Like many things, I think it's less of one key factor and more of an accumulation of many. Like what Ray said.
On a semi related tangent while poking through some research articles on artificially suspending spike growth; that initiating spikes by lowering the ambient temperature by about 10 to 15 degrees for a short period was a common theme. However, if the plants not healthy you'd probably wouldn't be doing it any favors.
There also was supposedly no change in response to different lengths of photoperiods, though that study had some issues.
*shrugs* It might be something to consider.
Personally I'm all for "It happens when it happens".