
Core stability and back pain: the best exercise, really?
By Kalorya · November 8, 2026 · 8 min read
Shixart1985 · CC BY 2.0
Is core training the best exercise for back pain? It helps, but no more than other exercises: out of 29 trials and 2,431 patients, the major Cochrane review does not find it more effective than other forms of exercise, regardless of the follow-up duration. Its slight lead over general exercise fades after three months. What provides relief is moving regularly — weight training, Pilates, walking or core training, take your pick. Here is what twenty-one studies say.
What we are talking about
Almost all studies focus on common low back pain, known as "non-specific": pain in the lower back with no identified cause, described as chronic when it lasts more than 12 weeks. "Core training" in these trials goes by several names: trunk stabilization, "core stability", "motor control" exercises. The 2016 Cochrane review describes it as work that first activates the deep trunk muscles and learns to control them, before moving on to more complex and functional movements. In a 2007 trial, this relearning was even done with an ultrasound to see the muscle contract.
First: exercise relieves the back
The 2021 Cochrane review, 249 trials, finds that exercise reduces pain by 15.2 points out of 100 compared to no treatment, usual care or a placebo, a difference that the authors consider clinically important. The gain on daily limitations is smaller, 6.8 points, below their threshold of importance. Compared to other non-drug treatments, exercise performs slightly better than simple information or physiotherapy without exercise, but no better than manual therapy. Reanalyzing the individual data of 3,514 participants, exercise reduced pain by 10.7 points out of 100 and limitations by 10.2; it worked better in those who did not have heavy physical work, who took medication for their back, and in those whose body mass index was lower.
Core training compared to other exercises
This is where the promise crumbles. Across 29 trials and 2,431 patients, the Cochrane review finds core training effective compared to a minimal intervention (12.97 fewer points of pain in the long term), but no more effective than other exercises, at any follow-up time point, and equivalent to manual therapy. Its authors conclude that the choice of exercise should depend on patient preferences, the therapist's training, cost and safety. For acute back pain, three small trials (197 participants) also find no advantage over other exercises.
Meta-analyses comparing core training to "general exercise" find an advantage, but a short-lived one. Across 5 trials and 414 patients, core training performed better at three months (1.29 fewer points of pain and 7.14 fewer points of disability), then no significant difference at 6 and 12 months, which a 2017 critical re-evaluation takes up as is. Across 11 trials, stabilization performed better than general exercise (1.03 fewer points of pain and 5.41 fewer points of disability), but no better than manual therapy. In the trial of 240 adults, core training performed slightly better than general exercise (strengthening, stretching, endurance) after 8 weeks; at 6 and 12 months, all three groups were equal.
Meta-analysis of 5 randomized trials, 414 patients (2012); only the difference at 3 months is significant
An older review goes in the other direction: in 2015, its authors conclude that trunk strengthening relieves better than traditional weight training and recommend targeting deep muscles. But this is a review without a meta-analysis, and the more recent major syntheses above and below do not find this advantage.
Is it the exercise, or the fact of being monitored?
A trial on 154 patients compared 12 core training sessions over 8 weeks to a fake treatment (turned-off ultrasound and shortwave devices). Core training improved activity (1.1 points on an 11-point scale) and the impression of recovery (1.5 points), but pain did not drop markedly: 0.9 points, at the limit of significance. The gains held at 6 and 12 months. And in 148 adults, therapy working on fear of movement and daily gestures performed slightly better than core training combined with manual therapy on disability at 8 weeks (4.75 points), with no difference on pain, and no difference at 6 and 12 months.
Did you know?
Weight training, deadlift, Pilates, walking
Network meta-analyses, which compare all exercises among themselves, do not place core training alone at the top. Out of 89 studies and 5,578 patients, Pilates came out first for pain, weight training and core training tied for physical function, weight training and endurance for mood; stretching and the McKenzie method did not do better than no treatment — but the quality of the evidence was judged low. Out of 118 trials and 9,710 participants, all exercises helped, except stretching for pain and the McKenzie method for disability; Pilates, mind-body exercises and core training were the most effective against pain, Pilates, weight training and core training against disability. Out of 217 trials and 20,969 participants, on the other hand, it was Pilates, the McKenzie method and "functional restoration" that came out ahead (15 to 19 fewer points of pain out of 100 compared to minimal treatment): the rankings do not agree, except on Pilates.
Weight training holds its ground well. Out of 10 trials and 434 patients, it reduced pain and disability, with certainty judged moderate. Out of 8 trials and 408 patients, work centered on the posterior chain (the muscles that extend the back and hip, as in the deadlift) performed better than general exercise on pain, disability and strength, especially over 12 to 16 weeks rather than 6 to 8, with no more adverse effects. Out of 6 trials and 333 patients, endurance and weight training lowered pain just as much as each other, with weight training additionally improving psychological well-being. And out of 5 trials, walking relieved as much as other exercises; adding it to an exercise program brought nothing extra in the short term (on daily walking, see how many steps per day}.
The deadlift, not for everyone right away
And how to avoid pain?
In prevention trials, exercise alone reduced the risk of back pain by 33% (8 trials, 1,634 people), and by 27% when combined with information. The authors recommend strengthening combined with stretching or endurance, two to three times a week. On the other hand, these trials showed no clear effect on sick leave or consultations. After an acute episode, core training added to medical treatment reduced recurrences at one year by 64%, but with very low-quality evidence: the authors consider the question unresolved.
The limitations, all the same
In practice
You don't need to do core training in particular. The 2021 network meta-analysis states this in its conclusion: it is better to encourage the person to do the exercise they enjoy, so that they keep doing it. The most effective programs, according to the 2022 one, featured at least 1 to 2 training sessions per week of Pilates or weight training, sessions lasting less than 60 minutes, and lasted 3 to 9 weeks for Pilates and core training. If you already weight train, posterior chain work has the best results over 12 to 16 weeks. To build your training sessions, the workout generator and the exercise library give you the movements; to choose between barbells and machines, read free weights or machines.
In summary
| Your question | What the studies say |
|---|---|
| Does exercise relieve back pain? | Yes: 15 fewer pain points out of 100 on average |
| Does core training do better than the rest? | No: no more effective than other exercises (29 trials) |
| And compared to general exercise? | Slightly better at 3 months, no more difference at 6 and 12 months |
| Weight training? | Effective; the posterior chain does better over 12 to 16 weeks |
| Deadlifts? | For a sufficiently resilient back and a fairly low level of pain |
| Walking? | As effective as other exercises |
| To prevent? | Exercise reduces the risk by a third; core training no better than any other |
To build a week of workouts that you want to stick with, try the workout generator.
Core training, weight training or Pilates: choose what you like, and stick with it for several weeks.
Build my workoutsEvery figure in this article comes from a study you can open: the underlined mentions lead to the summary of the work it is taken from, written solely from the abstract. If your back pain is severe, recent or unusual, talk about it to a doctor before training. Informational content, which does not replace the advice of a health professional.