The ARDL model supports that we can only use those independent variables that are either stationary at level or stationary after first-difference. After finding that all of the independent variables are stationary after first-difference, is it okay if I directly use the first-differenced variables (i.e., D.GDP for GDP) in my ARDL regression model. Normally, we use the variables without any effect of first difference. For example,

CO2=f(GDP, Export, Import)

Now, with first-difference the variables become, D.CO2, D.GDP, D.Export and D.Import

Normally, the ARDL command in stata is run as: ardl CO2 GDP Export Import

Now, if I use this command: ardl D.CO2, D.GDP, D.Export and D.Import instead of the above one, will it be okay?