How to calculate emmeans using get() in a loop

I am looping over about 20 columns, running mixed models. That all works fine. But when I try to calculate emmeans, I get an error because of the get(thisvar) in the model. The code is:

for(thisvar in thesevars){
mod_linear <- lmer(edss ~ time + get(thisvar) + age + gender + relapse_prior_1yr + new_lesions_prior_1yr + weight + (1|id), data = dat_long)
emms_mod_linear <- emmeans(mod_linear, ~ time, at = list(time = seq(0, 12, 1)))
}

and the error is:

Error in (function (object, at, cov.reduce = mean, cov.keep = get_emm_option("cov.keep"),  : 
  We are unable to reconstruct the data.
The variables needed are:
    time thisvar age gender relapse_prior_1yr new_lesions_prior_1yr weight
Are any of these actually constants? (specify via 'params = ')
The dataset name is:
    dat_long
Does the data still exist? Or you can specify a dataset via 'data = '

How can I use emmeans in a loop with get()? Thanks

A fix for your issue would be to get rid of get() and instead use e.g. reformulate to create the formula object in your loop.

Using a minimal reproducible example based on mtcars and using lm:

library(emmeans)

thisvars <- c("hp", "cyl")

emms_mod_linear <- list()

for (thisvar in thisvars) {
  fml <- reformulate(c("am", thisvar), response = "mpg")
  mod_linear <- lm(fml, data = mtcars)
  emms_mod_linear[[thisvar]] <- emmeans(mod_linear, ~am, at = list(am = c(0, 1, 2)))
}

emms_mod_linear
#> $hp
#>  am emmean    SE df lower.CL upper.CL
#>   0   17.9 0.676 29     16.6     19.3
#>   1   23.2 0.822 29     21.5     24.9
#>   2   28.5 1.796 29     24.8     32.2
#> 
#> Confidence level used: 0.95 
#> 
#> $cyl
#>  am emmean    SE df lower.CL upper.CL
#>   0   19.0 0.753 29     17.5     20.6
#>   1   21.6 0.938 29     19.7     23.5
#>   2   24.2 2.128 29     19.8     28.5
#> 
#> Confidence level used: 0.95