Psychoneuroendocrinology (2014) 49, 47—53

Available online at www.sciencedirect.com

ScienceDirect

j ourna l h om epa ge : www.elsev ier .com/ locate /psyneuen

Multiple time courses of salivaryalpha-amylase and dimensions of affect inadolescence

Leah D. Doane ∗, Scott A. Van Lenten

Department of Psychology, Arizona State University, United States

Received 7 March 2014; received in revised form 27 May 2014; accepted 10 June 2014

KEYWORDSSalivaryalpha-amylase;Positive affect;Negative affect;Arousal;Adolescence

Summary Previous research has illustrated associations among daily experiences, emotionsand stress-responding physiological systems. Recently, investigators have examined salivaryalpha-amylase (sAA), a surrogate marker of the autonomic nervous system, and its associationswith affect. The current study examined associations among affective valence, arousal andsAA across three different time courses at the momentary, daily and inter-individual level tounderstand varying influences of adolescents’ daily emotional experiences on sAA reactivity anddiurnal sAA activity. Adolescents (N = 82) provided salivary samples and diary reports of affectand experiences five times a day for three consecutive days. They also completed self-reportquestionnaires on trait affect. Findings from multilevel growth curves demonstrated that ado-lescents in our sample displayed typical sAA diurnal rhythms with levels dropping 30 min afterwaking and then increasing across the day to a peak in the late afternoon. Within person momen-tary experiences of high arousal positive affect were associated with momentary sAA reactivity.Prior day experiences of high arousal negative affect were associated with a greater amylaseawakening response (i.e., greater decrease) and flatter slopes the next day. Trait positive affectwas also associated with flatter sAA slopes. Our findings suggest that both affective arousal and

valence should be accounted for when examining differences in sAA reactivity and diurnal pat-terns. Further, our results indicated that emotion-physiology transactions among adolescentsoccur over varying time scales for salivary alpha-amylase as well as cortisol.

ts re

© 2014 Elsevier Ltd. All righ

∗ Corresponding author at: Department of Psychology, ArizonaState University, P.O. Box 871104, Arizona State University, Tempe,AZ 85287-1104, United States. Tel.: +1 480 965 5289;fax: +1 480 965 8544.

E-mail address: Leah.Doane@asu.edu (L.D. Doane).

1

Dcmohm

http://dx.doi.org/10.1016/j.psyneuen.2014.06.0070306-4530/© 2014 Elsevier Ltd. All rights reserved.

served.

. Introduction

aily socioemotional experiences of children and adoles-ents are associated with changes in physiology across

ultiple time courses: momentary, day-to-day, yearly, and

ntogenetically (Adam, 2012). Alterations across time areypothesized to be adaptations to sociocultural environ-ents and biological states. Specific momentary and daily

4

efFptmwnsnh

gsnstmraRl(

r(tN2teviseda(he

irsracpltnuhidwvttpOs

g(htdowes(ddwoD(cstua

2

2

DpstaPutstdAer(qpfia

2

PptAcpart

8

vents, such as experiencing discrete emotions, may dif-erentially affect physiological stress-responding systems.or instance, prior research investigating hypothalamic-ituitary-adrenal (HPA) axis function in adolescents foundhat momentary loneliness was associated with increasedomentary cortisol and high levels of loneliness one dayere related to increased cortisol awakening responses theext day (Doane and Adam, 2010). Examining physiologicalystems in response to socioemotional experiences simulta-eously across multiple time scales can provide insight intoow dimensions of emotion and physiology are related.

Salivary alpha-amylase (sAA) is another valuable surro-ate marker to assess stress-related physiological changes.AA is a digestive enzyme secreted in response to autonomicervous system (ANS) activation (Granger et al., 2007). Bothympathetic and parasympathetic nervous system innerva-ions stimulate the secretion of sAA via �- and �-adrenergicechanisms. Therefore, sAA can be conceptualized as a sur-

ogate marker of ANS activation and central noradrenergicctivity (Bosch et al., 2011; Cubała and Landowski, 2014).esearchers have also demonstrated causal links betweenevels of sAA and ANS activity using experimental designse.g., Rohleder et al., 2004).

Studies utilizing lab paradigms have shown that sAAesponds to a variety of stress tasks and affective stimuliNater and Rohleder, 2009). For example, evidence indicateshat sAA increases in response to both cognitive tasks (e.g.,oto et al., 2005) and psychosocial stress tasks (Takai et al.,004; Gordis et al., 2006). Further, sAA is sensitive to emo-ional stimuli such as affectively-valenced pictures (Boscht al., 2003; Sánchez-Navarro et al., 2012) and aversiveideos (Segal and Cahill, 2009), suggesting that sAA is anndicator of distress. Research has also demonstrated thatAA levels increase in response to stress-provoking experi-nces in naturalistic settings. For example, researchers haveocumented increases in sAA concentrations prior to takingn exam (Bosch et al., 1996) and before one’s first skydiveChatterton et al., 1997). Elevated momentary levels of sAAave also been associated with greater chronic stress (Natert al., 2007).

Despite a growing sAA literature, few researchers havenvestigated relations between momentary affective expe-iences and sAA reactivity outside of controlled laboratoryettings (i.e., naturalistic sAA reactivity). Nater et al. (2007)eported associations between momentary positive affectnd increased sAA, indicating that emotional valence mayontribute to specific changes in sAA. A study examininghysiological stress reactivity during a naturalistic fear chal-enge found that sAA concentrations increased in responseo fear, but only for those who perceived the experience asegative, indicating that both arousal and valence may stim-late sAA activity (Buchanan et al., 2010). Adam et al. (2011)ypothesized that emotional arousal may be a more salientndicator of sAA reactivity than valence. Their findingsemonstrated that momentary sAA levels were associatedith high arousal affective states, regardless of emotionalalence. sAA exhibits a diurnal pattern opposite that of cor-isol: sAA concentrations have a substantial decrease during

he 30 min after waking, then increase across the day witheaks in the late afternoon or evening (Nater et al., 2007;ut et al., 2013). Understanding the coupling of affectivetates and sAA diurnal activity is critical in adolescence

tetA

L.D. Doane, S.A. Van Lenten

iven developmental changes in affect and ANS activityLarson et al., 2003; Adam et al., 2011). Notably, researchersave not yet identified whether daily experiences of emo-ion are associated with sAA diurnal rhythms the nextay. The chronometric model of emotion-stress physiol-gy transactions (Adam, 2012) has only been substantiatedith cortisol (e.g., Doane and Adam, 2010); therefore, wexamined whether changes in affective states and corre-ponding sAA activity unfolded over a variety of time coursesAdam, 2012). This hypothesis is supported through researchemonstrating relations between interpersonal stress anday-to-day changes in diurnal sAA (Savla et al., 2013), asell as associations between ‘‘trait-like’’ emotional dis-rders (e.g., Post Traumatic Stress Disorder, PTSD; Majorepressive Disorder, MDD) and altered diurnal sAA patternsThoma et al., 2012; Cubała and Landowski, 2014). Thus, theurrent study examined associations among several dimen-ions of affective experiences and sAA across three differentime courses: momentary, daily, and inter-individually, tonderstand the varying influences of adolescents’ dailyffective experiences on diurnal sAA activity.

. Method

.1. Participants

ata came from a diverse sample of adolescents antici-ating college enrollment in the subsequent fall (N = 82;ee Doane and Zeiders, 2014). Participants were recruitedhrough orientation activities for the psychology departmentt a large southwestern university and/or through email.articipants were required to live within 35 miles of theniversity, be a senior in a local high school, and confirmhey were coming to the university the subsequent fall. Theample was racially/ethnically diverse and representative ofhe university: 54% Non Hispanic White, 23% Latino/Hispanicescent, 13% multiracial, 5% African American and 5% Asianmerican/Pacific Islander (see Table 1). Participants werexcluded from analyses for one or more of the followingeasons: fibromyalgia (n = 1), corticosteroid medication usen = 1), noncompliance with protocols (n = 2), insufficientuestionnaire data (n = 2), or 75% or more of salivary sam-les with sAA levels below measurement range (n = 9). Thenal analytic sample consisted of 68 adolescents (25% male),ged 17—18 years (M = 18.04, SD = .37).

.2. Procedure

articipants selected three typical consecutive weekdays toarticipate. Weekdays rather than weekends were selectedo minimize the influence of varying sleep schedules onNS activity (e.g., Meerlo et al., 2008). Participants signedonsent forms upon delivery of materials to their homes;arental consents were collected for participants under thege of 18. Participants were compensated $40. Study mate-ials included three diary booklets, an actigraph watch,rack cap compliance device with straws, vials, and ques-

ionnaires. Participants provided salivary samples and diaryntries immediately after waking, 30 min later, twice duringhe day, and at bedtime for three consecutive days (e.g.,dam et al., 2011). In total, participants were required to

Multiple

time

courses of

salivary alpha-am

ylase and

dimensions

of affect

in adolescence

49

Table 1 Descriptive statistics and zero order correlations of salivary alpha-amylase, affect, and covariates.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Mean (SD) or N(% of sample)

Min Max

1. Waking levels of salivaryalpha-amylase

1 56.65 (48.50) 2.62 193.80

2. Amylase awakeningresponse

−.43** 1 −.74 (1.30) −4.25 3.71

3. Amylase diurnal slope −.38** .38** 1 .01 (.06) −0.11 0.184. Positive affect high

arousala−.11 .33** −.01 1 .83 (.59) 0.07 2.75

5. Positive affect low arousala −.06 .29* .08 .90* 1 1.07 (.60) 0.17 2.966. Negative affect high

arousala.07 .08 .01 .22 .22 1 .34 (.33) 0.00 1.67

7. Negative affect lowarousala

.08 .04 .06 .21 .22 .87** 1 .23 (.26) 0.00 1.50

8. Trait positive affect (ATQ)b −.03 −.13 −.02 .14 .19 −.31** −.30** 1 5.27 (1.08) 2.36 7.009. Trait negative affect

(ATQ)b−.01 .02 −.04 −.25* −.25* .09 .12 −.59** 1 3.51 (.99) 1.11 6.00

10. Gender (male = 1) −.09 .16 .09 .18 .12 −.12 −.07 .10 −.25* 1 17 (25%)11. Non Hispanic White .19 −.25* −.29* −.08 −.08 .13 .08 −.07 .06 .03 1 38 (57%)12. Parent education level .09 .05 −.04 .25* .25* −.06 −.01 .12 −.11 .04 −.29* 1 3.39 (1.57) 1.00 6.0013. Oral contraceptive use .01 −.18 −.16 −.11 −.03 .13 .02 −.07 .33** −.34 .16 −.11 1 17 (25%)14. Age .04 .07 .06 −.15 −.22 .19 .17 −.26* .09 .06 .24 .07 −.05 1 18.04 (.37) 17.17 18.6715. Caffeine use prior to

sampling.12 −.15 .03 −.23 −.20 .07 .01 −.14 .23 −.20 .05 .10 .15 .04 1 .06 (.10) 0.00 0.47

16. Nicotine use prior tosampling

−.05 .07 −.06 −.20 .06 .23 .37** −.13 .10 .23 .10 .14 −.01 .18 −.07 .01 (.05) 0.00 0.36

a Aggregate between person diary indicators of high and low arousal positive and negative affect.b ATQ: Adult Temperament Questionnaire.* p < .05.

** p < .01.

5

fipopw(aec

2

2svesdodsLacvw

2PN2panaPbsiNMtpsw

2Piftsy((

2

Aat

dmtpsv2iAoaap

3

PedapawmnaPsrmi

(Nwdwea(� = −.016, p < .05), when controlling for trait affect. Trait PAwas associated with a flatter sAA slope (� = −.017, p < .05).

1 This linear time variable was constructed by subtracting thewake-up time from each of the individual time points such thattime could be interpreted as 0 = wake-up, .5 = wake up + 30Ymin,2.5 = 2.5 hours after waking, etc.

2 General models were conducted as follows: Level 1:sAA = �0 + �1(AAR) + �2(Time Since Waking) + �3(Time SinceWaking2) + �4 to �5(Momentary Affective × Arousal States) + �6 to�8(Momentary Covariates) + εijk Level 2: �0 to �3 = ˇi0 + ˇi1 to ˇi2(Prior Day Affect × Arousal) + �ij Level 3: ˇi0 to ˇij = � ij0 + � ij(TraitAffect) + � ij2 to � ij7 (Demographic and Health Covariates) + rijk.

3 The outcome of interest, sAA level, was log transformed, thusthe effect sizes for all of the variables can be interpreted as a per-

0

ll out fifteen diary entries (M = 14.18, SD = 1.27). Strict com-liance parameters were used to ensure accurate modelingf sAA diurnal patterns. We considered participants com-liant if (1) their track-cap-detected waking sample wasithin 15 min of their actigraph-detected wake time, and

2) their track-cap-detected second sample was between 23nd 37 min after their first sample. Based on these param-ters, 32 waking samples and 48 wake + 30 samples wereonsidered noncompliant and removed from analyses.

.3. Measures

.3.1. Salivary alpha-amylaseAA was collected by passive drool. Participants labeledials with time and date of sampling. Samples were refrig-rated until collected from participants’ homes and weretored at −20 ◦C. Samples were sent by courier over threeays on dry ice to Biochemisches Labor at the Universityf Trier (Trier, Germany). sAA samples were assayed inuplicate using a kinetic reaction utilizing a chromogenicubstrate, 2-chloro-4-nitrophenyl-D Maltrotriosid (CNP-G3;orentz et al., 1999). The intraassay coefficient of vari-tion ranged between 4.0% and 6.7%, and the inter-assayoefficients of variation ranged between 7.1% and 9.0%. sAAalues were log transformed and outliers (>3 SD from mean)ere winsorized.

.3.2. Momentary and daily affectarticipants rated their levels of affect from the Positive andegative Affect Schedule (PANAS; Watson et al., 1988). Using0 items, participants responded with how much they felt aarticular emotion in the past hour (0 = very slightly or not atll to 4 = extremely). Momentary and daily affect were orga-ized to signify dimensions of both emotional valence androusal. Four scales were created representing High Arousalositive (PAH; e.g., excited, enthusiastic, inspired), Cron-ach’s alpha = .81; High Arousal Negative (NAH; e.g., upset,cared, nervous), ˛ = .76; Lower Arousal Positive (PAL; e.g.,nterested, strong, attentive), ˛ = .83; and Lower Arousalegative (NAL; e.g., guilty, ashamed, irritable), ˛ = .64.omentary dimensions of affect were within-person cen-

ered by subtracting an individual’s level at a given timeoint from their person-level aggregated scores. Day-levelcores were created by averaging across the five momentsithin a day.

.3.3. Trait positive and negative affecterson-level measures of affect were assessed using a mod-fied version of the Adult Temperament Questionnaire shortorm (ATQ; Evans and Rothbart, 2007). Participants ratedhe extent to which each of the statements described them-elves (1 = extremely untrue of you to 7 = extremely true ofou). The present study utilized the Positive Affect subscale˛ = .88) and the factor scale for Negative Affect-Frustration˛ = .84).

.4. Analytic plan

three-level hierarchical linear growth model was used tossess associations among affect and sAA and to account forhe nested nature of the data. At Level 1 each individual’s

cu

m

L.D. Doane, S.A. Van Lenten

iurnal rhythm and amylase awakening response (AAR) wereodeled. This was done by including linear and quadratic

ime variables indicating how long since waking the sam-le was given,1 and a dummy variable representing the AARample (Adam et al., 2011). Within-person centered affectariables were also included at Level 1 (Model 1). At Level, prior-day affect was included in the prediction of wak-ng level, linear time (e.g., slope), and the AAR (Model 2).t Level 3, person-specific parameters including trait levelsf positive and negative affect were added (Model 3).2 Allnalyses included covariates of gender, momentary caffeinend nicotine use, oral contraceptive use, race/ethnicity,arents’ educational status, and age.

. Results

articipants showed the expected diurnal pattern of mod-rate morning levels (� = 3.367, p < .001), a significantecrease 30 min after waking — the AAR (� = −.832, p < .01),n increase in sAA across the day (linear time: � = .100,

< .01) at a rate of 10.5% at waking3 and slower increasescross the day (quadratic time: � = −.006, p < .01). Next,e included low and high arousal PA and NA in theodel. Momentary PAH was the only affect variable sig-

ificantly associated with momentary sAA therefore lowrousal variables were not included in subsequent models.4

AH was significantly associated with momentary levels ofAA (� = .081, p < .05), suggesting that when an individualeported greater PAH (compared to his/her own cross-timeean), the participant demonstrated a corresponding 8.7%

ncrease in sAA levels (Table 2, Model 1).When prior-day indicators of NAH and PAH were included

Table 2, Model 2), significant associations emerged forAH but not PAH. Prior-day NAH was marginally associatedith flatter sAA slopes (i.e., less of an increase across theay; � = −.013, p < .10). When trait indicators of NA and PAere included (Model 3), further day-to-day associationsmerged. Prior-day NAH was significantly associated with

greater AAR (� = −.259, p < .05), and a flatter sAA slope

ent change per one unit change in the independent variable aftertilizing the following transformation: �%change = ((eˆ�) − 1).4 All results were consistent with and without these low affectomentary measures included in the models.

Multiple time courses of salivary alpha-amylase and dimensions of affect in adolescence 51

Table 2 Multilevel growth model regression estimates predicting daily salivary alpha-amylase activity from momentary, priorday and trait positive and negative affect.

Model 1 momentary Model 2 prior day Model 3 trait

Coefficient SE Coefficient SE Coefficient SE

Intercept: waking level alpha amylase, �0 3.367*** .130 3.373*** .129 3.380*** .137Male −.013 .278 −.030 .297 −.046 .300Non Hispanic White .313 .267 .296 .274 .195 .292Parent education −.026 .081 −.032 .078 −.029 .087Oral contraceptive use −.203 .253 −.237 .252 −.170 .297Age −.563** .209 −.593** .212 −.613** .218

Level 2: Prior day high arousal PA .040 .181 .068 .724Level 2: Prior day high arousal NA .172 .115 .176 .112

Level 3: Trait positive affect −.021 .149Level 3: Trait negative affect .078 .118

Amylase awakening response, �1 −.892** .143 −.897** .142 −.890*** .140Male .216 .205 .201 .325 .218 .335Non Hispanic White −.280 .302 −.253 .325 −.296 .327Parent education .001 .089 .001 .089 .004 .094Oral contraceptive use −.150 .358 −.116 .356 −.075 .397Age −.045 .314 .006 .324 .133 .331

Level 2: Prior day high arousal PA −.170 .139 .069 .286Level 2: Prior day high arousal NA .015 .232 −.259* .110

Level 3: Trait positive affect −.267 .196Level 3: Trait negative affect −.044 .189

Time since waking: slope, �2 .100** .031 .099** .031 .099** .032Male .030* .015 .032* .015 .032* .016Non Hispanic White −.044* .016 −.043** .016 −.041* .017Parent education .001 .005 .003 .005 .005 .005Oral contraceptive use .016 .019 .019 .019 .024 .020Age .051** .016 .051** .017 .060** .017

Level 2: Prior day high arousal PA −.008 .008 −.015 .010Level 2: Prior day high arousal NA −.013† .008 −.016* .008

Level 3: Trait positive affect −.017* .007Level 3: Trait negative affect −.014 .009

Time since waking squared, �3 −.006** .002 −.006** .002 −.006** .002Momentary high arousal positive affect, �4 .081* .04 .082* .04 .077* .039Momentary high arousal negative affect, �5 .01 .048 .004 .048 −.005 .046Momentary caffeine consumption, �6 .327* .149 .309* .151 .328* .153Momentary nicotine use, �7 −1.158** .334 −1.102** .357 −.977* .315

All sAA levels reflect log10 U/mL, bold values indicate p < .05.† p < .10.* p < .05.

** p < .01.

ae

rca(ss

*** p < .001.

4. Discussion

Prior research has identified associations among socioemo-tional experiences of adolescents and changes in physiology,specifically HPA axis activity, across multiple time courses:momentary, day-to-day, and yearly (Adam, 2012). To thebest of our knowledge, our findings are the first to demon-strate significant within- and between-person associationsamong adolescents’ daily affective experiences and momen-

tary and diurnal variation in sAA, a surrogate marker ofANS activation and central noradrenergic activity. Fur-ther, results suggested that both emotional arousal andvalence play key roles in momentary, daily, and trait

hr

N

ssociations, but differentially depending on the time coursexamined.

In line with previous studies investigating naturalisticeactivity (Nater et al., 2007), we found a significant asso-iation between increases in PAH and momentary sAA, afterccounting for typical diurnal variation and daily behaviorse.g., caffeine and nicotine use). We did not find momentaryAA reactivity in relation to NAH or low arousal affectivetates. This finding partially supports Adam et al.’s (2011)

ypothesis that sAA is a more global indicator of arousalather than distress.

Although we did not find sAA reactivity to momentaryAH, we did detect daily associations with NAH. To the best

5

otvmteaAiPspelh(mfitaNabsicts

brtgtcd

sctpceplnomaiowsdiotA(r

a

ctrstcp2attafaFrtotp

aetpnpatraesift

R

TSPitt

C

Ni

A

TTcontributed to this research. We would also like to thank

2

f our knowledge, we were the first to examine the con-ributing role of prior-day affective states in sAA diurnalariation. Prior-day NAH was associated with an approxi-ately 29% greater AAR decrease and a 2% flatter sAA slope

he next day. We hypothesize that adolescents who experi-nced high levels of negative arousal one day exhibited andaptive physiological response the next day with a greaterAR decrease and less overall sAA across the day. Lower wak-

ng values and increases in the AAR have been documented inTSD samples (Thoma et al., 2012), therefore our findingsuggest greater decreases in AARs may be more adaptivehysiologically in helping individuals cope with their dailyxperiences. Specifically, we hypothesize that healthy ado-escents who experienced high levels of NAH one day mayave physiologically reacted differently the next morningi.e., greater decreases in the AAR) as a potential copingechanism in preparation for the upcoming day. Further, ourndings are consistent with another study of older adultshat found daily associations between interpersonal stressnd flatter sAA slopes (Savla et al., 2013). Increased dailyAH may also indicate distress such that our findings arelso in line with research demonstrating that low morningaseline levels of sAA are found in those with major depres-ive disorder (Cubała and Landowski, 2014). Previous dayncreases in negative mood may be associated with lowerentral noradrenergic release the following day as seen inhose with affect-related disorders such as major depres-ion.

Our study was the first to identify direct associationsetween trait affect and components of the sAA diurnalhythm. Higher trait positive affect was associated with flat-er sAA slopes, indicating those who typically experiencedreater positive affectivity had less increases in sAA acrosshe day. This finding complements research indicating thathronic stress is associated with increases in sAA across theay (Nater et al., 2007).

Despite innovative contributions, several limitationshould be considered. Although we collected intensive indi-ators of daily experiences and sAA, we were restricted tohree days of measurement. Future studies across a greatereriod of development could uncover complexities inhanging physiological processes and adolescents’ affectivexperiences. Our sample was dependent on self-selection,redominately female and from a small geographic area,imiting the generalizability of our findings. Although we didot detect any sex differences in the diurnal course of sAAr sAA reactivity, our recruitment and sampling proceduresay have been biased toward females (e.g., participating in

daily diary study). Further, because our protocol was orig-nally designed to collect salivary cortisol we did not assessr adjust for salivary flow rate, which has been associatedith estimates of sAA (Bosch et al., 2011). Some research

uggests these biases are less likely to occur when passiverool is used to collect saliva, which was the method utilizedn this study (Rohleder et al., 2006; DeCaro, 2008). Finally,ur momentary indicators of affective states reflected par-icipants’ experiences over the previous hour. Given thatNS activity can respond over a relatively short time frame

Granger et al., 2007), it is possible we missed momentaryeactivity to affective states for some cases.

The contrasts between momentary, daily, and trait affectnd sAA patterns are notable and support Adam’s (2012)

Mlwa

L.D. Doane, S.A. Van Lenten

hronometric model of emotion-stress physiology transac-ions. These contrasts further suggest there are differentialelations depending on the marker examined (e.g., cortisol,AA). For example, studies examining naturalistic momen-ary physiological reactivity have demonstrated that higherortisol levels are associated with momentary negative andositive affect (e.g., Jacobs et al., 2007; Matias et al.,011). We found that only high arousal positive affect wasssociated with momentary sAA reactivity. Further, nega-ive affect has been associated with greater increases inhe cortisol awakening response the next day (e.g., Doanend Adam, 2010). In this study we found similar patternsor sAA, whereby high arousal negative affect one day wasssociated with greater AARs and flatter slopes the next day.inally, we uncovered differences in sAA diurnal patternselated to trait positive affect. In contrast, both trait nega-ive and positive affect have been related with componentsf the cortisol diurnal rhythm (e.g., greater positive affec-ivity and steeper diurnal slopes; Hoyt et al., submitted forublication; Polk et al., 2005).

Importantly, our findings also suggest that both arousalnd valence should be accounted for when examining differ-nces in sAA reactivity and diurnal patterns. We hypothesizehat sustained negatively-valenced arousing emotions mightrolong physiological sensitivity to arousal (e.g., more pro-ounced AAR the next day) as compared to momentaryositive emotions. In contrast, trait-level positive emotion-lity may reflect a dimension of personality not relatedo transient changes in affect and sAA. Those who expe-ience more positive emotions in general might possess

global protective characteristic that helps to modulatemotion-stress physiology transactions. In sum, our findingsuggest that emotion-sAA transactions occur over vary-ng time courses in the daily lives of adolescents, anduture research should consider accounting for both affec-ive arousal and valence in the examination of these links.

ole of funding sources

his research was partially supported by the Institute forocial Science Research at Arizona State University (L.D.D.,rinciple Investigator). The grant agency had no further rolen the study design, data collection, analysis and interpreta-ion of the data, in the writing of manuscript or the decisiono submit the article for publication.

onflicts of interest

one of the authors (L.D.D., S.A.V.) have any conflicts ofnterest to declare with respect to this manuscript.

cknowledgements

he authors would like to thank the participants of the ASUransition to College Study for the time and effort they

ichael R. Sladek and Reagan Styles for comments on ear-ier drafts of this manuscript. This research was conductedith the support of the Institute for Social Science Researcht Arizona State University (L.D.D., Principal Investigator).

ons o

L

L

M

M

N

N

N

O

P

R

R

S

S

S

T

T

Multiple time courses of salivary alpha-amylase and dimensi

References

Adam, E.K., 2012. Emotion-cortisol transactions occur over multipletime scales in development: implications for research on emo-tion and the development of emotional disorders. Monogr. Soc.Res. Child Dev. 77, 17—27.

Adam, E.K., Till Hoyt, L., Granger, D.A., 2011. Diurnal alpha amylasepatterns in adolescents: associations with puberty and momen-tary mood states. Biol. Psychol. 88, 170—173.

Bosch, J.A., Brand, H.S., Ligtenberg, T.J., Bermond, B., Hoog-straten, J., Nieuw Amerongen, A.V., 1996. Psychological stressas a determinant of protein levels and salivary-induced aggre-gation of Streptococcus gordonii in human whole saliva.Psychosom. Med. 58 (4), 374—438.

Bosch, J.A., de Geus, E.J., Veerman, E.C., Hoogstraten, J., Ameron-gen, A.V.N., 2003. Innate secretory immunity in response tolaboratory stressors that evoke distinct patterns of cardiac auto-nomic activity. Psychosom. Med. 65 (2), 245—258.

Bosch, J.A., Veerman, E.C., de Geus, E.J., Proctor, G.B., 2011. �-Amylase as a reliable and convenient measure of sympatheticactivity: don’t start salivating just yet! Psychoneuroendocrinol-ogy 36 (4), 449—453.

Buchanan, T.W., Bibas, D., Adolphs, R., 2010. Salivary �-amylaselevels as a biomarker of experienced fear. Commun. Integr. Biol.3 (6), 1—3.

Chatterton Jr., R.T., Vogelsong, K.M., Lu, Y.C., Hudgens, G.A.,1997. Hormonal responses to psychological stress in men prepar-ing for skydiving. J. Clin. Endocrinol. Metab. 82 (8), 2503—2509.

Cubała, W.J., Landowski, J., 2014. Low baseline salivary alpha-amylase in drug-naïve patients with short-illness-duration firstepisode major depressive disorder. J. Affective Disord. 157,14—17.

DeCaro, J.A., 2008. Methodological considerations in the use of sali-vary �-amylase as a stress marker in field research. AmericanJournal of Human Biology 20 (5), 617—619.

Doane, L.D., Adam, E.K., 2010. Loneliness and cortisol: momentary,day-to-day, and trait associations. Psychoneuroendocrinology35, 430—441.

Doane, L.D., Zeiders, K.H., 2014. Contextual moderators of momen-tary cortisol and negative affect in adolescents’ daily lives. J.Adolesc. Health. 54 (5), 536—542.

Evans, D.E., Rothbart, M.K., 2007. Developing a model for adulttemperament. J. Res. Pers. 41, 868—888.

Gordis, E.B., Granger, D.A., Susman, E.J., Trickett, P.K., 2006.Asymmetry between salivary cortisol and �-amylase reactivityto stress: relation to aggressive behavior in adolescents. Psy-choneuroendocrinology 31 (8), 976—987.

Granger, D.A., Kivlighan, K.T., El-Sheikh, M., Gordis, E.B., Stroud,L.R., 2007. Salivary �-amylase in biobehavioral research. Ann.N. Y. Acad. Sci. 1098, 122—144.

Hoyt, L.T., Craske, M.G., Mineka, S., Adam, E.K., in press. Adoles-cent cortisol diurnal rhythms: Cross-sectional and longitudinalassociations with high arousal positive affect (submitted for pub-lication).

Jacobs, N., Myin-Germeys, I., Derom, C., Delespaul, P., Van Os, J.,Nicolson, N.A., 2007. A momentary assessment study of the rela-tionship between affective and adrenocortical stress responsesin daily life. Biol Psychol. 74 (1), 60—66.

W

f affect in adolescence 53

arson, R.W., Moneta, G., Richards, M.H., Wilson, S., 2003. Conti-nuity, stability, and change in daily emotional experience acrossadolescence. Child Dev. 73, 1151—1165.

orentz, K., Gütschow, B., Renner, F., 1999. Evaluation of a direct�-amylase assay using 2-chloro-4-nitrophenyl-�-D-maltotrioside.Clin. Chem. Lab Med. 37, 1053—1062.

atias, G.P., Nicolson, N.A., Freire, T., 2011. Solitude and cortisol:associations with state and trait affect in daily life. Biol. Psychol.86 (3), 314—319.

eerlo, P., Sgoifo, A., Suchecki, D., 2008. Restricted and disruptedsleep: effects on autonomic function, neuroendocrine stress sys-tems and stress responsivity. Sleep Med Rev. 12 (3), 197—210.

ater, U.M., Rohleder, N., 2009. Salivary alpha-amylase as a non-invasive biomarker for the sympathetic nervous system: currentstate of research. Psychoneuroendocrinology 34, 486—496.

ater, U.M., Rohleder, N., Schlotz, W., Ehlert, U., Kirschbaum,C., 2007. Determinants of the diurnal course of salivary alpha-amylase. Psychoneuroendocrinology 32, 392—401.

oto, Y., Sato, T., Kudo, M., Kurata, K., Hirota, K., 2005. The rela-tionship between salivary biomarkers and state-trait anxietyinventory score under mental arithmetic stress: a pilot study.Anesth. Analg. 101 (6), 1873—1876.

ut, D., Granger, D.A., Sephton, S.E., Segerstrom, S.C., 2013. Dis-entangling sources of individual differences in diurnal salivary�-amylase: reliability, stability and sensitivity to context. Psy-choneuroendocrinology 38, 367—375.

olk, D.E., Cohen, S., Doyle, W.J., Skoner, D.P., Kirschbaum, C.,2005. State and trait affect as predictors of salivary cortisol inhealthy adults. Psychoneuroendocrinology 30 (3), 261—272.

ohleder, N., Nater, U.M., Wolf, J.M., Ehlert, U., Kirschbaum, C.,2004. Psychosocial stress-induced activation of salivary alpha-amylase: an indicator of sympathetic activity? Ann. N. Y. Acad.Sci. 1032, 258—263.

ohleder, N., Wolf, J.M., Maldonado, E.F., Kirschbaum, C., 2006.The psychosocial stress-induced increase in salivary alpha-amylase is independent of saliva flow rate. Psychophysiology 43(6), 645—652.

ánchez-Navarro, J.P., Maldonado, E.F., Martínez-Selva, J.M.,Enguix, A., Ortiz, C., 2012. Salivary alpha-amylase changespromoted by sustained exposure to affective pictures. Psy-chophysiology 49 (12), 1601—1609.

avla, J., Granger, D.A., Roberto, K.A., Davey, A., Blieszner, R.,Gwazdauskas, F., 2013. Cortisol, alpha amylase, and daily stress-ors in spouses of persons with mild cognitive impairment.Psychol. Aging 28, 666.

egal, S.K., Cahill, L., 2009. Endogenous noradrenergic activationand memory for emotional material in men and women. Psy-choneuroendocrinology 34 (9), 1263—1271.

akai, N., Yamaguchi, M., Aragaki, T., Eto, K., Uchihashi, K.,Nishikawa, Y., 2004. Effect of psychological stress on the salivarycortisol and amylase levels in healthy young adults. Archives ofOral Biology 49 (12), 963—968.

homa, M.V., Joksimovic, L., Kirschbaum, C., Wolf, J.M., Rohleder,N., 2012. Altered salivary alpha-amylase awakening response inBosnian War refugees with posttraumatic stress disorder. Psy-

choneuroendocrinology 37, 810—817.

atson, D., Clark, L.A., Tellegen, A., 1988. Development and vali-dation of brief measures of positive and negative affect: thePANAS scales. J. Pers. Soc. Psychol. 54, 1063.