800 dude wrote:
By aerobic homeostasis, I mean that at MP you've started to "embarrass" your aerobic system, so the anaerobic system starts to make a measurable contribution, and muscle oxygenation starts to decrease somewhat. Essentially, your aerobic system is actually starting to get stressed. Running at these paces is great aerobic training, which "pushes" your LT up.
Obviously you can also increase LT by increasing your total anaerobic capacity. The problem with too much running right at LT is that it has actually been shown to DECREASE your anaerobic capacity, which means that, while you might be running the same pace as before, with a lower blood lactate level, you aren't able to accelerate to the blood lactate level that corresponded with your previous LT. Or, rather, you are, but that lactate level is actually supra-threshold now, so you'll blow up quickly.
As for 3k paced work, I'm inclined to think that the benefits are (1) neuromuscular, (2) anaerobic (depending on the structure of the workout), and (3) psychological. All three are really important if you're racing shorter distances. The psychological component in particular is often overlooked. The kind of pain you experience racing a hard 3k or 5k is so qualitatively different from the pain at the end of a fast finish long run that if you haven't gotten a bit familiar with it, it's very hard to push through.
Thanks for the answers so far. This is very close to what I was looking for. I had no idea that LT could be "pushed up" by MP training, but it makes sense. High-end aerobic runs are, if I understand it correctly, "widening the gap" between your aerobic and anaerobic states.
Given this construction, it would make sense to progress from high-end aerobic running to LT running...
But I'm concerned about the comment that eventually too much training at LT pace could decrease anaerobic capacity. I had always heard something different...that too much running at sub-LT pace could harm your lactate threshold...but you're suggesting it can also work the other way around....
Is this an argument for linear (periodized) training?