It depends on what you want.
Missing results in proc freq means that you do have missing values in the variable. Proc Freq is well tested and just because you think there aren't any missing values does not mean they aren't in the actual data.
Run this code to create a data set where you can see which observations in that data set have missing values for the Waistline variable.
data temp;
set lib.dresses;
where missing(waistline);
run;
You can place * between variables on the tables statement to see how two (or more) variables interact in the data. You could add the option Missing to the tables statement so the row in the table shows the missing values instead of the note at the end of table.
This shows a simple listing of the combinations of neckline and waistline with the missing included.
proc freq data=lib.dresses;
tables Neckline * Waistline / missing list;
run;
Consider: If a dress does not have any sleeves what value would expect to see for Sleevelength? A value of 0 would imply that there was a sleeve but the length was 0, which seems odd to me.