Functions tbl_regression() and tbl_uvregression() will pass any arguments passed in the ... to broom.helpers::tidy_plus_plus(...).

There are restrictions, however, when passing arguments to tbl_uvregression(). Keep in mind that serially constructs model objects and passes these model objects to tbl_regression(). Lastly, this series of 'tbl_regression' are stacked using tbl_stack().

Because the arguments passed in ... will be passed to tbl_regression() for each model, you must ensure that the arguments passed will not result in an error in any of the calls to tbl_regression(). For this reason, when passing arguments that take, for example, column names, you must pass the names in any_of(). For example, if you wanted remove reference rows, you would use:

trial %>%
  tbl_uvregression(
    method = lm,
    y = age,
    include = c(trt, grade),
    no_reference_row =  any_of(c("trt", "grade"))
  )

For more flexible inputs that accept all tidyselect notation, you must utilize tbl_regression() individually for each univariable model, and stack them with tbl_stack().