I don't know, but I do know that there are other factors possibly more important, i.e. age of the participant (particularly if the study includes children), and timing of the measurement (e.g. if they've eaten recently). These may be important to stratify on, or explore heterogeneity caused by them.

Hi Darren, 

You're absolutely right. Type of Ghrelin (acyl/Desacyl), Fasting state, BMI (of fatmass), Age, diurnal phase and Type 2 Diabetes are all known factors and I'm going to stratify on them. I'm not sure how large the variation caused by laboratory technique is. I would guess that this is smaller than the the factors listed above, but I'm not sure. 

Thanks for you answer!

Maybe you may find some information on the manufacturer's website.

https://www.merckmillipore.com/NL/en/product/Human-Ghrelin-%28active%29-ELISA,MM_NF-EZGRA-88K

Dear Noëlle, please see the following publications, which all are freely available at the journals website.

Gröschl M, Wagner R, Dötsch J, Rascher W, Rauh M. (2002). Preanalytical influences on the measurement of ghrelin. Clinical Chemistry 48(7):1114-1116.

[http://www.clinchem.org/content/48/7/1114.long]

Gröschl M, Uhr M, Kraus T. (2004). Evaluation of the comparability of commercial ghrelin assays. Clinical Chemistry 50(2):457-458.

[http://www.clinchem.org/content/50/2/457.full.pdf+html]

Hosoda H, Doi K, Nagaya N, Okumura H, Nakagawa E, Enomoto M, Ono F, Kangawa K. (2004). Optimum collection and storage conditions for ghrelin measurements: octanoyl modification of ghrelin is rapidly hydrolyzed to desacyl ghrelin in blood samples. Clinical Chemistry 50(6):1077-1080.

[http://www.clinchem.org/content/50/6/1077.full.pdf+html]

Rauh M, Gröschl M, Rascher W. (2007). Simultaneous quantification of ghrelin and desacyl-ghrelin by liquid chromatography-tandem mass spectrometry in plasma, serum, and cell supernatants. Clinical Chemistry 53(5):902-910.

[http://www.clinchem.org/content/53/5/902.full.pdf+html]

Dear Michael, 

Thank you so much! This is very helpful. 

It will be worth adding RIA/ELISA and serum/plasma for subgroup analyses in the meta-analysis but I can near guarantee that there will be no sig difference. 

Dear Noëlle,

Depending on the assay manufacturer the concentration is highly different. We have run an internal comparison and the value are ranking from 10 to 1000 for the same sample (according to the manufacturer). It doesn't really help for meta-analysis. You have to take care about the trends which should be the same, but not the absolute values reported.

Sincerely