There seems to be a lot of buzz in the fitness community recently about how situps and crunches are bad for people's backs.

For example, Harvard Medical School makes the claim, but they don't cite any scientific studies.

One reason is that sit-ups are hard on your back — by pushing your curved spine against the floor and by working your hip flexors, the muscles that run from the thighs to the lumbar vertebrae in the lower back. When hip flexors are too strong or too tight, they tug on the lower spine which can be a source of lower back discomfort.

Is there any evidence that doing crunches or situps to exercise abdominal muscles is harmful to the lower back?

  • 1
    To be fair, they provide a cite: to their "Special Health Report", which they are selling.
    – Oddthinking
    Oct 18, 2013 at 22:08
  • @Oddthinking they don't site that as the source of the information that situps cause lower back pain
    – Celeritas
    Oct 19, 2013 at 2:49
  • "For more on the benefits of strengthening your core, buy [our book]." I consider that a reference. The obvious next step is to borrow that book from a library and check they have more specific references there.
    – Oddthinking
    Oct 19, 2013 at 3:14
  • I've modified the title for two reasons: (1) lower-back pain is the only notable claim you have provided. (2) asking the more general question "is it harmful?" is far more difficult to demonstrate. It is plausible to show it doesn't cause lower-back pain, but showing it has absolutely no harm would require major epidemiological studies which are unlikely to exist.
    – Oddthinking
    Oct 19, 2013 at 5:23
  • I read the title of this question too quickly and my brain read something along the lines of "Do startups and churches cause lower-back pain?", and I HAD to stop and read it. Too bad it was something else...
    – T. Sar
    Dec 29, 2015 at 18:08

1 Answer 1



  • There is a demonstrated link in the literature between both cumulative compressive loads of the spine and low duration, high peak compression with low back pain incidence and spondylosis.
  • All types of sit-ups result in high levels of peak compression which meets the threshold of peak compression levels studied within the literature.
  • Athletes have a higher incidence spondylosis than the general population.
  • Pain is not a reliable indicator of the presence of spondylosis or other spinal abnormalities.
  • There is evidence that compressive forces can lead to other structural changes in intervertebral discs—for example, decreased disc thickness—as well as to changes in intervertebral disc cell metabolism.

Dr. Stuart McGill has demonstrated in his lab that sit-ups produce compression forces of around 3400 Newtons on the lumbar spine.

In this 1995 article, they predict the compression loads for various types of sit-ups. In 1997, they demonstrated this experimentally (discussed further below).

The mechanics of torso flexion has not been thoroughly investigated due to the lack of knowledge of tissue force-time histories during the performance of torso flexion exercises. An anatomically detailed, three-dimensional model that was sensitive to lumbar curvature and muscle activation patterns was used to assess tissue load distribution during the performance of both isometric and dynamic situps and standing flexion manoeuvres in 12 young men. Situps were performed starting with a flexed torso which was lowered to horizontal and then raised again (instrumentation restrictions did not allow subjects to rest their torsos when in the lowered position). Specific variables measured were air flow during inhalation-exhalation, intra-abdominal pressure, myoelectric activity of the torso muscles, intercostals, and rectus femoris, three-dimensional dynamic curvature of the lumbar spine and body segment displacements. Lumbar compressive loads greater than 3000 N were predicted for both straight leg and bent knee situps. No biologically significant differences were found between bent knee and straight leg situp techniques.


Here is an interview with Dr. McGill where he mentions this research.

CW: Let's talk about sit-ups. Stu, there are so many guys I know that just can't get enough of them. I honestly believe that we have the Rocky movies to blame for them being so ubiquitous.

Your book Low Back Disorders was one of the first clinical-based texts that taught us sit-ups aren't very good for our discs. Your lab determined one of the primary reasons and it's due to the high compressive forces they induce, on the order of 3400N or 764 pounds. Tell us more about that.

SM: The National Institute for Occupational Safety and Health (NIOSH) sets exposure limits for substances that are known to compromise the health of the American worker. Along with obvious limits for things like poisons, radiation exposure, etc., they've found values for low back compression that lead to elevated rates of disabling back disorders.

The NIOSH have sponsored research that surveyed workers, and their data show that when lumbar loads exceed 3400N, the injury rates go up and interventions are recommended. The problem is that these are for average workers. Who's average?

Now, the sit-up recruits torso and hip muscles that impose compressive load on the spine for the average male of about 3400N. Performing sit-ups uses some training capacity – compression plus repeated bending – to achieve training of the torso and hip flexors. But remember that this combination of load and motion is a potent cause of annulus collagen delamination.


Here is the 1997 paper where they published the results of the work referred to in the interview and 1995 paper above.

  • Axler C, McGill S. Low back loads over a variety of abdominal exercises: searching for the safest abdominal challenge. 1997. Med Sci Sports Exerc. 29(6):804-11.


In this 1997 study, they state in the introduction:

Abdominal exercises are prescribed for a variety of reasons, but mainly for rehabilitation of low back injury and as a component of fitness training programs (6,19). In the past, abdominal exercises have been recommended for their capacity to maximize muscle activity (3,21). However, some investigators have raised concern regarding the safety of abdominal exercise programs, with suspicion that tissue damage can occur (5,7,11,20), especially through compressive loading on the lumbar spine (16,18,20). Unfortunately, the level of risk of low back injury from performing various abdominal exercises, quantified by measures of tissue loading, has not yet been sufficiently examined (7). The purpose of this study was to identify quantitatively abdominal exercises that optimize the amount of abdominal muscle recruitment (or “challenge” to the abdominal muscle), with a simultaneous minimization of compressive load, or “penalty” to the lumbar spine.

Since then, more work has been done concerning the link between peak and cumulative compressive lumbar loads with low back pain and injury.

Village J et al. Electromyography as a measure of peak and cumulative workload in intermediate care and its relationship to musculoskeletal injury: an exploratory ergonomic study. 2005. Applied Ergonomics. 36(5):609-18.


Some quotes from the Discussion section:

  • Norman et al. (1998) developed a “low back pain index” suggesting from their data the percent of workers who are likely to report back pain at a given level of cumulative loading. From their data, a cumulative compression of 23 MN s corresponds with 50% of workers likely reporting low back pain. [...] This must be viewed with caution however as the data are from different populations of workers and cumulative compression is obtained with different methods.

  • Kumar (1990) also estimated cumulative compression in a study of 161 institutional CAs (14 males, 147 females) with and without back pain using a structured questionnaire/interview in a retrospective study. Spinal loading estimates were obtained by using recall, line drawings and/or a manikin model to obtain estimates of working postures and then analyzing these postures with a two dimensional biomechanical model. Cumulative compressive and shear loads were then estimated based on estimates of task duration and frequency. The pain groups had significantly greater average estimates of cumulative spinal compression (males=15.6 MN s, females=14.5 MN s) than the no pain groups (males=6.6 MN s, females=9.3 MN s).

  • Seidler et al. (2001) used a modification of the Kumar (1990) approach to evaluate cumulative occupational exposure of the lumbar spine to lifting, carrying and working postures with extreme forward bending. A case control study was conducted between 229 male patients with symptomatic osteochondrosis or spondylosis of the lumbar spine and 197 control subjects. [...]

  • Seidler et al., found that working postures with extreme forward bending for up to 1500 h (calculated over all working years) was associated with the diagnosis of osteochondrosis or spondylosis (OR 2) and the odds ratio increased to 4.3 for more than 1500 h exposure. Combined exposures to lifting or carrying with working postures with extreme forward bending yielded odds ratios of 16:1. This is one of the first studies to use cumulative risk factor exposure as an independent variable. The authors suggest that the pathogenic concept of chronic increases in intervetebral pressure has long been considered an important cause of lumbar spinal disease, yet it has been difficult to quantify. All these studies have evaluated cumulative loading over occupational exposures. Since compressive loading does not end with the work day, the next goal in research may be to develop ways of measuring the non-occupational component of cumulative compression.

  • Few instances of peak compression above the 3400 N guidelines (Waters et al., 1994) were experienced by CAs in IC. During an average of 75 min pre-breakfast, compression exceeding 3400 N is experienced for an average of 11.5 s. Other periods of the day have even less peak compressions. Yet, peak compression was significantly correlated to injury rates, MSI rates, numbers of tasks performed in the day and perceived exertion in the day. It appears that even though the time spent in peak exertions is small, the perception of workload is highly influenced by these exertions. Norman et al. (1998) found strong correlations within the peak spinal loading variables and within the cumulative loading variables, but poor correlation between the two. They suggested that peak and cumulative loading are measuring different aspects of risk for these jobs. Peak and cumulative compression measured among CAs in this study were highly correlated with each other, but were independently correlated with different variables supporting Norman et al.'s assertion that both are important.

From the Seidler et al. study:


To calculate cumulative forces to the lumbar spine over the entire working life, the Mainz-Dortmund dose model (MDD), which is based on an overproportional weighting of the lumbar disc compression force relative to the respective duration of the lifting process was applied with modifications: any objects weighing [gt-or-equal, slanted]5 kg were included in the calculation and no minimum daily exposure limits were established.

The calculation of the sum of forces to the lumbar spine is a useful tool for risk assessment for symptomatic osteochondrosis or spondylosis of the lumbar spine. The results suggest that cumulative occupational exposure to lifting or carrying and extreme forward bending increases the risk for developing symptomatic osteochondrosis or spondylosis of the lumbar spine.

Several experimental studies suggest that compressive forces can lead to structural changes in intervertebral discs—for example, decreased disc thickness—as well as to changes in intervertebral disc cell metabolism.

A review of the literature concerning ergonomics can be found here:


  • We find that the relationship between cumulative spinal compression and low back pain incidence is linear as per Norman et al 2008.

So a there is a definite established link between spinal compression -- both peak and cumulative -- and low back pain, spondylolysis, and osteochondrosis

What relevance does this have to athletes or the more fit population? Do they suffer from these issues as well? Do they have a lower incidence due to being more fit?

Soler T, Calderon C. The Prevalence of Spondylolysis in the Spanish Elite Athlete. 2000. Am J Sports Med. 28 no. 1 57-62.


  • The incidence of spondylolysis in elite Spanish athletes(~8%) was somewhat higher than in the general population, which ranges from 3.5 - 7%.
  • Incidence was found to vary significantly based on sport.
  • Incidence among weightlifters was 12.94% (N=11). This rate is lower than other authors who report rates between 15 and 36%.
  • Their findings of the incidence among Spanish elite athletes were much lower than incidence among athletes found among the other literature which range from 12-15% in studies with a high number of subjects across a large number of sports. One study with 28 subjects across 13 sports found an incidence rate of 100%.

And importantly:

Pain is not a reliable indicator of spinal disorder.

  • In the Soler and Calderon study, only about half of the people with spondylolysis were symptomatic.

  • Another study found that of a group of 98 people with no low back pain incidence, only 36 percent had normal spinal disks.


Thirty-six percent of the 98 asymptomatic subjects had normal disks at all levels. With the results of the two readings averaged, 52 percent of the subjects had a bulge at at least one level, 27 percent had a protrusion, and 1 percent had an extrusion. Thirty-eight percent had an abnormality of more than one intervertebral disk. The prevalence of bulges, but not of protrusions, increased with age.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .