I read in a data file from Excel, like I do all the time. But this time, SPSS added the "@" symbol at the beginning of all variable names that started with a number. When I tried to remove the symbols, they would reappear and I would get an error message that said the variable name started with a symbol that was not allowed. The correct names of the variables were put in the Variable Labels slot. Although I could not remove the "@" from the start of each variable name, I could change it to another letter/number. That is a work-around for the moment - I don't wish to have to change the variable names in all my documentation just because SPSS is adding a "@" and won't let me get rid of it. Any thoughts on why this is happening and how I can prevent it?

#SPSS Statistics

Note this information on variable names in SPSS.

Thanks so much -- 

I looked up that possibility before posting and found several on-line sites that said they could start with either a letter or number. I guess I didn't look far enough as just now I found a site that said they can only start with letters. Thanks for the definitive answer, Jon! Guess I am changing all the documentation anyway. 

I've seen this happen a lot when Excel reads in headers with special characters or Unicode symbols and silently normalizes them into valid variable names. It's similar to what happens with styled text copied from social platforms, where fonts or symbols get altered on import. If you're dealing with names that include decorative characters, it helps to sanitize or map them early. I ran into this while testing styled names from tools like instanamesstyle where preprocessing avoids unexpected renaming.