Although I generally avoid reviewing training techniques in this blog, I feel strongly about the potential of the polarized technique as the most promising modality to improve ones results. Polarized training is a very simple concept which revolves around the usage of many hours of boring, low intensity exercise (below VT1) with a small amount of very high intensity (above VT2) interval work (notice the "High aerobic shortage", is intentional). Although some time can be spent at VT2 or between VT1 and VT2, it should be minimal. Is there any logical rational behind this?
Several physiologic factors come to mind about why this is logical. Let's list them:
Junk miles are not worthless. High volume, low load exercise benefits mitochondrial mass, plasma volume, blood volume, type 1 muscle conversion and local capillary growth. More importantly, even if these effects were more pronounced with HIT training (with some evidence supporting), one simply can't do more than a given amount of very high intensity work without either over reaching or over training. This becomes even more critical with aging athletes, they need more time to recover after a HIT session. So instead of sitting around taking rest days, continuing your training with low load, higher volume makes sense.
Most studies show multiple benefits of HIT sessions and intervals. These range from improved stroke volume, cardiac contractility, VO2 max, MLSS, lactate disposal as well as overlap with some of the other benefits mentioned above. So HIT intervals are great to do, but will cause fatigue and overall body stress, so there is a limit to what can be done.
What about the middle zone between? Are there benefits to spending time at or just below VT2 power levels? Does introducing some middle zone training actually hurt your performance if you were otherwise just doing low and very high intensity training? It may be so - perhaps adding the middle zone (just under VT2) would cause enough fatigue and stress to interfere with proper HIT intervals. That is my concern and something that I have noted from my own training.
With that preamble, the following study is very enlightening.
Methods: The authors recruited a wide variety of young athletes:
Participants were members
of the Austrian cross-country skiing national team(n = 8), run-
ning (n = 21), triathlon(n = 4) or cycling(n = 15) teams during
or since the year before the current study.
The training intensity was con-
trolled by HR based on the baseline incremental test:(i)LOW
(HR at blood lactate value <2 mmol·L−1); (ii)LT(HR corresponding to a blood lactate of 3–5mmol·L−1); (iii)HIGH(>90% HRpeak)]
It is a bit regrettable that the LOW group was defined as such since that may have been over LT1/VT1, but lets be optimistic and call it LT1 power.
Here is the protocol map
Each cycle was repeated x 3 for a total of 9 weeks:
The first group A (LOW) was essentially training at LT1 with 2 sessions of LT2 intervals over 2 weeks.
The "threshold" group B (THR) trained at both the LT1 and LT2 without any HIT.
The polarized group (POL) labeled as C trained primarily at LT1 with several HIT sessions.
Finally the HIT group D just did high intensity work and as expected, was of shorter duration.
How did this turn out with respect to the actual training intensity distribution?
Here is the breakdown:
A few observations:
The numbers don't add up. For instance look at the HVT (high vol at low intensity). About 49% of the training was at low intensity, 9% at LT2 but what about the other 40% or so? The reason (thanks to Dr Stoggl for the clarification) is a typesetting error. The number in the brackets is the true percent:
- The POL group spent 68% at the "low" intensity, only 6% at LT2 and 26% at HIT values.
- The THR group spent about 54% of their training at the LT2. That surely must have been exhausting.
- The HVT group spent 16% at the LT2, the remainder at "low". Despite this group being labeled as low intensity, they did train almost 16% at the LT2 which is a commonly used routine for endurance sports.
What were the outcomes?
Before discussing this, the results can be deceptive in some respects. For an analogy, remember strength training. If you train for strength (high load, few reps) you will have a better 1 rep max. However if you train conventionally (moderate load x 10+ reps) you will get better muscle mass. So depending on what you are looking at, it may be easy to get fooled. It is possible that if they tested for Wingate 30 power, the HIT group would have done the best. The authors did look at established parameters such as time to exhaustion and VO2 max which is helpful. The other concern is the relative short nature of the study (9 weeks). Endurance exercise related physiologic improvement may take many years to manifest fully. However, given the practical reality of athlete schedules, study funding and commitments the 9 weeks is admirable and probably the best that can be practically done
Here are some breakdowns, with particular attention to VO2 max:
- The POL group had the largest increase in VO2 max. This will be explored below.
- Despite over twice the HIT sessions, the HIT group did not have as good a VO2 max response as POL. Therefore fewer HIT sessions with a large amount of low intensity training was superior to just HIT.
- Although I'm not quite certain about statistical significance, the THR group had a decline in VO2 max. This was spoken of in the discussion:
Time to exhaustion and Power/velocity at 2 or 4 mmol lactate:THR improves VO2peak, lactate or ventilatory thresholds and
endurance performance in untrained persons(Denis etal.,1984;
Londeree,1997;Gaskilletal.,2001). These findings contrast those
of the current study, as we did not observe improvements in
VO2peak, V/P4, TTEorV/Ppeak in our elite athletes in response to
THR. Additionally, it is possible that in well-trained endurance
athletes,repeated training bouts at LT might generate unwar-
ranted sympathetic stress.
In addition to the above VO2 parameters, a TTE test was done as well as a comparison of either power or velocity at a fixed lactate (2 or 4 mmol). For example during exercise at a lactate of 2 mmol, the POL group improved their power by 9.3% but the HVR did not change.
- We see the biggest improvements in TTE with POL, and the least with THR (although that one is not significant).
- Power at a lactate of 2 or 4 mmol improved with POL or HIT.
- Despite over twice the HIT sessions, the HIT group did not have as good a TTE response as POL
So what some final thoughts?
Obviously, polarized training seems advantageous in this population of young elite athletes.
This protocol lead to better VO2 max, TTE and power at fixed lactate than the other training distributions. Does this extend to other populations, especially older, non elite subjects? My suspicion is not only does it extend, it probably is even more critical. The THR group did relatively poorly in VO2, TTE and power at fixed lactate. I think this is a good example of what happens when one trains hard but takes very little time to have recovery sessions. As discussed previously, runners did best when their easy volume was high, not from their HIT sessions. The THR group probably was training too hard too often. In contrast the POL group spent some time at even higher work rates but compensated by a large fraction of training time at "easy" levels. The key issue here is what is the definition of easy. For these young, creme of the crop, national team athletes, "easy" may be quite different than what a 50 year old masters cyclist or runner will consider easy. As discussed in my posts on lactate and ventilatory thresholds, there is controversy on where these limits are. My personal viewpoint is that the easy zone should be well away from uncorrelated DFA a1 values, which turn out to be near VT1. Since a potential rational of easy training is to still get some benefit while recovering enough to perform HIT thereafter, the easy training must be of low cardiac stress. In regards to the near 25% training time in the HIT zone for the POL group, that may not be possible for most. Even if it was achievable, longer term maintenance at this intensity distribution (75:25 low:HIT) could be detrimental to non elite or older subjects.
Since it appears desirable to spend the majority of training at low intensity levels, what unique physiologic benefits occur in this zone?
Increased capillary density - A prime goal of endurance exercise training is improving net fuel delivery (O2) to the muscle cell. Having a more extensive network of capillaries will certainly be great benefit. An interesting study done several years ago looked at the changes in capillary density and vascular endothelial growth factor (VEGF) after either constant moderate vs high intensity cycling. The HIT intervention consisted of multiple 1 minute intervals of 120% VO2 max power versus the constant cycling group of 60% VO2 max (probably near VT1). Capillary density increased after a initial 4 week conditioning session that consisted of cycling at 64% of VO2 max power. However, the performance of the multiple HIT sessions did not further enhance density:
VEGF levels derived from muscle biopsy also supported this with higher post moderate vs post HIT intervals:
VO2 max changes:
Maximal oxygen uptake. The intense intermittent
training period led to an increase (P <0.05) in maximal
oxygen uptake from 3.59±0.21 to 3.87±0.20 lmin−1 and
from 41.43±1.45 to 45.43±1.80 ml min−1 kg−1.
- Therefore despite the enhancement in capillary density in the constant group, only the HIT subjects improved VO2 max.
- Low intensity training stimulates VEGF secretion and capillary growth. HIT does not appear to do so.
- Undoubtedly both HIT and LOW are useful modalities for improving fitness parameters.
- We should be reassured that low intensity exercise is not simply "junk miles".
Optimizing low intensity training:
Now that we see that low intensity work should be a focus of one's training distribution, is there any "optimization" that can be done? For instance, is it better to do low intensity exercise in heat vs cool conditions or fasting vs fed? What I am more interested here is heat/low carb training as a way to boost the metabolic benefits of low intensity training. As will be seen with low carb exercise, there may be actual detrimental effects if performing this in a low carb state. In addition, HIT during severe heat can be extremely dangerous to multiple organ systems and from the medical perspective if done at all, should be brief.
A recent review discussed the effects of heat on endurance factors in detail. There are numerous biologic changes that occur after either passive or during exercise heat exposure. Unfortunately, the test tube studies and animal models may not be pertinent to humans. The authors did state the following however:
Heat stress per se can stimulate HIF1-a and its downstreamAn interesting study was done about 12 years ago subjecting athletes to a hot sauna immediately after exercise for about 30 minutes. They showed increased blood and plasma volume as well as TTE in the sauna treated runners:
target genes Vegf, heme oxygenase-1 in rat myocardium, and
epo and epo receptor in rat kidney (Maloyan et al., 2005).
Long-term passive heat acclimation increased HIF1-a protein
levels and thus induced larger upregulation of gene activation
for Vegf, HO1, epo, and epo receptor in response to a single
bout of heat stress (Maloyan et al., 2005). Heat stress increases
circulating epo concentration in humans, although this finding is
not universal despite substantive heat stress (Akerman et al.,
2017), perhaps reflecting the important potentiating effect
gained from long-term heat acclimation. The success of runners
from Kenya and Ethiopia in endurance-based events in the
context of thermal adaptation warrants consideration. Athletes
who train in these countries are chronically exposed to lowgrade
heat stress that might be considered ‘‘at risk’’ for physically
active individuals (i.e., annual mean daytime dry bulb temperatures
>22 C). From an evolutionary perspective, warm and
relatively dry climates helped homo sapiens gain a major advantage
in terms of superior thermoregulation (outweighing a poorer
running economy) compared to other large mammals, conferring
superior endurance capabilities (Bramble and Lieberman, 2004)
The question then arises whether there is an advantage to exercising in the heat in preparation for a race at normal temperature. This is quite different than training in the heat for a race at high temps where acclimatization is essential. There has been controversy on this matter but two recent abstracts (full papers pending) are interesting.
The first one addresses this question exactly:
Heat acclimation involves physiological adaptations that directly promote exercise performance in hot environments. However, for endurance-athletes it is unclear if adaptations also improve aerobic capacity and performance in cool conditions, partly because previous randomized controlled trial (RCT) studies have been restricted to short intervention periods.The intervention:
Participants were instructed to maintain total training volume and complete habitual high intensity intervals in normal settings; but HEAT substituted part of cool training with 28 ± 2 sessions in the heat (1 hour at 60% VO2max in 40°C; eliciting core temperatures above 39°C in all sessions), while CON completed all training in cool conditions.That translates to 1 hour at probably zone 1 (VT1) at 104 degrees F. So pretty extreme!
I have done this and believe me, it's not fun.
When tested in cool conditions, both peak power output and VO2max remained unchanged for HEAT (pre 60.0 ±1.5 vs. 59.8±1.3 mL O2/min/kg)
Based on the present findings, we conclude that training in the heat was not superior compared to normal (control) training for improving aerobic power or TT performance in cool conditions.
The second abstract seems like it used the same subjects (same authors, protocol) as above but looked at hemoglobin mass and plasma volume:
Heat acclimation is associated with plasma volume (PV) expansion that occurs within the first week of exposure. However, prolonged effects on hemoglobin mass (Hbmass) are unclear as intervention periods in previous studies have not allowed sufficient time for erythropoiesis to manifest. Therefore, Hbmass, intravascular volumes and blood volume (BV)-regulating hormones were assessed with 5½ weeks of exercise-heat acclimation (HEAT) or matched training in cold conditions (CON) in 21 male cyclistsMethods:
HEAT (n=12) consisted of 1h cycling at 60%VO2peak in 40°C for 5 days/week in addition to regular training, whereas CON (n=9) trained exclusively in cold conditions (<15°C)Results:
PV increased (p=0.004) in both groups, by 303 ± 345ml in HEAT and 188 ± 286ml in CON. There was also a main effect of time (p=0.038) for Hbmass with +34 ± 36g in HEAT and +2 ± 33g in CON and a tendency towards a higher increase in Hbmass in HEAT compared to CON (time*group interaction: p=0.061). The Hbmass changes were weakly correlated to alterations in PV (r=0.493, p=0.023). Reticulocyte count and BV-regulating hormones remained unchanged for both groupsConclusion:
Hbmass was slightly increased following prolonged training in the heat and although the mechanistic link remains to be revealed, the increase could represent a compensatory response in erythropoiesis secondary to PV expansion.
Taking both of the studies together it appears that there is a subtle hemoglobin mass improvement, yet no noticeable improvement in power outputs or VO2 max. It is certainly possible that a few seconds could have been shaved off a 40 K TT after the heat intervention however, is it worth training under such miserable conditions? In addition, unless you live in the tropics, you would need an indoor set up with a strong heater.
Without question there may be a genetic component where some athletes will have a better heat effect than others. However it seems like a fair amount of trouble for a minimal effect.
Training under low carbohydrate conditions:
Another area that has received much attention is that of training with either a low carb diet or at least low carbohydrate conditions (fasting). As with heat training, both cellular, animal and human models are somewhat confusing. On one hand, it makes sense to optimize fat utilization by withholding carbs, but at the same time, one may be handicapped by not being able to perform with enough intensity and duration to get the usual training benefits. Here is a helpful diagram taken from an excellent review on the subject:
From the text:
Short-term (3–10 weeks) trainingImportantly the authors state that a mixture of training intensities and carbohydrate usage is needed for optimal performance goals:
programs in which some workouts are commenced with either
low muscle glycogen and/or low exogenous glucose availability
increase the maximal activities of selected genes and proteins
involved in carbohydrate and/or lipid metabolism and promote
mitochondrial biogenesis to a greater extent than when all workouts
are undertaken with normal or elevated glycogen stores
(Hulston et al., 2010; Yeo et al., 2008). These adaptations accrue
despite 7%–8% lower self-selected training intensities (Hulston
et al., 2010; Yeo et al., 2008).
As noted, a loss of training quality or intensity has been associatedIn the conclusion of the review some important points are raised:
with ‘‘train-low’’ sessions (Hulston et al., 2010; Yeo et al.,
2008) and must be balanced against aspects of ‘‘molecular upregulation’’
within the muscle cell. Overall, it appears important
to carefully periodize and integrate the number and type of sessions
completed with low carbohydrate availability within a
training program so that the overall goals of training are
achieved. This can be accomplished by a balance of
sessions undertaken with low carbohydrate availability to drive
molecular adaptations, while high-intensity workouts should be
commenced with high carbohydrate availability to allow an
athlete to mimic competition pace and habituate to competition
strategies (Bartlett et al., 2015; Stellingwerff, 2013; Jeukendrup,
2017). This ‘‘balance’’ appears to be more easily achieved with
sub-elite athletes (Marquet et al., 2016) than their elite counterparts
(Burke et al., 2017; Gejl et al., 2017).
Another caveat involves the possibility that a training strategyErgogenic potential of PPAR agonist therapy:
that promotes one attribute may endanger or impair another. An
impairment rather than improvement of metabolic flexibility
might result from the complex interactions between pathways
of substrate utilization; an upregulation of one pathway may
result in a simultaneous and reciprocal downregulation of others.
For example, in previous investigations of either short-term
(5 days) exposure to high-fat, low-carbohydrate diets or longer term (3 weeks) ketogenic diets (Burke et al.,2017), we found robust changes in muscle characteristics that promoted substantial increases in maximal rates of fat oxidation during exercise. However, this was associated with a reduction in the performance response to a training block (Burke et al.,
2017) and a failure to acutely enhance performance, even
when carbohydrate availability was restored before or maximized
during exercise (Burke et al., 2002; Carey et al., 2001).
Likely mechanisms underpinning these findings include a loss
of exercise economy (i.e., an increased O2 cost of exercise)
associated with the reliance on fat-based fuels (Burke et al.,
2017) and an impairment in muscle glycogenolysis underpinned
by decreased activity of the rate-limiting enzyme in carbohydrate
metabolism, pyruvate dehydrogenase (Stellingwerff et al., 2006).
These effects are observed during models involving sustained
(Burke et al., 2017) or periodic (Havemann et al., 2006) higher-intensity
exercise that forms a critical component of high-level
The above pathway caught my interest since a well known diabetes drug family leads to enhancement of of the PPAR pathway, namely pioglitazone (Actos). Over the years I have used this drug extensively for patients with type 2 diabetes. The other drugs in the family are off the market and since the class has "seen better days" in regards to clinical use, little new research has been done. However, a paper came out looking at Actos exercise related effects in mice that we can examine.
Rational for study:
Experimental evidence reveals that TZDs can induce the sameResults and conclusion:
biological effects in glucose metabolism, mitochondrial biogenesis,
and skeletal muscle fibre type as the prohibited drugs
GW1516 and AICAR. This study assessed the effects of
pioglitazone on performance and on skeletal muscle mitochondrial
biogenesis. For this purpose, blood glucose levels, the
protein expression of the intermediates involved in the mitochondrial
biogenesis pathway (cytochrome C, PGC-1α, NRF-1,
and TFAM), and citrate synthase activity in both soleus and gastrocnemius
muscles were measured. Maximal aerobic velocity
(MAV), endurance capacity, and grip strength were also determined
before and after a training period.
Treatment of exercising mice with either drug or placebo did not show any notable enhancement in maximal aerobic velocity, endurance capacity, citrate synthase or other markers of mitochondrial biogenesis
In our model, exercise-induced mitochondrial biogenesis is not
due to PPARgamma agonist pioglitazone administration. Overall,
oral pioglitazone administration enhances neither training adaptations
I include this for two reasons:
- Attempting to augment endurance exercise performance by PPAR manipulation probably is ineffective and could be detrimental (Actos causes fluid retention, weight gain and altered fat distribution).
- Even though molecular pathways appear simplistic, the translation to actual performance effects may be very complex and even contradictory.
- Polarized training appears to improve VO2max, power at fixed lactate and TTE in comparison to a regime of near 50:50 LT1:LT2 intensities. Perhaps a take home message here is to spend plenty of time at VT1 to be fresh for high quality sessions of HIT. There seems little benefit to adding zone 2 work loads or spending significant time at the MLSS. In other words if your Garmin watch says you need to train in the "High aerobic zone", ignore it or take it as a compliment that you are polarized.
Here is my "power zone" distribution over the year:
Based on these zones:
Zone 1 up to VT1
Zone 2 up to LT2
Zone 3 > LT2
- I may be spending too much time in zone 2 but much of that consisted of blog post testing and multiple ramps for lactate curves. On the other hand, at my age, I may not be able to handle more time in zone 3. When planning out your training distribution, multiple factors need to be taken into account.
- Changes in capillary density and vascular endothelial growth factor are more pronounced after constant moderate exercise than HIT.
- Training in the heat (40C or 104F) can increase plasma and blood volume but does not appear to translate to better TT performance or VO2 max values. This subject should continue to be monitored for other study results. There may be subsets of individuals who can benefit from heat exposure in regards to performance in thermo neutral conditions.
- Training under low carb conditions does improve multiple metabolic parameters that should lead to better endurance performance. However, it can be detrimental if not mixed with moderate carb intake during HIT sessions. From my perspective, low carb training is an ideal way to boost the benefits of a pure low intensity session - 1 to 2 hours at VT1 on awakening, while fasting (water only). If HIT is to be done, carbs should be present.
- If using a Garmin fitness tracker, take their training recommendation with a critical eye.
- Many thanks to Drs Stoggl and Sperlich for not only doing the polarized training study but their other contributions to training intensity distribution.