Endocrine Abstracts (2008) 16 P405

Performance characteristics of serum and salivary hormone quantification for luteal phase confirmation in behavioural studies

Gunnar Brandhorst1, Kathrin Behrens2, Sophie Hinrichs2, Kirsten Jordan2, Michael Oellerich1 & Nicolas von Ahsen1


1Department of Clinical Chemistry, Göttingen, Germany; 2Department of Medical Psychology and Medical Sociology, Göttingen, Germany.


Objective: To investigate gender differences in spatial orientation and verbal memory we conducted a study in 21 healthy young women and men (eleven women). Participating women were required to be in the mid-luteal phase of the cycle. Hormone profiling for behavioural studies is often performed in saliva samples to avoid venous blood sampling. We have compared simultaneously collected blood and saliva samples to confirm the presence of the luteal phase.

Methods: Serum and saliva samples were collected from eleven healthy women without oral contraceptive pill (age range 21–26 years) in mid-luteal phase (calculated from self-stated cycle lengths) at 0 and 2 h. Serum progesterone, estrogen, LH and FSH were determined on a Modular E170 system (Roche, Mannheim). Salivary progesterone and estrogen were determined by LIA (IBL, Hamburg).

Biochemical evaluation of the cycle phase was done according to the manufacturers’ reference intervals. First-tier was progesterone concentration. If inconclusive, estrogen and gonadotropin concentrations were second-tiers.

Results: By combined judgement of all parameters 7/11 women were in the luteal phase. All 7 women were correctly classified using only the serum progesterone concentration (100% sensitivity). Non-luteal phase women (4/4) were correctly excluded (100% specificity). Salivary progesterone identified 5/7 women in the luteal phase (71% sensitivity). Non-luteal phase women (3/4) were mostly excluded (75% specificity).

Conclusion: Only in 7/11 (64%) of the participating women the presence of the luteal phase was confirmed by hormone testing. Therefore, hormone testing seems to be justified in neuropsychological research. Classification by serum progesterone alone was as good as classification by progesterone, estradiol, LH and FSH combined. Serum progesterone determination seems to be superior to salivary quantification. In our study an adjusted reference interval for salivary progesterone would improve sensitivity of the salivary progesterone assay, however specificity would not change.

Article tools

My recent searches

No recent searches.