essential8 0.9.0

Major Changes

• score_le8() now scores children aged 4 months to 19 years with the pediatric LE8 definitions (Table 1 and Supplemental Appendix 2 of the 2022 advisory). Rows are routed by age, so adults and children can be scored in one call, and the new le8_population column records which definitions were used. Components AHA does not define at a child’s age (for example, blood pressure before age 1) are NA and are excluded from min_components.
• Pediatric MEPA uses the same input columns as adults, plus milk_source at 6-23 months: 6 items at 6-23 months, 10 items at 2-5 years (10 points per item), and all 16 items with adult bands from age 6. mepa_total reports items met on the age-appropriate screener.
• New pediatric inputs: age_months, milk_source, active_play_minutes, bmi_percentile, bmi_percent_of_p95, infant_weight_percentile, infant_weight_z_change, and the AAP 2017 blood-pressure references systolic_bp_p90, systolic_bp_p95, diastolic_bp_p90, and diastolic_bp_p95.
• New add_pediatric_references() derives the growth-chart and blood-pressure reference inputs from sex, age, height, and weight using the suggested packages cdcanthro and pedbp.
• New vignette("pediatric-scoring").
• score_le8() results now carry the class essential8_scores on top of the input’s class (data frames, tibbles, and grouped tibbles are otherwise unchanged), with new methods:
  • summary() reports the number of rows scored, the distribution of the composite and each component score, and the cardiovascular health categories.
  • tidy() reshapes the scores into long format, one row per input row and measure. tidy() is re-exported from generics.
  • plot() draws the mean scores, their distributions (type = "distribution"), or a histogram of the composite score by category (type = "composite") with ggplot2. type = "forest" draws the mean scores with confidence intervals, optionally compared across the groups in by, using ggforestplotR::ggforestplot(); other arguments to plot() are passed on to ggforestplot().
  • With dplyr, summarise() (or summarize()) without expressions returns the summary() statistics as a data frame, by group with .by. Summaries with expressions are handled by dplyr as before.
• Multiple imputation of missing inputs with mice:
  • score_le8(impute = TRUE) imputes missing inputs, scores every imputed data set, and returns them stacked with an .imp column. summary(), summarise(), tidy(), and plot() on the result pool across imputations with Rubin’s rules.
  • le8_imputation_control() sets the number of imputations, iterations, a seed, auxiliary predictors, and other mice::mice() arguments: score_le8(impute = le8_imputation_control(m = 50, seed = 1)).
  • The imputation imputes an input only for participants whose age uses it, limits each model’s predictors to inputs available for everyone it applies to, and never imputes a combination AHA does not score (for example, diagnosed diabetes without an HbA1c value, or current combustible smoking with current inhaled-NDS use). le8_imputation_setup() shows these settings, and score_le8_imputed() scores a mice run of your own.
• New vignette("case-study") works through a complete project: scoring 500 simulated children and adults, choosing min_components, imputing missing laboratory values, and summarizing and plotting the results.
• New as_le8_scores() restores the essential8_scores class to scored data after group_by(), as_tibble(), or reading saved results.

Breaking Changes

• Component score columns no longer have the le8_ prefix: they are now diet_score, physical_activity_score, nicotine_score, sleep_score, bmi_score, blood_lipids_score, blood_glucose_score, and blood_pressure_score. le8_n_components is now n_components. Columns that describe the LE8 composite keep the prefix: le8_composite_score, le8_complete, le8_category, and le8_population.
• Because output names are reserved, data that already contains one of the new names (for example, a sleep_score from another instrument) or one of the old le8_-prefixed names now raises an error. Rename those columns before scoring.

Minor Changes

• Errors now report score_le8() as the call rather than an internal helper.

essential8 0.3.0

Major Changes

• Added DASH and HEI-2015 support to score_le8(). Both methods accept raw food and nutrient inputs, existing diet scores, or percentiles (#9).
• Added packaged NHANES 2015-2018 references to convert diet scores to LE8 points, with options to supply a custom reference or scoring cutpoints.
• Raw scoring returns dash_total or hei2015_total. HEI-2015 also returns all 13 component scores.

Bug Fixes

• Diagnosed diabetes with a missing glucose_measure is no longer scored as though glucose_value were HbA1c; the glucose score is now missing.
• apply_lean_muscular_bmi_override = TRUE no longer awards 100 BMI points when bmi or bmi_profile is missing.
• Missing values in apply_lean_muscular_bmi_override and apply_prediabetes_metformin_penalty now blank a score only for records where the adjustment could apply.
• Entirely missing input columns stored as logical NA (such as blank CSV columns) are treated as missing instead of raising a type error.
• le8_physical_activity_score is now always double, including when activity inputs are entirely missing.
• score_le8() now returns an object of the same class as data, so tibbles and grouped tibbles are preserved.

Breaking Changes

• Input column names that differ from a case-sensitive LE8 input name only by case (for example BMI) now raise an error instead of being ignored.
• Supplying some but not all 16 MEPA items, or omitting any raw DASH or HEI-2015 column once raw scoring is selected, now raises an error naming the missing columns instead of silently returning missing diet scores.
• mepa_columns now requires diet_method = "mepa", and dash or hei2015 index columns are rejected under diet_method = "percentile".

Minor Changes

• Raw DASH scoring now returns the eight dash_<field>_score component scores, matching the raw HEI-2015 output.
• An entirely missing diet_value column no longer blocks scoring of a dash or hei2015 index column.
• A logical female column is accepted as the MEPA sex field.
• Removed unused internal rule-schema scaffolding.
• Expanded the diet vignette to explain MEPA, DASH, and HEI-2015, their role in LE8, and how their scores are calculated.
• Simplified the README and examples, using consistent tidy workflows.

essential8 0.2.0

Major Changes

• score_le8() can now work with data that have missing values (#7).
• Added min_components, an argument to control how many available components are required for le8_composite_score. Valid options range from 1-8.

Minor Changes

• Added le8_n_components and le8_complete outputs for transparent reporting of the available components for LE8 composite scores.
• Added consolidated, classed warnings when one or more LE8 components cannot be calculated for any observation, including a stronger warning when structural missingness makes min_components impossible to reach.

essential8 0.1.0

Major Changes

• Added score_le8() for complete adult records aged 20 years or older, using the eight component definitions in the AHA 2022 Presidential Advisory.
• Added raw 16-item MEPA derivation with exact case-insensitive screener-column matching, optional explicit column mappings, an auditable mepa_total, and population-percentile diet scoring.
• Added moderate-equivalent activity minutes, general or Asian-Pacific BMI profiles, and explicit glucose-measure selection.
• Changed score_le8() to take one scalar diet_method argument per call, defaulting to "mepa"; percentile calls require diet_value, and data using both methods must be scored in separate calls.
• Renamed the composite output from le8_score to le8_composite_score.
• Expanded MEPA sex input to accept trimmed, case-insensitive m/f and male/female, plus numeric or character 0/1 (0 is male and 1 is female). The default field is sex, female is recognized when sex is absent, and other field names can be supplied through mepa_columns.
• Added a reproducible Get Started vignette.