Renato, thanks for the info.
I am not now sure what to think as I have always used the "conventional" threshold training but this idea of using high intensity efforts to improve the threshold is altogether new.
But I must say that I think the types of examples you gave seem far too intense to do regularly. The following is a piece by the coach I referenced earlier, Wejo's coach John Kellogg. He speaks about threshold training as I have understood it; please read it and tell me what you think (from paragonrunning.com):
"Training in the zone which influences the lactate threshold(LT) is perhaps the single most important sort of hard training a distance runner can do, although it does not by itself cover every base. The key speed involved with this fast aerobic running is often called lactate threshold velocity (LTV). This is roughly the speed which can be maintained for one hour if the pace is even throughout. Hovering at or just below this speed will bring the blood lactate level up to about 4.0 mM/L, while the heart rate will generally rise to near 85% of the way from its resting value to its max value. Breathing frequency at this pace will normally be 5-6 steps per exhalation. If a breath is required every 4 steps, the pace is probably too fast, and a slightly tight, uncomfortable feeling may shortly ensue.
The critical zone in which to run to improve LTV is between 95% and 105% of LTV. For example, if LTV is 5:00 per mile, 95% of this speed would be 5:15.7 pace and 105% would be 4:45.7 pace. Training within this range has several purposes.
Continuous running at LTV for 18-22 min. once or twice per week will raise the LT over time, while running at slightly slower speeds (95%-96% of LTV) for about an hour will have an impact on running economy. 95% of LTV will put the HR at about 80% of its range and will induce a blood lactate level of about 3.5 mM/L. The world's best marathoners can often run at this pace for the full marathon distance, with blood lactate levels at 3.5 mM/L for near 20 miles (a small oxygen deficit is meted out over the last 4-6 miles of the race). Most runners can manage no better than 93%-94% of LTV for their marathon race pace.
Serious runners may have trouble holding back to the correct speeds for workouts such as these and might be tempted to turn the runs into time trials. Time trials and races are important, too, but they are too stressful to use year-round as training devices such as these threshold workouts are meant to be. It's important to stay within the most effective zone to improve LTV and economy. Many elite runners (particularly Kenyans) have developed the ability to approach (but not reach) the lactate threshold in training almost every day, but this is done by feel and not by timing the runs. While runners with less experience should limit their efforts at a near-LTV to 2-4 workouts per week, running by feel and not by time is the ideal way to use threhold (or "high-end" aerobic) training, even during repetition running. The Conversion Chart gives the predicted LTV based on current race times, but it’s wise not to be a slave to such numbers, because (as mentioned in the 'Continuous Easy Runs' page) coming very close to the threshold too often or exceeding it on a regular basis can be ruinous.
Lactate threshold training also indirectly affects VO2max and reinforces preferential muscle fiber recruitment if the speed is increased near the end of the workout. This principle of working two critical speeds (in this case LTV and vVO2max) is one of the most important, yet least practiced, procedures of training. It is vital to negative split everything when running all hard workouts except pure sprint work. A "negative split" refers to running the last section of the workout the fastest. Runs at slower than LTV are considered moderate and need not be done with negative splits.
Performing the bulk of a workout at faster than LTV (e.g., 105%) should be reserved for repetition running; otherwise, lactate levels will rise too quickly to have enough running time to be effective. Repetition running can be used at any speed within the LT training zone, but it must employ extremely short rest periods to be of use. These rest periods can be as short as one-third or even one-sixth of the run periods for the particular workout being performed (e.g., 3 min. runs at LTV with 30 secs. rests).
Remember to warm up well (at least 12 min. of jogging plus a few buildups/strides is suggested), and do a sufficient cool-down jog and post-workout stretching in conjunction with any faster-paced workout effort.
Examples of Lactate Threshold Workouts
Continuous run of 26-32 minutes with the first 5-7 min. at 10-15 seconds/mile slower than LTV, the next 18-22 min. at exactly LTV, and the last 2.5-3 min. smoothly accelerated from 95% of VO2max speed down to 105% of VO2max speed.
Suppose Dave is a good collegiate runner with a 14:20 5,000 PR and a 30:00 10,000 best. A 14:20 5,000 time predicts a LTV of 5:02 mile pace and a VO2max pace of 4:30 per mile. Dave would thus probably benefit most from a LT workout beginning at roughly 5:10-5:15 pace for a mile, 5:02 pace for 18-22 min. (4 miles would take 20:08), then about 0.6 miles at an average speed of 4:30 mile pace (allowing the first min. of that last segment to squeeze the pace down). The total distance covered would be about 5.6 miles, give or take a bit, with a total time of around 28 min. (average pace of 5:00), plus or minus a few minutes if Dave decided to go longer or shorter in the middle.
Since Dave could theoretically run 5:02 pace for an hour, he could obviously run harder during those middle 18-22 min., but the effect on his aerobic development would not be any better. In this case, going harder would only mean getting an anaerobic workout which would require more recovery time and which might actually undermine full aerobic development. Remember, economy is promoted through a steady VO2, which can be controlled perceptually by keeping the effort aerobic, the breathing pattern consistent, and the heart rate constant. Heart rates during the time spent at LTV should stay below 85% effort. If Dave's resting pulse is 40 and his maximum is 200, his range is 160. 85% of 160 is 136, so by staying below 40 + 136 = 176 beats/min., Dave would be running within 85% of his maximum effort. A little higher pulse near the end of the workout is acceptable.
These LT workouts are not time trials and the reins should be kept on the pace until the very end. Notice, however, that two critical speeds (95%-105% of LT pace and 95%-105% of VO2max pace) are being attained. This ensures that correct neurological feedback is achieved by recruiting more ST fibers early and more FT fibers later.
It is best to run all LT workouts in racing flats in order to develop ankle power and flexibility. This not only improves running performance, it also reduces the risk of injury. Heavy training flats restrict the full range of motion of the ankle, and performing every single continuous run in "trainers" is a prescription for injury.
Continuous run of 60-70 min. at 95%-96% of LTV with only a minor increase in tempo during the last few minutes.
This duration at this pace is comparable in intensity of effort to a run of 18-22 min. at LTV. It is therefore not as hard a workout as the continuous run described in the previous section, since additional running at a faster pace is performed in that shorter effort. Both workouts require similar recovery times, however. This is due to impact stress and the fact that the longer run expends more calories and requires more carbohydrate reloading. Heart rates should hover around 80%-82% of maximum effort.
This slower, longer tempo run was employed by the late Kiyoshi Nakamura in his coaching of 2:08:27 marathoner Toshihiko Seko, and it is still used as a staple workout in the training of Japanese marathoners. Statistical evidence (and experience) has shown that running at 95.2% of LTV for 65 min. provides the optimum training effect for improving both LTV and running economy. Elite runners generally run at 93% to 95% of LTV for a marathon race, so this 60-70 min. effort also provides a good pace workout without excessive impact stress and without risking glycogen depletion.
Collegiate runner Dave, with a 10,000 PR of 30:00 and a LT pace of 5:02 per mile, would run at 5:15-5:18 pace for a 60-70 min. tempo outing, and might get as far as 13.3 miles in 70 min. (the Japanese marathoners often go 20,000 meters, or 12.43 miles, in around 61-62 min.). Since Dave has a resting heart rate of 40 and a max of 200, 82% of his max effort would result in a pulse of 171. By keeping his heart rate around 170 or slightly under, he will derive the desired benefit from this particular workout.
Dave's 30:00 10K PR is equivalent to a 2:21:44 marathon, which requires a 5:24.3 mile pace, so by running a few secs./mile faster than this, Dave is also developing efficiency at the rhythm needed for the marathon in case he wants to contest this distance.
2 x 15-18 min. at LTV (no variation in pace during either run)/5-7 minutes easy jog between runs.
This workout is a slightly harder effort than either of the previous two LT workouts, especially if the maximum run length of 18 min. is used on both reps (2 x 15 min. is normally used by runners with less base mileage or less experience). No attempt should be made to pick up the pace at the end of either rep, although the entire second run can be done faster than the first if the first seems excessively easy. This workout allows more time spent at LTV than would be prudent to spend in any continuous effort (without the 5-7 min. rest) short of a time trial. The 5-7 min. recovery jog period gives the time to clear any accumulated lactate and lets the neuromuscular system relax briefly.
For 30:00 10K runner Dave, 2 runs of 3.5 miles each at a 5:02 pace provides the optimum effect directed toward improving his LTV. Each 3.5-mile segment would require about 17:38 at Dave's LTV.
This type of workout is also excellent for improving economy (owing to the overall time spent at LT pace), provided the pace remains even throughout both runs and never exceeds LTV. Going faster than LTV for 18 min. twice would probably become too strenuous to affect running economy.
It is usually a good policy to add 4-6 buildups during the post-workout cool-down period in order to recruit FT fibers. This ensures that FT fibers will be mobilized after ST ones are fatigued from the workout. The purpose is to influence correct sequential fiber recruitment, which is one of the most essential components of fast-paced running.
10-15 x 3 min. runs at LTV (first 1-2 reps slightly slower, last 2-3 gradually faster)/30-35 secs. rest periods between.
This workout provides 30-45 min. of running at LT speed, which would probably be too close to a time trial effort to maintain efficiency were it not for the rest periods. Notice how short the rest periods are in comparison to the run periods. This type of workout is sometimes referred to as being of high density. It is crucial to stay within the pace guidelines on high density sessions, since going too fast with very little rest would adversely affect relaxation and economy.
Dave, our college distance runner with PRs of 14:20 and 30:00, has a LT pace of 5:02 per mile. Dave would cover 950 meters in 2:58 at his LT pace (75 secs. per 400), so he could run the indicated workout on a track, then shuffle the remaining 50 meters to the next half-lap line before starting his next rep. This workout could also obviously be performed away from the track, either on a measured course or by effort. As usual, it's best to start slower and finish faster. So Dave could begin with 77 second 400s, warming up a little, and finish with 73 pace or even faster. The middle 2/3 of the reps should be done right at LT speed in order to promote efficiency.
Ronaldo DaCosta (2:06:05 marathoner) has reportedly run 15 x 1,000 meters at 3:00 each with 30 secs. rest periods. This is actually about 1 second per mile slower than his marathon pace! This is, however, just a workout (and no mention is made as to whether it was done at altitude). The moral is that relaxation and efficient rhythm are more important than "tying up like a big dog" when the objective of the workout is to train the aerobic component. If a former World Record holder runs his 3 min. repeats that slowly in comparison to his LT pace (which is probably about 4:35 mile pace), then going a little easier than possible can't be all that bad! Actually, DaCosta's 1,000s are run at about 95% of his predicted LTV, which is still within the zone which has an impact on raising his LT, and the high volume at this pace will certainly improve his economy.
16-20 x 400 at 105% of LTV (about 3-4 secs. per 400 faster than LTV)/20-25 secs. rest periods.
This workout gets to the upper limit of the training zone which works on raising the LT, and it actually approaches the speed which trains the VO2max. Its effect on LTV results mainly from the very short rest periods. The pace involved is a little faster than 10K pace for most runners, and people slower than 35:00 for 10,000 should tend toward 16 reps rather than 20 (some world-class runners do more than 20) at this pace.
Our hypothetical 30 min. 10K runner Dave, whose LTV is 5:02 mile pace (75 secs. per 400), runs at 71-72 secs. for his 400s on this workout. A few can be slower at first, some gradually faster at the end (as usual), but about 3/4 of the reps should stay at 71-72 to reinforce the rhythm.
Since the rests are so short, the heart rate will not decrease much (if any), but this is good by reason of the fact that lactate is used as a fuel by the heart as long as the rate of its accumulation does not increase sharply. If performed correctly, this workout produces a virtual steady state of cardiac output at a much faster pace than could be attained on a continuous run without going into the anaerobic zone. The heart rates during this workout generally stay above 85% of maximum effort but below 90%.
Ronaldo DaCosta's published workouts include up to 25 x 400 at a 66.2 average with 15-20 secs. recovery periods. Because of the higher number of reps and shorter rest periods, this should represent 103% of his LTV, and indicates that he could run at a 4:34-4:35 pace for an hour (about a half marathon). This is equivalent to "only" 2:07:30 for the marathon, but is consistent with DaCosta performing better as the distance increases. DaCosta also appears to work out a little below his utmost capabilities, possibly to work more on economy or possibly because of the cumulative loading effect of his two long runs per week. He also reputedly performs drills regularly and would presumably want to remain fresh enough to make them productive. In any event, there's a lesson in this inasmuch as workouts which are meant to train the aerobic system should not be run as hard as possible! "
I believe Mr. Cabral also has these articles.