Indirect Effect Model with Surge Function for Describing Melatonin Circadian Rhythm: Quantitative Comparison and Application between Normal Sleepers and Patients with Delayed Sleep-Wake Phase Disorder.
Level 4 - case-series / case-control
Comparative observational cohort and pharmacokinetic modeling study
PubMed 40288357 · doi:10.1159/000546125
What was done
Researchers developed an indirect effect pharmacodynamic model incorporating a surge function to describe endogenous melatonin circadian rhythms using data from 10 normal sleepers and 26 patients with delayed sleep-wake phase disorder (DSWPD). Model accuracy and stability were evaluated using diagnostic plots, visual predictive check, and bootstrap methods. Monte Carlo simulations (n = 1,000) compared circadian rhythm characteristics between groups. Bayesian estimation was tested using external data from 3 normal sleepers and 3 DSWPD patients to predict melatonin concentrations and dim light melatonin onset (DLMO).
What was found
The model estimated population typical values (relative standard error %): amplitude 7.95 (15.47%), peak time 23:59 (4.13%), peak width 4.12 (5.78%), elimination rate constant 1.23 h-1 (21.82%), and baseline melatonin concentration 3.21 pg/mL (23.27%). Monte Carlo simulations revealed DSWPD patients had an approximately 7-hour delay in DLMO, a significantly lower melatonin production rate constant, and a similar elimination rate constant compared to normal sleepers. Bayesian estimation predicted DLMO with an error margin within ±10%.
Why it matters
The study offers a quantitative pharmacokinetic-pharmacodynamic model showing that circadian phase alterations in DSWPD are characterized by reduced production rates and phase delay rather than altered clearance. The model may allow estimation of DLMO from sparse sampling data.
Limits
The total sample size was small (36 model development participants and 6 external validation participants). The abstract does not provide demographic details, assay type, biological matrix (saliva or blood), sampling intervals, or exact p-values for differences between groups. Findings depend substantially on model assumptions and Monte Carlo simulations.
Cited by
- contradicts Endogenous melatonin begins rising roughly two hours before habitual sleep time, peaks about one hour into sleep, and reaches its lowest level about two hours after waking.