After showing the squat a little too much love in Strong Science #1 and #2, I’ll turn my attention to the deadlift. I’ve got two classic studies that I’ve mentioned briefly elsewhere (here and here), but I’ll talk about them again here because they don’t get anywhere near the attention they should.
But before I dive into the specifics, it’s worth noting the efforts the researchers of the first two studies (1,2) have made to create scientifically valid data that is applicable in the gym. In doing this kind of research, there’s usually a trade-off when it comes to methodology. Lab-based studies with the strictest controls performed to rigorous scientific standards lose both appeal and applicability to those interested in elevating their performance, as the conditions are considered too artificial or the participants are usually of too low an initial training status. On the other end, field studies often capitalize on more advanced lifters but these studies don’t always have access to the fancy equipment found in labs, so crude measures are observed compared with what is capable within the lab. In this case it’s not that one methodology is preferred over another but that both are a necessary evil to grasp the bigger picture. In reality, it’s up to us to interpret these various studies, reconcile them with our personal experiences in the gym, and to come up with the best conclusions we can with the given information.
The two studies we have here strike a good balance between the scientific and practical domains of performance. The first, a field study that’s a basic biomechanical analysis of deadlift technique capitalized on the 1989 Canadian Powerlifting Championship for footage. The second study takes similar competitive lifters and moves them into the lab for more ‘intensive’ investigation. In this case the lifters are doing what they do (the deadlift) and we get the benefit of actually being able to view their individual lumbar vertebrae at the same time. Definitely the best of both worlds.
Lumbar spine loads during the lifting of extremely heavy weights
Methodology diatribes aside, in this study the researchers used competitive powerlifters and observed them as they actually competed. Because this was a sanctioned meet (1989 Canadian National Powerlifting Championship), no EMG electrodes or biomechanical markers could be used to facilitate the analysis, so landmarks on the body were estimated. Nevertheless, researchers taped all successful attempts of male and female lifters and captured many attempts of both sumo (wide stance, arms within knees) and conventional style (narrow stance, arms outside of knees) deadlifts.
An unfortunate and often confusing aspect of these studies is that when forces are calculated in these ‘extreme’ lifts, they can exceed what is considered or known to be the limit of the tissue of interest. Even though many fear mongers will twist the research to support arguments that full depth squats are bad, or that deadlifts will cause your spine to shatter, it really suggests that there is probably more to it than our models can account for. Are the estimates provided by these models deficient, or is this tolerance to increased loading the result of years of hard training? The fact that lifters everywhere subject themselves to these forces on a regular basis suggest that something is missing here.
Putting the absolute forces aside, the data does contain a valuable comparison between conventional and sumo deadlifts. The sumo deadlift resulted in a reduced moment at L4/L5 and reduced joint shear as well. This isn’t too surprising as this technique allows for the bar to be kept closer to hip and lumbar spine, reducing the moment arm for the barbell at these joints. This is in agreement with later studies that have demonstrated pretty similar mechanics (3). While switching to sumo may be advantageous to an individual lifter based on their body proportions (4), in this particular competition there was no difference between the total weight lifted between those who selected a sumo or conventional technique.
But the best lesson from this paper was that there is no universal style of deadlifting. The authors attempted to predict deadlift performance with a multiple linear regression, however because of the varying levels of hip, knee and lumbar moments, no discernible pattern appeared, stated best by the authors:
“The best deadlifters did not necessarily exhibit the highest moment about any particular joint; thus, no preferred strategy of lifting was evident on this or other variable analyzed in this study”
Looks like there’s a little bit more to the deadlift than simply picking a bar up off the floor.
Lumbar posterior ligament involvement during extremely heavy lifts estimated from fluoroscopic measurements
Cholewicki and McGill (2) recruited four powerlifters that were actively competitive (qualified for the national championships) to perform three trials: an unweighted full flexion trial followed by two near-maximal deadlift attempts. All trials were filmed with a fluoroscopic imager that allowed for direct, sagital viewing (from the side) of the lumbar vertebrae to determine the amount of flexion during the movement and use this data to predict ligament loading. The paper doesn’t explicitly state whether they were performing sumo or conventional style, however the diagrams in the paper lead me to believe that they used the conventional technique. Don’t worry about radiation either, these guys lifted with a lead jock on, good for stopping x-rays and any stray baseballs that might have been flying around the lab!
While the paper goes into intense detail about the length change of each vertebral ligament during the deadlift, it’s sufficient to say that most ligaments retained lengths less than what was achieved in the unloaded, full flexion trial. The lifters performed each deadlift with the appearance of full lumbar flexion (visually assessed), however the fluoroscopic measurements found that in all but one trial, the degree of intervertebral flexion was less than the maximum range of motion. From a technique perspective, it seemed that this degree of flexion was more or less set at the beginning of the lift. The authors noted that a small amount of flexion did occur when the lifters brought their hips through to complete the ‘lock-out’ phase of the lift, as the pelvis is brought under a relatively rigid spine.
The authors conclude that the main function of the lumbar ligaments in heavy lifting is not to support the load itself (which was the debate at the time), but rather act to limit range of motion between vertebral levels. From a practical standpoint what’s important here is that even though the lifters used what appeared to be a fully flexed lumbar spine to lift the weight, the imaging analysis showed they actually didn’t achieve full flexion at each lumbar vertebral level. That’s not to say that you can deem the flexed-lumbar spine deadlift universally safe, but it raises the question of how these lifters develop the awareness to flex their spines just enough to promote optimal mechanics for the lift without compromising the integrity of the ligaments (and intervertebral disc) at each lumbar vertebral level.
Given the level of debate between the lumbar flexion and anti-flexion camps, even though this paper is from the early 1990s, anyone on either side of the debate should take the time to read this paper as Dr Stu McGill posits some interesting points on the potential advantages of a moderately flexed but rigid spine in the deadlift.
A biomechanical analysis of straight and hexagonal barbell deadlifts using sub maximal loads
Seeing the high forces involved in the deadlift study mentioned above (1), it’s not surprising that someone took the time to develop a bar that could minimize these forces by placing the lifter in a more upright position, with the load positioned closer to the hip joint and lumbar spine. And so the trap bar was created. Originally named for it’s trapezoidal appearance (now most are hexagonal), this bar has established its place in most gyms, and is usually the deadlift variation of choice for new lifters that are intimidated by the straight bar deadlift. Many wouldn’t question that the bar is different than a conventional deadlift however an analysis of whether or not this bar actually achieves what it’s proposed to do hasn’t happened yet. Until now.
Swinton and colleagues (5) recruited 19 male powerlifters and had them perform conventional deadlifts with either a hex (trap) bar and straight bar at varying percentages of their max (10-80% 1RM) in a randomized order. All lifts were performed on a force platform and twelve markers were placed to mark various anatomical landmarks from the spinous process of C7 down to the ankle. Cameras recorded all attempts and three dimensional movements of the bar were tracked through each attempt.
The hex bar deadlift was found to increase the moment at the knee and reduce the peak moments at the hip and lumbar spine as compared to the straight bar deadlift. Differences in the path of the bar resulted in the load being kept behind the knee with the hex-bar, opposed to the straight-bar deadlift where the load was in front of the leg. Bar velocity was higher at each increment of %1RM with the hex bar between 40-80%1RM which resulted in higher power outputs from 30-80%1RM as well. All lifters also had a higher 1RM with the hex bar as well.
There’s no question that the hex-bar deadlift is the easiest of the full ROM deadlift variations to teach beginners. After reading this study, I suspect the hex-bar is almost more like a squat than a deadlift given the increased loading at the knee, and you may be short changing your new clients if you rely too heavily on the hex-bar as a deadlift variant. I know many clients prefer the hex-bar, and in these cases you may want to spend more time focusing on additional posterior chain exercises emphasizing the hips and lumbar region if this is your deadlift of choice.
The continuum of lumbar loading
It’s clear from this small collection of papers that we can significantly alter the moments and the required forces at the hip and lumbar spine by making modifications to both the technique of the exercise (conventional vs sumo deadlift) and equipment (straight or hex bar). In this case, if your goal is to minimize effects on the lumbar spine while still keeping the load anterior to the knees, the sumo deadlift would be preferred to the conventional. You can achieve even lower hip and lumbar moments with the hex-bar, although this will result in a shift in the load to behind the knee, more resembling a squat than a deadlift. Those in the later stages of recovery from back pain could employ a multi-lift approach, moving from less aggressive lumbar loading (hex bar, sumo deadlift) before returning to conventional, straight bar deadlifts.
- Cholewicki, J et al. (1991). Medicine and science in sports and exercise, 23(10), 1179–1186.
- Cholewicki, J, & McGill, SM (1992). Journal of biomechanics, 25(1), 17–28.
- Escamilla, RF et al. (2000). Medicine and science in sports and exercise, 32(7), 1265–1275.
- Hales, M. (2010). Strength & Conditioning Journal. 32(4), 44-51.
- Swinton, PA et al. (2011). Journal of strength and conditioning research 25(7), 2000–2009.