Aardal E, Holm AC. Cortisol in saliva-reference ranges and relation to cortisol in serum. Eur J Clin Chem Clin Biochem 1995;33:927-932.

Parallel serum and saliva samples were acquired in 197 individuals, in the morning and evening. A slight decrease in morning cortisol was seen in the oldest individuals tested (61-70 years). As has been demonstrated in other work, once serum cortisol exceeded ≈ 450 nmol/l, a marked increase in salivary cortisol was noted. This is attributed to a rapid increase in free plasma cortisol once the available sites on cortisol binding globulin are filled. The authors conclude by enumerating the advantages of saliva over serum (simplicity, decreased stress, convenience, sample mail-in, temperature stability).

Aardal-Eriksson E, Karlberg BE, Holm AC. Salivary cortisol- and alternative to serum cortisol determinations in dynamic function tests. Clin Chem Lab Med 1998;36:215-222.

Serum and salivary cortisol responses were compared in various dynamic tests of HPA axis function including insulin tolerance, CRH stimulation, and ACTH challenge. In 42 of 45 tests performed, consideration of salivary cortisol led to the same conclusion as serum cortisol. The authors conclude that due to a more pronounced cortisol response in saliva, and a closer correlation between salivary cortisol and serum ACTH, salivary cortisol may be used as an alternative parameter in dynamic endocrine tests.

Allolio B, Hoffmann J, Linton EA, Winkelmann W, Kusche M, Schulte HM. Diurnal salivary cortisol patterns during pregnancy and after delivery: relationship to plasma corticotrophin-releasing hormone. Clin Endocrinol 1990;33:279-289.

The circadian rhythm of salivary cortisol levels was studied in 10 pregnant women every 4 weeks throughout pregnancy. In 12 women, diurnal variation of salivary cortisol, serum cortisol, plasma ACTH, plasma CRH and serum progesterone was examined late in the third trimester and again 3-5 days after delivery. Mean salivary cortisol levels increased throughout pregnancy, reaching levels almost twice that of controls by late pregnancy, and falling rapidly with delivery. There was no correlation between plasma CRH, plasma ACTH, serum and saliva cortisol in late pregnancy. The authors conclude that although CRH levels increase in late pregnancy (due to placental production), this CRH has very little effect on baseline adrenocortical function. Instead, they hypothesize that the high progesterone levels in late pregnancy induce a state of glucocorticoid resistance, due to the antiglucocorticoid effect of progesterone.

Barrou Z, Guiban D, Maroufi A, Fournier C, Dugue MA, Luton JP, Thomopoulos P. Overnight dexamethasone suppression test: comparison of plasma and salivary cortisol measurement for the screening of Cushing’s syndrome. Eur J Endocrin 1996;134:93-96.

Salivary cortisol was compared to plasma cortisol for screening of Cushing’s Syndrome via overnight dexamethasone suppression test. Compared to plasma, salivary cortisol demonstrated superior performance with 100% specificity and 94% sensitivity (no overlap between controls and patients with Cushing’s Syndrome). In addition, salivary cortisol levels in women taking oral contraceptives were normal whereas the corresponding plasma levels were elevated (false positive) in some cases.

Belkien LD, Bordt J, Moller P, Hano R, Nieschlag E. Estradiol in saliva for monitoring follicular stimulation in an in vitro fertilization program. Fertil Steril 1985;44:322.

Serial serum and saliva measurements were compared in 23 patients undergoing ovulation induction in an in vitro fertilization program. Serum and saliva were well correlated (r=0.77) throughout, except for salivary estradiol levels less than 4 pg/ml. (Note that estradiol can easily be detected below this level with modern methods.) The authors conclude that, “assessment of ovarian function in clomiphene citrate (CC) or CC/hMG stimulated cycles can be performed precisely with the saliva estradiol assay.”

Bolaji II, Tallon DF, O'Dwyer E, Fottrell PF. Assessment of bioavailability of oral micronized progesterone using a salivary progesterone enzymeimmunoassay. Gynecol Endocrinol 1993;7:101-110.

These authors looked at salivary progesterone levels in 40 postmenopausal women taking 100 mg oral micronized progesterone in cyclical fashion (23 days/month) for one year. These levels fell into the range observed in the luteal phases of a control group of 40 normally cycling women. Five patients had simultaneous serum and saliva levels of progesterone measured for 12 hours after ingestion of a 100 mg dose of progesterone. Levels in both cases followed the same time curve, with both peaking at 2 hours. This supports the notion that progesterone diffuses rapidly from serum to saliva.

Booth A, Johnson D, Granger D, Crouter A, McHale S. Testosterone and child and adolescent adjustment: the moderating role of parent-child relationships. October 2000. Unpublished manuscript.

This study examined the relationship of salivary testosterone to self-reported risk behavior or depressive symptoms in adolescents. It was found that correlations between testosterone level and behavior were modulated by the quality of the parent-child relationship. The authors note, “Salivary measures seem to be especially suited to studies of youth because sampling is painless, acceptance rates are higher, and bias caused by self-selection may thus be reduced.”{mospagebreak}

Campbell BC, Ellison PT. Menstrual variation in salivary testosterone among regularly cycling women. Horm Res 1992;37:132-136.

Daily first morning saliva samples were collected from 20 regularly cycling women over the course of an entire menstrual cycle and assayed for testosterone and progesterone. Cycles were classified as ovulatory if salivary progesterone exceeded 100 pg/ml on any day in the cycle. Average testosterone level was significantly greater in anovulatory cycles. In both types of cycle, the same general features were observed: a luteal trough, a midcycle peak, and an early follicular peak, although in non-ovulatory cycles, the midcycle peak was longer, and the luteal trough shorter. These findings are consistent with what is known about changes in ovarian steroid metabolism throughout the menstrual cycle. Salivary assay is particularly well-suited to studies of this nature, which involve multiple measurements, since samples can be acquired with minimal inconvenience.

Castro M, Elias PC, Quidute AR, Halah FP, Moreira AC. Out-patient screening for Cushing’s Syndrome: the sensitivity of the combination of circadian rhythm and overnight dexamethasone suppression salivary cortisol tests. J Clin Endocrinol Metab 1999;84:878-882.

These authors measured salivary cortisol levels in normal individuals and patients with Cushing’s Syndrome. A single measurement at 23:00 hours produced a sensitivity and specificity of 93% for detection of Cushing’s Syndrome. When a 23:00 hour measurement was combined with an overnight 1 mg dexamethasone suppression test, 100% sensitivity was achieved. The authors conclude, the collection of saliva rather than blood is a practical, simple and noninvasive out-patient alternative to screen for hypercortisolism.”

Choe JK, Khan-Dawood FS, Dawood MY. Progesterone and estradiol in the saliva and plasma during the menstrual cycle. Am J Obstet Gynecol 1983;147:557-562.

Daily saliva and blood samples were obtained from 9 women throughout an entire menstrual cycle. Saliva and plasma levels of progesterone correlated reasonably well (r=0.58) although the saliva and blood samples were not acquired at the same time. The percentage of protein binding of progesterone in serum and saliva was also measured by equilibrium dialysis. Forty to fifty percent of the progesterone in saliva was protein bound, compared to three to six percent in plasma. However, unbound progesterone levels in plasma and saliva were comparable, supporting the contention that salivary progesterone represents unbound progesterone in the blood.

The authors also measured salivary and plasma estradiol but did not display or mention the plasma data, commenting only that there was no correlation between levels in the two matrices. This casts suspicion on the validity of their plasma estradiol data.

Christiansen K, Knussmann R. Androgen levels and components of aggressive behaviour in men. Horm Behav 1987;21:170-180.

Saliva and serum testosterone as well as serum dihydrotestosterone were measured in normal males and correlated to measures of aggression. Weak but significant correlations (r ≈0.2) were seen between self-rated spontaneous aggressive tendency and both saliva and serum testosterone levels.

Christiansen K, Knussmann R. Sex hormones and cognitive functioning in men. Neuropsychobiol 1987;18:27-36.

Parallel blood and saliva samples were obtained from 117 men. Cognitive function (verbal and spatial ability ) was then assessed and correlated to androgenic hormone levels. Serum testosterone and dihydrotestosterone levels, as well as salivary testosterone levels correlated positively, though weakly to spatial ability scores, and negatively with verbal measures, in keeping with the known sex differences in cognitive function.

Clements AD, Parker CR. The relationship between salivary cortisol concentrations in frozen versus mailed samples. Psychoneuroendocrinol 1998;23:613-616.

Saliva specimens were obtained from 17 volunteers. Half of each specimen was immediately frozen. The other half was subjected to varying temperatures and physical handling intended to mimic conditions experienced during mailing. Both sample halves were ultimately assayed for cortisol, and a correlation coefficient of 0.92 was derived between the quick-frozen and “mailed” specimens. The authors conclude that mailing saliva samples intended for cortisol assay is acceptable.

Dabbs JM. Salivary testosterone measurements: collecting, storing and mailing saliva samples. Phys Behav 1991;49:815-817.

The authors assessed the effect of saliva storage time and storage temperature, the use of stimulants to collect saliva, as well as the effect of cotton swabs for sample collection in salivary testosterone assays. Cotton swabs were found to contribute an apparent testosterone concentration ranging from 20 to 70 pg/ml. Sugar-containing gums elevated apparent testosterone levels by ≈ 30 pg/ml. Saliva samples could be stored unrefrigerated for up to 1 week with no effect on apparent testosterone level (specimens from women experienced an increase in apparent testosterone concentration if stored for more than a week at room temperature). The effect of storage time and temperature on apparent testosterone concentration was only studied on 2 men and 2 women, however. The authors conclude, “Salivary measurements provide a robust indicator of subjects’ testosterone levels. “

Dabbs JM. Salivary testosterone measurements: reliability across hours, days and weeks. Phys Behav 1990;48:83-86.

In this study, correlations between sequential measurements of salivary testosterone in men and women across intervals of hours, 2 days, 2 weeks, 5 weeks and 9 weeks were studied. The expected circadian variation, with highest levels in the a.m. was observed. Not surprisingly, correlations between measurements taken weeks apart were only modest, and worsened as the sampling interval increased. Random salivary testosterone levels were obtained from 126 cycling women and plotted as a function of time since last menstruation. No tendency toward a midcycle peak was observed, leading the authors to conclude that in behavioural studies, saliva samples could be acquired without regard to timing in the menstrual cycle.

Dabbs JM, Campbell BC, Gladue BA, Midgley AR, Navarro MA, Read GF, Susman EJ, Swinkels LM, Worthman CM. Reliability of salivary testosterone measurements: a multicenter evaluation. Clin Chem 1995;41:1581-1584.

Results of a multicenter evaluation (8 University laboratories, 1 commercial laboratory) of the reliability of salivary testosterone assay are presented in this paper. Each laboratory assayed aliquots of saliva specimens collected from 100 male and 100 female subjects. The intraclass correlation coefficient, which measures overall agreement on individual scores was r=0.87 for men and r=0.78 for women, Mean agreement between each laboratory and the combined set of all other laboratories was r=0.6 (both sexes). The authors concluded that the level of expertise in immunoassay of saliva was gradually increasing.

Davies RH, Harris B, Thomas DR, Cook N, Read G, Riad-Fahmy D. Salivary testosterone levels and major depressive illness in men. Br J Psych 1992;161:629-632.

Salivary testosterone levels were obtained on 11 men admitted to hospital for major depression. Mean levels in depressive were lower than controls, although the difference was not significant due to wide variances. There was significant negative correlation between salivary testosterone and various self-rated scales of anxiety and depression (Montgomery and Asberg, Leeds Anxiety and Depression Scale, Spielberger State Anxiety Inventory).

Davis EP, Donzella B, Krueger WK, Gunnar MR. The start of a new school year: individual differences in salivary cortisol response in relation to child temperament. Dev Psychobiol 1999;35:188-196.

Twice-daily (noon and evening) salivary cortisol levels were measured in 70 elementary school children during the first week of a new school year, in order to study the relationship between temperament and HPA axis functioning. Considering the group as a whole, there was no significant increase in cortisol during the first week of school. Individual changes in cortisol were examined, and there was no evidence that shy/anxious children produced more cortisol during the first week of school. In fact, consistent with studies on preschool children, more exuberant, extroverted children tended to exhibit the largest increases in cortisol. As in other studies of this type, particularly where small changes in cortisol might be seen, a non-invasive sampling method is critical for the proper interpretation of results, and avoids ethical issues of multiple venipunctures in children.

Delfs TM, Klein S, Fottrell P, Naether OG, Leidenberger FA, Zimmermann RC. 24-Hour profiles of salivary progesterone. Fertil Steril 1994;62:960-966.

Three groups of seven women were studied on days 5, 7 and 8 after urinary LH surge. Blood and saliva specimens were obtained at regular intervals for the ensuing 24 hours, and salivary progesterone levels were compared to plasma (total) levels. On average, salivary levels were ≈ 1% of plasma levels. The authors report a correlation of r=0.6 between group mean plasma total progesterone and salivary progesterone, and also report that for a given individual, saliva progesterone levels fluctuate by at least a factor of 3 over 24 hours. They conclude that salivary progesterone is not suitable for diagnosing luteal insufficiency due to the poor correlation between total plasma and salivary progesterone.

Duclos M, Corcuff JB, Arsac L, Moreau-Gaudry F, Rashedi M, Roger P, Tabarin A, Manier G. Corticotroph axis sensitivity after exercise in endurance-trained athletes. Clin Endocrin 1998;48:493-501.

ACTH or CRH-stimulated saliva and plasma (total) cortisol were measured before and after 2 hours of strenuous exercise in 8 trained athletes. Both matrices gave the same basic information, although salivary cortisol responded more dramatically. The authors commented, “ the greater ability of saliva assays to detect a cortisol increase strongly supports its use to study HPA axis physiology, especially when studies are made during the morning when plasma cortisol levels and the CBG (cortisol binding globulin) buffer capacity are high and might blunt the plasma cortisol response…the need to minimize sampling stress favours the use of saliva as opposed to venous samples.” {mospagebreak}

Filaire E, Duche P, Lac G, Robert A. Saliva cortisol, physical exercise and training: influences of swimming and handball on cortisol concentrations in women. Eur J Appl Physiol 1996;74:274-278.

This study examined the effect of physical training in general and the type of exercise in particular, on salivary cortisol levels in elite women athletes. Multiple samples were acquired for cortisol assay over a 24 hour period which included a training workout (swimming or handball). Circadian variation was seen in all groups. Salivary cortisol increased significantly in handball players after exercise, but not in swimmers. These differences were explained with reference to differential rates of water loss during exercise, the aerobic or anaerobic nature of the workout, thermal stress, gravitational factors and body position during exercise, and psychological factors (handball players must continually adapt to rapidly changing conditions, and this may be more stressful). Salivary cortisol levels therefore differed in ways which correlated to various physiologic factors known to affect cortisol secretion in general.

Filaire E, Lac G. Dehydroepiandrosterone (DHEA) rather than testosterone shows saliva androgen responses to exercise in elite female handball players. Int J Sports Med 2000;21:17-20.

Saliva testosterone and DHEA levels were measured in elite female athletes upon arising, as well as before, and after a simulated handball match, and again the following morning. These same measurements were made in a group of sedentary controls. Resting salivary testosterone and DHEA levels were significantly lower in the trained women, in keeping with findings of other researchers studying other steroid hormones. Levels did not change with exercise. Saliva testosterone and DHEA levels were tightly correlated, and the authors conclude that DHEA could serve as a surrogate for testosterone in sports medicine studies of androgen status.

Finn MM, Gosling JP, Tallon DF, Baynes S, Meehan FP, Fottrell PF. The frequency of salivary progesterone sampling and the diagnosis of luteal phase insufficiency. Gynecol Endocrinol 1992;6:127-134.

The study sought to determine the minimum number of salivary progesterone determinations which would adequately describe luteal function and properly diagnose luteal phase insufficiency. Daily saliva samples were collected from 83 women, for 3 consecutive cycles, and analyzed for progesterone. It was determined that a frequency of one sample every three days during the luteal phase (5-6 samples total) was necessary to pick up problems such as a short luteal phase (delayed rise in progesterone) or poor progesterone surge (early termination of progesterone elevation).

Gibson EL, Checkley S, Papadopoulos A, Poon L, Daley S, Wardle J. Increased salivary cortisol reliably induced by a protein-rich midday meal. Psychosom Med 1999;61:214-224.

This study looked at the effect of a protein-rich midday meal on salivary cortisol in 10 men. When the subjects fasted at midday, cortisol levels declined steadily over the 2 hour period. When the subjects consumed a protein-rich meal, cortisol levels were significantly increased by 1 hour after eating (compared to fasting). In a second experiment, the effect of protein composition was studied. Increased protein content of the midday meal significantly increased the post-prandial rise in cortisol, and significantly improved subjective wellbeing. Note that a non-stress-provoking means of assaying cortisol was important for the design and interpretation of this study.

Herbert J, Goodyer IM, Altham PM, Pearson J, Secher SM, Shiers HM. Adrenal secretion during major depression in 8- to 16-year-olds, I. Altered diurnal rhythms in salivary cortisol and dehydroepiandrosterone (DHEA) at presentation. Psychol Med 1996;26: 245-256.

Saliva cortisol, DHEA, and DHEAS levels were measured in 88 children (8-16 years old) determined to be suffering from major depression, and 25 children free of psychiatric illness. Serum was collected in parallel to saliva in the control subjects, and significant correlations between serum and saliva were seen for all 3 analytes. Higher evening cortisol levels and lower morning DHEA levels were found in the saliva of depressed children. Note that potential ethical problems related to frequent venipuncture in children were avoided by the use of saliva hormone testing.

Granger DA, Schwartz EB, Booth A, Arentz M. Salivary testosterone determination in studies of child health and development. Horm Behav 1999;35:18-27.

The characteristics of a salivary testosterone assay developed by the authors are described. Correlation of saliva values to serum is performed and r values of 0.96 and 0.78 are obtained for males and females, respectively. Morning levels correlated to pubertal status ratings for both males and females. Consistent with other studies reported by these authors, their assay values are artifactually elevated by the use of cotton or polyester swabs as well as the acidity of the saliva. The results of the assays were not affected by storage at -20•C or colder for up to 1 year. Some of the artifactual results reported are likely due to failure to extract the specimens before analysis.

Granger DA, Schwartz EB, Booth A, Curran M, Zakaria D. Assessing dehydroepiandrosterone in saliva: a simple radioimmunoassay for use in studies of children, adolescents and adults. Psychoneuroendocrinology 1999;24:567-579.

Performance of a radioimmunoassay for DHEA in saliva is reported. The authors note, “Levels in matched serum and saliva samples showed strong linear relationships for adult males and females”. Reference ranges are provided for children, adolescents and adults. Use of cotton swabs or sugared drinks to aid sample collection resulted in 50-150% overestimation of DHEA content. Failure to extract and preconcentrate specimens likely contributed to the susceptibility of this assay to outside factors.

Granger DA, Weisz JR, Kauneckis D. Neuroendocrine reactivity, internalizing behavior problems, and control-related cognitions in clinic-referred children and adolescents. J Abn Psychology 1994;103:267-276.

Salivary cortisol samples were taken from children and adolescents before and after a stressful interaction with a parent. Subjects who experienced the largest increases in cortisol had lower baseline levels, were more anxious and withdrawn, had more social problems and perceived themselves to have less control over their lives. Note that a non-stress-provoking method of cortisol assay was important for this study.

Granger DA, Weisz JR, McCracken JT, Ikeda SC, Douglas P. Reciprocal influences among adrenocortical activation, psychosocial processes, and the behavioral adjustment of clinic-referred children. Child Dev 1996;67:3250-3262.

Salivary cortisol measurements were made on a group of children before and after conflict-provoking parent-child interactions. The change in cortisol induced by conflict was stable over time and correlated to children’s perceptions of the locus of control over their successes and failures. A criticism of this study is the use of cotton swabs and drink crystals to collect and stimulate saliva flow; these sample collection techniques can perturb the results in some saliva assays, particularly those in which the samples are not extracted prior to analysis.

Harris B, Lovett L, Newcombe RG, Read GF, Walker R, Riad-Fahmy D. Maternity blues and major endocrine changes:Cardiff puerperal mood and hormone study II. BMJ 1994 Apr;308:949-953.

These authors studied the relationship of salivary progesterone and cortisol to mood in postpartum women. Salivary progesterone levels peaked just before delivery at roughly 1100 pg/ml, falling to 30 pg/ml by 10 days postpartum. Evening salivary cortisol levels peaked at delivery at around 2400 pg/ml, falling to approximately 600 pg/ml by 15 days postpartum. There were modest, statistically significant correlations between self-rated postpartum depressive symptoms, immediate antepartum progesterone levels and also the rate at which salivary progesterone increased in the 2 weeks preceding delivery. Interestingly, no correlation between postpartum blues and total plasma progesterone ( or changes thereof) was seen, demonstrating the utility of saliva as a vehicle to measure clinically significant parameters.

Heim C, Ehlert U, Hanker JP, Hellhammer DH. Abuse-related posttraumatic stress disorder and alterations of the hypothalamic-pituitary-adrenal axis in women with chronic pelvic pain. Psychosom Med 1998;60:309-318.

Women suffering from chronic pelvic pain (CPP) underwent CRF stimulation tests, with parallel measurement of salivary cortisol and serum ACTH levels. They also underwent an overnight dexamethasone suppression test with next-day measurement of salivary cortisol levels at 2 hour intervals over 16 hours. Compared to controls, CPP patients exhibited lower cortisol response to CRF in the face of normal ACTH output, and cortisol output was suppressed to a significantly greater extent by dexamethasone. Note that in this study, a non-invasive, non-stressful assay technique was mandatory for proper interpretation of the findings.

Heine RP, McGregor JA, Dullien VK. Accuracy of salivary estriol testing compared to traditional risk factor assessment in predicting preterm birth. Am J Obstet Gynecol 1999;180:S214-218.

Baseline risk of preterm labor was assessed in 601 pregnant women at 21-25 weeks gestation via conventional multifactorial risk assessment (Creasy score). Weekly salivary estriol assays were performed until delivery, and were deemed positive if the result exceeded a threshold of 2100 pg/ml. Salivary estriol screening correctly predicted preterm birth in 91% of cases versus 75% with Creasy scoring. In patients deemed high risk at baseline, saliva estriol screening was vastly superior, correctly predicting outcome in 87% of cases versus only 7% using Creasy scoring.

Johansson A, Henriksson A, Olofsson BO, Olsson T. Adrenal steroid dysregulation in dystrophia myotonica. J Int Med 1999;245:345-351.

In this study, salivary cortisol levels were measured in 15 patients with dystrophia myotonica (DyM) (Clinical features of DyM include myotonia, skeletal, cardiac and smooth muscle atrophy, cardiac rhythm disturbances, insulin resistance and hyperinsulinemia, as well as hypogonadism.) Diurnal variation of cortisol, and response to dexamethasone were measured. First morning cortisol in DyM patients was lower than controls, but decline throughout the day was less. Consequently, end of day cortisols were higher in the DyM patients. There was no difference in response to dexamethasone suppression between Dym subjects and controls. Serum hormone levels (androstenedione, DHEAS, 17-hydoxyprogesterone, testosterone) levels were lower in DyM patients.

Khan-Dawood FS, Choe JK, Dawood MY. Salivary and plasma bound and “free” testosterone in men and women. Am J Obstet Gynecol 1984;148:441-445.

Testosterone was measured in matched saliva and plasma samples in a group of men and women. Correlation coefficient for salivary on plasma testosterone was r=0.71. Diurnal variation of testosterone levels was exhibited. Hourly variation of measured values was on the order of 10-15%. On average, dialyzable testosterone in saliva was approximately 60% of that measured in plasma. The authors conclude, “The present investigation showed that salivary testosterone correlates significantly with plasma testosterone and reflects the response of the latter to some of the physiologic variables examined.”

Kiess W, Meidert A, Dressendorfer RA, Schriever K, Kessler U, Konig A, Schwarz HP, Strasburger CJ. Salivary cortisol levels throughout childhood and adolescence: relation with age, pubertal stage and weight. Pediatr Res 1995;37:502-506.

Salivary cortisol was correlated to age, pubertal status and weight in this study of 138 infants, children and adolescents. The expected circadian variation was observed in subjects older than 1 year. The lack of such variation observed in subjects less than 1 year old agreed with the findings of other researchers. Salivary cortisol concentrations also correlated with pubertal stage and weight. The authors note that, “The measurement of salivary free cortisol levels is a reliable, stress-free, and attractive way of testing adrenal function in infants and children.”

Kudielka BM, Schmidt-Reinwald AK, Hellhammer DH, Kirschbaum C. Psychological and endocrine responses to psychosocial stress and dexamethasone/corticotropin-releasing hormone in healthy postmenopausal women and young controls: the impact of age and a two-week estradiol treatment. Neuroendocrinol 1999;70:422-430.

The response of serum and salivary cortisol to a stressful experience, and to a CRH challenge after dexamethasone suppression was measured in estradiol and placebo-treated post-menopausal women, and in untreated younger controls. Estradiol treatment reduced the serum and saliva cortisol response in a 2 hour period subsequent to CRH administration, although the cortisol response was still greater than that registered for younger controls. There was no difference in cortisol response to stress however, among the groups. These results point to a role for estradiol as an important modulator of HPA axis reactivity. {mospagebreak}

Lac G, Lac N, Robert A. Steroid assays in saliva: a method to detect plasmatic contaminations. Arch Int Physiol Biochim Biophys 1993;101:257-262.

The authors measure concurrent saliva and plasma (total) steroid hormone levels (DHEAS, Testosterone, DHEA, cortisol and androstenedione) and report correlation coefficients of 0.51, 0.77, 0.73, 0.74 and 0.92 respectively. They calculate that between 2.5-10 microliters of plasma per ml of whole saliva must be present to perturb salivary hormone results by 20%. (Note: This is a substantial volume of exudate and would not be seen under normal circumstances such as avoidance of toothbrushing before sample acquisition, and absence of periodontal disease.)

Lac G, Marquet P, Chassain P, Galen F. Dexamethasone in resting and exercising men. II. Effects on adrenocortical hormones.

Matched serum and saliva specimens were obtained from men who performed a cycle ergometer workout after 4 days of dexamethasone suppression or placebo treatment. DHEA, DHEAS, cortisol and androstenedione were measured in both sample types, and testosterone, ACTH and aldosterone were measured in plasma. Dexamethasone prior to exercise suppressed the increase of all steroids except testosterone. In all cases, salivary gave the same information as plasma, and the authors conclude, ”the consistent results obtained from the 2 matrices allow us to consider salivary assays as a useful tool for steroid abuse detection.”

Lachelin GC, McGarrigle HH. A comparison of saliva, plasma unconjugated and plasma total oestriol levels throughout normal pregnancy. Brit J Obstet Gyn 1984;91:1203-1209.

Parallel plasma and salivary estriol levels were measured weekly in 25 women throughout their second and third trimesters of pregnancy. Levels increased throughout pregnancy and correlation coefficients between results in the two matrices were in excess of 0.9. The percentage of protein binding of unconjugated estriol in plasma was measured by equilibrium dialysis and found to be roughly 12%. This allowed comparison of salivary estriol to free estriol in plasma. Salivary estriol ranged from 85% (early pregnancy) to 120% (late pregnancy) of the plasma free estriol. Hence, salivary estriol levels are very similar, but not identical to free estriol levels in plasma. The authors conclude that, “salivary estriol concentrations parallel those of plasma unconjugated estriol throughout pregnancy and could probably replace them in the assessment of fetal wellbeing.”

Lechner W, Marth C, Daxenbichler G. Correlation of oestriol levels in saliva, plasma and urine of pregnant women. Acta Endocrinol 1985;109:266-268.

The authors measured conjugated and unconjugated estriol in plasma, urine and saliva in 50 women in their last trimester of pregnancy. They reported significant correlations between unconjugated salivary and total plasma estriol, and between unconjugated plasma and total urinary estriol, but only weak correlation between saliva and urinary estriol. They concluded that no definite conclusion could be drawn regarding the biological significance and clinical importance of salivary estriol.

Lipson SF, Ellison PT. Normative study of age variation in salivary progesterone profiles. J Biosoc Sci 1992;24:233-244.

Luteal phase salivary progesterone levels were measured serially in 124 women ages 18-44. Results for six groups of 20 women were reported. The expected variation of progesterone level according to menstrual cycle day was observed. Average luteal progesterone was lowest in the 18-19 year old group (≈ 60 pg/ml), increased to a maximum of ≈ 80 pg/ml in the 25-29 year old group, and returned almost to 18 year old levels in the 40-44 year old group. The authors commented, “This study contributes to an increased understanding of human reproductive ecology by demonstrating that luteal progesterone production varies with age, and by showing that the pattern of age variation in ovarian function corresponds closely with the observed pattern of age variation in fecundity.”

Lipson SF, Ellison PT. Development of protocols for the application of salivary steroid analyses to field conditions. Am J Human Biol 1989;1:249-255.

Practical parameters of interest for the performance of salivary hormone assays (testosterone, progesterone, androstenedione and cortisol) were examined in this study. Samples were centrifuged and extracted prior to analysis. Samples collected in polystyrene tubes gave the same results as samples collected in glass. Storage at room temperature for 1-6 months did not affect results. Freezing introduced a constant negative offset. Stimulants to saliva flow including lemon juice and sugared gum skewed results, as did coffee or milk, although these effects varied widely between individuals.

Lo MS, Ng ML, Azmy BS, Khalid BA. Clinical applications of salivary cortisol measurements. Sing Med J 1992;33:170.

The circadian variation of cortisol in plasma (total cortisol) and saliva was measured in 108 normal individuals. The correlation coefficient for cortisol in the two matrices was 0.81. Comparison was also made between salivary and plasma cortisol levels in normal pregnant females, hyperthyroid patients, hypertensive patients and diabetics. The marked elevations in total plasma cortisol in these instances were not observed with salivary cortisol. The authors conclude that salivary cortisol is a better measure of adrenal status than plasma.

Lu Y, Bentley GR, Gann PH, Hodges KR, Chatterton RT. Salivary estradiol and progesterone levels in conception and nonconception cycles in women: evaluation of a new assay for salivary estradiol. Fertil Steril 1999; 71:863-868.

This study attempted to replicate earlier work by Lipson and Ellison. Daily salivary estradiol and progesterone levels were measured in consecutive menstrual cycles of 11 women, until conception occurred. The authors report that mean estradiol level in the preovulatory period in conception cycles (averaged across the group and over an unspecified number of days) was not statistically different (p=0.16) from non-conception cycles. There appeared to be a clear difference between groups when the data were graphed. Paired analysis (more appropriate for small n) was not performed on the preovulatory data, although such analysis on post-ovulatory data did reveal a difference between conception and nonconception cycles. Overall, the conclusion reached by the authors appears to be inconsistent with the data they present. Correlations between serum and saliva estradiol levels were reported in 7 other women. Unfortunately, serum samples were acquired in the morning, and saliva was acquired in the evening. Not surprisingly, individual correlation coefficients varied widely.

Lu YC, Chatterton RT, Vogelsong KM, May LK. Direct radioimmunoassay of progesterone in saliva. J Immunoassay 1997;18:149-163.

This study looked at various aspects of a salivary progesterone assay. The influence of the composition of the specimen collection tubes, and the composition of the analytical standards on final results was studied. Progesterone was found to adsorb on to the walls of some types of collection tubes, and standards made up in stripped saliva gave more accurate results than standards made up in a buffer. Daily salivary progesterone levels were measured in 10 women throughout the menstrual cycle, and the expected luteal phase increase was seen. Finally, matched serum and saliva specimens were collected from 48 women in the luteal phase of their cycle, and the correlation coefficient between the two sample types was found to be r=0.75.

Martel FL, Hayward C, Lyons DM, Sanborn K, Varady S, Schatzberg AF. Salivary cortisol levels in socially phobic adolescent girls. Depression Anxiety 1999;10:25-27.

Saliva samples were collected from 27 adolescent girls diagnosed with social phobia (SP) and 21 controls, before and after a standardized anxiety-provoking situation (Trier Social Stress Test or TSST). There was no difference between SP subjects and controls in baseline cortisol levels, or in the circadian variation of these levels. No difference was seen between the groups either before or after the TSST, and this is consistent with studies of adults with SP. Cortisol levels for both groups were higher in anticipation of the TSST. The authors comment that, “salivary cortisol levels appear to be a sensitive measure of anticipatory anxiety.”

McGregor JA, Hastings C, Roberts T, Barrett J. Diurnal variation in saliva estriol level during pregnancy: a pilot study. Am J Obstet Gynecol 1999; 180:S223-225.

The circadian variation of estriol level in saliva was studied in 14 pregnant women. A marked nocturnal surge in estriol occurred, with a peak between 2 and 4 a.m. These findings agree with other human and animal research on plasma estriol levels. The study demonstrates that salivary estriol correlates to plasma estriol. The authors conclude that if salivary estriol is measured for pregnancy screening purposes, samples should be acquired during daylight hours to avoid false positives.

McGregor JA, Jackson GM, Lachelin GC, Goodwin TM, Artal R, Hastings C, Dullien V. Salivary estriol as risk assessment for preterm labor: a prospective trial. Am J Obstet Gynecol 1995;173:1337-1342.

Serial salivary estriol monitoring in pregnancy was shown to be a better predictor of pre term labor and delivery compared to conventional risk assessment based on history. When a threshold of 2300 pg/ml was used, salivary estriol correctly predicted the occurrence of pre term labor in 77% of cases versus 37% for clinical risk score. The ability of salivary estriol to predict pre term delivery was lower (50%) but still better than risk score prediction (33%). The authors conclude, “Serial collection and measurement of salivary estriol was easy to perform and relatively effective as a means to identify women at heightened risk of pre term labor and delivery.”

Metcalf MG, Evans JJ, Mackenzie JA. Indices of ovulation: comparison of plasma and salivary levels of progesterone with urinary pregnanediol. J Endocr 1984;100:75-80.

This is an early study comparing salivary progesterone with plasma progesterone and with 24 hour urinary excretion of pregnanediol in the follicular and luteal phases of the menstrual cycles of 20 women. (One sample was collected in each half of the cycle, per woman.) Saliva progesterone levels in the luteal phase were seen to increase by a factor of 3, less than the increases seen for plasma progesterone and 24 hour urinary pregnanediol. The authors question the integrity of their assay however, and note that falsely elevated progesterone results tended to occur in samples acquired in the a.m. This suggests their antibody might have been cross-reacting with other hormones including cortisol. Nevertheless, they concluded that salivary progesterone was not as reliable an indicator of the occurrence of ovulation compared to 24 hour urinary pregnanediol or plasma progesterone.

Meulenberg PM, Hofman JA. Salivary progesterone excellently reflects free and total progesterone in plasma during pregnancy. Clin Chem 1989;35:168-172.

Salivary progesterone was compared to free plasma progesterone (via equilibrium dialysis) and total serum progesterone in 36 pregnant women, sampled four times throughout pregnancy. Overall, salivary progesterone was well correlated to both free (r=0.88) and total serum progesterone (r+0.9), although salivary progesterone was on average ≈ 50% of free progesterone in serum. The authors felt that centrifugation of saliva prior to analysis may have sequestered a fraction of the progesterone in saliva. Both free serum progesterone and salivary progesterone represented ≈ 0.5-1% of total serum progesterone throughout pregnancy. In the postpartum period, binding of progesterone in serum decreased markedly (% free progesterone increased to 12% from ≈ 1%), and overall, levels of progesterone fell drastically. The authors conclude, “Despite the dramatic increase in concentrations of total progesterone and binding proteins in plasma during pregnancy, we found highly significant correlations between total and free progesterone in plasma and salivary progesterone in the group as a whole as well as individuals.”

Moran DJ, McGarrigle HH, Lachelin GC. Lack of normal increase in saliva estriol/progesterone ratio in women with labor induced at 42 weeks’ gestation. Am J Obstet Gynecol 1992;167:1563-1564.

Salivary estriol and progesterone levels were measured in 16 pregnant women from gestational day 280 to day 294. The ratio of estriol to progesterone increased steadily (due to increasing estriol) over the 14 days in 6 women who went into labour spontaneously. The ratio did not change in 10 women whose labour was induced at 294 days. These findings are in agreement with other studies regarding the change in hormone levels and the initiation of labour.

Moss HB, Vanyukov MM, Martin CS. Salivary cortisol responses and the risk for substance abuse in prepubertal boys. Biol Psych 1995;38:547-555.

Salivary cortisol in prepubertal boys anticipating a stressful task was measured. Sons of fathers who used drugs or exhibited antisocial behavior had lower levels of cortisol while anticipating the stressor, compared to controls, despite similar professed levels of anxiety in each group. Note that the results of this study would likely have been skewed significantly if serum cortisol measurements had been made: anticipation of acquisition of the blood sample would likely have equaled or superseded anticipation of the stressful task. Hence saliva sampling was a critical aspect of this study.

Nahoul K, Rao LV, Scholler R. Saliva testosterone time-course response to hCG in adult normal men, comparison with plasma levels. J Steroid Biochem 1986;24:1011-1015.

Testosterone levels were measured in blood and saliva specimens collected for 96 hours in 13 men after injection of hCG. Salivary testosterone, on average, was approximately 2% of serum testosterone. Testosterone levels in both fluids traced out the same response curves over time, and were significantly correlated (r=0.65). The authors suggest that saliva could be employed instead of blood to perform this standard assay of testicular function.

Nahoul K, Scholler R. Comparison of saliva and plasma 17-hydroxyprogesterone time-course response to hCG administration in normal men. J Steroid Biochem 1987;26:251-257.

The response of salivary and plasma 17-hydroxyprogesterone to an injection of hCG was measured in 12 men, and very similar time behavior was noted in both cases. The hormone level in saliva was approximately 2.5% of that measured in plasma.

Navarro MA, Nolla JM, Machuca MI, Gonzalez A, Mateo L, Bonnin RM, Roig-Escofet D. Salivary testosterone in postmenopausal women with rheumatoid arthritis. J Rheumatol 1998;25:1059-1062.

Salivary testosterone, as well as serum free and total testosterone and SHBG levels were measured in 44 postmenopausal women with rheumatoid arthritis (RA), and compared to levels in controls. The correlation coefficient between saliva and serum free testosterone was 0.45. There was no difference in levels between controls and RA patients not treated with glucocorticoids. In glucocorticoid-treated RA patients, salivary testosterone, total testosterone and SHBG were significantly lowered.

Navarro MA, Villabona CM, Blanco A, Gomez JM, Bonnin RM, Soler J. Salivary excretory pattern of testosterone in substitutive therapy with testosterone enanthate. Fertil Steril 1994;61:125-128.

Serial saliva and serum testosterone levels were measured in 16 men after a single intramuscular injection of testosterone enanthate. Testosterone levels in both fluids followed a similar time course, with an initial peak or plateau, then a gradual decline. Levels remained above the low end of the eugonadal range for ≈ 2 weeks. The authors conclude that, “Salivary testosterone concentration may be applied to assess the effectiveness of testosterone substitutive therapy. The simplicity of the salivary sampling technique means that it could replace serum in the monitoring of this type of therapy.” {mospagebreak}

O'Leary P, Feddema P, Chan K, Taranto M, Smith M, Evans S. Salivary, but not serum or urinary levels of progesterone are elevated after topical application of progesterone cream to pre and postmenopausal women. Clin Endocrinol 2000;53:615-620.

This study measures the response of serum, salivary and urine progesterone and progesterone metabolites in premenopausal (n=6) and postmenopausal (n=6) women after a single application of a cream containing 64 mg of micronized progesterone. Serum levels rose marginally after cream application. Urine progesterone metabolites did not change. Saliva progesterone levels increase substantially within the first 3 hours of application. There was an approximately 7-fold greater increase in salivary progesterone levels in premenopausal women compared to postmenopausal women. On this basis, the authors conclude that no useful relationship between saliva progesterone levels and transdermal progesterone dose exists. This disparity between pre and postmenopausal transdermal progesterone levels has not been seen at ZRT Laboratory. Measurements taken after several cycles of progesterone use (2-3 weeks per cycle) might have been more relevant, giving postmenopausal women time to adapt to higher progesterone exposure.

O’Rourke MT, Ellison PT. Salivary estradiol levels decrease with age in healthy, regularly-cycling women. End J 1993;1:487-494.

Daily salivary estradiol and progesterone levels were measured over one full menstrual cycle in 53 women ranging in age from 25 to 48 years. The average saliva estradiol profiles corresponded closely to published serum and plasma estradiol profiles. Average salivary follicular and luteal estradiol levels declined with age, although the mid-cycle peak level did not. These findings are consistent with the literature on fertility and aging.

Petsos P, Ratcliffe WA, Heath DF, Anderson DC. Comparison of blood spot, salivary and serum progesterone assays in the normal menstrual cycle. Clin Endocrin 1986;24:31-38.

This 1986 study compares serum, blood spot and salivary progesterone levels in 6 patients. Salivary progesterone displays the expected luteal phase increase and tracks serum progesterone well. Salivary levels correlated well (r=0.89) with serum. Despite the low sensitivity (detection limit ≈ 3000 pg/ml) and high CV (12-14%), this study demonstrates the congruence of saliva and serum progesterone measurements.

Pruessner JC, Hellhammer DH, Kirschbaum C. Burnout, perceived stress, and cortisol responses to awakening. Psychosom Med 1999;61:197-204.

The salivary cortisol response to awakening was studied in 66 teachers, to explore the relationship of stress, HPA axis function and “burnout”. On 3 consecutive days, subjects submitted saliva specimens on awakening, and 15, 30 and 60 minutes thereafter. All subjects took 0.5 mg of dexamethasone the night before acquisition of the third set of saliva specimens. These data were correlated to data gathered from questionnaires pertaining to psychological state. On days 1 and 2, saliva cortisol peaked 15-30 minutes after waking, then declined. On day 3, salivary cortisols increased throughout the hour, rather than peaking. Subjects who reported higher levels of stress and burnout had significantly lower cortisol secretion on all 3 days. Lower cortisol output correlated well with feelings of exhaustion, detachment from work, and a loss of drive, increased somatic complaints and low self esteem. Note that a non-stress-inducing method of cortisol assay was crucial to the design of this study.

Quissell D. Steroid hormone analysis in human saliva. Ann N Y Acad Sci. 1993;694:143-145.

In this very brief overview of salivary steroid hormone analysis, the author mentions several possible pitfalls to analysis including: metabolization of hormones by cells of the salivary gland or oral flora, contamination of saliva by plasma from traumatized oral tissue or gingival crevicular fluid. Nevertheless, the author concludes, “Saliva collection has provided the medical and research community with an excellent medium for the monitoring of plasma steroid levels.”

Raff H, Raff JL, Findling JW. Late-night salivary cortisol as a screening test for Cushing’s Syndrome. J Clin Endocrinol Metab 1998;83:2681-2686.

Salivary cortisol levels were studied in 73 normal individuals and 78 patients suspected of having Cushing’s Syndrome. Patients with spontaneous Cushing’s Syndrome were found to have a markedly elevated late-night salivary cortisol, and could be identified with a sensitivity of 97%. The authors conclude, “the measurement of late-night salivary cortisol is a simple, convenient, and reliable way to screen patients for Cushing’s Syndrome.”

Read GF. Status report on measurement of salivary estrogens and androgens. Ann NY Acad Sci 1993:146-160.

This paper reviews the status of salivary hormone testing. Some of the tentative conclusions have since been validated. This paper is primarily of historical interest.

Read GF, Harper ME, Peeling WB, Griffiths K. Changes in male salivary testosterone concentrations with age. Int J Androl 1981;4:623-627.

Salivary testosterone levels were measured in males ranging in age from 20 to 80 years. Mean salivary testosterone decreased gradually with age, falling by a factor of approximately 2.4 between age 25 and 75. These findings agree with the age related decline also measured in plasma free testosterone.

Read G, Fahmy D, Wilson D, Griffiths K. A new approach for breast cancer research, assays for steroids in saliva.

This paper outlines studies validating some of the basic analytical characteristics of salivary steroid hormone assays. Levels of 17-hydroxyprogesterone and cortisol in mixed saliva correlated well (r= 0.98) with fluid collected directly from the parotid duct outlet. Serial simultaneous measurements of 17-hydroxyprogesterone in plasma, saliva and parotid fluid correlated well and displayed circadian variation. Levels of 17-hydroxyprogesterone and cortisol were equal in unstimulated and stimulated saliva samples. Plasma and salivary progesterone levels correlated well over a menstrual cycle, with salivary progesterone exhibiting the expected luteal phase increase. Women with suspected luteal phase deficiency were seen to have low or erratic levels of salivary progesterone.

Read GF, Walker RF, Wilson DW, Griffiths K. Steroid analysis in saliva for the assessment of endocrine function. Ann NY Acad Sci 1993: 260-274.

A number of facets of salivary steroid hormone analysis are discussed in this paper. Salivary hormone levels (progesterone, 17-hydroxyprogesterone, cortisol and estriol) were shown to be independent of rate of saliva production. Saliva responded rapidly to change in serum levels; following IV administration of cortisol to dexamethasone-suppressed subjects, salivary cortisol levels peaked within 3 minutes of injection (serum levels peaked within 1 minute). The utility of salivary 17-hydroxyprogesterone levels for monitoring children with Congenital Adrenal Hyperplasia was discussed, as was the use of salivary progesterone for studying induction of ovulation in infertility. An excellent correlation between free plasma estriol and salivary estriol (r=0.95) was presented, and the role of salivary estriol for pregnancy monitoring discussed. Data on the correlation of salivary testosterone and progesterone to stage of sexual maturity were also presented.

Read GF, Wilson DW, Campbell FC, Holliday HW, Blamey RW, Griffiths K. Salivary cortisol and dehydroepiandrosterone sulphate levels in postmenopausal women with primary breast cancer. Eur J Cancer Clin Oncol 1983;19:477-483.

Serial salivary DHEAS and cortisol measurements were taken for 2 consecutive days in women with breast cancer, and age-matched controls. Circadian variation, with an a.m. peak was seen in both groups of women, for both hormones. Women with cancer appeared to have marginally elevated cortisol and marginally lower DHEAS levels compared to controls, but due to within-group variance, the differences were not significant. The authors comment on the advantage of using saliva specimens when large numbers of samples have to be acquired.

Riad-Fahmy D, Read GF, Walker RF. Salivary steroid assays for assessing variation in endocrine activity. J Steroid Biochem 1983;19:265-272.

The correlation between plasma and saliva hormone levels in a variety of situations are presented in this paper. Plasma and salivary cortisol levels after an injection of ACTH in a normal subject, measured at 15 minute intervals were well-correlated. Cortisol measurements in saliva and plasma were also acquired every 15 minutes over a 12 hour period in 3 individuals, and again the values were very well correlated. Other data validating salivary hormone assay are also presented, including serial measurements of salivary testosterone levels illustrating the expected circadian variation over 12 hours, and serial salivary progesterone and estradiol measurements over the course of menstrual cycles, once again demonstrating expected variation. The authors note, “Studies like these illustrate the usefulness of salivary sampling regimens for monitoring short term changes in secretory activity. Since these samples unlike plasma, are easily collected by stress-free, non-invasive techniques, volunteers are more readily recruited, and undue disturbance of endocrine systems is avoided.”

Rosmond R, Bjorntorp P. The hypothalamic-pituitary-adrenal axis activity as a predictor or cardiovascular disease, type 2 diabetes and stroke. J Int Med 2000;247:188-197.

Circadian variation in salivary cortisol, and salivary cortisol response to dexamethasone were measured in a group of men, along with blood lipids, insulin, blood pressure, body mass index, and waist-hip ratio. Preliminary work with the cortisol assay indicated excellent correlation (r=0.92) between plasma and saliva cortisol levels. The study demonstrated that HPA axis dysfunction (typified by low a.m. cortisol, low circadian variability, blunted suppression of cortisol by dexamethasone, and poor postprandial cortisol rise) was associated with a tight clustering of other established risk factors for diabetes, coronary artery disease and stroke.

Ruutiainen K, Sannikka E, Santti R, Erkkola R, Adlercreutz H. Salivary testosterone in hirsutism:correlations with serum testosterone and the degree of hair growth. J Clin Endocrinol Metab 1987;64:1015-1020.

Testosterone was measured in saliva and serum samples in 53 females attending a hirsutism clinic. Testosterone in saliva was significantly correlated ( r=0.41) to facial hirsutism whereas serum free testosterone was not correlated. The authors conclude,” On the basis of the results, salivary testosterone seems to relate to the bioavailable fraction of the hormone and thus appears to be an optimal method for studying hirsute females”. {mospagebreak}

Sannikka E, Terho P, Suominen J, Santti R. Testosterone concentrations in human seminal plasma and saliva and its correlation with non-protein-bound and total testosterone levels in serum. Int J Andrology 1983;6:319-330.

This is an early paper on salivary hormone analysis. Salivary testosterone in males was shown to be independent of saliva flowrate (correlation between testosterone concentration in stimulated and unstimulated saliva samples was 0.96.), and whole saliva values correlated well (r=0.86) with saliva collected directly from the parotid duct. The correlation coefficient for free serum testosterone on salivary testosterone was 0.75. The salivary testosterone level correlated well with Tanner stage in adolescent boys. The authors conclude, “Determination of testosterone in saliva could thus provide a convenient and accurate index of the non-protein-bound concentration of testosterone in serum i.e. the availability of hormone to tissues.”

Scheer FA, Buijs RM. Light affects morning salivary cortisol in humans. J Clin Endocrinol 1999;84:3395-3398.

This study examined the effect of light on early morning salivary cortisol levels. In 14 males, levels were measured every 20 minutes for 1 hour after awakening, under 2 sets of conditions: 0 lux (for which subjects wore a dark cap over their eyes) and 800 lux (for which subjects wore an illuminated visor). Cortisol levels were significantly higher in the light exposed subjects, at least for the first 40 minutes after waking. Sixteen hours after waking,, the presence or absence of light did not influence salivary cortisol. These findings agree with the results of animal experiments. The authors chose saliva as a measurement vehicle because it was deemed to be, “a convenient and reliable measure for the unbound, and therefore active, plasma cortisol levels.”

Schramm W, Smith RH, Craig PA, Grates HE. Testosterone concentration is increased in whole saliva but not in ultrafiltrate after toothbrushing. Clin Chem 1993;39:519-521.

Testosterone levels were measured in whole saliva and an ultrafiltrate collected via an osmotically-pumped collection device before and after toothbrushing, in a group of males. On average, the concentration of testosterone in whole saliva was increased 14% by toothbrushing, compared to 5% for the ultrafiltrate, and these differences were statistically significant despite wide individual variation. It is evident that toothbrushing releases additional protein-bound testosterone into saliva in some individuals; saliva samples should be collected before toothbrushing to ensure highest possible accuracy.

Schramm W, Smith RH, Craig PA, Paek SH, Kuo HH. Determination of free progesterojne in an ultrafiltrate of saliva collected in situ. Clin Chem 1990;36:1488-1493.

This paper details the use of a saliva collection device consisting of a semipermeable membrane enclosing an osmotically-active material (sucrose). An ultrafiltrate of saliva is drawn into the collector over the course of 15 minutes. Progesterone levels in whole saliva correlated well with that in the ultrafiltrate (r=0.95), although progesterone in ultrafiltrate was ≈84% of that in whole saliva. The progesterone level in the ultrafiltrate also correlated well (r=0.86) to serum progesterone in 2 women over the course of two menstrual cycles. The authors felt that analysis of ultrafiltrate obviated the need for centrifugation and extraction of specimens prior to analysis.

Schurmeyer T, Nieschlag E. Effect of ketoconazole and other imidazole fungicides on testosterone biosynthesis. Acta Endocrinol 1984;105:275-280.

Serial serum and saliva testosterone levels in 5 men were measured for 24 hours after ingestion of ketoconazole. Testosterone levels in serum and saliva began to decline significantly within 1 hour, reached a minimum between 4 and 6 hours later, and had returned to pre-ingestion levels at 24 hours. The effect was determined to be exerted at the level of testosterone synthesis by Leydig cells, rather than through an effect on gonadotrophins. The study indicates that saliva testosterone was reflective of changes in serum testosterone in this unusual setting.

Schwartz EB, Granger DA, Susman EJ, Gunnar MR, Laird B. Assessing salivary cortisol in studies of child development. Child Dev 1998;69:1503-1513.

This study assessed the effect of a salivary flow stimulant (powdered drink crystals) on salivary cortisol levels. Cortisol levels in unstimulated saliva samples spiked with drink crystals after collection (to simulate concentrations achieved in saliva with field use of the stimulant) resulted in stimulant concentration-dependent over-estimation of cortisol levels. Saliva specimens obtained under field conditions (drink crystals and cotton swab) returned cortisol levels which were significantly higher than expected. This study illustrates the pitfalls of salivary hormone assays which do not employ extraction methods.

Shirtcliff EA, Granger DA, Schwartz EB, Curran MJ, Booth A, Overman WH. Assessing estradiol in biobehavioral studies using saliva and blood spots: simple radioimmunoassay protocols, reliability, and comparative validity. Horm Behav 2000;38:137-147.

The authors describe a method for RIA of estradiol in saliva and blood spots, and compare both to serum. Salivary values correlated to serum values with r=0.68 for females and for males there was no correlation due to poor detection limits. Average salivary levels for males and females were 0.5 pg/ml and 1.0 pg/ml respectively (significantly lower than the normal ranges at ZRT Laboratory). Use of cotton swabs or chewing gum resulted in falsely elevated readings. Serial assays throughout the menstrual cycle revealed the expected fluctuations in estradiol levels. The low sensitivity and susceptibility to error of the asay is due to the small sample volume (300 microliters) and failure to extract or preconcentrate the specimens. In general however, the results validate the use of saliva for estradiol measurement.

Shirtcliff EA, Granger DA, Schwartz E, Curran MJ. Use of salivary biomarkers in biobehavioral research: cotton-based sample collection methods can interfere with salivary immunoassay results. Psychoneuroendocrinol 2001;26:165-173.

This paper outlines the performance of a commercially available salivary hormone assay (Salimetrics). Results for DHEA were seen to be artificially elevated when samples were collected via cotton swabs or plugs. Saliva DHEA levels correlated well to serum (r=0.73) when no cotton was used, but the correlation was lost when cotton was used, due to the spurious elevation. This study demonstrates one of the potential problems introduced when saliva is analyzed directly, without extraction.

Steptoe A, Cropley M, Griffith J, Kirschbaum C. Job strain and anger expression predict early morning elevation in salivary cortisol. Psychosom Med 2000;62:286-292.

Salivary cortisol levels at intervals throughout the day were measured in teachers and correlated to perception of job stress level and patterns of anger expression. Individuals experiencing high levels of job strain and exhibiting high levels of expressed anger had almost 22% higher salivary cortisol in the early part of the working day. Saliva sampling facilitated the acquisition of multiple samples with little or no disruption of work routines and minimal additional stress.

Stones A, Groome D, Perry D, Hucklebridge F, Evans P. The effect of stress on salivary cortisol in panic disorder patients. J Affect Disorders 1999;52:197-201.

Salivary cortisol measurements were used to assess the response to stress in patients suffering with panic disorders, and controls. Compared to control subjects, salivary cortisol levels before a stressful situation were lower, and did not change, indicating a difference in HPA axis regulation in these patients. (In controls, levels were higher before stress, and dropped afterward.). The authors note that a non-stressful sampling technique was mandatory in this work, to avoid perturbation of the findings.

Sumiala S, Tuominen J, Huhtaniemi I, Maenpaa J. Salivary progesterone concentrations after tubal sterilization. Obstet Gynecol 1996;88:792-796.

This study looked at the effect of laparoscopic tubal ligation on salivary progesterone levels in 55 women. Daily salivary progesterone levels were measured in 3 cycles: the cycle immediately before surgery, and cycles 3 and 12 months after the procedure. Cumulative mean progesterone and peak luteal progesterone were both decreased at 3 and 12 months, indicating that tubal ligation, even by minimally invasive means, has a significant effect on ovarian function. A longer follow-up period would be necessary to determine if this effect is reversible.

Swinkels LM, Ross HA, Smals AG, Benraad TJ. Concentrations of total and free dehydroepiandrosterone in plasma and dehydroepiandrosterone in saliva of normal and hirsute women under basal conditions and during administration of dexamethasone/synthetic corticotropin. Clin Chem 1990;16:2042-2046.

Salivary and plasma DHEA levels (free and total plasma DHEA) were measured in 22 normally cycling women. Correlation coefficients for salivary vs total and free DHEA in plasma averaged r=0.9 in both luteal and follicular phases. There was near-quantitative agreement between mean plasma free DHEA and salivary DHEA. The authors also made these same comparisons in dexamethasone-suppressed women with hirsutism given CRH, and in general, found the same good correlations. They concluded, “The concentration of DHEA in saliva is a reliable index of adrenal DHEA secretion, allowing a stress-free multiple sampling regimen for the study of time-related hormonal events.”

Thijssen JH, Gispen-de Wied CC, van Heeswijk GM, Veeman W. Determination of dexamethasone in saliva. Clin Chem 1996;42:1238-1242.

Assay of dexamethasone in saliva after oral ingestion was reported in this paper. Serum and saliva dexamethasone concentrations correlated extremely well (r=0.99). Moreover, free dexamethasone is known to be ≈ 18% of total serum dexamethasone, and salivary dexamethasone was ≈ 17% of total serum dexamethasone. Therefore salivary dexamethasone appears to quantitatively represent unbound dexamethasone in plasma.

Tschop M, Behre HM, Nieschlag E, Dressendorfer RA, Strasburger CJ. A time-resolved fluorescence immunoassay for the measurement of testosterone in saliva: monitoring of testosterone replacement therapy with testosterone buciclate. Clin Chem Lab Med 1998;36:223-230.

In this paper, the treatment of male hypogonadism via testosterone replacement was monitored via salivary measurements. Parallel serum and saliva samples were taken at increasing intervals after a single injection of testosterone buciclate, for a total of 16 weeks. The concentration versus time profiles for serum androgens (total testosterone + total DHT) and saliva testosterone were very similar. A weak, but significant correlation between total serum androgens and salivary testosterone (r=0.22) was seen, but no details on the timing of acquisition of saliva and blood samples were provided. Samples acquired at different times of the day would not be expected to correlate for a hormone such as testosterone for which significant circadian variation is seen. The authors conclude, “The time resolved fluorescence immunoassay for salivary testosterone provides a useful tool for monitoring androgen status in men and women, and is well suited for the follow up of testosterone replacement therapy on an outpatient basis.”

Tulppala M, Bjorses UM, Stenman UH, Wahlstrom T, Ylikorkala O. Luteal phase defect in habitual abortion: progesterone in saliva. Fertil Steril 1991;56:41-44.

Daily salivary progesterone levels were measured in consecutive menstrual cycles in women with a history of multiple miscarriages, and controls. Serum progesterone levels were also measured along with endometrial biopsies and LH measurements to establish accurate timing within the cycles. The saliva progesterone levels exhibited an ovulatory pattern in all cases, but there was no difference in salivary progesterone between the study groups, and the same was true for serum progesterone. Salivary progesterone measurements are not helpful in the study of women with frequent miscarriages.

Tunn S, Mollmann H, Barth J, Derendorf H, Krieg M. Simultaneous measurement of cortisol in serum and saliva after different forms of cortisol administration. Clin Chem 1992;38:1491-1494.

Cortisone was administered to 8 dexamethasone-suppressed males by intravenous, oral and rectal routes, and the resulting serum and saliva cortisol levels were measured over the ensuing 8 hours. (In serum, free and bound cortisol levels were measured.) These levels followed a nearly identical time course over the whole 8 hours for both IV and oral administration. A saturation effect was seen in which saliva levels increased rapidly once a threshold serum concentration was exceeded. This supports the contention that salivary cortisol reflects the non-protein-bound cortisol in serum. (Once all available cortisol binding sites on cortisol binding globulin and albumin are occupied, the concentration of free cortisol in serum will increase rapidly.) Also, the close agreement in time between behaviour between serum and saliva indicates that delivery of hormone to saliva is blood-based, rather than lymphatic-based. {mospagebreak}

Vedhara K, Hyde J, Gilchrist ID, Tytherleigh M, Plummer S. Acute stress, memory, attention and cortisol. Psychoneuroendocrinol 2000;25:535-549.

Salivary cortisol levels were measured in university students in two 48 hour periods before, and during examinations. Paradoxically, cortisol levels were lower during the presumably more stressful exam period. The authors commented that confounding effects of venipuncture stress were avoided by the use of saliva as a measurement vehicle.

Vining RF, McGinley RA, Symons RG. Hormones in saliva: mode of entry and consequent implications for clinical interpretation. Clin Chem 1983;29:1752-1756.

This is a landmark paper for salivary hormone analysis in which measurements of non-protein bound, unconjugated estriol and DHEA in serum are compared to “salivary estriol” and “salivary DHEA”. Direct numerical agreement between the two matrices indicates that salivary hormone levels represent the non-protein bound, unconjugated, or bioavailable hormone. Consideration of the anatomy of the salivary gland, along with solubility considerations, and comparison of the concentration of conjugated hormone in saliva and serum allow the following conclusions to be drawn: Unconjugated steroids make their way into saliva by partitioning between the membranes of the acinar cells and saliva. Unconjugated steroids are largely excluded from saliva since they primarily gain entry by passage through acinar cell tight junctions. Because the concentration of conjugated estriol in serum is roughly ten times that of unconjugated estriol however, the amounts of conjugated and unconjugated estriol in saliva are roughly equal despite their different mechanisms of transport into saliva.

Vining RF, McGinley RA. The measurement of hormones in saliva: possibilities and pitfalls. J Steroid Biochem. 1987;27(1-3):81-94.

This paper is an excellent review of the fundamental principles of salivary hormone analysis (saliva formation, transport of hormones into saliva, effect of saliva flow rate) and well summarizes the body of knowledge circa 1986 regarding the clinical applicability of saliva hormone assays. Various tentative conclusions (such as the utility of salivary testosterone and estradiol measurements) have been confirmed. The authors also enumerate a number of practical and scientific advantages of saliva analysis: minimal stress on the HPA axis, good representation of free plasma hormone levels, avoidance of ethical problems in studies of children, convenience and low cost, ease of acquisition of multiple samples.

Vittek J, L’Hommedieu D, Gordon G, Rappaport S, Southren L. Direct radioimmunoassay (RIA) of salivary testosterone: correlation with free and total serum testosterone. Life Sciences 1985;37:711-716.

This is an earlier paper on salivary hormone testing which compares salivary testosterone levels with free and total serum testosterone in men and women. Testosterone concentration in submandibular and mixed (submandibular and parotid) saliva were quantitatively equal (r=0.99). Free serum testosterone and salivary testosterone were also quantitatively equal (r=0.97). Salivary testosterone and total testosterone were well correlated (r=0.8) and salivary testosterone was ≈ 2% of total serum testosterone. The authors conclude that, “determination of salivary testosterone is a reliable method to detect changes in the concentration of available biologically active hormone.”

Voss HF. Saliva as a fluid for measurement of estradiol levels. Am J Obstet Gynecol 1999;180:S226-231.

This paper is a good general overview of the underlying physiology of saliva as it relates to steroid hormone analysis. It describes the characteristics of a commercially available salivary estriol assay and discusses the utility of salivary estriol for pregnancy monitoring.

Vuorento T, Lahti A, Hovatta O, Huhtaniemi I. Daily measurements of salivary progesterone reveal a high rate of anovulation in healthy students. Scand J Clin Lab Invest 1989;49:395-401.

Daily salivary progesterone levels were measured on 17 healthy, regularly menstruating university students, and compared to profiles obtained from 15 other women for whom the occurrence of ovulation was verified by ultrasound, LH surge and endometrial biopsy. On two occasions (one in the follicular phase and one in the luteal phase) parallel serum progesterone measurements were made. Serum and saliva measurements correlated well (r=0.93) with saliva progesterone roughly 1% of serum progesterone. Salivary progesterone profiles demonstrated the expected luteal phase peak. Interestingly, 47% (6/15) of the students did not ovulate as judged by the absence of a luteal phase progesterone rise. The authors conclude, “Normal (indicating ovulation) and pathological (luteal dysfunction or no ovulation) progesterone profiles can be reliably discriminated. Sample collection is easy and makes ling-term monitoring of corpus luteum function a feasible means of diagnosis and follow up of treatment in infertility patients.”

Walker RF, Read GF, Wilson DW, Riad-Fahmy D, Griffiths K. Chronobiology in laboratory medicine: principles and clinical applications illustrated from measurements of neutral steroids in saliva. Chronobiology: its role in clinical medicine, general biology, and agriculture, Part A, pages 105-117.

The authors present salivary cortisol and testosterone data from subjects ranging in age from newborn to adolescence. These data illustrate the expected circadian variation demonstrated in serum and plasma studies, as well as the variation of salivary testosterone with stage of genital maturity, and confirm that salivary hormone measurements correlate well to physiology.

Wang DY, Fantl VE, Habibollahi F, Clark GM, Fentiman IS, Hayward JL, Bulbrook RD. Salivary oestradiol and progesterone levels in premenopausal women with breast cancer. Eur J Cancer Clin Oncol. 1986 Apr;22(4):427-33.

Daily saliva specimens were submitted for estradiol and progesterone analysis by 24 women for one complete menstrual cycle. Twelve of the women had a past history of breast cancer, but were disease free at the time of study, and twelve were controls. Parallel saliva and serum specimens were also obtained on 19 women with no history of cancer. In these 19 women, salivary estradiol levels correlated well to free serum estradiol levels (r=0.78). Serum free estradiol was determined by centrifugal ultrafiltration dialysis. Salivary levels were on average, ≈ 30% of free plasma values. Salivary estradiol also correlated well (r=0.79) to total serum estradiol. When profiles of estradiol and progesterone between breast cancer survivors and unaffected controls were compared, no differences were seen.

Wang DY, Knyba RE. Salivary progesterone: relation to total and non-protein-bound blood levels. J Steroid Biochem 1985;23:975-979.

Daily saliva samples were taken from 9 women throughout an entire menstrual cycle, and the expected luteal phase increase was observed. Paired random blood and saliva samples were also analyzed and a correlation coefficient of r=0.78 was derived between salivary progesterone and total serum progesterone. The authors also measured the percent binding of progesterone in serum to be roughly 3%. From the total serum progesterone values, they computed free plasma progesterone values and note that these computed values are approximately 5 times greater than the salivary levels. They conclude that salivary progesterone does not represent unbound progesterone in blood. However, their saliva levels are significantly lower than other reported values, including those reported by contemporary assay labs including ZRT.

Wellen JJ, Smals AG, Rijken JC, Kloppenborg PW, Benraad TJ. Testosterone and androstenedione in the saliva of patients with Klinefelter’s Syndrome. Clin Endocrinol 1983;18:51-59.

This study compared saliva and plasma (total) levels of testosterone and androstenedione in controls and patients with Klinefelter’s Syndrome. Both saliva and plasma assays yielded the same result: testosterone levels, but not androstenedione levels were lower in Klinefelter’s patients. Correlation coefficients between saliva and plasma for testosterone and androstenedione were 0.58 and 0.7 respectively. The authors conclude, “The data illustrate that measurement of salivary steroids by reflecting the free hormone fraction in blood may be useful in evaluating endocrine function in both health and disease.”

Wong YF, Mao K, Panesar NS, Loong EP, Chang AM, Mi ZJ. Salivary estradiol and progesterone during the normal ovulatory menstrual cycle in Chines women. Eur J Obstet Gynecol Reprod Biol 1990;34:129-135.

Daily consecutive estradiol and progesterone measurements were made throughout one cycle in 10 women. The profiles generated agree with the findings of other researchers. Parallel serum samples were also acquired in 5 of the 10 women, and a correlation coefficients of r=0.93 was obtained for both estradiol and progesterone. The authors conclude that, “salivary steroids may serve as an assessment of follicular dynamics especially in relation to ovulation induction, artificial insemination or in vitro fertilization.”

Worthman CM, Stallings JF, Hofman LF. Sensitive salivary estradiol assay for monitoring ovarian function. Clin Chem 1990;36:1769-1773.

A method for salivary estradiol assay is discussed in this paper. Samples were obtained from 15 normally-cycling women at various times in their cycles, and the expected pre-ovulatory peak was readily detected. Matched serum and saliva specimens were assayed in another group of women undergoing ovulation induction, and a correlation coefficient r=0.76 was generated. The authors conclude that, “monitoring of estradiol concentrations in saliva may provide a useful diagnostic tool that allows better resolution of some clinical problems than do values for serum alone.”

Wren B, McFarland K, Edwards L, O’Shea P, Sufi S et al. Effect of sequential transdermal progesterone cream on endometrium, bleeding pattern, and plasma progesterone and salivary progesterone levels in postmenopausal women. Cliamcteric 2000;3:155-160.

Twenty-eight postmenopausal women received cyclic transdermal progesterone and continuous transdermal estradiol for 3 consecutive 28 day cycles. Saliva progesterone was measured serially through days 71-84 and compared to serum. Serial endometrial biopsies were taken. There were several aberrant findings in this study. Firstly, the authors observed that the saliva progesterone results fluctuated by up to 2 orders of magnitude within 24 hours. They also detected less than 1 pg/ml of progesterone in serum before progesterone cream usage, and very little change after. Their baseline serum progesterone levels are far below accepted postmenopausal levels for commercial progesterone assays (e.g. DPC Immulite). No details on verification of compliance with cream usage were given, or the timing of saliva sampling relative to cream application. Both of these factors may exert a significant effect on variability of measured levels. The authors conclude that salivary progesterone levels were too variable to offer any assistance in the management of women receiving HRT, and that transdermal progesterone failed to induce a discernable response in the endometrium.

ANDROSTENEDIONE

Lac G, Lac N, Robert A. Steroid assays in saliva: a method to detect plasmatic contaminations. Arch Int Physiol Biochim Biophys 1993;101:257-262.

The authors measure concurrent saliva and plasma (total) steroid hormone levels (DHEAS, Testosterone, DHEA, cortisol and androstenedione) and report correlation coefficients of 0.51, 0.77, 0.73, 0.74 and 0.92 respectively. They calculate that between 2.5-10 microliters of plasma per ml of whole saliva must be present to perturb salivary hormone results by 20%. (Note: This is a substantial volume of exudate and would not be seen under normal circumstances such as avoidance of toothbrushing before sample acquisition, and absence of periodontal disease.)

Lac G, Marquet P, Chassain P, Galen F. Dexamethasone in resting and exercising men. II. Effects on adrenocortical hormones.

Matched serum and saliva specimens were obtained from men who performed a cycle ergometer workout after 4 days of dexamethasone suppression or placebo treatment. DHEA, DHEAS, cortisol and androstenedione were measured in both sample types, and testosterone, ACTH and aldosterone were measured in plasma. Dexamethasone prior to exercise suppressed the increase of all steroids except testosterone. In all cases, salivary gave the same information as plasma, and the authors conclude, ”the consistent results obtained from the 2 matrices allow us to consider salivary assays as a useful tool for steroid abuse detection.”

Lipson SF, Ellison PT. Development of protocols for the application of salivary steroid analyses to field conditions. Am J Human Biol 1989;1:249-255.

Practical parameters of interest for the performance of salivary hormone assays (testosterone, progesterone, androstenedione and cortisol) were examined in this study. Samples were centrifuged and extracted prior to analysis. Samples collected in polystyrene tubes gave the same results as samples collected in glass. Storage at room temperature for 1-6 months did not affect results. Freezing introduced a constant negative offset. Stimulants to saliva flow including lemon juice and sugared gum skewed results, as did coffee or milk, although these effects varied widely between individuals.

Wellen JJ, Smals AG, Rijken JC, Kloppenborg PW, Benraad TJ. Testosterone and androstenedione in the saliva of patients with Klinefelter’s Syndrome. Clin Endocrinol 1983;18:51-59.

This study compared saliva and plasma (total) levels of testosterone and androstenedione in controls and patients with Klinefelter’s Syndrome. Both saliva and plasma assays yielded the same result: testosterone levels, but not androstenedione levels were lower in Klinefelter’s patients. Correlation coefficients between saliva and plasma for testosterone and androstenedione were 0.58 and 0.7 respectively. The authors conclude, “The data illustrate that measurement of salivary steroids by reflecting the free hormone fraction in blood may be useful in evaluating endocrine function in both health and disease.”

Young MC, Walker RF, Riad-Fahmy D, Hughes IA. Androstenedione rhythms in saliva in congenital adrenal hyperplasia. Arch Dis Child 1988;63:624-628.

Salivary androstenedione levels were measured in 17 patients with Congenital Adrenal Hyperplasia, along with plasma levels of other steroids. The authors demonstrate that in plasma, androstenedione, testosterone and 17-hydroxyprogesterone are all well correlated, and that salivary androstenedione correlates well to plasma testosterone (and by inference, the other plasma markers). They generated saliva androstenedione nomograms which could be used to guide cortisol dosing in pubertal patients.