Regular physical activity (PA) has been found to contribute to lower HbA1c in T1D pediatric populations, but this relationship is not conclusive.

In a recent study, we examined associations between PA and HbA1c over 6 months in 1T1D pediatric patients. PA was measured with Fitbit monitors and summarized as average weekly steps and hours of moderate to vigorous PA (MVPA). Linear regression was used to analyze associations and adjust for variables including baseline HbA1c.

Counter to expectations, PA across patients was positively associated with HbA1c. An increase of 10K average steps/week was associated with a 0.115 mmol/mol increase in HbA1c (P=0.017). In addition, an increase of 1 hour of average weekly MVPA was associated with a 0.088 mmol/mol increase in HbA1c (P=0.038). However, within patients, weekly PA levels were negatively associated with weekly BG levels in 75% of patients.

This reversal of trend between within and between patient measures is an interesting example of what is called Simpson’s paradox in statistics. The Figure illustrates how these relationships coexist: weekly data show a negative trend for 3 patients (dashed lines); however, averaged data show a positive trend (solid line).

Although across patients, higher PA tended to be associated with higher HbA1c, the within patient association suggests an encouraging effect in which PA among pediatric patients may actually help to lower HbA1c levels.


D.A. Watson: None. T.L. Barnes: None. L.M. Gandrud: Consultant; Self; UnitedHealth Group.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at