I'm going to assume that you scan to .pdf (and I do it all the time) in a way that it creates a bitmapped image of the page, rather than a formatted text copy. In other words, the computer doesn't recognize the letters as letters, just as series of dots. That's not an efficient way to digitally record and store text.

One solution is to use Optical Character Reader (OCR) software. Some probably came bundled with your printer. These programs look at the scanned page and convert the image into text. They're not perfect, but their errors are rare and almost always caught by spell-checkers. OTOH, an OCR scan loses the graphics and formatting of the page, giving you just the text. Whether that's a problem or not depends on what you're scanning.

You've already figured out one solution: Scanning at a lower resolution. Also, scanning as a black-and-white document will probably help.

And if you just can't get the file small enough, see A Better Way to Send Large E-Mail 'Attachments'.

Lincoln