Ben Parsons givin' er at Jackson - photo: Mark GockeIntervals. The very thought of the word makes my pulse rise. From my first exposure to them as a 14 year old swimmer to my weekly date with them now, intervals inspire varying degrees of pain and satisfaction while preparing me to perform at my best. If you want to go fast in an endurance event, there is no substitute for regular exposure to high intensity training. While I maintain that the bulk of one's training should consist of relatively low intensity effort for reasons best covered by Mark Twight here, twice-weekly bouts of threshold or higher work prepares us for competitive effort. Zone 2 training may lay the aerobic foundation but zone 4 and 5 efforts are what bring out our best performances.
There's more than one way to skin this intensity cat. Athletes and coaches alike spend lots of time thinking of ways to structure these efforts and get the work done. Many athletes obsess about rest/work schemes. After all these years, I'm not sure that little differences matter all that much.
More rest, higher work intensity, less rest, more pain, long, short, pyramids, etc. They all stress our physiology in ways that bring about adaptation which allows us to go faster. On the other hand, I have come to believe that longer intervals (20-40 minutes) are valuable in sports that require lengthy threshold efforts (skimo and MTB). Other than that, break them up as you like. Vary the mix from week to week and try to stay entertained. But the pain will be there and it must be faced head on, regardless of the structure.
As a target event approaches (1-3 weeks out), paring down the work interval and increasing the rest will keep your efforts sharp and fatigue at a minimum. The right mix and volume likely varies from athlete to athlete. Cookie cutter recommendations offer a place to start but probably stray from each individual's ideal. Trial and error are the way.
Earlier this winter, as I prepared for the National Skimo Championships in Jackson, I made some observations about interval structure and my response to them that I want to share here. It raises some questions to which I don't really have an answer but thought that putting it out there may yield some insight from readers (MFT? WW? You out there?).
Nothing earth shattering here but simply some thoughts about the appearance of two HR graphs from two different threshold-type sessions. Both workouts took place on the same stretch of terrain on Snow King ski hill in town. Basically, the "loads" were similar. Ski gear was the same. Heart rate is measured using a Suunto T6 wrist computer and downloaded to Movescount. My speed is unknown as I did not track overall time over the distance. But my HR response is nearly identical. For what it's worth, the efforts "felt" the same. I know, a nearly worthless statement but... whatever.
The first involves a steady state effort at threshold. As you can see, my HR ramps up predictably and is maintained relatively constant through the duration of the bout. 
A second effort reveals the same response. A couple of sharp spikes in HR are the result of brief supraventricular arrhythmias, probably stemming from catacholamine release during the hard interval. As long as these are brief and not symptomatic, there is nothing for me to worry about.
The second graph is the result of a similar length effort but done as a series of 30 seconds hard, 30 seconds easy efforts. Initial HR response rises more slowly but then is maintained throughout the bout. Small dips in HR are seen giving the graph the sawtooth appearance.
At a glance, the trends of the graphs are nearly identical.
So, with two essentially identical HR responses to two exercise bouts, can I assume that their impact on my physiology is the same? I honestly don't know. I will say that the effort during the 30/30 is of higher quality. However, I use a different movement pattern to drive HR up during the short work interval. This certainly has a different motor patterning impact on technique. This style of "running" is similar to what we do for the first 30-40 seconds of a race but not something sustainable after that.
One could ask if it's worth "practicing" this type of pattern since it's not the one I will use for the majority of the race. I don't know. The movement pattern is one used at the start of events and I have broken into it during flat sections of climbs or charging into transitions. It has merit. But, really, I use it in the 30/30s due to its higher metabolic demand. Plus, it's kinda fun.
The other part of this equation is the "power output" of the different stride patterns. If we had power meters like cyclists do I think my wattage would be higher with the running style stride. This certainly creates a different physiologic demand, even at a short duration, than the more sustainable stride pattern. How this impacts overall conditioning is unknown to me.
For now, varying these motor patterns on the way to creating metabolic stress simply feels right to me. It also breaks up the monotony of training. Both of these are worthy psychological factors that probably have a positive impact on my overall preparation. Hard to really know for sure. Other opinions welcome. - Brian