Original Research: Marathon Pace Prediction

Marathon Running

All marathoners, from fast to last, predict their race pace, usually by comparing shorter race times, time trials, previous marathons or “gut” feelings. But studies show that the best predictor of race performance may be submaximal performance. The importance of accurate prediction is more than a game — it can help monitor training progress, increase race confidence, provide an invaluable pacing strategy and even predict injury.

Dr. Philip Maffetone

Just after the 2016 U.S. Olympic Trials in Los Angeles, coach Alberto Salazar claimed that Galen Rupp (his star 10K runner and now marathon winner) had recently clocked a 20-mile training run at a 4:52 minutes per mile (min/mi) pace, with a heart rate of 150 beats per minute (BPM). If this is true, Rupp would be the clear favorite to beat the East Africans at this summer’s Olympic Games in Rio, Brazil. But the normally secretive Salazar rarely surrenders specifics about his athletes.

However, it must be considered that when running at any given heart rate (HR), the onset of fatigue causes later miles to be slower than earlier miles. If HR remains constant, pace drops. And if pace remains constant, HR rises. Therefore, it’s highly unlikely that Rupp can perform a workout where the 1st and 20th mile are run at the same pace, while maintaining a constant submax HR.

Regardless, Salazar may have spilled the proverbial beans: Speed at an aerobic submax HR is highly predictive of marathon performance. Let’s suppose that only Rupp’s 1st mile was run at 4:52 min/mi, with a submax heart rate of 150 BPM. This time could predict that Rupp might not only win an Olympic medal in the marathon, but, on a fast course with cool temperatures, also could establish a new marathon record. He could potentially even break the 2-hour mark — the greatest remaining barrier in the sport since Sir Roger Bannister broke the 4-minute mile.

(In the book “1:59,” I present the evidence and details for how a runner could break the 2-hour marathon barrier: by improving submax performance to under 4:50 minutes per mile.)

For generations, marathoners and their coaches have pondered predictions of how fast athletes might run. Early estimations were notably based on instinct and intuition, “gut” feelings, and the many evolving formulas, all of which I experimented with early in my career with inconsistent results. Regardless of the prediction, runners are either thrilled to achieve success or disappointed by the results. It’s not much different today: Sometimes a well-run, properly-paced race leads some finishers to claim it seemed too easy — that they could have had a better time, had they only run faster earlier in the race. However, due to the body’s physiological setup, that supposition is unlikely to be accurate. Later in the text, we discuss the reasons why.

Laboratory Testing

Since the 1970s, scientists have been trying to accurately predict marathon times in runners of all levels of talent. Methods of prediction are included but not limited to:

  • The correlation of marathon performance with maximal oxygen consumption (VO2max) is a long-standing estimate, but not a very good one.
  • A rather complex formula that includes the oxygen cost of running, VO2max, and from the largest fraction of VO2max that can be sustained throughout the race.
  • Dr. Michael Joyner, who authored the first published scientific paper addressing the possibility of a sub-two-hour marathon in 1991, used an equation including VO2max, lactate threshold percentage, and running efficiency to predict marathon time. Joyner estimated an elite athlete could theoretically run the marathon in 1:57:58 based on this formula.
  • The maximal lactate steady state (MLSS) has also been used to predict marathon times. It is defined as the workload (MLSSw) at the highest blood lactate concentration that can be maintained over time without a continual blood lactate accumulation. Marathon average paces are just below this level.

For most runners, collecting this and other data required for marathon prediction requires lengthy evaluations in a laboratory facility with proper equipment and protocols. For progress to be monitored across time, it is necessary to perform regular testing. As a result, most runners — even elite athletes — do not utilize this approach due to these and other factors such as cost, availability, and often, inconvenience. As important as laboratory testing is, it may not be necessary: an easy and accurate test can be performed by anyone with a heart-rate monitor.

Submax Field Testing

A common feature of all sports lasting more than a few minutes is that higher aerobic submax capacity results in higher competitive performance. In the marathon, race paces are usually only seconds faster than submax training paces in runners of all abilities. This means that the faster one can run while maintaining a lower-intensity submax HR, the faster the race pace. (This phenomenon is applicable to all endurance sports.)

The Aerobic System

The aerobic system is the collection of various systems and processes that intake, transport, and utilize oxygen, in particular, to oxidize fat for fuel. Fat provides the body with a stable supply of long-term energy that complements glucose utilization and reduces training and racing fatigue while conserving glycogen.

In the marathon, about 99 percent of race energy is provided by the aerobic system. Unlike shorter endurance events such as the 5k, where intensity is closer to one’s VO2max, marathoners perform at lower intensities such as 80-85 percent of VO2max. The intensity at which the marathon is run must remain relatively low. As exercise intensity increases, the percentage of energy provided by sugar increases, while the energy provided by fat decreases. The body’s sugar stores are far too small to provide energy for the duration of such a long race. To the degree that an athlete relies too much on sugar as a primary source of fuel (due to an untrained or dysfunctional submax aerobic system), running a marathon will be an extremely stressful, challenging — and slower — endeavor.

This means that by developing maximum aerobic function — MAF — one can optimize both submax speed and race performance. In addition, studies show that submax tests are the best predictors of endurance performance in runners (and for other endurance athletes, such as cyclists and race walkers, as well as for untrained people). Measuring submax performance can be accomplished regularly through a simple field test using a heart-rate monitor on a reasonably flat running course, such as a track.

MAF HR and Test

The MAF HR is a submax intensity useful for both training and submax testing. It corresponds closely with physiological laboratory measures, including:

  • Aerobic threshold (Aer T).
  • Onset of Blood Lactate Accumulation (OBLA).
  • Fatmax (the highest level of fat oxidation, which occurs during submax activity).

The MAF Test is a submax evaluation that measures pace at a given HR. For example, if an athlete can run one mile in 8 minutes while maintaining 140 HR, the MAF Test result is 8 minutes per mile. (Anyone can perform an MAF Test in his or her particular sport.)

Both the MAF HR and MAF Test were developed by the author in the early 1980s and are described in detail on my website. Monthly, measurable improvement in MAF Test scores (running faster at the same HR) is the most important measure of increasing health and fitness in an athlete. These improvements should also correspond to improving performances (even in shorter endurance races). Figure 1 is an example of a runner’s 18-month progress of MAF Tests with three corresponding marathons.

FIGURE 1. A graphical dipiction of an athlete’s 18-month MAF Test first mile progress with results of three marathons (average pace). Marathon dates are aligned with their closest MAF Test.

Time-Tested Results

Examples of elite runners comparing first mile MAF Test times and marathon paces:

  • In the early 1980s when I put a heart monitor on Norwegian Grete Waitz, who would become a nine-time winner of the New York Marathon, she ran a 6:05 aerobic pace, which corresponded to her then 2:32 New York marathon — averaging 5:48 pace.
  • A few years later, England’s Priscilla Welsh developed her MAF HR pace to 6:00 minutes per mile, and ran a 2:30 marathon averaging 5:44 pace.
  • Not long after his 2011 Boston Marathon 2:04:58 finish, a 4:46 pace, American Ryan Hall clocked a 5:07 MAF Test mile at altitude, estimated to be sub-5-minutes at sea level, which would correspond to his Boston finishing time.

Clinical observations by the author since the early 1980s demonstrated that in a healthy athlete running a typical 26.2-mile course (without significant changes in elevation, closer to sea level, and without excess weather stress such as higher temperatures or humidity, or increased winds), most could average about 15 seconds per mile faster than their MAF Test pace (within a range of 10 above and 10 below on average). This applied to age-group runners as well as elite marathoners.

Additional data was recently collected to assess the relationship between MAF Test and marathon race pace. The MAF Tests from seven female and 10 male runners of varying performance levels were analyzed. Results demonstrated that average marathon paces ranged from -17 sec/mi to +1 sec/mi, relative to 1st mile MAF paces, with a mean time of 4 seconds. This corresponds well with past clinical observations. It is possible that the slightly faster marathon pace relative to MAF Test times observed by the author is due in part to:

  • Differences in marathon courses (elevation change) and weather.
  • The author’s use of physical therapies such as biofeedback to improve the athlete’s muscle balance and gait immediately before races (in most but not all cases).
  • Well-defined dietary recommendations (particularly no refined carbohydrates and lower overall carbohydrate intake).
  • Combinations of these or other factors.

The Value of Prediction

Why is it important to predict marathon race times? Accurate prediction has real benefits for an individual athlete, including providing an important pacing strategy, as well as helping assess the balance of health and fitness.

Pacing Strategy

Lambert et al. (2004) defined pacing as the subjective competitive strategy in which an individual manipulates speed to achieve his or her performance goal. Pacing can help reduce fatigue and improve performance. Maintaining a consistent marathon pace throughout the race has been shown to be an effective performance strategy. (Pacing can be performed by the individual or with the help of another runner.)

Naturally, pacing strategy must consider the details of each particular course regarding elevation changes. For example, one would avoid running an average pace for a first mile that is significantly uphill; likewise, during a downhill mile pace may be faster than average.

By using the 1st mile of the MAF Test to predict average marathon race pace, one could create an optimal pacing strategy. This could help runners reduce their pace variability throughout the race. A lower pace variability than that observed in age-group runners is a hallmark of the elite runner. Even during championship marathons run exclusively by elite runners, top finishers showed a more even pace pattern than the less successful contenders.

Psychological and Physiological Factors

Adoption of optimal pacing strategies in a marathon is of such great importance that it could be said that the race is won in the first 5k rather than the last. In describing athletes who adopt effective pacing strategies St. Clair Gibson and Renfree (2013) write that runners who employ more even pacing throughout a race “will be able to record faster times and finish ahead of athletes with superior physiological capacities who paced themselves less effectively.”

Most coaches, clinicians and athletes have known for decades that fast starts in most events from about 800 meters and longer can impair overall performance (similar situations exist in cycling, triathlon and others). Yet, most athletes run initial miles too fast for their ability, and relative to their personal best times, and must therefore slow down too much later in the race resulting in poor performance.

Subjective factors, especially those of a psychological nature, can interfere with a runner’s ability to avoid faster, early paces. Whether in the lead or at the back of the pack, marathoners are more likely to follow other runners in the initial stages of the race and run too fast rather than follow their own perceived abilities. Sometimes referred to as “herd mentality,” this is a social phenomenon seen not only in marathon runners but in other sports (and across all levels of society). Denes-Raj and Epstein (1994) describe this as “conflict between intuitive and rational processing: when people behave against their better judgment.”

Pacing strategy is a decision-making issue which occurs long before the race begins. With an objective pace strategy plan based on submax testing, runners can follow their own inherent abilities — racing “within themselves” — rather than that of others. The result can be:

  • What usually seems like a “too easy” first 5K.
  • A “negative split” — a natural faster second half of the race — which is also associated with increased race success.
  • The ability to run faster at the end.
  • Performing at, near or above a personal best.

In the author’s experience, pacing success appears to work best in healthy runners, including those who oxidize higher amounts of fats for race energy and those with better running economy.

Health and fitness

The close relationship of the MAF Test to a person’s marathon pace can answer an important question about the balance of health and fitness. Average marathon race paces that are much slower or much faster than the relative MAF Test may be associated with a physiological imbalance:

  • Too slow a race may be indicative of less than optimal fat oxidation with reduced long-term energy to maintain a fast pace, or an irregular gait (often due to neuromuscular imbalance) reducing running economy.
  • Too fast a race could be an artificially inflated pace, commonly seen in the early stage of the overtraining syndrome where excess sympathetic tone creates artificial strength and speed.

The first indicators of worsening health and fitness may be observed in training as reductions in MAF Test speed (which sometimes initially appears as a lengthy plateau in monthly MAF Tests), with the potential of predicting various physical, biochemical or mental-emotional injuries. This often occurs even before the onset of pain, fatigue, mood changes, or other symptoms of poor recovery and physiological breakdown. Preventing this very common injury pattern may be Galen Rupp’s biggest challenge.

Considering that Rupp’s MAF HR may be 150, with the possibility that he can run a 4:52 MAF Test, it reflects great marathon potential. So, why didn’t he run faster than a 5-minute mile — his average pace at the Olympic Trials? Perhaps his primary goal was to qualify for the U.S. Olympic Team. In addition, his relatively “slow” 2:11 winning time could easily be attributed to race-day conditions: temperatures reached 76°F by the time he crossed the finish line. A 50°F day could have improved Rupp’s marathon time significantly — possibly by several minutes. Can he stay healthy, run his potential in Rio, break a world record on a faster course and flirt with 1:59? Time will tell.

Bibliography

Alghannam AF, Jedrzejewski D, Tweddle M. Reliability of time to exhaustion treadmill running as a measure of human endurance capacity. Int J Sports Med. 2016; 37(3):219-23.

Angus SD. Did recent world record marathon runners employ optimal pacing strategies? J Sports Sci. 2014; 32(1):31-45.

Banerjee AV. A simple model of herd behavior. Q J Econ. 1992; 107(3):797-817.

Billat VL, Sirvent P, Py G, Koralsztein JP, Mercier J. The concept of maximal lactate steady state: a bridge between biochemistry, physiology and sport science. Sports Med. 2003; 33(6):407-26.

Dantas JL, Doria C. Detection of the lactate threshold in runners: What is the ideal speed to start an incremental test? J Hum Kinet. 2015; 45:217–224.

Denes-Raj V and Epstein S. Conflict between intuitive and rational processing: when people behave against their better judgment. J Pers Soc Psychol. 1994; 66:819-829.

Di Prampero PE. The energy cost of human locomotion on land and in water. Int J Sports Med. 1986; 7:55–72.

El Helou N, Tafflet M, Berthelot G, et al. Impact of Environmental Parameters on Marathon Running Performance. PLoS ONE. 2012; 7(5):e37407.

Ely MR, Martin DE, Cheuvront SN, Montain SJ. Effect of ambient temperature on marathon pacing is dependent on runner ability. Med Sci Sports Exerc. 2008; 40(9):1675–1680.

Farrel, PE, Wilmore JH, Coyle EF, Billing JE, Costill DL. Plasma lactate accumulation and distance running performance. Med Sci Sports Exerc. 1979; 11:338–344.

Galy O, Ben Zoubir S, Hambli M, et al. Relationships between heart rate and physiological parameters of performance in top-level water polo players. Biol Sport. 2014; 31:33-38.

Hartung GH, Krock LP, Crandall CG. Prediction of maximal oxygen uptake from submaximal exercise testing in aerobically fit and nonfit men. Aviat Space Environ Med. 1993; 64(8):735-40.

Hartung GH, Blancq RJ, Lally DA, Krock LP. Estimation of aerobic capacity from submaximal cycle ergometry in women. Med Sci Sports Exerc. 1995; 27(3):452-7.

Joyner MJ. Modeling optimal marathon performance on the basis of physiological factors. J Appl Physiol. 1991; 70:683– 687.

Kolbe T, Dennis SC, Selley E, Noakes TD, and Lambert MI. The relationship between critical power and running performance. J Sports Sci. 1995; 13:265–269.

Lambert MI, Dugas JP, Kirkman MC, Mokone GG, Waldeck MR. Changes in running speeds in a 100 KM ultramarathon race. J Sports Sci Med. 2004; 3(3):167-173.

Lambert EV, St. Clair Gibson A, Noakes TD. Complex systems model of fatigue: integrative homeostatic control of peripheral physiological systems during exercise in humans. Br J Sport Med. 2005; 39(1):52-62.

Loy SF, Holland GJ, Mutton DL. Effects of stair-climbing vs run training on treadmill and track running performance. Med Sci Sports Exerc. 1993; 25(11):1275-8.

Maffetone, Philip. 1:59: The Sub-Two-Hour Marathon Is Within Reach — Here’s How It Will Go Down, and What It Can Teach All Runners about Training and Racing. New York, NY: Skyhorse Publishing; 2014.

Maffetone P. White Paper. An introduction to MAF: Maximum Aerobic Function. Independent; 2016.

Maffetone P. White Paper. MAF exercise heart rate: How it can help improve sports performance. Independent; 2016.

Meyer T, Lucía A, Earnest C, Kindermann W. A Conceptual Framework for Performance Diagnosis and Training Prescription from Submaximal Gas Exchange Parameters: Theory and Application. Int J Sports Med. 2005; 26:S38-S48.

Santos-Lozano A, Collado PS, Foster C et al. Influence of sex and level on marathon pacing strategy: Insights from the New York City race. Int J Sports Med. 2014; 35(11):933-8.

St. Clair Gibson A, Renfree A. Influence of different performance levels on pacing strategy during the women’s world championship marathon race. Int J Sports Physiol Perform.
2013; 8:279-285.

Yoshida T, Chida M, Ichioka M, Suda Y. Blood lactate parameters related to aerobic capacity and endurance performance. Eur J Appl Physiol Occup Physiol. 1987; 56:7–11.

Special thanks to Ivan Rivera for assistance in writing and editing, Hal Walter for editorial and Simon Greenland for formatting.

Join the discussion 57 Comments

  • Andy says:

    Hello

    Above you are writing that MAF HR corresponds closely with physiological laboratory measures, including:
    Aerobic threshold (Aer T), Maximal lactate steady state (MLSS) and Fatmax (the highest level of fat oxidation, which occurs during submax activity).
    I agree that AerT and Fatmax occur at similar intensities but MLSS should be far more intensive from my understanding.

    Your referenced paper “Billat VL, Sirvent P, Py G, Koralsztein JP, Mercier J. The concept of maximal lactate steady state: a bridge between biochemistry, physiology and sport science.” states that MLSS occurs at intensities around 4 mmol Lactate and the time to exhaustion at this intensity is about one hour (faster than marathon pace).

    Could you please elaborate further.

    Best Regards
    Andy

  • Mike Smith says:

    So is there a handy formula we can use to predict marathon race pace based on the MAF test pace?

  • Mike Smith says:

    I ran three marathons in the past three years before breaking down. As a result, I missed my second Boston Marathon last year. That’s when I realized I had been overtraining with weekly intervals, tempo runs, and long runs over this duration. Even my easy runs were not easy enough. I’m now 64. After taking a couple of months off from running, I began the MAF method in January of this year. After 6 months, my MAF tests have improved from over 13 minutes to 10:53, a far cry from my pace for the three marathons, which ranged from 8:03 to 8:11.

    If I subtract 15 seconds from my MAF test pace, my goal marathon pace (10:37) doesn’t come close to my old marathon times (sub 8:10). Is it reasonable to expect continued improvement in my MAF pace tests after 6 months of MAF training?

    • Mike:

      Yes, it is. Typically, you can expect improvements of ~15 seconds per mile per month for about a year or two. It’s entirely reasonable for your MAF pace to come down to 10:00 or below in another few months.

    • Adam says:

      I got my maff test down to 5.50mile pace and I can tell u I could not run a marathon at 5.35mile that’s for sure. My pb o’s 2.50. There has to be something more to it then what the article claims.

      • Stuart says:

        Hi Adam,

        Not an expert but I’m sure I read somewhere that the quicker you get (like your fast pac!) the smaller the gap between MAF and MP. But at the slower pacers it rings more true.

        Experts may confirm!

        Thanks

        Stu

        • Stuart:

          Yes, that’s true. What likely happens is that Adam has a lot of aerobic power (given his speed) but not a lot of aerobic endurance, meaning that he can’t really extend that speed for 26 miles. MAF-15sec works well as a predictor provided that someone’s aerobic system is conditioned to run a marathon distance.

          • Sergio Davila says:

            Hi Ivan, interesting and useful topic. Question: since the performance prediction would work well for someone whose aerobic system is conditioned for the distance, what would be a good or valid predictor for that conditioning? How could I assess during training if my own aerobic system’s condition is good enough to sustain the MAF-15sec predictor for the whole marathon distance?

          • Sergio:

            Thanks for your question. It’s very good. Usually, when you look at someone’s MAF tests, you’ll see that their MAF speed trends downward shallowly, and then suddenly takes a sharp downward turn where it drops by a minute or more within a mile or two, to stabilize at a much slower speed. We don’t have a lot of hard data on this, but what we’ve been using is that this downward turn should occur within a mile or 2 of marathon distance. After you have some experience with MAF tests, you can kind of intuit where this downward turn will happen just by looking at a 5-mile MAF test. Usually, when someone’s speed doesn’t drop substantively between mile 1 and 5, you can expect it to drop very little between miles 5 and 10, 10 and 15, and so on. It’s the people whose speed drops 15 or 10 seconds (or more) between miles 1 and 2, and so on, that you can expect to be walking at mile 15.

  • Anthony Marsh says:

    Could you please clarify what you mean by “first mile MAF test”? Is this the first mile you run straight off the bat or after the 2 mile warm up that is prescribed on the full MAF test?
    Why would you be only using the first mile as the gauge of progression rather than the full test?
    Thanks
    Anthony

    • Anthony:

      The speed during the warm-up doesn’t factor in to our measurement of aerobic development. The only miles we test during the MAF Test are after our test begins—after the warm-up has concluded. (In other words, the warm-up miles aren’t “tested” miles). So, when we say 1st mile MAF Test, we are referring to the first mile we run once our warm-up has concluded and the test has actually begun.

      We wouldn’t be using 1st mile as gauge of progression. We are using our 1st mile (-15 sec) as our marathon pace predictor. (This is an important distinction). Our gauge of aerobic progression is comparison between averages of multiple MAF tests, and the slope between the initial and final miles of our MAF test. When 2 tests have the same 1st mile MAF, but the latter one has shallower slope, it indicates that aerobic endurance, but not aerobic power, has improved. If you did have greater aerobic power, your 1st mile MAF would be faster (which would mean that your marathon speed as a whole would be higher).

      In other words, we use the first first mile of a MAF test (-15 sec) and not the full test because the first mile is a better predictor of marathon pace. Phil has found that a per mile pace 15 seconds faster than the 1st mile of the MAF test correlates far better with marathon speed than any other variable (MAF Test average speed; MAF Test slope) put through any other set of operations. The reasons the 1st mile MAF (-15 sec) works partly has to do with how many miles it takes to deplete muscle glycogen at an intensity that is 15 sec/mile faster than fully aerobic (~26 miles). One of the reasons this number (-15 sec) doesn’t change depending on MAF test times is because a 15 sec/mile is 5% faster than a (world elite) 5 minute MAF mile, but an 3.6% faster than a (blazing) 7 minute MAF mile. So, the extremely fit world-record athlete gets to run a marathon at 105% of their MAF pace, while the very fit ~2:45 marathoner gets to run at 103.6% of their MAF pace.

      • Anthony Marsh says:

        Ivan,

        Great explanation, thanks so much. It all makes sense now.
        I don’t suppose any studies have been done to predict half marathon pace based on 1st mile MAF pace?
        Regards

        Anthony

        • We’ve done a bunch of data collection and correlations. That’s how everyone comes up with number-based formulas. All you have to do is get a bunch of runners, look at a bunch of variables (in this case we only looked at MAF-test related variables), and see how they correlate with marathon speed. Of course, why the correlation works is pretty transparent: the marathon is 99% aerobic. For that reason, marathon pace was always going to fall just above your aerobic threshold speed (MAF) when muscle glycogen is full (at the 1st mile). The only real question is whether marathon pace was 10, 20, 30 seconds faster than MAF pace.

    • Marco says:

      Ivan:

      This is very interesting. But how do you increase aerobic endurance in addition to aerobic power? Longer MAF runs perharps?

      • Marco:

        Yes. How much so and when is untested from a laboratory/formal research standpoint. But what some athletes do (and we here at MAF have started to use) is to do very long MAF Tests, to find the point where MAF speed starts dropping more quickly. For someone who runs a marathon, this point should occur at around mile 15-20. So what you’ll find is that MAF pace slows relatively steadily (for an experienced marathoner, probably around 0-10 seconds per mile), until suddenly it starts slowing much faster (15-30 seconds per mile). I’m not sure it’s completely generalizable to all people, meaning that I’m not sure that everyone experienced a marked drop-off point (hence the need for more formal research). So, while in my experience this drop-off is very common, that is my personal coaching experience (plus Dr. Maffetone’s and a few others) so it may or may not fit your case.

        That point marks the aerobic system’s real endurance limit. It’s not its absolute physiological limit, or even close (for the simple reason that you can run far beyond it, and then walk much, much further), but it does mark how much that aerobic system wants to go. So that’s the best indicator I know of what your long run should be in terms of your body’s physiological capabilities (rather than your race plans).

        How I would go about training endurance is to find that point (or if you can’t find it, find a point where your speed has dropped so much that it’s very difficult to maintain a cadence of 160-180 spm) and try to reach it about once a week to once every 10 days. So that’ll essentially be your long run. For your other runs, I’d go 60-80% of that distance except when it means that you’ll be running more than 90 minutes. So, unless you’re an elite athlete and are monitoring a lot of other metrics, don’t run more than 90 minutes habitually, even if it means that you’re doing 40-50% of that mileage. For example, after aerobic base training, my drop-off tends to happen around mile 19. My pace goes from about 8:15 to 8:45 from miles 1-19, and down to 9:45 by mile 22. Go figure. So I’m just not going to run 14 miles a day, because I simply don’t have the time for it, or the energy. I habitually run 6-8 miles a day for 5 days with a long run of 13-20 miles (I hit that drop-off every 2 weeks), and 1 day of rest.
        Does this make sense?

  • Michelle says:

    What are your thoughts on nasal breathing on health and fitness levels?

  • Philip Mooney says:

    Hi Ivan,

    Figure 1 is interesting. What are your thoughts on the large difference in MAF pace and marathon pace on the October – November ’13 data point? Wasn’t expecting each difference in MAF pace vs. marathon pace to be identical, but a difference THAT large vs. the other two data points differently grabs my attention.

    • Phillip:

      It’s hard to say. The difference between 1st mile MAF pace and pace for a perfectly flat marathon is 15 secs/mile. So I’d say it has to do with idiosyncrasies of the course itself, perhaps compounded by a few other rapidly-changing work and lifestyle variables.

  • Benedict Dugger says:

    Ivan,

    Really appreciate the sharing of this formula and I am looking forward to testing it on myself and my running clients.

    Would you or Phil be willing to share or speculate on any related formulas or predictions with regards to ultra marathons, specifically 50k’s, 50m’s, 100k’s and 100m’s?

    It would seem that sustainable pace levels for those distances should even closer resemble submax aerobic HR levels than (road) marathons and the need for efficient fat burning would be even more important.

    I’m also interested in the variables of elevation change (since the great majority of (non-road/track or loop) ultras involve elevation change levels that have significant impact on race pace), and “fatigue” factors (e.g. increased bio mechanical failure, less efficient energy regulation, sleep deprivation) which for many runners can start to display diminishing performance returns in 12+ hour 50, 62, and 100 mile races. Be great to get more predictability on these factors and variables as well (in addition to heat and altitude, other frequent pace influence factors of ultra marathons)

    Love to get your thoughts on this! Thanks.

    • Benedict:

      Thanks for your comment.

      We are coming out with more formulas, but don’t have the research to back it up yet. Typically, you can think of a road 50k as the race to run at your MAF HR.

      Ultramarathons, particularly trail ultras, are such idiosyncratic animals that it’d be very hard to be formulaic about how to run one. Guidelines on flats are relatively straightforward, but it’s another thing entirely to calculate at what pace you should run a particular .8 mile stretch of road with a 7% slope when the last stretch as 4.5 miles of gravel at 4% and they had a 3.3 mile stretch of -2% downhill at rocky trail. Very, very hard to know.

      • Benedict Dugger says:

        Ivan,

        I totally agree; that’s why I brought it up in the first place.

        I think it would be very interesting to look further into it, especially because these types of events in some ways are the most ideal endurance events given that the length of these races even more so than marathons dictates the need for lower intensity aerobic effort (or risk bonking or blow up). This makes the need for a MAF type method high relevant.

        Through my coaching business and being quite well connected in the ultra running community I’d be more than happy to contribute whatever I can to learning more about how a MAF method could be applied to and benefit thousands of regular folks (as well as elite athletes) who now, for all intents of purposes, primarily over rely on too much anaerobic training without a proper aerobic base. I have access to hundreds of ultra runners who may be interested in doing tests or a study on this.

        With regard to the challenges of trail ultras, because of the added variables and increased complexity probably few have tried to develop a results predictable and reproducible formula for it; however, that’s exactly why I think it’s such a great opportunity. Someone will, sooner or later. GPS course info could relatively easily be translated into grade and altitude variables, and even terrain surfaces could be calculated in terms of increased technical trail deviations from the mean (for specified trail segments). I have some more thoughts about it as well, but without getting bogged down by minutiae, I think the MAF approach is the best one I have seen so far that could be applied towards these types of events.

        Doesn’t this sound like a fun, beneficial and meaningful project to take on? 🙂

        thanks, Benedict

      • Ivan, Benedict,
        The ideal effort measure that would handle gradient, head-wind, and trail conditions would be a power measure (as in cycling power-measurement). However, power in running is currently unavailable (aside from one product which appears to be using an algorithm on HR + gradient). Eventually a power-meter in the shoe (the contact point for transfer of power) will likely be developed.
        Until that time, HR is a good proxy (perhaps RPE is even better?). In this realm, for ultras, it is advisable to try and work out the average HR you think you can sustain for a given time-on-course. Then run to this HR (from the start) +/- 5bpm or so. This will indicate a slowing on the hills, or in head-wind, or when the trail is very loose, and a speeding up in the reverse situations. You’ll most likely find the competition go harder than you are in the early 10-30kms, but by the last 10-30km of an ultra, you will find your pace is higher and experience the positive boost of passing people. This is simply applying the ‘balanced exertion principle’ alluded to in the main article. In my experience this approach is very effective, with a pre-determined allowance for some higher HRs on particularly hard gradients, but within limits.
        What is not yet very well understood is how the HR responds to excessive time on course, and other more complex factors such as mental fatigue or even down-recruitment through neurological channels (ala Noakes et al’s central govenor). But for the moment, using a calculation as discussed here for HR and trying to stick with this (rather than pace-watching) is a strong candidate for how to run complicated gradient/trail ultras.
        Simon.
        (Ultra runner and author of Angus (2014) in the Bibliography)

        • Dr. Simon:

          Thank you very much for your comment.

          We have a bit of a problem with power measurements as they pertain to endurance training and endurance events. We certainly do not discount them entirely—they are extremely useful measures—but we do prefer the heart rate as our primary measure of intensity. The main reason we prefer the heart rate is the same reason a lot of people treat it as a confound (because of its response to excessive time on course, mental fatigue, and down-recruitment, etc.). The parenthesized factors, and many others, are implicated in the sum total of physiological stress the body experiences as a result of the event. Particularly in relation to endurance events, we are interested in this sum total of stress more than we are in power.

          Let me use a couple of examples to explain why. Take 2 different runs. Both runs have the same mileage at the same “low-intensity” power output (say, 45% of maximum). However, run 1 was run in the afternoon, after a very easy work day, in balmy weather, while run 2 was run after a stressful work day, and while managing visiting relatives at home. The heart rate in the second run was much higher. However, because of the stressful nature of the second run, it is far more likely that the body, in its stressed state, reacts from run 2 as if it were an “adverse” experience. Importantly, the elevated heart rate manifest during the duration of this run has a very strong correspondence with the body’s gestalt perception of adversity.

          For these reasons, the ability for the body to improve and grow from run 1 is not equivalent to run 2. Similarly, the onset of recovery after “race” 1 would come much more quickly than after “race” 2 (runs of an identical “high” power output). So, while these runs may both occur at the same power, they set the body on different paths towards diverging athletic futures. This is particularly true of endurance training and racing, where the level of stress that the body finds itself in (as reflected by the heart rate) is quite protracted and becomes increasingly impactful to the body’s well-being as it increases in duration.

          Incidentally, I recognize at least one implication of your point about excessive time on course: sympathetic fatigue might suppress the heart rate, giving readings that overestimate physiological thresholds of intensity. So in that sense, heart rate is not an absolute measure that works for every situation, and it is imperfect even in those (many) situations where it does work. But in the attempt of capturing the body’s gestalt experience of the training or racing event (within reasonable bounds), it works quite well.

          Riffing off of Noakes’ theory of predictive governance, I believe that RPE becomes an increasingly good measure—particularly for racing—commensurate to the experience that the athlete has with this distance or distances like it.

          Thanks again for your comment, and I’d love to hear your thoughts on my response.

  • Bruce says:

    Interesting article. I’m 51 and planning to do my first marathon. A hilly one in Co. Kerry, Ireland… a splendid and beautiful place not too far from where I live. Happy to accept that it may be a mid-life crisis thing but it really has given me a fresh purpose. I did go to the trouble and expense of getting some lab testing done to determine the pace at which my lactate levels started to climb. Also had my VO2max measured (52.1 ml x -kg-1 x min-1). Had my body fat measured too (18%) which is high and I attribute to bad habits (Guinness, Jameson, crisps). I’ve lost 3kg since (hills are easier). The 2 mmol/L mark seems to be the magic number from these tests for aerobic threshold – that happened for me at around a 8:25 min/mi pace. Should I equate that with my MAF and expect to pace myself at 15sec faster than that for the marathon? 8:10/mi seems quick to me but I may have completely missed what you were getting at. Note: I’ve done a few 10K’s (also in hilly terrain) most recently one in 45:30 (hilly). That’s just under 7:30/mi pace. Ave HR 182, Max 193… Thanks!

    • Bruce:

      No, the 2 mmol mark has been used to mark the anaerobic threshold, not the aerobic threshold. The aerobic threshold is the onset of blood lactate accumulation (OBLA), which is the point where lactate starts to increase above resting levels (way below 2 mmol). If you want a laboratory measure of the aerobic threshold, look for FAT MAX (heart rate at the highest rate—not percentage—of fat-burning).

      • Bruce says:

        Hi Ivan,

        I had equated OBLA with 4 mmol, i.e., the exercise intensity corresponding to 4 mmol lactate in the blood (a quick Google search confirms). That was another important number that came up in my test. OBLA is what some people call the anaerobic threshold. This terminology is all a bit confusing, but I was given to believe that up until lactate levels reached 2 mmol, exercise intensity would pretty light/easy (which was certainly the case in my test) and that once you exceed this 2 mmol threshold, intensity increases and you start using carbs (along with fat) as an energy source. I understood that after 4 mmol, you are using mostly carbs. Have I missed something?

        • Bruce:

          When doing light activity, your blood lactate levels are far less than 1 mmol. Rates of fat oxidation are inversely correlated to blood lactate, because lactate stops lipolysis (fatty tissue breakdown). So rates of fat-burning peak at the point that blood lactate levels increase above resting (OBLA), and decrease as lactate levels increase. So, by the time that you hit 2 mmol, your rates of fat-burning have already been decreasing for a while.

          This is why the MAF HR coincides with the maximum rate of fat-burning (fat max). Check out our other paper on the MAF HR, and the sources cited within.

          • Shey Doane says:

            Hi Ivan,
            On the issue of fat burning peaking where blood lactate levels increase above resting OBLA…I would love to know your thoughts on the FASTER study by Dr Volek, where it was concluded that Peak fat oxidation occurred at a much higher % VO2max in Fat Optimized athletes (aka “LC” subjects) than HiCarb athletes. Other studies have also stipulated that the body can do many things with lactate depending on training (ie, lactate can also be converted into energy for muscles to utilize). Just not sure that the OBLA is always the point at which you’ve crossed over the anaerobic threshold.

          • Shey:

            OBLA is the point where you cross over the aerobic, not anaerobic threshold. It is true that peak fat oxidation occurs at a much higher percentage of Vo2 Max (relative VO2) in fat-optimized athletes. However, those same athletes also hit very high percentages of their total VO2 max at relatively low heart rates. This is because the heart rate is primarily controlled by the same stress hormones that increase sugar oxidation and decrease fat oxidation. So, while the fat oxidation of these athletes is high and their relative VO2 is also high, all this is happening at a moderate heart rate.

            (In other words, you can’t look at a high percentage of VO2 Max and conclude that the athlete is working at a high heart rate. A lot of these athletes hit VO2 Max only barely above their anaerobic threshold).

            The reason that you hit peak fat oxidation at OBLA is because lactate is a direct inhibitor of lipolysis, and the various stress hormones associated with lactate clearance, sugar availability, etc. are all indirect inhibitors of lipolysis, AND they all counterregulate the hormones that increase lipolysis. So the body is effectively hardwired to bring down lipolysis as soon as lactate starts increasing. Why?

            Simple: mitochondria don’t actually burn “fats” OR “sugars”. They burn a molecule called Acetyl-CoA. Lactate and pyruvate (which both come from sugar) get transformed into Acetyl-CoA, and fatty acids get broken down into Acetyl-CoA. Let’s back away for a second. Your body only started producing lactate because it didn’t have enough mitochondria to aerobically burn the sugars you were using. So, if your body kept lipolysis at the same level while also producing lactate, one of 3 things would happen:

            1) you’d flood the bloodstream with free fatty acids
            2) you’d flood the bloodstream with lactate
            3) you’d flood the bloodstream with Acetyl-CoA

            So no human body out there that would benefit from increasing fat-burning while also being in anaerobic function. That’s the short of why the hormones I discussed above work the way they do. Hope this helps.

  • Michael says:

    Ivan,

    This is a really informative article. I use kilometres instead of miles where I come from, and based my MAF test on km 1-5 (after warming up). In order to predict my marathon pace, would I subtract 15 seconds from my first km MAF test time, or 9 seconds (15 seconds per mile converted to kilometres)?

  • Sebastian says:

    Hi Ivan,

    I remember your advice to run a marathon 10-15 bpm faster than MAF HR. Could you please share which prediction method (10-15bpm above MAF HR or 10-15 sec faster than MAF pace) is better? Thank you.

    • Sebastian:

      This estimate (10-15 sec faster than 1st mile MAF pace) is way better.

      My 10-15 BPM higher was my way of helping people test out their marathon race pace before I had this data.

      • Marco says:

        I just ran my first marathon in Brussels and I started in zone MAF HR + 10/15 sec (162-167 bpm). I was able to remain at my target race pace (MAF pace – 15 sec) up to mile 13. From that point I started to slow down quite a bit in this HR zone. So I had to push my HR around 172 bpm (MAF HR + 20) up to mile 20, and around 182 bpm (MAF HR + 30) up to the finish line.

        I should add that the marathon course was very hilly and quite tough (according to some seasoned marathoners).

        Also cramps started to threaten me from mile 20 up to the finish line. Luckily I never had a real one but they were around, ready to trigger. I guess it’s probably muscle overwork rather than dehydration or something else.

        Regarding nutrition, I ate a Phil’s bar 1 hour before the race, a Phil’s shake 30 min before, and a gel every hour during the race. I also drank water every time I had the opportunity.

        Overall I’m quite happy with my performance (3h59) and it was an amazing journey but I’m trying to analyze my race:
        – Is the HR increase normal?
        – I guess it is due to muscle fatigue but maybe it means I lack aerobic endurance or something?
        – Next time should I race at MAF pace – 15 sec instead of racing at MAF HR – 10/15 sec? And not worrying too much about HR?
        – What can I do to prevent cramps in the future? More long runs? Strength workouts? For the past few months I’ve been only doing aerobic running workouts a few times a week, nothing more.

        Thanks in advance for your help

        • Marco:

          – Yes, that HR increase is perfectly fine and normal for a race.
          – An increase in HR means that you are just using more and more of your physiology (since you are using anaerobic channels to maintain the same power output). This is ideal in a race since you want to produce the quickest time that your physiology can safely handle—and that is always going to mean increasing your heart rate to maintain pace. (what is NOT healthy is to treat training sessions as they are races, as your physiology cannot sustain this kind of output chronically. But sustaining it every once in a while for a race is natural and well within its capabilities).
          – Yes, more runs of steadily increasing length. Also strength workouts such as squats may help. Also, work on your form and make sure you are not overstriding.

  • Jo C says:

    Following on from Benedict’s question, do you have a predictive method for the Ironman marathon? I recently race IM Frankfurt and finished slower than I would have predicted from my MAF pace. I am assuming this was accumulated fatigue?
    A second question I have is regarding MAF on the bike? IF I hit MAF on the bike (and this is extremely difficult for me) I am pretty much at my 20km race power. Obviously this is not sustainable for any decent periods in training rides. My equivalent perceived effort to my run MAF for the bike is way, way lower. Is this due to lack of muscle development on the bike compared with the run? I have been racing Ironmans for 10 years now, and have seen no significant improvements on the bike in about the past 5 years

    • Jo:

      Yes—check out this podcast.

      Do you mean to say that perceived effort is higher on the bike than for the run at the same HR? This is because perceived effort is the effort the brain is making to power the muscles. Heart rate is the effort that the metabolism is making overall. Let me explain this with an example: suppose you want to force open a jar lid that has been stuck. Your perceived effort as you attempt to twist the lid might near 100%. However, your heart rate might not even break 90.

      This is because while your brain needs to drive a few muscles (in the hand and forearm) to a very high percentage of their maximum power output, it doesn’t really need to increase your metabolic rate to do so: it’s just a few muscles, and they’re quite small.

      Cycling is similar to this: in cycling, you’re almost only using the core and leg muscles, while in running you need a much more generalized use of the body’s musculature in order to be able to stabilize and stop from tipping over in any direction. (On the bike you have 5 points of support at all times: seat, 2 pedals, and 2 handlebars. Running you only have 1 foot on the ground for a small percentage of gait). So, in running, your body has to distribute the same metabolic power across a lot more muscles, while in biking it can pour that same metabolic energy into relatively few muscles. This means that in biking, those few muscles contract much more powerfully than in running, where a lot of muscles contract at a lower rate. In order to contract those muscles much more powerfully, the brain has to increase its effort, and that is what you perceive.

  • Stuart says:

    Hi,

    I have been trained ning MAF in preparation of my 2nd Marathon, late octobre. I’m seing improvements after being frustrated at the beginning with the slow pace and limited improvement. However, diet adjustments have seen a recent change.

    I believe I am on the right track and have a fairnunderstanding of whats involved to continue my training. However, my main issue is nutrition for marathon. I feel uneducated on what to do when for my long runs and particualarly race day. I know its trial and error and long runs are ideal to test nutrition but do you have some tips of what kind of carbs or nutrition would be suitable for race day to gain optimum performance combined with OFM?

    Thanks!

    Stu

    • Stuart:

      Have a small, balanced breakfast 2-3 hours before the race, with a relatively small amount of low-glycemic carbs, some fats, and some protein. You want to keep it light and easily-digestible. Think of it as being about eliminating hunger, rather than about getting full. For during the race, don’t start consuming anything until 30-45 minutes after the onset of activity (which includes warm-up). After 30 minutes, the aerobic system is revved up, which means that you can start taking high-glycemic carbs without dampening fat-burning. I usually recommend people starting by testing out 120 calories/hour (more or less 1 race goo), and scale down and up from there.

  • Mike C says:

    This is very interesting but I wonder about the effects of nutrition and “carb dependency” or “keto-adaptation” on formulas such as this. Unless I am misreading it seems like the pace – 15 seconds is roughly the pace it would take run out of glycogen stores at 26.2 miles. One of the goals of become more fat-adapted (or even going into ketosis) is to burn a greater percentage of fat at the given paces. Therefore, if we have 2 runners with the same MAF pace, one who is fat-adapted and one who is not, then the runner who is fat-adapted should still have greater glycogen stores left at the end of a marathon while running at the same pace – 15 seconds. Would it therefore be advisable to adjust this pacing for the different athletes?

    • Mike:

      You don’t adjust it. Or put another way, you don’t adjust it unless you have laboratory evidence that says otherwise. That’s because the more fat-adapted runner would be burning a greater percentage of fats at any given time, but he would also be burning energy OVERALL at a much higher rate, which means that the more fat adapted runner will be burning as much sugar, because that highly trained runner is exceeding their MAF HR by a similar amount. This means that the same amount of sugar is equivalent to a smaller percentage of overall fuel consumption. Here’s a great research article with corroborating evidence.

      Think about it this way: if running 15 seconds faster than MAF means that in some time period X, you are burning 1 unit of sugar and 9 units of fat (10% sugar and 90% fat), becoming more fat adapted you can be burning 1 unit of sugar and 14 units of fat (7% sugar and 93% fat). You are still running 15 seconds faster than your MAF, but your MAF might be 45 seconds faster than before. So, in our little thought experiment, that 1 unit of sugar is what produces that 15 sec faster pace.

      Now let’s apply it to the real world.

      A 5 minute MAF miler can run MAF at 137% the speed of a 7 minute MAF miler (12 mph to 8.6 mph). By running 15 seconds/mile above their MAF HR, the 7 minute miler is running at 101% of their MAF speed. Conversely, a 5 minute MAF miler is running at 105% of their MAF speed. (A 15 second difference turns 12 MPH into 12.6 MPH, but turns 8.6 MPH into only 8.9 MPH. For comparison, 15 seconds turns 60 MPH into 75 MPH.)

      So there’s another phenomenon here, which makes the real world different from the thought experiment above: to account for this speed increase, a runner running that marathon world-record 4:45 pace of 12.6 MPH—World-record holder Dennis Kimetto averaged 4:47 in Berlin—would be burning sugar at a greater rate than that 7 minute MAF miler, and their rate of fat-burning would be high enough that the rate at which they burn sugar would still mean that sugar burning accounts for a smaller percentage of total fuel consumption. In other words, the 7 minute MAF miler might be burning 1 unit of sugar and 9 units of fat (10% sugar and 90% fat) but the 5 minute miler might be burning 1.5 units of sugar and 18.5 units of fat (7.5% sugar and 92.5% fat). Even though it accounts for a smaller percentage, this theoretical world-record holder is burning sugar at 1 1/2 times the rate of the 7 minute MAF miler.

      So, for a faster MAF runner to run 15 seconds faster than their MAF HR, their sugar-burning systems, which provide increasing energy above their MAF HR, have to provide energy at a greater rate than for a slower MAF runner to run 15 seconds faster than their MAF HR. Faster MAF runners also have livers with a greater ability to provide glucose, which means that it all evens out: the faster MAF runner ends up depleting their glycogen supply at a similar rate than a slower MAF runner, when both are running 15 seconds faster than their MAF. We’ve got a lot of observational data to corroborate this. In conclusion, adding 15 seconds to MAF time is a calculation that seems to work for a vast majority of runners across a huge range of speeds.

  • Ryan Spearman says:

    A couple of questions- firstly, does the prediction formula work for minutes per km just the same?

    Also, how does this formula relate to half marathons? I am guessing I could keep my heart rate well above my MAF for the duration of a half marathon, how do I figure out what that rate is before the race?

    Finally, you talk about pace strategies throughout the race, are you essentially saying figure out what the pace is according to the test then stick to it for the whole race, or is there a more complex technique?

    • Ryan:

      Yes, it does. Just do the appropriate conversions.

      For a half marathon, you’d look for your pace to be around 30 seconds faster than your 1st mile MAF.

      There’s a bunch of different techniques, but the simplest strategy is to figure out a pace and stick to it for the whole race. Figuring out a pace is also the first step towards using a more complex strategy, such as negative splits.

      • Ryan Spearman says:

        That seems like it would be quite slow- my current MAF speed is 8:13 per km (13:11 min/mile) meaning the 1/2 marathon would take 2:53!!
        Before I started MAF training I was doing between 10k and 14k always with less the 6 min/km, sometimes as low as 5:30.

        Could I still run the marathon at that pace?

        • Ryan:

          The issue with that is this: it’s the aerobic system that runs the marathon. That means that you probably won’t be able to run a marathon much faster than MAF-15 seconds because doing so would mean too much recruitment of your anaerobic system. Typically, that means you’ll hit the wall, have to slow down anyway, and miss your target pace by a larger margin. But it’s also possible to force yourself to recruit the anaerobic system for longer periods of time by putting your head down and hammering away (and downing caffeine and sugar chews). That’s a bad idea. Please don’t do it. You might manage to hit your target pace, but activating your anaerobic system that way means that you ran a marathon far beyond your physiological capabilities. It’ll show post-race: you’ll be tight in strange places, you’ll be systemically inflamed, you’ll probably come down with something respiratory, etc.

          So the MAF-15 isn’t just about how the body can potentially run a marathon without considering that it has an athletic future ahead of it. MAF-15 (give or take a few seconds to account for individual differences) is a formula to produce a highly-successful marathon running career. At the peak of that career, the runner will be able to produce race performances much faster than the fastest race of the runner who attempted to run MAF-45 or MAF-115 and hoped to be fast that way.

  • Billy says:

    Hello.. If I do once a Month MAF 5 mile test…. What Training should I do for the Rest of the Training week or weeks ?? Tempo 88% ?? Intervals 98% ?? Fartlek 85% ?? Long Runs on 70%?? Can you be specific please ?? Thank you 🙂

    • Billy:

      It depends wholly on your training goals, training history, medical history, injury history, anthropometry, genetics, etc. In effect, we simply can’t “be specific” to what kind of training you should or shouldn’t do. We can only be specific while discussing your particular situation (whatever that may be) at the exclusion of all others. But there is no way to be specific when talking about how all (or even most) people should train. What we can do is be general: if you’re an endurance runner, for example, I can tell you that your training will probably start to become unhealthy when your high-intensity training exceeds around 25% of your total volume. What I can also tell you is how, generally, your training should eventually end up looking like if you want to be prepared for a 5k, 10k, marathon, trail ultra, etc.

      • Billy says:

        Thank you for your Reply… 🙂
        My Current MAF is 135bpm as am 50 years old… I compete on a 10km 45mn with HR of 170 BPM of max 187 & 25km 2:00 & HalfMarathon of 1:35 with 164BPM & Marathon of 3:30 with 162 BPM… & offcourse I want to get faster.
        I Hope MAF will help me run Better .. but dont know how many Slow runs & MAF’s should I do per week ?? or how many anaerobic runs to do per week !!?? I just need a small guide of a Weekly training 🙂 ..
        Can you Please help ??
        Thank you
        BV

        • Billy:

          What I would do is first figure out how many hours per week you regularly train.

          Then, introduce about 80% aerobic, 20% anaerobic training. If you usually train 6h/week, a typical marathon training week can look something like this:

          m 1h aerobic
          t 30m anaerobic (intervals: 400m sprint, 800m interval)
          w 1h aerobic
          th 1h aeobic
          f 30m anaerobic (hills)
          s 1-1.5h aerobic
          su rest

          Usually, what do (for marathons) is train 2 weeks like the above and then 2 weeks 100% aerobic (see below). I’ve found it very useful to staggering training so that (a) within the same week you have short days and long days, and (b) so that within the same month you have 2 hard weeks and 2 easy weeks. So what happens is that every month, your body has 2 weeks to get completely, absolutely recovered from the easy training weeks. This, in turn, gives you leeway to play around with the volume a lot more and change your routines on the fly. So, the first month, even if your planned training seems conservative, commit to it and see how the entire month plays out. Then take an inventory of the month, and plan out the next month based on that.

          A typical 100% aerobic week (for the same runner) would look like:

          m 30m
          t 1h
          w 30m
          t 1h
          f 30m
          s 1-1.5h
          s rest

          • Billy says:

            I cannot thank you Much 🙂 …. I will try that & maybe increase the Duration as my weekly Km is 80 -100km 🙂 .. Thank you for the Enlightenment.
            BV

  • loqueelvientoajuarez says:

    Hi,

    Thank you very much for this interesting article. I have a question, though, for my marathon pace is much, much faster than MAF pace.

    I did my first MAF test yesterday (taking formula FC = 180 – age + 5) and it yielded a pace of 5’20″/km (8:34/mi) at 145 bpm with little variations between 2000 m segments. I have screwed up a marathon six weeks ago at 4:15/km (6:50/mi) and 170 bpm (measured during training runs at 4:15 close to the race), that’s quite a huge difference. While I have lost some fitness during the post-race two-weeks break, I was not faster than 4:55/km (7:50/mi) at MAF HR (it’s the higher end of my easy pace range, so I run it and monitor it often) on terrain similar to the race.

    How come run a disappointing marathon one minute per mile faster than MAF test pace at peak fitness?

    Can I draw some conclusions from this discrepancy?

    For reference:
    8 years running, 4 competitively, 40 years.
    max HR probably ~ 200, 192/195 obtained at the end of shorter races.
    10K 38:19 HR=176-180 (intervals @ 10K), HR=183 (race)
    HM 1:26 HR=172-176 (training runs @ HMP), HR=179 (race)
    FM 2:59 HR=166-172 (training runs @ MP), HR not measured during race
    previous base training = 2 moderate runs at HR=155 (6 weeks) to 160 (6 weeks), 5 days of “commuting doubles” at HR=125-145, high carb diet. Programmes used for races: Hansons, Jack Daniels. PR’d in the marathon each year from 3:38 to 2:59.

    • loqueelvientoajuarez:

      It is quite possible that the 180-Formula underestimates your physiological MAF HR (aerobic threshold). With your speed and your improvement history, that sounds relatively likely. What that means is that your MAF HR might be closer to 150. The 180-Formula deals with population averages, and works best for people with metrics (such as VO2 max) that are within 2 standard deviations of the mean. When you suspect you may be outside of this (and it’s reasonable to suspect in your case), it’s best to take laboratory measures such as aerobic threshold, VT1, and FAT MAX. As stated above, your “real” MAF HR is the aerobic threshold/FAT MAX, and your VT1 serves as a confirming variable: VT1 occurs a few BPM above the aerobic threshold.

      Another possibility (that I think is less likely in your case) is that you run your marathons at anaerobic speeds (which are unhealthy at that distance), meaning that your aerobic system should grow into that speed. That means that you would keep that speed or improve it mildly, but while running at a slightly lower heart rate.

  • loqueelvientoajuarezr says:

    Thanks for your insight Ivan.

  • Mark Reading says:

    Great thread, lots of good information.

    I recently undertook lactate and VO2 testing. My LT1 (2 nmol) and LT2 ( 3.5 nmol) are 147 bpm and 162 bpm respectively. Yet my MAF is 135 bpm (45 year old runner, back training for 18 months) and my marathon pace is working out at 7:20 with a target HR of 147-153 bpm. But my MAF pace is averaging more like 7:55.

    So, do I go at 7:20, or revise down and add 10 minutes to my marathon time? 😉

    • Mark:

      The 180 formula is never as accurate as laboratory tests. The MAF HR is usually a few BPM below your LT1. So it’s likely that your real physiological MAF HR (which corresponds with the maximum rate of fat-burning) is at around 142-145 BPM, which might put your MAF pace at around 7:45 – 7:35.

      Subtract 15 seconds from that and it seems to me like your marathon pace (7:20) corresponds very well with our estimates.

Leave a Reply