Background: Circadian and ultradian rhythms in BG have been observed in individuals without diabetes and are regulated both by time of day and meal intake, with ultradian fluctuations generally of smaller amplitude than circadian fluctuations. We assessed relative stability of circadian and ultradian rhythms in individuals without T1D (n=16), and individuals with T1D using two different therapies: sensor-augmented pump therapy (SAPT, n=100) and DIY closed loop (n=30).

Methods: Frequency decomposition using continuous morlet wavelet transformation (A to C) were created to assess change in rhythmic composition of normalized blood glucose data.

Results: Example Wavelet (C) in T1D illustrates high power in the 1-6H range, and lower power in the ~24H range for nearly all time points. The ratio of power in the 23-25H (circadian) range to power in the 1-6H (ultradian) range was calculated to estimate relative stability of circadian to ultradian rhythms. Those without T1D showed higher circadian compared to ultradian power, whereas both groups with T1D showed a comparably lower ratio (D).

Conclusion: Wavelet analysis of BG in individuals with and without T1D reveals a difference in relative strength of daily and hourly rhythms, independent of therapy type. These results suggest current therapies do not recreate expected daily rhythmicity in blood glucose, perhaps due the lifespan of rapid-acting insulin and meal patterns.


D.M. Lewis: Research Support; Self; JDRF. A.D. Grant: None. L. Kriegsfeld: None.



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