Are flexibility and muscle-strengthening activities associated with a higher risk of developing low back pain?
I initially commented on facebook the following:
"Umm 'observational cohort study' - maybe people with back pain are more likely to use resistance machines because other exercises exacerbate their pain? Haven't read the full text yet but the abstract conclusions are misleading "...use of weight training machines is associated with increased risk of developing low back pain compared to use of free weights..." The word 'developing' should be 'reporting' as this misleading gives an indication of temporal direction which this study can't determine."I've managed to get hold of the full text now and read through so I thought I'd comment on it here as although the authors wordings may not be as misleading as I first thought, there are still issues with the study (or at least how it is reported) that give cause for concern with such strong conclusions.
The vague reporting in the methods section leaves me with a number of questions. In essence the researchers in the study drew upon data from he Aerobics Centre Longitudinal Study (ACLS) for reanalysis in order to examine the incidence of LBP in those using muscle-strengthening and flexibility exercises. Participant data was used as long as they had undergone at least 2 medical examinations as part of the ACLS between 1977 and 2005. Participants with a number of morbidities reported during their last examination were excluded so that the population consider was essentially healthy. The authors also report that they excluded any participants who reported LBP at 'baseline.' During examinations participants were asked to indicate yes/no as to whether they had ever had a problem with LBP or whether they were currently suffering from it. They were also asked to answer the following questions:
- "Are you currently involved in a muscle-strengthening program?"
- "Can you specify the muscle-strengthening activity as "Callisthenics", "Free Weights", "Weight Training Machines" or "Other"?"
- "Are you currently involved in exercises to maintain or improve your joint flexibility?"
- Can you specify the flexibility activity as "Stretching", "Callisthenics", "Exercise Class, "Yoga" or "Other"?"
Now the reason I italicised 'baseline' is because based upon the methods and the reporting of their results it is very difficult to discern when, relative to the follow-up period, baseline actually was.
The population were followed for 4.9+3.8 years. The results notes that 'new incidence of LBP' was reported and examined, however it is not indicated during what examination (out of the 'at least 2' across the follow-up) this data was taken from relative to the baseline nor how many examinations were even conducted on average for each participant. So we have no idea as to 'when' over the follow-up period LBP was actually reported and based upon the incongruity between the question asked for determining the presence of LBP reported in the methodology (ever had a problem with LBP or currently suffering from it) and the reported 'new incidence of LBP' it is difficult to discern.
As baseline reporting of LBP (whenever baseline indeed was) was considered exclusion criteria it is reasonable to conclude that any relationship between baseline variable and LBP at follow-up (whenever that indeed was) was prospective, despite not knowing any more details.
So what relationships were reported? At baseline there were differences between those reporting that they engaged in either muscle-strengthening or flexibility exercises compared with those who did not. Generally those engaging in either were 'healthier' across a range of markers in their health examination. However, what was not reported was whether within the groups engaging in muscle-strengthening or flexibility exercises there were any other differences between those using different approaches.
The researchers did however adjust for many of these health markers in 3 other models and found quite consistently that those reporting they engaged in muscle-strengthening using either weight training machines or 'other' methods were significantly more likely to report LBP ('new incidence' or past problems is unclear?). Stretching exercises also significantly increased the likelihood of reporting LBP though the results were not as consistent across the models.
Though the analyses of these relationships appears robust enough my concerns about the lack of clarity in the methodology for reporting of LBP, and thus the validity of the data collected, leaves me questioning this. It's unclear to me as to whether this relationship is indeed truly longitudinally prospective and that for whatever reason engaging in exercise using resistance machines increases risk of 'developing' LBP. Despite the authors comments in the discussion regarding free weight exercise potentially doing a better job of strengthening the paraspinal musculature there is actually little to no evidence to suggest that they are indeed better at doing so. The conclusions are also extremely broad in nature and do not consider the specific exercises engaged in. There are very likely exercises, both free weight and machine based, that when compared to others are either better or worse in terms of development of LBP. It would have been interesting to understand what 'other' methods of muscle-strengthening were used also in order to better speculate regarding a potential mechanism for how LBP might develop from engagement in both resistance machines and 'other.'
All in all this study in my humble opinion doesn't really tell us very much and it concerns me that it may put people of the use of resistance machines for reasons not wholly valid.