Article by Michel Y. Farhat, Ph.D and Elias F. Ilyia, Ph.D
The use of saliva as a vehicle for determination of plasma steroid hormone levels has increased dramatically in recent years. Since 1983, more than 2500 papers and research articles dealing with salivary diagnostic tests have been published1. Both clinicians and investigators have used saliva to assess numerous clinical problems including, digitalis toxicity, celiac disease, liver function and immunodeficiency. Saliva has also been used for pharmakicokinetic studies and therapeutic drug monitoring in a variety of clinical situations. The value of saliva as a monitoring medium resides in the fact that it is easy to collect, store and ship, and is non-invasive, thus convenient for multiple sampling.Plasma-saliva transfer
The reliability of saliva testing depends on establishing a direct correlation between saliva and plasma concentration of a particular substance. The transfer of substances from plasma into the saliva is dependent on their physiochemical properties1. A small molecular weight and a great lipid solubility are normally associated with a faster transfer rate. A good correlation has also been established between saliva/plasma ratio of substances, their pKa and salivary pH. Salivary flow rate and the existing pathophysiology of the oral cavity have also been shown to affect salivary distribution of substances.
Saliva concentration of a particular hormone is dependent on the affinity and total binding capacity of various binding proteins in plasma. As blood passes through salivary glands, free "unbound" and weakly bound (low affinity binding protein) forms of hormones will diffuse through the salivary gland epithelium into the saliva. As in other clearing organs, membrane transfer occurs in both directions, is passive for most substances and equilibrium is governed by the transmembrane concentration gradient. Thus, saliva levels reflect the free concentration of hormones in plasma, and in the absence of high affinity, high capacity-binding proteins, these levels correlate with plasma concentrations. On the other hand, hormones that have high affinity, high capacity-binding proteins, such as thyroxine, are difficult to assay in the saliva. These hormones have a very high plasma total to free hormone ratio and exist in small amounts in saliva. Salivary steroid hormone analysis
Monitoring plasma steroid levels is essential for the clinical assessment of a patient's endocrine function. Saliva becomes an important diagnostic tool, since in many instances, the standard plasma and urine sampling techniques may not provide the optimum sampling conditions. Some of the problems that diagnostic laboratories had to overcome in establishing the validity of salivary steroid assays, were to determine whether steroid concentration in saliva can be measured with accuracy and whether these small values are meaningful and correlate with plasma levels or any other physiological parameter. ,Salivary steroid levels tend to be much lower than those in plasma, because they reflect the level of unbound steroid, which represents about 2-5% of total plasma concentration. Salivary glands may also transform certain steroids during their passage across the salivary epithelium3. Highly polar molecules, such as sugars and conjugated steroids do not cross the lipid lining of cells and can only get into whole saliva via blood contamination from small cuts and ulcers or from gingival fluid. Here we review the pharmacodynamics and partitioning of few hormones, that have been characterized in saliva and evaluate the correlations between their saliva and plasma concentrations.
Progesterone was one of the first steroids to be reliably assayed in human saliva. Once saliva samples are collected, salivary progesterone concentrations remain stable under a wide range of handling conditions. Because of its high circulating blood levels (ng/ml during luteal phase), saliva concentrations of the hormone usually remain within the limits of sensitivity of most conventional assays. Average salivary progesterone concentrations vary from 20-100 pg/ml during the follicular phase to 100-500 pg/ml during the luteal phase (Figure 1). Progesterone levels are also affected by women's age, degree of activity, nutrition and race 4,5. A high correlation coefficient (between 0.8 and 1.0) has been established between plasma and salivary levels of the hormone 6,7, making saliva a useful diagnostic tool. Serial measurement of salivary progesterone has been used to assess ovarian function, and in particular for diagnosis of defective or inadequate luteal function, as well as to monitor response to hormone therapy. Other clinical applications may include monitoring placental function by repetitive measurement of salivary progesterone during pregnancy 9. Estradiol
Salivary estradiol (E2) is about 1-2% of total plasma values, Earlier studies, using assays with variable sensitivities and specificities, have described a wide range of salivary estradiol levels. Three independent studies using radioimmunoassay10-12, enzyme immunoassay13 and chemi-luminescence immunoassay14 have reported comparable salivary E2 levels during normal nonstimulated menstrual cycles ranging from 5-15, 10-30 and 7-20 pmol/L during the follicular, periovulatory and luteal phase, respectively (Figure 2). These levels in stimulated cycles ranged from 10 to 120 pmol/L15. Moreover, Wong et al.10 showed a mid-cycle salivary E2 peak, corresponding to the mid-cycle LH surge, followed by a midluteal rise. In these women, the changes in salivary E2 were similar to the E2 pattern in serum, with a high degree of correlation (r=0.93). A similar correlation (r=0.96) was observed between salivary E2 + estrone and free plasma E2 in FSH-stimulated cycles13. Testosterone Under physiological conditions, a very good correlation exists between salivary and serum testosterone (T) in men16,17. Wang et al. have demonstrated that following exogenous T
Salivary cortisol exhibits clear diurnal variation and circadian rythmicity both in normal and depressed individuals. Salivary cortisol is highest in the morning and varies between 13 and 23 nmoles/L These levels decrease significantly during the day and reach their lowest value at night (1-3 nmoles/L) (Figure 3). Dehydroepiandrostenedione (DHEA)
DHEA ia an adrenal steroid produced in abundant amounts and is conjugated to sulfate to form DHEAS before its release into the circulation 32. DHEA and DHEAS are interconvertable and exist in dynamic equilibrium with each other33. Salivary DHEAS is found in saliva at about 0.1% of its plasma concentration. Serum fluctuations in DHEA(S) concentrations are accurately and rapidly reflected in salivary levels.34Other hormones
Salivary estriol levels are also high, easy to measure and correlate well with plasma unbound unconjugated estriol 35,36. Androstenedione can also be measured with sufficient accuracy in saliva. The absence of specific high affinity binding proteins for androstenedione, results in a linear relationship between plasma total and plasma free fraction, and hence an excellent correlation between plasma and saliva levels of the hormone 37.
The above studies clearly demonstrate that saliva is a useful diagnostic tool for measurement of steroid hormones. Salivary concentration represents the free form of a particular hormone, and thus is a true reflection of its bioactivity. Moreover, the non-invasive nature of saliva collection and the convenience of multiple sample facilitate the design of functional assays for the assessment of various endocrine functions. In a recent review, Mandel38 has likened saliva to a mirror reflecting the emotional, hormonal, immunological as well as nutritional and metabolic status of the body. The broad spectrum of interactions and relationships among these factors opens a whole field of diagnostic possibilities worth exploring and evaluating.
1. Malamud D. Br Med J. 305: 207-208, 1992.
2. Mandel ID. J Oral Pathol 19: 119-125, 1990.
3. Ferguson DB. Ed. Front Oral Physiol. Karger, Basel: 5: 1-162, 1984.
4. Lipson SF and Ellison PT. Am J Human Biol 1:249-255, 1989.
5. Frisch RE. Hum Reprod 2: 521-533, 1987.
6. Meulenberg PM and Hoffman JA. Clin Chem 35:168-172, 1989.
7. Vuorento T, Lahti A, Hovatta O and Huhtaniemi I. Scand J Clin Lab Invest 49:395-401, 1989.
8. Walker SM, Walker RF and Riad-Fahmy D.Horm Res 20: 231-240, 1984.
9. Connor ML, Sanford LM and Howland BE. Can J Physiol Pharmacol 60: 410-413, 1982.
10. Wong YF, Mao K, Panesar N, Loong EPL, Chang AMZ and Mi ZJ. Eur J ObstetGynecol Reprod Biol 34: 129-135, 1990.
11. Evans JJ, Steward CR, Merrick AY. Br J Obstet Gynaecol 87:624-626, 1980.
12. Berthonneau J, Tanguy G, Janssens Y, et al. Human Reprod 4: 625-628, 1989.
13. Mounib N, Sultan CH, Bringer J, Hedon B, Nicolas JC, Cristol P, Bressot N and xxxDescomps. J Steroid Biochem 31:861-865, 1988.
14. Deboever J, Kohen F, Bouve J, Leyseele D and Vanderkerckhove D. Clin Chem 36: 2036-2041, 1990.
15. Smith RG, Besch PK, Dill B, Buttram Jr VC. Fertil Steril 31: 513,1979.
16. Walker RF, Wilson DW, Read GF, Riad-Fahmy D. Int J Androl 3: 105,1980.
17. Gaskell SJ, Pike AW, and Griffiths K. Steroids 36: 219-228, 1980.
18. Baxendale PM and James VHT. In: Immunoassays for Clinical Chemistry. Hunter WM and Corrie JET, eds. Churchill Livingstone, New York. pp. 430-444, 1983.
19. Gould VG, Turkes AO and Gaskell SJ. J Steroid Biochem 24:563-567, 1986.
20. Read GF, Harper ME, Peeling WB and Griffiths K. Int J Androl 4: 623-627, 1981.
21. Cook NJ, Read GF, Walker RF, Harris B and Riad-Fahmy D. Eur J Appl Physiol 55: 634-638. 1986.
22. Vining RF, McGinley RA, Maksvytis JJ and Ho KY. Ann Clin Biochem 20: 329-335, 1983.
23. Laudat MH, Cerdas S, Fournier C, Giuban D et al. J Clin Endocrinol Metab 66: 343-348, 1988.
24. Vining RF, McGinley RA and Symons RG. Clin Chem 29: 1752:1756, 1983.
25. Umeda T, Hiramatsu R, Iwaoka T, Shimada T, Miura F and Sato T. Clin Chem Acta 110: 245-253, 1981.
26. Peters JR, Walker RF, Riad-Fahmy D and Hall R. Clin Endocrinol 17: 583-592, 1982.
27. Kahn JP, Rubinow DR, Davis CL, Kling M and Post RM. Biol Psychiatry 23: 335-349, 1988.
28. Tunn S, Mollmann H, Barth J, Derendorf H and Krieg M. Clin Chem 38: 1491-1494, 1992.
29. Cook NJ, Read GF, Walker RF, Harris B and Riad-Fahmy D. Eur J Appl Physiol 55: 634-638, 1986.
30. O'Connor P and Corrigan DL. Med Sci Sports Exdrcise 19: 224-228, 1987.
31. Tarui H and Nakamura A. Aviat Space Environ Med 58:573-575, 1987.
32. Guechot J et al. Neuropsychobiology 18: 1-4, 1987.
33. Ratcliffe JG. In: Adrenal Cortex, Eds. Anderson DC, Winter JD. pp. 188-205, 1985.
34. Walker R et al. 9th Tenovus Workshop, Cardiff, UK, 1982.
35. Robinson J, Walker S, Read GF and Riad-Fahmy D. Lancet i: 1111-1112, 1981.
36. Fischer-Rasmussen W, Gabrielsen MV and Wisborg T. Acta Obstet Gynecol 60: 417-420, 1982.
37. Turkes AO and Read GF. In: Immunoassays of steroids in saliva. Read GF, Riad-Fahmy D, and Walker RF and Griffiths K, eds. pp. 228-238, 1984 Alpha Omega, Cardiff.
38. Mandel ID. Ann NY Acad Sci 694: 1-10, 1993.