Terminology is critical in IT, where similar terms may mean very differnt things. DB2 databases have "rows", they do not have "records". You can refer to the 5000th record of a file, for example, as there is an order imposed on the file. However, if you issue the same query to a database three times to retrieve 5000 rows, you might get back 3 different sets of 5000 rows -- or the same set of 5000 rows. You do not know what you will get back, though, unless you use an ORDER clause as part of the query.
So basicaly the very concept of the question you asked makes no sense as you asked it -- you could equally delete every 10th row of the database. Either reformulate the question to make sense, or give up on the idea of doing whatever you think you are accomplishing by asking for nonsense.