Training

Why Increasing Weekly Mileage Improves Performance

By Richard Gibbens

Introduction

One of the most dominant beliefs of conventional training wisdom is that increasing weekly mileage results in improved performance.  Numerous runners have anecdotally reported that when they increased their weekly run mileage that their performance improved, sometimes significantly.  Additionally, it is commonly believed that the fastest distance runners in the world, the elites of the sport, generally all train with relatively high weekly mileages.  It is not unusual for elite runners to report running 100 or more miles per week.  Finally research studies consisting of runner surveys generally find that the faster runners tend to run higher weekly mileages than do slower runners, which has generally been interpreted to mean that the higher mileage produces the faster runners.  These observations - personal experiences, anecdotal data of elite training mileages, and research surveys - have been sufficient to convince many runners that increasing mileage not only results in improved performance but that it is the best way to do so.

However, when scientists have conducted controlled research studies examining the influence of training volume on endurance performance the results have often contradicted conventional training wisdom.  Multiple studies have shown that weekly mileage exerts little influence on performance and that other training variables exert much more influence on performance than does volume of training.  There is little controlled research supporting the belief that volume of training is an important performance enhancer.

Why is there such a discrepancy between the anecdotal observations of the influence of weekly mileage and the research studies on the same topic?  Why do the controlled research studies find that volume of training plays a relatively minor role in performance while in everyday life many runners swear by the effects of increasing weekly mileage?  Are all the research studies wrong?  Or are the anecdotal observations of runners in error?

I believe that both groups are right.  Increasing weekly mileage can absolutely result in measurably improved performance.  And the research studies that find no improvement due to increases weekly mileage are also correct.  How can this be?  The two appear to contradict each other.  In this article we will examine the apparent contradiction to see if we can discover a) if increasing mileage can produce enhanced performance and, if it does, b) how it does so.

Part 1:  Research

Increasing weekly mileage

Consider for a moment the following scenario.  A runner says to you that he increased his weekly mileage from the 40 miles per week (mpw) to 70 mpw and his performance improved.  Statements like this are not uncommon.  When questioned, this same runner may also say that he tried higher intensity training prior to increasing mileage but that his performance had plateaued with high intensity training and it was only when he switched to higher volume that improvement began to occur.

Let’s assume that the runner did improve when he upped his mileage to 70 mpw.  It is very tempting to say that simply increasing mileage from 40 to 70 mpw caused the improvement.  It’s easy to see how a runner could say “I ran more and I got better” and thus attribute the improvements to the increased weekly mileage. 

However, there is more to the story than simply saying “I increased my weekly mileage and got better.”  What specifically did the runner do to increase weekly mileage?  Did he run more frequently?  Did he run the same frequency while increasing the distance he ran?  Did he combine the two – running more frequently for further distances?  Did he dramatically increase the distance of his longest run?  Did his training pace change?  Did he train more intensely?  Did he increase the volume of intensity?

The point is that some combination of the above must occur when an increase in weekly mileage occurs.  Simply saying that “I ran higher mileage and I got better” doesn’t provide us with answers to any of our questions.  Why did a higher weekly mileage make you a better runner?  That’s the question we want to answer.  Was it really the increase in weekly mileage that made you better?  Could other factors be responsible for the improved performance while mileage gets all the credit?  Simple math tells us that increasing weekly mileage means either running more frequently or for a longer duration or a combination of the two.  Is it possible that either one of these factors (running more frequently or running further distances each time you run) caused the improved performance irrespective of changes in weekly mileage?  Furthermore, saying that you ran more mileage doesn’t provide any indication as to the intensity of any of your runs.  Did you run at a higher level of intensity or an increased volume of intensity?  I realize that referring to weekly mileage is often a shorthand way of communicating training information but to simply lump all of these training variables together without distinguishing the effect any one individual variable has on training is not very illuminating.  Furthermore, if all you understand is that to improve you must increase mileage, then in order to continue to improve you will be forced to continually increase weekly mileage.  What happens when you reach the point where you can’t increase your weekly mileage either due to injury, overtraining, limited time, or some other reason?  What do you do when you get to the point that increasing mileage no longer improves performance?

On the other hand, what if you were able to distinguish the effect each individual training variable or combination of variables have on performance?  For example, what if you found out that increasing the duration of your runs, without increasing frequency, accounted for 80% of the improvements you experienced?  Or what if you discovered that increasing the volume of intensity produced 90% of your improvements?  Both of these are hypothetical examples, but pretending for a moment that one of them was true, would this knowledge cause you to alter your training?  If duration played as significant of a role as in my example, would you concentrate your efforts on increasing overall mileage or would you concentrate your training on increasing the duration of your workouts?

You cannot increase weekly mileage without some change in the training variables.  We live in a cause-effect world.  Weekly mileage is not a cause; it is simply the sum of frequency and duration.  Training variables are the cause.  Performance is the effect.  Without understanding the causes you will be limited in your ability to maximize the effect.  By knowing the influence of each training variable and combination of variables you can then more precisely design your training program to place the appropriate emphasis on each training variable.  Conversely, if your primary focus is on increasing mileage, without an understanding of the variables that actually produce an increase in weekly mileage, then you are much less likely to produce the absolute best performance you are capable of.  Therefore our first task is to define and examine the training variables.

The Training Variables

What are the training variables?  The exercise science community recognizes four training variables, summarized in the acronym FITT:

F - frequency of training

I -  intensity of training

T - time or duration of training

T - type or mode of training

Frequency of training is usually expressed in terms of a week, which is how we will use it in this article.  “I worked out 4 times last week” expresses a frequency of training of 4 times per week.

Intensity is a measure of the difficulty of exercise.  In running, intensity is commonly expressed as some percentage of either VO2peak or maximum heart rate (HRmax).  “I ran at 80% HRmax” would be an expression of training intensity.  Another way of expressing intensity is known as rate of perceived exertion, or RPE.  This is simply the runner’s subjective expression of how hard they feel they are working.  In this article we will generally use %VO2peak or %HRmax as a measure of intensity.

Time simply refers to the duration of an exercise session and can be expresses in several ways.  “I ran for 30 minutes” would be an expression of training time.  Typically though, duration is expressed as a distance rather than as an amount of time that has been run.  It is much more common to say “I ran 4 miles” than it is to say “I ran 33 minutes.”  For our purposes, we will generally use distance in miles or kilometers as our measure of duration of exercise.

Type expresses the mode, or form, of training.  Running, cycling, and skiing are all different modes, or forms, of exercise.  Different forms of exercise produce different results.  For example, exercising by lifting heavy weights will produce measurably different results than will running long distances.  Our focus in this article is specific to running and we are not comparing the results of exercising achieved by running to the results of exercising in other ways.  As it relates to our purposes in this article type will not vary.  There will be cases where we will examine data from research studies that use exercises other than running, but we will not be comparing how other forms of exercise stack up against running.

These four variables determine the specifics of any and every workout. Since our focus in this article is exclusively on running, we have just three variables to examine – frequency, intensity, and duration.  And when it comes to the term “weekly mileage” only 2 of our variables apply – frequency and duration.  If you increase your weekly mileage then you have manipulated frequency and/or duration so that the result is an increase in weekly mileage.  Though you may have changed the intensity of your workouts, intensity is not factored in when calculating total weekly mileage.

Now that we have defined the 3 training variables that comprise our running workouts, our next step is to gain an understanding of the influence each of them has on performance.

Systematically Examining the Training Variables

To best determine the influence each training variable has on performance we would need to systematically examine each variable in turn.  The simplest way to do this would be to conduct a series of experiments holding two of the variables constant while modifying the third.  Any changes in performance could then reasonably be attributed to the changed variable.  Table 1 graphically illustrates this simple process of experimentation.

Table 1:  Systematically measuring the influence of the three primary training variables

In Case 1, by holding duration and frequency constant but modifying intensity, differences in performance would be a result of the changes in intensity.  In Case 2 frequency and intensity are held constant and duration is modified to determine the effect it has on performance.  In Case 3, frequency of training is modified while duration and intensity are held constant, exposing the influence training frequency has on performance.  From the data gathered we would gain an understanding of the magnitude of performance changes caused by each training variable and the relative influence each variable has.

There are research studies that fall within our matrix above, in that they hold two of the training variables constant while modifying one, and we will examine these research studies in this paper.  However, there are also studies that do not strictly follow our straightforward, systematic course, studies that examine the influence of combinations of the variables, such as the role of weekly mileage (which is just the combination of frequency and duration in a particular week).  From these studies we will learn what we can while doing our best to elicit the influence of each of the variables.

Case 1 – Measuring the influence of Intensity

This is the easiest of the three cases to examine because there is a fairly sizeable body of research on this topic and few would debate the significant influence intensity plays in performance.  Virtually all training programs today that are focused on improving performance include either regular intense training or one or more phases of high intensity training.

Hickson et al

Robert Hickson et al, in a series of 4 research studies, individually examined the influence of each of the training variables.  In the first of these studies he recruited healthy, untrained or moderately trained subjects and had them train 40 minutes per day, 6 days per week for 10 weeks (1).  For 3 days/wk the subjects cycled at VO2peak for 5 minute intervals with 2 minutes of easy cycling at 50-60% VO2peak between each interval.  On the alternate 3 days they ran as far as they could in 40 minutes.  The important thing about this study was that frequency and duration were held constant while intensity was increased relative to improvements in the subjects.  As the subjects improved the training pace (intensity) was also increased.

As a result of this training regime the subjects improved a significant and surprisingly large amount.  The researchers wrote that “the magnitude of the increase in VO2max was also larger than expected.”  Additionally the authors noted that “Rather surprisingly, average VO2max increased linearly during the entire 10 wk of training without showing a tendency to level off.”  They also note that “Endurance time and time to attainment of peak heart rate during the entire endurance test also increased linearly and were both closely correlated with VO2max.”

This study provides strong evidence of the importance intensity plays in performance.  Despite no changes in either frequency or duration, performance improved steadily and linearly for 10 weeks, all due to the influence of intensity.  In a follow up study Hickson further examined the role of intensity by again holding duration and frequency constant while reducing intensity by either 1/3 or 2/3 (2).  In this study Hickson had moderately active subjects follow a similar training program as in his first study, consisting of 40 minutes of exercise per day, 6 days per week, for 10 weeks.  After 10 weeks of training the subjects continued to train with either 1/3 or 2/3 decreases in intensity for an additional 15 weeks.  As a result of the decreased intensity VO2peak and performance declined in both groups.  Just a 1/3 decrease in intensity resulted in measurable decreases in fitness and performance.  Increasing training intensity while holding duration and frequency constant resulted in improved performance.  Decreasing the training intensity while holding duration and frequency constant caused performance to decline significantly.

In their summary of the 5 training studies the researchers stated, “When taken together and in relation to the training protocol employed, these results show that training intensity plays a principle role in regulating the maintenance of the increased aerobic performance.”

Mikesell and Dudley

A question that remained unanswered with Hickson’s series of studies was whether the results were applicable to trained subjects (recall that Hickson used only untrained or moderately training subjects).  To answer this question two researchers at Ohio University, Kevin Mikesell and Gary Dudley, decided to replicate Hickson’s first study while using trained runners as their subjects (3).  Their stated purpose was to “determine the nature of the adaptive response to progressively intense aerobic training in ‘trained’ subjects.” 

They recruited well-conditioned, competitive male runners who were training an average of 82 km/wk and averaged 5 years of running experience.  Performance tests revealed that these subjects initial level of fitness was equal to or greater than that attained by the subjects in Hickson’s study.

Like Hickson, Mikesell and Dudley kept frequency and duration constant while increasing intensity relative to improvements in performance.  They had the subjects train 1 hour/day, 6 days/wk for 6 weeks, alternating 3 days of running and 3 days of cycling.  Note that though frequency and duration of training were held constant during the study that the total running volume of these subjects was 60% below their usual training regime.  In effect then, these subjects were exercising at a significantly decreased average training volume and an increased training intensity.

For 5 weeks the subjects experienced a significant, linear increase in performance and fitness.  In spite of their already high fitness and VO2peak levels at the start of the training study the average weekly improvement of these subjects was very similar to those in Hickson’s study.  The authors concluded that intense training induced a significant and linear increase in performance in trained athletes too.

In the 6^th week of the study the performance of the subjects began to drop.  This was attributed to overtraining and since then some have suggested that athletes limit high intensity training to 6 weeks or less.

Mikesell and Dudley’s research answered the question - increasing intensity does result in improvements in already highly trained athletes too, even when the training volume is fully 60% below the usual training volume of the subjects.

Gaskill et al

The studies of Hickson and Mikesell/Dudley lasted 25 weeks and 6 weeks respectively.  Based on the data from this and other similar studies, runners generally accept that short-term increases in intensity produce measurable changes in performance.  However, does it follow that high intensity training is equally effect in the long term?  Does increasing intensity results in a short term improvement only?  Will increasing intensity for a longer period of time produce continuing improvements or will performance plateau or even drop as in the study by Mikesell and Dudley?  A study by Gaskill et al that examined significant changes in training intensity in cross country skiers during a two-year research study provides an answer to this question (4).

In this study researchers designed a 2 year project to study a group of 14 competitive cross-country skiers.  These subjects had been training and competing for an average of 8 years for the males and 11 years for the females.  During the first year of training “all 14 athletes used a similar training program following the current paradigm of high volume and low intensity.”  Training averaged 660 hours per year with 16% (106 hours) of the training at high intensity following a standard base building, high volume, periodization program.    The researchers also reviewed the training data for the previous three years for all the athletes and noted yearly increases in training volume of 5-10% annually.

At the end of the first year of training only half of the athletes improved their performance.  Increasing weekly mileage produced improved performance in only 7 of the subjects.  The 7 athletes that did improve during the first year were assigned to the control group and for the 2^nd year of training they repeated the previous year’s training but increased their total yearly training volume by 6% to 688 hours.  Yearly training volume was increased by adding additional base building workouts, with no changes in the overall volume of intense training. 

Half of the athletes in this study did not improve from the increased training volume during the 1^st year of training.  The researchers also noted that this group’s performance had plateaued in the three years prior to the beginning of this study.  Four years of steadily increasing training volume following a traditional base building, high volume, periodized training program had not produced improvements in this group of athletes, nor were they able to outperform the athletes in the control group.  These lower performing athletes were assigned to the treatment group.  They decreased overall training volume slightly while significantly increasing the volume of intense training by 236% (increase from 100 hours to 236 hours).  This was accomplished by increasing the volume of intense training while decreasing the volume of easy paced, base building training.  The increased volume of intense training was spread throughout the entire training and competitive year.

At the end of the 2^nd year the control group, which had increased training volume during the 2^nd year with no increase in training intensity, had an insignificant increase in performance as measured by their best finish in the U.S. National Championships and their season long ranking by the United States Ski Association.  In contrast, the treatment group, which had decreased their base building training and increased their volume of intense training, improved their performance dramatically and significantly in both measures and they even outperformed the control group in total USSA points.

This study shows that long-term intense training can not only be sustained but can result in improved performance versus a conventional high volume, base building, periodized training program, at least for some athletes.  It also shows us that some athletes do not respond well to a conventional base building, increasing volume training program.

Billat et al

Finally, two studies by Billat et al examined the training characteristics of the national marathon teams of Portugal and France and elite Kenyan long distance runners (5, 6).  In both studies Billat discovered that the better performing athletes trained at a higher volume of high intensity training than did the lower performing athletes.  In the case of the marathon runners, Billat noted the faster marathoners “trained more total kilometers per week and at a higher velocity.”  Conversely, for the elite Kenyan runners Billat discovered that the faster Kenyans ran less weekly mileage but a higher total volume of intense training.  In both cases, the volume of high intensity training was associated with better performance, with total high intensity training, not volume of training, distinguishing between the faster and slower athletes.

In summary, the research on intensity of training shows that from untrained subjects to elites and from training programs lasting just 6 weeks to those as long as two years intensity of training exerts a principle role on performance.  Table 6 presents the results of the six studies on changes in intensity.

Case 2 – Measuring the influence of Duration

In case 2 we examine the role of duration on performance.  Again, a systematic approach means we want to hold intensity and frequency constant while modifying duration.  Any changes in performance will allow us to determine the effect duration has on performance.

Hickson et al

Robert Hickson and a team of researchers examined changes in performance due to changes in training duration (7).  In this study the subjects trained 40 minutes per day, 6 days per week, for 10 weeks.  Then they reduced training duration to either 26 or 13 minutes per day for an additional 15 weeks.  Intensity and frequency were maintained for the entire 15 week reduced duration training period.

The subjects’ short-term endurance was measured with a maximal test that resulted in exhaustion within 4-8 minutes.  Long-term endurance performance was tested by having the subjects exercise at 80% VO2peak until exhaustion, which was reached in about 2 hours.

Short-term performance times during the short-term test remained constant for both the 26 and 13 minute groups during the entire 15 week reduced duration training period, with no decrease from peak values.  There was even a small increase in performance for the 26 minute group, though the improvement did not rise to a level of significance.

Long term performance was a different story though.  Following 15 weeks of reduced duration training there was no decrease in long-term performance in the 26 minute group.  However, there was a significant loss of performance by the 13 minute group, with an average decrease of 10%, from 139 to 123 minutes.

It’s not surprising that short term performance did not decline since, in terms of specificity of training, both groups were training longer than the duration of the short term test.  However, with only 13 or 26 minutes of training per session it is more difficult to make the case that specificity of training played a significant role in a performance test that took more than 2 hours to complete.  The difference in performance of the 13 and 26 minute groups in the long-term test indicate that duration of training exerts a powerful influence on performance.

Interestingly, based on data from this and an unpublished study, Hickson concluded that a greater duration of training was required to improve performance than to maintain performance.

Dudley et al

In a classic study Gary Dudley and team examined changes in mitochondrial enzyme density amongst 19 different animal training groups (8).  In this study he had rats training at one of 6 different intensity levels for either 5, 10, 15, 30, 60, or 90 minutes per day, 5 days per week, for 8 weeks.  Within any particular intensity level, there were 3 or 4 different training durations.

The results of this study were unequivocal.  At any intensity training adaptations, as measured by changes in mitochondrial enzyme density, were significantly increased with increases in duration.  Though different types of muscle fibers responded to different intensity levels, adaptations peaked at 60 minutes of training.  Adaptations did not decrease with 90 minutes of training, but 90 minutes of training also didn’t produce greater adaptations either.

Unfortunately, performance was not measured during this study so we can’t compare if performance would have been different for the different training durations.  However, since changes in mitochondrial enzyme density are correlated with endurance performance we can reasonably say that increasing duration does result in improved performance, at least up to a duration of one hour of running anyway. 

Pollock et al

Supporting the research of Dudley is a research study by Pollock et al that examined improvements in fitness and risk of injury (9).  Researchers recruited 99 untrained volunteers and had them run at one of three durations (15, 30, 45 minutes), 3 days per week, for 20 weeks, all at the same intensity. 

As would be expected all three groups significantly improved performance.  Additionally, changes in performance increased with each increase in duration of training, with the 45 minute duration group improving the most.  This study clearly demonstrates the influence duration of training has on performance.  The results are shown in table 2.

Table 2:  effects of duration on changes in performance

Hansen et al

In a recent study, Danish researchers explored the effects of training twice per day versus training once per day (10).  In this study they had subjects perform leg extensor exercise for either 1 or 2 hours per day, at 75% of maximal power output, for 10 weeks.  One leg was trained one hour per day for 5 days per week.  The other leg was trained for 2 hours per day 3 times per week, though training for this leg was not continuous.  The subjects trained this leg for 1 hour, took a 2 hour rest in a fasting state, and then trained it again for an additional 1 hour.

Prior to beginning the training program the subjects were tested for peak power output and time until exhaustion at 90% of peak power.  At the end of the 10 week training period the subjects were tested again for both peak power and time to exhaustion.  Peak power had improved significantly for both legs as compared to pre-training levels, with both legs equally improved and matched post-training.  However, time to exhaustion was nearly double for the 2 hr leg as compared to the 1 hr leg (19.7 minutes vs. 11.9 minutes respectively).

Two facts need to be noted.  The leg that was trained 2 hours per day trained 1 hour more per week than did the leg that was trained 1 hour per day (6 hrs. per week for the 2 hr leg vs. 5 hrs. per week for the 1 hr. leg).  Additionally, the 2 hr. leg was not trained continuously, with 2 hours of rest (but not food) occurring mid-exercise.  One additional hour of training per week likely had no measurable effect on performance and the 2 hr rest with fasting probably allowed little long-term recovery of the exercising muscles.  Therefore, in spite of these factors, this study argues that the increased duration is responsible for the markedly higher performance of the 2 hr leg.

Additionally, the improvement in performance was measured in a maximal test that took less than 20 minutes to complete.  As in the Hickson study, specificity of training likely had little influence on performance because both legs were trained for a longer duration than the length of the test.  All things considered, this study provides perhaps the strongest evidence of the principle role duration of training exerts on performance.  The results of this study are graphically shown in figure 1.

Fig. 1:  Differences in performance from training either 2 hr day for 3d/wk or 1 hr day for 5d/wk

Wenger and Bell

Two researchers from the University of Victoria conducted an extensive review of the available research on the effects of different factors, such as frequency, intensity, and duration, on exercise performance and fitness (11).  “This review has grouped many studies on different populations with different protocols to show the interactive effects of intensity, frequency, and duration of training…”  They reviewed a total 78 individual studies pertaining to their subject matter.    Of the 78 total studies 37 pertained to the effect of duration on performance.  Based on the data found in the 37 studies the researchers wrote, “Thus, improvements in aerobic power can be achieved across all durations from 15 to 45 minutes.  However, the longer duration (> 35 minutes) are more beneficial.”

Considering the results of all 5 studies it is clear that duration of training has a principle affect on performance.  This is a significant point.  Previously, duration of exercise has been seen as important for two main reasons.  First is the requirement for specificity in training - specificity demands that the athlete perform workouts of similar duration as the competitive event.  Second is the need to build an “aerobic base”.   The aerobic base theory suggests that a high volume of long duration, easy paced runs result in improved aerobic capacity.  This improved aerobic capacity, or base, is supposed to enable the athlete to achieve a better performance when high intensity if added back into the training program than if a “base” had not been built or if a smaller “base” had been built.  Note that this theory suggests that the “base” is not the direct cause of improved performance, but that it provides a foundation that allows the athlete to achieve a better performance than if the foundation had not been built. 

Our review of the research shows that duration of training has a direct, primary influence on performance.  Certainly specificity of training is important, but the influence that duration has on performance is different than just the requirement for specificity of training.  Note also that the research shows that even though duration exerts a principle influence on performance, its influence is measurably less than that of intensity.  Intensity exerts the most powerful influence on performance, followed by duration of training.  The results of the five studies on duration are summed in table 6.

Case 3 – Measuring the influence of Frequency

Now we turn our attention to the influence of frequency of training.  Again, the goal would be for research studies to hold duration and intensity constant while varying frequency.  Any changes in performance then can be logically attributed to frequency of training.

However, as logical as the above may sound it is actually difficult to implement.  The challenge is that you can’t have a workout without having duration and intensity also.  If one runner trains 3 days per week and another trains 5 days per week even though they may both train for the exact same duration and frequency on the 3 common training days, the runner training 5 days per week has 2 additional days of duration and intensity.  This makes it more difficult to hold duration and intensity equal while modifying frequency.  Different researchers have tackled this problem in a variety of ways.  For our purposes the important thing is to simply be aware that measuring the influence of frequency is more difficult to assess than the other two for the reason stated above.  We will consider duration and intensity to be equal if during the common training days the subjects train with the same duration and intensity.

Hickson et al

Robert Hickson and team examined the effect reducing frequency of training had on performance (12).  In this study subjects trained 40 minutes per day, 6 days per week, for 10 weeks which resulted in a large improvement in fitness and VO2peak.  At the end of the 10 week train up period, subjects reduced training frequency to either 4 or 2 days of training per week for an additional 15 weeks.  Duration and intensity were held constant during the 15 weeks of reduced frequency training.

Despite reducing frequency by 1/3 or 2/3, the subjects’ performance and VO2peak were maintained for the entire 15 week period.  Interestingly, 2 subjects that had reduced training to 4 day per week for 15 weeks then volunteered to further reduce training to 2 days per week for an additional 15 weeks.  Their performance remained stable the entire 30 week period of reduced training.  Based on data from this and an unpublished study Hickson concluded that less frequency was required to maintain performance than was required to improve performance.  I would note that in every Hickson study where intensity was plateaued performance plateaued also.  Each time the subjects increased intensity in relation to improved fitness, performance continued to improve.  This being the case it would have been illuminating to see the results of a reduced frequency study where intensity was allowed to increase.   In any case, this study reveals that, at a minimum, athletes can maintain performance for at least 30 weeks even when training has been reduced to just 2 days per week if intensity and duration are maintained.

Crews and Roberts

Two researchers wanted to examine the effect of different intensities and frequencies of training on performance (13).  To do so they recruited 46 subjects and trained them at one of two training intensities for either 1, 3, or 5 days per week for 7 weeks.  Duration of weekly training was equalized with all subjects training for 50 minutes each week (the duration of a training session was dependent on the frequency of training with less frequent training resulting in a longer individual duration of training).  The end result was 3 lower intensity training groups and 3 higher intensity training groups.    Prior to beginning training the subjects were tested for various indices of fitness and given a performance test.

The results of this study were the 5 and 3 days per week groups both improved significantly more than the 1 day per week group but that there was no significant differences in improvement between the equal intensity 5- and 3-day groups (the higher intensity 3 day group improved slightly more than the higher intensity 5 day group but not to a level of significance).  The study also showed that the higher intensity groups improved more than did the low intensity groups for all training frequencies.  In fact the higher intensity 1-day per week group improved more than did all the lower intensity groups (though not significantly more than the 5-day group). 

This study indicates that frequency of training plays a role in performance, albeit a minor one, if intensity is equal for all groups.  It further shows that intensity and duration of training exerts more influence than frequency since the higher intensity 1-day of training per week group improved more than all the day lower intensity groups.  Finally, it illustrates the different results produced by the different blends of frequency, intensity, and duration. The results of this study are graphically displayed in figure 2.

Fig. 2: Effect of intensity and frequency on performance (physical work capacity)

Pollock et al

Pollock and team wanted to examine improvements in fitness with increasing frequency of training in relation to risk of injury (14).  In this study 71 subjects trained either 1, 3, or 5 days per week, 30 minutes per session, all at the same intensity.  The results are shown in table 3.

The results of this study provide some support for Hickson’s belief that more frequent training is required to improve performance than is required to maintain it.  In this study each increase in frequency resulted in an improvement in performance though there is not a significant difference between either 1- and 3 d/wk or 3- and 5 d/wk training.  Even though there is not a significant difference between adjacent groups the trend in improvement is clear.

Table 3:  effects of frequency on changes in performance

Busso et al

In contrast to the study by Pollack is a study by Thierry Busso and a team of French researchers.  They conducted a study to evaluate variations in performance response due to changes in training frequency (15).  More specifically their aim was to discover “whether an increase in training frequency and thus a decrease in recovery time between training sessions would induce a progressive increase in the magnitude and duration of long-term fatigue induced by an identical training load.” 

They had subjects train for 15 weeks via cycling.  The first 8 weeks of the study the subjects trained 3 days per week.  Following a week off from training the subjects then trained 5 days per week for 4 weeks.  Performance was measured throughout the training period.

As would be expected, performance at the end of the 3 days per week of training was significantly elevated over pre-training values.  When training was increased to 5 days per week, average performance improved only slightly, but insignificantly, over values obtained from training 3 days per week (P_lim5, W = 355 vs. 348 respectively).  The researchers also noted a much wider variation in performance during 5 days per week of training and a longer recovery time.

Dolgener, Kolkhorst, Whitsett

Also in stark contrast to the results of Pollack, Dolgener et al found no difference in performance between two groups of novice marathon runners who trained either 4 or 6 days per week (16).

This group of researchers wanted to evaluate two common components of marathon training – easy training runs and long runs.  They recruited 71 subjects to participate in an 18 week study which culminated with running a marathon.  The subjects were divided into 2 groups, with one group running 6 days per week and the second group running just 4 days per week, with equal intensity of training and duration of long run for both groups.  The researchers hypothesized that as long as the subjects all performed the same long run each week and all trained at the same intensity level that performance would be the same for both groups despite the differences in training frequency and overall training volume.  The differences in weekly training volume peaked at 90 minutes of training for 4 weeks.  Table 4 summarizes the weekly differences in training volume between the two groups.

Table 4: Difference in weekly training volume between the lower mileage and higher mileage marathon training groups

Despite the differences in training frequency and overall training volume both groups had identical performance in the marathon.  The researchers concluded that the additional training volume and frequency of the 6 days/wk group were not necessary in order to successfully complete a marathon and provided no additional physiological benefits.  (In the time since this study was published these researchers have trained over 500 subjects to successful marathon completion following the 4 days per week training program.)

Mutton et al

All of the above studies on modified frequency used untrained subjects.  It is possible that subjects that have a training history will differ from untrained subjects in their response to changes in frequency.  A group of researchers wanted to investigate whether replacing half of weekly run volume with cycle training would result in a change in fitness and 5km performance (17).  They recruited twelve moderately trained runners (16-30km per week of training with an average 5k time of 23 minutes) and divided them into two groups.  One group ran 4 days per week and the other ran 2 days per week and cycled 2 days per week.  Both groups trained for 5 weeks with the same duration, frequency, and total volume of training.

At the end of 5 weeks of training, both groups significantly and equally improved 5k performance by 1.7 minutes, an average 7.4% improvement.

While this study was designed to test the effectiveness of cross training it also suggests that frequency of specific event training might not be as important as frequency of training for all activities combined. 

Wenger and Bell

Two researchers from the University of Victoria conducted an extensive review of the available research on the effects of the factors of frequency, intensity, and duration on exercise performance and fitness (11).  “This review has grouped many studies on different populations with different protocols to show the interactive effects of intensity, frequency, and duration of training…”  They reviewed a total of 78 individual studies pertaining to their subject matter.  After reviewing 44 studies pertaining to frequency of training the researchers wrote, “The optimal frequency for all intensities of training is 4 times per week.”

Summary of Frequency Studies

In review, five of the above studies that reported changes in performance all found little to no differences in performance due to higher frequency of training.  One study (Pollock et al) found that performance improved insignificantly more in the group that trained with the most frequency.  Finally, a review of 44 studies examining the changes in performance showed 4 days to be the optimal training frequency.

Based on the research we see that frequency of training does not exert a strong influence on performance, especially in comparison to the influence of duration and intensity.  Training frequency exerts a mild influence on performance and its influence is secondary to that of duration and intensity.

Additional Studies

Other studies are available that examine different combinations of the training variables on performance and fitness.  Though these studies do not fall within our matrix of controlling two variables while modifying the third variable they do provide useful data for our investigation.

Franklin et al

This group of researchers wanted “a) to determine the accuracy of predicted marathon finishing time among experienced and inexperienced marathoners, and b) to examine the relationship between training mileage per week and marathon performance”(18).  The researchers interviewed 158 contestants of the 1977 RWR Marathon.  Subjects were classified according to marathon experience as either first time marathoners, second time marathoners, or experienced marathoners (more than 2 marathons).  127 subjects completed the marathon, consisting of 63 first timers, 29 second timers, and 35 experienced marathoners.

The research showed that weekly mileage influenced marathon performance only slightly.  Correlation coefficients indicated that 16.7%, 28%, and 9.6% of marathon performance among first timers, second timers, and experienced marathoners respectively was explained by differences in weekly run mileage.  It was also discovered that for any given training mileage per week, experienced marathoners ran faster than did first time marathoners.

The researchers commented, “…the variation show(s)…that training expressed in miles/week (without control for specific frequency, intensity, or duration) offers little explanation for the differences in marathon performance amongst experienced or inexperienced runners.  Since distance is a function of intensity (speed), duration, and frequency, the present results highlight the inadequacy of training distance alone as a major determinant of performance.”

Grant et al

A study by researchers from the University of Glasgow confirms the data from the Franklin marathon study.  In this study researchers examined the influence of weekly training mileage on marathon performance by gathering data from 88 first time marathoners (but not necessarily novice runners) at the 1982 Glasgow Marathon (19).

As in the Franklin study, the data from this research showed that weekly training mileage was a poor predictor of marathon performance.  The correlation coefficient for marathon performance and average training mileage indicated a prediction coefficient of just 14%.  In other words, weekly mileage could only predict marathon performance in 14% of the cases.

In their conclusions the researchers wrote, “There is no practical relationship between average weekly training mileage and race time, for novices at least.” 

The researchers also noted that “First time marathoners do not need to run exceptional mileages during training and do not suffer dramatic slow down in the latter stages because of their moderate training.”  This finding has been confirmed from the marathon data from Dolgener et al (16).

King et al

Abby King and team designed a study “to determine the 2-year effects of differing intensities and formats of endurance exercise” on performance and various indices of fitness and health (20).

They recruited 259 sedentary subjects between the ages of 50 – 65 and randomly assigned them to either a control group or one of three exercise groups:

Exercise consisted mostly of running/walking with some use of treadmills or stationary cycling.  Adherence to the exercise program was regularly monitored throughout the 2 year period and the subjects submitted monthly training data.

Performance data at the end of two years was compared to initial performance levels.  Subjects in all three training groups showed significantly greater improvements in both VO2peak and performance than the control subjects.  Performance for the two higher intensity, 3-training-days-per-week groups improved more than the lower intensity, 5-training-days-per-week group (18% higher intensity home training, 16% higher intensity group training, and 14% lower intensity home training).  The results are summed in table 5.

Table 5:  Comparison of 3 training prescriptions for intensity, frequency, daily duration, weekly duration, total duration, and performance from 2 years of training

The really interesting thing about this study is that it is basically a competition of duration and intensity versus frequency and volume.  Two groups trained with up to 10% higher intensity for a 1/3 longer duration while the other group trained with higher frequency and training volume.  The results were that the higher intensity, longer duration groups outperformed the higher volume and frequency group.  This study supports the idea that volume and frequency influence performance mildly, while intensity and duration exert a primary influence on performance.

Research Summary

The research on the influence of the three training variables of frequency, duration, and intensity shows that duration and intensity exert a strong, principle influence on performance, while frequency of training exerts a smaller, secondary influence on performance.  The research shows that intensity influences performance the most, followed by duration, with frequency a distant third.  Weekly mileage, as a measure of training, a goal or training, or as a determinant of performance, without considering the influence of intensity, duration, and intensity, is a poor predictor of performance.  Table 6 summarizes the 18 studies on intensity, duration, and frequency reviewed in this section.  Table 7 reviews the three additional studies examining difference combinations of intensity, duration, and frequency on performance.

Table 6:  Summary of research studies examining effects of intensity, duration, and frequency on performance

Table 7:  Summary of additional research examining effects of combinations of different intensity, duration, and frequency on performance