Would be good to add info to make your example reproducible, but something like this might work for you. Looking in the data, if you're referring to RIDAGEYR as the variable, then replace it with the "agevariable" below so it'd be:
library(tidyverse)
newdf <- yourdata %>%
select(variable1, variable2, agevariable) %>%
filter(agevariable <=15) %>%
arrange(agevariable)
head(newdf) #to check to see if it worked for you.