Dazult's Dietary Intake Evaluation Tool, DaDiet, is a web-based software tool that allows accurate estimate of exposure to nutrients and to substances added to foods, including contaminants, food additives and novel ingredients. Used by scientists and regulatory specialists in some of the world's leading food technology, manufacturing and consultancy companies, DaDiet provides easy to use tools to investigate the effect of changes to food compositions on a population's diet.

The Fundamentals

Dietary assessments involve combining a number of different data sources to estimate the intake of substances in food. Virtually all dietary assessment involve 3 essential data sources:

1. Consumption Survey
2. Food Classification System
3. Concentration Data

Consumption Surveys

DaDiet can be used with multi-day 24 hour recalls and multi-day diary based consumption surveys. These dietary data instruments involve the collection of data at the eating event level, typically with food code and the amount of food consumed recorded or estimated. The best example of a multi-day 24 hour recall is the US NHANES WWEIA surveys, while the UK's NDNS is now a 4 day food diary.

In addition to dietary intake records, both 24 hour recalls and food diaries contain data on the subject, including age, gender, bodyweight and statistical weighting. The statistical weighting is used as a measure of how representative that subject is of the total population, given the number of similar subjects within the survey population, while other subject data can be used to filter results and examine subsets of the population.

The Person-Day Concept

A person-day is the combination of a subject and the day on which the eat. For example in the 4-day UK survey a subject can be associated with up to 4 "person-days". For most modelling these person-days are treated as being correlated (as you would expect), but for some modelling efforts there are treated as independent (e.g. EFSA's Guidance on the Use of Probabilistic Methodology for Modelling Dietary Exposure to Pesticide Residues)

Food Classification Systems

Food Classification Systems allow foods to be grouped into logical selections. Nearly all surveys come with at least one default food classification system. While these are useful, you will often want to classify foods according to your own needs. This tends to be the most time consuming element of dietary intake and exposure assessments, and DaDiet provides a wizard to simplify the process as much as possible.

Substance Concentration Data

Concentration data are used to set the concentrations of substances (including nutrients, contaminants, food additives and novel ingredients) in foods or groups of foods.

Nutrients

Many consumption surveys provide default nutrient data for individual foods consumed in the survey. In DaDiet it is possible to model situations where the nutrient concentration:

• changes to a fixed (absolute) value, e.g. 5mg/100g;
• is changed by a percentage, relative to the default value e.g. +5% or -7%;
• is changed by an absolute amount, relative to the default value e.g. -5mg/100g or +10 mg/100g.

Other Substances

Baseline data for concentrations of other substances in foods tend not to be supplied with consumption surveys. In this case DaDiet allows concentrations to be set for substances in food in absolute terms, since there is no baseline for relative changes.

In DaDiet it is possible to run intake assessments using multiple substances at once, e.g. reducing sugar while increasing stevia.

Basic Population Intake Algorithm

When calculating dietary intakes, DaDiet calculates the dietary intake of various chemicals, ingredients and nutrients for every subject within the survey, as well as recording the food group responsible for each intake. The reporting wizards then uses each subject's statistical weighting to calculate various statistics, including means and percentiles, for the total population, or sub-populations such as age bands or consumers of certain foods. DaDiet also allows for the per subject data to be downloaded, for custom statistical modelling and analysis.

Advanced Modelling

Food Replacement Modelling

DaDiet's Food Replacement Model allow modification of the original survey in a number of ways:

• Replacing one food with another/new food, with possibly different quantities;
• Portion size changes;
• Adding in food consumption events, based on person-day consumption counts, which can be useful, for example, to reach "five-a-day" targets.

Portion size changes allow a portion size to be set to a particular fixed value for every subject, or can be set to a percentage change from its current value in the consumption survey.

Adding consumption events

The replacement model can be configured to ensure that each person-day receives a minimum number of consumption events of foods from individual foods or food groups listed in the replacement model. Multiple replacement models can be run together to ensure a more realistic scenario, where the quantity of other selected foods is reduced, for example, to approximately maintain energy intake levels.

Ingredient Modelling

DaDiet's Ingredients Model enables foods to be broken into their raw ingredients. Substance concentrations can be set in either the raw ingredient or the food as eaten. When the food is eaten in the survey, the concentration of the substance in the food is set by:

• the concentration provide for the food as eaten in concentration data file, where available;
• if the concentration is not set in then it is derived from the concentration of the subtance in its raw ingredients;
• otherwise it is set to the default value for its food classification group, or zero if this is not set.

Where ingredient data is used in an assessment, the report includes intake of the ingredient itself and the ingredients contributing most to the intake of the substance, in addition to the food classification breakdown.

Custom Modelling

In addition to the above models and algorithms, which are available to all DaDiet users, Dazult have developed custom models for users with particular, often project specific, needs.

A bit of chance...

DaDiet supports probablistic modelling for most of the input variables. The following distributions are available and can used to build custom input data models.

• bernoulli(p)
• binomial(n, p)
• neg_binomial(n, p)
• hypergeometric(M, n, N)
• poisson(mu)
• beta(a, b)
• chi_squared(df)
• gamma(shape)
• lognormal(s)
• normal()
• pareto(b)
• rayleigh()
• students_t(df)
• triangular(a,b,c)
• uniform(a,b)
• pick_one([a,b,c,...]): Pick one value from the comma seperated list.
• pick_min([a,b,c,...]): Pick the min value from the comma seperated list.
• pick_max([a,b,c,...]): Pick the max value from the comma seperated list.

In addition, certain models support alternative methods of producing probablistic results, including concentration data and replacement models.

Version History

DaDiet version numbering is derived from the year and month of the version release. Release updates are roughly every six months. When referencing DaDiet in your work, we recommend the following:

Dazult Limited. (yyyy). DaDiet - The Dietary Intake Evaluation Tool (Version yy.mm) [Software]. Available from https://dadiet.daanalysis.com.

where yyyy is the 4 digit year, yy the two digit year, and mm the two digit month.

Version 24.09

• Between July 2017 and June 2024, DaDiet was updated continously as requested features came in. These include:
• Special support for uniquely designed surveys from certain countries.
• HESS and CEDEM methodologies for estimating high intakes for EFSA surveys.
• Uncertainty model add-ons.
• Custom plots.
• General linear models.
• Custom Excel result download formats.
• Support for other exposure sources.
• In July 2024 we reverted to planned roughly six monthly updates.
• There is full backward compatitibilty - any files created in 17.04 will work in 24.09.
• New tools for endpoint comparison.
• New advisory checks on validity of statisitcal results.

Version 17.04

• New Default Excel format, including a single sheet with all population level data. Food group breakdowns are provided in additional sheets.
• EFSA Comphrensive Analysis methods.
• Three new calculation outputs to help analysis intake diversity: Meal Count, Distinct Food Codes per day, Distinct Food Code thoughout survey, with breakdowns by food group.
• IESTI pesticide screening tool [on request, requires initial setup to meet specific survey inputs].

Version 16.11

• Support for complex survey variables (strata and psu) for confidence intervals and standard errors.
• Pesticide Models: Support for variability and processing factors.
• Support for Market Share and Brand Loyalty Models.
• Addition of new Mexican, Brazilian and US datasets.
• Additional Consumption Filters.
• Quantile Analysis by Food Group.
• New Default Excel Format, includes consumer only statistics often used in EFSA styled reports.

Version 16.05

• A New File Management System, allowing drag and drop uploading of PDFs and other files.
• Support for Food Group Hierarchies.
• Easier to find files during assessment setup.
• New Audit Access Only Licence.

Version 15.10

• Complex Food Replacement Modelling via a simple command based language.

Version 15.05

• Additional calculation types (per kcal etc.).
• Additional replacement modelling functionality.
• Added food code intake breakdown.
• Added built-in intakes comparison tools.
• Added assessment auditing tools.
• Improved file creation wizards.
• Added personalisation settings.

Version 14.10

• Added ingredient level analysis.
• Added meal breakdown.
• Added advanced assessment filtering.

Version 14.05

• Added usual intakes.
• Added acute intakes based on EFSA's person-day.
• Added replacement modelling.

Version 13.10

• Added probablistic support.
• Added food classification wizard.
• Added concentration data wizard.

Version 13.05

Initial release:

• Deterministic intake assessments.
• Absolute and per-kg bodyweight results.
• Food Classification intake breakdown.
• NHANES public data pre-installed.
• Support for installing confidential data, ensuring compliance with data licences.
• Support for custom Excel reporting.