Methods. Anandamide levels were measured in plasma from 45 healthy pregnant nonsmokers (TABLE) presenting to the early pregnancy assessment unit at 6 to 12 weeks’ gestation with a threatened miscarriage (painless vaginal bleeding associated with a viable pregnancy). The assay was conducted with a high-performance liquid chromatography–mass spectrometry isotope dilution method (Waters Micromass Quattro Premier Mass Spectromer; Waters Corporation, Milford, Massachusetts). 6 The extraction and quantification were undertaken within 2 hours of blood collection. Patients, clinicians, and researchers were blinded to the results during the follow-up period. Based on published anandamide data, 5, 6 a minimum of 6 spontaneous miscarriages and 6 live births would allow a clinically significant difference of 40% in anandamide concentration to be observed with 80% power, assuming 2-sided=. 05. Ethics committee approval for the study and written informed consent from each volunteer were obtained; participants did not receive financial compensation. Groups were compared using the Mann-Whitney U test; significance was set at 2-sided P=. 05. Data were analyzed using InStat Version 3.01 (GraphPad Software Inc, San Diego, California), and the area under the receiver operating characteristic curve was calculated using Stata 9 (StataCorp, College Station, Texas).

Results. Of the 45 women, 9 subsequently miscarried and 36 had live births. Both groups had similar characteristics (Table) and underwent similar treatment. The median plasma anandamide concentration in the miscarriage group (3.47 nM; interquartile range, 2.83-3.86) was approximately 3-fold that in the live birth group (1.07 nM; interquartile range, 0.81-1.45; P. 001)(FIGURE). All women who miscarried had anandamide values greater than 2.0 nM; 34 of the 36 in the live birth group (94.4%) had anandamide values less than 2.0 nM. One of the 2 live-birth patients with values greater than 2.0 nM developed severe preeclampsia and delivered a 1.85-kg growth-restricted infant at 33 weeks; the other had an uncomplicated delivery of a 3.7-kg full-term infant. There were no distinguishing characteristics of outliers.