Table 1.

Summary of Studies

Citation	Sample characteristics	Factors controlled	Stressor type	Sampling timeline	Collection method / assay	Results by biomarker	Limitations
Filaire et al., 2010	N = 9 university professors (M = 12.1 ± 1.3 years of experience in teaching and lecturing) (males: n = 7; age: M = 42.5 ± 2.4 years; height: M = 181.0 ± 3.8 cm; weight: M = 81.2 ± 2.7 kg; experience teaching: M = 12.3 ± 1.8 years) (females: n = 2; age: M = 39.2 ± 2.5 years; height: M = 167.1 ± 1.5cm; weight: M = 54.2 ± 3.0 kg; experience teaching: M = 11.9 ± 0.8 years)	all healthy and free of cardiovascular and inflammatory diseases, allergies, dental issues, and substance abuse; no women pregnant or taking oral contraceptives; no extreme physical activity 48 hours prior; no sports, anti- histamines, or anti- inflammatory medication 24 hours before testing; rescheduled if infection on test day; no teeth-brushing before morning saliva samples; no smoking, eating, or alcohol, caffeine, or fruit juice 60 minutes before sampling	within-subjects design: 2 hour lecture to 200 students (2nd class period of the year) vs. a control day without lecture	right before (10am) , 0 minutes after (12pm), 120 minutes after (2pm), 8 hours after (8pm) the completion of the stressor or the control period	TNF-α, IL-10, IL-2, and IL-4: cotton swab Salivettes (Sarstedt Co., Nümbrecht, Germany; cytometric bead array kit (BD Biosciences Pharmingen, San Diego, CA, USA)	TNF-α: effect for sampling time [F(3, 24) = 4.7; p = .04, g = 2.53], with higher TNF-α concentrations 120 minutes after the completion of the lecture compared with pre-lecture levels IL-10: no effect for day or time of sampling IL-2: effect for sampling time [F(3, 24) = 5.1; p = .05, g = 1.73], with higher IL-2 concentrations 120 minutes after the completion of the lecture compared with pre-lecture levels IL-4: effect of sampling time [F(3, 24) = 4,2; p = .04, g = 1.33], with higher IL-4 concentrations 120 minutes after the completion of the lecture compared with pre-lecture levels	very small sample size, primarily men; no control for menstrual status in the two women; stress paradigm not been validated in previous studies; professors likely of uniformly high SES and thus results may not be generalizable; unclear to what degree participants found the manipulation stressful
Usui et al., 2012	N = 10 (all males; age: M = 23 ± 3 years; height: M = 176.4 ± 3.4 cm; body mass M = 66.8 ± 7.8 kg; V4 O2max M = 49.7 ± 4.8 ml/kg/minute)	non-smokers, active, no respiratory or anti-inflammatory diseases (incl. asthma); no recent psychological issues or traumas; no dental issues; no medication use 4 weeks prior; no caffeine and alcohol 24 hours before task	within subjects design: exercised on recumbent ergometer at 75 % V4 O2 max for 60 minutes (exercise session) or sat reading or writing quietly (resting session)	0 minutes before, 0 minutes after, 60 minutes after, 120 minutes after the completion of the stressor or the control period	TNF-α, IL-6, and IL-1β: cotton swabs; ELISA kits (Bio- Rad, Bio-Plex Pro Assays, CA, USA).	TNF-α: effects of time: at baseline vs. immediately after (p < .01, g = 2.05) and 60 minutes (p < .05, g = .77) after the completion of the stressor; effect of condition control vs. stress immediately after the completion of the stressor, (p < .01); levels peak immediately after the completion of the stressor IL-6: effect of time: at baseline vs. immediately after (p < .01, g = 1.65) and 60 minutes (p < .05, g = .39) after the completion of the stressor; effect of session: control vs. stress immediately after (p < .01); and 60 minutes (p < .05) after the completion of the stressor; levels peak at immediately after the completion of the stressor IL-1β: effect of time: at baseline vs. immediately after (p < .01, d = 9.69) and 60 minutes after (p < .05, g = 3.03) the completion of the stressor; effect of session: control vs. stress immediately after (p < .01) and 60 minutes after (p < .05) the completion of the stressor; levels peak immediately after the completion of the stressor	very small sample size; no female participants
Dugue et al., 1996	N = 14 (males: n = 6, 22-38 years, M = 33.8 years) (females: n = 8; age 30-39 years, M = 35.1 years)	participants were “subjectively healthy”	within subjects design: stress condition: 15 minutes rest period and then 10 minutes sauna exposure (90°C, , 40% humidity) control condition: resting periods for 15 minutes and then an additional 30 minutes	stress condition: 0 minutes after rest period and 0 minutes after completion of the sauna exposure control condition: 0 minutes after the completion of the first rest period and 0 minutes after the completion of the second rest period	TNF-α, IL-6, IL-1β, and IL- 2: Salivettes; ELISA kits (Immunotech, Luminy, France)	TNF-α: tendency to increase after completion of the sauna stressor; statistically significant increase in the males (p < .05, g = .12) IL-6: detectable in only 60 percent of the specimens investigated IL-1β: showed a tendency (p = .11) to increase after the completion of the sauna stressor as compared to baseline IL-2: levels not detectable	very small sample size; no objective measurement of health status; no control for periodontal health in all participants; no control for menstrual cycle and hormonal contraceptives in female participants; stress paradigm has not been validated; only one post-stimulus saliva sample taken soon after the stressor
Groer et al., 2010	N = 141 (n = 27 female), M = 37 years old (range, 22–64 years); “motorcycle scenario” (n = 49), “workplace scenario” (n = 92)	majority of the officers had >3 years or more of police experience; two groups were comparable on demographic variables	between subjects design: two virtual reality scenarios: motorcycle chase (2 minutes) and workplace gun confrontation (6 minutes) scenarios	0 minutes before, 10 and 30 minutes after completion of the stressor	IL-6: passive drool ELISA (eBioscience, San Diego, CA)	IL-6: significant increase at 10 minutes after the completion of the stressor in the workplace group (t = 2.02, p = . 05, g = .16)	smaller proportion of women than men; no control for periodontal health; no control for menstrual cycle and hormonal contraception in females; samples only taken up to 30 minutes post- stressor; virtual reality simulation not well-validated and may not have been considered stressful
Minetto et al., 2005	N = 17 spinning activity: n =7 endurance- trained athletes; age M = 29.5 ± 8.0 years; 42.9% female, n = 3) isokinetic activity: n = 10 athletes of different disciplines; age M = 27.0 ± 6.9 years; all males	complete medical exam performed; no intense physical activity for 24 h before the study	between subjects design: 7 endurance athletes did controlled spinning activity for 3 hours; 10 athletes from different disciplines did isokinetic exercise test on a Cybex 6000 device (Cybex, Division of Lumex, Ronkonkoma, USA)	spinning activity: 15 and 5 minutes before, and 0 minutes after completion of the exercise isokinetic activity: 15 and 5 minutes before the warm-up, 0 minutes after the completion of the exercise activity, and 7, 15, 30, 45, 60, 90 and 120 minutes after completion	IL-6: Salivettes (Sarstedt, Numbrecht, Germany); ELISA (Quantikine High Sensitivity human IL-6 immunoassay, R&D Systems) and immunoradio metric assay (IRMA; Biosource- Europe, Nivelles, Belgium)	IL-6: not detectable in IMRA methods; for the isokinetic test salivary levels only showed a non- significant increase immediately after the completion of the test; salivary and serum levels were not correlated at any time point	very small sample size; no control for periodontal health; no control for menstrual cycle and hormonal contraception in females
Minetto et al., 2007	N =15 elite athletes (all male) median age = 23 years, median weight = 70 kg	none of the subjects was a current smoker or taking any medication; no exercise 24 hours before experiments	within subjects design: (only the athletes completed the exercise task, which was 15 minutes warm- up cycling and three isometric maximal voluntary contractions of the knee extensors; then 160 isometric contractions of the knee extensor muscle completed in ~25 minutes	all samples collected between 10am and 1pm; 0 minutes before exercise (in all 32 participants), and then 0 minutes after, and 30, 60, 90, and 120 minutes after completion of the stressor (in the athletes)	IL-6: passive drool and Salivettes (Sarstedt, Numbrecht, Germany); ELISA (Quantikine High Sensitivity human IL-6 immunoassay, R&D Systems, Abingdon,UK)	IL-6: levels peak immediately after the completion of the stressor (statistically significant from baseline, p < .05) and return to baseline by 120 minutes; no significant correlation between serum and salivary levels; cotton- interference effect with cotton Salivettes, passive drool more accurate	small sample size; no female athlete participants; no control for periodontal health; no control for menstrual cycle and hormonal contraception in females
Lester et al., 2010	N = 36 first year undergraduate students (males: n =2, age = 21–22 years) (females: n = 34, age = 20–35 years)	excluded if pregnant; no eating or drinking one hour before saliva sampling	within subjects design: 1 baseline period and 3 timed anatomy practical exams ~3 weeks apart	all samples collected between 12:30 and 1:00 pm at a baseline time point (two weeks before the first test) and then within 60 minutes prior to each of the three exams, each ~3 weeks apart	IL-6 and IL-12: passive drool; ELISA (R&D Systems, Minneapolis, MN)	IL-6: increase from the first to the third test (p < .05) IL-2: increase from the first to the third test (p < .05) IL-12: increase from the first to the third test (p < .05)	small sample size; no control for periodontal health in all participants; no control for menstrual cycle and hormonal contraception in females; samples only taken at one time point during the exam stressors
Ilardo et al., 2001	N = 30 normal healthy individuals (males: n = 20, age = 20-28, M = 20.1 years) (females: n = 10, age = 19-23, M = 21.2 years)	excluded if: history of psychological disorders, chronic illness, ongoing medical treatment, regular drug intake, alcohol use or smoking	within subjects design: rafting competition; subjects were assigned to six- member rafting teams and paddled for an average time of 1 hour	0 minutes before (at 2pm) and 0 minutes after (at 3pm) the completion of the first one- hour training session on day 1; 0 minutes before (at 10am) and 0 minutes after (at 11am) the completion of the second one-hour training session on day 2; and 0 minutes before (at 3pm) and 0 minutes after (at 4pm) the completion of a one-hour competitive rafting session on day 3; two samples also collected at 8am & 6pm on each day	IL-1β: Salivettes (Sarstedt, Germany); ELISA (Immunotech, Luminy, France)	IL-1β: concentration significantly higher in males (12.6 pmol/l) than in females (5.1 pmol/l) after the completion of the competition period (z = 1.96, p < .05); significant increase after the completion of the one hour competitive rafting in both sexes (p < .05)	small sample size; no screening for periodontal health in all participants; no control for menstrual cycle and hormonal contraception in females; non- validated stressor paradigm; only one sample was taken immediately post- task; unclear whether participants had any previous experience with rafting
Zefferino et al., 2006	N = 30 emergency policemen (all males: age M = 44.5 ± 5.8 years; experience M = 17.1 ± 6.4 years)	no mention of exclusion criteria; no eating or drinking (except for water) 1 hour before saliva collection	within subjects design: emergency work shift vs. control vacation period	at the beginning and end of the completion of a work shift (8am and 1:30pm, respectively) or at the same times during the vacation period	IL-1β: Salivettes (Salivette- Sarstedt); ELISA (Roche)	IL-1β: concentrations higher at the beginning of the shift than at the end, (p < .05, g = .35); non- significant trend for IL-1β reduction during the vacation period compared to work shift period	small sample size; no screening for periodontal health in all participants; no female participants; participants may not have seen the work day situation as particularly stressful
Campisi et al. 2012	N = 15 healthy college undergraduates (males: n = 4, age = 18-22) (females: n = 11, age = 18-22)	no chronic or acute illness (including periodontal disease), no regular medication (with the exception of contraceptives); good health prior to study; no exercise, meals, or beverages at least 1h prior to study	within subjects design: Trier Social Stress Test	0 minutes before stressor, and 0 minutes after and 30 minutes after the completion of the stressor	CRP: Salivettes (Sarstedt, Newton, NC); ELISA (Salimetrics, State College, PA)	CRP: no statistically significant differences in levels between each time point of sampling	very small sample size; no control for periodontal health in all participants; no control for menstrual cycle and hormonal contraception in females; samples only taken 30 minutes post- stressor
Mastro-lonardo et al. 2007	N = 50 (as part of a larger study on those with and without psoriasis; only controls used in this review) age- and sex- matched healthy controls: 22 men and 28 women; age M = 39.8 ± 10.6)	controls: matched for age, sex, educational level, smoking habits, and use of oral contraceptives; no exercise, smoking, alcohol, or eating for at least 1 h before the session	between subjects design: 5- minute relaxation period, two stressful tasks (mental math and Stroop Color Word Naming Test) of 5-minutes duration each	all sessions occurred between 2:30pm and 3:30pm; collected at 0 minutes before stressor and 10 minutes after the completion of the stressor	IL-1β: Salivettes (Sarstedt, Germany); ELISA (Euroclone Ltd., city, UK)	IL-1β: levels increase after completion of the stressor among healthy controls with a significant group-by-time interaction (p < .01)	no control for periodontal health in all participants; no control for menstrual cycle and hormonal contraception in females; only one post-stressor sample was taken
Izawa et al. 2013	N = 50 healthy young adults (males: n = 39, age M = 21.4 ± 2.4 years) (females: n = 11, age M = 21.6 ± 3.4 years)	no physiological or psychological disorders and no HPA–axis and immune system- affecting drug; all females were in late luteal or early follicular phase of menstrual cycle; no eating , drinking or exercising 1 hour before experiment	within subjects design: Trier Social Stress Test	all sessions occurred between 2:00pm and 7:30pm; collected after 10 minutes of resting baseline, after a 10 minute preparation period, after a 5 minute speech, after a 5 minute mental math task, and 10, 20, 30, 45, and 60 minutes after the completion of the stressor	IL-6: passive drool; ELISA (R & D Systems, Abingdon, UK)	IL-6: levels significantly higher than baseline immediately after (p < .01, g = .37), and 10 (p < .01, g = .33) and 20 minutes after (p < .01, g = .27) the completion of the TSST; levels return to near baseline 60 minutes after the completion of the TSST	no control for periodontal health in all participants; no control for hormonal contraception in females
Mahmood et al., 2013	N = 24 dental students (males: n = 12, females: n = 12)	non-smoking, not taking antibiotics, no chronic diseases or pregnant women; no psychotropic medications	within subjects design: three exam periods-- one month before mid-year exam, during the mid-year exam; and one month after the mid-year exam	all sessions occurred between 8:00am and 12:00pm	IL-1β: passive drool ; ELISA (Salimetrics, State College, PA)	IL-1β: levels significantly higher during mid-year exam period as compared to the pre-exam period (p < .01, g = 1.24) and the post-exam period (p < .01, g = 2.14)	small sample size; no control for hormonal contraception in females