Summary: Salivary alpha-amylase: More than an enzyme - Investigating confounders of stress-induced and basal amylase activity (Dipl.-Psych. Jana Strahler) The hypothalamus-pituitary-adrenal (HPA) axis and the autonomic nervous system (ANS) are two of the major systems playing a role in the adaptation of organisms to developmental changes that threaten homeostasis. The HPA system involves the secretion of glucocorticoids, including cortisol, into the circulatory system. Numerous studies have been published that introduced salivary cortisol to assess HPA axis activity and therefore strengthens its role as an easy obtainable biomarker in stress research that can be monitored easily and frequently. Recent findings suggest a possible surrogate marker of autonomic activity due to autonomic innervation of salivary glands: salivary alpha-amylase (sAA). Up to date, additional methodological research is needed for a better understanding of the advantages and disadvantages of sAA activity in comparison to already established markers of ANS activity. The aim of the present thesis is to further our knowledge of confounders of sAA activity under basal and acute stress conditions and to strengthen the validity of this enzyme as an easy obtainable alternative for ANS testing. After introducing classical and modern stress concepts and stress system physiology (chapter 2), the reader is acquainted with anatomical basics of salivary gland innervation and secretion of salivary proteins, including sAA, due to autonomic innervation (chapter 3 and 4). Afterwards, a more nuanced review of methodological considerations of sAA determination shows gaps of knowledge concerning its usefulness as a marker of ANS activity (chapter 5). Given the fact that the integration of sAA into developmental and aging research is a relative recent phenomenon, several issues have to be addressed before a final conclusion could be drawn. Therefore, we conducted a series of studies incorporating these considerations regarding behavioral correlates of inter- and intraindividual differences in sAA activity with a special emphasis on older adults. Chapter 7 deals with sAA activity under psychological stress conditions in different age groups. Since vulnerability to disease and disease prevalence patterns change with age, it is important to investigate stress reactivity of people in different age groups. We therefore investigated children between 6 and 10 years, because childhood is a sensitive period of growth and development, and thus plays an important role for later life health. Young adults were included to represent the most studied human age group as a reference. Older adults between 59 and 61 years were investigated, because at this age the course is set for the further development of a person’s health in later life, and because autonomic stress responses in older age might be important determinants of cardiovascular and inflammatory aging. Our goal is to test for associations of sAA with more established stress system markers, i.e., salivary cortisol as outcome measurement of HPA reactivity, heart rate (HR) and heart rate variability (HRV) as markers for autonomic reactivity, and to directly compare these responses between different age groups across the life span. Secretion of sAA and cortisol was repeatedly assessed in 62 children, 78 young adults, and 74 older adults after exposure to a standardized psychosocial stressor, the Trier Social Stress Test. In addition, cardiovascular activity was measured in both adult groups. Older adults showed attenuated sAA, HR, and HRV responses. Furthermore, we found higher sAA but lower cortisol at baseline as well as lower sAA and cortisol responses in children. Age by sex interactions were observed only for cortisol with higher responses in older male participants. No associations between the parameters were found. Results in children and young adults confirm previous results. Overall, findings implicate sAA as an alternative or additional autonomic stress marker throughout the life span, with marked and rapid responsiveness to stress in three relevant age groups. The impact of age and chronic stress on basal sAA activity is the center of interest in chapter 8. We therefore assessed diurnal profiles of sAA and salivary cortisol in 27 younger and 31 older competitive ballroom dancers as well as 26 younger and 33 older age- and sex-matched controls. According to the Allostatic Load concept, repeated, non-habituating responses to social-evaluative conditions, which characterize the lives of competitive ballroom dancers, should be associated with stress system dysregulations. Furthermore, we expect to see an increased sympathetic drive associated higher overall alpha-amylase activity in older adults. Analyses revealed an elevated daily overall output of sAA in older adults while there was no effect of age on mean cortisol levels. Alterations of diurnal rhythms were only seen in younger male dancers showing a flattened diurnal profile of sAA and younger dancers and female older dancers showing a blunted diurnal rhythmicity of cortisol. Furthermore, we found a negative correlation between summary indices of basal sAA and the amount of physical activity. In conclusion, higher overall output of sAA in older adults was in line with the phenomenon of a “sympathetic overdrive” with increasing age. Furthermore, a lower output of sAA in people who are more physical active was in line with the hypothesis of an exercise-induced decrease of sympathetic activity. Taken together, results of chapter 7 and 8 show a clear impact of age on sAA activity, either under acute stress or basal conditions. One problem when integrating sAA into developmental and aging research is the use of adrenergic agonists and antagonists what is very common in older adults, i.e. antihypertensive drugs (AD). As well, the previously shown sympathetic overactivity that occurs with normal aging is associated with higher blood pressure (BP). Therefore, chapter 9 deals with a possible impact of high BP and AD on diurnal sAA activity in 79 older adults (33 normotensive adults, 16 medicated vs. 45 hypertensive adults, 34 medicated). Results showed a pronounced rhythm of sAA in all groups. Diurnal profiles differed significantly between men and women with men lacking the typical decrease of sAA in the morning and showing more pronounced alterations throughout the day. An effect of AD on sAA profiles and area under the curve values indicates that subjects not using AD´s show a heightened diurnal profile and a higher total output of sAA. Descriptively, this was also true for hypertensive older adults. Hypertensive subjects and those not using AD showed the highest diurnal output of sAA and the steepest slope. In sum, our results show an impact of antihypertensive medication and a difference between normotensive and hypertensive subjects on characteristics of diurnal sAA activity. Hence, findings are of particular interest in research using sAA as a prognostic indicator of pathological states and processes. Given the fact that hypertension was also shown to be associated with substantial changes of transmitters within the suprachiasmatic nucleus (SCN) - the “biological clock” that receives photic input from retinal glands via the retinohypothalamic pathway - and an altered output from the SCN to the sympathetic nervous system, we broaden the idea of a possible effect of different lighting conditions on morning sAA profiles in chapter 10. In a counterbalanced within-subjects design six men and 16 women of different ages collected sAA morning profiles on two consecutive days with leaving their shutters closed on the one day (= dark) and open their shutters on the other day (= bright). We were able to replicate earlier findings of light-induced changes of salivary cortisol with higher responses during the bright condition. On either day, women showed larger cortisol increases than men. Despite multisynaptic autonomic connections arising from the SCN projecting to multiple organs of the body, we could not find an effect of sunlight on sAA morning profiles. Evidence for circadian clock gene expression in human oral mucosa might account for this result and indicates that peripheral oscillators may act more like integrators of multiple different time cues, e.g. light, food intake, instead of a “master” oscillator (SCN). Results of chapter 7 to 10 provide clear evidence that sAA is heightened in states of autonomic arousal, i.e. stress, aging and hypertension, and that its circadian rhythmicity seems to be regulated rather integrative than directly via efferent input from hypothalamic SCN neurons. In chapter 11 this thesis tries to approach one central question: What is the biological meaning of the findings made? According to this enzyme´s anti-bacterial and digestive action short term changes might not have a biological meaning itself but rather reflect just a small part of multiple coordinated body responses to stressful stimuli. While the sympathetic branch of the ANS mainly stimulates protein secretion, the parasympathetic branch stimulates saliva flow. Acute stress responses might therefore be interpreted as reflecting predominant sympathetic activity together with parasympathetic withdrawal. The same mechanism could also be suitable for the finding of higher diurnal levels of sAA in older adults or hypertensive subjects reflecting a higher peripheral sympathetic tone in these groups. Diurnal profiles of sAA itself may reflect circadian changes in autonomic balance. Circadian rhythms are of great advantage since they enable individuals to anticipate. This pre-adaptation enables the individual to cope with upcoming demands and challenges. Our finding of a relationship between sAA and salivary cortisol what strengthens the relevance of glucocorticoids that were previously shown to be able to phase shift circadian rhythms in cells and tissue. Within a food-related context there is evidence that decreasing levels of sAA in the morning could reflect increases of feeling hungry since sAA systematically increases during food consumption and with the subjective state of satiety. So far, much more research is needed to identify underlying physiological mechanisms of circadian sAA rhythmicity. Taking the next step, future studies will have to focus on the integration of sAA assessment into longitudinal studies and different disease states to prove its applicability as a marker of sympathetic neural functioning in the genesis and prognosis of disease.

Summary: Salivary alpha-amylase: More than an enzyme - Investigating confounders of stress-induced and basal amylase activity (Dipl.-Psych. Jana Strahler) The hypothalamus-pituitary-adrenal (HPA) axis and the autonomic nervous system (ANS) are two of the major systems playing a role in the adaptation of organisms to developmental changes that threaten homeostasis. The HPA system involves the secretion of glucocorticoids, including cortisol, into the circulatory system. Numerous studies have been published that introduced salivary cortisol to assess HPA axis activity and therefore strengthens its role as an easy obtainable biomarker in stress research that can be monitored easily and frequently. Recent findings suggest a possible surrogate marker of autonomic activity due to autonomic innervation of salivary glands: salivary alpha-amylase (sAA). Up to date, additional methodological research is needed for a better understanding of the advantages and disadvantages of sAA activity in comparison to already established markers of ANS activity. The aim of the present thesis is to further our knowledge of confounders of sAA activity under basal and acute stress conditions and to strengthen the validity of this enzyme as an easy obtainable alternative for ANS testing. After introducing classical and modern stress concepts and stress system physiology (chapter 2), the reader is acquainted with anatomical basics of salivary gland innervation and secretion of salivary proteins, including sAA, due to autonomic innervation (chapter 3 and 4). Afterwards, a more nuanced review of methodological considerations of sAA determination shows gaps of knowledge concerning its usefulness as a marker of ANS activity (chapter 5). Given the fact that the integration of sAA into developmental and aging research is a relative recent phenomenon, several issues have to be addressed before a final conclusion could be drawn. Therefore, we conducted a series of studies incorporating these considerations regarding behavioral correlates of inter- and intraindividual differences in sAA activity with a special emphasis on older adults. Chapter 7 deals with sAA activity under psychological stress conditions in different age groups. Since vulnerability to disease and disease prevalence patterns change with age, it is important to investigate stress reactivity of people in different age groups. We therefore investigated children between 6 and 10 years, because childhood is a sensitive period of growth and development, and thus plays an important role for later life health. Young adults were included to represent the most studied human age group as a reference. Older adults between 59 and 61 years were investigated, because at this age the course is set for the further development of a person’s health in later life, and because autonomic stress responses in older age might be important determinants of cardiovascular and inflammatory aging. Our goal is to test for associations of sAA with more established stress system markers, i.e., salivary cortisol as outcome measurement of HPA reactivity, heart rate (HR) and heart rate variability (HRV) as markers for autonomic reactivity, and to directly compare these responses between different age groups across the life span. Secretion of sAA and cortisol was repeatedly assessed in 62 children, 78 young adults, and 74 older adults after exposure to a standardized psychosocial stressor, the Trier Social Stress Test. In addition, cardiovascular activity was measured in both adult groups. Older adults showed attenuated sAA, HR, and HRV responses. Furthermore, we found higher sAA but lower cortisol at baseline as well as lower sAA and cortisol responses in children. Age by sex interactions were observed only for cortisol with higher responses in older male participants. No associations between the parameters were found. Results in children and young adults confirm previous results. Overall, findings implicate sAA as an alternative or additional autonomic stress marker throughout the life span, with marked and rapid responsiveness to stress in three relevant age groups. The impact of age and chronic stress on basal sAA activity is the center of interest in chapter 8. We therefore assessed diurnal profiles of sAA and salivary cortisol in 27 younger and 31 older competitive ballroom dancers as well as 26 younger and 33 older age- and sex-matched controls. According to the Allostatic Load concept, repeated, non-habituating responses to social-evaluative conditions, which characterize the lives of competitive ballroom dancers, should be associated with stress system dysregulations. Furthermore, we expect to see an increased sympathetic drive associated higher overall alpha-amylase activity in older adults. Analyses revealed an elevated daily overall output of sAA in older adults while there was no effect of age on mean cortisol levels. Alterations of diurnal rhythms were only seen in younger male dancers showing a flattened diurnal profile of sAA and younger dancers and female older dancers showing a blunted diurnal rhythmicity of cortisol. Furthermore, we found a negative correlation between summary indices of basal sAA and the amount of physical activity. In conclusion, higher overall output of sAA in older adults was in line with the phenomenon of a “sympathetic overdrive” with increasing age. Furthermore, a lower output of sAA in people who are more physical active was in line with the hypothesis of an exercise-induced decrease of sympathetic activity. Taken together, results of chapter 7 and 8 show a clear impact of age on sAA activity, either under acute stress or basal conditions. One problem when integrating sAA into developmental and aging research is the use of adrenergic agonists and antagonists what is very common in older adults, i.e. antihypertensive drugs (AD). As well, the previously shown sympathetic overactivity that occurs with normal aging is associated with higher blood pressure (BP). Therefore, chapter 9 deals with a possible impact of high BP and AD on diurnal sAA activity in 79 older adults (33 normotensive adults, 16 medicated vs. 45 hypertensive adults, 34 medicated). Results showed a pronounced rhythm of sAA in all groups. Diurnal profiles differed significantly between men and women with men lacking the typical decrease of sAA in the morning and showing more pronounced alterations throughout the day. An effect of AD on sAA profiles and area under the curve values indicates that subjects not using AD´s show a heightened diurnal profile and a higher total output of sAA. Descriptively, this was also true for hypertensive older adults. Hypertensive subjects and those not using AD showed the highest diurnal output of sAA and the steepest slope. In sum, our results show an impact of antihypertensive medication and a difference between normotensive and hypertensive subjects on characteristics of diurnal sAA activity. Hence, findings are of particular interest in research using sAA as a prognostic indicator of pathological states and processes. Given the fact that hypertension was also shown to be associated with substantial changes of transmitters within the suprachiasmatic nucleus (SCN) - the “biological clock” that receives photic input from retinal glands via the retinohypothalamic pathway - and an altered output from the SCN to the sympathetic nervous system, we broaden the idea of a possible effect of different lighting conditions on morning sAA profiles in chapter 10. In a counterbalanced within-subjects design six men and 16 women of different ages collected sAA morning profiles on two consecutive days with leaving their shutters closed on the one day (= dark) and open their shutters on the other day (= bright). We were able to replicate earlier findings of light-induced changes of salivary cortisol with higher responses during the bright condition. On either day, women showed larger cortisol increases than men. Despite multisynaptic autonomic connections arising from the SCN projecting to multiple organs of the body, we could not find an effect of sunlight on sAA morning profiles. Evidence for circadian clock gene expression in human oral mucosa might account for this result and indicates that peripheral oscillators may act more like integrators of multiple different time cues, e.g. light, food intake, instead of a “master” oscillator (SCN). Results of chapter 7 to 10 provide clear evidence that sAA is heightened in states of autonomic arousal, i.e. stress, aging and hypertension, and that its circadian rhythmicity seems to be regulated rather integrative than directly via efferent input from hypothalamic SCN neurons. In chapter 11 this thesis tries to approach one central question: What is the biological meaning of the findings made? According to this enzyme´s anti-bacterial and digestive action short term changes might not have a biological meaning itself but rather reflect just a small part of multiple coordinated body responses to stressful stimuli. While the sympathetic branch of the ANS mainly stimulates protein secretion, the parasympathetic branch stimulates saliva flow. Acute stress responses might therefore be interpreted as reflecting predominant sympathetic activity together with parasympathetic withdrawal. The same mechanism could also be suitable for the finding of higher diurnal levels of sAA in older adults or hypertensive subjects reflecting a higher peripheral sympathetic tone in these groups. Diurnal profiles of sAA itself may reflect circadian changes in autonomic balance. Circadian rhythms are of great advantage since they enable individuals to anticipate. This pre-adaptation enables the individual to cope with upcoming demands and challenges. Our finding of a relationship between sAA and salivary cortisol what strengthens the relevance of glucocorticoids that were previously shown to be able to phase shift circadian rhythms in cells and tissue. Within a food-related context there is evidence that decreasing levels of sAA in the morning could reflect increases of feeling hungry since sAA systematically increases during food consumption and with the subjective state of satiety. So far, much more research is needed to identify underlying physiological mechanisms of circadian sAA rhythmicity. Taking the next step, future studies will have to focus on the integration of sAA assessment into longitudinal studies and different disease states to prove its applicability as a marker of sympathetic neural functioning in the genesis and prognosis of disease.:Table of Contents 1. Introduction 1 2. Stress 3 2.1. Stress concepts 3 2.1.1. Traditional concepts of stress 3 2.1.2. Allostasis and Allostatic Load 4 2.2. Stress system physiology 6 2.2.1. The hypothalamic-pituitary-adrenal (HPA) axis 6 2.2.1.1. Physiology 6 2.2.1.2. HPA axis activity indicators 6 2.2.2. The autonomic nervous system (ANS) 7 2.2.2.1. Physiology 7 2.2.2.2. ANS activity indicators 8 2.2.3. Relationships between stress systems 10 3. Saliva and salivary glands 11 3.1. Physiology 11 3.1.1. Anatomy, origin, and composition 11 3.1.2. Innervation 12 3.1.3. Salivary gland physiology with aging 13 3.2. Saliva and salivary flow 13 3.3. Protein secretion 14 4. Alpha-amylase in saliva 15 4.1. Chemical characteristics 15 4.2. Secretion of alpha-amylase 15 4.3. Diagnostic value of alpha-amylase 16 5. Methodological considerations of alpha-amylase determination 17 5.1. Collection methods and preparation 17 5.1.1. Saliva collection 17 5.1.2. Impact of flow rate 17 5.1.3. Impact of pH-value 18 5.2. Biochemical determination 18 5.3. Interindividual differences in sAA activity 19 5.3.1. Basal activity 20 5.3.2. Acute responses 20 5.3.3. Age effects 21 5.3.3.1. Basal amylase activity 21 5.3.3.2. Stress-induced amylase activity 21 5.3.4. Sex differences 22 5.3.4.1. Basal amylase activity 22 5.3.4.2. Stress-induced amylase activity 23 5.3.5. Modulating factors influencing amylase (re-)activity 24 5.3.5.1. Impact of smoking 24 5.3.5.2. Impact of alcohol 25 5.3.5.3. Impact of caffeine 25 5.3.5.4. Impact of high body fat and obesity 26 5.3.5.5. Impact of food intake 26 5.3.5.6. Impact of physical exercise 27 5.3.5.7. Impact of somatic and psychiatric diseases 27 5.3.5.8. Impact of medical drugs 29 5.3.5.9. Impact of sunlight on diurnal amylase 29 6. Aims and outline of the present work 31 7. Salivary alpha-amylase stress reactivity across different age groups 32 7.1. Introduction 32 7.2. Methods 35 7.2.1. Participants 35 7.2.2. Study Protocol 35 7.2.3. Measures 36 7.2.3.1. Saliva sampling 36 7.2.3.2. Heart rate and heart rate variability 36 7.2.3.3. Biochemical analyses 37 7.2.3.4. Psychometrical analyses 37 7.2.4. Statistical analyses 38 7.3. Results 38 7.3.1. Sample characteristic 38 7.3.2. Subjective stress response 39 7.3.3. Physiological stress response 39 7.3.3.1. Salivary alpha-amylase 39 7.3.3.2. Salivary cortisol 40 7.3.3.3. Heart rate 42 7.3.3.4. Heart rate variability 43 7.3.3.5. Determinants of the salivary alpha-amylase stress response 45 7.4. Discussion 45 7.5. Conclusion 48 8. Aging diurnal rhythms and chronic stress: Distinct alteration of diurnal rhythmicity of salivary alpha-amylase and cortisol 49 8.1. Introduction 49 8.2. Methods 52 8.2.1. Participants 52 8.2.2. Study protocol 53 8.2.3. Measures 53 8.2.3.1. Saliva sampling 53 8.2.3.2. Biochemical parameters 54 8.2.3.3. Psychological parameters 54 8.2.4. Statistical analyses 54 8.2.4.1. Preliminary analyses 54 8.2.4.2. Diurnal course of salivary alpha-amylase 55 8.3. Results 56 8.3.1. Sample characteristic 56 8.3.2. Preliminary analyses: impact of oral contraceptives, blood pressure, and lipid lowering medication on diurnal profiles 56 8.3.3. Diurnal course of salivary alpha-amylase 57 8.3.3.1. Salivary alpha-amylase over the day 57 8.3.3.2. Salivary alpha-amylase after awakening 58 8.3.4. Diurnal course of salivary cortisol 59 8.3.4.1. Salivary cortisol over the day 59 8.3.4.2. Salivary cortisol after awakening 60 8.3.5. Diurnal course of salivary biomarkers: associations and determinants 61 8.4. Discussion 62 8.5. Conclusion 65 9. Impact of blood pressure and antihypertensive drugs on diurnal alpha-amylase activity: A novel marker of sympathetic drive 67 9.1. Introduction 67 9.2. Methods 68 9.2.1. Participants 68 9.2.2. Study protocol 69 9.2.3. Measures 69 9.2.3.1. Saliva sampling 69 9.2.3.2. Biochemical parameters 69 9.2.3.3. Blood pressure assessment 70 9.2.4. Statistical analyses 70 9.3. Results 70 9.3.1. Participants 70 9.3.2. Salivary alpha-amylase 71 9.3.2.1. Salivary alpha-amylase over the day 71 9.3.2.2. Salivary alpha-amylase after awakening 74 9.4. Discussion 75 9.5. Perspectives 76 10. Light affects morning salivary cortisol, but not salivary alpha-amylase 77 10.1. Introduction 77 10.2 Methods 79 10.2.1. Participants 79 10.2.2. Study protocol 80 10.2.3. Measures 80 10.2.3.1. Saliva sampling 80 10.2.3.2. Biochemical parameters 81 10.2.4. Statistical analyses 81 10.3. Results 82 10.3.1. Sociodemographics 82 10.3.2. Salivary alpha-amylase 82 10.3.3. Salivary cortisol 84 10.3.4. Associations between biochemical parameters 85 10.4. Discussion 86 10.5. Conclusion 89 11. General discussion 90 11.1. Summary of the results 90 11.1.1. Salivary alpha-amylase stress reactivity across different age groups 91 11.1.2. Aging diurnal rhythms and chronic stress: Distinct alteration of diurnal rhythmicity of salivary alpha-amylase and cortisol 91 11.1.3. Impact of blood pressure and antihypertensive drugs on diurnal alpha-amylase activity: A novel marker of sympathetic drive 92 11.1.4. Light affects salivary morning cortisol, but not salivary alpha-amylase 93 11.2. Integration of main findings 93 11.3. Stress-induced amylase activity, basal rhythm, and its biological meaning 95 11.4. Methodological consequences 97 11.4.1. Circadian variation 97 11.4.2. Longitudinal variation 98 11.4.3. Short-term variation and stability 98 11.4.4. Long-term change 99 11.5. Outlook 100 12. References 102