Skip to contents

Calculate metabolic risk feature flags (pediatric-friendly thresholds)

Source: R/metabolic_risk_features.R
metabolic_risk_features.Rd

Compute four binary risk flags from routine clinical measures:

  • dyslipidemia

  • insulin_resistance

  • hyperglycemia (prediabetes-range glycemia)

  • hypertension (BP >=95th percentile via z > 1.64)

Usage

metabolic_risk_features(
  data,
  col_map = NULL,
  na_action = c("keep", "omit", "error", "ignore", "warn"),
  na_warn_prop = 0.2,
  verbose = TRUE
)

Arguments

data

A data.frame or tibble containing at least these numeric columns:

  • chol_total, chol_ldl, chol_hdl, triglycerides (mmol/L)

  • age_year (years)

  • z_HOMA (standardized HOMA-IR)

  • glucose (mmol/L)

  • HbA1c (mmol/mol; IFCC units)

  • bp_sys_z, bp_dia_z (BP z-scores)

col_map

Optional named list to map required keys to column names in data. Keys: c("chol_total","chol_ldl","chol_hdl","triglycerides","age_year", "z_HOMA","glucose","HbA1c","bp_sys_z","bp_dia_z"). Default NULL (use same names).

na_action

One of c("keep","omit","error","ignore","warn") controlling missing-data policy.

  • "keep": keep NA; outputs become NA where inputs are NA.

  • "omit": drop rows with NA in any required input.

  • "error": abort if any required input contains NA.

  • "ignore"/"warn": aliases of "keep"; "warn" also emits missingness diagnostics.

na_warn_prop

Numeric in \([0,1]\); per-variable threshold for high-missingness warnings. Default 0.2.

verbose

Logical; if TRUE, prints column mapping and computing messages.

Value

A tibble with four factor columns (levels c("0","1")):

  • dyslipidemia

  • insulin_resistance

  • hyperglycemia

  • hypertension

Details

By default, behavior matches prior implementation: required columns are validated, NA values are kept (propagate to outputs), no extreme-value checks or capping are applied, and a tibble with 0/1 factor flags is returned.

Units and criteria (no automatic unit conversion):

  • Lipids (mmol/L): total cholesterol > 5.2 OR LDL-C > 3.4 OR HDL-C < 1.0 OR triglycerides > 1.1 (age 0-9) OR > 1.5 (age 10-19) => dyslipidemia = 1. Note: no TG cutoff is applied for adults aged >= 20 years.

  • Insulin resistance: z_HOMA > 1.28 (~=90th percentile) => insulin_resistance = 1. z_HOMA is a within-sample or external z-score of HOMA-IR (see Matthews et al. 1985).

  • Hyperglycemia: fasting glucose in (5.6, 6.9) mmol/L OR HbA1c in (39, 47) mmol/mol => hyperglycemia = 1. Boundaries are EXCLUSIVE (open intervals); boundary values (exactly 5.6 or 6.9 mmol/L; exactly 39 or 47 mmol/mol) are not flagged. ADA criteria use inclusive lower bound (>= 5.6 mmol/L, >= 39 mmol/mol).

  • Hypertension: either BP z-score > 1.64 (~=95th percentile) for systolic or diastolic => hypertension = 1.

Note

These flags are heuristic screening rules derived from published clinical guidelines. They are not validated diagnostic criteria and should not replace clinical judgment. The dyslipidemia and hypertension thresholds are designed for pediatric populations (ages 0-19); for adults >= 20, only TC, LDL-C, and HDL-C criteria contribute to the dyslipidemia flag.

References

National Heart, Lung, and Blood Institute (2011). “Expert Panel on Integrated Guidelines for Cardiovascular Health and Risk Reduction in Children and Adolescents: Summary Report.” Pediatrics, 128(Suppl 5), S213–S256. doi:10.1542/peds.2009-2107C . American Diabetes Association Professional Practice Committee (2024). “2. Diagnosis and Classification of Diabetes: Standards of Care in Diabetes—2024.” Diabetes Care, 47(Suppl 1), S20–S42. doi:10.2337/dc24-S002 . Flynn JT, Kaelber DC, Baker-Smith CM, Blowey D, Carroll AE, Daniels SR, de Ferranti SD, Dionne JM, Falkner B, Flinn SK, Gidding SS, Goodwin C, Leu MG, Powers ME, Rea C, Samuels J, Simasek M, Tran VT, Urbina EM (2017). “Clinical Practice Guideline for Screening and Management of High Blood Pressure in Children and Adolescents.” Pediatrics, 140(3), e20171904. doi:10.1542/peds.2017-1904 . Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC (1985). “Homeostasis Model Assessment: Insulin Resistance and Beta-Cell Function from Fasting Plasma Glucose and Insulin Concentrations in Man.” Diabetologia, 28(7), 412–419. doi:10.1007/BF00280883 .

See also

liver_markers(), lipid_markers(), kidney_failure_risk(), inflammatory_markers()

Examples

df <- data.frame(
  chol_total = c(5.2, 6.4), chol_ldl = c(3.2, 4.1), chol_hdl = c(1.3, 1.0),
  triglycerides = c(1.8, 2.5), age_year = c(45, 60), z_HOMA = c(0.5, 1.2),
  glucose = c(5.5, 6.8), HbA1c = c(38, 46), bp_sys_z = c(0.2, 1.1),
  bp_dia_z = c(0.1, 0.9)
)
metabolic_risk_features(df)
#> metabolic_risk_features(): reading input 'df' — 2 rows × 10 variables
#> metabolic_risk_features(): col_map (10 columns — 10 inferred from data)
#>   chol_total        ->  'chol_total'    (inferred)
#>   chol_ldl          ->  'chol_ldl'    (inferred)
#>   chol_hdl          ->  'chol_hdl'    (inferred)
#>   triglycerides     ->  'triglycerides'    (inferred)
#>   age_year          ->  'age_year'    (inferred)
#>   z_HOMA            ->  'z_HOMA'    (inferred)
#>   glucose           ->  'glucose'    (inferred)
#>   HbA1c             ->  'HbA1c'    (inferred)
#>   bp_sys_z          ->  'bp_sys_z'    (inferred)
#>   bp_dia_z          ->  'bp_dia_z'    (inferred)
#> metabolic_risk_features(): computing markers:
#>   dyslipidemia       [chol_total, chol_ldl, chol_hdl, triglycerides, age_year]
#>   insulin_resistance [z_HOMA]
#>   hyperglycemia      [glucose, HbA1c]
#>   hypertension       [bp_sys_z, bp_dia_z]
#> metabolic_risk_features(): results: dyslipidemia 2/2, insulin_resistance 2/2, hyperglycemia 2/2, hypertension 2/2
#> # A tibble: 2 × 4
#>   dyslipidemia insulin_resistance hyperglycemia hypertension
#>   <fct>        <fct>              <fct>         <fct>       
#> 1 0            0                  0             0           
#> 2 1            0                  1             0