3. Psychobiological mechanisms of stress and restitution

The focus in this section is to increase the knowledge concerning the interplay between stress, sleep, health perception and fatigue, and the underlying main biological mechanisms. A certain focus is on inflammation, which is increasingly recognized as an important link between behavior and health, and coupled to subjective health and wellbeing. Here, inflammation represents a generalized response to threat in the internal rather than the external milieu, and is thus a generalized part of the stress response. In this work we have tried to delineate the physiological sleep response to real life stress, effects of sleep deprivation, and the links between sleep, fatigue, inflammation and subjective health. However, we have shown that behavioral manipulations (stress or sleep deprivation), may also stimulate the parts of the inflammatory response, which in turn is linked to both subjective health, sleep and behavior. This work includes studies with polysomnography, positron emission tomography, functional magnetic resonance imaging, immunology, subjective ratings and, in some cases, investigations to study if core observations are also detectable on the population level using epidemiological approaches.


3.1 Field studies of stress and sleep

It was observed in some of our epidemiological studies (see Long-term effects) that stress was related to disturbed sleep. This has been further investigated in several collaborative (S+P) studies using polysomnography (recording EEG/EOG/EMG during sleep – “PSG”). The results show that a working day with high stress is followed by slightly impaired (more stage 1 (poor)) sleep during the night sleep and that individuals under stress under a longer period of time show impaired sleep (increased Stage 1). The stress/PSG relations have been followed up with collaborative (S+P) longitudinal studies of sequential relations between key variables via the day- to-day variability across days. This has shown that stress during the day predicts reduced sleep quality during the subsequent night, sequentially across 42 days. Similarly, poor or short sleep predicts next day fatigue and sleepiness across a number of days and so does illness symptoms and subjective health. The latter effects are particularly strongly related to fatigue and sleepiness. The results on subjective health and fatigue are mirrored in an experimental study of partial sleep deprivation showing that health ratings deteriorate gradually in parallel with increased fatigue, with increased inflammatory activation from stimulated samples, but not in circulation. The studies of sleep in this section should also be seen in relation to the studies of insomnia diagnoses related to subsequent sickness absence, disability pension and mortality (see elsewhere).

3.2 Sleepy brain – neural/emotional effects of sleep reduction

One major collaborative effort (S+P+external partners (KI)) has been to build a consortium for functional magnetic resonance imag- ing (fMRI) for use in relation to stress, emotion and sleep loss. The initiation took two years and the project now delivers results, e.g. regarding how emotional functioning and regulation is affected by suboptimal sleep. We show that sleep deprivation caused increased subjective responses viewing others in pain, but less activation in early perceptual areas for face perception when viewing emotional faces. We are currently investigating the neural correlates of the increased emotional response, with preliminary observations related to increased activity in the parietal cortex. Across sleep conditions (full sleep and sleep loss), pictures of other people in pain (a model for empathy) entailed significantly increased activity in the anterior insulae and the medial cingulate cortex. The findings suggest that sleep deprivation affects processing of social stimuli in the brain, of obvious interest for person-to-person communication in working life and interaction in daily life. The findings on increased emotional reactivity as an effect of suboptimal sleep may have links to our positron emission tomography studies  showing reduced serotonin receptor affinity in patients exposed to long-term occupational stress (resulting in exhaustion syndrome), as well as decreased amygdala/prefrontal cortex connectivity. These findings are the first to indicate brain activity behind burnout symptoms and also link to our previous findings of frequent micro-arousals characterizing the poor sleep of burnout subjects. The decreased amygdala prefrontal connectivity could represent a physiological correlate to a decreased control of emotional responsivity, and has been shown both behaviorally and with fMRI after short sleep. We currently address this phenomenon with dedicated behavioral tests performed in the fMRI scanner. The effects of short sleep on emotional processing are important since these processes lay the ground for successful social interaction and could explain the finding that sleep disturbances predict poor social support at work at the population level. We therefore also directly test social interaction in behavioral tests that aim to be applied in working life in future studies. A first result is that partial acute sleep restriction leads to increased sensitivity to social exclusion, further strengthening the notion of a strong connection between sleep, emotion and social interaction of importance for working life. In spite of the obvious relevance, this subject area has hardly been investigated in this perspective previously. In this context should also be brought up the finding that strong reduction of sleep precedes referral for exhaustion symptom (burnout).

3.3 Sleep, sleepiness and the timing of sleep and work

An important focus has been to study how insufficient sleep and timing of work affects risk of occupational accidents. A series of studies have investigated these factors in relation to night time driving as well as studies on driving performance on real roads, together showing that night driving leads to very high levels of physiological, subjective and behavioral sleepiness, which is strongly related to lane departures (driving off the road). As many as 42% of the drivers reached so severe sleepiness (both subjectively and objectively), that they were taken off the road by the test leader. The studies are the first in the world of their kind. With respect to the effects of timing of sleep we showed in a collaborative study (P+S) that sleeping during the day, something common amongst night workers, results in a number of physiological changes, including profound effects on metabolic indices. In one study we simulated ships’ watch schedules for the bridge officers (EU-project) and found that the most common schedules (6 on – 6 off) lead to high levels of sleepiness and falling asleep incidents (EEG-recording) during early morning watches, particularly if the free watch had been used for work (which is common). In another attempt to understand non-day work scheduling we studied short and long-haul airline crews. Night flying and early mornings were particularly characterized by fatigue/sleepiness. The data was used to validate and improve our mathematical work schedule fatigue predictor. The mathematical model is now part of Boeing’s crew scheduling software (discussed in “Methods”).

3.4 Perception of own health/self-rated health

Subjective health is an established predictor of various types of morbidity and at the same time it is negatively affected by stress. A series of studies have shown subjective health to relate to inflammation. In two studies aiming to further delineate the mechanisms and trajectories of subjective health, we have found that a low-grade inflammation (either measured by high levels of IL-6 levels or a high erythrocyte sedimentation rate) is related to poor self-rated health independent of diagnosis. When studying factors affecting self-rated health, we have recently shown that it is affected by acute sleep loss and recovery, and inflammatory stress. We are presently pursuing how inflammation and self-rated health relates to brain activity. Our hypothesis is that the activation of the insular cortex, an area central for the integration of physiological changes in the body, increases in response to inflammatory stress.

3.5 Perception of the health of others

The ability to judge other peoples health is important in daily interaction as well as in clinical encounters. We have been one of the first groups in this new research field and have systematically investigated factors affecting how we perceive health in others. We have so far investigated how inflammation and insufficient sleep affects facial, gait-related and olfactory cues and how this relates to perception of health. We have in a collaborative study (P+S) shown that people with insufficient sleep are judged as being more tired, less attractive, have worse health and to be more sad. In addition, we have presented the first evidence that humans can smell an early innate immune response in a sick individual and that early sickness can be detected from trough changes in gait pattern. These findings support the hypothesis that humans have an ability to detect signals of acute sickness in others and we are now pursuing studies of how watching sick people affects stress, inflammation and social interaction. In all, we have taken a leading role in the studies of mechanisms affecting health perception in others, similarly to what we have previously done in the area of self-ratings of health. We believe that a better understanding of these factors will give important knowledge for everyday life as well as for clinical consultation.


Planned research

The work from the first half of the Center decade will be added to and extended, partly based on the new findings from the first half. Most of the work constitutes an intensive collaborative effort (S+P+other units at SU, KI, and other national and international units) and extends and allows specific investigation of mechanisms that are also investigated in field-studies in organizational and clinical settings that will include a combined methodological approach covering biomarkers, activity and sleeping patterns, diary data along with self-reports in questionnaires and organizational record data on sickness absence. The funding will be used for key personnel in this collaboration.

3.1 Brain function, social interaction and immunity after reduced/impaired sleep, stress or inflammatory challenge.

One major effort of S+P will be to consolidate our brain imaging consortium (with other SU and KI partners) and continue (thefirst studies focused on restricted sleep) to investigate the effects of sleep quality on cognitive processing (e.g. emotion regulation) by suppressing certain sleep stages (slow wave sleep, REM-sleep and fragmented sleep). The project will allow us to tease out how different sleep disturbances (time, fragmentation and certain sleep stages) affect cognitive functioning, such as emotion regulation, and also test whether vulnerability to sleep disturbances differ with age and if women differ from men. We also aim to investigate the effects of stress at the work place, and patient groups with exhaustion using similar brain-imaging protocols, allowing for comparison between studies, as with unique research questions of mechanisms (eg. serotonin receptor affinity). We will continue our involvement in MR studies of exhaustion syndrome patients. In one study we will look at the regulation of emotion and on functional connectivity. In another study we will carry out a two-year follow-up of patients with exhaustion syndrome and serotonin receptor affinity (among other questions).

3.2 Immune effects of sleep loss

Following our previous findings, the brain imaging studies also include immune effects of sleep deprivation, and the relation of such changes to sleepiness and emotional and empathic processing in the brain. These efforts involve intense collaboration within the center (P+S) and with the Division of Biological Psychology at the department of Psychology, as well as with several researchers at the Osher Center for Integrative Medicine at KI. Moreover, we will also develop ecologically relevant tests of (stress related) social interaction between individuals, and to test the sensitivity of such tests to sleep deprivation. An important aim with a strong developmental poten- tial is to start a European consortium on studies of inflammatory stress and brain imaging, and to use such techniques in adequately powered settings to understand the effects of inflammation on social interaction. Members are so far Manfred Schedlowski’s and Sigrid Elsenbruch’s respective group in Essen, Germany, and Neil Harrison and colleagues in University of Sussex in UK.

3.3 Womens’ sleep quality

Women are more prone to sleep disturbances and their sleep seems more influence by stress than men. The second phase (10-year follow-up) of our study of womens’ sleep will be completed this year. We will use SSC resources in a collaborative (S+P) effort to help analyze how polysomnographically recorded sleep (and endocrinology) ages, how it is changed in high and low stress women (and males), how perceived sleep quality corresponds to physiological sleep – are there objective indications of women’s higher levels of sleep complaints, if so, are there are explanations for the higher levels of sleep complaints in women (total work load, children, responsibilities at home), what are the relations between PSG and outcomes like fatigue/sleepiness in men and women, can social factors explain women’s higher levels of sleep problems (responsibility at home). In both polysomnographical studies new methods will be tried to find physiological indicators of sleep quality (consolidation analysis of PSG, fragmentation, stage shifts, sleep spindles, K-complexes, REM density), different frequency ranges. The questions of women’s sleep will be the focus also of epidemiological studies.

3.4 Sleep loss as a modifier of the stress response

A few studies have begun to test the possibility that sleep loss affects/increases stress responses. It is very obvious in our recent paper based on SLOSH. We will therefore carry out a collaborative (S+P) study of total sleep loss and exposure to the Trier stress test (a public presentation) in older and younger individuals. Saliva samples for cortisol, alpha amylase and other stress indicators will be used, as well as heart rate monitoring, self-rating and performance tests.

3.5 Sleep, disease and aging

Some of the SSC support will be used in a series of epidemiological studies (S+I+E) which will relate stress to sleep and these two factors to resulting disease, mortality, dementia, general morbidity, disability pension, drug consumption (all through linkage to official registers for the entire population). This also involves the aging process and dementia, in relation to sleep and stress. We have in preliminary analyses found that while sleep shortens with age, alertness rises and fatigue falls. Some of the latter may be due to loss of stress with increasing age (including retirement). The aging trajectories may predict the development of dementia, but in a paradoxical way; at one stage sleep duration may be increased in subsequent dementia patients (new preliminary analyses). This will now be tested in the Betula (aging, cognition) cohort (Lars-Göran Nilsson) with risk markers for dementia (APOE), and PSG recordings will be made in select groups in that cohort. Also gender differences in sleep will be investigated. This study also has a link to the previous study men- tioned and to the retirement studies we are carrying out.

3.6 Poor sleep and work performance

Closely related to the sleep quality / sleepiness issue is the link between poor (stress related) sleep and work performance. There are reasons to believe that much of the day-to-day variations in fatigue may be linked to poor work performance. To establish this link, as well as the one with stress, would change the view of sleep disturbances in working life and probably give treatment for disturbed sleep a new role at work. This issue will be addressed through longitudinal epidemiology (via the SLOSH database, S+E) and via field studies using actigraphy, polysomnography and performance measurement and will be carried out in collaboration (S+P).

3.7 Mechanisms and trajectories of subjective health as studied by observational and experimental designs.

We will extend the present analysis of dynamics of self-rated health in a U.S sample of elderly individuals in an invited collaboration (University of Kentucky). As noted above, we will also carry out several studies of inflammatory stress which will further advance knowledge on the immune and neural (fMRI) mechanisms of impaired self-rated health.

3.8 Perception of health in others: facial, gait-related and olfactory signals of sleep debt and inflammation.

We will carry a series of studies to investigate how we perceive health and sickness in other people. This includes analyses of odor (molecular changes of odor from t-shirts worn by sick individuals), gait (measured with 3D multichannel gait analyses) and appearance (from photographs). We will study both natural sickness and experimental models where we will stimulate the immune system to induce acute sickness. These studies will be done in collaboration with experts in gait analyses (at Harvard Medical School), molecular analyses (University of Pennsylvania) and analyses of photographs (St:Andrews University).