Geriatric Sleep Research Program

Home > Research > Comparison of Dim Light Melatonin Onset Algorithms for Serum Melatonin in Older Adults

Title

Comparison of Dim Light Melatonin Onset Algorithms for Serum Melatonin in Older Adults

Abstract

One of the major challenges in chronobiology is accurately determining a subjectís circadian phase in order to identify if it is advanced or delayed. A variety of tools have been proposed to study this, including constant routine protocols which measure core body temperature or overnight melatonin sampling in dim light conditions. While other markers of circadian phase, such as core body temperature, are influenced by food and sleep-wake statues, the melatonin secretion profile is relatively unaffected by these considerations, thus making it particularly useful. In particular, the dim light melatonin onset (DLMO), which is the initial surge in melatonin release in the early part of the night under low light conditions (<50 lux), has been found to be a consistent and reliable measure of the intrinsic circadian phase.

A variety of algorithms exist to identify the DLMO. Initial studies used a fixed threshold of 10 pg/ml: the DLMO was defined as the first interpolated value to exceed that threshold. Recent work based on saliva melatonin has included alternative algorithms such as defining the DLMO as the first specimen that is two standard deviations above the baseline mean. However, it is unclear if the current algorithms used to define the dim light melatonin onset are robust enough to calculate the DLMO in low melatonin conditions. Given that older subjects tend to have lower melatonin than younger subjects, the former would serve as an ideal population in which to study the feasibility of DLMO calculation in low melatonin conditions. Furthermore, it is unclear if saliva melatonin can be a useful marker of DLMO in this group because of issues such as hyposalivation and low melatonin levels. In this study, we examined the feasibility of a variety of DLMO algorithms to identify the DLMO in elderly subjects using both saliva and serum specimens. We limited our analysis to DLMO algorithms that do not require a complete 24-hour melatonin profile and do not involve curve/model fitting. Development of robust, standardized and communicable techniques for DLMO analysis is a crucial component of advancing the field of melatonin research. Our intents was to identify DLMO algorithms that may be applicable for subjects with low melatonin levels, involve minimal subject burden, and are thus more robust tools for large-scale epidemiology studies.