Brisk activity tied to slightly lower insulin and blood fats in 12,874 children

The World Health Organization asks children and teenagers to be active for an average of at least an hour a day. Most fall short: worldwide, 80% of adolescents do not meet the recommended levels of physical activity. Part of the case for that hour is the heart and metabolism. Active children have long been tied to better blood pressure, insulin and cholesterol, and WHO links more sitting to poorer heart and metabolic health.

A new pooled study suggests the link with activity is much weaker than that case assumed. Researchers combined 14 groups of children, 12,874 in all, whose activity was measured with the same kind of motion sensor and who were checked again a few years later. More moderate to vigorous activity went with slightly lower insulin, LDL cholesterol and triglycerides, the main fats in the blood, and was not linked to body mass index, or BMI. Total time spent sitting was not linked to a worse overall profile in the main analysis. The analysis was led by Anders Husøy, with Ulf Ekelund and Jakob Tarp of the Norwegian School of Sport Sciences as senior authors. It was published on October 6 in JAMA Pediatrics.

Why the evidence on children’s activity needed a longer look

Most of what was known came from snapshots. Earlier studies mostly measured activity and health at the same moment, and the few that followed children forward were small. The new paper’s authors make the same point, describing the evidence as resting on single-moment and small prospective studies, and set out to do better.

A snapshot cannot say which came first. A child carrying extra weight may move less because of it, and that alone can make activity look protective. The International Children’s Accelerometry Database, a research consortium that also supplied part of the new data, found the classic pattern in 2012. Pooling sensor data from 20,871 children at a single point in time, it reported in the Journal of the American Medical Association that activity went with better readings on every marker measured. Time spent sitting, by contrast, was not associated with any outcome once brisk activity was taken into account.

The World Health Organization’s own 2020 evidence review was cautious too. It found the benefits of activity in children lay mostly in risk markers, not in illness itself, and said there was still insufficient evidence available to fully describe the dose-response relationship: nobody could yet say how much activity buys how much benefit.

How 12,874 children wore accelerometers in 14 cohorts

The team gathered individual records from six groups in the International Children’s Accelerometry Database and eight more found through a search of published research. The children came from 16 countries on four continents. Each had worn an ActiGraph accelerometer, a small motion sensor clipped to the hip, for at least three days. Activity was counted by the device, not recalled by a parent.

The researchers reprocessed all the raw sensor data the same way and analyzed the children as one large group. That makes this an individual-participant meta-analysis, which is stronger than averaging the published results of separate studies, because every child is handled by the same rules.

The children were 5 to 17 years old, 10 on average, and about half were girls. At the start, 18.5% were overweight and 8.7% had obesity. Six of the studies were school-based interventions that added physical education classes, promoted activity across a whole school or provided school meals, and the analysis allowed for which children had been in them. The children’s markers were measured again after a mean follow-up of 3.1 years. The markers were BMI, systolic blood pressure, and fasting levels of blood sugar, insulin and the two blood fats. The team also combined them into a single risk score, scaled so that one unit equals the usual spread between children.

The main analysis took account of age, sex, puberty, family income or education, birth weight and ethnicity. It also allowed for each child’s BMI at the start, a choice that turned out to matter a great deal.

What 26 extra minutes of activity a day was linked to

Results were expressed per 26 extra minutes a day of moderate to vigorous activity, the usual spread between children. In the main model, each extra 26 minutes was tied to modestly lower insulin and blood fats three years on. The combined score improved by 0.04 of a unit.

To see how small that is, consider a pooled study of childhood cohorts from the United States, Finland and Australia, published in the New England Journal of Medicine. It followed 38,589 people from childhood for a mean of 35 years. A full unit higher on a similar childhood risk score went with 2.71 times the risk of a fatal cardiovascular event in adulthood. The change linked to activity in the new study is about one twenty-fifth of that unit.

Body mass index did not budge with more activity. Systolic blood pressure was a fraction of a point higher, an unexpected direction. Vigorous activity on its own looked much the same, without the blood pressure link.

The links showed up only in some children. They appeared in under-13s and in groups followed for less than two years, and they were strongest in children who started out heavier. In teenagers and over longer follow-up there was nothing. The authors suggest that the normal changes of puberty, in body fat, insulin, blood pressure and blood fats, could potentially mask effects of activity.

There was a dose effect of sorts. Children in the least active fifth averaged about 19 minutes a day of brisk activity. Only those in the most active fifth, at about 93 minutes a day, showed clearly lower odds of high clustered risk at follow-up. That is well above the hour WHO recommends.

The authors write that their results do not confirm the strong links between activity and risk markers that snapshot studies had shown.

Sitting time and children’s heart and metabolic health

The result for sitting was more surprising. WHO’s fact sheet links more sedentary time in children and adolescents to more body fat and to poorer cardiometabolic health, fitness and behavior. In this study, each extra 106 minutes a day of sitting went with slightly lower blood sugar and blood pressure. The combined score was no worse.

The pattern split by sex and weight. More sitting went with a slightly better score in girls and the leanest children, and a slightly worse one in boys and the heaviest.

The authors conclude that sedentary time “is unlikely an important determinant of cardiometabolic risk in youth.” They add a limit to that, since a hip sensor captures total sitting, whatever a child is doing. Their result, they write, “does not preclude negative, possibly indirect, consequences of specific sedentary behaviors such as screen time.” Screens, in other words, could still matter in other ways.

Why small changes in children’s risk markers may still matter

Children’s risk markers matter because they track into adult disease. In that New England Journal of Medicine study, the risk of a fatal cardiovascular event in adulthood rose with every childhood risk factor it measured. A 2024 analysis of 11,550 people from seven of those groups found that overweight, raised blood pressure, smoking and high cholesterol in the teenage years were predictors of cardiovascular events after age 25.

The Norwegian-led team grants that small shifts across a whole population are potentially important for public health, and that modest effects might add up if they last for years. Randomized trials also lend some support to a direct effect. A meta-analysis of 24 exercise trials in young people found a small to moderate effect on fasting insulin from exercise training. Those trials were short and supervised, though, and the new study’s authors point out that brief trials are less muddied by the changes of puberty and may not reflect everyday life.

What the pooled activity data leave open

This is an observational study, and the authors expect leftover confounding, the influence of unmeasured differences between children, to make the links look stronger than they are. They could adjust only for factors every group had recorded. Diet quality, sleep, and other behavioral and socioeconomic factors were left out.

The choice of model mattered a great deal. Without allowing for children’s starting body mass index, more activity went with lower later BMI. More sitting went with higher insulin and blood fats. Once starting BMI was included, most of those links shrank, disappeared or reversed. The authors chose the adjusted model as their main one because they judged the simpler model likely to be confounded. The authors acknowledge that if activity had already lowered children’s weight before the study began, adjusting for that weight removes part of a real effect.

The groups used standardized but not identical lab methods, body mass index cannot tell fat from muscle, and follow-up was short. The study says nothing directly about adult heart disease, and 94.5% of the children were White or born in the study country.

What it means for activity guidelines for children

The authors do not argue against activity. They suggest that if heart and metabolic markers are the goal, more than the WHO’s 60 minutes may be needed. They also write that activity may not prevent excess weight gain in youth. WHO’s case for activity rests on more than these markers. Its fact sheet notes that in children and adolescents, physical activity promotes bone health, healthy growth and motor and cognitive development.

Whether a particular child’s activity, weight or blood results need attention is assessed by their doctor.

The pooled data show children’s activity going with slightly better heart and metabolic markers over a few years, more faintly than earlier studies suggested, and lend little support to the idea that sitting itself is the problem.