Durango Software’s
Foods Analysis and Residue Evaluation(FARE) is a multi-objective research
program for calculating dietary intake,
nutrient intake, and dietary exposure to
agricultural chemicals and food additives
for the U.S. population based on food consumption
data from the 1994-96, 1998 Continuing Survey
of Food Intake for Individuals (CSFII).
FARE allows users to specify foods
of interest at the foods-as-eaten or ingredient
levels, nutrients (for nutrient analyses),
residue amounts (for exposure analyses),
sources of foods, location of consumption,
times of consumption, and populations of
interest (by their demographics), and generates
reports with full distributions of intake
amounts for the selected populations, both
for users and at the per-capita level. The
computational capabilities of FARE
reach far beyond any other existing software
package currently available anywhere in
the world. The following summary is only
a partial list of the wide range of analysis
and reporting capabilities provided by FARE:
(1) FARE can compute
dietary intake (whether consumption, nutrient,
or exposure amounts) based on the foods-as-eaten
level reported by individuals in the CSFII,
at the USDA NDB (5-digit) ingredient level,
or at the RAC (raw agricultural commodity)
level.
(2) FARE can apply
either the Novigen recipe set or EPA-FCID
recipe set “on the fly” to the
foods-as-eaten reported in the CSFII for
each individual in order to translate these
foods to their RAC (raw agricultural commodity)
equivalents. Alternative recipes sets or
partial recipe sets can also be used if
they are properly formatted. Durango Software
has the capability of generating alternative
recipes sets for use with FARE.
(3) The FARE DRS
file editor can be used to explore the entire
8-digit NFCS food code hierarchy (e.g.,
1xxxxxxx is milk and milk products, 12xxxxxx
is creams and cream substitutes, etc.),
to specify foods of interest at the NFCS
foods-as-eaten level, at the RAC level (either
based on the older Novigen/EPA commodities
or the newer FCID commodities), or to specify
RACS derived from specific NFCS codes (at
any level in the eight-digit NFCS hierarchy).
The resulting “DRS” files are
saved to the disk and used later as the
basis for specifying the foods of interest
in FARE analyses. (Some advanced analytical
features in FARE require the user to modify
the DRS file using a text editor.)
(4) Daily RAC consumption
and related dietary exposure analysis can
be computed for each individual in the CSFII
and the distribution of consumption and
exposure amounts across specific demographic
groups can be reported (similar to DEEM).
(5) Intake amounts can be
evaluated on a daily basis, hourly basis,
for different periods of time during the
day, e.g., 4 six-hour periods, for specific
meals, by “line-item” (e.g.,
each item of food consumption evaluated
separately, by length of the eating period,
or on a two-day average basis.
(6) Intake amounts for each
CSFII individual in a specific demographic
group can be saved to a CSV file (comma-delimited
ASCII file), along with specific demographic
data for that individual, for post processing
using Excel, SAS, or other tools.
(7) FARE can import
a DEEM or DEEM-FCID residue file,
including those requiring Monte Carlo analysis,
and perform an exposure analysis nearly
identical to DEEM, with more information
on hourly, part-day and 2-day average performance.
(8) FARE can also
import a DEEM or DEEM-FCID residue
file and perform an exposure analysis for
a restricted set of foods-as-eaten (e.g.,
for milk products only). This requires a
specially formatted DRS file as well as
a DEEM file.
(9) FARE has the
capability of dynamically selecting “recipes”
for oils, sugars, margarine, shortening,
and beer using Monte Carlo analysis, rather
than assuming some proportional allocation
of ingredients from all recipes for these
items, as has been done in the past. (This
feature works only with the Novigen Sciences
recipes since only this recipe set identifies
“auxiliary recipes” as the basis
for resolving these items. The EPA-FCID
recipes have already broken down these items
into their final ingredients and thus have
no auxiliary recipes.)
(10) Because the time of consumption
for most food items is included in the CSFII
database, FARE can be modified
for special studies, such as reversibility
analyses, which require a time element.
While only the FARE programmer
can make such modifications to the program,
the FARE platform is quite amenable
to such modifications.
(11) FARE can perform
Monte Carlo analyses (MCA) for selecting
residue amounts from a distribution or a
list of residue amounts assigned by the
user to specific foods-as-eaten or ingredients.
When MCA is used, multiple iterations of
exposure can be computed for each individual
in the population of interest (up to 5000
iterations per individual).
(12) FARE can perform
consumption and exposure analyses with drinking
water only, or food-preparation water only,
or both, without using a DRS file or DEEM
residue file. If food-preparation water
analyses are to be conducted for specific
foods-as-eaten only, a DRS file is required.
Drinking water consumption can also be included
in an analysis of specific foods-as-eaten
only or of specific RACs only. (Drinking
water cannot be included in an analysis
of RACs from specific foods-as-eaten as,
by definition, drinking water cannot be
an ingredient in a food-as-eaten. FARE
program logic will warn the user if this
is attempted and lock out the analysis.)
(13) The FARE report
includes user means and standard deviations,
per capita means and standard deviation,
and user and per capita distribution reporting
from 10 to 100 percent (including 95, 97.5,
99, 99.5, and 99.9%) for each time period
included in the analysis. (There may not
be enough observations in each time period
to calculate a full distribution, for example,
in the hours between 11 pm and 6 am.). The
FARE report also provides the number
of observations in each “bin”
(consumption or exposure interval) in which
the individual results are stored, so that
the basis for calculating the percentiles
can be verified and further refined (outside
of FARE) is required.
(14) FARE allows
the user to undertake nutrient analyses
using the demographic and food consumption
data from the CSFII, including means and
distributions for specific demographic groups.
Nutrient data is available for 52 different
nutrients, including (but not limited to)
fats, protein, carbohydrates, food energy,
vitamins, elements, and fatty and trans-fatty
acids. Nutrient analysis can be performed
at the NFCS code level for foods as eaten,
at the 5-digit NDB level, and at the 3-digit
RAC level (based on the older Novigen/EPA
RACs), but not at the FCID commodity level.
(15) FARE allows
the user to set up batch files for performing
multiple analyses based on different DRS
files so that a number of time-consuming
analyses can be run sequentially while the
user is away from his/her computer, with
each report saved to a different file named
by the user.
(16) A new “special
purpose” nutrient file (e.g., a file
containing new transfatty acid data indexed
for different types of foods) can be referenced
by a specially formatted DRS file that contains
specific foods-as-eaten and RACs along with
pointers to the indexed nutrient data in
the special purpose file for the purpose
of computing nutrient intake beyond the
capabilities of the CSFII nutrient database.
The FARE program
is only available by special licensing arrangements
from the Food
and Chemicals Practice of Exponent, Inc.
For more information on licensing requirements,
please contact Exponent at (202) 772-4900.
