If all of your data (in the varchar fields) are *regular* characters, then you shouldn't lose any real data.
The 'data' you are losing is unicode representation of your data.
Since the files are 15 times larger than the original, there must be more going on here. Unicode data takes 2 bytes per character. Non-unicode takes 1 byte per character. So, at best, this would only account for a doubling in size (not 15 times larger).
I recommend that you backup your database before continuing this process. In Enterprise Manager, right click on the database, click on 'all tasks' then click on backup.
Then, do a shrink... right click on database, click 'all tasks', click shrink database, click the check box for 'Move pages to beginning of file before shrinking', then click OK.
There is another forum for SQL Server (Microsoft SQL Server: Programming). The guys over there are VERY sharp, and can probably help more than me.
Hope this helps.
-George
Strong and bitter words indicate a weak cause. - Fortune cookie wisdom