Skip to main content
. 2023 Jul 31;3(8):100545. doi: 10.1016/j.crmeth.2023.100545

Figure 4.

Figure 4

Amplitudes and peak times of circadian baseline levels of glucose are highly heterogeneous between participants

(A) The amplitude of the 24-h sinusoidal circadian rhythm in baseline glucose levels after model fitting to the CGM data for all participants. The boxes represent the 25th, 50th (median), and 75th percentiles of the posterior distribution and the whiskers represent the 5th and 95th percentiles.

(B) The circadian peak time of the glucose circadian rhythm across all participants.

(C) Examples of participants with a high (ID 07) and a low (ID 03) amplitude glucose circadian rhythm. Blue, CGM data; black, fitted model of circadian baseline (using mean posterior parameter values).

(D) Examples showing two participants with large phase difference in underlying glucose rhythm (ID 20 peak phase 10:00, ID 17 peak phase 20:00).

(E) The explained variance in glucose levels using just the meal component of the model (light orange) compared with the inclusion of the circadian rhythm (dark orange). Error bars represent the 5th and 95th percentiles of the posterior distribution. Participant order is the same in Figures 3A, 3B, 4A, and 4B.

See also Figure S10.