Intervals versus continuous training


As of September 2017, new Sweat Science columns are being published at Check out my bestselling new book on the science of endurance, ENDURE: Mind, Body, and the Curiously Elastic Limits of Human Performance, published in February 2018 with a foreword by Malcolm Gladwell.

- Alex Hutchinson (@sweatscience)


The age-old debate: which is “better,” interval training or continuous exercise? It’s a stupid debate — but I’ll get to that in a sec. First, a new study in the August issue of the Journal of Strength & Conditioning Research, from researchers in Spain. They put 22 physically active non-runners through one of three different eight-week training programs:

  1. Intervals: Three workouts a week. Mondays were 4-7 x 2:00, Wednesdays were 3-5 x 3:00, Fridays were 2-5 x 4:00, with rest equal to the length of the interval.
  2. Continuous: Three workouts a week, starting with 16:00 at 75% of vVO2max and building up to a high of 40:00 at 75% of vVO2max.
  3. Control: Nuthin’.

Here’s how the three groups progressed (MAS is their speed at VO2max):

As you can see (and as the paper concludes), the interval training and the continuous training produced virtually identical results. Which proves… well… nothing, really. Comparing a steady diet of 100% intervals to a steady diet of 100% continuous runs is like one of those “If you could only bring one album to a desert island to listen to for the rest of your life, what would it be?” conversations. Every different workout provides a slightly different stimulus to the body, so trying to identify “the best” is a pointless exercise. For optimal performance and health, we need a mix of different workouts.

The authors of the new study make an important point when trying to explain why previous comparisons of interval and continuous training have produced mixed results:

[I]t may be suggested that the exercising intensity and the subjects’ training background influence subsequent endurance training adaptations.

This is key! Take someone who has been “jogging” five days a week for a few years and have them start doing hard interval sessions a couple of times a week, and you’ll see dramatic improvements. But if you have someone who has been doing sprint training but no sustained running for a few years, and then add a tempo run and a long run each week, you might see equally dramatic improvements. In neither case does this “prove” that one type of workout is best — it’s context-dependent.

So what’s the perfect mix of workout types? Science doesn’t have an answer, but elite athletes have settled into some consistent patterns through trial and error. A reader (thanks, Marc!) recently sent me a link to an interesting review published a couple of years ago in Sportscience (full text freely available) that analyzes this question very thoroughly based on studies of elite endurance athletes in many different sports. They conclude that there’s an “80:20” rule for intensity:

About 80 % of training sessions are performed completely or predominantly at intensities under the first ventilatory turn point, or a blood-lactate concentration. The remaining ~20 % of sessions are distributed between training at or near the traditional lactate threshold (Zone 2), and training at intensities in the 90-100 %VO2max range, generally as interval training (Zone 3). An elite athlete training 10-12 times per week is therefore likely to dedicate 1-3 sessions weekly to training at intensities at or above the maximum lactate steady state.

Other researchers like Carl Foster break it down into three zones rather than two, and say that athletes do about 70% of their training below threshold, about 20% at or near threshold, and 10% above threshold. That’s a pretty small diet of intervals. Of course, elite athletes have different goals (and more time to train) than the recreationally active volunteers in the Spanish study — so the breakdown of three workouts a week might be quite different from 10-12 workouts a week. Still, my advice for anyone at any level is to include at least one interval session and at least one continuous session in your weekly routine — even if you’re just training twice a week!

13 Replies to “Intervals versus continuous training”

  1. Curious if they looked at results relative to muscle fiber composition? In other words, might a predominantly fast twitch runner thrive more on intervals, and a slow twitch runner thrive more on LSD? But when you don’t control for this, you get no difference? No idea if there is a relationship, just a thought – individual results may vary.

  2. Interval-training can mean different things to different sports as well. Swimming is a sport that is trained, at least at the elite level, almost entirely with interval training, but much differently – you can do intervals that will give you 5 seconds rest, repeatedly, for a given distance. For instance, a set might be 20 x 200 @ 2:05 if you are a swimmer that can hold 2:00 for every 200 (there is almost no way a runner can do this as they don’t just stop at a wall like a swimmer so they can’t be ready to go again in 5 seconds or less). Obviously this is far different than a 1:1 swim:rest ratio, but it is also different than doing a straight 4000. A huge percentage of elite swimmers’ training is done in sets like this where they are right at lactate threshold… maybe 60% if you are a distance swimmer. I never hear of other sports doing that kind of percentage of LT training, but they are also limited by how they can do interval training, and I suspect that is a huge reason why.

  3. I was about to bring up the swimming thing, but BMan beat me to it. When I was a kid during the 70’s, one of our typical 2hr workouts might have totaled 6500 meters, structured as 1500 warm-up, 4 sets of 10 x 100 intervals, followed by 1000 warm-down. This is very close to the 60% interval ratio that BMan mentions. The only thing different was that our intervals were designed to have more than 5s rest; usually about 20 or 25s. And note that the composition of all workouts was similar; we did this twice a day, 6 days a week. There was not really a mix of interval workouts, tempo workouts, etc.

    In my adult years I have switched to cycling, where there is usually a much smaller component of interval work. I assumed that my experience during the 70’s was obsolete, and that training methods for swimming had no doubt changed. But BMan is not the first to disabuse me of that notion. What’s with that?

  4. I can’t tell from the post or the article abstract the total time spent training over 8 weeks by the interval versus continuous group.

    If 45 minutes a week running intervals gets me the same training effect as three hours of steady, less intense running, I’ll opt for the intervals.

    Note: I’m not a competitive athlete. I’m just trying to gain the long-term health benefits of regular exercise, such as greater lifespan and lower rates of heart disease and cancer.

    These benefits, so well-documented by Blair et al, are linked to endurence on standardized maximal treadmill tests. They’re also linked to weekly hours of self-reported physical activity; the surveys back then didn’t delineate interval training versus slow, steady cardo work.


  5. Good points here — thanks for the comments.

    @Pete: Yeah, very interesting question. They didn’t look at muscle fibres, but it’s not hard to imagine that the experiment might produce different results with Usain Bolt than it would with Haile Gebrselassie. And of course, there are many other individual variations: lots of studies showing that some people respond to endurance training with big changes in VO2max, while others barely move the needle with the same training. Could be that those genetic differences are more fine-grained in terms of differing response to intervals versus continuous training.

  6. @Bman and @Phil: Very interesting point, and I honestly don’t know what the explanation is. I’ve always thought about the differences between running training and other sports like swimming in terms of impact: runners simply can’t do the same volume and intensity and recover in time. But biking also doesn’t have that jarring impact, so that doesn’t entirely explain it. I don’t think it’s just a question of what’s practical: there’s no reason runners and/or bikers couldn’t train in a 2:00 on / 10s off pattern — they just don’t.

    Two vague suggestions. One, that’s it primarily a cultural difference. After all, the main point of the study discussed above is that the difference forms of training are simply different routes to the same endpoint. Perhaps the cultures of the different sports have simply evolved to favour particular training styles. Second, I wonder if the use of different muscle groups plays any role. Running and cycling rely primarily on the very large muscles of the legs. Swimming is nominally a whole-body activity, but my assumption/guess is that the arms supply the majority of power. Could the limiting factors in supplying the different combination of muscles with fuel and oxygen somehow dictate different training limitations?

  7. @Steve Parker, M.D.
    Excellent point, Steve, and thanks for bringing it up. The authors of this study attempted to equalize the total workload in both cases. Workload was defined as duration in minutes multiplied by the “%MAS” (which is the workout running speed expressed as a percentage of the running speed at VO2max). So workload reflected a combination of duration and intensity.

    In practical terms, the actual durations were also identical when you include the recovery time. E.g. when the interval group was doing 7×2:00 at 100%MAS with 2:00 recovery at 50%MAS (a total of 28 minutes with average 75%MAS), the continuous group was doing 28 minutes at 75%MAS.

  8. @alex
    I can see cycling doing 2:00 on, :10 rest, even :05 rest, but runners cannot just stop instantly and rest for :05 I don’t think. There is a deceleration stage (and for that matter, an acceleration stage) that really limits it. Also for running you couldn’t stop on a dime so that you stop and start at the same place on the track every time so that you know that you are keeping an exactly consistent speed (with cycling you can almost do it w/ your speedometer, but even that is not going to be exact b/c you also will have some acceleration and some small variances that are not as noticeable and doing exactly 200 yards in exactly 2:00.0 and pushing off exactly :05 later – when I say exact, I mean within a few tenths of a second).

    Another comment is that swimmers do sometimes train with huge amounts of rest as well – when they are doing anaerobic training. Of course the percentages are dictated by what events you race (and your coach’s philosophy, among other things). Which brings me to your next comment – legs vs. arms. Interestingly that is probably at least half of the difference from swimming in the 1970s to swimming now (it was a gradual progression from probably 1980 to 1995 and rule changes were even made to prevent the sport from becoming entirely legs). In the 70s and early 80s swimming was predominantly an arm sport. But the massive increases in speed at the top level since then have been largely due to using more and more legs. Sprinting is considered a leg-dominated event now. As the events get longer there is a higher percentage of propulsion from your arms, though even in the 1500 (the longest pool event) the top swimmers, especially on the men’s side but somewhat on the women’s, use a very strong 6-beat kick (and interestingly studies show that the elite ones manage to use more anaerobic energy than aerobic even over a 15-minute event). Once you move to the swim marathon, though there is certainly some kicking (and more from men than women), it is much more leg-dominated, and it is definitely dominated by aerobic metabolism. Not surprisingly, training will reflect this as well, and pretty much every swimmer thinks of kicking as sprinting and pulling as aerobic.

    Though I have never seen a study to show this, my personal experience is that legs are not suited nearly as well as arms for aerobic activity. They are larger muscles that therefore require a lot more oxygen for aerobic activity and are also located much farther away from the heart and lungs and cannot be replenished with oxygen nearly as fast.

  9. Alex, cultural differences, yes obviously. But those are what we are trying to explain; they can only be the explanation if it doesn’t matter whether you train intervals or not.

    One thing is certain: *something* has changed over the last 3 decades. When I was a swimmer, it was normal for male athletes to be washed up (if you’ll forgive the expression) at age 22 or 23; women much younger. The limiting factor was the ability to recover from 2-a-day intense interval workouts. Some aspects of this training regime have relaxed; I recall that when Michael Schmidt won a gold medal at the ripe old age of 25, his secret was said to be limiting training to 2 hours a day. As for the rest, who knows? It could be as boring as doping for recovery.

    Regarding the use of legs in swimming, I can see how this explanation might appeal to a non-swimmer, but it is completely unrealistic. When swimming crawl, whether front or back, kicking is less *efficient* than pulling. But swimming races are short, and it is total power that matters, not efficiency. To be be able to compete with athletes of similar ability, you have to kick hard. An the point is simply irrelevant when applied to breast stroke, which relies heavily on kick, or fly, which recruits all significant skeletal muscle. The reality is that when in training, one experiences DMS and hypertrophy of leg muscles in all strokes.

  10. “Regarding the use of legs in swimming, I can see how this explanation might appeal to a non-swimmer, but it is completely unrealistic.”

    Regardless of the details of which muscles are used in what proportion, the fact remains that swimming uses different muscles compared to running and cycling. Bman says that swimmers spend a higher proportion of time near lactate threshold. What determines lactate threshold? The ability of muscle cells to get enough oxygen to fulfill the demands placed upon them.

    A tougher question is: what limits aerobic performance? Is it the ability of the lungs to supply enough oxygen? The ability of the heart to pump blood powerfully enough? The ability of arteries and veins to circulate that blood to where it’s needed? The rate at which oxygen can be transferred from blood to muscle cells? The rate at which mitochondria in muscle cells can use the oxygen to produce ATP? Nobody really knows, but there’s evidence that a bottleneck anywhere along that chain of events can limit performance.

    So what happens if you’ve got a cyclist exercising just below lactate threshold, and you suddenly and magically transform him into a swimmer and ask him to maintain exactly the same effort level? His heart is beating the same, his arteries are the same, but now the same amount of oxygen is being circulated to arm AND leg muscles, and (roughly) twice as many muscle cells are trying to get that oxygen. Does this mean that there will be an oxygen shortage at the cellular level and lactate levels will rise? This could mean that swimmers would be at a higher lactate level than cyclists at a comparable effort level.

    I don’t know the answer — I’m just speculating wildly here. My point, though, is that the use of arms and legs rather than just legs could be physiologically relevant and offer some explanation for why the training isn’t directly transferable.

    The other point I’d make is that the line between interval and continuous training isn’t necessarily that bright. Doing 20 x 2:00 intervals on 0:05 rest strikes me as being culturally but not physiologically different from doing 40:00 continuous. Those 0:05 rests may be significant in swimming because they allow a brief period of unconstrained breathing. But in cycling, I don’t think that stopping pedalling for 0:05 every 2:00 would change the physiological stimulus of the workout significantly.

  11. “My point, though, is that the use of arms and legs rather than just legs could be physiologically relevant and offer some explanation for why the training isn’t directly transferable.”

    It could, but then shouldn’t that apply to other whole-body sports like classic-style Nordic skiing?

    “The other point I’d make is that the line between interval and continuous training isn’t necessarily that bright.”

    That is obviously true as a generalization, but it was BMan who mentioned 5s rest. I’m talking about winding your pulse up to 240 bpm here. If that’s not an interval, then intervals don’t exist.

  12. My comment about interval training was that the term is ill-defined for the purpose of talking generally about what type of training has what benefits. You can do interval training with :05 rest (which is not the same as straight swimming – I’l argue that later) and you can have interval training with :15 seconds rest, or with a minute’s rest, or 6 minutes rest. All of these train your body differently, stressing different energy systems different amounts.

    Swimming 2 hours or 1 hour or 40 minutes continuously is definitely not the same as breaking the swim, even for :05 every 2:00. Though I cannot cite any studies on this, I can tell you from extreme experience that there is a definite reset in your body that takes place, on a system-wide level, with even a :05 rest. You push off the wall and not only are your arm and leg muscles feeling a bit more speed, but your skeletal system can hold you higher in the water without effort and your gut does not develop the utter pain of consistent swimming (OK, the gut/diaphragm pain could be related to the restricted breathing of swimming). Perhaps they all are related to the restricted breathing issue – it makes me think about how your body performs better just by tasting carbs, as opposed to actually digesting them. Perhaps your body senses air and therefore is not just mentally but physically fooled into some sort of a reset.

    As far as swimming when you are older, swimmers workout harder now than they did in the time of the ultra-distance training of the late 1970s and early 1980s. And, in fact, training distances have risen to about the same (from a low in the 90s), but intensities have gone up and time in the pool has gone way up. You hear about a few exceptions (e.g. – Jason Lezak), but those are mostly sprinters. A lot more is known about recovery now, a lot more resources are put into recovery, and a lot more is known about how to plan a season or multiple seasons so that one’s body can handle it. Like any sport there are a few people doping, but it is far less prevalent than in cycling or many other sports IMO. But legal supplements are used heavily, timing of ingestion is perhaps the biggest change, and other legal methods are used. (And who is Michael Schmidt? Do you mean Michael Gross? He was 24 in 1988 when he won the 200 fly).

    No question that legs supply the power and arms supply the endurance (obviously simplified). Not all swimming races are short, however. The open water 10K isn’t so short… basically a 2 hour race (and there is a 25K at World Championships that is not in the Olympics yet). And of course, there is a reduced use of legs in this event. What is curious is the 1500 – a little under 15-minutes for a good male swimmer, and about 16 minutes for a good female swimmer. All the top men have a fairly vigorous 6-beat kick, but with the women it is a mix, and even those women that 6-beat, it is not as vigorous as with the men. Is that minute or 1:20 difference in time the cutoff for using legs significantly. Possibly related to this is that some recent studies on elite male swimmers have shown that the top men are now swimming the entire 1500 at En3 (above LT – lactate levels above 4.0 mM). Conventional wisdom is that is impossible for that long.

    Anyway, arms are much smaller and much closer to the heart and lungs, so those bottlenecks are reduced. I am not sure if arms have a higher percentage of slow-twitch fibers or not, but I wouldn’t be surprised. I would also agree that training both arms and legs simultaneously can cause greater aerobic adaptation. I would also suggest that training horizontally likely changes something, or perhaps more importantly racing horizontally is the bigger key.

Comments are closed.