Dietary environmental impact metrics for R • reiod

Dietary environmental impact metrics from Intake’s Environmental Impacts of Foods framework, for R.

What it does

Family	Question	Source	Function
Absolute threshold	Is this diet below the per capita daily share of a planetary boundary?	Table 3	`eiod_absolute_threshold()`
Relative benchmark	Where does this diet sit in a fixed global reference distribution?	Tables 4a–4g	`eiod_relative_band()`
Composite — HESD	Healthy and environmentally sustainable?	Table 5a	`eiod_hesd()`
Composite — HDLEI	Healthy and low environmental impact?	Table 5b	`eiod_hdlei()`

Seven indicators: total, arable and pasture land use; greenhouse gases; eutrophication potential; water use; biodiversity loss. All 616 cut-offs are age- and sex-specific, and ship with the package.

Installation

# install.packages("pak")
pak::pak("nicholustintzaw/reiod")

No dependencies beyond base R. reiod imports only stats and utils, which ship with every R installation, so it installs with no dependency tree. The consequence is a plain vector API: pass df$biodiv_pdf, not biodiv_pdf.

Your columns can be called anything. "biodiv" is the indicator key reiod knows; what the column is named in your data is a separate matter, and the two never have to match.

Example

eiod_demo is a synthetic dataset shipped with the package: 600 respondent-days with age, sex, a GDQS score and the seven impact totals. eiod_demo_map says which of its columns holds which indicator.

library(reiod)

d <- eiod_demo
d$grp <- eiod_age_sex_group(d$age, d$sex)   # the eight age x sex groups

# check the inputs before computing anything
summary(eiod_check_data(d, impacts = eiod_demo_map,
                        age = "age", sex = "sex", gdqs = "gdqs"))
#> PASS - 0 error, 0 warning, 33 note

# Table 3 - % of diets below each absolute threshold
abs7 <- eiod_absolute_all(d, d$grp, eiod_demo_map)
round(100 * colMeans(abs7, na.rm = TRUE), 1)
#>   abs_landtotal  abs_landarable abs_landpasture         abs_ghg     abs_eutroph 
#>            70.2            80.8            68.7            22.2            53.5 
#>       abs_water      abs_biodiv 
#>            95.2            57.1

# Tables 4a-4g - relative band (1-10) for greenhouse gases
table(eiod_relative_band(d$ghg_kgco2e, d$grp, "ghg"))
#> 
#>   1   2   3   4   5   6   7   8   9  10 
#> 249  30  65  78  29   1   4  63  23  58

# Table 5a - healthy and environmentally sustainable diet
table(eiod_hesd(d$biodiv_pdf, d$gdqs, d$grp))
#> 15 observation(s) aged under 2 years set to NA: the GDQS is not validated below age 2, so the HESD metric does not apply to them.
#> 
#>         Healthy and env. sustainable     Not healthy but env. sustainable 
#>                                  206                                  125 
#>     Healthy but not env. sustainable Not healthy and not env. sustainable 
#>                                  124                                  110

The percentages are unweighted and the data are synthetic: they demonstrate the code, not a result. With survey data, estimate prevalences with a design-aware estimator (see below).

What it does not do

reiod does not contain the EIOF database. It implements the metric cut-offs defined in the Intake methods paper and applies them to impact values you have already derived. Matching food items to impact factors, applying edible-portion, processing and cooking-yield adjustments, and summing to the respondent-day is upstream work that this package starts after.

It does not calculate the Global Diet Quality Score. The composite metrics take a GDQS value as input.

It applies no survey design. reiod returns respondent-level indicator vectors. Every prevalence you report should come from a design-aware estimator — the survey package or equivalent — or both your point estimates and your standard errors will be wrong for a complex design.

It cannot see incomplete EIOF coverage. Not every food has an impact factor, and a food without one contributes nothing to the day total, so the total is understated. reiod receives one number per respondent-day; the aggregation has already discarded the evidence. Count the share of consumed weight carrying a factor when you aggregate, and report it with any prevalence.

Cautions before you report results

1. Eight eutrophication cut-offs differ from the published PDF, on Intake’s instruction. The paper prints the Table 3 eutrophication row in kilograms while Table 4f and the EIOF databases use grams; Intake confirmed in October 2026 that both tables are meant to be in grams and supplied corrected values. They are the printed figures × 1000 — same quantity, different unit — so no result changes, and the value_source column marks the eight rows. Supply eutrophication in grams and convert nothing.

2. A low relative band is not evidence of sustainability. For greenhouse gases even band 1 exceeds the planetary boundary; for water even band 10 is below it. The two metric families answer different questions and will not agree. Never use one as a proxy for the other.

3. HDLEI is nested inside HESD. The HDLEI cut-off is uniformly 1.52× stricter, so every HDLEI diet is also a HESD diet. They are not two independent dimensions and must not be presented as a 2×2.

4. The relative reference distribution is country-level. It is built from Food Balance Sheet data — national availability averages, not individual consumption. A band is a statement about position relative to national average availability patterns, not relative to other people or other days. Worth a sentence in any methods section.

5. The data checks are opt-in. Only one test runs on the default metric path: an impact vector must not be an integer, because is.numeric(1L) is TRUE and nothing else would catch it. Sign, zeros, unit magnitude, retained precision and cut-off proximity are checked only if you call eiod_check_data().

Documentation

Start with Get started for what is in the package and what the cut-offs are. Then the articles, in order:

	Article
1	Check your data first
2	Demographic groups
3	Table 3 — absolute thresholds
4	Tables 4a–4g — relative benchmarks
5	Tables 5a & 5b — composite metrics

Each metric article works through a single value, a column from a dataset, adding results to a data frame, all seven indicators at once, and only the indicators you have.

Reference notes: the precision pre-flight check, a broken dataset fault by fault, and eutrophication’s corrected cut-off.

Source

Intake (2026). Methods for Assessing Healthy, Environmentally Sustainable Diets Globally. Tables 3, 4a–4g, 5a and 5b.

The cut-offs ship as eiod_cutoffs, a 616-row long-format dataset built from the same source file as a companion Stata implementation, so the two cannot drift apart at the data layer.

{riycf} — infant and young child feeding indicators, same dependency-free design.

Contributing

Bug reports and suggestions are welcome at https://github.com/nicholustintzaw/reiod/issues. See the contributing guide before opening a pull request. Note that the cut-off values change only on Intake’s instruction.

Code of Conduct

Please note that this package is released with a Contributor Code of Conduct. By contributing to this project, you agree to abide by its terms.

reiod