Non-weight-bearing aerobic cross training

All forms of low-impact aerobic cross training provide the opportunity to enhance some aspects of fitness while reducing risk of injury due to lack of the impact at footfall.  As described in my previous post, cross-training that entails weight-bearing enhances endurance of the postural muscles that provide a stable core to support the driving force exerted by the legs.  For this reason, most of my cross-training is weight-bearing.  However under some circumstances, non-weight bearing training is preferable.

If the main goal is recovery, non-weight-bearing cross-training minimises stress on the musculo-skeletal system.    In particular, swimming has been shown to produce more effective recovery than complete rest after heavy training.  It is probable that when swimming the increased blood circulation promotes removal of the debris resulting from the muscle damage during heavy training.   If the goal is maintaining fitness while recovering from injury, avoidance of weight bearing might be essential in the early phase of recovery, and a judicious combination of weight bearing and non-weight-bearing training cross-training subsequently introduced as the injury heals.  Emily Infeld’s log of her training after suffering a hip stress fracture three months before the US Trials for the Rio Olympics is a very informative description of judicious integration of non-weight bearing and weight-bearing cross training.  Finally, non-weight bearing cross training such as cycling or swimming might be preferable simply because these activities are enjoyable and mentally stimulating.

Let us consider several of the most popular forms of non-weight bearing cross training in greater detail:


One of the great virtues of cycling is that with a few precautions and a gradual build up, you can cycle for many hours with minimal risk of injury.  It also offers great opportunities for an enjoyable alternative to running.  It is therefore a great way to develop cardiac output and endurance.   However, a crucial question is whether or not it is a good way to build up the leg muscles that are most important for running.   Like all other forms of low impact cross training, it does not condition the muscles to cope with the eccentric contractions at footfall.  Nonetheless, as in running, the main generators of power in cycling are the extensors of hip and knee, with a contribution form calf muscles in the final stage of the down-stroke.  At low and moderate intensities cycling has the potential for developing important features such as fat burning, capillary supply and the ability to shuttle lactate between type 2 and type 1 fibres, in these muscles, while at high intensity it can enhance their power.

However, the notorious challenge of the transition from bike to run in the triathlon raises the possibility that cycling might use these muscles in a different manner that is actually antagonistic to running.   Although I am not a triathlete, for more than sixty years the bike has been my major means of transport for local travel, including travelling home from work before an  evening training session and even more critically, for travelling to and from races. My experience leaves little doubt that in the short term, cycling does impede running.  What causes this and does it have implications for longer term influence of  cycling on running?  Unfortunately, I have never been able to get a convincing answer to this question in my discussions with triathletes, but I have developed some speculations based on my own experiences and on the anatomy of the hip and knee extensors.

The first point to note is that the range of hip extension differs between running and cycling.  When running, the hip extensors come into play in arresting the flexion of hip and knee in late swing (initially an eccentric contraction) and continue to act until late stance, by which stage the hip is extended beyond the neutral position and the leg is angled downwards and back.  During late stance, the hip extensors undergo concentric contraction, though the fact that the fact that the hamstrings cross both hip and knee complicates matters; we will return to the implications of this later.   When cycling the hip extensors are passively streched during the upstroke and contract concentrically throughout the down-stroke. The hip extension ends with the foot below the torso and the hip still slightly flexed. Thus, throughout the period of extension the hip is actually in a state of full or partial flexion.  Although the hip extensors are not subject to active stretching during the down-stroke, the extensors are nearly fully stretched at the beginning of the down-stroke and remain slightly lengthened relative to the neutral position at the end of the down-stroke.

Thus, despite the fact that cycling avoids the potentially seriously eccentric contraction that occurs during running, the hip extensors are at least somewhat lengthened relative to neutral and the flexors slightly shortened throughout the stroke, and there is a risk that protracted periods of cycling will lead to a tendency for shortening of hip flexors and stretching of hip extensors.  For a person with a desk job, this exacerbates the tendency for shortening hip flexors induced by hours of sitting.   Such a tendency for shortened hip flexors and stretched extensors might impede the extension of hip and knee in later stance that is crucial for powerful running.

However, the situation is actually a little more complex. The major hip extensors are the gluteus maximus and the hamstrings.  The long hamstrings cross both hip and knee, acting as flexors of the knee in addition to extending the hip.  During both cycling and running, the knee and hip extend together, so the length of the hamstrings changes relatively little.  The issue of shortened hip flexors and stretched extensors appears at first sight to apply only the flexors and extensors crossing a single joint (psoas and gluteus maximus).  However, the picture is even more complex, because the hamstrings, apart from the short head of biceps femoris, take their origin from the ischial tuberosity which is below and slightly behind the hip joint, exacerbating the tendency for stretching as the hip flexes, while most of the fibres of the long head of biceps femoris and semimembranosis are inserted only a short distance below the knee joint, such that knee flexion produces little tendency towards shortening.  The net effect of the location of the origin and insertions of the various components of the hamstrings is that both semimembranosis and the long head of biceps femoris are stretched passively during the hip and knee flexion during the upstroke of cycling.


Illustration of the passive stretching of semimembranosis muscle as the hip and knee flex to approximately 90 degrees. The muscle is about 6 cm shorter than the femur in the neutral position, but only 2-3 cm shorter during flexion. In this illustration the pelvis remains neutral. When cycling, forward lean of the trunk displaces the ischial tuberosity backwards, adding to the stretch of semimembranosis.

Overall, there is a tendency for shortening of psoas, a hip flexor,  and lengthening of gluteus maximus and two major components of the hamstrings. At least in the short term this will impede the powerful extension of the hip in late stance required when running.   If cycling forms a large part of training, it might produce a sustained imbalance between hip flexion and extension, resulting in sustained impediment of the hip extension crucial for powerful running.

I think there are two ways of minimizing the risk of impeding hip extension.  First of all, it is probably useful to stretch the hip flexors after cycling. Static stretching should only be done while the muscles are warm. I rarely engage in passive stretching. However I regularly do dynamic hip swings (standing on one leg, swinging the other leg forward and back, with knee extended, emphasising on a good back-swing.  If I ever do a triathlon, I will be inclined to spend 30 seconds in the second transition mobilizing the hip.  The second strategy for reducing the problem of dominance of flexion over extension is cycling at a high cadence.  This favours the development of muscle properties such as capillary supply and fat metabolism rather than the building up of powerful type 2 fibres, thereby reducing risk of developing a strong imbalance between flexors and extensors.

Yet another potentially important factor is the type and fitting of the saddle.  On my commuter bike I have a saddle that is wide enough to support both ischial tuberosities (the ‘sit bones’).  On one of the few occasions when I have cycled vigorously for a sustained period, I was amazed to find that that on dismounting I could scarcely walk, let alone run. It appeared that my hamstrings were partially paralysed.  This was almost certainly due to pressure on the upper part of the hamstrings, which are attached to the ischial tuberosities.  The problem was only transient but emphasized to the importance of a well fitted saddle of the correct width.

In summary, cycling is a potentially valuable form of cross training. It is possible to cycle for far longer periods than feasible when running; it is good for developing cardiac endurance and also for the attributes of skeletal muscles important for running in the aerobic zone, but it is necessary to avoid developing an imbalance between hip flexors and extensors. This might be achieved by cycling at high cadence and at doing dynamic stretching of the hip flexors for least a short period afterwards.


Over the years I have only swum sporadically, though for several months after I had injured the lateral ligaments of my left knee in a cycling accident last year, swimming became the mainstay of  my cross training.  I consider that the front crawl is the most useful stroke because the flutter kick and core strength required for a well-balanced position in the water help maintain the endurance of the gluteals and trunk muscles engaged during running.

Unless you devote some attention to swimming technique, front crawl can become an anaerobic activity (despite a relatively low heart rate). It is noteworthy that in the study led by Peter Peeling at University of Western Australia, in which a recovery session including 2 km of moderate intensity swimming produced more effective recovery than passive rest of similar duration after intense running interval sessions, the participants were triathletes.  I doubt that swimming would produce such a beneficial recovery in runners who were not technically accomplished swimmers.

For most people, and especially for male distance runners, the centre of mass of the body is near to the hips while the centre of buoyancy is in the chest.  As a result the body tends to rotate to a feet- down, head-up position in the water, increasing drag and tending to make swimming an anaerobic activity. The streamlined position necessary for a sustained aerobic front crawl requires a flutter kick and engagement of trunk muscles, actions which in themselves are directly beneficial to the distance runner.  I consider the Swim Smooth site is a very helpful source of guidance on front crawl technique.


Aqua jogging using a flotation belt for buoyancy, or deep water running, involve similar neuromuscular action to running, with zero or minimal impact.  However, the relative activity in the quads and hamstrings differs between different styles of deep water running, and is also likely to differ from ‘on land’ running.   For example, Mercer and colleagues demonstrated that when running at a stride frequency that the runners had self-selected during ‘on land’ running, activation of quads and hamstrings was lower during a high knees style of deep water running than a ‘cross-country’ style.  However, the high knees style produced greater activation of hamstrings that a body weight-supported treadmill with either 60 or 80% support, but similar activation of quads.  Furthermore, it is subjectively harder to achieve a given heart rate, and maximum achievable heart rate tends to be lower during aqua jogging or deep water running than when running on land.  This might be because of greater venous return of blood to the heart and consequently increased stroke volume, though I am not aware of direct evidence for this.

Overall the evidence indicates that aqua jogging or deep water running can produce useful gains in fitness in previously untrained individuals, and can help maintain fitness in injured athletes, but it should not be assumed to be very similar to ‘on land’ running  in either the relative activation of different muscle groups, or in cardiovascular responses.

In light of the greater perceived effort required to achieve a given heart rate, and also the potential for boredom, I consider that aqua jogging and deep-water running lend themselves better to interval style sessions, if the goal is to increase fitness. On the other hand, if the goal is recovery after hard training, aqua aerobics (perhaps best done in a group led by an enthusiastic leader and accompanied by lively music) might be enjoyable and relaxing.

Similar to the evidence that swimming can promote better recovery than passive rest after intense running, Takahashi and colleagues demonstrated that 30 minutes of walking, jogging and jumping in water daily for three days following a down-hill running session produced better recovery of muscle, evidenced by less soreness and stiffness, than were observed in a control group.


Body-weight-support treadmill

In the so-called anti-gravity treadmill, the lower body is encased in an airtight bag.  Air-pressure in the bag is increased thereby tending to lift you off the treadmill.  The principle is similar to aqua-jogging but with the advantage that the reduction in effective body weight can be set at any desired level from 0 to 80%.  In both principle and practice, this is can be an effective device for promoting recovery from injury, though accessibility is limited and the cost is probably prohibitive for use simply as a form of cross training.


Low impact, aerobic cross training is a useful way in which to increase volume of training, with beneficial effects on features such as capillary supply to heart and skeletal muscle, ability to metabolise fats, ability to shuttle lactate between type 2 and type 1 fibres, endurance of postural muscles and other aspects of fitness relevant to distance running.  It greatly reduces the risk of injury arising from impact at foot fall, but conversely, cannot enhance the ability to cope with the eccentric contraction of leg muscles at footfall that plays a cardinal role in getting airborne.  Furthermore different forms of cross training achieve the various physiological goals of cross-training to differing degrees.   The optimum choice between them depends on the specific training goals and on other circumstances.

In general, I favour weight-bearing cross training over non-weight-bearing on account of the benefits to postural muscles, bones and other connective tissues. In particular, I favour the elliptical cross trainer used in the hands-free mode, because it provides a very effective workout for postural muscles and upper body actions that are relevant to running.  However, many individuals find it boring and would prefer to be out-doors.

If you have a large amount of time available and enjoy being out-doors, walking, especially hill-walking is a good option for conditioning the legs.  Similarly, cycling is potentially a great form of cross training on account of the fact that, after adequate preparation, you can cycle virtually all day with minimal risk of injury.  However, as discussed in my speculative account of the differences in neuromuscular activity between running and cycling, I think prolonged cycling creates a risk of shortening of hip flexors and stretching of hip extensors, that might impede the hip extension in late stance that plays a key role in running.  This risk might be diminished by cycling with a high cadence and by regular hip-mobility exercises.

If the primary goal is promoting recovery from a hard session, or during the initial phases of mobilisation after injury, swimming, aqua jogging or aqua aerobics might be preferable.   Other devices such as the zero-runner or the anti-gravity treadmill are potentially useful because they allow a pattern of muscle recruitment that more closely resembles that of running. However limited accessibility and cost might be limitations for many runners.


8 Responses to “Non-weight-bearing aerobic cross training”

  1. malcolmbalk Says:

    What about cycling standing up…would it minimize the shortening of the hip flexors …you mention?

    • canute1 Says:

      Malcolm, Thanks for your comment. Standing up on the pedals does decrease the stretching of hip extensors and shortening of flexors, but does not abolish the problem entirely. Typically the stretch of semimembranosis is about 5-7cm when seated and about 3-5 cm when standing. (I have added a figure illustrating this)

  2. John Says:

    Good article, a great read. Interestingly, the abstract from the Peeling study speculates that a “swimming-based recovery session enhanced following day exercise performance, possibly due to the hydrostatic properties of water and its associated influence on inflammation.” As inflammation is part of the adaptation process post training stress, swimming recovery appears to permit greater performance in the very short-term but, potentially, to the detriment of longer term performance improvement?

    • canute1 Says:

      John, Thanks. Yes, swimming in cold water has the potential to decrease inflammation which is part of the mechanism by which the benefis of training are achieved. The question of the optimum amount of inflammation for optimum training benefit is a challenging one. It is porbaly best to avoid swimming in cold water

  3. Ewen Says:

    Thanks Canute. Malcolmbalk mentioned stand-up cycling “out of the saddle” and I think that’s a good variation to use in a cycling cross-training program for two reasons:
    1. The best place to do it is on a sustained hill where you’re working against gravity, so HR will usually be higher than seated cycling.
    2. The hardest push on the pedals is downward which is perhaps more similar to running. There is also ‘side to side’ movement of the body/core when cycling out of the saddle which could be beneficial.
    I agree that cycling “too much” (and too little running) in a x-training program can be detrimental to running performance. Not only the reasons you mention but also because there is no “springing off the ground” action for the muscles when cycling.

    BTW, saw a basic but interesting article on how lack of exercise affects aging muscles:

    • canute1 Says:


      Thanks for your comments about cycling and also for the link to the article about the general health benefits of maintaining the muscles in good condition.

      The good news is that regular moderate exercise is great for maintaining health though the passing years. For those of us interested in maintaining running ability into old age, the challenge is identifying what are the key physiological changes responsible for age-related deterioration in running, and finding ways of slowing these changes. Decrease in HRmax is one inexorable change. Loss of skeletal muscle power is also an issue, but it is not clear that arresting the loss of power is enough to arrest the loss of running performance.

      The main point of low-impact cross training is minimising accumulation of damage to muscle. Maybe the challenge is finding the optimum balance between stress and recovery for muscle in order to maximize running longevity.

  4. russtyd Says:

    what leads you to believe that the length of the muscle is a factor? hurdlers and steeple chasers are not significantly slower than other runners over the same distance. surely its a neuro/motor issue?

    • canute1 Says:


      Thanks for your comment. You raise an interesting point

      Hurdlers and steeplechasers produce greater hip flexion during the stride in which the leg reaches forwards to clear the hurdle, and unlike cyclists, this hip flexion is accompanied by less knee flexion than occurs during cycling, so there is an even greater tendency for passive stretching of the hip extensors on that stride than during cycling. However, on every stride there is an active contraction of hip extensors back beyond the neutral point and a corresponding stretching of the hip flexors in late stance, whereas cyclists do not extend the hip back behind neural at all. In fact even when the pedal is at its lowest point, the hip remains slightly flexed. The point I made in my post is that cycling creates a tendency for sustained shortening of hip flexors and stretching of hip extensors. However, hurdlers and steeplechaser regular both stretch and extend the hip flexors due to the larger range of motion at the hip, so a persisting imbalance is less likely.

      I agree that this imbalance is likely to be a largely neuromuscular recruitment problem, rather than a fixed deformity due to structural change such as collagen deposition..

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: