Remote exercise program speeds youth concussion recovery, but mainly if started early

The advice for concussion has shifted over the past decade from strict rest toward carefully dosed activity, but most of the trials behind that change studied fresh injuries and required repeated clinic visits. A randomised trial published on 19 October in JAMA Pediatrics asked a different, more practical question: can a supervised aerobic-exercise program delivered entirely by video help young people whose concussion symptoms have dragged on — and does it still help once weeks or months have passed [s1]?

What was tested

The Mobile Subthreshold Exercise Program, or MSTEP, was compared head-to-head with a stretching program in 11- to 18-year-olds who had a concussion up to a year earlier [s1]. Participants were recruited from outpatient settings across four US states — Washington, Utah, Wisconsin and Oregon — with recruitment and data collection running from March 2021 to October 2025 [s1][s2].

Teenagers assigned to MSTEP were given an activity tracker and an aerobic program that began at 60% of their maximum heart rate, with intensity and duration stepped up over six weeks through weekly videoconferencing calls [s1]. The control group received stretching exercises and similar weekly calls, so both arms got the same contact and attention — isolating the effect of the aerobic exercise itself [s1]. The two primary outcomes were concussive symptoms and health-related quality of life; sleep, depression and anxiety were tracked as secondary outcomes, and everyone was followed for six months [s1].

What the trial found

The trial enrolled 210 participants [s1]. Their mean age was 15.1 years, 122 (58.1%) were female, and the mean time since injury was 73.3 days — well over two months, confirming this was a group with prolonged, not acute, symptoms [s1].

Both groups improved over time, as concussed young people generally do. The question was whether the exercise program improved things faster. At the three-month mark it did: compared with stretching, MSTEP showed greater improvement in both concussive symptoms (β, −3.15; 95% confidence interval, −6.05 to −0.25; adjusted P=.04) and quality of life (β, 4.15; 95% confidence interval, 0.64 to 7.66; adjusted P=.04) [s1]. A separate model of how symptoms decayed over time confirmed a significantly faster rate of symptom recovery in the exercise group (difference in rates, −0.68; 95% confidence interval, −2.22 to −0.04; P=.002) [s1].

By six months, however, the two groups had converged: no significant differences remained between MSTEP and stretching [s1]. The secondary outcomes — sleep, depression and anxiety — improved in all participants but showed no advantage for either group [s1]. So the exercise program appears to speed recovery rather than to reach a better final destination; the control group largely caught up.

The timing catch

The most clinically useful finding is about who benefited. When the researchers looked for factors that changed the effect, only one mattered: time since injury [s1]. The benefit of the exercise program was statistically significant only for young people who were less than three months from their concussion (concussive symptoms: β, −4.44; 95% confidence interval, −7.99 to −0.90; HRQoL: β, 6.64; 95% confidence interval, 2.55 to 10.72) [s1]. There was no evidence that sex or age changed the response [s1].

That pattern argues for starting a structured exercise program sooner rather than later in a prolonged recovery, and tempers expectations for those already many months out.

How to read it

The headline is encouraging and practical: a rehabilitative exercise program for concussion can be delivered remotely, through a tracker and weekly video calls, and still speed symptom recovery in young people [s1]. For families far from a specialist clinic, that accessibility matters.

The caveats are equally clear. The advantage was modest and temporary — real at three months, gone by six — and confidence intervals that nearly touched zero signal an effect that is genuine but not dramatic [s1]. The program also worked only for the more recently injured, and the exercise was individually calibrated and supervised, not unstructured activity to push through at home [s1]. It is a telehealth protocol run by a trial team, not a licence to self-prescribe.

Our broader explainer on youth concussion and returning to play and school sets out the staged, symptom-guided approach these findings sit within. This article is informational and not medical advice; decisions about activity after a concussion should be made with a qualified clinician.