    • Energy: Doubly labeled water for total energy expenditure [18].
    • Protein: Urinary nitrogen for protein intake [18].
    • Micronutrients/Food Groups: Serum carotenoids for fruit and vegetable intake, or erythrocyte membrane fatty acids for fatty acid consumption [18] [75].

Data Processing and Statistical Analysis

Objective: To process the collected data and perform the statistical analyses to evaluate validity. Procedure:

• Data Cleaning and Nutrient Calculation: Enter and clean all dietary data. Convert food consumption into nutrient intakes using appropriate food composition databases (e.g., the Canadian Nutrient File [27] or the National Food Database of Sweden [23]).
• Perform Statistical Analyses: Use statistical software (e.g., STATA, R) to conduct the following analyses on energy and key nutrients, as demonstrated in the studies reviewed [72] [73] [27]:
  • Correlation: Calculate Spearman's rank correlation coefficients between the test and reference method.
  • Bland-Altman Analysis: Generate Bland-Altman plots and calculate the mean difference (bias) and limits of agreement.
  • Cross-Classification: Classify participants into quartiles or tertiles based on intake from each method. Calculate the percentage classified into the same, adjacent, and opposite categories.

Table 2: Exemplary Data from Validation Studies Using Multiple Statistical Approaches

Study Context (Tool vs. Reference)	Nutrient/Food	Correlation Coefficient (Spearman's r)	Bland-Altman Mean Bias (%, if reported)	Cross-Classification (% same/adjacent)
Web-based 24-HDR vs. Interview [27]	Energy	0.31	+4.7% (p<0.05)	76.2%
	Carbohydrates	0.52	Not specified	Not specified
FFQ vs. 3-day 24-HDR [74]	Various Nutrients	0.40 - 0.70	Bland-Altman showed acceptable agreement	78.8% - 95.1%
Triple-pass 24-HDR vs. Weighed Record [73]	Energy	Not specified	-2.8%	79% (same quartile)
	Protein	Not specified	-9.4%	89% (same quartile)
Web-based Tool vs. 24-h Recalls (T1 Diabetes) [23]	Energy	0.79	No significant bias	Not specified
	Carbohydrates (% E)	0.94	No significant bias	Not specified

The Scientist's Toolkit: Research Reagent Solutions

The following table lists key materials, tools, and software essential for implementing the validation protocol described above.

Table 3: Essential Reagents and Tools for Dietary Validation Studies

Category	Item	Function/Application in Protocol
Dietary Assessment Software	R24W, ASA24, Nutrition Data, PCN Pro	Web-based or software platforms used as the test method or for processing dietary data from recalls and records [27] [23] [76].
Food Composition Database	Canadian Nutrient File (CNF), USDA FoodData Central, National Food Database of Sweden	Standardized databases used to convert reported food consumption into estimated nutrient intakes [27] [23].
Biomarker Assays	Doubly Labeled Water, Urinary Nitrogen, Serum Carotenoids, Erythrocyte Membrane Fatty Acids	Objective biological measurements used as reference methods to validate energy, protein, and specific nutrient intakes without self-reporting bias [18] [75].
Portion Size Estimation Aids	Volumetric Aids, Food Photographs, Household Measures, Fiducial Markers	Critical tools to improve the accuracy of portion size estimation during dietary recalls. Used in both test and reference methods [73] [27] [76].
Statistical Analysis Packages	STATA, R, SPSS	Software used to perform Spearman correlations, generate Bland-Altman plots, and conduct cross-classification analyses [72] [74].

The rigorous validation of 24-hour dietary recall methods is fundamental to ensuring data quality in nutritional research, clinical practice, and public health monitoring. The integrated use of correlation coefficients, Bland-Altman plots, and cross-classification analysis provides a comprehensive picture of a method's performance, assessing the strength of relationship, the extent of agreement, and its utility for categorizing subjects. By adhering to the detailed experimental protocol and utilizing the essential tools outlined in this application note, researchers can robustly evaluate new and existing 24-HDR tools, thereby strengthening the evidence base for the critical role of diet in health and disease.

The 24-hour dietary recall (24HR) is a cornerstone method for assessing individual food and nutrient intake in nutritional epidemiology, public health monitoring, and clinical research. Its validity, however, is not absolute and can vary significantly across different populations, technological platforms, and cultural contexts. Validation studies are therefore critical to understanding the measurement properties of this tool in specific use cases. This article synthesizes lessons from recent validation research conducted across adolescent, adult, and international cohorts, providing a comparative analysis of methodologies and outcomes. We present structured protocols and data to guide researchers in designing and interpreting 24HR validation studies, framed within the broader context of dietary assessment methodology.

Case Study Summaries and Comparative Data

The following table summarizes key characteristics and validation outcomes from selected recent studies across diverse populations.

Table 1: Summary of 24-Hour Dietary Recall Validation Studies Across Various Cohorts

Population & Location	Tool / Method Tested	Reference Method	Sample Size	Key Validity Outcomes
Adolescents, Québec, Canada [27]	Web-based self-administered 24HR (R24W)	Interviewer-administered 24HR	272 adolescents (12-17 years)	Energy intake 8.8% higher vs. reference; Significant correlations for most nutrients (0.24-0.52); 36.6% classified in same quartile.
Adolescent Females, Vietnam [77]	Mobile AI-Assisted App (FRANI)	Weighed Food Records	36 females (12-18 years)	Energy, protein, fat equivalent at 10% bound; CCCs for energy and nutrients 0.60-0.81.
Adults with T1D, Sweden [23]	Web-based Program (Nutrition Data)	Unannounced 24HR	42 adults (Median 46.5 years)	No significant mean differences in energy/macronutrients; Strong correlations (Energy: r=0.79, Carbs: r=0.94).
Multi-National Adults, Ireland [6]	Expanded Web-based Tool (Foodbook24)	Interviewer-led 24HR	Brazilian, Irish, Polish adults	Strong correlations (r=0.70-0.99) for 15/26 nutrients and 8/18 food groups.
Healthy Adults, Denmark [9]	Web-based Tool (myfood24)	Biomarkers & 7-day Weighed Food Records	71 adults (53.2 ± 9.1 years)	87% acceptable reporters; Strong correlation between folate intake and serum folate (ρ=0.62).
Women & Children, Niger [78]	In-person 24HR Survey	Repeat 24HR for usual intake modeling	3367 total (Children, Adolescents, Women)	Documented energy intake & critical micronutrient deficiencies (e.g., Calcium, Vitamins A, B12).

Detailed Experimental Protocols

Protocol 1: Validating a Web-Based Recall Against an Interviewer-Administered Standard in Adolescents

This protocol is derived from the study validating the R24W tool among French-Canadian adolescents [27].

• Objective: To assess the relative validity of a web-based, self-administered 24HR (R24W) for estimating energy and nutrient intakes against a traditional interviewer-administered 24HR.
• Population: Adolescents aged 12-17 years.
• Design:
  • Test Method: Participants complete the R24W on up to three occasions within one month. The tool uses a multiple-pass method, a pre-defined food list, and portion-size images.
  • Reference Method: A single interviewer-administered 24HR conducted by a trained dietitian using the USDA Automated Multiple-Pass Method (AMPM) in a controlled setting.
  • Administration Order: To avoid bias, the order of administration (web-based vs. interview) is randomized.
• Key Procedures:
  • Training: Participants complete a mandatory tutorial for the web-based tool before first use.
  • Supervision: Some recalls are completed under supervision (e.g., in a school lab) to ensure protocol adherence and provide technical support.
  • Data Entry: Dietitians enter data from the interviewer-led recalls into analysis software (e.g., Nutrific) linked to a national nutrient database.
• Statistical Analysis:
  • Paired t-tests to compare mean intakes of energy and nutrients.
  • Pearson correlation coefficients to assess the association between methods.
  • Cross-classification analysis to determine the proportion classified into same or adjacent quartiles.
  • Bland-Altman plots to visualize agreement and identify proportional bias.

Protocol 2: Validating a Mobile AI Tool Against Weighed Food Records in an LMIC Context

This protocol outlines the method used to validate the FRANI app in Vietnam [77].

• Objective: To validate a mobile AI application for dietary assessment against the gold standard of Weighed Food Records (WR) in adolescent females in a low- and middle-income country (LMIC) setting.
• Population: Adolescent females, aged 12-18 years.
• Design:
  • Test Method: FRANI mobile app. Participants are trained to take pictures of all foods and beverages using a "pop-socket" as a reference object for portion size estimation. The AI identifies food and estimates portion size, which the user confirms or corrects.
  • Reference Method: Weighed Food Records (WR). Enumerators shadow participants throughout the day, weighing all foods and beverages before and after consumption with digital scales.
  • Study Duration: 3 non-consecutive days (including weekdays and a weekend day).
• Key Procedures:
  • Enumerator Training: Intensive 7-day training for data collectors and supervisors.
  • Simultaneous Data Collection: WR and FRANI app data are collected on the same days.
  • Independent 24HR: A multi-pass 24HR is conducted the day after the WR/FRANI day by a different enumerator to compare all three methods.
• Statistical Analysis:
  • Mixed-effects models to test for equivalence at 10%, 15%, and 20% bounds.
  • Concordance Correlation Coefficients (CCC) to measure agreement.
  • Error analysis to identify food omissions (memory error) and portion-size estimation bias.

Protocol 3: Usual Intake Assessment in a Large-Scale Nutritional Survey

This protocol describes the methodology used to collect and analyze dietary data for public health monitoring in Niger [78].

• Objective: To describe the usual nutrient intakes and their adequacy in vulnerable population groups in a resource-constrained setting.
• Population: Children (2-5 years), adolescent girls (10-18 years), and women (19-49 years) from five regions of Niger.
• Design:
  • Primary Method: A single in-person 24HR survey.
  • Repeat Recall: A second 24HR is conducted on a random 20% sub-sample within 3-10 days of the first recall to account for intra-individual variation.
• Key Procedures:
  • Pre-Survey Preparation:
    • Food Listing: Inventory of foods and recipes commonly consumed in the target regions.
    • Recipe Standardization: Standardizing hundreds of local recipes with defined ingredients and cooking methods.
    • Pilot Survey: Conducted to refine tools and procedures.
  • Data Collection: The 24HR is conducted in households, using a chronological approach for the previous day's intake.
  • Data Processing: Use of the National Cancer Institute (NCI) method to model the distribution of usual daily intakes from the within- and between-person variance data.
• Statistical Analysis:
  • Calculation of median energy and nutrient intakes.
  • Comparison of nutrient intakes to Estimated Average Requirements (EARs) to determine prevalence of inadequacy.

Visualizing a 24HR Validation Study Workflow

The following diagram illustrates the core workflow common to many 24-hour dietary recall validation studies, integrating elements from the cited protocols.

The Scientist's Toolkit: Key Reagents & Materials for 24HR Validation

Table 2: Essential Materials and Tools for 24-Hour Dietary Recall Validation Studies

Item	Function / Description	Examples from Literature
Nutrient Composition Database	Converts reported food consumption into nutrient intake data. Critical for consistency between test and reference methods.	Canadian Nutrient File (CNF 2015) [27], UK CoFID [6], Swedish Food Database [23].
Standardized Food List & Recipes	A comprehensive list of region-specific foods and standardized recipes to ensure all consumed items can be accurately coded.	4,288 standardized recipes in Niger [78]; 113-item FFQ list in PERSIAN study [15].
Portion Size Estimation Aids	Visual tools to help participants estimate the quantity of food consumed, reducing a major source of measurement error.	Food photographs [27] [6], food models [15], household measures (cups, spoons) [23], AI with reference object ("pop-socket") [77].
Dietary Assessment Software/Platform	The technological interface for data entry, whether a web-based tool, mobile application, or data management system.	R24W [27], Foodbook24 [6], Nutrition Data [23], myfood24 [9], FRANI App [77].
Reference Method Instruments	The tools required for the chosen reference standard (e.g., interview guides, scales for weighing food).	Digital kitchen scales (e.g., Tanita) for Weighed Records [77], structured interview scripts for 24HR [27] [78].
Biomarker Assay Kits	For objective validation, kits to analyze biomarkers in blood, urine, or other samples that reflect nutrient intake.	Serum folate tests, 24-hour urinary nitrogen/potassium assays [9].

The validation of 24-hour dietary recalls is an essential, multi-faceted process that must be tailored to the specific population, tool, and research objective. Key lessons from recent studies underscore that while web-based and AI-assisted tools show great promise in improving scalability and user engagement, their performance varies. Successful validation requires rigorous study design, including appropriate reference methods, careful attention to portion size estimation, and the use of statistical techniques that go beyond simple mean comparisons to assess agreement and misclassification. The protocols and data summarized here provide a framework for researchers to conduct robust validations, ensuring that dietary data collected in diverse cohorts are fit for purpose, whether for clinical research, public health monitoring, or nutritional epidemiology.

Conclusion

The validation of 24-hour dietary recalls is not a one-size-fits-all process but a multifaceted endeavor critical for generating reliable data in nutritional science and clinical research. A robust validation strategy integrates a deep understanding of foundational protocols, thoughtful application tailored to the study population, proactive mitigation of measurement errors, and rigorous comparison against appropriate reference standards. The future of dietary assessment is being shaped by technological advancements, particularly web-based and automated self-administered tools like ASA24 and R24W, which show promising validity and acceptability. For biomedical and clinical research, this evolution promises enhanced scalability, reduced cost, and improved precision in capturing dietary exposures. Future efforts must focus on the continued refinement of these tools, the development of novel biomarkers for intake, and the standardization of validation protocols across diverse global populations to strengthen the evidence base linking diet to health and disease outcomes.

References

• 1. 24-hour Dietary Recall (24HR) At a Glance [https://dietassessmentprimer.cancer.gov/profiles/recall/]
• 2. Dietary Assessment Methods: What is a 24 hour recall? [https://www.myfood24.org/blog/dietary-assessment-methods-what-is-a-24hr-recall/]
• 3. 24-hour dietary recalls [https://www.measurement-toolkit.org/diet/subjective-methods/24-hour-dietary-recall]
• 4. Accuracy and Cost-effectiveness of Technology-Assisted ... [https://www.researchprotocols.org/2021/12/e32891/]
• 5. ASA24® Dietary Assessment Tool | EGRP/DCCPS/NCI/NIH [https://epi.grants.cancer.gov/asa24/]
• 6. Expansion and Assessment of a Web-Based 24-Hour Dietary ... [https://ojphi.jmir.org/2025/1/e52380/]
• 7. A Preliminary Study on a Form of the 24-h Recall That ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC9269212/]
• 8. 24-Hour Dietary Recall in the Expanded Food and Nutrition ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC10574531/]
• 9. Validity and reproducibility of a web-based dietary ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12187487/]
• 10. AMPM - USDA Automated Multiple-Pass Method [https://www.ars.usda.gov/northeast-area/beltsville-md-bhnrc/beltsville-human-nutrition-research-center/food-surveys-research-group/docs/ampm-usda-automated-multiple-pass-method/]
• 11. AMPM - Features [https://www.ars.usda.gov/northeast-area/beltsville-md-bhnrc/beltsville-human-nutrition-research-center/food-surveys-research-group/docs/ampm-features/]
• 12. Dietary misreporting: a comparative study of recalls vs ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12034172/]
• 13. Minimum days estimation for reliable dietary intake ... [https://www.nature.com/articles/s41430-025-01644-8]
• 14. Evaluation of the Reliability and Validity of a Food ... [https://pubmed.ncbi.nlm.nih.gov/40732895/]
• 15. Comparisons to 24-Hour Dietary Recalls and Selected ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12569992/]
• 16. NHANES Dietary Analyses Module [https://wwwn.cdc.gov/nchs/nhanes/tutorials/dietaryanalyses.aspx]
• 17. Data Analysis for the 2025 Guidelines Committee Report [https://www.dietaryguidelines.gov/2025-advisory-committee-report/data-analysis]
• 18. Protocol of the validation of the experience sampling-based ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12587619/]
• 19. Overview of Dietary Assessment Methods for Measuring ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC8338737/]
• 20. Measurement Errors in Dietary Assessment Using Self- ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC5683000/]
• 21. Data Processing & Data Analysis - Dietary Assessment Primer [https://dietassessmentprimer.cancer.gov/profiles/recall/processing.html]
• 22. Validation of the web-based self-administered 24-h dietary ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC8437850/]
• 23. Validation of a web-based dietary assessment program ... [https://www.frontiersin.org/journals/nutrition/articles/10.3389/fnut.2024.1395252/full]
• 24. Comparison of self-reported dietary intakes from the ... [https://www.sciencedirect.com/science/article/pii/S0002916522027459]
• 25. Comparison between Self-Completed and Interviewer ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC9781276/]
• 26. Comparison of an interviewer-administered with ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC10261653/]
• 27. Assessing the relative validity of a web-based self ... [https://nutritionj.biomedcentral.com/articles/10.1186/s12937-024-00954-0]
• 28. Validation of Interviewer-Assisted Recall for Measuring ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC5072126/]
• 29. Comparison of Interviewer-Administered and Automated ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC4462333/]
• 30. Validation of an automated self-administered 24-hour ... [https://pubmed.ncbi.nlm.nih.gov/35057638/]
• 31. Comparing Self-Administered Web-Based to Interviewer- ... [https://www.mdpi.com/2674-0311/4/2/17]
• 32. Emerging Technologies for Investigating Food Consumer ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12620339/]
• 33. a review and evaluation of online 24-h dietary recall tools [https://pmc.ncbi.nlm.nih.gov/articles/PMC11705012/]
• 34. Comparison of ASA24® Versions | EGRP/DCCPS/NCI/NIH [https://epi.grants.cancer.gov/asa24/comparison.html]
• 35. ASA24® Publications | EGRP/DCCPS/NCI/NIH [https://epi.grants.cancer.gov/asa24/resources/publications.html]
• 36. Automated Self-Administered 24-H Dietary Assessment ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC8194205/]
• 37. Evaluation and Comparison of Food Records, Recalls ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC3202154/]
• 38. Validity and reliability of an online self-report 24-h dietary recall method (Intake24): a doubly labelled water study and repeated-measures analysis - PMC [https://pmc.ncbi.nlm.nih.gov/articles/PMC6722486/]
• 39. Validation of the 2 × 24 h recall method and a 7-d web- ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC10511678/]
• 40. Accuracy of 24-hour dietary recalls among free-living older ... [https://www.sciencedirect.com/science/article/abs/pii/S0022316625006868]
• 41. The standardized computerized 24-h dietary recall method ... [https://pubmed.ncbi.nlm.nih.gov/21731006/]
• 42. Development and feasibility of a 24-hour recall software to ... [https://www.sciencedirect.com/science/article/abs/pii/S0889157525004752]
• 43. An Image-Voice Dietary Assessment System for Estimating ... [https://www.jmir.org/2025/1/e65939]
• 44. Validation of 24‐h dietary recall for estimating nutrient intakes ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC7503205/]
• 45. Reliability and validity of the 24-hour recall [https://www.sciencedirect.com/science/article/abs/pii/S000282232139410X]
• 46. Nutrient intakes from a 24-hour recall survey among ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12400888/]
• 47. Measurement Error in Dietary assessment [https://nutritionalassessment.org/errors/]
• 48. Impact on Nutrition Research and Adjustment for its Effects [https://prevention.cancer.gov/tools-resources/measurement-error-impact-nutrition-research-and-adjustment-its-effects]
• 49. Evaluation of measurement error in 24-hour dietary recall ... [https://www.sciencedirect.com/science/article/pii/S0002916522032257]
• 50. Validity of an online 24-h recall tool (myfood24) for dietary ... [https://bmcmedicine.biomedcentral.com/articles/10.1186/s12916-018-1113-8]
• 51. Validity of Interviewer-Administered 24-h Dietary Recalls in ... [https://www.mdpi.com/2072-6643/15/7/1757]
• 52. Dietary misreporting: a comparative study of recalls vs energy ... [https://bmcmedresmethodol.biomedcentral.com/articles/10.1186/s12874-025-02568-4]
• 53. Predictive equation derived from 6497 doubly labelled ... [https://www.nature.com/articles/s43016-024-01089-5]
• 54. Doubly labelled water [https://www.measurement-toolkit.org/physical-activity/objective-methods/doubly-labelled-water]
• 55. A doubly-labeled water validation study [https://www.medrxiv.org/content/10.1101/2025.04.02.25325112v1]
• 56. Evaluation of new predictive equations for energy ... [https://www.sciencedirect.com/science/article/pii/S0261561425001682]
• 57. A systematic review of dietary data collection methodologies ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC10992480/]
• 58. Protocol for a Mixed Methods Study [https://www.researchprotocols.org/2025/1/e70194]
• 59. Learn More about Season Effect - Dietary Assessment Primer [https://dietassessmentprimer.cancer.gov/learn/season.html]
• 60. Progressive 24-Hour Recall: Usability Study of Short ... [https://www.jmir.org/2020/2/e13266/]
• 61. Comparison of the 24 h Dietary Recall of Two Consecutive ... [https://www.mdpi.com/2072-6643/14/9/1960]
• 62. The Use of Pictorial Recall Aids Modifies Dietary ... [https://www.sciencedirect.com/science/article/pii/S2475299124023862]
• 63. Validation of the web-based self-administered 24-h dietary ... [https://link.springer.com/article/10.1007/s00394-021-02547-7]
• 64. Accuracy of energy and nutrient intake estimation versus ... [https://www.sciencedirect.com/science/article/pii/S0002916524004568]
• 65. A Primer on the Validity of Assessment Instruments - PMC [https://pmc.ncbi.nlm.nih.gov/articles/PMC3184912/]
• 66. Usability Metrics [https://www.nngroup.com/articles/usability-metrics/]
• 67. 12 Key Usability Metrics to Unlock User Insights [https://maze.co/collections/reporting-analysis/measure-usability-metrics/]
• 68. The relative validity of nutrition assessment methods for ... [https://www.sciencedirect.com/science/article/pii/S0271531724000800]
• 69. Accuracy of 24-hour dietary recalls among free-living older ... [https://pubmed.ncbi.nlm.nih.gov/41173375/]
• 70. Tools for Evaluating the Content, Efficacy, and Usability of ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC8686474/]
• 71. 18 Best Usability Testing Tools: Features & Pricing [https://maze.co/guides/usability-testing/tools/]
• 72. Dietary assessment methodologies in eating disorders [https://pmc.ncbi.nlm.nih.gov/articles/PMC12482415/]
• 73. of triple pass Validation - 24 in Ugandan children by... hour dietary recall [https://pubmed.ncbi.nlm.nih.gov/27795836/]
• 74. Evaluation of the Reliability and Validity of a Food ... [https://www.mdpi.com/2072-6643/17/14/2270]
• 75. Biochemical validation of a food frequency questionnaire ... [https://www.sciencedirect.com/science/article/pii/S0939475325000882]
• 76. Relative Validity and Reliability of the Remind App as an ... [https://www.mdpi.com/2072-6643/15/8/1824]
• 77. Relative validity of a mobile AI-technology–assisted dietary ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC9535545/]
• 78. Nutrient intakes from a 24-hour recall survey among ... [https://jogh.org/2025/jogh-15-04247/]