Salivary Cortisol Results Obtainable within Minutes of Sample Collection Correspond with Traditional...

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/authorsrights

http://www.elsevier.com/authorsrights

Author's personal copy

Clinical Therapeutics/Volume 37, Number 3, 2015

Original Research

Salivary Cortisol Results Obtainable Within Minutes of SampleCollection Correspond With Traditional Immunoassays

Elizabeth A. Shirtcliff, PhD1; Robert L. Buck, PhD2; Mary J. Laughlin, PhD2;Thomas Hart, MS3; Craig R. Cole, PhD2; and Paul D. Slowey, PhD2

1Department of Human Development and Family Studies, Iowa State University, Ames, Iowa;2Oasis Diagnostics, Vancouver, Washington; and 3Middleton Research, Middleton, Wisconsin

ABSTRACT

Purpose: Cortisol is frequently assayed as a stress-responsive biomarker which changes over the course ofminutes to meet the demands of a person’s socialcontext. Salivary cortisol is often used as a noninvasivesampling method that possesses important health im-plications. A critical barrier to psychobiological re-search that involves salivary cortisol is a time delay ofdays to months before cortisol results are obtained viaimmunoassay, long after the person is no longerproximate to the social context in which they providedthe sample. The present study was designed to addressthis critical barrier through creation of a lateral flowtest (LFT) cortisol device capable of measuring salivarycortisol within minutes of sample collection. The LFT isfrequently used within commercial point-of-care set-tings to obtain rapid answers to the presence/absence ofa biomarker. The present study extends the LFT intothe research domain by presenting performance char-acteristics of a quantitative LFT that measures salivarycortisol within 20 minutes of sample collection.

Methods: Saliva samples from 29 adults (15 men)were obtained in the morning and afternoon by usingPassive Drool and then the Super � SAL Extra CollectionDevice (hereafter Super � SAL) and later assayed withLFT and a commercially available enzyme immunoassay.

Findings: Results indicate the LFT correlated wellwith these collection methods (R ¼ 0.872 withSuper � SAL, R ¼ 0.739 with Passive Drool, P o0.0001) and at comparable levels to correspondence of

Super � SAL with Passive Drool (R ¼ 0.798, P o0.0001) which were measured with the same assay.

Implications: These results open an exciting newpossibility to integrate this technologic advance intostress research, including knowing and potentiallychanging the person’s social context in a time-sensitive manner. Methodological improvements suchas this have the possibility of refining conceptualmodels of stress reactivity and regulation. (Clin Ther.2015;37:505–514) & 2015 Elsevier HS Journals, Inc.All rights reserved.

Key words: collection methods, cortisol, immuno-assay, lateral flow technology, stress regulation.

INTRODUCTIONStress is a leading cause of morbidity and mortality in theUnited States.1 A small set of biomarkers provideinformation about chronic and acute stressors.2 Cortisolis putatively the most frequently investigated stressbiomarker3 because cortisol is linked with many physiol-ogic processes such as neural development and cell death,4

immune function,5 learning and memory,6 sleep,7 meta-bolism and fat distribution,8 growth and development,9

reproduction,10 and aging.11 The powerful role of cortisolin shaping health outcomes is illustrated by its clinicalvalue in treatments that range from mild rashes in over-the-counter creams to life-saving efforts.12

Cortisol is the end product of the hypothalamic-pituitary-adrenal (HPA) axis. After appraisal of the

Scan the QR Code with your phone to obtainFREE ACCESS to the articles featured in theClinical Therapeutics topical updates or textGS2C65 to 64842. To scan QR Codes yourphone must have a QR Code reader installed.

Accepted for publication February 12, 2015.http://dx.doi.org/10.1016/j.clinthera.2015.02.0140149-2918/$ - see front matter

& 2015 Elsevier HS Journals, Inc. All rights reserved.

March 2015 505


social context in limbic and paralimbic neural circui-tries,13,14 the hypothalamus releases corticotropicreleasing hormone that initiates a hormonal cascadethat culminates with adrenocorticotropic hormonethat stimulates the release of cortisol from the adrenalcortex and into the blood.15 After �15 minutes fromthe onset of a stressor, cortisol concentrations peak.16–18

As a lipid-soluble hormone, cortisol easily crossesthrough cellular membranes which allows it to traveldirectly to cell nuclei to change gene expression,19,20

especially in the brain where it is responsible forterminating the stress response4 via negativefeedback.21,22 Cortisol also acts throughout the bodywhere it influences physiology over seconds, minutes,hours, and days.23 Salivary measurement of smallsteroids such as cortisol take advantage of the factthat free cortisol is lipid soluble; this biologicallyactive fraction of total cortisol passes through theacinar cells to enter saliva via passive diffusion inproportion to cortisol’s entrance into cell nuclei.24

Measuring cortisol in saliva has opened a window ofopportunity to conduct stress-related research that in-volves many repeated measures25 or applications withvulnerable populations26,27 or in unique settings.28,29

Salivary measurement has even indicated unique diagnos-tic and treatment information about cortisol-related dis-eases such as Addison syndrome or Cushing diseasewhich previously required much more invasive measure-ments throughout treatment.30–33 First-generation sali-vary cortisol assays relied on radiolabeled cortisolantibodies in radioimmunoassays that were highly sensi-tive and specific. This sensitivity was enhanced with thenext generation of assays that quantified cortisol concen-trations via optical density measurements in commerciallyavailable enzyme (EIAs) or luminescence immunoassays(LIAs) that required much lower sample volumes to beeffective. These particular methods rely on the degree ofcolor change of bound horseradish peroxidase to estimatehormone concentrations relative to a standard curve. Therange of sensitivities with the use of these assays provideda limit of detection in the range of picogram per milliliter,a necessary prerequisite for salivary steroid hormonedetection whereby concentrations are low.

A critical barrier in the stress field is that currentimmunoassay methods do not provide reportableresults for cortisol for days to months after collection.Recent studies are beginning to explore the feasibilityof point-of-care measurement of salivary cortisol,34–37

typically using lateral flow test (LFT) assays to deliver

qualitative or quantitive test results within minutes ofsample collection.38 The present study describes thefirst in a new generation of assays for measuringsalivary hormones such as cortisol by using an LFTdevice (Oasis Diagnostic Corporation, Vancouver, WA).The first US patent that used the term lateral flow wasfiled in 1988 and was awarded in 1990,39 althoughseveral companies had patented elements of thatconcept earlier. The first successful commercial useof an LFT was the EPT urine human chorionicgonadotropin hormone dipstick pregnancy test. LFThas been widely used in industry since the late 1980sand early 1990s, primarily to target the commercialneeds for point-of-care assays.40 In similar fashion toEIA and LIA, LFT relies on the detection of an emittedsignal (in this case fluorescence) to provide a fullyquantitative cortisol readout, but, unlike EIA and LIAor even technologies such as mass spectrometry ormicroarrays, LFTs are not widely used in the academiccommunity. The reasons for this may be related toeducation, intellectual property considerations, and alack of LFT suppliers that cater to the specific needsand recommendations of researchers.

The cortisol LFT technology is a unique proprietaryformat that improves on several related patented tech-nologies.41–43 We designed this LFT specifically for theresearch community to provide major advantages overcurrently available methods for saliva and include thefollowing: (1) the assay takes minutes from samplecollection to end results; (2) the assay and reading unit(Litebox Image Analysis Module, [LIAM]) are portable,allowing real-time cortisol assessment to be performed ina variety of point-of-care and nontraditional settings; and(3) the assay retains many of the advantages of availablemethods (eg, high sensitivity and quantitative output), sofew sacrifices in assay quality are required for real-time,point-of-care cortisol measurement. The present studydescribes the methodology and performance character-istics of real-time LFT cortisol, including comparison ofthis technology with another commercially availablenon-LFT assay. The goal was to determine correspond-ence and, by extension, viability of obtaining real-timecortisol scores within the near future.

METHODSParticipants

All procedures were approved by the institutionalreview board at a research university, and all participantsprovided informed consent. Participants were excluded if

Clinical Therapeutics

506 Volume 37 Number 3


they were taking oral steroids, had eaten or consumedalcohol within 1 hour, or were 436 years of age. Afterexclusion, the final sample size was N ¼ 29.

ProceduresParticipants provided four saliva samples, two in the

morning before 11:30 AM (mean = 8:54 AM; range =6:11–11:27 AM) and two in the afternoon/evening after1 PM (mean = 4:46 PM; range = 1:29–11:11 PM).These rough times were selected to obtain high and lowcortisol concentrations anticipated by the diurnal rhythmof the hormone. A total of three participants providedmorning or afternoon/evening samples only, leaving atotal of 55 samples. Order of sampling was not counter-balanced to avoid carry-over effects. First, participantsprovided an unstimulated saliva sample via Passive Droolby expectorating directly into a 2-mL polypropylenecryovial (mean [SD] = 3.43 [3.07] minutes for collectiontimes). Immediately after the Passive Drool sample andquestionnaire (10.53 [15.97] minutes later), participantsprovided a second saliva sample by using the Super � SALExtra Collection Device (Oasis Diagnostic Corporation),hereafter termed Super � SAL. This device serves as thefirst stage of LFT cortisol collection because it containsan absorbent pad that collects up to 2 mL of saliva (2.63[1.52] minutes for collection times). The absorbent padfilters many extraneous substances from saliva which isnecessary with LFT, because these types of assays aresensitive to viscosity and sample quality. Once the samplevolume adequacy indicator on Super � SAL indicated thatample volume was collected (by changing from lightgreen to blue), participants removed the collection device,inserted it into the compression tube, and expressed thesaliva into a cryovial by using force on the handle. Thecompression tube contains a proprietary filtration me-dium, which removes additional extraneous materialsand unwanted particulates that can compromise theLFT. In this way, the Super � SAL provided purifiedsaliva for the LFT and a confirmation specimen for latertesting. Samples were immediately frozen at –801C andshipped on dry ice to Oasis Diagnostic Corporation forassay by LFT. Residual samples were then shipped ondry ice to Middletown Research (Middletown, WI) forindependent confirmation by using the EIA method.

MethodsIBL Cortisol Assay

Both Passive Drool and Super � SAL specimens wereassayed with a solid-phase enzyme-linked immunosorbant

assay by using reagents with the same lot and expira-tion date purchased from IBL International (Ham-burg, Germany; www.IBL-International.com). TheIBL kit uses 50 uL of saliva with an expected rangeof 0.015 to 5 mg/dL cortisol and reported correspond-ence with IBL cortisol luminescence assay of R = 0.96.Intra- and inter-assay CVs are, on average, 4.9% and8.2%, respectively.

LFT Cortisol SystemSaliva from Super � SAL was assayed with the LFT;

Passive Drool was not used, given viscosity concerns.Given the novelty of this assay, nonproprietary detailsare provided. The basic platform is a positive read smallmolecule competitive quantitative LTF strip with severalcomponents. The first major component is a test housingthat incorporates the two LFT strips. A sample padlocated directly below the housing sample well acts as areservoir to distribute saliva to a conjugate pad. Theconjugate pad then accepts the saliva as a fluid mediumfor hydration of the dried conjugate. The VerOFy LFTpoint-of-care cortisol system then uses two europiumfluorescent particle-based conjugates that react with freeand conjugate bound cortisol in the saliva sample (here-after termed fluid because the properties of saliva arepurposely altered to accommodate the LFT). A nitro-cellulose membrane then accepts reacted conjugate fromthe conjugate pad. The nitrocellulose membrane thenbinds the reaction products to the two capture areas thateach contain two capture bands that are immobilizedonto the nitrocellulose membrane. The first pair of bandsconsists of immobilized bovine serum albumin-cortisol.The second pair of bands is a combination of bindingmolecules that scavenge any conjugate that does not bindto the bovine serum albumin-cortisol bands. An absorp-tion pad then accepts fluid from the nitrocellulosemembrane, providing the capillary engine to continu-ously pull fluid through the strip.

The second major component is LIAM which is afluorescent LFT cassette reader with onboard capa-bility to analyze and transfer cortisol data via blue-tooth connection. After the LFT strips are exposed tothe sample, the VerOFy test housing is inserted intothe LIAM reader because the test housing is specifi-cally engineered to expose the nitrocellulose mem-brane on the LFT strip to exciting UV light in a preciseorientation to ensure reproducible image analysis. TheLIAM uploads test information, including lot-to-lotcalibration curve parameters by reading a QR code on

E.A. Shirtcliff et al.

March 2015 507


the housing. The LIAM is also equipped with a digitaltemperature sensor that allows kinetic temperaturecorrections to be made within the operating range of181C to 301C because LFT is known to be sensitive totemperature. Ultimately, the LIAM will generate analgorithm to perform the quantitation to reportcortisol scores in units of picogram per milliliter withdata transferred to a computer or smartphone. Atpresent, results are reported as test-to-reference ratio(T/R) in which low ratio values indicate cortisol nearzero and higher ratio scores indicate greater binding inthe secondary capture zone relative to the primarycapture zone on the strip. That is, binding of cortisolto the primary zone is displaced and captured by thesecondary band when cortisol is high (Figure 1).

Statistical AnalysisFirst, we examined performance specifications of

the assay by using t tests and the CV. Correspondenceacross assay type was calculated as Pearson bivariatecorrelations and then Fisher R-to-Z transformation todetermine whether the magnitude of one correlationwas significantly larger than another. Finally, a seriesof linear regressions relied on the R2 to elucidate howmuch variance in LFT or Passive Drool, respectively,were accounted for by the other assays. We thenadded collection time or high-versus-low concentra-tion, respectively, to the linear regression to determinewhether these explained or changed the correspond-ence across assays.

RESULTSPerformance SpecificationsTime Course

To determine the time course of the LFT, that is theoptimal time to read the cortisol scores after samplecollection, saliva samples were run through ourcollection and purification system and read at5-minute intervals. The rise in cortisol concentrationsas the sample flows down across the nitrocellulosemembrane and binds to cortisol is shown in Figure 2.Stability appeared to be achieved at 20 minutes aftercollection. To probe further, a series of t tests werecalculated in which samples assayed before a timethreshold were compared with samples assayed after atime threshold. Samples assayed at 10 minutes, (t ¼4.7; P o 0.002) and 15 minutes (t ¼ 2.8; P o 0.025)were significantly lower than the samples assayedlater. The sample assayed at 20 minutes (t ¼ 1.91;P ¼ 0.10) and 25 minutes (t ¼ 1.35; P ¼ 0.22) and30 minutes (t ¼ 0.97; P ¼ 0.36) were not significantlydifferent from one another, suggesting stability in con-centration was achieved at �20 minutes. Hereafter,all LFT is reported according to a read time of20 minutes.

VariationPerformance specifications for LFT cortisol were

good. Of the possible 55 LFT devices, 4 failed qualitycontrol (QC) specifications (7.2%) on the basis of areaof the strip or image brightness.

Figure 1. Illustration of the LIAM reader and testhousing in which the VerOFy LFT stripsare integrated, as well as the Super � SALExtra Collection Device. LFT ¼ lateralflow test; LIAM ¼ Litebox Image Ana-lysis Module.

2.5

2

1.5

1

0.5

01 5 10

Time (in Minutes) before Reading LFT Cortisol

T/R

Rat

io f

or C

orti

sol

15 20 25 30 35 40

Replicate 1Replicate 2Mean

Figure 2. Time course of LFT cortisol suggeststhe optimal read time is 20 minutesafter collection, but stable reads areachieved earlier. LFT ¼ lateral flowtest; T/R ¼ test-to-reference.




Strips were selected a priori for quality control onthe basis of total area of the strip and image brightness.All strips were developed with a saliva pool purifiedwith Super � SAL and run five times by using strips fromthe same lot (CORT140528CB-2). The interassay CVwas excellent, with 6.99% variation across the fivereplicates. The interassay CV for low QC strips was notacceptable, given the low area of the strip and poorlight quality (49.15%), indicating the importance ofhigh QC. The reading for one replicate run to illustratethe calculation of the T/R is shown in Figure 3.

To determine the lower limit of detection, theprecision of read was measured with a purified salivapool that contained low cortisol in T/R units on thebasis of five bright readings. A calculated %CV wasdetermined to be 7% (ie, SD of 0.032), the lowestconcentration of cortisol that can be distinguishedfrom zero is 0.513 T/R (2 SDs from zero T/R). Withthe use of our current algorithm as a reference, thiscorresponds to 0.91 ng/mL (ie, 2.51 nmol/L), which iswell within the expected range of salivary cortisolwith the use of the IBL EIA kit.

Participant CharacteristicsFor the human saliva component, participants in-

cluded 32 adults (15 men) between the ages of 18 and36 years (27.5 [4.704] years). Participant ethnicitiesincluded white (65.6%), Hispanic (18.8%), African

American (3.1%), Asian or Pacific Islander (3.1%), andother (6.3%). For all of the saliva samples collected,only 1 sample of 51 (2%) fell below the lower limit ofdetection of 0.91 ng/mL (ie, 2.51 nmol/L).

Correspondence with EIACorrespondence across assay types is based on 55

samples for the immunoassays and 51 samples forLFT cortisol. Bivariate correlations of LFT with othermethods were highly significant (R ¼ 0.872 withSuper � SAL; R ¼ 0.739 with Passive Drool; both,P o 0.0001). With the use of Fisher R-to-Z trans-formation, we found that LFT indicated a trend for abetter correlation with Super � SAL than with PassiveDrool (Z ¼ 1.93, P ¼ 0.054). The intercorrelation ofPassive Drool with Super � SAL was highly significant(R ¼ 0.798 with Super � SAL, P o 0.0001). With theuse of Fisher R-to-Z transformation, we determinedthat the correlations of Passive Drool with Super � SALor LFT were comparable in magnitude (Z ¼ 0.24, P ¼0.47). These intercorrelations are shown in Figure 4.

A series of linear regressions were then run inwhich Passive Drool and then Super � SAL predictedLFT cortisol to further determine whether Super � SALor Passive Drool was best related to LFT. PassiveDrool significantly predicted LFT cortisol (β ¼ 0.74,P o 0.0001), explaining 55% of the variance in LFT

Example LIAMTM output used togenerate T/R Ratio

Den

sity

(pi

xels

)

Distance (pixels)0 100

120

100

80

60

40

20200 300 400 500 600 700

Figure 3. Example LIAM output used to generateT/R ratio. Like other assays, initialresults are reported in optical densityand then later converted to concentra-tion values. LIAM ¼ Litebox ImageAnalysis Module; T/R ¼ test-to-reference.

2 LFT T/RSsal1.8

1.8

1.6

1.4

1.2

1

0.80.6

0.40.2

0

1.6

1.4

1.2

1

0.8

0.6

0.40 0.5 1 1.5

Supe

rSal

Cot

isol

Passive Drool Cortisol ng/mL

LFT

Cor

tiso

l T/R

Rat

io

2 2.5 3

Figure 4. Intercorrelations between techniquesfor measuring cortisol are high. BothPassive Drool and Super � SAL cortisolare obtained with a commerciallyavailable assay from IBL; LFT cortisolwas obtained within 20 minutes ofassay start. LFT ¼ lateral flow test;Ssal, Super � SAL; T/R ¼ test-to-reference.


March 2015 509


cortisol (F[1,49] ¼ 58.9, P o 0.0001). WhenSuper � SAL was added as a predictor, an additional22% of the variance in LFT cortisol was explained(F[1,48] ¼ 44.49, P o 0.0001), suggesting Super-� SAL is related to LFT cortisol beyond the correlationwith Passive Drool. More specifically, although Super-� SAL predicted LFT cortisol independently of PassiveDrool (β ¼ 0.79, P o 0.0001), Passive Drool nolonger contributed independently to the prediction ofLFT cortisol (β ¼ 0.10, P ¼ 0.40). Conversely, wefound that Super � SAL alone explained 76% of thevariance in LFT cortisol, (F[1,49] ¼ 155.82, P o0.0001), but Passive Drool did not add further in theprediction of LFT cortisol, explaining 0.4% additionalvariance in LFT cortisol (F[1,48] ¼ 0.73, P ¼ 0.4).These findings suggest that LFT cortisol is related toother methods of measuring cortisol and may be bestrelated to Super � SAL cortisol that shares a collectiondevice methodology.

Given that Passive Drool may be considered thegold standard, we then determined which method wasbest correlated with Passive Drool cortisol. We con-ducted a parallel series of linear regressions to seewhether Super � SAL or LFT best correlated withPassive Drool. Beyond the 55% of the variance inPassive Drool cortisol explained by LFT cortisol,Super � SAL explained an additional 11% of thevariance in Passive Drool (F[1,48] ¼ 15.29, P o0.0001), and only Super � SAL remained an independ-ent predictor of Passive Drool for Super � SAL (β ¼0.68, P o 0.0001) and for LFT (β ¼ 0.15, P ¼ 0.4).LFT cortisol did not explain the variance in PassiveDrool cortisol beyond Super � SAL (R2 ¼ 0.5%,F[1,48] ¼ 0.73, P ¼ 0.4). Findings suggest PassiveDrool cortisol is significantly related to LFT cortisol,but Super � SAL explains additional variance in corti-sol beyond the LFT method perhaps related to the factthat both Passive Drool and Super � SAL were assayedwith the same IBL assay. Taken together, all threemethods are highly intercorrelated, but Super � SALmay be uniquely related to the other technologies,likely through shared collection devices (ie, LFT) orassay techniques (ie, Passive Drool).

To determine whether intercorrelations signifi-cantly differed between morning and afternoon/eve-ning collection times (correspondence of LFT withSuper � SAL was R ¼ 0.857 in the morning and R ¼0.719 in the afternoon/evening; correspondence ofLFT with Passive Drool was R ¼ 0.567 in the

morning and R ¼ 0.729 in the afternoon/evening; allPs o 0.05), we conducted a series of linear regressionsin which the main effect of assay type and time(morning or afternoon/evening) were included in step1, and the interaction between assay type and time ofday was added in step 2. With Passive Drool as theoutcome, we did not find that the intercorrelation withLFT differed by time of day (F[1,47] ¼ 0.54, P ¼0.47) nor did we find that the intercorrelation withSuper � SAL differed by time of day (F[1,51] ¼ 1.0,P ¼ 0.32). Similarly, we did not find that theintercorrelation of LFT with Super � SAL differed bytime of day (F[1,47] ¼ 0.03, P ¼ 0.87). Findingssuggest that the intercorrelations across assay type didnot systematically change across the day.

Parallel analyses contrasted specifically with low-versus-high cortisol concentrations across each assaytype by using a median split. We did not find that theintercorrelation of Passive Drool with LFT (F[1,47] ¼0.29, P ¼ 0.59) or Passive Drool with Super � SAL(F[1,51] ¼ 2.24, P ¼ 0.14) or LFT with Super � SAL(F[1,47] ¼ 0.73, P ¼ 0.4) differed by low-versus-highconcentrations, respectively. In summary, these threeassay types are highly intercorrelated, and the magni-tude of their association does not systematically varyby time of day or concentration.

DISCUSSIONThe biomarker cortisol is frequently investigated inpsychobiological research, with hundreds of studiesacross the past few decades suggesting its utility. Acritical barrier in the stress field is that currentimmunoassay methods do not provide reportableresults for cortisol for days to months after collection.Failure to obtain time-sensitive information is espe-cially problematic for this hormone because cortisolconstantly changes to allow the person to efficientlyencode and filter salient environmental cues.44 Withthe use of LFT and a portable LIAM we were able toassay cortisol reliably within 20 minutes with strongassay performance specifications and correspondencewith other methods.

Real-time cortisol measurement has the potential toadvance the field of stress physiology. Real-timetechnology with other psychobiological measures(eg, blood pressure, heart rate, or skin conductance)indicate the tremendous effect that real-time resultscan have on the stress field, including permittingbiofeedback studies,46 clinical trials to improve




relaxation,47 and enhancing cardiac diagnosis andclinical outcomes.48 We set out to confront the time-sensitivity barrier with the HPA axis because thesesystems can be dissociated.49 We believe the field ispoised to consider new HPA biomarkers, given thepowerful effect of advances in cortisol technology (eg,hair cortisol45).

Obtaining salivary cortisol scores within 20 mi-nutes of sample collection is advantageous becausethis time course matches well with the time course ofpeak cortisol reactivity after an acute stressor andprecedes the time in which behavioral effects ofcortisol are anticipated23 or before negative feedbackemerges.50 Practically speaking, this technology maybe useful for ensuring that the person is no longerstressed by laboratory arrival before initiating atargeted stressor. More interestingly, this techno-logy may open the possibility of identifying stressresponders while still within the laboratory. Success-ful laboratory stressors frequently trigger a stressresponse in �50% to 70% of participants.13,17 Evenfor robust laboratory challenges, it is problematicwhen a large percentage of persons are nonrespond-ers. Some persons may appear as nonresponders whenthey appraise the challenge, but successfully regulateresponsivity before crossing the relatively high stressthreshold of the HPA axis.51 Conversely, it may bemaladaptive for some persons to appear asnonresponders.52 In this case, failure to find an HPAstress response signifies that the person washyporesponsive to stress, with a low capacity tomount a stress response even in situations that callfor it. This lack of malleability in the HPA axis isproblematic because such persons would not be ableto recalibrate their physiologic functioning to meet thedemands of a changing environment, increasing theperson’s risk of stress-related diseases53 andproblems.54 Early identification of nonresponderswithin 20 minutes is an important step towarddistinguishing between these types of nonrespondersand further advancing the understanding of psycho-pathologic risk.

The time course of 20 minutes of delay is typical ofLFT,38 with shorter read times for qualitative resultsand longer read times for quantitative results.36 Werecognize that shorter delays may be desirable whilestill maintaining quantitative results as for the VerOFysystem. Future experiments will determine whetherearlier readings on the LIAM produce cortisol scores

that can be reliably extrapolated to inform theperson’s final cortisol reading. It is possible atpresent that each sample could be read multipletimes, beginning as early as 5 minutes aftercollection and then finalized by 20 minutes aftercollection. In the research setting, for example, thispossibility would allow stressors to be modifiedduring the challenge after a nonresponder isidentified so that the experimenter could enhance theacute challenge. A handful of studies has exploredhow modifications to social evaluative threat51,55–59

or uncontrollability57,60,61 can be experimentallyadjusted to influence rates of cortisol responsivity.The next step is to implement such modifications withinformation about the person’s current physiologybefore initiation of a modification.

Correspondence of LFT cortisol with Passive Drooland Super � SAL cortisol scores obtained with the IBLcommercial assay was good with correlations near orabove R ¼ 0.80. Future analyses will determinewhether similar correspondence is obtained with massspectrometry and other commercially available assays.Much of the degradation in correspondence wasintroduced before LFT, but this was necessary, giventhat LFT is sensitive to sample viscosity and partic-ulates. With traditional assays, particulates in salivaare reduced by the first freeze-thaw cycle, centrifuga-tion, and (in some cases) dilution or extraction. Withthe LFT, none of these initial steps are required toavoid the time delay in protocols that these stepsnecessitate (ranging from 10 minutes to 2 days).Indeed, costly laboratory equipment such as ultracoldfreezers or centrifuges are not needed, which opens apossibility of sample collection and assay in a range ofsettings without access to refrigeration or reliableshipping.62,63 The current methodology requires theportable fluorescence LIAM reader (whose cost is onpar with an inexpensive laptop computer) to obtainresults but no further large equipment. This technol-ogy thus reduces cost by eliminating the need forcostly laboratory equipment, shipping, and storage.Moreover, cortisol is currently analyzed at substantialcost per sample, regardless of its utility. This practicalburden is large, given that stress protocols fail in 30%to 50% of participants.13,17,64 Real-time analyticalinformation about this stress biomarker has the abilityto exert a powerful effect on the stress field. Althoughfinal costs are not yet determined, we anticipate thatthe cost of VerOFy cortisol will be substantially


March 2015 511


cheaper than current laboratory-based methods thatare typically $12 to $25 for external laboratories.

The present study has several limitations. First, werecognize that correspondence of LFT with tradi-tional assays is not perfect, and further validationmay be desirable on a study-by-study basis. There-fore, the collection device was designed to allow for aconfirmation specimen to be obtained and shipped toa laboratory, if desired. Second, we recognize theinherent variability in results and ongoing calibrationof the algorithm to convert LTF cortisol to stand-ardized units, and, consequently, we recommend thatthe device is used for research purposes at this time.Indeed, we believe that a major advantage of thistechnology is that it is being created through collab-oration between researchers and LFT experts tocreate a device that best accommodates the needsand desires of the research community. This divergesfrom the traditional applications of LFT. Nonethe-less, we acknowledge that at present the technology isnot yet likely to be applicable to diagnostic or treat-ment adherence domains, such as for Addison syn-drome or Cushing disease, in which laboratory- orclinic-based approaches are more widely ac-cepted.32,65 Third, time-course data require havingthe portable LIAM on site. This was not necessary toobtain correspondence data in the present study;nevertheless, planned studies will obtain cortisol datain real time across different sites. Finally, futuredirections include extension of this technology toother biomarkers frequently investigated in saliva.Regardless of these limitations, we believe that meas-urement of this time-varying stress biomarker can beenhanced through targeted exploration of its time-sensitive nature. The advantage of obtaining cortisolvalues within 20 minutes of saliva sample collectionrepresents an exciting new methodological directionwith the potential to contribute to conceptuallyeffective research directions.

ACKNOWLEDGMENTSWe thank the Stress Physiology Innovative Team(SPIT laboratory) for all their help as well as theemployees at Oasis Diagnostics and MiddletonResearch.

CONFLICTS OF INTERESTThis research was supported by a phase I (R43AT006634) and phase II (R44 AT006634) Small Business

Innovative Research Award to Dr. Slowey. Dr. Shirtcl-iff does not have a financial stake in the productand sits as a volunteer on the board of advisors ofOasis Diagnostics. Drs. Buck, Laughlin, Cole, andSlowey are owners or employees of Oasis Diagnostics,the company developing the VerOFy technology.Dr. Shirtcliff collected all samples and conducted allstatistical analyses to ensure the analyses were inde-pendent of the Oasis team. The authors have indicatedthat they have no other conflicts of interest regardingthe content of this article.

REFERENCES1. McEwen BS, Seeman T. Protective and damaging effects

of mediators of stress. Elaborating and testing theconcepts of allostasis and allostatic load. Ann N Y Acad

Sci. 1999;896:30–47.2. Juster RP, McEwen BS, Lupien SJ. Allostatic load bio-

markers of chronic stress and impact on health andcognition. Neurosci Biobehav Rev. 2010;35:2–16.

3. Lupien SJ, Ouellet-Morin I, Hupbach A, et al. Beyond thestress concept: allostatic load–a developmental biologicaland cognitive perspective. In: Cicchetti D, Cohen D, eds.Developmental Psychopathology, Vol 2. 2nd ed. Hoboken, NJ:John Wiley & Sons; 2006:578–628.

4. Uno H, Eisele S, Sakai A, et al. Neurotoxicity of gluco-corticoids in the primate brain. Horm Behav. 1994;28:336–348.

5. Segerstrom SC, Miller GE. Psychological stress and thehuman immune system: a meta-analytic study of 30 yearsof inquiry. Psychol Bull. 2004;130:601–630.

6. Heffelfinger AK, Newcomer JW. Glucocorticoid effects onmemory function over the human life span. Dev Psychopa-thol. 2001;13:491–513.

7. McMillen IC, Mulvogue HM, Kok JS, et al. Circadianrhythms in sleep and wakefulness and in salivary melato-nin and cortisol concentrations in mothers of term andpreterm infants. Sleep. 1993;16:624–631.

8. Dallman MF, la Fleur SE, Pecoraro NC, et al. Minireview:glucocorticoids–food intake, abdominal obesity, andwealthy nations in 2004. Endocrinology. 2004;145:2633–2638.

9. Johnson AE, Bruce J, Tarullo AR, Gunnar MR. Growthdelay as an index of allostatic load in young children:predictions to disinhibited social approach and diurnalcortisol activity. Dev Psychopathol. 2011;23:859–871.

10. Tilbrook AJ, Turner AI, Clarke IJ. Effects of stresson reproduction in non-rodent mammals: the role ofglucocorticoids and sex differences. Rev Reprod. 2000;5:105–113.




11. Epel ES, Lin J, Wilhelm FH, et al.Cell aging in relation to stressarousal and cardiovascular diseaserisk factors. Psychoneuroendocrinol-

ogy. 2006;31:277–287.12. Frey BM, Frey FJ. Clinical pharma-

cokinetics of prednisone and pre-dnisolone. Clin Pharmacokinet. 1990;19:126–146.

13. Pruessner JC, Dedovic K, PruessnerM, et al. Stress regulation in thecentral nervous system: evidencefrom structural and functionalneuroimaging studies in humanpopulations - 2008 Curt RichterAward Winner. Psychoneuroendocri-

nology. 2010;35:179–191.14. Kern S, Oakes TR, Stone CK, et al.

Glucose metabolic changes in theprefrontal cortex are associatedwith HPA axis response to a psy-chosocial stressor. Psychoneuroen-

docrinology. 2008;33:517–529.15. Vazquez DM. Stress and the devel-

oping limbic-hypothalamic-pituitary-adrenal axis. Psychoneuroendocrinology.1998;23:663–700.

16. Dickerson SS, Kemeny ME. Acutestressors and cortisol responses: atheoretical integration and synthe-sis of laboratory research. PsycholBull. 2004;130:355–391.

17. Kirschbaum C, Pirke KM, Hellham-mer DH. The 'Trier Social StressTest'–a tool for investigating psy-chobiological stress responses in alaboratory setting. Neuropsychobiol-ogy. 1993;28:76–81.

18. Gunnar MR, Vazquez D. Stressneurobiology and developmentalpsychopathology. In: Cicchetti D,Cohen DJ, eds. Developmental Psy-chopathology. Vol 2. 2nd ed. Hobo-ken, NJ: John Wiley & Sons;2006:533–577.

19. De Kloet ER. Hormones and thestressed brain. Ann N Y Acad Sci.2004;1018:1–15.

20. Gunnar M, Quevedo K. The neuro-biology of stress and development.Annu Rev Psychol. 2007;58:145–173.

21. Yehuda R, Southwick SM, KrystalJH, et al. Enhanced suppression of

cortisol following dexamethasoneadministration in posttraumaticstress disorder. Am J Psychiatry.1993;150:83–86.

22. Spijker AT, van Rossum EF. Glu-cocorticoid sensitivity in mooddisorders. Neuroendocrinology. 2012;95:179–186.

23. Sapolsky RM, Romero LM, MunckAU. How do glucocorticoids influ-ence stress responses? Integratingpermissive, suppressive, stimula-tory, and preparative actions.Endocr Rev. 2000;21:55–89.

24. Granger DA, Kivlighan KT, Fortu-nato C, et al. Integration of salivarybiomarkers into developmental andbehaviorally-oriented research:problems and solutions for collect-ing specimens. Physiol Behav. 2007;92:583–590.

25. Shirtcliff EA, Allison AL, Arm-strong JM, et al. Longitudinalstability and developmental prop-erties of salivary cortisol levels andcircadian rhythms from childhoodto adolescence. Dev Psychobiol. 2012;54:493–502.

26. Gunnar M, Connors J, Isensee J.Lack of stability in neonatal adre-nocortical reactivity because ofrapid habituation of the adreno-cortical response. Dev Psychobiol.1989;22:221–233.

27. Skinner ML, Shirtcliff EA, HaggertyKP, et al. Allostasis model facili-tates understanding race differen-ces in the diurnal cortisol rhythm.Dev Psychopathol. 2011;23:1167–1186.

28. Larina IM, Bystritskaya AF, Smir-nova TM. Psychophysiologicalmonitoring under conditions ofreal and simulated microgravity.Hum Physiol. 1999;25:574–579.

29. Chatterton RT Jr, Vogelsong KM,Lu YC, Hudgens GA. Hormonalresponses to psychological stressin men preparing for skydiving.J Clin Endocrinol Metab. 1997;82:2503–2509.

30. Ross IL, Levitt NS, Van der Walt JS,et al. Salivary cortisol day curves

in Addison's disease in patients onhydrocortisone replacement. HormMetab Res. 2013;45:62–68.

31. Smans L, Lentjes E, Hermus A,Zelissen P. Salivary cortisol daycurves in assessing glucocorticoidreplacement therapy in Addison'sdisease. Hormones (Athens). 2013;12:93–100.

32. Raff H. Utility of salivary cortisolmeasurements in Cushing's syn-drome and adrenal insufficiency.J Clin Endocrinol Metab. 2009;94:3647–3655.

33. Restituto P, Galofre JC, Gil MJ,et al. Advantage of salivary corti-sol measurements in the diagnosisof glucocorticoid related disor-ders. Clin Biochem. 2008;41:688–692.

34. Kaushik A, Vasudev A, Arya SK,et al. Recent advances in cortisolsensing technologies for point-of-care application. Biosens Bioelec-

tron. 2014;53:499–512.35. Nara S, Tripathi V, Singh H, Shriv-

astav TG. Colloidal gold probebased rapid immunochromato-graphic strip assay for cortisol.Anal Chim Acta. 2010;682:66–71.

36. Zangheri M, Cevenini L, Anfossi L,et al. A simple and compact smart-phone accessory for quantitativechemiluminescence-based lateralflow immunoassay for salivary cor-tisol detection. Biosens Bioelectron.2015;64:63–68.

37. Choi S, Kim S, Yang J-S, et al. Real-time measurement of human sali-vary cortisol for the assessment ofpsychological stress using a smart-phone. Sensing Bio-Sensing Res.2014;2:8–11.

38. Posthuma-Trumpie GA, Korf J, vanAmerongen A. Lateral flow (im-muno)assay: its strengths, weak-nesses, opportunities and threats.A literature survey. Anal Bioanal

Chem. 2009;393:569–582.39. Eisenger RW, Khalil MH, Katz DH,

Sargeant RB, inventors. Lateral flow,non-bibulous membrane assay proto-

cols. 1990.


March 2015 513


40. Wong R, Tse H. Lateral Flow Immu-

noassay. New York, NY: HumanaPress; 2009.

41. Bauer JS, Hyatt TP, Wang H, BuckRL, inventors. Positive detectionlateral flow apparatus and methodfor small and large analytes. 2009.

42. Buck RL, inventor. Lateral flow teststrip with migrating label. 2009.

43. Boehringer H, Rowley G, PronovostAD, inventors. Quantitative lateralflow assays and devices. 2005.

44. Del Giudice M, Ellis BJ, Shirtcliff EA.The Adaptive Calibration Model ofstress responsivity. Neurosci Biobehav

Rev. 2011;35:1562–1592.45. Sharpley CF, McFarlane JR,

Slominski A. Stress-linked cortisolconcentrations in hair: what weknow and what we need to know.Rev Neurosci. 2012;23:111–121.

46. McGrady A. Effects of group relax-ation training and thermal biofeed-back on blood pressure andrelated physiological and psycho-logical variables in essential hyper-tension. Biofeedback Self Regul. 1994;19:51–66.

47. van Santen M, Bolwijn P,Verstappen F, et al. A randomizedclinical trial comparing fitness andbiofeedback training versus basictreatment in patients with fibromyal-gia. J Rheumatol. 2002;29:575–581.

48. Rokos IC, French WJ, Mattu A,et al. Appropriate cardiac cathlab activation: optimizing electro-cardiogram interpretation and cli-nical decision-making for acuteST-elevation myocardial infarction.Am Heart J. 2010;160:995–1003.1003.e1-8.

49. Andrews J, Pruessner JC. The com-bined propranolol/TSST paradigm–anew method for psychoneuroendoc-rinology. PLoS One. 2013;8:e57567.

50. Jahn AL, Fox AS, Abercrombie HC,et al. Subgenual prefrontal cortexactivity predicts individual differ-ences in hypothalamic-pituitary-adrenal activity across differentcontexts. Biol Psychiatry. 2010;67:175–181.

51. Bosch JA, de Geus EJ, Carroll D,et al. A general enhancement ofautonomic and cortisol responsesduring social evaluative threat. Psy-chosom Med. 2009;71:877–885.

52. Petrowski K, Herold U, Joraschky P,et al. A striking pattern of cortisolnon-responsiveness to psychosocialstress in patients with panic disorderwith concurrent normal cortisolawakening responses. Psychoneuroen-docrinology. 2009;35:4147–4421.

53. Raubenheimer PJ, Young EA, An-drew R, Seckl JR. The role ofcorticosterone in human hypotha-lamic-pituitary-adrenal axis feed-back. Clin Endocrinol (Oxf). 2006;65:22–26.

54. Shirtcliff EA, Vitacco MJ, Graf AR,et al. Neurobiology of empathy andcallousness: implications for thedevelopment of antisocial behavior.Behav Sci Law. 2009;27:137–171.

55. Het S, Rohleder N, Schoofs D, et al.Neuroendocrine and psychometricevaluation of a placebo versionof the ‘Trier Social Stress Test’. Psy-choneuroendocrinology. 2009;34:1075–1086.

56. Wiemers US, Schoofs D, Wolf OT.A friendly version of the trier socialstress test does not activate theHPA axis in healthy men andwomen. Stress. 2013;16:254–260.

57. Westenberg PM, Bokhorst CL,Miers AC, et al. A prepared speechin front of a pre-recorded audi-ence: subjective, physiological, andneuroendocrine responses to theLeiden Public Speaking Task. Biol

Psychol. 2009;82:116–124.58. Taylor SE, Seeman TE, Eisenberger

NI, et al. Effects of a supportive or

an unsupportive audience on bio-logical and psychological respon-ses to stress. J Pers Soc Psychol.2010;98:47–56.

59. Dickerson SS, Mycek PJ, Zaldivar F.Negative social evaluation, but notmere social presence, elicits corti-sol responses to a laboratory stres-sor task. Health Psychol. 2008;27:116–121.

60. Taylor SE, Burklund LJ, EisenbergerNI, et al. Neural bases of moder-ation of cortisol stress responsesby psychosocial resources. J Pers

Soc Psychol. 2008;95:197–211.61. Creswell JD, Welch WT, Taylor SE,

et al. Affirmation of personal val-ues buffers neuroendocrine and psy-chological stress responses. Psychol

Sci. 2005;16:846–851.62. Worthman CM, Panter-Brick C.

Homeless street children in Nepal:use of allostatic load to assess theburden of childhood adversity. DevPsychopathol. 2008;20:233–255.

63. Decker SA. Low salivary cortisoland elevated depressive affectamong rural men in Botswana:reliability and validity of laboratoryresults. J Physiol Anthropol. 2006;25:91–101.

64. Gunnar MR, Talge NM, Herrera A.Stressor paradigms in developmen-tal studies: what does and does notwork to produce mean increases insalivary cortisol. Psychoneuroendocri-nology. 2009;34:953–967.

65. Ceccato F, Barbot M, Zilio M,et al. Performance of salivary cor-tisol in the diagnosis of Cushing’ssyndrome, adrenal incidentaloma,and adrenal insufficiency. Eur J

Endocrinol. 2013;169:31–36.

Address correspondence to: Elizabeth Shirtcliff, PhD, Iowa State University,4380 Palmer, Suite 2330, Ames, Iowa 50011-4380. E-mail: [email protected]



Salivary Cortisol Results Obtainable within Minutes of Sample Collection Correspond with Traditional...

Documents

Transcript of Salivary Cortisol Results Obtainable within Minutes of Sample Collection Correspond with Traditional...