Polarised Cycling Training: A Detailed Guide
Polarised cycling training is a hot topic for those with their ears to the ground on cycling training and performance. Here, we’ll present a comprehensive overview of training with a polarised intensity distribution, looking specifically at:
What's Polarised Training?
In general terms, ‘polarisation’ means to be at one extreme or another, generally avoiding a medium, middle-ground area.
In cycling, this applies to training intensity distribution and the act of spending most of your training time at either a low or higher intensity.
It’s a concept that was most famously brought to light by the work of Stephen Seiler and has since been studied further by many other sports scientists.
Polarised training is almost diametrically opposed to the classic ‘threshold’ training approach, where the vast majority of training is done at a middle intensity at or around the maximal lactate steady state (also known as lactate threshold).
The polarised model defines three training zones, as illustrated in the figure below, where the vast majority of training (~75%) is spent in Zone 1, a considerable portion (~20%) is performed at Zone 3, and relatively little (~5%) is spent in Zone 2. Often this is simplified as 80% low intensity, 20% high intensity – hence the name “80:20” training, although this is an over-simplification that can lead to confusion, which we’ll get onto later.
Polarised Training Zones
The three zones are usually defined based on the relationship between lactate levels and power output.
For completeness, we’ve included a scientific explanation of how the zones are defined, although if you’re not interested in the science, you can skip to the next section on how these zones relate to a classic six/seven zone model.
LT1/VT1
The first lactate threshold (LT1) is the workload at which lactate levels become substantively elevated above resting levels, reflecting an increased contribution to power production from glycolysis (the break-down of carbohydrates). The first ventilatory threshold also occurs in a similar “area” to LT1, and can be observed by an increase in an athlete’s breathing rate, which is related to the intake of oxygen and especially the expiration of carbon dioxide (CO2)
LT2/FTP/CP
The second lactate threshold (LT2) is the workload at which the rate of lactate accumulation begins to exceed the rate of lactate clearance, meaning lactate levels accumulate rapidly and do not reach a steady state.
It is sometimes referred to as the ‘Maximal Lactate Steady State’ or MLSS. This is the point at which the aerobic energy system can no longer keep pace with the energy demand.
It’s what we’re typically trying to estimate when we do an FTP test, although FTP will almost always be slightly different to LT2. Critical power is a related concept and often sits in the same intensity region, though again will be slightly different to LT2 and FTP in most cases.
These thresholds are hard to determine precisely, given that lactate levels vary continuously with both power and time, and that blood lactate samples can only be taken as discrete measures.
There are therefore a number of protocols – none of which are perfect - that can be used to establish the cut points. These will produce varying results. A common approach is to define LT1 and LT2 as the workload where lactate levels are 2mmol/L and 4mmol/L respectively. However, in reality the actual turn-points might be higher or lower than these workloads.
Relation To Traditional Training Zones
It’s all well and good defining these training zones based on lab measures of lactate or expired gas.
However, most people don’t have access to these types of measures. We can try to relate the polarised model to a typical six or seven-zone model of heart rate or FTP training zones (see here for our recommended zoning system):
LT2 would sit around the mid to top end of Zone 4 in the six/seven-zone model.
LT1 is a little harder to define. On average, LT1 occurs at around 52% of Maximal Aerobic Power (MAP) (Cerezuela-Espejo et al., 2018), which would sit somewhere around ~65-70% of FTP, or around the top of Zone 2 in a six/seven-zone model. You may, however, see some sources suggesting LT1 occurs at the top of Zone 3. This is because the position of LT1 varies considerably, and will be higher for people with better fat oxidation abilities.
In our view, the best approach if you can’t do lab testing is to assume LT1 occurs at around the middle of Zone 3 in a six/seven zone model. So, a polarised model might manifest as an intensity distribution of 75% of training sessions in Zones 1, 2 and lower Zone 3 (below ~85% FTP), 5% in higher Zone 3 and Zone 4 and 10% in Zones 5+. For clarity, when we are referring to the polarised model zones, we’ll call these the ‘low’ ‘medium’ and ‘high’ zones, rather than zones 1-3 so we’re not confusing zone numbers between the different systems!
Ratings of perceived exertion can also be used fairly reliably to estimate the lactate threshold turn-points (Dantas et al., 2015), so it’s also worth checking your perceived exertion during training sessions to ensure these zone boundaries feel right for you. The low zone should be around a 1-4 out of 10 in terms of effort, the medium zone will feel like a 5 or 6, and the high zone will feel like 7+.
Why Use A Polarised Model?
Traditionally, endurance athletes and coaches have followed the ‘threshold model’, where the majority of training time is spent at or around the lactate threshold. It was thought that these sessions, being at a moderately hard, yet tolerable intensity, would produce the greatest training stress and thus stimulate the biggest training adaptations.
However, around two decades ago, a series of observational studies on the actual training practices of top-level athletes found that across many different endurance sports and countries, elite and Olympic-level athletes are not using the threshold approach in their training.
Instead, they were found to be spending large amounts of training time at low intensities and considerable time at high intensities. In other words, they largely minimise threshold training…
Evidence From Observational Studies
Going into these studies in more detail, Billat et al. (2001) for example, found that during the 12 weeks in the run-up to the Olympic qualifying trials, 20 elite marathon runners on average completed 78% of running distance below marathon pace, 18% above marathon pace, and only 4% at marathon pace. Marathon pace for these top-level athletes would be very close to the lactate threshold.
A polarised intensity distribution also has been found to be present during other phases of the training cycle. For example, Seiler & Kjerland (2006) conducted an observational study of junior Norwegian cross-country skiers over a 32-day period of the preparatory training phase (October/November) and showed that 75% of sessions were conducted at heart rates below VT1 (the first ventilatory threshold), 17% were conducted with significant portions above VT2 (the second ventilatory threshold), and only 8% were conducted with significant portions between VT1 and VT2.
Fiskerstrand and Seiler (2004) conducted a three-decade long study of Norwegian rowers, and found that as training practices evolved to incorporate increased polarisation (alongside greater training volume and more altitude camps), athletes’ performance and physiological fitness markers (e.g. VO2max) also improved. Studies like these offer some support for a polarised training approach.
Evidence From Intervention Studies
The studies above are observational, and not interventional studies (where training is manipulated to look at the impact of a given program), so they do not prove that polarised training causes good performance results.
Interventional studies are hard to conduct given the reluctance of many athletes and coaches to modify existing training approaches. However, a couple of interventional studies on polarised training do exist, and both support the use of a polarised approach as compared to a threshold model.
One study by Muñoz et al. (2014) compared 10 weeks following a polarised training plan (75/5/20 low/medium/high zones) versus a threshold plan (45/35/20 low/medium/high zones) in 32 recreational athletes. The plans were matched for training load (i.e. volume * intensity), meaning the total training time was slightly higher in the polarised group, but not excessively.
Overall, the polarised group improved their 10km running times by a greater amount than the threshold group (5.0% vs 3.5%) although this was not statistically significant. Interestingly though, when comparing the 6 athletes who had the closest adherence to the polarised and threshold plans, this difference was much more pronounced, with the polarised group improving by 7.0% and the threshold group only improving by 1.6%.
Another study (Stöggl & Sperlich, 2014) compared 9 weeks following a polarised training plan to several other training approaches, including the threshold model. The participants were 41 well-trained endurance athletes (~ national level). The polarised plan resulted in the largest improvement (11.7%) in VO2peak, whereas the threshold program resulted in an average 4% decrease in VO2peak.
Physiological Basis Of Polarised Training
Let’s look at some of the physiological reasons why a polarised approach might be beneficial.
In our view, the most compelling arguments are that:
By emphasising low-intensity training, development of the aerobic energy system is prioritised. This low-intensity training brings about adaptations within the muscles such as increased mitochondrial density and activity, improved capillarisation, and increased fat oxidation abilities, all of which contribute to improved lactate threshold (or FTP), and may also help to improve aerobic capacity. The strength of the aerobic system is the largest determinant of performance across all endurance disciplines, so it’s appropriate that this low-intensity work should be given due attention.
By minimising moderate intensity training, polarised training allows a higher overall training volume. Many aerobic training adaptations occur as a function of training duration rather than intensity, and so by maximising total training time, larger aerobic adaptations can occur.
Related to the above, by keeping intensity low most of the time, high-intensity training can be completed to a high quality. This is important, because some adaptations require very high intensities to be reached and sustained. For example, to bring about cardiovascular adaptations related to improved aerobic capacity, we need to be able to hit and sustain heart rates close to maximum. When a large portion of training is done at threshold, athletes are typically too fatigued to hit these high heart rates.
Some other arguments include:
Training in the Medium Zone can be very stressful due to the high amount of time spent with high lactate/acid levels (Seiler & Kjerland, 2006). It’s been argued that the adaptive responses to Medium Zone training are not sufficient to warrant this associated training stress. In other words, you don’t get much bang for your buck with Medium Zone training as you do with the Low and High Zones.
A polarised approach can help prevent non-functional overreaching or overtraining by moderating the amount of stressful high intensity training and allowing for ample time for recovery and adaptation between key high intensity workouts.
A polarised approach ensures training is varied, and incorporates a mixture of intensities. This avoids training stagnation, where all training sessions are ultimately very similar.
Our Experience
Anecdotally in our coaching and consulting experience, a polarised model has proven successful for the vast majority of our own athletes. This includes those who have lots of time to train (and importantly, lots of time to recover) and those who have limited time and/or lead stressful lives externally to training.
However, we do feel that sticking rigorously to a polarised approach at all times of the season is too restrictive, as we will discuss in the next section.
Should You Avoid 'Threshold' Training?
A common misconception with polarised training is that threshold training should be avoided entirely. This is perpetuated by the ’80:20’ mantra that’s often associated with polarised training.
However, this would be ill-advised…
If we look at the studies of training intensity distribution, we can see that approximately 5-10% of training sessions are done in the Medium Zone. That’s roughly one session every 2-3 weeks assuming one training session per day (or more often if training more than once per day). So, threshold training is clearly not eliminated entirely.
There are some very good reasons to do threshold/Medium Zone training:
Training at or just below the lactate threshold, particularly when combined with fasted training and low-cadences, can be an effective way to reduce the propensity for glycolysis, and thereby improve the lactate threshold. In particular, the high force demands of this type of training can bring about aerobic adaptations in Type IIa muscle fibres, which may not be activated in lower intensity training, particularly in time-crunched athletes who can’t ride for as long. Combining these sessions with carbohydrate restriction may be particularly beneficial in this regard.
This low-cadence, moderate intensity work can also help to improve muscular endurance (the point at which muscle fibres begin to fatigue) by placing a higher strain on the muscle fibres for an extended period.
Training at or just above the lactate threshold also seems to improve the ability to ‘shuttle’ (i.e. move) lactate to other parts of the body where it can be oxidised (i.e. removed). This also improves the lactate threshold, and can be particularly important for punchy disciplines where lactate is frequently accumulated after short high-power efforts, and must then be cleared.
Threshold training helps to familiarise an athlete with riding for an extended time at an intensity similar to race pace. It’s typically good to include a bit of race-pace riding before a race to build confidence riding at this intensity.
Threshold training can also be useful a “bridge” earlier in the season to help in the transition from training at lower intensities to higher intensities above the lactate threshold, especially for athletes who are new to interval training.
Pyramidal vs Polarised Training
More and more attention has recently focused on a similar training intensity distribution commonly known as a “pyramidal”, and many have asked if this is in fact the most optimal approach and better than the polarised model.
What’s interesting here is that both intensity distributions can often be one and the same!
The polarised model originally came from the observations of notable exercise physiologists and researchers like Stephen Seiler, who looked at the training intensities of top performers in cycling, rowing and cross-country skiing to name a few.
In these early stages when polarised training as a concept was in its infancy, observations were being made by using what Seiler describes as a “session goal” approach. This is where workouts are classified as low, medium or high intensity based on the main portion of the workout and/or the goal of the workout.
As an example, even though a workout featuring 4x8 minute intervals may also include a plenty of low intensity time too (in the warm up, the recovery intervals between the work intervals and a cool down), the overall workout would be labelled as high intensity due to this main block of work and the purpose/goal of the session.
Seiler and others using the “session goal” approach generally saw that for every 4x low intensity workouts, top performers completed 1x high intensity workouts, i.e. 1/5 were high intensity or in other words 80% low intensity and 20% high intensity.
However, looking at the same data but using an alternative approach, one we might call “time in zone”, where the actual second-by-second time spent in each training intensity zone is logged (now widely possible with mass market power meters, HR monitors and accompanying head units), a pyramidal intensity distribution is often seen, i.e. where most of the time is still spent at a low intensity (e.g. 60% total of time), a lower but still significant amount of time spent at a middle intensity (e.g. 30% of total time) and a comparatively small amount of time at high intensities (e.g. 10% of total time).
What’s important to note here is that both pyramidal and polarised training intensity distributions are generally spoken about in the context of the 3-zone model discussed above, where the separators are thresholds like LT1/VT1 and LT2/FTP/CP. Even when riding in the same zone (e.g. middle intensity, between LT1 and LT2) there’s a large range of power outputs or heart rates that exist within that zone. Riding at the bottom of the zone can result in a big changes is tolerable duration compared to riding at the top of the zone, and arguably can result in different adaptive responses too.
Intensity needs to be viewed as a spectrum, and not something which features highly distinct silos (where once a certain threshold is crossed, energy systems suddenly switch off and others turn on and wildly different physiological responses occur).
Periodising A Polarised Training Plan
A question we are often asked relates to how to periodise a polarised training plan – or in other words, how should the structure of the plan change between different training phases, such as the winter off-season versus the summer competition season.
This question follows nicely from the previous section, where we suggested you might want to include differing proportions of Medium and High Zone training at different times of the year. Indeed, many observational studies have noted differences in the proportions of Medium and High Zone training time at different phases of the season (Stöggl & Sperlich, 2015), and this is also something we adjust with our own athletes.
Exactly how to balance the medium/high-intensity work at different times of the season can be a little confusing. Some coaches tend to make training more polarised (i.e. the low intensity work lower, and the high-intensity work higher) as race season approaches. Others make training less polarised, incorporating more work at threshold as competition approaches.
Unfortunately, there’s no set rule – both of these approaches can be appropriate depending on the athlete and their race discipline.
The best approach to periodising a plan is to perform testing to understand your strengths and limiters (e.g. using power profile testing or lab testing). Then you can work backwards from your target races, to determine how much time you have to devote to different aspects of your fitness (such as VO2max and lactate threshold development). Repeated testing can then ensure you are progressing in the right way, and allow you to adjust your plan if needed.
Practically-speaking, we find it useful to use a six or seven zone model when thinking about training periodisation, as there is more discrimination between training intensities. In this sense, the goals of a given training phase might be to improve (i) muscular endurance, (ii) lactate threshold, (iii) VO2max, (iv) anaerobic capacity and (v) neuromuscular power, which (in a very simplified sense) would correspond to including training in zones 2/3, 4, 5, 6 and 7 respectively.
As an example, if a focus of a training phase was to improve VO2max, you might allocate ~20% of training sessions to interval sessions at Zone 5. Whereas, if the goal was to build anaerobic power, then these sessions would instead be allocated mostly to Zone 6 intervals.
It’s entirely ok to combine goals, provided that they are compatible. So, for example, building muscular endurance alongside developing the lactate threshold or improving VO2max would be compatible aims, and you could distribute the ~20-25% of sessions across these aims. So, for example, you might include 5% of sessions working at Zone 3 (muscular endurance), 10% at Zone 4 (Threshold) and 10% at Zone 5 (VO2max).
Combining the goals of improving the lactate threshold and anaerobic power at the same time would probably be incompatible goals, as one broadly requires a reduction in lactate production and the other requires an increase in lactate production.
In relation to training in the Low Zone, this should be present throughout all phases of training, and should remain at around 75% of all training sessions. However, depending on the training phase, the overall intensity and/or length of the sessions might change (e.g. shorter and lower-intensity sessions during the competition phase).
Polarised Training For Time-Crunched Cyclists
A trap that many athletes with restricted training time fall into is that they make all training moderately hard. This is often to ‘compensate’ for limited training time.
Many time-crunched athletes also worry that, with limited training time, a polarised approach won’t be effective due to the high proportion of low-intensity training.
However, the study discussed above (Muñoz et al., 2014) showed a polarised approach to be more effective than a threshold approach among recreational athletes. Crucially, the training volume was low in both groups (averaging 39.1H and 36.3H of training over the 10 weeks respectively).
From our own experience working with over 100 athletes that fall into the time-crunched camp, we’ve also found that a polarised model is effective when training time is limited (in the region of 5-8 hrs/week).
What’s often overlooked among time-crunched athletes is that there is very limited recovery time, life can be hectic, and sleep may be restricted. We think this is one of the key reasons that a polarised approach - where training sessions are only stressful when they really need to be – is a good approach for this type of athlete. Many training adaptations occur as a function of duration rather than intensity, so there’s little point in adding more stress and fatigue by riding harder when you don’t need to.
There are, however, some adaptations we would make to a polarised training approach with a time-crunched athlete, which help to make up for the lower overall volume of riding, and in particular the likelihood that these athletes can only do one or two long endurance rides per week:
We’d often include a slightly higher proportion of training time in Zone 3 (of a six/seven zone model), particularly done at a relatively low cadence (~60-75rpm) and around ‘sweetspot’ (i.e. ~85-95% FTP). This type of training puts an increased load on the muscle fibres, which can help to provide some of the muscular endurance benefits that a long ride would usually provide. We might include this in 1-2 sessions per week, depending on how well the athlete recovers from these sessions and how restricted the athlete’s training time is.
We often also include some fasted or ‘low glycogen availability’ rides, which help to simulate the metabolic conditions experienced towards the end of a long ride, where glycogen stores are depleted and the body is forced to use a higher proportion of fat as a fuel source. Again, we’d include these at most twice a week, and always combine them with low-intensity or muscular endurance rides only.
To balance out these broader concepts above, here are some more implementable tips for successfully training with a polarised model when training time is limited:
First, try to find time for a long ride somewhere in the week, so as not to neglect aerobic fitness. It’s usually easiest to include this kind of ride on a weekend, since this is when you’re most likely to have a few hours to string together. Long endurance rides really are the best type of training you can do to improve your aerobic base.
Next, you’ll want to be doing the most potent and stressful training sessions with the hour or so you have here and there throughout the rest of the week. The high quality sessions should include training at or above your lactate threshold. When scheduling these workouts, you want to try and make sure you’re rested before each hard session, where using a hard/easy approach to your week can work very well.
Using an indoor trainer for some of the sessions can be a good way to cut down the preparation time associated with a ride, and also to increase the quality of a training session by minimising dead time (e.g. coasting) and maximising time in the intended zone.
Stay on top of your recovery so that in the days off between your hard sessions, you can recover well and hit the next workout ready to go hard. Recovery is arguably more important for the time-crunched athlete given the typically hectic schedule outside training.
Calculating Polarised Intensity Distribution
A key point of confusion surrounds how exactly to calculate polarised training intensity distribution – should we be looking for 75% of training time to be low intensity, or 75% of training sessions?
The short answer to this is that it’s the number of training sessions that should follow the 75/5/20 low/medium/high distribution.
In this approach, you’d simply label every workout as low, medium or high intensity based on the intended adaptations and the main goal of the workout/session.
As an example, a workout targeting anaerobic power with intervals well above the lactate threshold would be classed as a high intensity workout, even though the time spent at high intensities within the workout would relatively low, since the work intervals are short, the recovery intervals are relatively long, and the warm up and cool down will also be mostly at lower intensities.
Achieving polarised intensity distribution then simply means about 1 in 5 workouts will be classified as high intensity, and about 4 in 5 workouts are low intensity. This offers a very easy and simple way to polarise training, but has obvious shortcomings in how all low and high intensities are handled the same.
An alternative and arguably easier way of tracking intensity distribution is using a time-in-zone approach. Training devices like power meters can track, to the second, how much time is spent at different intensities, which may be delineated in several different ways depending on the training zone system the athlete uses. Using training software like TrainingPeaks, it’s easy to look at exact intensity distributions over any time period desired.
In contrast to the approach of classifying sessions as high/medium/low, using a time-in zone method will often show far more time spent at lower intensities and less time spent at higher intensities (for the reasons discussed around warm ups, cool downs, recovery intervals etc.). Something like a 90% low, 3% medium and 7% high distribution would be typical of a polarised plan, although this would depend greatly on the goal of your current training phase. For example, if you’re including more threshold work, then this percentage would be higher and the percentage at high intensities would likely be lower.
Overall, the best approach is probably to look at your intensity distribution both at the level of sessions (to check you’re following the 75/5/20 distribution), and at the level of time in zones (to check you’re accumulating time in the intended zone e.g. Zone 5 if you’re working on VO2max).
Should Each Workout Be Polarised?
Another common question we receive is whether each workout should be ‘polarised’ individually.
In other words, should each training session include ~75% of training time at a low intensity, ~5% at threshold, and ~20% at high intensity.
While there may be some workouts that happen to possess this intensity distribution, we certainly wouldn’t recommend building workouts to try to achieve this type of distribution intentionally.
Trying to polarise every workout can needlessly constrain the design of training sessions, and can compromise the specific purpose of a given session. For instance, making 10-20% of a recovery-focused workout high intensity would diminish the workout’s ability to help the athlete rest. Similarly, an interval workout that restricted the high intensity to only 10-20% of the total time likely wouldn’t provide the high intensity volume necessary to cause an adaptive response.
Instead, a polarised intensity distribution should be assessed over a period of at least a week (if not longer), verifying for example that ~75% of training sessions over a 1-2 week period are low intensity.
Polarised Training Calculator
Whilst calculating your training intensity distribution is fairly straightforward, and simply requires looking at a broad distribution chart covering the time period you wish to analyse, a program we use and love called intervals.icu has a handy tool which analyses your training intensity and classifies it as either Polarised, Pyramidal, Threshold, HIIT (High Intensity Interval Training), Base and “Unique”:
Polarised Training Weeks
Time Crunched (5-8H/week)
At the risk of being too generic, here’s a polarised weekly structure that you might want to use as a basic framework for your training if you’ve got ~5-8 hours/week to train. This particular week’s intervals are focussed on VO2max development, but this could be adjusted depending on what you’re looking to achieve with a given training block.
MONDAY: Rest day (since you probably trained the previous day)
TUESDAY: 1 hour: VO2Max (e.g. 5-7x 3mins @ Zone 5 heart rate or power, with 3-4mins recovery between each. Warm up/warm down either side of intervals @ Zone 2).
WEDNESDAY: Rest day or light 30-40min spin @ Zone 2.
THURSDAY: 1 hour: VO2max (e.g. 3x 9M microburst blocks alternating between 30S ‘on’ and 15S ‘off’. The ‘on’ intervals @ Zone 6 power, or 9/10 effort. The ‘off’ intervals @ easy spinning or coasting. Recover for ~4 mins between each block. Warm up/warm down either side of intervals @ Zone 2).
FRIDAY: Rest day or light 30-40min spin @ Zone 1.
SATURDAY: 2-3 hours: Endurance ride @ Zone 2 heart rate or power, or 2-3/10 effort level.
SUNDAY: 1 hour: fasted muscular endurance intervals (e.g. 3-4x 8-9mins @ top end of Zone 3 or 6/10 effort, with low cadence (60-75 rpm). 3-4mins recovery between each interval and warm up/warm down @ Zone 2 either side of intervals).
Of course, the days can be switched around to suit your own unique circumstances and should be thought of as a guide rather than a rigid structure, with the high intensity sessions especially needing to evolve over time as discussed above.
Time Rich (10-16H+/week)
In contrast to the plan above, we’ve also included an example polarised training week for an athlete with a greater amount of training time available. We’ve still planned this to largely tie in with a typical weekly structure, where most training time availability is over the weekend.
Monday: recovery day (complete rest or 1 hour @ Zone 1)
Tuesday: 1.5-2 hours: VO2max (e.g. session as above, but with extended Zone 2 warm-up and cool down, and optionally more intervals depending on fitness level)
Wednesday: 2.5-3 hours: endurance ride @ Zone 2
Thursday: recovery day (complete rest or 1 hour @ Zone 1)
Friday: 1.5-2 hours: VO2max (e.g. session as above, but with extended Zone 2 warm-up and cool down, and optionally more intervals depending on fitness level)
Saturday: 3-4 hours: endurance ride @ Zone 2
Sunday: 1.5-3 hours: endurance ride @ Zone 2
Strategies such as muscular endurance work and fasted training can also be built into the endurance rides for a time-rich athlete, depending their fitness levels and ability to recover.
Summary
Polarised (and pyramidal) training has been shown to be an effective approach to intensity distribution among athletes of all levels of performance and time availability.
It presents a more sustainable, effective and time-efficient means of training distribution compared to the threshold model, and could really help you reach your cycling goals sooner.
However, the polarised approach doesn’t have to be strictly enforced at all times. If you take anything from this article, it should be that the most important aspect of polarised training is that you include a high volume of training in the Low Zone.
How you distribute the remainder of sessions across medium/high intensities is not too critical, and you can include a higher proportion of threshold training at different times of the season, depending on your own specific training goals.
Get Fast, Faster:
Free Workout Guide
Get your Key Workouts Guide; a free collection of 10 highly-effective, fundamental workouts you can use today to begin improving your endurance, threshold power, VO2max and other vital cycling abilities.
References
Billat, V. L., Demarle, A., Slawinski, J., Paiva, M., & Koralsztein, J. P. (2001). Physical and training characteristics of top-class marathon runners. Medicine & Science in Sports & Exercise, 33(12), 2089-2097.
Cerezuela-Espejo, V., Courel-Ibáñez, J., Morán-Navarro, R., Martínez-Cava, A., & Pallarés, J. G. (2018). The relationship between lactate and ventilatory thresholds in runners: Validity and reliability of exercise test performance parameters. Frontiers in physiology, 9, 1320.
Dantas, J. L., Doria, C., Rossi, H., Rosa, G., Pietrangelo, T., Fanò-Illic, G., & Nakamura, F. Y. (2015). Determination of blood lactate training zone boundaries with rating of perceived exertion in runners. The Journal of Strength & Conditioning Research, 29(2), 315-320.
Kenneally, M., Casado, A., & Santos-Concejero, J. (2017) The Effect of Periodisation and Training Intensity Distribution on Middle-and. International Journal of Sports Physiology and Performance.
Muñoz, I., Seiler, S., Bautista, J., España, J., Larumbe, E., & Esteve-Lanao, J. (2014). Does polarized training improve performance in recreational runners?. International journal of sports physiology and performance, 9(2), 265-272.
Rosenblat, M. A., Perrotta, A. S., & Vicenzino, B. (2019). Polarized vs. threshold training intensity distribution on endurance sport performance: A systematic review and meta-analysis of randomized controlled trials. The Journal of Strength & Conditioning Research, 33(12), 3491-3500.
Seiler, K. S., & Kjerland, G. Ø. (2006). Quantifying training intensity distribution in elite endurance athletes: is there evidence for an “optimal” distribution?. Scandinavian journal of medicine & science in sports, 16(1), 49-56.
Stöggl, T., & Sperlich, B. (2014). Polarized training has greater impact on key endurance variables than threshold, high intensity, or high volume training. Frontiers in physiology, 5, 33.
Stöggl, T. L., & Sperlich, B. (2015). The training intensity distribution among well-trained and elite endurance athletes. Frontiers in physiology, 6, 295.