The relationships of working conditions, recent stressors and childhood trauma with salivary cortisol levels

Transcription

1 Psychoneuroendocrinology (2012) 37, Available online at j our na l h omepa g e: lse vie r.c om/l oca te/ psyne ue n The relationships of working conditions, recent stressors and childhood trauma with salivary cortisol levels Michiel Holleman a, *, Sophie A. Vreeburg b, Jack J.M. Dekker a,c, Brenda W.J.H. Penninx b,d,e a Research Department of Arkin Mental Health Institute, Amsterdam, The Netherlands b Department of Psychiatry, VU University Medical Center, Amsterdam, The Netherlands c Department of Clinical Psychology, VU, Amsterdam, The Netherlands d Department of Psychiatry, Leiden University Medical Center, Leiden, The Netherlands e Department of Psychiatry, University Medical Center Groningen, Groningen, The Netherlands Received 8 March 2011; received in revised form 19 September 2011; accepted 21 September 2011 KEYWORDS HPA axis; Cortisol; Cortisol awakening response; Dexamethasone suppression test; Stress; Demand control model; Job strain; Life events; Childhood trauma Summary Background: An etiological model has been suggested where stress leads to high cortisol levels and hypothalamic pituitary adrenal (HPA) axis dysregulation, resulting in somatic diseases and psychopathology. To evaluate this model we examined the association of different stressors (working conditions, recent life events and childhood trauma) with various cortisol indicators in a large cohort study. Methods: Data are from 1995 participants of the Netherlands Study of Depression and Anxiety (NESDA). Most of the selected participants had a current or remitted anxiety and/or depressive disorder. Working conditions were assessed with self-report questionnaires, life-events and childhood trauma were assessed with interview questionnaires. Cortisol levels were measured in seven saliva samples, determining the 1-h cortisol awakening response (CAR), evening cortisol levels and cortisol suppression after a 0.5 mg dexamethasone suppression test (DST). Results: Regression analyses adjusted for covariates showed two significant associations: low social support at work and high job strain were associated with more cortisol suppression after the DST. No other associations were found with any of the cortisol variables. Conclusions: Working conditions, recent stressors and childhood trauma were not convincingly associated with cortisol levels. # 2011 Elsevier Ltd. Open access under the Elsevier OA license. * Corresponding author at: Research Department of Arkin Mental Health Institute Amsterdam, Strekkerweg 77, 1033 DA Amsterdam, The Netherlands. Tel.: address: michiel.holleman@arkin.nl (M. Holleman) # 2011 Elsevier Ltd. Open access under the Elsevier OA license. doi: /j.psyneuen

2 802 M. Holleman et al. 1. Introduction Stress is thought to lead to a hyperactive hypothalamus pituitary adrenal (HPA) axis and, when chronic, the resulting high cortisol levels may contribute to somatic diseases and psychopathology (Sapolsky, 1998). Evidence for elevated cortisol levels are for example found among persons with obesity or ischemic heart disease (Reynolds et al., 2010), hypertension or diabetes mellitus (Schoorlemmer et al., 2009) and atherosclerosis (Eller et al., 2005; Dekker et al., 2008). In addition, hyperactivity of the HPA-axis was found among persons with depression (Vreeburg et al., 2009a) and anxiety disorders (Vreeburg et al., 2010a). A central assumption is that underlying stress exposure and experience are related to altered HPA axis activity. To assess basal HPA-axis activity, salivary cortisol measures are often used. The cortisol awakening response (CAR) reflects the natural response of the HPA-axis on awakening (Wilhelm et al., 2007), is associated with the circadian rhythm of cortisol secretion and is considered to be regulated in anticipation of demands of the upcoming day (Fries et al., 2009). The CAR is preferable to single morning cortisol measurements since it has a greater intra-personal stability (Coste et al., 1994; Wüst et al., 2000) and is considered to be a more reliable measure for the acute reagibility of the HPAaxis (Schmidt-Reinwald et al., 1999). Evening cortisol levels reflect basal activity, whereas the dexamethasone suppression test (DST) examines the reactivity to a pharmacological challenge of the stress system. Prior studies have examined the impact of various different types of stressors on HPA-axis functioning in adulthood. Stressors examined so far differ in their recency and timing across the lifespan (e.g. recent stressors versus stressors that occurred in early life) and in their severity (traumatic versus more normative), which make research findings not easily comparable across studies. Moreover, some studies have indicated that chronic hyperactivity of the HPA-axis in response to early chronic or traumatic stressors results in a possible adaptational pattern of decreased cortisol levels in later life. This may explain why also lower cortisol levels have been found among persons with stress-related conditions like chronic fatigue syndrome, fibromyalgia and posttraumatic stress disorder (Heim et al., 2000a; Fries et al., 2005; Miller et al., 2007; Pervanidou, 2008) and would suggest that the impact of early-life stressors (such as childhood trauma) may even be reversed as compared to recent stressors that take place in later life (such as job stressors or recent life stressors) (Gerritsen et al., 2010). To assess the negative effects of work stressors, the job demand-control model is frequently used. This model assumes that job strain is the combination of high job demands with low job control, also called decision latitude. Job strain has been positively associated with morning cortisol (Steptoe et al., 2000; Alderling et al., 2006) and CAR levels (Maina et al., 2009a,b). The CAR also has been positively associated with high job demands in men and with low decision latitude in women (Maina et al., 2008). Furthermore, higher evening saliva cortisol levels were associated with high job strain (Rystedt et al., 2008). In contrary, others found higher saliva cortisol levels among subjects reporting low job strain (Steptoe et al., 1998; Fujiwara et al., 2004). Also when considering other theoretical concepts for job stress such as the effort reward imbalance model, overall results for HPA-axis activity are conflicting (Chandola et al., 2010). The clearest evidence for job characteristics on HPAaxis activity in a recent meta-analysis appeared to exist between high job stressor levels and a high CAR (Chida and Steptoe, 2009). Studies on recent life events and HPA-activity are difficult to compare, since the assessment of life events vary in time. Associations have been found between increased morning cortisol and life events during the last three years (Gerritsen et al., 2010) and between increased evening cortisol and life events within a month but not within six months (Strickland et al., 2002). In contrast, another study found no associations between morning or evening cortisol and life events within one to six months (Harris et al., 2000). It has also been hypothesized that particularly childhood trauma leads to alterations in HPA-axis regulation (Nemeroff, 2004; Carpenter et al., 2007). However, results are conflicting. Some studies have indicated that persons exposed to childhood trauma in the past have indications of higher basal or reactive cortisol levels later in life (Heim et al., 2001, 2000b, 2008; Rinne et al., 2002; Faravelli et al., 2010). On the contrary, others showed negative childhood events to be associated with lower morning or reactive cortisol levels (Elzinga et al., 2008; Gerritsen et al., 2010), which could be explained by the fact that adaptational processes result in hypocortisolism later in life. These inconsistent findings have not been clearly explained yet, but it has been hypothesized that exact timing, severity of trauma, or co-occurring psychopathology such as PTSD or depressive disorder might play a moderating role in the link between trauma and HPA-axis functioning (Heim et al., 2000b, 2008; Elzinga et al., 2008). Taken together, there is evidence that different types of stressors generally are associated with HPA-axis functioning. Mainly higher HPA-axis activity has been observed for recent stressors such as job stressors and recent life events, but for childhood trauma both higher and lower HPA-axis functioning have been observed. Since results have not always been consistent, there is a need for further research. Studies on the association of stressors and HPA-activity with a large cohort are scarce, especially with a broad set of stressors and cortisol measurements and covariates. Therefore, we examined the relationship between work stressors, important life events and childhood trauma with various salivary cortisol measures (cortisol awakening response, evening level and suppression after dexamethasone), correcting for detailed covariates. 2. Methods 2.1. Study sample Data are from the Netherlands Study of Depression and Anxiety (NESDA), a large cohort study on the course of depressive and anxiety disorders. In total 2981 respondents participated in the study. Included were 652 subjects with no current or past psychiatric disorders ( healthy controls ) obtained through a screening approach conducted among 65 general practitioners, as well as 628 subjects with earlier and 1701 subjects with current episodes of depressive or

3 Work, recent stress, childhood trauma and salivary cortisol 803 anxiety disorder. The latter group was recruited through the community, the general practitioner practices using a screening approach as well as through outpatient mental health care settings in order to recruit a representative cohort of patients reflecting all treatment settings. Psychiatric diagnoses were established by using the lifetime version 2.1 of the Composite Interview Diagnostic Instrument (CIDI, Wittchen, 1994). Across recruitment setting, uniform in- and exclusion criteria were used. A general inclusion criterion was an age of 18 through 65 years. In order to maintain representativity, only two exclusion criteria used: (1) a primary clinical diagnosis of a psychiatric disorder not subject of NESDA which will largely affect course trajectory such as psychotic disorder, obsessive compulsive disorder, bipolar disorder, or severe addiction disorder, and (2) not being fluent in Dutch since language problems would harm the validity and reliability of collected data. For objectives and more on methods of NESDA see Penninx et al. (2008). The research protocol was approved by the Ethical Committee of participating universities and all respondents provided written informed consent. We subsequently excluded a total of 19 pregnant or breastfeeding women, 113 participants on corticosteroids and 71 participants on tricyclic antidepressants because of their demonstrated effects on the HPA-axis, leaving a sample of 2778 participants. Of these 2000 (72.0%) returned saliva samples, of whom 1829 (90.6%) returned all seven samples, 131 (6.5%) six, 24 (1.2%) five and 16 (0.7%) four samples or less. Responders on saliva collection did not differ from non-responders in gender, job stressors and childhood trauma exposure, but were older (43.6 versus 38.3 years, p <.001), more educated (12.3 versus 11.8 years, p <.001), had less negative life events the past year (0.59 versus 0.74, p <.001) and were less likely to have an anxiety disorder (41.0% versus 49.7%, p <.001) or depressive disorder (38.6% versus 46.0%, p <.001) Measurements Stressors Working conditions and work stress. Only participants currently working at least 8 h a week (63.2%) were included for analysis on working conditions. Based on Karasek s demands/control model, perceived working conditions were measured with a Dutch questionnaire based on the Job Content Questionnaire (Houtman, 1995). The questionnaire consisted of 29 dichotomous items, forming five scales: job demands (5 items; Cronbach s a =.77), decision latitude or job control (13 items; Cronbach s a =.78), social support at work (8 items; Cronbach s a =.83) and job insecurity (3 items; Cronbach s a =.78). Job strain is based on the median split of job demands and decision latitude. Conform the quadrant definition, participants perceiving high job strain (high demands combined with low decision latitude) were distinguished from the others, forming a dichotomous variable (Landsbergis et al., 1994) Important life events. The number of important life events during the past year, were recorded by an interview with the list of threatening experiences containing 12 life event categories with long-term contextual threat plus the possibility to declare an other category (Brugha et al., 1985). Defined categories were a serious illness or injury to subject; serious illness or injury to a close relative; death of first-degree relative; death of close family friend ore seconddegree relative; separation due to marital difficulties; broke off a steady relationship; serious problem with a close friend, neighbor or relative; unemployed or seeking work for more than one month; subject sacked from job; major financial crisis; problems with police and court appearance; something valuable lost or stolen Childhood trauma. Traumas during childhood were inventoried by a structured interview using the NEMESIS questionnaire (de Graaf et al., 2002). Respondents were asked whether they had experienced emotional neglect, psychological abuse, physical abuse or sexual abuse on one or more occasions before the age of 16. Respondents were asked to report these adversities with the questions: Do you consider that in your case you can speak of emotional neglect? psychological abuse? physical abuse? sexual abuse? Emotional neglect was described to respondents as: At home one did not listen to you, your experiences or problems were ignored, you felt that you could not get your parent s attention and support. Psychological abuse was described as: You were cursed, unjustly punished, your brothers and sisters were favored or you were cheated on. Physical abuse was described as: You were kicked, hit with or without an object, or you were physically maltreated in any other way. Sexual abuse was defined as: You were sexually touched against your will, or you were forced to sexually touch the other. On each category participant were also asked how often these adversities occurred by choosing between: once, sometimes, regularly, often and very often. We considered the experience of an adversity more than once (i.e. sometimes or more often) as an indication for the presence of a childhood trauma as compared to no abuse or once. In line with van Harmelen et al. (2010), we defined childhood emotional abuse as the presence of emotional neglect or psychological abuse more than once (i.e. sometimes or more often) as compared to no abuse. Besides considering the types of childhood trauma as described above, also a cumulative index, called childhood trauma index, was calculated as the sum of experienced number and frequency (0 never, 1 once or sometimes, 2 regularly to very often) of childhood trauma for each participant ranging from 0 to 8 (Wiersma et al., 2009). In our earlier studies, we described associations between childhood abuse with anxiety and depression, (Hovens et al., 2010) chronicity of depression (Wiersma et al., 2009) and independent of psychopathology reduced medial prefrontal cortex volume (van Harmelen et al., 2010) Salivary cortisol As described in more detail elsewhere (Vreeburg et al., 2009a), respondents were instructed to collect saliva samples at home on a regular (working) day shortly after the interview. The median time between the interview and saliva sampling was 9 days (25th 75th percentile: 5 23). The median time between saliva sampling and the job content questionnaire was 2 days (25th 75th percentile: 0 10). Saliva samples were obtained using Salivettes (Sarstedt, Germany) at seven time points; at awakening (T1) and 30 (T2), 45 (T3) and 60 (T4) min later, at 2200 h (T5) and 2300 h (T6), and the next morning at awakening (T7) after ingestion

4 804 M. Holleman et al. of 0.5 mg dexamethasone, taken directly after the saliva sample at 2300 h (T6). After return by mail, samples were stored at 80 8C. Cortisol analysis was performed by competitive electrochemiluminescence immunoassay (E170 Roche, Switzerland). The functional detection limit was 2.0 nmol/l and the intra- and interassay variability coefficients in the measuring range were less than 10%. In data cleaning, missings were assigned to cortisol values >2 SDs above the mean. This way we eliminated respectively 12 (0.6%), 24 (1.2%), 31 (1.6%), 23 (1.2%), 16 (0.8%), 12 (0.6%) and 13 (0.7%) samples from T1 to T h awakening cortisol. We calculated the area under the curve with respect to the increase (AUCi) and with respect to the ground (AUCg) (Pruessner et al., 2003). The AUCg is an estimate of the total cortisol secretion and predicts mean cortisol levels throughout the day, whereas the AUCi is a measure of the dynamic of the cortisol awakening response (CAR), more related to the sensitivity of the system, emphasizing changes over time (Schmidt-Reinwald et al., 1999; Edwards et al., 2001; Pruessner et al., 2003; Fekedulegn et al., 2007). Morning cortisol analyses included all persons with four valid morning cortisol values (n = 1680) Evening cortisol. Since cortisol levels at 22h00 and 23h00 strongly correlated (r = 0.74, p <.01), we used the mean of both cortisol levels (n = 1984) Dexamethasone suppression test (DST). This test provides information on the negative feedback system of the HPA-axis, since dexamethasone reduces cortisol level by acting on the pituitary (Carroll et al., 1981). Sample T7 was provided by 1916 subjects. Of these, 1870 subjects reported they had taken the 0.5 mg dexamethasone. As an indicator for suppression a cortisol suppression ratio (T1/T7) was calculated (n = 1845). In sum, 1995 respondents provided at least one usable saliva measurement and constitute the present study sample Covariates Effects of sociodemographics (sex, age), sampling factors (awakening time, working, season, sleep duration) and health indicators (cardiovascular disease, physical activity and smoking) on salivary cortisol variables are found in former analyses of our data (Vreeburg et al., 2009a). These identified determinants are considered as covariates. Respondents reported time of awakening and working status (yes/no) on the sampling day. Season was categorized in less (October through February) and more daylight (March through September) months. Average sleep duration during the past four weeks was dichotomized as < = 6 or > 6 h/ night, and smoking status as current versus no smoker. Physical activity was assessed using the International Physical Activity Questionnaire and expressed in 1000 MET-minutes (metabolic energy turnover per minute) a week (Craig et al., 2003). Cardiovascular disease (CVD) included stroke, angina pectoris, myocardial infarction, percutaneous transluminal coronary angioplasty and coronary artery bypass grafting and was adjudicated using standardized algorithms considering self-report and medication use (based on drug container inspection and ATC coding). 3. Statistical analyses AUCg and AUCi showed normal distributions. The distribution of evening cortisol and DST were skewed and therefore logtransformed for analyses. Multivariable linear regression analyses were performed; each stressor was entered in separate analyses without and with adjustment for covariates. Since we earlier described increased CAR and evening cortisol levels among persons with depressive and anxiety disorders (Vreeburg et al., 2009b, 2010a) we verified whether adding psychopathology as covariate (yes or no having a current diagnosis of depressive or anxiety disorder) or adding an interaction term of psychopathology * stressor made a difference. Analyses were conducted using SPSS version Results Sample characteristics of the 1995 respondents are presented in Table 1. The mean age was 43.1 years (SD = 13.1, range 18 65) and 65.1% was female persons (63.2%) were employed during the period of the interview and 1234 worked on the day of cortisol sampling persons (76%) had a current or remitted depression and/or anxiety disorder. In Table 2 the results of the multivariable regression analyses are presented. Unadjusted and adjusted analyses yielded largely similar results. Adjusted analyses found significant associations for social support at work and cortisol suppression after dexamethasone ingestion (b =.06, p =.05) suggesting that persons with more social support are less likely to suppress cortisol after dexamethasone ingestion. A similar type of association was found for job strain with cortisol suppression (b =.06, p =.03), suggesting that persons with more job strain are more likely to suppress cortisol after dexamethasone ingestion. None of the other stressors showed any association with cortisol variables. We checked whether adding a variable for depressive or anxiety disorders (which we previously associated with higher CAR and evening cortisol levels (Vreeburg et al., 2009a, 2010a)) changed any of our reported findings in Table 2. However, no different associations emerged. In addition, we could not demonstrate the presence of moderation by psychopathology since entering stressor * psychopathology interaction terms did not yield any significant interaction terms for the four cortisol outcomes Post hoc analyses We checked whether the impact of job strain would be stronger when this variable was calculated as a continuous variable by using the quotient method that divides job demands by decision latitude. For this continuous indicator, we could not confirm any significant associations with cortisol variables. Since it has been shown that age and gender may have an impact on the relation between stress and cortisol (Kudielka et al., 2004; Otte et al., 2005) we systematically tested for potential gender and age interactions in our results. We added gender * stressor and age * stressor interaction terms for all stressor variables examined (n = 10) for each of the four main cortisol variables (AUCg, AUCi, Evening, DST). Only four of the 80 interaction terms yielded significant findings,

5 Work, recent stress, childhood trauma and salivary cortisol 805 Table 1 Sample characteristics; N = % Female 65.1 Age (mean years, SD) 43.2 (13.1) Years of education (mean, SD) 12.4 (3.3) Time of awakening (mean, SD) 7h30 (1h12) % Employed 63.2 % Working on a day of sampling 59.1 % sampling in month with more daylight 55.0 % 6 h of sleep 30.1 % Smoking 33.1 Physical activity 3.7 (3.1) (mean MET-minutes 1000, SD) % Cardiovascular disease 5.2 Psychopathology % Lifetime depression and/or anxiety 76.0 % Current depression and/or anxiety a 54.7 Dystymia 8.7 MDD 33.8 Social phobia 21.3 Panic disorder 20.1 GAD 13.2 Stress indicators Number of negative life events in the past year % One event 22.6 % Two or more events 4.7 % Job strain (high demands with low control) b % Childhood emotional abuse 39.1 % Childhood physical abuse 11.8 % Childhood sexual abuse 9.6 Cortisol characteristics Cortisol awakening response (CAR) AUCg (mean, SD) 19.1 (7.0) AUCi (mean, SD) 2.3 (6.3) Evening cortisol Mean evening cortisol (mean, SD) 5.4 (3.4) Dexamethasone suppression test Cortisol suppression ratio (mean, SD) c 2.8 (1.7) Abbreviations: SD = standard deviation, MET-minutes = metabolic energy turnover per minute, AUCg = area under the morning curve with respect to the ground, AUCi = area under the morning curve with respect to the increase. a Current depression and/or anxiety during the past six months. b Percentage of employed participants N = c Cortisol suppression ratio = cortisol at awakening/cortisol at awakening after dexamethasone ingestion. which is what can be expected on the basis of statistical chance alone. Consequently, we conclude that there was no convincing evidence for gender or age interaction in the impact of stressors on HPA-axis functioning. Subsequently, we tested whether the impact of early life adversity could have created vulnerability for later stressor exposures as suggested by Heim et al. (2002) and Elzinga et al. (2008). For this purpose, we tested six interaction terms for the childhood trauma index with 5 job stressor variables and the recent life event variable. Significant interactions were found for childhood trauma index with decision latitude on evening cortisol (b =.25, p =.004), the AUCi (b =.23, p =.02) and AUCg (b =.21, p =.03), of the childhood trauma index with social support on AUCi (b =.19, p =.01) and AUCg (b =.21, p =.001) and of the childhood trauma index with job demands on AUCi (b =.13, p =.03). Those with more childhood trauma had lower cortisol values when reporting more job demands and higher cortisol values when experiencing job control or social support at work. 5. Discussion This study examined the association of working conditions and stressors with HPA axis activity, while adjusting for important covariates. Overall, we observed few significant associations, suggesting that the stressors examined were not convincingly related to salivary cortisol levels. The only significant associations after adjustment for covariates were found for low social support at work and high job strain with more cortisol suppression after dexamethasone ingestion. However, in this large cohort study where we examined ten stressors on four cortisol outcomes, two results with a p-value under the significance threshold of.05 is what can be expected on the basis of chance. Consequently, this study could not convincingly confirm associations between various types of stressors with HPA-axis functioning. Earlier results of studies on the association of work stressors with HPA axis activity have not been consistent, but associations of high job strain have been found with an elevated CAR, higher morning cortisol or evening cortisol (Steptoe et al., 2000; Alderling et al., 2006; Rystedt et al., 2008; Maina et al., 2009a, 2009b). Others found inverse associations (Steptoe et al., 1998; Fujiwara et al., 2004). Most of these studies were not based on a large population but had specific populations (e.g. only call-center employees), which makes them hard to compare to our study sample. Also, most studies did not consider such a broad spectrum of covariates, although we also could not reveal many associations in unadjusted analyses. In a meta-analysis of the CAR and psychosocial factors, Chida and Steptoe (2009) found a positive association between job stressor level and the CAR. They also observed a negative association between the CAR and fatigue, burn out and exhaustion, which could explain our null finding since we did not include these factors in our study. We should acknowledge that a reason for our null findings might also be that retrospective measurements of work stressor may not be very reliable, since in a study of Metzenthin et al. (2009) salivary cortisol levels seemed to be more related to subjective work stress when rated at the time of salivary sampling than when rated retrospectively. With respect to aberrant HPA axis activity reflected by differences in the amount of suppression after dexamethasone ingestion, we did find two significant associations with persons reporting less social support and reporting high job strain showed more cortisol suppression. In light of the many associations we explored, these associations should be considered as null findings. Nevertheless, it is interesting to relate them to earlier literature. Hypersuppression after dexamethasone ingestion suggests increased sensitivity of the HPA axis to negative feedback and is hypothesized to be a mechanism behind hypocortisolism, possibly after chronic stress (Heim et al., 2000a). Although to our knowledge there is no comparable study on the relationship of job

6 806 M. Holleman et al. Table 2 Results of multivariable analyses associating stress measurements with cortisol indicators a : regression coefficients (b) and p-values. AUCg AUCi Evening DST Unadjusted Adjusted a Unadjusted Adjusted a Unadjusted Adjusted a Unadjusted Adjusted a Working conditions b p b p b p b p b p b p b p b p N = 1048 N = 1048 N = 1252 N = 1174 Job demands Social support Job insecurity Decision latitude Job strain N = 1680 N = 1680 N = 1984 N = 1845 No. negative life events Emotional abuse Physical abuse Sexual abuse Trauma index Abbreviations: AUCg = area under the morning curve with respect to the ground, AUCi = area under the morning curve with respect to the increase, Evening = mean cortisol at 22h00 and/or 23h00, DST = cortisol suppression ratio (salivary cortisol T1/salivary cortisol T7). a Adjusted for sex, age, education, working, time of awakening, sleep, month with more daylight, smoking, physical activity, and cardiovascular disease. characteristics and DST, Pruessner et al. (1999) did find an association of burnout with more suppression. Burnout might be seen as a consequence of chronic work stressors (Maslach and Jackson, 1981). It is likely that a considerable part of the persons in our study with low social support and high job strain might have suffered from chronic burnout feelings, and therefore the two significant associations we found could be in line with earlier findings that especially chronic work stressors are associated with more suppression after DST (Miller et al., 2007). Despite all this, we have to acknowledge that in line with our overall interpretation of null findings a recent review could not show evidence that work stress measured longitudinally was associated with lower cortisol reactivity (Chandola et al., 2010). With respect to recent life events we did not find any associations with the cortisol measurements. This is in concordance with another study (Harris et al., 2000), but others did find associations for morning cortisol or evening cortisol (Strickland et al., 2002; Gerritsen et al., 2010). Results with former studies are difficult to compare since they all use different criteria for the recency definition of life events, ranging from one month to three years, which may add to heterogeneity of findings. Cortisol levels might be heightened during a stressful episode and with the passage of time the body might counter-regulate, resulting in cortisol levels below normal (Miller et al., 2007). It could theoretically be that higher cortisol levels of persons with very recent life events (e.g. one month ago) are mixed with low to normal cortisol levels of persons who experienced less recent events (e.g. eleven months ago), leaving no overall correlation between cortisol levels and life events. This idea can be supported by Strickland et al. (2002) who found increased evening cortisol with life-events within the past month and not with life-events that occurred longer than six months ago. Another explanation for our null findings for recent life events, may be that the life events considered were either too diffuse or integrating life events that had not resulted in high enough levels of stress experience. Finally, we could also not confirm that recent life events become more important for HPA-axis functioning when a person has been exposed to childhood trauma, since no convincing evidence for interactions between childhood and recent stressors were found. For the relationship of childhood trauma with HPA axis activity we did not find any significant results while others found higher basal or reactive cortisol levels later in life (Heim et al., 2000b, 2001; Rinne et al., 2002; Heim et al., 2008; Faravelli et al., 2010) or lower morning or reactive cortisol levels (Elzinga et al., 2008; Gerritsen et al., 2010). Also, here the outcomes of earlier studies are conflicting, not based on large populations and furthermore difficult to compare since various studies were based on specific psychiatric populations (e.g. PTSD). In our earlier studies, we described associations between childhood abuse with anxiety and depression (Hovens et al., 2010), chronicity of depression (Wiersma et al., 2009) and independent of psychopathology reduced medial prefrontal cortex volume (van Harmelen et al., 2010). Considering these findings, we do not have reasons to doubt the childhood trauma assessment validity. Also considering that unadjusted analyses did not reveal associations either, we believe that childhood trauma in our study truly did not impact on HPA-axis functioning. However, it remains to be determined in other studies

7 Work, recent stress, childhood trauma and salivary cortisol 807 whether exact timing and severity of childhood trauma, or co-occurring psychopathology such as PTSD might play a moderating role (Heim et al., 2000a, 2002, 2008; Elzinga et al., 2008). It has also been hypothesized that particularly the interaction of childhood trauma with recent stressors is important for the reactivity of the HPA-axis (Heim et al., 2002; Elzinga et al., 2008). Based on earlier findings, higher as well as lower cortisol levels might be expected when persons with childhood trauma are exposed to recent stressors. Our post-hoc analyses suggested that persons with more occurrence of childhood trauma had a lower CAR or evening cortisol levels when reporting more demands or less decision latitude or less social support at work. The presence of five out of 24 significant interactions although slightly more than to be expected could still implicate null findings. Alternative explanations that deserve further future exploration could be that these blunted associations are due to underlying genetic or psychological interactions since genetic factors and personality or coping styles may play a role in the individual sensitivity to life events (Bet et al., 2009). Also, a blunted association might reflect consequences of chronic stress exposure since childhood trauma is associated with a variety of psychosocial stressors later in life (Kendall-Tackett, 2002). In a previous study with the same sample, we observed a higher CAR in persons with remitted as well as current MDD and panic disorder (Vreeburg et al., 2009b, 2010a). Therefore, a higher cortisol awakening curve might represent a trait indicator for affective disorders potentially already present before the onset of these disorders. This would suggest that underlying genetic vulnerability could be responsible for these earlier observations, which is in line with the finding that also healthy controls with a family history of affective disorders already showed an elevated CAR similar to patients (Vreeburg et al., 2010b). The present findings suggest that exposure to stressors does not further differentiate HPA-axis function in our sample. It could be that stressors have more impact on HPA-axis functioning in healthy controls who do not have the same underlying increased genetic vulnerability. The absence of significant psychopathology * stressor interaction terms in our study does not provide evidence for this suggestion. Also earlier findings are inconsistent with this suggestion since Strickland et al. (2002) did not find different association in depressed or non-depressed persons, and both Heim et al. (2000b, 2008) and Yehuda et al. (2004) found stronger effects among depressed patients. To our knowledge, our study is unique with respect to its large cohort and our extensive assessment of stressors and cortisol variables as well as covariates. However, our study also had some limitations. First, non-compliance with the sampling instructions or dexamethasone ingestion could have resulted in measurement error, which may have led to underestimation of associations. However, it should be noted that even when awakening is closely monitored, still at least 15% of all persons do not respond with a cortisol rise (Dockray et al., 2008). We checked for compliance to the DST test in a subset of our cohort by measuring dexamethasone levels with a radioimmunoassay using the anti-dexamethasone antibody from IgG Corporation (Nashville, TN, functional detection limit is 0.4 nmol/l and reported cross-reactivity for cortisol is.04% (Weijtens et al., 1997)). Among 47 respondents with a T1/T7 ratio < 1.5 (indicative of non-suppression) who reported dexamethasone ingestion, we found detectable dexamethasone levels (>0.4 nmol/l) in the T7 saliva samples among 90%, indicating that non-compliance with dexamethasone ingestion is not likely to be frequent. Second, the responsiveness of the HPA axis might be associated with the severity, duration, and exact timing of onset of the stressor (Heim et al., 2008) which we could not further consider. Third, retrospective measurements of stressors might not be reliable and mix up chronic and acute stressors. However, our assessments of stressors were based on widely used and validated assessments, and were earlier found to be associated with the development and chronicity of psychopathology (Wiersma et al., 2009; Hovens et al., 2010; Spinhoven et al., 2011). Fourth, the group of healthy persons might be undersized in our sample and therefore our results might not be representative for the general population. All in all, we conclude that in this large cohort study the relationship between childhood trauma, recent negative life events and work stressors with different indicators of saliva cortisol could not be convincingly established in a mixed population of depressed/anxious and healthy individuals. These null findings could indicate that the impact of stressors on HPA-axis functioning becomes visible only when more detailed aspects of severity, chronicity and timing of stressors are considered. Role of funding source The infrastructure for the NESDA study is funded through the Geestkracht program of the Netherlands Organisation for Health Research and Development; this organisation has no further role in study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication. Conflict of interest All authors reported no conflict or financial competing interests. Acknowledgements The infrastructure for the NESDA study ( is funded through the Geestkracht program of the Netherlands Organisation for Health Research and Development (Zon-Mw, grant number ) and is supported by participating universities and mental health care organisations (VU University Medical Center, GGZ ingeest, Arkin, Leiden University Medical Center, GGZ Rivierduinen, University Medical Center Groningen, Lentis, GGZ Friesland, GGZ Drenthe, IQ Healthcare, Netherlands Institute for Health Services Research (NIVEL) and Netherlands Institute of Mental Health and Addiction (Trimbos). References Alderling, M., Theorell, T., de la Torre, B., Lundberg, I., The demand control model and circadian saliva cortisol variations in a

8 808 M. Holleman et al. Swedish population based sample (The PART study). BMC Public Health 6, 288. Bet, P.M., Penninx, B.W.J.H., Bochdanovits, Z., Uitterlinden, A.G., Beekman, A.T.F., van Schoor, N.M., Deeg, D.J.H., Hoogendijk, W.J.G., Glucocorticoid receptor gene polymorphisms and childhood adversity are associated with depression: new evidence for a gene-environment interaction. Am. J. Med. Genet. B: Neuropsychiatr. Genet. 150B, Brugha, T., Bebbington, P., Tennant, C., Hurry, J., The list of threatening experiences: a subset of 12 life event categories with considerable long-term contextual threat. Psychol. Med. 15, Carpenter, L.L., Carvalho, J.P., Tyrka, A.R., Wier, L.M., Mello, A.F., Mello, M.F., Anderson, G.M., Wilkinson, C.W., Price, L.H., Decreased adrenocorticotropic hormone and cortisol responses to stress in healthy adults reporting significant childhood maltreatment. Biol. Psychiatry 62, Carroll, B.J., Feinberg, M., Greden, J.F., Tarika, J., Albala, A.A., Haskett, R.F., James, N.M., Kronfol, Z., Lohr, N., Steiner, M., de Vigne, J.P., Young, E., A specific laboratory test for the diagnosis of melancholia. Standardization, validation, and clinical utility. Arch. Gen. Psychiatry 38, Chandola, T., Heraclides, A., Kumari, M., Psychophysiological biomarkers of workplace stressors. Neurosci. Biobehav. Rev. 35, Chida, Y., Steptoe, A., Cortisol awakening response and psychosocial factors: a systematic review and meta-analysis. Biol. Psychol. 80, Coste, J., Strauch, G., Letrait, M., Bertagna, X., Reliability of hormonal levels for assessing the hypothalamic pituitary adrenocortical system in clinical pharmacology. Br. J. Clin. Pharmacol. 38, Craig, C.L., Marshall, A.L., Sjöström, M., Bauman, A.E., Booth, M.L., Ainsworth, B.E., Pratt, M., Ekelund, U., Yngve, A., Sallis, J.F., Oja, P., International physical activity questionnaire: 12- country reliability and validity. Med. Sci. Sports Exerc. 35, Dekker, M.J., Koper, J.W., van Aken, M.O., Pols, H.A., Hofman, A., de Jong, F.H., Kirschbaum, C., Witteman, J.C., Lamberts, S.W., Tiemeier, H., Salivary cortisol is related to atherosclerosis of carotid arteries. J. Clin. Endocrinol. Metab. 93, Dockray, S., Bhattacharyya, M.R., Molloy, G.J., Steptoe, A., The cortisol awakening response in relation to objective and subjective measures of waking in the morning. Psychoneuroendocrinology 33, Edwards, S., Clow, A., Evans, P., Hucklebridge, F., Exploration of the awakening cortisol response in relation to diurnal cortisol secretory activity. Life Sci. 68, Eller, N.H., Netterstrom, B., Allerup, P., Progression in intima media thickness the significance of hormonal biomarkers of chronic stress. Psychoneuroendocrinology 30, Elzinga, B.M., Roelofs, K., Tollenaar, M.S., Bakvis, P., van Pelt, J., Spinhoven, P., Diminished cortisol responses to psychosocial stress associated with lifetime adverse events a study among healthy young subjects. Psychoneuroendocrinology 33, Faravelli, C., Amedei, S.G., Rotella, F., Faravelli, L., Palla, A., Consoli, G., Ricca, V., Batini, S., Sauro, C.L., Spiti, A., Dell osso, M.C., Childhood traumata. Dexamethasone suppression test and psychiatric symptoms: a trans-diagnostic approach. Psychol. Med. 40, Fekedulegn, D.B., Andrew, M.E., Burchfiel, C.M., Violanti, J.M., Hartley, T.A., Charles, L.E., Miller, D.B., Area under the curve and other summary indicators of repeated waking cortisol measurements. Psychosom. Med. 69, Fries, E., Hesse, J., Hellhammer, J., Hellhammer, D.H., A new view on hypocortisolism. Psychoneuroendocrinology 30, Fries, E., Dettenborn, L., Kirschbaum, C., The cortisol awakening response (CAR): facts and future directions. Int. J. Psychophysiol. 72, Fujiwara, K., Tsukishima, E., Kasai, S., Masuchi, A., Tsutsumi, A., Kawakami, N., Miyake, H., Kishi, R., Urinary catecholamines and salivary cortisol on workdays and days off in relation to job strain among female health care providers. Scand. J. Work Environ. Health 30, Gerritsen, L., Geerlings, M.I., Beekman, A.T., Deeg, D.J., Penninx, B.W., Comijs, H.C., Early and late life events and salivary cortisol in older persons. Psychol. Med. 40, de Graaf, R., Bijl, R.V., Smit, F., Vollebergh, W.A., Spijker, J., Risk factors for 12-month comorbidity of mood, anxiety, and substance use disorders: findings from the Netherlands Mental Health Survey and Incidence Study. Am. J. Psychiatry 159, van Harmelen, A.L., de Jong, P.J., Glashouwer, K.A., Spinhoven, P., Penninx, B.W., Elzinga, B.M., Child abuse and negative explicit and automatic self-associations: the cognitive scars of emotional maltreatment. Behav. Res. Ther. 48, Harris, T.O., Borsanyi, S., Messari, S., Stanford, K., Cleary, S.E., Shiers, H.M., Brown, G.W., Herbert, J., Morning cortisol as a risk factor for subsequent major depressive disorder in adult women. Br. J. Psychiatry 177, Heim, C., Ehlert, U., Hellhammer, D.H., 2000a. The potential role of hypocortisolism in the pathophysiology of stress-related bodily disorders. Psychoneuroendocrinology 25, Heim, C., Newport, D.J., Heit, S., Graham, Y.P., Wilcox, M., Bonsall, R., Miller, A.H., Nemeroff, C.B., 2000b. Pituitary adrenal and autonomic responses to stress in women after sexual and physical abuse in childhood. JAMA 284, Heim, C., Newport, D.J., Bonsall, R., Miller, A.H., Nemeroff, C.B., Altered pituitary adrenal axis responses to provocative challenge tests in adult survivors of childhood abuse. Am. J. Psychiatry 158, Heim, C., Newport, D.J., Wagner, D., Wilcox, M.M., Miller, A.H., Nemeroff, C.B., The role of early adverse experience and adulthood stress in the prediction of neuroendocrine stress reactivity in women: a multiple regression analysis. Depress. Anxiety 15, Heim, C., Mletzko, T., Purselle, D., Musselman, D.L., Nemeroff, C.B., The dexamethasone/corticotropin-releasing factor test in men with major depression: role of childhood trauma. Biol. Psychiatry 63, Houtman, I., Reliability and Validity of the Dutch Version of the Karasek Job Content Questionnaire. APA, Wahington, DC. Hovens, J.G., Wiersma, J.E., Giltay, E.J., van Oppen, P., Spinhoven, P., Penninx, B.W., Zitman, F.G., Childhood life events and childhood trauma in adult patients with depressive, anxiety and comorbid disorders vs. controls. Acta Psychiatr. Scand. 122, Kendall-Tackett, K., The health effects of childhood abuse: four pathways by which abuse can influence health. Child Abuse Negl. 26, Kudielka, B.M., Buske-Kirschbaum, A., Hellhammer, D.H., Kirschbaum, C., HPA axis responses to laboratory psychosocial stress in healthy elderly adults, younger adults, and children: impact of age and gender. Psychoneuroendocrinology 29, Landsbergis, P.A., Schnall, P.L., Warren, K., Pickering, T.G., Schwartz, J.E., Association between ambulatory blood pressure and alternative formulations of job strain. Scand. J. Work. Environ. Health 20, Maina, G., Palmas, A., Filon, F.L., Relationship between selfreported mental stressors at the workplace and salivary cortisol. Int. Arch. Occup. Environ. Health 81, Maina, G., Palmas, A., Bovenzi, M., Filon, F.L., 2009a. Salivary cortisol and psychosocial hazards at work. Am. J. Ind. Med. 52,

9 Work, recent stress, childhood trauma and salivary cortisol 809 Maina, G., Bovenzi, M., Palmas, A., Filon, F.L., 2009b. Associations between two job stress models and measures of salivary cortisol. Int. Arch. Occup. Environ. Health 82, Maslach, C., Jackson, S.E., The measurement of experienced burnout. J. Organizat. Behav. 2, Metzenthin, P., Helfricht, S., Loerbroks, A., Terris, D.D., Haug, H.J., Subramanian, S.V., Fischer, J.E., A one-item subjective work stress assessment tool is associated with cortisol secretion levels in critical care nurses. Prev. Med. 48, Miller, G.E., Chen, E., Zhou, E.S., If it goes up, must it come down? Chronic stress and the hypothalamic pituitary adrenocortical axis in humans. Psychol. Bull. 133, Nemeroff, C.B., Neurobiological consequences of childhood trauma. J. Clin. Psychiatry 65, Otte, C., Hart, S., Neylan, T.C., Marmar, C.R., Yaffe, K., Mohr, D.C., A meta-analysis of cortisol response to challenge in human aging: importance of gender. Psychoneuroendocrinology 30, Penninx, B.W., Beekman, A.T., Smit, J.H., Zitman, F.G., Nolen, W.A., Spinhoven, P., Cuijpers, P., de Jong, P.J., van Marwijk, H.W., Assendelft, W.J., van der Meer, K., Verhaak, P., Wensing, M., de Graaf, R., Hoogendijk, W.J., Ormel, J., van Dyck, R., The Netherlands study of depression and anxiety (NESDA): rationale, objectives and methods. Int. J. Methods Psychiatr. Res. 17, Pervanidou, P., Biology of post-traumatic stress disorder in childhood and adolescence. J. Neuroendocrinol. 20, Pruessner, J.C., Hellhammer, D.H., Kirschbaum, C., Burnout, perceived stress, and cortisol responses to awakening. Psychosom. Med. 61, Pruessner, J.C., Kirschbaum, C., Meinlschmid, G., Hellhammer, D.H., Two formulas for computation of the area under the curve represent measures of total hormone concentration versus timedependent change. Psychoneuroendocrinology 8, Reynolds, R.M., Labad, J., Strachan, M.W., Braun, A., Fowkes, F.G., Lee, A.J., Frier, B.M., Seckl, J.R., Walker, B.R., Price, J.F., Elevated fasting plasma cortisol is associated with ischemic heart disease and its risk factors in people with type 2 diabetes: the Edinburgh type 2 diabetes study. J. Clin. Endocrinol. Metab. 95, Rinne, T., de Kloet, E.R., Wouters, L., Goekoop, J.G., DeRijk, R.H., van den Brink, W., Hyperresponsiveness of hypothalamic pituitary adrenal axis to combined dexamethasone/corticotropin-releasing hormone challenge in female borderline personality disorder subjects with a history of sustained childhood abuse. Biol. Psychiatry 52, Rystedt, L.W., Cropley, M., Devereux, J.J., Michalianou, G., The relationship between long-term job strain and morning and evening saliva cortisol secretion among white-collar workers. J. Occup. Health Psychol. 13, Sapolsky, R.M., Why Zebras Don t Get Ulcers: An Updated Guide to Stress, Stress-related Diseases, and Coping. W.H. Freeman and Co., New York. Schmidt-Reinwald, A., Pruessner, J.C., Hellhammer, D.H., Federenko, I., Rohleder, N., Schurmeyer, T.H., Kirschbaum, C., The cortisol response to awakening in relation to different challenge tests and a 12-h cortisol rhythm. Life Sci. 64, Schoorlemmer, R.M., Peeters, G.M., van Schoor, N.M., Lips, P., Relationships between cortisol level, mortality and chronic diseases in older persons. Clin. Endocrinol. 71, Spinhoven, P., Elzinga, B.M., Hovens, J.G., Roelofs, K., van Oppen, P., Zitman, F.G., Penninx, B.W., in press. Positive and negative life events and personality traits in predicting course of depression and anxiety. Acta Psychiatr. Scand. doi: /j x. Steptoe, A., Cropley, M., Griffith, J., Kirschbaum, C., Job strain and anger expression predict early morning elevations in salivary cortisol. Psychosom. Med. 62, Steptoe, A., Wardle, J., Lipsey, Z., Mills, R., Oliver, G., Jarvis, M., Kirschbaum, C., A longitudinal study of work load and variations in psychological well-being, cortisol, smoking, and alcohol consumption. Ann. Behav. Med. 20, Strickland, P.L., Deakin, J.F., Percival, C., Dixon, J., Gater, R.A., Goldberg, D.P., Bio-social origins of depression in the community. Interactions between social adversity, cortisol and serotonin neurotransmission. Br. J. Psychiatry 180, Vreeburg, S.A., Kruijtzer, B.P., van Pelt, J., van Dyck, R., DeRijk, R.H., Hoogendijk, W.J., Smit, J.H., Zitman, F.G., Penninx, B.W., 2009a. Associations between sociodemographic, sampling and health factors and various salivary cortisol indicators in a large sample without psychopathology. Psychoneuroendocrinology 34, Vreeburg, S.A., Hoogendijk, W.J., van Pelt, J., Derijk, R.H., Verhagen, J.C., van Dyck, R., Smit, J.H., Zitman, F.G., Penninx, B.W., 2009b. Major depressive disorder and hypothalamic pituitary adrenal axis activity: results from a large cohort study. Arch. Gen. Psychiatry 66, Vreeburg, S.A., Zitman, F.G., van Pelt, J., Derijk, R.H., Verhagen, J.C., van Dyck, R., Hoogendijk, W.J., Smit, J.H., Penninx, B.W., 2010a. Salivary cortisol levels in persons with and without different anxiety disorders. Psychosom. Med. 72, Vreeburg, S.A., Hartman, C.A., Hoogendijk, W.J., van Dyck, R., Zitman, F.G., Ormel, J., Penninx, B.W., 2010b. Parental history of depression or anxiety and the cortisol awakening response. Br. J. Psychiatry 197, Weijtens, O., van der Sluijs, F.A., Schoemaker, R.C., Lentjes, E.G., Cohen, A.F., Romijn, F.P., van Meurs, J.C., Peribulbar corticosteroid injection: vitreal and serum concentrations after dexamethasone disodium phosphate injection. Am. J. Ophthalmol. 123, Wiersma, J.E., Hovens, J.G., van Oppen, P., Giltay, E.J., van Schaik, D.J., Beekman, A.T., Penninx, B.W., The importance of childhood trauma and childhood life events for chronicity of depression in adults. Clin. Psychiatry 7, Wilhelm, I., Born, J., Kudielka, B.M., Schlotz, W., Wust, S., Is the cortisol awakening rise a response to awakening? Psychoneuroendocrinology 32, Wittchen, H.U., Reliability and validity studies of the WHO composite international diagnostic interview (CIDI): a critical review. J. Psychiatr. Res. 28, Wüst, S., Wolf, J., Hellhammer, D.H., Federenko, I., Schommer, N., Kirschbaum, C., The cortisol awakening response normal values and confounds. Noise Health 2, Yehuda, R., Halligan, S.L., Golier, J.A., Grossman, R., Bierer, L.M., Effects of trauma exposure on the cortisol response to dexamethasone administration in PTSD and major depressive disorder. Psychoneuroendocrinology 29,