Saliva pH as a biomarker of exam stress and a predictor of exam performance

Miri Cohen, Rabia Khalaila

PII: S0022-3999(14)00253-0
DOI: doi: 10.1016/j.jpsychores.2014.07.003
Reference: PSR 8841

To appear in: Journal of Psychosomatic Research

Received date: 16 April 2014

Revised date: 1 July 2014
Accepted date: 4 July 2014

Please cite this article as: Cohen Miri, Khalaila Rabia, Saliva pH as a biomarker of exam
stress and a predictor of exam performance, Journal of Psychosomatic Research (2014),
doi: 10.1016/j.jpsychores.2014.07.003

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
 ACCEPTED MANUSCRIPT

Saliva pH as a biomarker of exam stress and a predictor of exam performance

Miri Cohena and Rabia Khalailab

PT
a
School of Social Work, Faculty of Social Welfare and Health Sciences,

University of Haifa, Haifa, Israel

RI
b
School of Nursing, Zefat Academic College, Zefat, Israel.

SC
NU
MA
Corresponding Author:
Miri Cohen, Faculty of Social Welfare and Health Sciences,
ED

University of Haifa, Mount Carmel, Haifa 31905, Israel.

Tel: 972-4-8240565 Fax: 972-4-8240578
PT

Email: cohenm@research.haifa.ac.il
CE
AC

1
 ACCEPTED MANUSCRIPT

ABSTRACT

Background: Salivary pH is regulated by the sympathetic and parasympathetic nervous

PT
system; therefore, it may serve as a biomarker of stress.

Aims: To assess the associations between the cognitive and emotional dimensions of

RI
exam stress and pH levels, and the predictability of salivary pH in relation to test

SC
performance.

NU
Methods: A prospective study. Eighty-three nursing students answered a questionnaire

on stress appraisals, experienced stress, test anxiety (including worry and emotionality
MA
subscales) and health behaviors, and gave a saliva sample for measuring pH on the

morning of their first term exam and three months later. Their performance on the test
ED

(grades) was also recorded.

PT

Results: Levels of pH in saliva were higher (levels of acidity were lower) in the post

exam compared to the exam period, in parallel to lower threat appraisal, experienced
CE

stress, and test anxiety levels post exam. Controlling for smoking, physical activity and
AC

working hours per week, pH levels at both time points were predicted by appraised threat

regarding the exam situation, experienced stress, and the emotionality dimension of test

anxiety. pH at Time 1 predicted performance on the exams and mediated the associations

of experienced stress and emotionality subscale with test performance.

Conclusions: the present study indicates that pH levels may serve as a reliable,

accessible and inexpensive means by which to assess the degree of physiological

reactions to exams and other naturalistic stressors.

Key words; Saliva pH, academic stress, test anxiety, stress appraisals, test performance

2
 ACCEPTED MANUSCRIPT

INTRODUCTION

Stress is an inherent part of human life and penetrates almost every human experience

PT
[1]. Stressful encounters, as well as one’s psychological reactions to them, activate the

RI
hypothalamic-pituitary-adrenocortical (HPA) axis and the sympathetic nervous system,

SC
resulting in the secretion of excessive levels of stress hormones, mainly cortisol and

catecholamines [2]. Stress hormones may lead to the dysregulation of bodily functions,

NU
including the immune, cardiovascular and metabolic functions [2].
MA
The outcomes of the these effects may be measured as stress response biomarkers

[3,4]. A wide array of stress biomarkers has been proposed and empirically assessed to
ED

different extents (e.g., cortisol natural killer activity, in-vivo or in-vitro levels of pro- or

anti-inflammatory cytokines [2,4,5]. The development of new stress biomarkers was

PT

recently recognized as an important aim in stress research [3,6]. The use of salivary
CE

biomarkers, in particular, has gained increased interest over the past few years [6], due to

the richness of soluble proteins and enzymes in saliva, and the relative easiness of
AC

collecting samples [7,8].

Saliva is secreted from the three major salivary glands and from the minor salivary

glands in the oral mucosa [9]. It consists of 99.5% water and 0.5% electrolytes,

glycoproteins, enzymes and secretory antibodies, such as secretory immunoglobulin A

(sIgA). These components perform multiple roles in the process of digestion, oral health,

and initial protection against bacterial or viral factors [9]. Salivary secretion is regulated

by a reflex arc that consists of afferent receptors and nerves carrying impulses that are

induced by the actions of gustation and mastication, salivation center, and an efferent part

3
 ACCEPTED MANUSCRIPT

that consists of parasympathetic and sympathetic autonomic nerves [9]. Therefore, the

secretion of saliva, as well as its composition and functions, are controlled by the

sympathetic and parasympathetic autonomic nervous systems. It has been suggested that

PT
the parasympathetic nerves mainly regulate fluid secretion, while the sympathetic nerves

mainly regulate salivary protein secretion. However, more recent findings suggest that

RI
the two systems work together to evoke salivary secretions comprised of both fluid and

SC
protein [10].

NU
The most frequently-studied saliva biomarkers are levels of cortisol [7,11], sIgA

[12,13] and alpha-amylase [8,14]. Recently, the pH level in saliva has been suggested as
MA
a possible useful non-expensive biomarker indicating psychological stress levels [15].

Saliva’s pH represents its degree of acidity, whose balance (pH =7) is needed for the
ED

optimal functioning of its various components [9]. The regulation of saliva volume and
PT

its composition by the sympathetic and parasympathetic nervous systems under stress

may lead to a lower rate of secretion from the saliva glands in the mouth, often expressed
CE

by dry mouth (xerostomia) in reaction to stress [16]. This lowered rate of saliva secretion
AC

leads to a decrease in the bicarbonate secreted in saliva (alkaline), which results in

increased acidity and a decrease in oral pH [9]. Indeed, salivary flow rate was found to be

significantly lower during stressful periods [16].

A decrease in oral pH may be responsible for the dysregulation of other saliva

biomarkers of stress, due to its effect on dysregulation of saliva components such as

cortisol [17], sIgA [18] and alpha-amylase [18]. Therefore, pH may be an antecedent of

stress-induced dysregulation in levels of components of the saliva previously measured as

4
 ACCEPTED MANUSCRIPT

biomarkers of stress. In addition, saliva pH was found to reliably reflect skin or urinary

pH [19].

An extensive search of databases revealed that only three prior studies have assessed

PT
the associations of psychological stress and pH levels [15,19,20]. Sandin and Chorot [19]

examined changes in levels of skin, salivary, and urinary pH in reaction to academic oral

RI
examination stress in 32 women. During the stress situation, statistically significant

SC
reductions in the skin, salivary, and urinary pH levels were produced, and these were

NU
associated with anxiety levels. Morse et al. [20] showed that practicing meditation

reduced anxiety significantly, increased salivary volume, and raised salivary pH.
MA
Recently, pH levels were examined among spouses caring for cancer patients and age-

matched individuals in the community. Lower levels of pH saliva were found among
ED

caregivers of cancer patients than in the comparison group. Controlling for background
PT

variables, being a caregiver of a cancer patient, experiencing depression, and having a

lower level of perceived mastery predicted lower pH levels [14].

CE

Therefore, the use of salivary pH as a stress biomarker may constitute an answer to

AC

the need to establish non-invasive and feasible methods with which to measure the

physiological indicators of stress [3,6]. However, this subject should be studied further,

preferably using established models of human stress, such as academic exams.

Academic exams are an example of naturalistic stressors, which are time-limited and

typically perceived as aversive, and are often studied as a model of psychological and

physiological reactions to stressful encounters. Therefore, they may be a useful model for

assessing novel biomarkers. Based on cognitive theories of stress (e.g., [21]), the present

study assessed the effects of cognitive appraisals of the stress inherent in the exam

5
 ACCEPTED MANUSCRIPT

situation, test anxiety, and levels of experienced stress on saliva’s pH levels. According

to the cognitive approach, stressful encounters, such as academic exams, are subjected to

cognitive processing in which individuals appraise the threat or challenge embodied in

PT
the stressful situation, the coping resources available to cope with it [21]. These

appraisals affect the coping strategies individuals employ in response to the stress

RI
situation, and shape the psychological (e.g., experienced stress, test anxiety) and

SC
physiological outcomes of the stress encounter.

NU
Test anxiety consists of the specific cognitive, emotional and physiological reactions

evoked by the stimuli of testing, and includes cognitive aspects (i.e., worry) and
MA
emotional and physiological arousal (i.e., emotionality) components [22,23].

Previous studies reported that academic exams are related to neuroendocrine and immune
ED

alterations [15,24,25]. In addition, studies found that heightened physiological stress

PT

levels, as measured by biomarkers, are related to lower performance on exams (24,

27,28), suggesting that heightened physiological stress might influence

CE

cognitive/emotional processes related to performance. Only one study described above

AC

[19] researched the effect of exams on salivary pH and found a decrease in pH in relation

to exam stress.

Physiological responses to stress may be buffered by health behaviors and health

status [2,29,30]. Low physical activity, cigarette smoking, or poorer nutrition were found

to be related to the degree of activation of the HPA and sympathetic nervous system with

consequent effects on stress hormones, immune parameters, as well as saliva biomarkers

[2,29-31]. For example, saliva pH was lower in smokers than in non-smokers [31].

6
 ACCEPTED MANUSCRIPT

The aim of the present study was to determine whether salivary pH can be a biomarker of

stress, using stressful exams as a model for naturalistic stressors. More specifically, based

on the cognitive model of stress and coping, the study attempted to determine

PT
1. The relations between challenge and threat appraisals, experienced stress and test

anxiety and levels of pH; 2. The predictive value of pH on exam performance; and

RI
3. whether levels of salivary pH change from exam measurement to the 3 months post-

SC
exam measurement.

NU
The main hypotheses were as follows:

Hypothesis 1. a) levels of threat appraisal, experienced stress and test anxiety will
MA
be lower and b) challenge appraisal and c) pH levels will be higher at Time 2 (post

exam) compared to Time 1 (exam).

ED

Hypothesis 2. At both time points, a) threat appraisal, experienced stress and test
PT

anxiety scores will be negatively associated with pH levels, while b) challenge appraisal

will be positively associated with pH levels, c) experienced stress and test anxiety will
CE

mediate the effect of threat appraisals on pH.

AC

Hypothesis 3. pH levels at Time 1 will be negatively associated with exam

performance (exam grades) and will predict performance on the examination. d. pH

levels at Time 1will mediate the effects of experienced stress and test anxiety on exam

performance.

METHOD

Participants

The study was approved by the university ethics board. Participants included 83 students

7
 ACCEPTED MANUSCRIPT

from first and second-year nursing studies for a Bachelor degree in a Northern Israeli

college. The students who participated in the study did final exams (T1) in two major

courses: microbiology in the first year and pharmacology in the second year. These were

PT
the students’ first term semester exams (13 exams in total for the first-year students and 8

exams in total for the second-year students); the mandatory minimum passing grade was

RI
60 and 65 (out of 100), respectively. Therefore, these exams were perceived as very

SC
difficult and stressful. A total of 121 students (71 from first-year studies and 50 from

NU
second-year studies) were asked to participate. The research assistant explained the

experiment and assured students that participation would be on a voluntary basis and that
MA
confidentiality would be preserved. Eighty-three students (49 and 34, respectively)

agreed and signed an informed consent form. At T1, one hour before the beginning of the
ED

exam, the participants answered the questionnaires and gave saliva samples (after a 2-
PT

hour fast) before beginning the exam. All participants were then approached three months

later, during a non-exam period, for the T2 measure. An SMS message was sent to all
CE

students who had participated in T1, inviting them to participate in the T2 measure. The
AC

data was collected in college about an hour before the beginning of the study day, in the

morning under fasting conditions. Sixty-eight students agreed to participate in the T2 part

of the study. These students were asked to complete the questionnaire and give a saliva

sample. As in T1, the data collection took about 15 minutes for each student. Therefore,

the participation rates were 68.6% and 81.9%, respectively. T2 participants and those that

dropped out were not different in background characteristics.

Measures

Saliva pH level represents the level of acidity in the saliva. A saliva specimen given by

participants prior to the interview was deposited in a special receptacle. The test was
8
 ACCEPTED MANUSCRIPT

taken in the morning, after a minimum rest of 10 minutes. Students were instructed not to

ingest any food or drink, or smoke for at least two hours before the saliva test. After

giving the saliva sample, we offered coffee to all participants. The research assistant

PT
asked each participant to fill his or her mouth with saliva, swallow it, repeat the action

two more times consecutively, and then, on the third time, to spit into a small receptacle

RI
(a plastic container). The amount of saliva collected was between 3 to 5 ml; this amount

SC
was obtained for all participants and was sufficient for pH assessments, without adding

NU
water. A valid digital device was used, the CyberScan pH 501 made by El-Hamma

Instruments Ltd., Israel, which comprises a kit of multi-purpose sensors. The pH level
MA
measurement was carried out immediately to minimize the environmental impact on the

measurement. The sensor connected to the device was dipped into the collection
ED

receptacle with the saliva, until a beep was heard several seconds later, indicating that the
PT

measurement was completed [15]. Then, the salivary pH level displayed on the device

screen was recorded on the questionnaire form. The optimum pH measurement for saliva
CE

is between 6.5 and 7.5 maintained all day (pH=7), which means it is at a neutral level of
AC

balance between acid and alkaline in the body. A measurement of pH>7 shows alkaline

(base) saliva, while pH<7 indicates acidic saliva.

Challenge and threat appraisals were measured using a four-item scale, ranging from

1=not at all to 9=to a very high degree. The scale was created based on Folkman and

Lazarus’s conceptualization of threat and challenge appraisals [21], and was adapted to

the specific stress of academic exams, as was done in previous studies [32,33]. Threat

and challenge appraisals were not significantly associated (r = 0.09 for Time 1 and 0.08

for Time 2).

9
 ACCEPTED MANUSCRIPT

Test Anxiety [34] . The Hebrew version of the test [35] consists of 20 items that depict a

variety of cognitions and feelings that characterize test situations, rated on a 1-4 scale

(1=hardly ever; 4= almost always). The inventory has two subscales: Worry and

PT
Emotionality. The internal reliabilities (Cronbach's alphas) in the presents study at T1

were .92 for the whole scale, .83 for worry, and .81 for emotionality. At T2, the internal

RI
reliabilities were .94, .88, and .91, respectively.

SC
Experienced stress is a single-item measure probing the degree of subjective feeling of

NU
stress in the previous week on a scale ranging from 1 (no stress at all) to 9 (extreme

feelings of stress).This measure of stress was used in several previous studies and was
MA
highly correlated with different measures of emotional distress [36-38].

Exam performance is represented by the exam grades. Since participants were students
ED

attending two courses that may differ in regard to content and instructors, the exam
PT

grades were transformed into a Z-score.

Background details included personal data, religion, perceived economic status

CE

(1=extremely poor to 5=excellent), employment (whether the participant had a job and
AC

the number of hours he/she worked per week), physical activity (degree of activity per

week: 1=less than once a week, 2= once a week, 3=several times a week) and whether or

not the participant smoked, plus the number of cigarettes smoked per day.

Statistical analysis

Data was analyzed using SPSS 21. Frequencies, means and standard deviations were

calculated for background variables. Differences between Time 1 (exam) and Time 2

(post-exam) were assessed using paired t-tests. Correlations were assessed using Pearson

10
 ACCEPTED MANUSCRIPT

correlations. Multiple regression analyses were conducted to assess the associations

between study variables and pH, and between pH and exam performance (controlling for

background and study variables). pH was normally distributed. Z-scores were obtained

PT
for test performance. Demographic variables and year of study were not associated with

pH and were not entered into the regressions. Nonparametric bootstrap procedure [39]

RI
was used to examine whether experience of stress and test anxiety and mediated the

SC
relation of stress appraisals and pH and whether pH mediated the relations of experienced

NU
stress and test anxiety on performance (grades), while controlling for confounders. The

mediation effect was measured using the PROCESS program for SPSS [40]. The
MA
observed dataset was randomly resampled in 5000 samples as previously suggested [39].

The null hypothesis was rejected when the 95% confidence interval did not include zero.
ED

A significance level of p<.05 was used.

PT

RESULTS
CE

Background characteristics
AC

Table 1 describes participants’ background characteristics. The mean age was about 22

years, mostly non-married; women comprised more than 60% of the sample. Only 22%

were employed, most held part-time jobs and working hours per week ranged from 0 to

24. The family income was moderate. Participants reported, on average, a high frequency

of physical activity. Most students were non-smokers and the average number of

cigarettes per day for the smokers was relatively low.

Differences in study variables: exam and post-exam

11
 ACCEPTED MANUSCRIPT

Table 2 presents the means (SDs) and differences between Time 1 (exam) and Time 2

(post exam) study variables. For both time point measures, the means for threat were

about average (scale range 1-9), while the mean for challenge appraisal was higher. Level

PT
of experienced stress related to exams was also high (scale range 1-9). Test anxiety sub-

scales and total scores were at the center of the scale (scale range 1-4). pH levels were in

RI
the normal range of 6.5-7.5. In accordance with Hypothesis 1a, degree of threat appraisal,

SC
experienced stress and total text anxiety score (but not each of the subscales of worry and

NU
emotionality) were significantly lower at Time 2 (post-exam) compared to Time 1

(exam). The difference between Time 1 and Time 2 in terms of challenge appraisal
MA
scores, were not significant; thus, Hypothesis 1b was not confirmed. According to

Hypothesis 1c, levels of pH were significantly higher (acidity lower) at Time 2 compared
ED

to Time 1.
PT
CE

Associations between appraisals, experienced stress, test anxiety and pH

Hypothesis 2 regarding the effects of appraisals, experienced stress and test anxiety on
AC

levels of pH were assessed using correlations and multiple regressions. At Time 1, pH

levels were negatively associated with threat appraisal (r = -.38, p<.001) and with

experienced stress (r = -.49, p<.001); meaning, the higher the threat appraisal and the

experienced stress, the lower the pH (the higher the saliva acidity). Levels of pH were not

significantly associated with challenge appraisals (r =.09, p>.05), test anxiety total score

(r = -.13, p> .05), worry subscale scores (r = -.18, p> .05), and emotionality subscale

scores (r = -.19, p>.05). At Time 2, pH levels remained negatively and significantly

associated with threat appraisal (r = -.48, p<.001) and experienced stress (r = -.34, p<.01),

12
 ACCEPTED MANUSCRIPT

with no significant associations with challenge appraisals (r = -.04, p>.05), test anxiety

total and subscale scores (r = .05 to 01, p>.05).

Four separate models of regression analyses were conducted to further assess

PT
Hypothesis 2. For all four models, none of the demographic variables was associated with

pH levels; hence, they were not entered into the analyses, whereas amount of working

RI
hours per week and health behaviors related to smoking and physical activity were

SC
controlled for. To assess the predicting effect of experienced stress on pH at Time 1

NU
(Model 1), pH level at Time 1 was regressed on health behaviors, threat appraisal and

experienced stress. Number of cigarettes smoked per day and the level of experienced
MA
stress were negatively and significantly associated with pH; namely, higher smoking and

higher experienced stress were associated with lower pH (higher acidity). It is important
ED

to note that threat appraisal was strongly associated with pH before experienced stress
PT

was entered, suggesting a possible mediation effect.

To assess the predicting effect of test anxiety on pH at Time 1 (Model 2), pH at Time
CE

1 was regressed on worry and emotionality subscales (with working hours, health
AC

behaviors and threat appraisal within the regression). Smoking, threat appraisal and

emotionality, but not worry, were significantly associated with pH levels.

To assess the predicting effects of experienced stress and test anxiety on pH at Time

2 (Models 3 and 4), pH at Time 2 was regressed separately on threat appraisal, pH at

Time 1 and either experienced stress or test anxiety subscales, controlling for working

hours and health behaviors. pH was found to be associated with smoking, threat appraisal

and experienced stress, while the association with text anxiety subscales was not

significant.

13
 ACCEPTED MANUSCRIPT

Hypothesis 2 regarding the effects of appraisals, experienced stress and test anxiety

on levels of pH was partially confirmed: pH levels at both time points were predicted by

threat appraisals and experienced stress. Regarding test anxiety, emotionality only

PT
predicted pH levels at Time 1. Challenge appraisal was not significantly associated with

pH.

RI
SC
The mediation effect of experienced stress and emotionality on the associations

between threat appraisal and pH were assessed using the parallel multiple mediation

NU
model [61; 62] (mediation effects were not assessed for challenge appraisal and worry,
MA
due to lack of associations with pH). The analysis for Time 1 showed that while

controlling for working hours, smoking and physical activity, threat appraisal had an
ED

indirect effect on pH, mediated by experienced stress and emotionality, (B=-.07, Boot SE

.03, 95%CI -.12 - -.03 for experienced stress; B=-.03, Boot SE .02, 95%CI -.08 - -.05 for
PT

emotionality). In the Time 2 analysis, experienced stress mediated the relation between
CE

threat appraisal and pH at Time 2 (B=-.04, Boot SE .02, 95%CI -.10 - -.01 for

experienced stress). Therefore, Hypothesis 2c was partially confirmed.

AC

pH and test performance

The range of grades was from 30 to 100; the mean grade was 69; and the median grade

was 72.

At Time 1, pH levels were positively and significantly correlated with test performance

(r=30, p<.01); meaning, the higher the pH, the higher the exam grades. At Time 2,

associations were not significant (r=.15, p>.05). Controlling for age, first-year or second-

year studies, working hours, threat appraisal and experienced stress, pH significantly

14
 ACCEPTED MANUSCRIPT

predicted exam performance ( =.31, p<.05, R2=.20, F(5, 81)=3.98, p<.05). The same

analysis was conducted with test anxiety instead of experienced stress, and pH once again

predicted exam performance ( =.35, p<.01, R2=.20, F(5, 81)=3.90, p<.01).

PT
Finally, the mediation effect of pH (Time 1) on the associations of experienced stress

RI
and emotionality with exam performance was assessed using the simple mediation model

SC
[62]. pH was found to mediate the associations of experienced stress (B=-.32, Boot SE

.12, 95%CI -.12, -.03) and emotionality (B=-.08, Boot SE .04, 95%CI -.57, -.10) with

NU
performance. Thus, Hypothesis 3 regarding the predicting effect of pH (Time 1) on

performance and its mediating effect between experienced stress and emotionality (but
MA
not worry) was confirmed.
ED
PT

DISCUSSION
CE

The present results show that pH levels in saliva were higher (levels of acidity were

lower), at Time 2 compared to Time 1. This difference was in accordance with the
AC

decrease in threat appraisal, experienced stress and test anxiety levels from the exam

period to the post-exam period. Controlling for smoking, physical activity and working

hours per week, pH levels at both time points were predicted by levels of appraised threat

of the exam situation and experienced stress, while the emotionality dimension of the

stress exam predicted pH at the exam only. Moreover, controlling for background and

study variables, pH predicted performance on the exams.

The present study supports a very recent study [15] and several previous ones [19,20]

that show a dose-response relation between experienced stress and pH: the higher the

15
 ACCEPTED MANUSCRIPT

stress experienced by the participants, the lower the pH. Moreover, stress levels

decreased and pH levels increased at Time 2, the non-exams period, but still remained

related in a dose-response manner. The same pattern of relations was found for other

PT
neuro-immune factors in studies that assessed students during exam periods and post-

exam periods, such as saliva cortisol (e.g., [25,41],) or various immune cellular functions

RI
or secretion of cytokines [25,42], blood pressure, heart rate or heart rate variability [43].

SC
Levels of pH, which represent saliva acidity, decrease in stress situations, due to

NU
activation of the HPA axis and the sympathetic nervous system in reaction to stressful

experience [2]. We suggest that changes in pH levels may be an antecedent to the

MA
dysregulation of other saliva components in reaction to stressors, such as cortisol [7,11],

sIgA or alpha-amylase [18]. Therefore, the present study provides additional support,
ED

indicating that pH levels may serve as a reliable, accessible and inexpensive means by
PT

which to assess the degree of physiological reactions to stressors.

Levels of test anxiety were high at the exam and low during the non-exam period, as
CE

was reported by previous studies [28]. However, only the emotionality scale, but not the
AC

worry scale, was associated with pH. Worry and emotionality responses were theorized to

be conceptually independent [22,23] and these two components were found to be aroused

and maintained by different situational conditions [44]. During an academic exam,

individuals scoring higher on the emotionality component may be more prone to

activation of the nervous system. It may be that emotionality and the perceived

experience of stress constitute a similar construct, and both provide an indication of level

of arousal. In contrast, the worry scale consists of thoughts and worries that may have

different effects on the individual’s stress reactions. Some worries may be related to

16
 ACCEPTED MANUSCRIPT

appraisals that constitute more of a challenge than a threat, activating the nervous system

in a different way. However, other studies found that the worry dimension, rather than the

emotionality dimension, was associated with various physiological markers, such as

PT
cardiovascular responses [43].

According to the psychological models of stress, threat and challenge appraisals have

RI
a major effect on the degree of stress experienced by individuals and their emotional

SC
reactions to the stress situation. In the past, the effects of these appraisals have often been

NU
assessed in relation to psychological outcomes of coping with various stressors [5,33],

including academic exams [45], but not in relation to physiological markers of exam-
MA
related stress. The present findings indicate that threat appraisals, but not challenge

appraisals, were associated with lower pH. Previous studies indicated that threat
ED

appraisals were consistently related to activation of the HPA axis and sympathetic
PT

nervous system. In contrast, challenge appraisals were found to buffer the effects of stress

on emotional and physiological reactions, but were also associated with short-term
CE

activation of physiological stress markers [46]. Perhaps the consequences of challenge

AC

appraisal differ among students, reducing the physiological stress reaction for some,

while for others serving to stimulate the stress reaction [46].

In accordance with the cognitive approach [21], the effect of threat appraisal on the

physiological measure of pH was mediated by the psychological experience of stress,

expressed in the variables of experienced stress and emotionality. These results support

the approach that appraisals have a substantial effect on health, but mainly indirectly

through the effect of emotional reactions, which activate physiological stress responses

[5,37].

17
 ACCEPTED MANUSCRIPT

pH levels were a significant predictor of exam performance, and mediated the

associations between the experienced stress and test anxiety with performance. These

results add to a paucity of previous studies assessing associations between biomarkers

PT
and performance in exams (e.g., [27,28,47]. These results may show that stress and

anxiety-related physiological alterations serve to worsen performance [48]. However, it

RI
may simply be that academically weak students experience more stress during an exam

SC
than their stronger counterparts.

NU
Amount of working hours per week was not related to pH levels. Perhaps this was

due to the fact that the mean working hours per week were low and only a few
MA
participants reported working more than 10 hours per week. In addition, it might be that

many of the students take days off work during exam periods; therefore, work stressors
ED

did not affect stress and of pH levels.

PT

Two main health behaviors were examined in the present study: cigarette smoking

and physical activity. A higher number of cigarettes smoked per day was found to be
CE

associated with lower pH at both time points. This result supports a previous study on the
AC

relations between smoking and saliva pH [31] and a large volume of previous studies that

reported increased stress hormone levels and an increase in pro-inflammatory markers in

smokers, as compared to non-smoking individuals, during stressful encounters (e.g.,

[30,49].

In contrast to smoking, frequency of engaging in physical activity was non-

significantly associated with pH during exam periods, and was not associated with pH

during the post-exam period. However, the trend towards association between frequency

of physical activity and pH may suggest that during periods of high stress, physical

18
 ACCEPTED MANUSCRIPT

activity serves as a buffer, and has a lower effect during a more relaxed time period.

Indeed, previous studies showed firm associations between physical activity, levels of

stress hormones or markers of inflammation [29].

PT
While exams and other naturalistic stressors are unavoidable in the modern era, the

present study suggests that the harmful effects of stress can be prevented through

RI
different pathways: by helping individuals reframe their appraisals of a stressor as a

SC
threat, through consultation and the teaching of cognitive means of reappraisal; by

NU
teaching students how to reduce anxiety or feelings of stress by means of peer or

professional support, or relaxation training [50]; by buffering the negative effects of

MA
stress on health through encouraging students to engage in healthy behaviors.

The strengths of the current study are its prospective design and assessment of
ED

performance as an outcome variable. Scholars often stress the need to use salient
PT

measures less susceptible to research bias than questionnaires [51]. The study’s main

limitations are the relatively small number of participants, which limit generalization of
CE

the results. In addition, more frequent assessments of pH could provide additional

AC

information about the relations between pH and exam stress. Another main limitation is

the single-point measurement of pH in saliva. Since pH levels change throughout the day,

especially in reaction to various stimuli and a subsequent flow index, variance of pH

value might be assumed. Although an effort was made in this study to control for pH

variability by conducting all assessments between 8-10 am, after at least two hours of

fasting and not smoking, it is possible that conducting several assessments during one day

or on several consecutive days could better control for the effects of pH variability. In

addition, further studies are needed to control for the effect of saliva flow rate [10].

19
 ACCEPTED MANUSCRIPT

In spite of these limitations, the study contributes to the existing knowledge on stress

biomarkers. It provides further evidence that pH is a reliable and sensitive stress

biomarker, using a naturalistic stressor - academic exams. Moreover, saliva pH is directly

PT
affected by activation of the sympathetic and parasympathetic nervous systems and may

affect levels of other saliva biomarkers, such as a-amylase. Therefore, it may be a reliable

RI
representative of nervous system activation and its effects. Thus, it is suggested that

SC
studies aimed at expanding the understanding of the various effects of psycho-social

NU
factors and processes on physiological responses to stress, can use and benefit from this

feasible and non-invasive measure.

MA
ED
PT
CE
AC

20
 ACCEPTED MANUSCRIPT

REFERENCES

1. Selye H. The stress of life. New York: McGraw-Hill; 1956.

PT
2. Kiecolt-Glaser JK, McGuire L, Robles TF, Glaser R. Psychoneuroimmunology and

psychosomatic medicine: back to the future. Psychosom Med 2002;64:15-28.

RI
3. Macaluso M, Preskorn SH. How biomarkers will change psychiatry: From clinical trials to

SC
practice. Part I: Introduction. J Psychiatr Pract 2012;18:118-21.

NU
4. Cohen M, Meir T, Klein E, Volpin G, Assaf M, Pollack S. Cytokine levels as potential

biomarkers for predicting the development of posttraumatic stress symptoms in casualties

MA
of accidents. Int J Psychiatr Med 2011;42:117–31.

5. Cohen M, Klein E, Kuten A, Fried G, Pollack S. Increased emotional distress in daughters

ED

of breast cancer patients associated with decreased natural cytotoxic activity, elevated

levels of stress hormones and decreased secretion of Th1 cytokines. Int J Cancer
PT

2002;100:347-54.
CE

6. Marques AH, Silverman MN, Sternberg EM. Evaluation of stress systems by applying
AC

noninvasive methodologies: measurements of neuroimmune biomarkers in the sweat,

heart rate variability and salivary cortisol. Neuroimmunomodulation 2010;17:205-8.

7. Hellhammer DH, Wust S, Kudielka BM. Salivary cortisol as a biomarker in stress research.

Psychoneuroendocrinology 2009;34,163-71.

8. Nater UM, Rohleder N. Salivary alpha-amylase as a non-invasive biomarker for the

sympathetic nervous system: current state of research. Psychoneuroendocrinology

2009;34:486-96.

9. Humphrey SP, Williamson RT, A review of saliva: normal composition, flow, and

function. J Prosthet Dent 2001;85:162-9.

21
 ACCEPTED MANUSCRIPT

10. Bosch JA, Veerman EC, de Geus EJ, Proctor GB. a-Amylase as a reliable and

convenient measure of sympathetic activity: Don’t start salivating just yet!

Psychoneuroendocrinology 2011;36:449-53.

PT
11. Levine A, Zagoory-Sharon O, Feldman R, Lewis JG, Weller A. Measuring cortisol in

human psychobiological studies. Physiol Behav 2007;90:43-53.

RI
12. Teeuw W, Bosch JA, Veerman EC, Amerongen AV. Neuroendocrine regulation of

SC
salivary IgA synthesis and secretion: implications for oral health. Biol Chem

NU
2004;385:1137-46.

13. Bosch JA, de Geus EJ, Ring C, Nieuw Amerongen AV, Stowell JR. Academic
MA
examinations and immunity: Academic stress or examination stress? Psychosom Med

2004;66:625-6.
ED

14. Nater UM, Rohleder N, Gaab J, Berger S, Jud A, Kirschbaum C, Ehlert U. Human
PT

salivary alpha-amylase reactivity in a psychosocial stress paradigm. Int J Psychophysiol

2005;55:333-42.
CE

15. Khalaila R, Cohen M, Zidan J. Is salivary pH a marker of depression among older spousal
AC

caregivers for cancer patients? Behav Med 2013 Nov 21. [Epub in press].

16. Matos-Gomes N, Katsurayama M, Makimoto FH, Santana LLO, Paredes-Garcia E,

Becker MA, et al. Psychological stress and its influence on salivary flow rate, total

protein concentration and IgA, IgG and IgM titers. Neuroimmunomodulation

2010;17:396-404.

17. Golden SH, Wand GS, Malhotra S, Kamel I, Horton K. Reliability of hypothalamic–

pituitary–adrenal axis assessment methods for use in population-based studies. Eur J

Epidemiol 2011;26:511-25.

22
 ACCEPTED MANUSCRIPT

18. Papacosta E, Nassis GP. Saliva as a tool for monitoring steroid, peptide and immune

markers in sport and exercise science. J Sci Med Sport 2011;14:424-34.

19. Sandin B, Chorot P. Changes in skin, salivary, and urinary pH as indicators of anxiety

PT
level in humans. Psychophysiology 1985;22:226-30.

20. Morse DR, Schacterle GR, Furst ML, Esposito JV, Zaydenburg M. Stress, relaxation and

RI
saliva: relationship to dental caries and its prevention with a literature review. Ann Dent

SC
1983;42:47–54.

NU
21. Lazarus R, Folkman S. Stress, appraisal and coping. New York: Springer; 1984.

22. Spielberger CD, Vagg PR. Test anxiety: A transactional process model. In Spielberger
MA
CD, Vagg PR, eds. Test anxiety: Theory, assessment, and treatment. Washington: Taylor

and Francis; 1995:3-14.

ED

23. Liebert RM, Morris LW. Cognitive and emotional components of test anxiety: A
PT

distinction and some initial data. Psychol Rep 1967;20:975-8.

24. Lindahl M, Theorell T, Lindblad F. Test performance and self-esteem in relation to

CE

experienced stress in Swedish sixth and ninth graders—saliva cortisol levels and
AC

psychological reactions to demands. Acta Paediatr 2005; 94:489-95.

25. Lacey K, Zaharia MD, Griffiths J, Ravindran AV, Merali Z, Anisman H. A prospective

study of neuroendocrine and immune alterations associated with the stress of an oral

academic examination among graduate students. Psychoneuroendocrinology

2000;25:339–56.

26. Cohen M, Ben-Zur H, Rosenfeld M. Sense of coherence, coping strategies, and test

anxiety, as predictors of test performance. Int J Stress Manag 2008;15:289-303.

23
 ACCEPTED MANUSCRIPT

27. Sarid O, Anson O, Yaari A, Margalith M. Academic stress, immunological reaction, and

academic performance among students of nursing and physiotherapy. Res Nurs Health

2004;27:370-7.

PT
28. Ng V, Koh D, Chia SE. Examination stress, salivary cortisol, and academic performance.

Psychol Rep 2003;93(3 Pt 2):1133-4.

RI
29. Hamer M. Psychosocial stress and cardiovascular disease risk: the role of physical

SC
activity. Psychosom Med 2012;74:896-903.

NU
30. Cohen M, Granger S, Fuller-Thomson E. The association between bereavement and

biomarkers of inflammation. Behav Med 2013 Nov 22. [Epub ahead of print]
MA
31. Parvinen T. Stimulated salivary flow rate, pH and lactobacillus and yeast concentrations

in non-smokers and smokers. Scand J Dent Res 1984;92:315-8.

ED

32. Rosenfeld, M. The relationship between sense of coherence, socio-demographic variables,

PT

coping strategies, test anxiety and test performance among students. Haifa: University of

Haifa; 2004. (Hebrew).

CE

33. Kimron L, Cohen M. Coping and emotional distress during acute hospitalization in older
AC

persons with earlier trauma: the case of Holocaust survivors. Qual Life Res 2012;21:783-

94.

34. Spielberger CD, Gonzalez HP, Taylor CJ, Anton WD, Algaze B, Ross GK,

Westberry LG. Test Anxiety Inventory. Palo Alto, CA: Consulting Psychologists

Press; 1980.

35. Zeidner M. Coping with examination stress: Resources, strategies, outcomes. Anxiety

Stress Coping 1995;8:279-98.

24
 ACCEPTED MANUSCRIPT

36. Van Lankveld W, Naring G, van der Staak C, van Pad Bosch P, van de Putte L. Stress

caused by rheumatoid arthritis: Relation among subjective stressors of the disease,

disease status, and well-being. J Behav Med 1993;16:309-21.

PT
37. Cohen M, Pollack S. Mothers with breast cancer and their adult daughters: The

relationship between mothers’ reaction to breast cancer and their daughters’ emotional

RI
and neuroimmune status. Psychosom Med 2005;67:64-71.

SC
38. Cohen M. First-degree relatives of breast-cancer patients: cognitive perceptions, coping

NU
and adherence to breast self-examination. Behav Med 2002;28:15-22.

39. Preacher, K. J., & Hayes, A. F. Asymptotic and resampling strategies for assessing
MA
and comparing indirect effects in multiple mediator models. Behav Res Methods

2008;40:879–891.
ED

40. Hayes, A.F. (2012). PROCESS [Macro]. Retrieved from http://afhayes.com/

PT

introduction-to-mediation-moderation-and-conditional-process-analysis.html.

41. Preuss D, Schoofs D, Schlotz W, Wolf OT. The stressed student: Influence of written
CE

examinations and oral presentations on salivary cortisol concentrations in university

AC

students. Stress 2010;13:221–229.

42. Kamezaki Y, Katsuura S, Kuwano Y, Tanahashi T, Rokutan K. Circulating cytokine

signatures in healthy medical students exposed to academic examination stress.

Psychophysiology 2012;49:991-7.

43. Conley KM, Lehman BJ. Test anxiety and cardiovascular responses to daily academic

stressors. Stress Health 2012;28:41-50.

44. Zhang N, Henderson CN. Test anxiety and academic performance in chiropractic

students. J Chiropr Educ 2013 Dec 18. [Epub ahead of print]

25
 ACCEPTED MANUSCRIPT

45. Skinner N, Brewer N. The dynamics of threat and challenge appraisals prior to stressful

achievement events. J Pers Soc Psychol 2002;83:678-92.

46. Olff M, Langeland W, Gersons BP. Effects of appraisal and coping on the neuroendocrine

PT
response to extreme stress. Neurosci Biobehav Rev 2005;29:457-67.

47. Keller PS, El-Sheikh M, Granger DA, Buckhalt JA. Interactions between salivary cortisol

RI
and alpha-amylase as predictors of children's cognitive functioning and academic

SC
performance. Physiol Behav 2012;105:987-95.

NU
48. Bardi M, Koone T, Mewaldt S, O'Connor K. Behavioral and physiological correlates of

stress related to examination performance in college chemistry students. Stress

MA
2011;14:557-66.

49. Badrick E, Kirschbaum C, Kumari M. The relationship between smoking status and
ED

cortisol secretion. J Clin Endocrinol Metab 2007;92:819-24.

PT

50. Yahav R, Cohen M. Evaluation of a cognitive-behavioral intervention for adolescents. Int

J Stress Manag 2008;15:173-88.

CE

51. Podsakoff PM, MacKenzie SB, Lee JY, Podsakoff NP. Common method biases in
AC

behavioral research: a critical review of the literature and recommended remedies. J Appl

Psychol 2003;88:879-903.

26
 ACCEPTED MANUSCRIPT

Table 1: Background characteristics of participants

Age, years (M, SD, range) 21.9 3.0 19-36

PT
Gender (N, %)

Male 32 38.6

RI
Female 51 61.4

SC
Familial status (N, %)

Married 6 7.2

NU
Not married 77 92.8
MA
Amount of working hours per week 2.20 5.4 0-24

Perceived economic status (M, SD) 3.2 0.7 1-5

Religion (N, %)
ED

Jewish 25 30.1
PT

Muslim 51 61.4

Druze 5 6.0
CE

Other 2 2.4

Frequency of physical activity (M, SD) 2.58 0.73 1-3

AC

Number of cigarettes smoked per day 3.45 7.83 0-40

(M, SD)
1
amount of hours per week ranged from 4 to 24

27
 ACCEPTED MANUSCRIPT

Table 2: Means (SDs) and range of study variables

T1 T2

PT
M SD M SD t (67)

Threat appraisal 5.93 1.56 4.70 1.61 5.55**

RI
Challenge appraisal 7.10 1.12 6.95 0.89 0.98

SC
Experienced stress 6.01 1.89 5.72 1.37 2.05*

Worry 2.28 0.70 2.18 0.75 1.17

NU
Emotionality 2.59 MA 0.68 2.45 0.74 1.70

Test anxiety (total score) 2.53 0.59 2.32 0.63 2.50*

pH 6.95 0.60 7.41 0.74 -4.33**

Test performance 69.33 14.11

ED

*p<.05, **p<.001
PT
CE
AC

28
 ACCEPTED MANUSCRIPT

Table 3: The associations of pH levels with threat appraisal, experienced stress, and test anxiety at

Time 1 and Time 2

PT
Model 1 Model 2 Model 3 Model 4
pH (Time 1) pH (Time 1) pH (Time 2) pH (Time 2)
experienced stress test anxiety as experienced test exam as

RI
as predictor predictor stress as predictor
predictor

SC
Working hours per week -.13 -.14 -.08 -.03

Physical activity .18 .19 .03 .06

NU
Smoking -.25** -.23* -.24* -.25*

Threat appraisal (Time 2) -.10 -.27* -.43*** -.57***

MA
Experienced stress (Time -.40*** __ -.24* __

2)
ED

Worry (Time 2) __ .11 __ -.11

Emotionality (Time 2) __ -.33* __ -.08

PT

pH (Time 1) __ __ -.07 .08

CE

R2 (Adjusted R2) .40 (.36) .33 (27) .32 (.25) .35 (.27)

F (df) 9.46 (74)*** 5.67 (74)*** 4.50 (60) 4.39** (60)

AC

29
 ACCEPTED MANUSCRIPT

Highlight

Salivary pH is regulated by the sympathetic and parasympathetic nervous system

Perceived threat, experienced stress, and emotionality predicted pH levels

PT
pH at pre-exam predicted performance on the exam

RI
pH mediated the association between stress measures and performance on the exam

SC
pH levels may serve as a reliable and accessible biomarker of stress

NU
MA
ED
PT
CE
AC

30