Measuring melatonin by immunoassay

The pineal gland hormone melatonin continues to be of considerable interest to biomedical researchers. Of particular interest is the pattern of secretion of melatonin in relation to sleep timing as well as its potential role in certain diseases. Measuring melatonin in biological fluids such as blood and saliva presents particular methodological challenges since the production and secretion of the hormone are known to be extremely low during the light phase in almost all situations. Active secretion only occurs around the time of lights out in a wide range of species. The challenge then is to develop practical high-throughput assays that are sufficiently sensitive and accurate enough to detect levels of melatonin less than 1 pg/mL in biological fluids. Mass spectrometry assays have been developed that achieve the required sensitivity, but are really not practical or even widely available to most researchers. Melatonin radioimmunoassays and ELISA have been developed and are commercially available. But the quality of the results that are being published is very variable, partly not only because of poor experimental designs, but also because of poor assays. In this review, I discuss issues around the design of studies involving melatonin measurement. I then provide a critical assessment of 21 immunoassay kits marketed by 11 different companies with respect to validation, specificity and sensitivity. Technical managers of the companies were contacted in an attempt to obtain information not available online or in kit inserts. A search of the literature was also conducted to uncover papers that have reported the use of these assays, and where possible, both daytime and night-time plasma or saliva melatonin concentrations were extracted and tabulated. The results of the evaluations are disturbing, with many kits lacking any validation studies or using inadequate validation methods. Few assays have been properly assessed for specificity, while others report cross-reaction profiles that can be expected to result in over estimation of the melatonin levels. Some assays are not fit for purpose because they are not sensitive enough to determine plasma or saliva DLMO of 10 and 3 pg/mL, respectively. Finally, some assays produce unrealistically high daytime melatonin levels in humans and laboratory animals in the order of hundreds of pg/mL. In summary, this review provides a comprehensive and unique assessment of the current commercial melatonin immunoassays and their use in publications. It provides researchers new to the field with the information they need to design valid melatonin studies from both the perspective of experimental/clinical trial design and the best assay methodologies. It will also hopefully help journal editors and reviewers who may not be fully aware of the pitfalls of melatonin measurement make better informed decisions on publication acceptability.