Background The effect of exercise to prevent low back pain (LBP) and associated disability is uncertain. We carried out a meta-analysis to address this question.

Methods Literature searches were conducted in PubMed, Embase, Cochrane Library, Google Scholar, and Research Gate from their inception through September 2016. Randomized controlled trials (RCT) and clinical controlled trials (CCT) were eligible for inclusion in the review if they compared an exercise intervention with usual daily activities and at least some of the participants were free from LBP at baseline.

Results Sixteen controlled trials including 13 RCTs and 3 CCTs qualified for meta-analyses. Exercise alone reduced the risk of LBP by 33% (risk ratio (RR)=0.67, CI: 0.53 to 0.85, I²=23%, 8 RCTs, N=1634) and exercise combined with education by 27% (RR=0.73, CI: 0.59 to 0.91, I²=6%, 6 trials, N=1381). The severity of LBP and disability due to LBP were also lower in the exercise than control groups. Moreover, results were not changed by excluding the CCTs, or by adjustment for publication bias. There were few trials on healthcare consultation or sick leave for LBP, and meta-analyses of these trials did not show statistically significant protective effects of exercise.

Conclusions Exercise reduces the risk of LBP and associated disability, and a combination of strengthening with either stretching or aerobic exercises performed 2–3 times/week can reasonably be recommended for prevention of LBP in the general population. However, education about back disorders, ergonomic principles or exercise effects appears to have no additional beneficial effect on LBP.

Funding Finnish Ministry of Education and Culture.