Abs and back pain: it is one of the worst-told pairs in rehabilitation. On one side there are people who have not done a crunch in years for fear of "ruining their discs". On the other there are people who grind through sets of sit-ups convinced that a strong belly holds the spine together. They look like opposite positions, but they come from the same idea: that there is a fixed list of good exercises and a list of bad ones, the same for everyone.
That list does exist and is circulated everywhere. Except that it was born in a biomechanics laboratory, not in a study of people with low back pain. It is worth knowing where it comes from before letting it decide what you can do.
⚠ Important: read this before continuing
The exercises and advice in this article are general guidance and may not suit your situation. Every person has a unique clinical history: what works for others could be counterproductive for you.
Want a professional's opinion? Start from the questionnaire: you sign up and see who will follow you, a physiotherapist registered with the Register chosen for your case, whom you can write to in chat. If you would rather hear their voice, the first consultation is free and we look at your situation together.
Before choosing exercises: the signs that have nothing to do with abs
The vast majority of back pain hides nothing serious. A small minority does, and for that minority the mat is the wrong answer. See a doctor, or go to the emergency department, if even one of these signs appears:
- Loss of bladder or bowel control, or sudden difficulty passing urine.
- Loss of sensation in the "saddle" area: inner thighs, perineum, genitals, anal area.
- Weakness in one or both legs that worsens over hours or days, a foot that gives way, trips that did not happen before.
- Fever or shivering together with back pain, especially if you have an infection elsewhere or a vascular catheter.
- History or suspicion of cancer, unintentional weight loss, pain that gives no relief even lying down and wakes you at night.
- Major trauma (a fall, an accident), or even a modest trauma if you have osteoporosis or have been taking corticosteroids for a long time.
An honest clarification: these warning signs serve to decide when to get seen, not to self-diagnose. None of them, taken alone, means something serious is going on, and not everything serious announces itself with a sign. The point is to take the decision away from the mat and bring it to the clinic.
The idea that holds everything up: "your core is weak, that is why it hurts"
It is the premise that almost nobody questions. If your back hurts, the wall of muscles that should protect it is weak; strengthen it and the pain goes away. It is a tidy, mechanical, reassuring story. The problem is that the studies that tested it did not confirm it.
A systematic review published in Pain gathered 26 prospective cohort studies, people measured when they were well and then followed over time to see who developed pain, 24 of which concerned exactly low back pain and only three neck or shoulder pain. The verdict on trunk muscle endurance is clear: strong evidence of no relationship with the risk of developing back pain. For trunk strength and spinal mobility, the results were too inconsistent between one study and another to conclude anything: inconclusive evidence.
Where does the opposite belief come from, then? Mostly from studies that take a single snapshot: they take people with back pain, take people without it, measure the trunk and find a difference. But a snapshot does not tell you what came first. Someone who has been in pain for months moves less, avoids loads and trains less: a trunk that tires sooner can very well be the consequence of the pain, not its cause.
This changes a lot about the relationship between abs and back pain. It does not mean "don't train": it means you are training the trunk to build capacity, not to repair a fault that hurt you. The same myth has a domestic version: the protruding belly of the small child attributed to weak abs. There too, where the association was measured, it was too small to do anything about it: what you really see in children's protruding bellies.
So do abs actually help?
Yes, but for different reasons from the ones you have been told, and with numbers worth knowing before you start, because expecting the wrong thing is the best way to give up after three weeks.
The Cochrane review on exercise for chronic low back pain included 249 randomised trials. Compared with no treatment, usual care or placebo, exercise reduces pain by 15.2 points on a 0-100 scale (95% confidence interval -18.3 to -12.2), with moderate certainty evidence: the authors considered a difference of 15 points clinically relevant, so the result only just reaches it. On functional limitation the gain is 6.8 points (from -8.3 to -5.3) and it does not reach the 10-point threshold that the same authors had set as the minimum important difference. Compared with other conservative treatments the advantages exist but are small and not clinically important.
In one sentence: exercise works, it works better than doing nothing, and nobody has shown that one form of exercise beats the others.
On this second point, the Cochrane review on motor control exercises, those built precisely on activating the deep trunk muscles, is even more explicit. Across 29 studies and 2,431 participants, they were not clinically more effective than other forms of exercise at any point in the follow-up. Compared with a minimal intervention they do work (for long-term pain, mean difference 12.97 points, from -18.51 to -7.42). Compared with manual therapy, moderate to high quality evidence of no clinically important difference. For fairness: there is a comparison in which motor control comes out ahead, and that is against exercise combined with instrumental physical therapies (electrotherapy and similar), but with very low to low quality evidence. The authors' conclusion is the most useful of all: the choice of exercise should depend on the preferences of the patient and the therapist, on training, costs and safety.
A more recent meta-analysis of 57 randomised trials and 7,705 participants compared Pilates, core stability and strength core training with each other in non-specific chronic low back pain. All three improve pain compared with controls (standardised mean difference 0.70; from 0.58 to 0.82). The figures for the individual methods differ (Pilates 0.75, strength core 0.68, core stability 0.53), but the statistical comparison between them does not reach significance (P = 0.24), neither for pain nor for function. To be fair, the authors nevertheless present Pilates in their conclusions as the best method for pain and strength core as the best for function (where it scores 0.76): this is a reading of the numerical differences that their own statistical comparison does not confirm. Put simply: the choice of "school" matters less than you may be told.
Where the list of "banned" exercises really comes from
Here we reach the point that almost no article on abs and back pain tells in full.
In 1997 Axler and McGill published a study in which nine volunteers performed twelve different abdominal exercises. For each one they measured the electrical activity of the abdominal muscles and estimated the compression on the lumbar spine, then built an index: how much muscular work you get for each unit of load you pay. It is from this study, and others like it, that the list of exercises to avoid that you find in every gym descends.
The trouble is that the list is reported badly, for at least three reasons.
The first: in the study, partial crunches achieved the best indices of muscular work to spinal load cost. The crunch, the exercise that is usually demonised first, comes out rather well. The authors also clarify that the exercises with the best ratio were not necessarily those with the lowest compression in absolute terms.
The second: the real conclusion of the study is not "here is the safe exercise". It is the opposite: no single exercise trained all the abdominal muscles optimally while imposing minimal loads at the same time, so a variety of different exercises is needed depending on the goal.
The third, the most important: these are loads estimated in a laboratory on nine volunteers, not on people followed over time. There is no study showing that people who do sit-ups develop more back pain than people who do not. The leap from "this exercise compresses more" to "this exercise will hurt you" is a leap the data do not authorise.
The same logical leap underlies the most famous ban of all, that on abs in pregnancy: it comes from anatomical reasoning, not from a study that saw harm. Which exercises become truly impractical in those months, and for what reasons, is in which abs you can do in pregnancy. After birth that ban lasts even longer, and there too the studies say something else: when abs really return, and what comes first, is in the sequence of returning to movement after birth.
Be careful, though: I am not saying biomechanics is useless. If you have an ongoing radiculopathy, a symptomatic disc, or pain that flares every time you bend forward, knowing which exercises load the spine most in flexion is practical information, and it makes sense to use it to choose where to start. What I am saying is something else: that information helps you order exercises over time, not divide them into permitted and forbidden forever.
An important point: before trying these exercises, make sure you have received a correct diagnosis. People often think they have one problem when the real cause of the pain is something else. Doing the wrong exercises can make things worse rather than better.
The criterion that replaces the list: your tolerance
If there is no universal list, the relationship between abs and back pain has to be decided case by case, and you need a personal criterion. The one I use in the clinic is simple and you can apply it yourself.
The 24-hour rule. Do the exercise, then see what happens the next day. Widespread muscle soreness that passes within a couple of days is normal and not a problem. Back pain more intense than usual that lasts beyond 24 hours, or morning stiffness worse than usual, is telling you the dose was too much: not that the exercise is wrong, that the dose was too much. Go back to half the volume and try again.
The sign that is a real stop. If during or after an exercise the pain moves downwards, from the back to the buttock, from the buttock to the thigh, from the thigh to the calf, or tingling and numbness appear in the leg, stop and do not push. This is heading in the opposite direction to the one we want.
The honest "can I breathe?" test. If during an exercise you have to hold your breath and clench your teeth to hold the position, you are above your current capacity. A trunk exercise done well makes you work, but it lets you breathe.
Where to start
I do not call these "safe" exercises, because that is not a category that exists. I call them low, controllable demands: the sensible place to start for most people, because the load can be graded and the position is easy to keep under control.
Dead bug. Lying on your back, knees and hips at 90 degrees, lower back resting on the floor. Slowly extend the opposite arm and leg, return, switch sides. If the lower back lifts off the floor you have extended too far: reduce the range. It is the exercise I start with most often, because it teaches you to keep the trunk still while the arms and legs move, which is what real life requires. The full technique of the dead bug deserves a few more minutes.
Bird dog. On all fours, extend the opposite arm and leg while keeping the pelvis stable. The classic mistake is lifting the leg too high, rotating the pelvis: better little and symmetrical than a lot and crooked. You will find the most common bird dog mistakes explained in detail.
Front plank. It serves to resist extension, not to run a duration contest. If you cannot keep your pelvis aligned, put your knees down: it is a regression, not a failure. On the question of how long to hold a plank we have written a separate piece, because it is the question we get asked most often.
Side plank. It trains the obliques and quadratus lumborum, that is, resistance to lateral bending, what you need when you carry the shopping in one hand. Here too you can start from the knees.
Partial crunch. Yes, the crunch. We lift only the shoulder blades and head a few centimetres, without hands behind the neck (the hands go under the lower back or crossed on the chest) and without momentum. It is the exercise that, in the laboratory study from which the "banned" list comes, came out best in the ratio between muscular work and spinal load, and there is no reason to exclude it on principle.
The ones I postpone, not the ones I ban
Full sit-up, weighted Russian twist, lifting straight legs from lying, crunches with hands behind the neck and neck pulls. For these the honest position is not "never again", it is not right now.
They are high demands, for reasons worth distinguishing. Sit-ups and straight-leg raises are, among the exercises measured in the laboratory, those that estimated the highest compression loads on the spine. The weighted Russian twist was not in that study at all: I postpone it for a practical reason, not a biomechanical one. It loads rotation under external load, and it is among the first movements where control is lost. This applies to everyone: they are often performed quickly and in long sets, and technique gives way before motivation does. If you are rebuilding tolerance after an episode of low back pain, putting them at the start is like going back to running starting from 400 metres. But if in a few weeks you reintroduce them gradually, in few, controlled repetitions, and your back does not protest in the following 24 hours, there is no reason based on evidence to keep excluding them from your life.
The reasoning works the other way for someone who is well and has done them for years without problems: there is no reason to stop out of fear.
Outside the mat, too, the sit-up serves no one for getting up from lying down: you roll onto your side and push with your arm, which is the same route every child finds by themselves when they learn to sit up without anyone teaching them.
How to perform them: the three things that change everything
Breathe. Holding your breath during trunk exercises is the most widespread mistake. Breathe out during the effort phase and keep a breath that is possible for the whole duration of the isometric hold.
Look for neutral position, not forced flattening. The lumbar spine has a natural curve and it should be respected. "Pressing the lower back into the floor" at all costs is not a goal: the goal is for the lower back not to move when you move the arms and legs.
Fewer repetitions, done better. In clinical practice the quality of performance matters more than the number: ten repetitions done slowly and with control are worth more than thirty done in a rush. Better to stop two repetitions early than two late.
How often to train
There is no proven dosage valid for everyone, and be wary of anyone who presents one as such. What we can say is which parameters the studied programmes had.
In the meta-analysis of 57 studies and 7,705 participants, the meta-regression analysis identified as optimal parameters: for strength core training 3-4 sessions a week of 30-45 minutes; for Pilates 2-3 sessions of 50 minutes for 8-12 weeks; for core stability 3-4 sessions of 40-60 minutes for 6-8 weeks. In the same work, weekly frequency turned out to be the strongest predictor of pain improvement for strength training.
These are the numbers of programmes run in studies, often under supervision. The practical translation for someone training at home is more modest but more realistic: little and regular beats a lot and occasional, and a programme you cannot follow for eight weeks is worth less than a shorter one you can complete.
What to expect, and what not to
One point that enthusiastic posts never mention: in the Cochrane review of 249 studies, the improvement in function, meaning how much your back limits you in daily activities, did not reach the threshold that the authors themselves had set as clinically important, while the improvement in pain barely did. It is a real benefit, but smaller than the one promised around.
There is, however, a chapter where the numbers are more encouraging, and that is prevention. A systematic review of 21 randomised trials and 30,850 participants published in JAMA Internal Medicine found moderate-quality evidence that exercise and education together reduce the risk of an episode of low back pain (relative risk 0.55; from 0.41 to 0.74). Exercise alone seems to reduce it (0.65; from 0.50 to 0.86), but with low to very low quality evidence. Education alone has no effect (1.03), and neither do lumbar belts (1.01) or insoles (1.01).
On adverse effects, again from the Cochrane review: they were reported in 37 of the 112 exercise groups (33%) and in 12 of the 42 comparison groups (29%), and they were mostly minor events such as muscle soreness. In short, exercise is not a risky intervention.
In short
The relationship between abs and back pain rests on a premise that prospective studies have not confirmed: there is no solid evidence that a trunk with poor endurance leads to low back pain. And the list of banned exercises that has frightened half the population comes from loads estimated in a laboratory on nine volunteers, in a study whose real conclusion was that variety is needed, and in which the partial crunch came out well.
What remains, and it is a lot, is this: training the trunk is part of exercise, exercise reduces pain and disability in chronic low back pain, and together with education it reduces the risk of new episodes. The way you choose it, however, changes: no longer a list of good and bad exercises the same for everyone, but your tolerance measured at 24 hours, a progression that starts from low demands and builds on them, and the willingness to reintroduce later what you postpone today.
If you want to understand which exercises make sense for your situation, with your history, your pain and your work, the physiotherapists at Ri-Hub assess the case through tele-rehabilitation and build the progression with you, from your home. If instead you are starting from the basics of the trunk, you will find everything in Core Strengthening Exercises: What the Research Says, and if the pain is already chronic in Chronic Low Back Pain: the Exercise Programme.
Sources
- A systematic review of the relation between physical capacity and future low back and neck/shoulder pain, Pain, 2007
- Low back loads over a variety of abdominal exercises: searching for the safest abdominal challenge, Medicine & Science in Sports & Exercise, 1997
- Exercise therapy for chronic low back pain, Cochrane Database of Systematic Reviews, 2021
- Motor control exercise for chronic non-specific low-back pain, Cochrane Database of Systematic Reviews, 2016
- Effects of different types of core training on pain and functional status in patients with chronic nonspecific low back pain: a systematic review and meta-analysis, Frontiers in Physiology, 2025
- Prevention of Low Back Pain: A Systematic Review and Meta-analysis, JAMA Internal Medicine, 2016
Sources last checked: August 2026.
You might also like:
Riabilitazione del Crociato: le Fasi e i Criteri Ergonomia Smart Working: Guida alla Postazione Perfetta Esercizi Rinforzo Core: Cosa Dice la Ricerca
Stay in touch with us Join our dedicated Facebook group: we answer all patients, give practical advice and clarify the most common doubts.
Need a consultation?
Start from the Ri-Hub questionnaire: a few questions about your case, you sign up, and you see who will follow you: a physiotherapist registered with the Register, whom you can write to in chat whenever you like. If you prefer to talk it through by voice, the first consultation is free, lasts twenty minutes and commits you to nothing: they arrive having already read what you wrote, and together you decide whether and how to continue.
