Revisiting the spinal flexion debate: prepare for doubt
Whether the spine should flex repeatedly and under load is an old debate that we are still having having and we should be having it because I don't think its settled. I thought it was settled 20 years ago but I should have challenged my biases more. The basic question is whether you are at less risk for pain/injury if you minimize the flexing movement of the lumbar spine during activities and if minimize the flexed position when lifting heavy. The case to minimize flexion is laid out in this article here detailing how the Canadian Military has removed the sit-up and trumpeted the death of the sit-up. Link here. I have probably taught 1000s of people to hinge at their hips and try to minimize spinal flexion during many activities...certainly activities that demand high loads on the spine...But is this right?
Would we see a dramatic difference in the prevalence of low back pain if suddenly the world attempted to minimize lumbar flexion during sports, exercise, lifting, sitting, picking up their kids or picking strawberries? Would injuries and pain decrease if the sit-up were banned and office workers were given electroshocks to stop them from slouching in their desks? If I take the advice of famous on-line strength coaches and rehab "experts" after their most recent weekend seminar that takes them from spine stability neophyte to stability expert faster than a certain ring changes Smeagol to Gollum then I'm going to scare the flexion out of all my patients immediately. But its never that simple and I'm hoping to share the multiple viewpoints that the research provides us.
Disclaimer of Debate and Doubt
I don't have a lot of answers here. I'm not going to criticize researchers or clinicians espousing any particular side because I've held all those positions too and nothing is black and white here. The point of this incredibly long post is to try and present the two sides to this debate. My bias is still to try to avoid spinal flexion under many conditions while at the same time caring very little about spinal flexion under other conditions. We will even discuss why the following video should not worry us or perhaps should worry us :).
[video width="426" height="426" mp4="http://www.greglehman.ca/wp-content/uploads/2016/01/full-spine-flexion-with-weight-einar.mp4"][/video]
Because this post is pretty long I will give you the Bottom Line First. We have two camps: Try to minimize flexion and Flex Away.
The Quick Case for Neutral
- great research by great biomechanists (Stu McGill, Jack Callaghan) have documented a potential Disc injury mechanism for repeated spinal flexion under low loads in in-vitro models. Meaning, if you want to damage a disc you'd better add some repetitive flexion and extension. Also supported by the work of Wade here. I should also stress that their recommendations for exercise and functional tasks are not only based on in-vitro models but also biomechanical modelling and epidemiology.
- repeated flexion is cumulative and we assume that the discs have a finite number of flexion cycles. That finite number will certainly vary across people but we can't measure someone's adaptability or estimate their tolerance
- biomechanists have also shown that the compressive strength of the entire disc and vertebra unit is stronger in neutral postures (links here)
- lifting from low heights (we assume that this requires a flexed spine) can increase the prevalence of low back pain (plenty of research on this but here is one link here)
- the great work of Solomonow showing that spinal flexion induces ligamentous creep and influences the role of the spinal musculature. The link to pain is unknown.
- great biomechanical models (they typically measure muscle activity, spinal posture, joint positions and then try to calculate the loads on the lumbar spine) show that lifting with a flexed spine versus a neutral spine leads to comparable levels of compression but increases in the amount of anterior shear (link here and here). Anterior shear and cumulative compressive loading is often linked with injury.
- there is NO research showing that a flexed spine is stronger in in-vitro studies or epidemiological studies when it comes to injury or pain.
- a neutral spine is often associated with a more hip dominant movement strategy which may be beneficial for performance in many activities.
So for those that knock the research supporting Neutral because "its all done on dead pig spines" they need to overdose on bacon (not a bad way to go). The in-vitro studies (pig spines) are one piece of a much larger and robust argument.
The Quick Case for Flexion
- some in-vitro studies show that disc herniation is possible even in a neutral spine. This suggests that Neutral alone is not wholly protective (link here)
- a neutral spine is not protective of other injuries (e.g. the vertebral endplate can fracture) and the loads where failure occurs are comparable between neutral and flexed position (link here)
- other biomechanists using different biomechanical models document less loading on the spine (specifically anterior shear) when lifting with a stooped/flexed back posture versus attempting minimize flexion (link here)
- repeated spinal flexion might be largely unavoidable. While we can advocate avoiding the sit-up for many reasons (performance choices being one) the alternatives may not even better when you compare the loads on the spine. Did you know that the Kettlebell swing will see an average of 26 degrees of lumbar flexion (link here). Or squatting, when trying to be in neutral, can see 40 degrees (link here)?
- flexion in sport may also be unavoidable. Think golf (about 48% of max flexion repeated 1000s of times a day), rowing, throwing, kicking, cycling, landing when snowboarding, skiing etc. Sports which are flexion biased do not have greater degrees of low back pain versus those in neutral (link here)
- if you are worried about the loads in a sit-up or the loads on the spine when lifting with a flexed spine you need to be worried about doing everything. Neutral is not protective against loads. The kettlebell swing again shows comparable compression and shear to the sit-up.
- disc herniation might be the least of our worries. Disc herniation is estimated to be involved in low back pain 2-5%.
- "damage" in the spine is poorly linked with pain. Loading of the spine is not very well linked with "damage" or degeneration. Our recommendations to avoid postures that lead to an assumed greater degree of loading may therefore have little influence on the pain prevalence.
- the body adapts: repeated flexion and repeated loading may be a good thing for the spine and the disc. Its not infinite and we can't measure people's tolerance and adaptability so are a bit in the dark here.
- a flexed spine during lifting is both metabolically more efficient and is often associated with greater force output (i.e. when you see someone lifting their best deadlift ever, does their spine stay in neutral or does it typically flex more?)
...and now some details.
The Dominant View: A Neutral Spine is a Safer Spine
Below I will outline the case to avoid spinal flexion under certain conditions and present some of the literature that favour a neutral spine.
A documented disc herniation injury pathway exists
Callaghan & McGill (2001) made the a laboratory observation that it was very difficult to herniate a disc when compressed with an in-vitro set up when the spine was compressed in a neutral posture. While compressing a motion segment (disc attached to the adjacent bones) they were able to induce posterior disc herniation with low levels of compression (260, 867 and 1472 Newtons) coupled with repeated cycles of flexion and extension. The amount of flexion motion was restricted to the neutral range. Meaning the motion segment flexed until the torque-angle curve started to deviate from neutral. The authors note that this was about 35% of the maximum flexion range of the motion segment. The authors found that in the low compression group 1/5 motion segments showed a disc herniation the others surviving 86400 cycles of movement. The likelihood of herniation increased as compression increased.
Wade et al (2015) looked at the influence of compressive loading rate and spinal flexion on the strength of ovine motion segments in an in-vitro model. The authors loaded the motions segments either slowly (2mm/min) or rapidly (40 mm/min) in neutral or flexed posture. The flexed posture was 10 degrees which was viewed as the maximal flexion angle of the ovine motion segment. The following injury types were found:
- Flexed with low compression rate: only endplate fractures occurred.
- Neutral with low compression rate: only endplate fractures occurred
- Flexed with high compression rate: endplate fractures and posterior disc wall failures (50% of specimens)
- Neutral with high compression rate: only endplate fractures.
Interestingly, there was no difference in the total amount of compression required to cause endplate fractures whether flexed (10900 Newtons) or in neutral (11500 Newtons). However, the average load to induce disc herniation (8900 Newtons) was less than the average load to induce endplate fracture (10900 Newtons) regardless of posture.
Both of these studies suggest that either repeated flexion or flexion and compression under rapid loading is required for disc herniation. Clinical and exercise recommendations from these studies recognize that many people will perform thousands of sit-ups in their lives, will flex their spine under load with abandon and will not have pain. It is assumed that they adapt and they have a spine that tolerates bending. However, we can't guess who has this ability so it is not unreasonable to look at solely at this research and think that the safest bet may be to avoid repeated flexion and compression.
You can also argue that why would anyone want to bother with a sit-up or lifting with a flexed spine? If there is slight injury risk and no performance or fitness benefit then choosing exercises or techniques that attempt to maintain a neutral spine is prudent and beneficial for performance. The caveat here is whether you are scaring people and creating "pain expectation". An individual's spine can be robust but that doesn't mean they might not be sensitive. Expectation (e.g I've been told to not flex my spine because it creates injury) can certainly sensitize some people. Expectations have been linked to poorer outcomes and greater pain (Benedetti et al 2007, Bialosky et al 2010, Lurie et al 2015). And just like many people can flex their spine's as much as they want many people can also not be sensitized and subject to the Nocebo or fear that might be created with advice to avoid flexion.
The entire motion segment might be stronger in Neutral
In contrast to Wade et al (2015), who showed that regardless of flexion the endplate failed at the same amount of compression other biomechanists have ashown that the compressive strength of the entire disc and vertebra unit is stronger in neutral postures (links here). Gunning et al (2001) using a porcine motion segment in-vitro testing model evaluated the ultimate compressive strength of motion segments in neutral and flexed postures (22 degrees, considered full flexion) after having been compressed at 1000 Newtons for 3 hours in either neutral or flexed static positions. The authors found that the average ultimate compressive strength of the motion segment when compressed in neutral was 12, 109 Newtons (S.D = 1703) whereas the compressive strength of the motion segment when compressed in a flexed posture was statistically less at 10118 Newtons. Of interest, was the lack of posterior disc herniation under all conditions.
Trunk bending is related to occupational low back pain.
Lifting from low heights (we assume that this requires a flexed spine) can increase the prevalence of low back pain (plenty of research on this but here is one link here). The limitation of this research is that it doesn't measure spinal flexion. It just measures that the entire trunk was tilted relative to vertical. It is unknown if the forward bend comes from the hips, thoracic spine or lumbar spine. Only that bending forward someway is related to low back pain incidence
Spinal flexion induced Creep influences lumbar function
The great work of Solomonow showing that spinal flexion induces ligamentous creep and influences the role of the spinal musculature. The link to pain is unknown.
Biomechanical modelling shows increased loading with spinal flexion
Great biomechanical models (they typically measure full body kinematics, ground reaction forces, muscle activity, spinal posture, joint positions etc and then try to calculate the loads on the lumbar spine) show that lifting with a flexed spine versus a neutral spine leads to comparable levels of compression but increases in the amount of anterior shear (link here and here). Anterior shear and cumulative compressive loading is often linked with injury (Coenen et al 2013) thus minimizing compression and shear via alterations in spinal posture are often rationalized when teaching lifting techniques. Increased loading via increased spinal flexion has been shown in other biomechanical labs in addition to the ones previously mentioned. Professor Shirazi-Adl has also documented this finding.
Does evidence exist to make recommendations to load the spine in a flexed posture?
There is NO research showing that a flexed spine is stronger in in-vitro studies or epidemiological studies when it comes to injury or pain. This is where I really understand the argument to avoid flexion. There is a lot of uncertainty. Without there being a compelling case to encourage spine flexion during activities of daily living or exercise it is not unreasonable to argue avoiding it if you think there is a slight risk of future injury.
A neutral spine may be better for performance in many tasks.
A neutral spine is often associated with a more hip dominant movement strategy which may be beneficial for performance in many activities. I won't get into this but I respect coaches and their wisdom. We don't see people doing maximal clean and jerks or snatch lifts with massive degrees of spine flexion.
Neutral is the dominant view and clinical expertise has a role in guidelines
Neutral is the view most espoused by experts. I realize the limits of expert opinion but we can't know everything and its not unreasonable to listen to the experts. I respect them and want their opinions.
But this post is about looking at all sides. Someone else could probably provide a stronger and more extensive case for Neutral and that certainly exists out there.
Below I want to provide the case for allowing more flexion in our lives.
The Case For Flexion
At its simples the argument is that flexion is not so bad...or it is not really an independent risk for injury under the majority of cases. Like many risk factors it is probably mediated by a number of factors. For example, low load flexion (e.g bending to pick up a ball, leaning over a chair, prolonged sitting) could certainly pose much less of a risk for injury/pain when comparing ballistically picking up an extremely heavy object with a fully flexed spine. That is the nuanced view that doesn't vilify spine flexion under all conditions. One can also take a much more extreme view and suggest that spinal flexion is no more of an injury risk under ANY condition. Spinal flexion might be a complete red herring and that the factors leading to pain/injury will exist regardless of the posture of the spine.
Lets look at the research that argues that spinal flexion is perhaps no more risky than a neutral spine for injury and pain. On purpose I have tried to stay in the mechanical realm and have avoided arguments related the psychosocial aspects of pain. They play a large role too but I find it interesting when biomechanical research often questions other biomechanical research.
Discs can herniate even in Neutral Conditions
Veres et al (2010) showed that a neutral spine is not wholly protective against disc damage. This is significant when we recall the earlier work of Callaghan and McGill (2001) suggesting that repeated flexion-extension is a mechanism for disc herniation. This papers suggests that flexion-extension is not the only mechanism to create herniation and maintaining neutral isn't sufficient to prevent a disc herniation under certain conditions (link here). Veres et al (2010) pressurized discs internally both slowly and rapidly and were able to document that disc herniation occurred even in neutral postures.
What the data suggest is that even if lifting heavy loads and you attempt to minimize your injury risk with a neutral spine you can still damage structures (in In-Vitro models) at comparable compressive loading regardless of whether you avoid flexion. What you see in the above table are the number of injuries that occurred in the motion segments when loaded to different amounts of compression (%UCT), under different repetitions (cycles/min) and at what degree of flexion expressed as % of the Neutral Zone (300 versus 100 % of NZ). Remember the neutral zone is that range where there is little resistance to movement. 300% indicates a lot of flexion. What we see is that even at 100% of NZ bending End Plate Fractures occur when loaded to 40% of the Ultimate Compressive Strength. Conversely, we also see zero injuries when we stay in neutral and load high repetitions at 40% of UCT but when you flex the spine there are more injuries.
Further, the authors did a histological analysis of the motion segments that did not fail. In essence, they looked to see if subtle injuries were occurring. They reported:
However, a noticeable effect of posture was observed across some conditions, with increased damage in tissue samples fromthe 300% NZ range condition. As expected, tissue samples excised fromthe posterior–lateral region were also found to have more disruption (i.e. clefts, fissures, etc.), in comparison to those excised from the anterior region of the IVD.
Confused? So, neutral is not wholly protective but perhaps there is a trend that flexion is still more injurious.
Different biomechanical models show different spinal loading
Other biomechanists using different biomechanical models document less loading on the spine (specifically anterior shear) when lifting with a stooped/flexed back posture versus attempting minimize flexion (link here). Kingma et al (2010) looked at spinal loading when lifting a large load. They had people lift stooped (full flexion), squat style and a "weightlifters" squat.
What Kingma et al (2010) found was interesting. With their biomechanical model they showed that stoop lifting had less anterior shear than lifting with a style that attempts to have a neutral spine. Naturally you will still notice that there is flexion in the spine. You can't avoid it even when you try. If you flex your hips your spine will go with it. See below the shear levels and spinal flexion values.
Shear Values Below
Spinal flexion might be largely unavoidable.
As evidenced in the above graphic from Kingma we see lots of spinal flexion even when trying to squat lift. We see between 45-52 degrees of flexion. This is comparable to what we saw in the Potvin and McGill (1991) paper that first looked at shear values in stooped versus neutral postures. Even in those lifters who tried to maintain neutral we say more than 40 degrees of flexion. Graph below.
If we can't avoid flexion during activities what are the implications?
While we can advocate avoiding the sit-up for many reasons (performance choices being one) the alternatives may not even better when you compare the loads on the spine. Did you know that the Kettlebell swing will see an average of 26 degrees of lumbar flexion (link here). The point being, if you are concerned about spinal flexion under load it may be impossible to avoid.
Flexion in sport may also be unavoidable. Think golf (about 48% of max flexion repeated 1000s of times a day), rowing, throwing, kicking, cycling, landing when snowboarding, skiing etc. Sports which are flexion biased do not have greater degrees of low back pain versus those in neutral (link here)
A very interesting Observation - you may not be able to avoid the assumed injury pathway to herniation
Remember that paper by Callaghan and McGill (2001) that showed that repeated flexion-extension is able to regularly create disc herniations in a porcine in-vitro model? It was an important paper because it establishes a potential injury mechanism. But there is something in that paper that I never noticed until recently. The amount of flexion was only to the end of the Neutral Range. The authors reported that this was approximately 35% of the maximal spinal range of each motion segment. Do see the implication here? Follow the logic here:
- repeated low-load flexion-extension movements going to the end of neutral range can create disc herniation in in-vitro models
- therefore we it is argued to avoid exercises or movements that do this
- therefore we should avoid sit-ups
- therefore we should avoid other exercises with repeated flexion-extension movements that take the spine to the end of neutral
- we should determine what other activities take the spine to the end of neutral?
From the previous sections we see that Kettlebell swings, squats, jumping, skiing, running, golf, throwing, kicking, sitting etc all take the spine to the end of its neutral range. The flexion movement that was documented to create a disc herniation might be almost unavoidable even in the tasks we are advocating instead of sit-ups. We could even expect that amount of load and that amount of flexion to occur in almost every sedentary individual who sits, walks, lies down, puts their shoes on, takes their socks off etc.
So, we have to ask. How common are herniations. Does everyone have them?
Disc herniation might be the least of our worries.
Disc herniation is estimated to be involved in low back pain 2-5%.
People have changes on their MRIs and other imaging but it is quite often poorly linked with pain. Disc herniations can also resolve on their own and nor do herniations explain back pain very well. Heck, people have surgical repairs of their herniations, re-herniate in the future and don't have pain. Herniations with neural compromise are better at explaining leg pain but quite poor at explaining back pain. (link here). However, in other papers we see a slightly different story. Disc herniations alone aren't related to pain but herniations with neural compromise are better related to pain. This is seen in the Boos et al (1995) study which showed a greater prevalence of disc herniation with neural compromise in the asymptomatic group. And when you have that neural compromise the pain is most likely explained by some type of neuroimmune response and a chemical mediator. Meaning pain becomes more about sensitivity rather than damage. We can have the damage but its only painful if there is sensitization. This is why so many other variables influence pain - because so many other variables influence sensitivity.
"Damage" in the spine is poorly linked with pain.
Loading of the spine is not very well linked with "damage" or degeneration. And load as a driver of degeneration has been questioned in the Twin Spine Studies. From Videman et al (2010)
Contrary to common beliefs, our findings suggest that cumulative or repetitive loading because of higher body mass (nearly 30 pounds on average) was not harmful to the discs. In fact, a slight delay in L1-L4 disc desiccation was observed in the heavier men, as compared with their lighter twin brothers
Therefore, our recommendations to avoid postures that lead to an assumed greater degree of loading may have little influence on the pain prevalence.
The body adapts
The repeated flexion and repeated loading may be a good thing for the spine and the disc. Its not infinite and we can't measure people's tolerance and adaptability so we are a bit in the dark here. This one is tough clinically for me. Because we really don't know how much flexion might be a good thing and how much might be sensitizing. This is obvious an individual limit and would be difficult to predict.
A quick summary of the case for flexion
- discs can herniate even in neutral
- other structures can be damaged even in neutral
- some studies suggest that ultimate compressive strength is not influenced by flexion
- the amount of flexion movement documented to create herniation may be unavoidable during many activities
- disc herniations may poorly explain the prevalence of low back pain
- the disc can adapt to flexion
Summary and a Recommendation
Thanks for making it this far and get ready for a murky and luke warm recommendation. I respect clinicians who take both positions on this issue. It is obviously not cut and dry. My bias through the years has been to go Neutral especially under heavy load conditions and I would stay with that recommendation for many reasons, the least of which being injury risk.
I would suggest that performance goals and symptoms should drive our clinical and movement strategies. I will break the suggestions down into different scenarios.
Repeated Low Load Activities
Suggestion: No flexion fear
These are things like bending over to tie your shoe, fluffing your pillows, sitting etc. The spine can handle bending. Its built for it. Don't freak out over flexion in these cases.
Repeated Flexion based activities
Suggestion: No flexion fear (perhaps my most contentious statement and I completely respect the opposite view)
If you love doing sit-ups, have done them for years and you don't have low back pain I would be hard pressed to tell someone to stop doing them. Being too worried about this movement is a little like "stranger danger" and never letting my kids out of my sight. I don't doubt that getting abducted is possible and it would be horrible - I just don't view it as that great of a risk. Same things goes with repeated spinal flexion under load. There are just too many sports and activities that have this movement that it just seems unavoidable and the potential injury seems to contribute to only a small percentage of low back pain cases. So to exclude the sit-up seems unnecessary. However, there are plenty of other reasons to avoid it. Surely there a better exercises than the sit-up?
If you are starting a training program and the sit-up can't be justified based on performance goals then why would you do it? And if you have flexion related pain then...
Flexion Related Pain
Suggestion: avoid flexion - at least temporarily
This is the easiest one. If it hurts don't do it. Calm shit down, build shit back up. If you have patients who consistently aggravate their pain with flexion related activities then this is the time to start training neutral. Or if they are in sports that require high degrees of repeated flexion then "sparing" the spine during their training is absolutely important. These could be the cases where the nociceptive driver is a disc with some nerve root involvement so avoiding the aggravating activity is helpful initially provided you also address the multidimensional nature of pain and all the contributing sensitizers. However, we can't create a long term fear of that movement. We teach that avoiding flexion is temporary measure to allow them to desensitize. And part of treatment is to slowly re-introduce some flexion over time. That is the part of "build shit back up".
These might also be cases where other "impairments" are relevant. For example, you might work with a rower who is consistently aggravated by repeated flexion. When looking at their rowing technique they appear to be more flexed than their team mates. They might also have less hamstring flexibility than average. Here you have a person who is sensitized to flexion, whose sport requires flexion and whose physical parameters dictate that they need more spine flexion to do their functional task. None of those previous factors are inherently injurious or faulty but under certain circumstances they contribute to the sensitivity. Addressing them in the person is relevant but might not be in an individual under different circumstances.
Heavy Load Activities
Suggestion: try to maintain neutral
Even though some research suggests that flexion may not be worse than neutral there is no research showing it is safer. I think we can certainly listen to our superstar biomechanists here. High load activities are examples where biomechanics are important. We are dealing with loads that might exceed the structural strength of the tissue. Psychosocial variables may have a smaller or non-existent role to play here - at least with the injury mechanism. I'm comfortable choosing the Neutral position here. You can also justify this based on performance with many high load activities and high performance activities being associated with a hip dominant movement strategy.
Well thats enough. I reserve the right to change my clinical opinion on all of this with new information as you should as well.