It depends on how you use it.
If you're pulling the same number of fields each and every time, you can loop through the records passing each of the fields from a single row as arguments to a function. Beacuse each result is going to have the same number of fields, you don't have to worry about blanks or lost data.
BUT, if this funciton is being called by different areas of your script that may not necessarily select the same number of columns all the time, you should add each of the fields to an array so you can pass one single item - the array - into the function. Then, one area of the script can pass in an array with 10 items, while another area of the script can pass an array of 20 items, and the function isn't going to complain about the number arguments. There's only one!