- Hypermobility and Early Development: Milestones, Motor Delay, and Predictions - 12 September 2026
- Fibromyalgia and Itching: Why It Happens, What the Science Actually Says, and What You Can Do About It - 9 September 2026
- Hypermobility in Children: Growing Pains, the Beighton Score, and What Actually Helps - 15 August 2026
This article is part of our comprehensive guide to hypermobility and Ehlers-Danlos syndrome.
Research shows that Bendy babies, hit their motor milestones later more often than other babies do, and that’s been sitting in the research since the late eighties (which makes it one of the few things in this area to have held up reasonably well) [1]. What usually gets left out though, is everything that comes afterwards. Most of those babies researched improved within a few months, hinting that being bendy on its own turns out to predict very little about where a child actually ends up [1][2].
Hypermobile children who were behind as babies, were still behind on gross and fine motor skills at five years old, and the children who were hypermobile without the delay weren’t [2]. So, the useful question was never really how far, say their elbows hyperextended, but rather whether or not their bendiness was turning up alongside anything else.
That distinction does an awful lot of work, and hardly anybody bothers to actually make it. Plenty of people arrive at a page like this as an adult, rather than as a parent. Generally, working backwards throughout their own childhood to see whether the late walking, the difficulty writing, or the years of just being called “clumsy” explain anything at all. Some of that will fit, a fair bit of it won’t, and the research is at least pretty good for telling you which parts of a childhood are worth reading something into, and which parts are just childhood.
Hypermobility is extremely common in childhood and becomes less common as children grow, which is exactly why the criteria used for children have to be different from the adult ones [3][4]. So, a lot of what gets flagged as a problem is a child, just being a child. Some of it genuinely isn’t though, and telling those two apart is considerably harder than either the reassuring version or the alarming version of this topic would ever admit.
What Actually Turns Up in Hypemobile Babies
The clearest early finding comes from babies between eight and fourteen months old, where motor delays turn up roughly three times as often in those who were hypermobile [1]. Three joints carried most of that: how far the foot pulled up at the ankle, how far the hips opened out, and how far the elbows travelled back into hyperextension [1]. And they weren’t simply adding together either, as having more than one of them looked like it did more than the sum of its parts [1].
Now, the reassuring half, which gets quoted a great deal less often! Most of those babies studied improved over the following six months [1], so late isn’t the same as stuck, and in infancy it very often sorts itself out without anybody doing anything clever at all.
The broad direction has held up since [5], although it isn’t unanimous, and that matters a great deal: because when babies born at full term were looked at on their own, joint laxity made no difference at all to motor development or to the age they started walking [6]. So, that’s two sets of babies and two opposite answers, and nobody has properly resolved which one generalises.
In practice, the babies that parents ask us about a lot of the time, are usually the ones who skipped crawling, bottom shuffled for months, or even sat late and then walked more or less on time. This is generally, after somebody has told them she’s just double jointed. None of that’s a diagnosis of anything on its own (plenty of thoroughly unbendy babies bottom shuffle too), and the research isn’t precise enough to tell you which pattern means what. What it is precise enough to say though, is that the delay described in that early work was ordinary infant motor development, running behind schedule, and that’s a very different thing to hear at a twelve month mile stone check in, than hearing your baby has a condition.
Hypermobile + Late Is the Combination That Matters
Following bendy babies forward is where this research earns it keep though. At age five, the children who had been both hypermobile and delayed as infants, were still behind on gross and on fine motor skills [2], whereas the children who had been bendy without being late, had caught up and looked like everybody else [2].
Take that with a pinch of salt though please, as the work behind it is small, and small is still small, no matter however sensible a finding sounds [2][7].
Why the combination should matter isn’t even settled either. As the he best available guess is that a baby who is bendy and late, is a baby whose nervous system is having a much harder time making sense of a wobblier body. As opposed to a baby who is both bendy and on time, that seems to be managing that perfectly well. It’s a guess though, like a lot of things in hypermobile research, and while it does fit what’s been measured since, nobody has tested it directly (surprise, surprise).
Still, if you’re trying to work out whether to push for an assessment, what matters is whether the bendiness has arrived alongside a child whose movement is genuinely behind where it should be, as a child who is wildly flexible and hitting everything on time is in a different group entirely.
Keep in mind that it can swing the other way, as being late is worth paying attention to regardless of hypermobility. Delay is worth looking at on its own terms, and whether the child happens to be bendy as well changes very little about that.
Born Early Changes the Picture, and Nobody Can Agree
Babies born preterm is a different situation entirely, and the findings there point in opposite directions as well. Preschool children born early who had joint laxity walked later and scored lower on movement testing than those without it [8]. Then, in toddlers born preterm, somewhere between two and three and a half years old, hypermobility showed no relationship at all with proprioception or even with developmental scores [9].
Both of those can’t be right in the same way though, surely, and the stance is that children who were born early, come with plenty of other reasons to be late on its own, so separating out the contribution of loose joints from everything else prematurity brings is genuinely difficult.
So, in short, If your child was born early and is both bendy and behind, the useful thing to know here is that the second fact is doing the work in any assessment, and the bendiness is a detail alongside it, rather than the explanation for it (or at least it would appear that way)
By School Age It Gets Harder to Discern
Once you get past the toddler years and into ordinary school aged children, the association gets very thin indeed. Generalised hypermobility in five and a half year olds showed no link with overall motor performance [10], and the same broad answer came back in other general groups of children [11][12]. In eight year olds, hypermobile children weren’t worse at motor tasks, and in places came out looking slightly better [13].
Which fits with the other thing worth knowing about childhood hypermobility, as it’s common and it fades, turning up in a sizeable share of school aged children, becoming less common as they get older [4]. In fact, the paediatric criteria exist precisely because applying an adult threshold to a nine year old, flags an enormous number of children whose joints are doing nothing unusual for their age [3].
So, a hypermobile seven year old who is running about, doing PE, sleeping fine and not in pain is, on the current evidence, a hypermobile seven year old. That isn’t a diagnosis waiting to happen, and the research doesn’t support treating it as one.
The Sensory and Motor Side, and What It Does and Doesn’t Explain
This is the part we find most interesting, as it’s the layer our own work sits on, and the newer research has moved off of milestones, and onto what the nervous system is actually receiving, which in our opinion is a much better place to be looking.
In toddlers around a year old, hypermobility came with lower scores across deep pressure response, adaptive motor function, visual and tactile integration and vestibular response, along with a lower overall score for sensory function [14]. In older children, it seems to show up in more specific ways though: hypermobile children had poorer tactile sense in the foot and a poorer sense of where the ankle was sitting [15]. Children with symptomatic hypermobility had poorer knee position sense and poorer kinaesthesia, together with weaker knee flexors and extensors [16], while the way the knee moved through the walking cycle even differed as well [17]. Dynamic balance and lower leg strength both came out lower too [18].
That’s fairly consistent, and it’s the one that makes the most mechanistic sense to us. As more compliant tissue around a joint means the signal coming in from joint is less crisp, and a brain making movement decisions on a noisier signal, makes worse ones. Prediction is only ever as good as the information going into it. Which is why our first move with a hypermobile body is almost never to load it harder, it’s to give the nervous system something clearer to work with and a better strategy to use, and then load it.
There’s a version of this that parents will recognise instantly: the bendy child who leans on furniture, sits on their feet, grips a pencil like it owes her money, or comes home from school completely shattered after a day of sitting upright at a desk, is doing something that costs her more than it costs the child next to her. Holding a joint steady with muscle, because the joint gives you less to work with on its own, is simply more work. Whether that’s the explanation for the tiredness hasn’t been shown yet, so treat it as the reasonable inference it is, and it’s far and away the most common thing we get asked about in this age group. There are two caveats though, and they both matter.
Not every sensory measure comes out that way, as the ability to recognise an object by touch alone, and the sense of your own body’s dimensions, showed no relationship with hypermobility, which argues fairly strongly against sensory processing being the single explanation for everything people report [19]. Proprioception showed nothing in those preterm toddlers either [9]. So, it’s a thread rather than a complete story, and anyone selling you the complete story is ahead of the evidence.
Now, the bigger caveat is timing, as all of this is measured alongside and never before. Nobody has shown that the sensory differences cause the motor delays, rather than following it, or that both are an effect of something else entirely. These things turn up together, which is a reason to take the sensory layer seriously in how you train, and not a reason to claim you know which way the arrow points. The underlying biology of hypermobility itself is still poorly understood [20], and that really isn’t a small gap to be standing next to.
Coordination Difficulties, and the Overlap That Keeps Turning Up
Developmental coordination disorder, the diagnosis given to children who are markedly clumsier than their age would predict, keeps appearing in the same conversation as hypermobility. Among children referred because of coordination difficulties, hypermobility was more common than in typically developing children, and within that referred group the bendier children did worse on motor testing [21]. In the general population, that relationship simply wasn’t there [21].
Some of that’s a referral pattern as much as a biological one, and it’s worth holding both possibilities at once, as children who are bendy and clumsy get sent to somebody, while children who are bendy and coordinated get told they’re flexible and packed off to gymnastics, and they never appear in anybody’s numbers.
The symptom overlap itself has been described often enough to take seriously as well: pain, falls, difficulty sticking with a motor task, handwriting problems, constipation, swallowing difficulties, and trouble telling a story in the right order [22][23]. Parents tend to recognise that cluster immediately, usually because they’ve been handed a separate explanation for each item on that list, by a different person each time.
The most sensible summary is that there are probably at least two groups of hypermobile children: one with entirely typical motor development, and one with a much broader developmental picture [24][25].
ADHD, Autism, and Where the Claim Runs Well Ahead of the Evidence
You’ll have seen the claim that hypermobile children are far more likely to be autistic or to have ADHD. It does has something real behind it, it’s just been badly overstated.
The links described in this space are fairly wide: ADHD and autism, but also learning, language, social, sleep, feeding and emotional difficulties, often recognised long before anybody mentions connective tissues [26][27]. Among children already under a specialist services for hypermobility spectrum disorder or hypermobile Ehlers-Danlos syndrome, a minority had a confirmed ADHD diagnosis and a smaller minority had a confirmed autism diagnosis, with ADHD more common in the hEDS group than in the HSD group [28].
Pull the autism work together and hypermobility does turn up more often in autistic people than you’d expect by chance, but the figures bounce around so much from one group to the next, that the pooled number isn’t really worth very much on its own [29][30][31].
There’s one finding in here that should keep everybody honest though, as when eleven year olds were looked at in ordinary schools rather than through a specialist service, there was no association between hypermobility, musculoskeletal pain, and neurodevelopmental difficulties at all [32].
So, the strong version of the claim “hypermobile children have a clear excess of autism and ADHD at population level” isn’t currently supported. The narrower version is: among children symptomatic enough to end up in a specialist service, these things travel together a lot [28][25]. Those are two very different statements indeed, and they get used interchangeably and the swapped around all of the time, which is usually why there is so much misinformation around the subject.
Why the Research Argues With Itself So Much
Most of the contradictions above come down to one unglamorous problem, which is who counts as hypermobile in the first place, as the same group of children can look extremely hypermobile or only mildly, so depending on the cut off used, the child’s age, sex, and whether bendiness in one or two joints counts or only bendiness more or less everywhere [33][3][4]. Which is not a great look for a measurement.
The second problem is where the children came from. Children seen in specialist services are described as having real disability and functional loss [35][36], and children pulled from the general population mostly aren’t [13]. Both of those descriptions are accurate about the children sitting in front of the person writing about them, neither generalises to the other, and a good deal of the arguing in this field is honestly two groups describing genuinely different children, all whilst assuming they’re describing the same ones.
Now, the third is that much of the broader neurodevelopmental material is built by looking backwards through the records of specialist services, rather than following children forwards, which is considerably the weaker way to find out what predicts what [34].
None of this makes the research useless, but it does mean that a confident headline in either direction, whether it’s the reassuring one or indeed the frightening one, is telling you more about which children got measured, rather than about your child.
What This Means for an Actual Child
– Bendy on its own: A child who is hypermobile, hitting her milestones roughly on time and not in pain is a child who is hypermobile. The evidence doesn’t support treating that as a developmental problem, and on the current numbers, the flexibility is most likely to reduce as they grow [4][3].
– Bendy and behind: This is the combination worth acting on, as it’s the one still visible years later [2]. It doesn’t mean it will persist though, it means this is the group where persistence has actually been seen, which is a much better reason to get an assessment.
– A Beighton score has a shelf life: Flexibility changes across childhood and the criteria used for children are deliberately different from the adult ones, so a score taken at six and a score taken at fourteen aren’t really the same measurement [3][4][33]. Ask which threshold was used and how old they were when it was taken.
– Ask about control rather than range: The things that came out lower in hypermobile children were balance, joint position sense, tactile sense and strength [14][15][16][18], and every one of those is about controlling range rather than having it, so an assessment that ends at how far a joint moves has measured the least useful thing available to it.
– School is where it usually surfaces: Handwriting, sitting still, PE and fatigue by the afternoon are the things that get raised first, and they were in the cluster described alongside coordination difficulties [22][23]. A child struggling with those is worth taking seriously whether or not anybody has put a name to her joints yet. In fact, if you are having trouble accessing the support you are legally entitles to, as we do know not all schools are great in this regard, check out our free “Parents School Guide“. It contains a wealth of legal information, in plain terms, template letters, word for word responses to common school push backs, the whole shebang .
– Chasing a label isn’t the same as getting help: The paediatric criteria is genuinely difficult to apply, the phenotype shifts as a child grows, and a score on its own settles very little [3][33]. What changes a child’s week is somebody looking at how they control their body, what they can be taught to do differently, and that doesn’t wait on a diagnosis.
– Symptomatic is a different conversation entirely: Where a child is in pain, exhausted, fainting or struggling at school, the hypermobility is part of a bigger picture, and it has been described that way for roughly twenty years [37]. Children with heritable connective tissue conditions report lower quality of life and poorer mental health than their peers, and that deserves exactly as much attention as the joints do [38].
What Nobody Knows Yet
The gap here is big one, as nobody can tell you which bendy toddlers, become the bendy fifteen year olds with pain, and which ones become the bendy fifteen year olds who are absolutely fine. Working out what marks that transition, early enough to be useful, is the single most valuable thing this field could produce [34][39].
Underneath that sits the same four holes in every corner of this topic: not enough children followed forwards for long enough (like I mentioned earlier), very little at population level, a mechanism that remains unproven, and criteria still applied inconsistently from one place to the next [34][39][33]. Whether treating early delay changes anything at fifteen is almost entirely unstudied, which is uncomfortable, given how much confident advice gets handed to parents.
The Fibro Guy
