Physical activity and cardiometabolic markers in children: real but weak associations

Design

Harmonised individual participant data meta-analysis of 14 prospective cohorts

Population

12,874 children and adolescents, mean age 10.4 years, mean follow-up 3.1 years

Primary outcome

BMI, insulin, glucose, LDL cholesterol, triglycerides and systolic blood pressure, individually and as composites

Effect

Per 26 min/day MVPA: insulin -0.24 µIU/mL (95% CI -0.38 to -0.10); LDL -1.54 mg/dL; BMI 0.00

This harmonised individual-participant-data meta-analysis combined 14 cohorts with 12,874 children and adolescents (mean age 10.4 years, 52% female, mean follow-up 3.1 years). Activity and sedentary time were measured by accelerometer, and later cardiometabolic markers were analysed with models adjusted for confounders chosen through causal diagrams.

Per standard deviation of moderate-to-vigorous activity (26 minutes a day), insulin was lower by 0.24 µIU/mL (95% CI -0.38 to -0.10), LDL cholesterol by 1.54 mg/dL and triglycerides by 1.77 mg/dL, and a composite risk score by 0.04. BMI was unchanged and systolic blood pressure was slightly higher (0.25 mm Hg, 0.04 to 0.46). Sedentary time was not linked to a worse profile; it was linked to lower systolic pressure and slightly lower glucose.

The authors describe the associations as weak, sensitive to the modelling approach and different across subgroups. These are observational associations, with small absolute differences. They support encouraging activity but do not show how much it would change any child's later risk, and they argue against assuming that sedentary time itself is a major driver.

• Encourage moderate-to-vigorous activity in children for general health; the cardiometabolic effect per extra 26 minutes a day was small.
• Do not expect BMI to change with more activity alone; no association was found.
• Sedentary time was not linked to a worse cardiometabolic profile in this analysis.
• Individual effects varied by sex, BMI, age and follow-up, so avoid precise promises.

Why it matters

It questions the assumption that sedentary time is a direct cardiometabolic hazard in childhood.

Don't overread it

These are observational associations, weak and sensitive to modelling, and they do not show that increasing activity lowers later risk.

The statistics, in plain English

The numbers are per standard deviation of activity, about 26 minutes a day. A change in LDL of 1.5 mg/dL is statistically clear in 12,874 children but small for an individual, which is why the authors call the associations weak.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.