.m files are text files and can be read by any version. There are functions defined in R2013a that were not defined in earlier versions so the code might not necessarily work, but the file will be readable.
The exception to this is that there is a possibility that you have saved your .m file in UTF-8 form; if you try to read a UTF-8 file in an early enough version then it will treat any extended characters as just sequences of bytes. This would affect you most if you have string constants that use characters that are not in ISO-8896-1 .
