High-High vs Low-Low Four-Group Logistic Regression with Additive Interaction

Analyzes joint effects of two continuous exposures by median dichotomization, creating four exposure groups and fitting logistic regression models with sequential covariate adjustment. Calculates additive interaction measures.

Usage

highlow_analysis(
  data,
  outcome,
  exposure_a,
  exposure_b,
  model2 = NULL,
  model3 = NULL,
  model4 = NULL,
  output_dir = NULL,
  save_format = c("none", "docx", "all"),
  recode = FALSE,
  filename_base = "highlow_analysis"
)

Arguments

data: Data frame containing outcome, exposures and all covariates
outcome: Character scalar, binary outcome variable name (0/1)
exposure_a: Character scalar, continuous exposure A variable name
exposure_b: Character scalar, continuous exposure B variable name
model2: Character vector, additional covariates for model 2
model3: Character vector, additional covariates for model 3
model4: Character vector, additional covariates for model 4
output_dir: Output directory for saving results, default NULL
save_format: Save format: "none", "docx", "all", default "none"
recode: Logical, whether to recode interaction variables, default FALSE
filename_base: Base filename for outputs, default "highlow_analysis"

Value

List containing analysis results, interaction measures, and optional saved file paths

Examples

if (FALSE) { # \dontrun{
result <- highlow_analysis(
  data = mtcars,
  outcome = "vs",
  exposure_a = "mpg",
  exposure_b = "wt",
  model2 = c("cyl", "gear"),
  model3 = c("carb"),
  output_dir = tempdir(),
  save_format = "all"
)
} # }