Really your question should be "how do I stop my emails being flagged as spam?"
The problem is that while every email service provider has some sort of spam filtering - nobody likes to say how its implemented in case the spammers use this information to get around the blocks. A lot of thought went into making IP and SMTP very highly fault-tolerant, so in the absence of end-to-end authentication there is no definitive way to separate spam from ham - leaving security by obscurity as the only solution.
I think its unlikely that the X-Authentication-Warning is explicitly causing the spam filter to kick in, however a very common feature in spam prevention is Bayesian filtering - and if that term occurs more often in spam than ham, then yes, it will contribute to the problem.
If you want to improve your deliverability, then do make sure you publish restrictive SPF 1.0 records, make sure you've masquerading configured on any internal MTAs, and make sure any external facing MTAs have PTR records published which include the mail domain in the IP address. It would also be a good idea to study how spamassassin works.