For example, the General Social Survey requests respondents to name up to six individuals with whom they discuss “important matters.” The assumption is that people discuss matters that are important to them with people who are important to them … However, a recent study by Bearman and Parigi (2004) shows that when people are asked about the so-called “important matters” they are discussing, they respond with just about every topic imaginable, including many that most of us wouldn’t consider important at all. Even worse, some topics are discussed with family members, some with close friends, some with coworkers, and others with complete strangers. … Bearman and Parigi also find that some 20% of respondents name no one at all. One might assume that these individuals are “social isolates” — people with no one to talk to — yet nearly 40% of these isolates are married! It is possible that these findings reveal significant patterns of behavior in contemporary social life — perhaps many people, even married people, really do not have anyone to talk to, or anything important to talk about.
— from Structure and Dynamics of Networks
Drunk Warhead 
I think this survey is really true or maybe realistic, to some extent. I guess you post it cos’ it really meets your situation.
I recently read a book called “Love Letter” by a Jap. writer and it shared a story how two penpals who have links with a dead man happen to be the lovers of the man himself.
It might sound ridiculous but the process is quite natural. Any way, all I want to say is that even though it seems rather safe to spit out the secrets to a stranger, you can’t prevent it from being a future hidden threat.
No. It does NOT apply to my situation. I posted it just because it is curious. In many cases, the benefit of communicating with a stranger is larger than the potential threat.