To reduce variables, there are different techniques like Stepwise Regression, Forward Selection, Backward Elimination. if you use stepAIC function and pass linear model lm as parameter then it will produce the final reduced set of features. it also takes care of multicollinearity.

you may refer this link to know how stepAIC works:

https://ashutoshtripathi.com/2019/06/10/what-is-stepaic-in-r/