Someone else had a similar-sounding issue with this package last week:
I guess it kind of went away on its own so they closed the issue. 
If you are able to reproduce the issue or collect some data points, I would encourage you to chime in on that thread and see if you can help figure out what is happening to you guys.
Edit to add a follow-up thought:
@glomar_response I just saw your comment in this other thread:
It occurred to me it seems possible these two issues could be related. If the checkupdates utility is repeatedly attempting to connect to repository mirrors that are currently having connectivity issues, it may lead to timeouts and resource consumption as the utility waits for responses.
I don’t know this for sure (just kind of thinking out loud here), but it could be that the Waybar utility is continuing to kick off additional instances of checkupdates while the old ones hang due to the mirror issue. Eventually all the hung checkupdates processes could start to pile up. That would explain the resource usage climbing out of control.
Edit 2
I ended up chiming in on this issue to see what the developer thinks about this theory, and if they would be willing to amend the code so only a single sync_database process will run at a given time (so they can’t pile up if they get hung on a mirror issue). https://github.com/coffebar/waybar-module-pacman-updates/issues/4#issuecomment-1925863817