Accepted Manuscript Title: Elevated hair cortisol levels in chronically stressed dementia caregivers Author: Tobias Stalder Antje Tietze Susann Steudte Nina Alexander Lucia Dettenborn Clemens Kirschbaum PII: DOI: Reference:
S0306-4530(14)00154-1 http://dx.doi.org/doi:10.1016/j.psyneuen.2014.04.021 PNEC 2686
To appear in: Received date: Revised date: Accepted date:
25-2-2014 4-4-2014 25-4-2014
Please cite this article as: Stalder, T., Tietze, A., Steudte, S., Alexander, N., Dettenborn, L., Kirschbaum, C.,Elevated hair cortisol levels in chronically stressed dementia caregivers, Psychoneuroendocrinology (2014), http://dx.doi.org/10.1016/j.psyneuen.2014.04.021 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Elevated hair cortisol levels in chronically stressed dementia caregivers
Tobias Stalder a*†, Antje Tietzea*, Susann Steudtea, Nina Alexandera, Lucia Dettenbornb, Clemens Kirschbauma
Department of Medical Psychology, University Hospital Hamburg-Eppendorf, Germany
†
an
us
b
Technische Universität Dresden, Department of Psychology, Dresden, Germany
cr
a
ip t
(* shared first authorship)
Corresponding author & reprints:
Dr. Tobias Stalder
Technische Universität Dresden
M
Department of Psychology Andreas-Schubert-Bau Zellescher Weg 19
d
01069 Dresden, Germany Fax: +49 351 463 -37274
te
Tel: +49 351 463 -39678
Ac ce p
E-mail: [email protected]
Running title: Hair cortisol in dementia caregivers
Keywords: Cortisol; hair; stress; chronic; caregiver; depression; human.
Word count: 1997
1
Page 1 of 14
Abstract: Hair cortisol concentrations (HCC) are assumed to reflect integrated long-term cortisol levels and have been proposed as a promising endocrine marker of chronic psychological stress. The current study examined HCC in relation to caregiving burden, a well-established naturalistic model of chronic stress in humans. HCC and relevant psychosocial data were examined in 20 caregivers of relatives with
ip t
dementia and 20 non-caregiver controls matched for age and sex. Results revealed elevated HCC in dementia caregivers compared to non-caregiver controls (F(1,38) = 4.4, p = .04, ηp2 = .10). Further, within caregivers, a trend for a positive association of HCC with self-reported caregiving burden (r =
cr
.43, p = .058) and a positive association with depressiveness (r = .48, p = .045) were observed. No other associations between HCC and subjective measures were seen. These findings concur with the
Ac ce p
te
d
M
an
us
notion that HCC sensitively capture endocrine aberrations in stress-exposed groups.
2
Page 2 of 14
Introduction: Long-term changes in the secretion of the glucocorticoid cortisol are considered a crucial mediator for the link between chronic stress and ill health. To obtain valid assessments of long-term cortisol levels, however, has been methodologically challenging in the past. Cortisol measured in blood, saliva or urine exhibits high situational variability and thus provides only a poor reflection of underlying long-
ip t
term secretory patterns. Over the past decade, the analysis of hair cortisol concentrations (HCC) has increasingly gained acceptance as an easily obtainable index of integrated long-term cortisol secretion
cr
(reviews: Russell et al., 2012; Stalder & Kirschbaum, 2012). Considerable evidence has now supported this method in terms of its overall validity, test-retest reliability and robustness to potential
us
confounding influences (Stalder & Kirschbaum, 2012).
Overall, these data highlight HCC as a promising marker for endocrine correlates of chronic stress (Staufenbiel et al., 2013). However, direct enquiries in the context of chronic stress have produced
an
diverging results: while research examining relationships of HCC with subjective stress measures has failed to report consistent results, initial studies conducted in stress-exposed groups have uniformly shown elevations in HCC (Stalder & Kirschbaum, 2012). The latter evidence concerning HCC in
M
stress-exposed groups, however, is still based on a smaller number of studies carried out in specialized samples. Specifically, increased HCC were reported in hospitalized and ventilated infants, chronic pain patients, unemployed individuals, students with major life stressors (review: Staufenbiel et al.,
d
2013) and members of Aboriginal First Nation communities in Canada (Henley et al., 2013). These
te
studies feature groups in which the effect of psychological stress may not be fully distinguishable from effects of underlying medical conditions (e.g., hospitalized infants or chronic pain patients) or ethnic group differences (e.g., hair characteristics of Aboriginal people vs. whites).
Ac ce p
This highlights the need to further investigate long-term endocrine alterations in groups with welldefined increased psychosocial stress exposure. A suitable human condition for doing so is provided by dementia caregiving burden. Caring for a demented relative constitutes a highly stressful task, characterized by changing role patterns, continuous mourning for the affected relative and demanding caring duties due to the uncontrollable nature of dementia (Schoenmakers et al., 2010). Consistently, dementia caregivers have been shown to experience higher rates of psychological stress, an increased risk for psychiatric diseases and poorer health outcomes compared to other caregivers or non-caregiver peers (Pinquart & Sorensen, 2003). Further, increased activity of the endocrine stress system in dementia caregivers has been indicated (e.g., Vedhara et al., 1999), although with some inconsistency (e.g., Irwin et al., 1997). The current study thus set out to provide a first investigation of HCC in groups of dementia caregivers and non-caregiver controls matched for age and sex. We predicted to find elevated longterm cortisol levels in dementia caregivers. To further foster interpretability of results, we also assessed dimensional relationships with self-reported caregiving burden, depressiveness, social support and sleep problems. 3
Page 3 of 14
Methods: Participants: Twenty-one elderly caregivers of relatives with dementia and 22 age and sex-matched non-caregiver controls were recruited from self-help groups for dementia caregivers (caregivers) and meeting centers for the elderly (controls). The dementia caregiver group comprised individuals who were currently
ip t
caring for a close relative with a general practitioner-verified diagnosis of dementia. Caregivers in this group had been caring for their spouse (n=20) or parent (n=1) for a mean (± SD) time of 40.8 (± 30.8)
cr
months. The care settings of most participants were classified with German nursing care levels I or II which signify severe or very severe care needs, respectively. High caregiving burden in caregivers was
us
further reflected by mean (± SD) scores of 38.2 (± 13.2) on the Zarit Burden Interview (ZBI; Braun et al., 2010). A ZBI cut-off value of 26 has been proposed to mark an increased mental health risk due to high caregiving burden (Schreiner et al., 2006). The non-caregiver group comprised individuals who
an
had never cared for a demented relative.
Participants were only included in the study (both groups) if they were not taking glucocorticoidcontaining medication and had not had any medical surgery within three-months prior to hair
M
sampling. To avoid creating an artificial sample of elderly participants, we refrained from imposing further exclusion criteria based on medical diseases or medication intake, but instead controlled for these effects statistically. All participants provided written informed consent prior to taking part in the
d
study. The study protocol was approved by the ethics committee of the TU Dresden Medical Faculty
Self-report measures:
te
and the study was conducted in accordance with the Declaration of Helsinki.
Ac ce p
Information on demographic and health-related variables (age, sex, body-mass-index (BMI), marital status, regular daily alcohol consumption, chronic diseases and medication intake) and hair-related characteristics (washes/week, hair treatment: coloration, bleaching, permanent wave) were measured using a self-developed questionnaire. Two ten-point rating scales were used to assess physical and mental well-being (endpoints: very low, very high) with high values signaling high well-being. The Social Readjustment Rating Scale (SRRS; German Version: Katschnig, 1980) was used to assess major life events over the past year. Depressive symptomatology over the past two weeks was assessed using the Beck Depression Inventory (BDI-II; German Version: Hautzinger, Keller and Kühner, 2009) and the ENRICHD Social Support Inventory (ESSI; German Version: Spaderna et al., 2009) was employed to measure general perceived social support. Finally, the Jenkins Sleep Problem Questionnaire (JSPQ, Jenkins et al., 1988) was used to assess self-reported sleep problems over the previous months.
4
Page 4 of 14
Hair cortisol analysis: Hair strands (~3 mm diameter) were cut scalp-near from a posterior vertex position. The proximal 3 cm hair segment was used for analyses. Based on an average hair growth rate of 1cm/month, this hair segment represents hair grown over the 3-month period prior to hair sampling. Wash and steroid extraction procedures followed the laboratory protocol described in Kirschbaum et al. (2009) with 10
ip t
mg of powdered hair being used for analyses in the current study. Cortisol levels were determined using a commercially available immunoassay with chemiluminescence detection (CLIA, IBL-
cr
Hamburg, Germany). Intra- and interassay coefficients of variance were below 10%.
Statistical analysis:
us
Analyses were performed using SPSS for Windows, version 21 (IBM, Chicago, Illinois). Box plot analyses revealed three extreme outlying HCC values exceeding three interquartile ranges from the
an
median which were excluded from subsequent analyses (one caregiver, two non-caregiver controls). Hair cortisol data were positively skewed and log10-transformations were applied and successfully established a normal distribution. Descriptive data in text and figures are reported in original units
M
(pg/mg). Initial group comparisons were conducted using t-tests (continuous variables) and Fisher’s exact tests (dichotomous variables). HCC were compared between dementia caregivers and noncaregiver controls using one-way ANOVA and followed up by ANCOVA adjusting for variables
Ac ce p
Results:
te
examined using Pearson correlations.
d
differing between groups. Dimensional associations between self-report measures and HCC were
Table 1 shows descriptive characteristics of the two study groups. Groups were well-matched on the examined demographic, health- and hair-related parameters, except for a significantly higher rate of married individuals in dementia caregivers. As expected for the elderly, a high percentage of participants in both groups reported chronic diseases and/or regular medication intake. Diseases included hypertension (n=12), back pain (n=3), ocular disorders/glaucoma (n=5), diabetes (n=4), allergies (n=3) and osteoarthritis (n=3). Most common medication types were antihypertensives (n=24), thyroxin (n=6) and antihyperglycemics (n=3). No group differences were observed in either prevalence of chronic disease types or medication categories (all p’s > .34), except for a nonsignificant trend for a higher rate of hypertension in non-caregiver controls (Fisher’s exact, p =.08). As expected, dementia caregivers reported lower well-being (physical and mental) and social support as well as higher depressiveness. No significant differences in self-reported life events or sleep problems were observed.
[Please insert Table 1 about here]
5
Page 5 of 14
Figure 1a shows HCC of dementia caregivers and non-caregiver controls. The univariate ANOVA revealed higher HCC in dementia caregivers (mean ± SD: 27.4 ± 11.3 pg/mg) than in non-caregiver controls (mean ± SD: 20.5 ± 7.3 pg/mg; F(1,38) = 4.4, p = .04, ηp2 = .10). This effect remained significant in an ANCOVA adjusting for differences in marital status and prevalence of hypertension
= .17) or ESSI scores (F(2,37) = 2.9, p = .097) were controlled for.
ip t
(F(3,35) = 4.9, p = .03, ηp2 = .12). The effect missed statistical significance when BDI-II (F(2,31) = 2.0, p Correlation analyses carried out across groups revealed no significant associations between HCC and SRRS, BDI-II, ESSI and JSPQ (all p’s > .14). Analyses conducted within the caregiver group
cr
revealed a trend for a positive correlation of HCC with ZBI-assessed caregiving burden (r = .43, p = .058, n = 20) and a positive association with BDI-II scores (r = .48, p = .045, n = 18; see Figures 1b
an
patients’ nursing care level were observed (all p’s > .34).
us
and 1c, respectively). No further associations of HCC with self-report measures, time of caregiving or
[Please insert Figure 1 about here]
M
Discussion:
This study investigated the relationship of integrated long-term cortisol levels in hair and caregiving burden, a well-established human model of chronic psychosocial stress. Our main finding showed
d
significantly higher HCC in dementia caregivers compared to non-caregiver controls. This concurs
te
with previous data showing elevated HCC across a range of stress-related conditions (Stalder & Kirschbaum, 2012; Staufenbiel et al., 2013). Besides this group effect, analyses conducted within dementia caregivers also revealed positive relationships of HCC with perceived caregiving burden (at
Ac ce p
trend level) and depressiveness.
Our main finding of increased HCC in dementia caregivers concurs with previous reports of elevated salivary or blood cortisol in this group (e.g., Vedhara et al., 1999). Interestingly, this pattern has not unanimously been found (e.g., Irwin et al., 1997), perhaps partly due to limitations in employed cortisol assessment methods. The fact that caring for a demented relative is highly stressful has repeatedly been shown (Pinquart & Sorensen, 2003) and was also reflected in the current data showing lower physical and mental well-being in caregivers. Given that the two groups were comparable on other examined variables, it is most likely that increased HCC in caregivers were mainly attributable to repeated cortisol reactions to stressful experiences in caregivers’ daily routines. This interpretation was further supported by our tentative dimensional finding suggesting that caregivers who reported greater caregiving burden also tended to exhibit higher HCC. Besides increased stress exposure in caregivers, the fact that the HCC group effect missed significance after controlling for BDI-II or ESSI scores further suggests that higher depressiveness and lower perceived social support in this group also partly contributed to the observed long-term endocrine aberrations. A potential role of depressiveness was further highlighted by our finding of a 6
Page 6 of 14
positive association of HCC and depressiveness within caregivers, although this relationship was not seen across the whole sample. The finding within caregivers could be seen as suggestive of increased long-term cortisol secretion in more depressed individuals which has also been indicated by some but not all previous HCC research (review: Staufenbiel et al., 2013). However, as this association was only seen within the relatively small group of caregivers, further investigation is needed until firm
ip t
conclusions can be drawn.
The latter fact that associations between HCC and self-report measures were only seen within caregivers but not across the whole sample is an interesting aspect of the current findings. It can only
cr
be speculated what may be responsible for this result. We have previously suggested that, amongst other things, individual differences in self-awareness, social desirability and retrospective bias may
us
have had a negative effect on the quality of self-report data in previous HCC research (Stalder & Kirschbaum, 2012). It is conceivable that this has applied less to the current caregiver sample who
an
may have been particularly sensitized to observing their stress-related living situation through frequent participation in caregiver self-help groups. This could have resulted in more accurate self-reports and thus increased the likelihood of detecting psychoendocrine covariance in this group. Still, other
M
explanations are clearly possible and future corroboration of this effect in a larger sample is expedient. Such testing may also benefit from the use of more elaborate self-assessment methods, such as ecological momentary assessments or the experience sampling method, which may further help to
Ac ce p
te
d
raise ecological validity of self-report data across the whole sample.
7
Page 7 of 14
References: Braun, M., Scholz, U., Hornung, R., Martin, M., 2010. Die subjektive Belastung pflegender Ehepartner von Demenzkranken. Hinweise zur Validität der deutschen Version des Zarit Burden Interviews. Z. Gerontol. Geriatr., 43, 111-119.
am Main: Pearson Assessment & Information GmbH.
ip t
Hautzinger, M., Keller, F., Kühner, C., 2009. BDI-II. Beck-Depressions-Inventar. Revision. Frankfurt
Henley, P., Jahedmotlagh, Z., Thomson, S., Hill, J., Darnell, R., Jacobs, D., Johnson, J., Williams, N.C., Williams, R.M., Van Uum, S., Bend, J.R., Koren, G., 2013. Hair Cortisol as a Biomarker of
cr
Stress Among a First Nation in Canada. Ther. Drug. Monit., 35, 595-599.
Jenkins, C. D., Stanton, B. A., Niemcryk, S. J., Rose, R. M., 1988. A scale for the estimation of sleep
us
problems in clinical research. J. Clin. Epidemiol., 41, 313-321.
Irwin, M., Hauger, R., Patterson, T. L., Semple, S., Ziegler, M., Grant, I., 1997. Alzheimer caregiver
an
stress: basal natural killer cell activity, pituitary-adrenal cortical function, and sympathetic tone. Ann. Behav. Med., 19, 83-90.
Katschnig, H., 1980. Sozialer Stress und psychische Erkrankung. München: Urban & Schwarzenberg.
M
Kirschbaum, C., Tietze, A., Skoluda, N., Dettenborn, L., 2009. Hair as a retrospective calendar of cortisol production – Increased cortisol incorporation into hair in the third trimester of pregnancy. Psychoneuroendocrinology, 34, 32-37.
d
Pinquart, M., Sorensen, S., 2003. Differences between caregivers and noncaregivers in psychological health and physical health: a meta-analysis. Psychol. Aging, 18, 250-267.
te
Russell, E., Koren, G., Rieder, M., Van Uum, S., 2012. Hair cortisol as a biological marker of chronic stress: current status, future directions and unanswered questions. Psychoneuroendocrinology, 37,
Ac ce p
589-601.
Schoenmakers, B., Buntinx, F., Delepeleire, J., 2010. Factors determining the impact of care-giving on caregivers of elderly patients with dementia. A systematic literature review. Maturitas, 66, 191200.
Schreiner, A. S., Morimoto, T., Arai, Y., Zarit, S., 2006. Assessing family caregiver's mental health using a statistically derived cut-off score for the Zarit Burden Interview. Aging Ment. Health, 10, 107-111.
Spaderna, H., Weidner, G., Zahn, D., Smits, J., 2009. Psychological Characteristics and Social Integration of Patients with Ischemic and Non-Ischemic Heart Failure Newly Listed for Heart Transplantation: The Waiting for a New Heart Study. App. Psychol. Health Well-Being, 1, 188210. Stalder, T., Kirschbaum, C., 2012. Analysis of cortisol in hair – State of the art and future directions. Invited Minireview. Brain Behav. Immun., 26, 1019-1029.
8
Page 8 of 14
Staufenbiel, S. M., Penninx, B. W., Spijker, A. T., Elzinga, B. M., van Rossum, E. F., 2013. Hair cortisol,
stress
exposure,
and
mental
health
in
humans:
a
systematic
review.
Psychoneuroendocrinology, 38, 1220-1235. Vedhara, K., Cox, N. K., Wilcock, G. K., Perks, P., Hunt, M., Anderson, S., Lightman, S.L., Shanks, N.M., 1999. Chronic stress in elderly carers of dementia patients and antibody response to
Ac ce p
te
d
M
an
us
cr
ip t
influenza vaccination. Lancet, 353, 627-631.
9
Page 9 of 14
Conflicts of interest
Ac ce p
te
d
M
an
us
cr
ip t
The authors have no conflicts of interest to declare.
10
Page 10 of 14
Contributors
Antje Tietze, Tobias Stalder, Lucia Dettenborn and Clemens Kirschbaum designed the study. Tobias Stalder chiefly wrote the manuscript. Susann Steudte and Nina Alexander also participated in writing
ip t
the manuscript as well as in analysing the data. All authors contributed and have approved the final
Ac ce p
te
d
M
an
us
cr
manuscript.
11
Page 11 of 14
Role of the funding source
The research described here was supported by a grant from the German Research Foundation (DFG;
Ac ce p
te
d
M
an
us
cr
ip t
KI 537/31-1).
12
Page 12 of 14
Table 1 Comparison of demographic, hair-related and self-report measures of dementia caregivers and noncaregiver controls Non-caregiver controls
(n = 20)
(n = 20)
71.2 (6.1) 19 (95) 20 (100) 26.66 (3.75) 2 (10) 12 (60) 16 (80)
72.2 (6.4) 17 (85) 13 (68.4) 25.13 (3.94) 0 (0) 16 (80) 17 (85)
1.59 (0.72) 10 (50)
1.68 (0.88) 4 (20)
t(35) = -.34 Fisher’s exact
.74 .10
4.8 (2.0) 4.9 (1.8) 111.5 (72.5) 16.3 (5.7) 11.6 (5.7) 10.5 (5.9)
6.7 (1.4) 6.9 (1.5) 84.8 (66.5) 9.8 (7.1) 15.4 (3.9) 8.6 (5.4)
t(38) = -3.41 t(38) = -3.77 t(34) = 1.15 t(32) = 2.95 t(38) = -2.48 t(36) = 1.06
.002 .001 .26 .006 .018 .29
p
t(38) = - 0.50 Fisher’s exact Fisher’s exact t(36) = 1.23 Fisher’s exact Fisher’s exact Fisher’s exact
.62 .61 .008 .23 .49 .30 1.0
cr
ip t
Test statistic
us
an
M
Demographics and health-related variables Age, year (M, SD) Female (n, %) Married (n, %) Body mass index (M, SD) Daily alcohol consumption (n, %) Chronic diseases (n, %) Regular medication (n, %) Hair-related variables Washes per week (M, SD) Hair treatment (n, %) Self-report measures Physical well-being (M, SD) Psychological well-being (M, SD) SRRS (M, SD) BDI-II (M, SD) ESSI (M, SD) JSPQ (M, SD)
Dementia caregivers
Ac ce p
te
d
SRRS, Social Readjustment Rating Scale; BDI-II, Beck Depression Inventory; ESSI, ENRICHD Social Support Inventory; JSPQ, Jenkins Sleep Problem Questionnaire
13
Page 13 of 14
p = .04
25
20
b)
50 40 30 20 10
r = .43, p = .058 0 0
15
10
20
30
40
50
60
Caregiving burden (ZBI)
5
0
Caregivers
Controls
c)
50
cr
10
60
40 30
us
Hair cortisol concentration (pg/mg)
Hair cortisol concentration (pg/mg)
30
60
ip t
a)
Hair cortisol concentration (pg/mg)
35
20 10
r = .48, p = .045
an
0
5
10
15
20
25
30
35
Depressiveness (BDI-II)
Ac ce p
te
d
M
Figure 1 Mean (± SEM) HCC of dementia caregivers and non-caregiver controls (a) as well as associations of HCC with perceived caregiving burden (b) and depressiveness (c) within dementia caregivers.
14
Page 14 of 14
S0306-4530(14)00154-1 http://dx.doi.org/doi:10.1016/j.psyneuen.2014.04.021 PNEC 2686
To appear in: Received date: Revised date: Accepted date:
25-2-2014 4-4-2014 25-4-2014
Please cite this article as: Stalder, T., Tietze, A., Steudte, S., Alexander, N., Dettenborn, L., Kirschbaum, C.,Elevated hair cortisol levels in chronically stressed dementia caregivers, Psychoneuroendocrinology (2014), http://dx.doi.org/10.1016/j.psyneuen.2014.04.021 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Elevated hair cortisol levels in chronically stressed dementia caregivers
Tobias Stalder a*†, Antje Tietzea*, Susann Steudtea, Nina Alexandera, Lucia Dettenbornb, Clemens Kirschbauma
Department of Medical Psychology, University Hospital Hamburg-Eppendorf, Germany
†
an
us
b
Technische Universität Dresden, Department of Psychology, Dresden, Germany
cr
a
ip t
(* shared first authorship)
Corresponding author & reprints:
Dr. Tobias Stalder
Technische Universität Dresden
M
Department of Psychology Andreas-Schubert-Bau Zellescher Weg 19
d
01069 Dresden, Germany Fax: +49 351 463 -37274
te
Tel: +49 351 463 -39678
Ac ce p
E-mail: [email protected]
Running title: Hair cortisol in dementia caregivers
Keywords: Cortisol; hair; stress; chronic; caregiver; depression; human.
Word count: 1997
1
Page 1 of 14
Abstract: Hair cortisol concentrations (HCC) are assumed to reflect integrated long-term cortisol levels and have been proposed as a promising endocrine marker of chronic psychological stress. The current study examined HCC in relation to caregiving burden, a well-established naturalistic model of chronic stress in humans. HCC and relevant psychosocial data were examined in 20 caregivers of relatives with
ip t
dementia and 20 non-caregiver controls matched for age and sex. Results revealed elevated HCC in dementia caregivers compared to non-caregiver controls (F(1,38) = 4.4, p = .04, ηp2 = .10). Further, within caregivers, a trend for a positive association of HCC with self-reported caregiving burden (r =
cr
.43, p = .058) and a positive association with depressiveness (r = .48, p = .045) were observed. No other associations between HCC and subjective measures were seen. These findings concur with the
Ac ce p
te
d
M
an
us
notion that HCC sensitively capture endocrine aberrations in stress-exposed groups.
2
Page 2 of 14
Introduction: Long-term changes in the secretion of the glucocorticoid cortisol are considered a crucial mediator for the link between chronic stress and ill health. To obtain valid assessments of long-term cortisol levels, however, has been methodologically challenging in the past. Cortisol measured in blood, saliva or urine exhibits high situational variability and thus provides only a poor reflection of underlying long-
ip t
term secretory patterns. Over the past decade, the analysis of hair cortisol concentrations (HCC) has increasingly gained acceptance as an easily obtainable index of integrated long-term cortisol secretion
cr
(reviews: Russell et al., 2012; Stalder & Kirschbaum, 2012). Considerable evidence has now supported this method in terms of its overall validity, test-retest reliability and robustness to potential
us
confounding influences (Stalder & Kirschbaum, 2012).
Overall, these data highlight HCC as a promising marker for endocrine correlates of chronic stress (Staufenbiel et al., 2013). However, direct enquiries in the context of chronic stress have produced
an
diverging results: while research examining relationships of HCC with subjective stress measures has failed to report consistent results, initial studies conducted in stress-exposed groups have uniformly shown elevations in HCC (Stalder & Kirschbaum, 2012). The latter evidence concerning HCC in
M
stress-exposed groups, however, is still based on a smaller number of studies carried out in specialized samples. Specifically, increased HCC were reported in hospitalized and ventilated infants, chronic pain patients, unemployed individuals, students with major life stressors (review: Staufenbiel et al.,
d
2013) and members of Aboriginal First Nation communities in Canada (Henley et al., 2013). These
te
studies feature groups in which the effect of psychological stress may not be fully distinguishable from effects of underlying medical conditions (e.g., hospitalized infants or chronic pain patients) or ethnic group differences (e.g., hair characteristics of Aboriginal people vs. whites).
Ac ce p
This highlights the need to further investigate long-term endocrine alterations in groups with welldefined increased psychosocial stress exposure. A suitable human condition for doing so is provided by dementia caregiving burden. Caring for a demented relative constitutes a highly stressful task, characterized by changing role patterns, continuous mourning for the affected relative and demanding caring duties due to the uncontrollable nature of dementia (Schoenmakers et al., 2010). Consistently, dementia caregivers have been shown to experience higher rates of psychological stress, an increased risk for psychiatric diseases and poorer health outcomes compared to other caregivers or non-caregiver peers (Pinquart & Sorensen, 2003). Further, increased activity of the endocrine stress system in dementia caregivers has been indicated (e.g., Vedhara et al., 1999), although with some inconsistency (e.g., Irwin et al., 1997). The current study thus set out to provide a first investigation of HCC in groups of dementia caregivers and non-caregiver controls matched for age and sex. We predicted to find elevated longterm cortisol levels in dementia caregivers. To further foster interpretability of results, we also assessed dimensional relationships with self-reported caregiving burden, depressiveness, social support and sleep problems. 3
Page 3 of 14
Methods: Participants: Twenty-one elderly caregivers of relatives with dementia and 22 age and sex-matched non-caregiver controls were recruited from self-help groups for dementia caregivers (caregivers) and meeting centers for the elderly (controls). The dementia caregiver group comprised individuals who were currently
ip t
caring for a close relative with a general practitioner-verified diagnosis of dementia. Caregivers in this group had been caring for their spouse (n=20) or parent (n=1) for a mean (± SD) time of 40.8 (± 30.8)
cr
months. The care settings of most participants were classified with German nursing care levels I or II which signify severe or very severe care needs, respectively. High caregiving burden in caregivers was
us
further reflected by mean (± SD) scores of 38.2 (± 13.2) on the Zarit Burden Interview (ZBI; Braun et al., 2010). A ZBI cut-off value of 26 has been proposed to mark an increased mental health risk due to high caregiving burden (Schreiner et al., 2006). The non-caregiver group comprised individuals who
an
had never cared for a demented relative.
Participants were only included in the study (both groups) if they were not taking glucocorticoidcontaining medication and had not had any medical surgery within three-months prior to hair
M
sampling. To avoid creating an artificial sample of elderly participants, we refrained from imposing further exclusion criteria based on medical diseases or medication intake, but instead controlled for these effects statistically. All participants provided written informed consent prior to taking part in the
d
study. The study protocol was approved by the ethics committee of the TU Dresden Medical Faculty
Self-report measures:
te
and the study was conducted in accordance with the Declaration of Helsinki.
Ac ce p
Information on demographic and health-related variables (age, sex, body-mass-index (BMI), marital status, regular daily alcohol consumption, chronic diseases and medication intake) and hair-related characteristics (washes/week, hair treatment: coloration, bleaching, permanent wave) were measured using a self-developed questionnaire. Two ten-point rating scales were used to assess physical and mental well-being (endpoints: very low, very high) with high values signaling high well-being. The Social Readjustment Rating Scale (SRRS; German Version: Katschnig, 1980) was used to assess major life events over the past year. Depressive symptomatology over the past two weeks was assessed using the Beck Depression Inventory (BDI-II; German Version: Hautzinger, Keller and Kühner, 2009) and the ENRICHD Social Support Inventory (ESSI; German Version: Spaderna et al., 2009) was employed to measure general perceived social support. Finally, the Jenkins Sleep Problem Questionnaire (JSPQ, Jenkins et al., 1988) was used to assess self-reported sleep problems over the previous months.
4
Page 4 of 14
Hair cortisol analysis: Hair strands (~3 mm diameter) were cut scalp-near from a posterior vertex position. The proximal 3 cm hair segment was used for analyses. Based on an average hair growth rate of 1cm/month, this hair segment represents hair grown over the 3-month period prior to hair sampling. Wash and steroid extraction procedures followed the laboratory protocol described in Kirschbaum et al. (2009) with 10
ip t
mg of powdered hair being used for analyses in the current study. Cortisol levels were determined using a commercially available immunoassay with chemiluminescence detection (CLIA, IBL-
cr
Hamburg, Germany). Intra- and interassay coefficients of variance were below 10%.
Statistical analysis:
us
Analyses were performed using SPSS for Windows, version 21 (IBM, Chicago, Illinois). Box plot analyses revealed three extreme outlying HCC values exceeding three interquartile ranges from the
an
median which were excluded from subsequent analyses (one caregiver, two non-caregiver controls). Hair cortisol data were positively skewed and log10-transformations were applied and successfully established a normal distribution. Descriptive data in text and figures are reported in original units
M
(pg/mg). Initial group comparisons were conducted using t-tests (continuous variables) and Fisher’s exact tests (dichotomous variables). HCC were compared between dementia caregivers and noncaregiver controls using one-way ANOVA and followed up by ANCOVA adjusting for variables
Ac ce p
Results:
te
examined using Pearson correlations.
d
differing between groups. Dimensional associations between self-report measures and HCC were
Table 1 shows descriptive characteristics of the two study groups. Groups were well-matched on the examined demographic, health- and hair-related parameters, except for a significantly higher rate of married individuals in dementia caregivers. As expected for the elderly, a high percentage of participants in both groups reported chronic diseases and/or regular medication intake. Diseases included hypertension (n=12), back pain (n=3), ocular disorders/glaucoma (n=5), diabetes (n=4), allergies (n=3) and osteoarthritis (n=3). Most common medication types were antihypertensives (n=24), thyroxin (n=6) and antihyperglycemics (n=3). No group differences were observed in either prevalence of chronic disease types or medication categories (all p’s > .34), except for a nonsignificant trend for a higher rate of hypertension in non-caregiver controls (Fisher’s exact, p =.08). As expected, dementia caregivers reported lower well-being (physical and mental) and social support as well as higher depressiveness. No significant differences in self-reported life events or sleep problems were observed.
[Please insert Table 1 about here]
5
Page 5 of 14
Figure 1a shows HCC of dementia caregivers and non-caregiver controls. The univariate ANOVA revealed higher HCC in dementia caregivers (mean ± SD: 27.4 ± 11.3 pg/mg) than in non-caregiver controls (mean ± SD: 20.5 ± 7.3 pg/mg; F(1,38) = 4.4, p = .04, ηp2 = .10). This effect remained significant in an ANCOVA adjusting for differences in marital status and prevalence of hypertension
= .17) or ESSI scores (F(2,37) = 2.9, p = .097) were controlled for.
ip t
(F(3,35) = 4.9, p = .03, ηp2 = .12). The effect missed statistical significance when BDI-II (F(2,31) = 2.0, p Correlation analyses carried out across groups revealed no significant associations between HCC and SRRS, BDI-II, ESSI and JSPQ (all p’s > .14). Analyses conducted within the caregiver group
cr
revealed a trend for a positive correlation of HCC with ZBI-assessed caregiving burden (r = .43, p = .058, n = 20) and a positive association with BDI-II scores (r = .48, p = .045, n = 18; see Figures 1b
an
patients’ nursing care level were observed (all p’s > .34).
us
and 1c, respectively). No further associations of HCC with self-report measures, time of caregiving or
[Please insert Figure 1 about here]
M
Discussion:
This study investigated the relationship of integrated long-term cortisol levels in hair and caregiving burden, a well-established human model of chronic psychosocial stress. Our main finding showed
d
significantly higher HCC in dementia caregivers compared to non-caregiver controls. This concurs
te
with previous data showing elevated HCC across a range of stress-related conditions (Stalder & Kirschbaum, 2012; Staufenbiel et al., 2013). Besides this group effect, analyses conducted within dementia caregivers also revealed positive relationships of HCC with perceived caregiving burden (at
Ac ce p
trend level) and depressiveness.
Our main finding of increased HCC in dementia caregivers concurs with previous reports of elevated salivary or blood cortisol in this group (e.g., Vedhara et al., 1999). Interestingly, this pattern has not unanimously been found (e.g., Irwin et al., 1997), perhaps partly due to limitations in employed cortisol assessment methods. The fact that caring for a demented relative is highly stressful has repeatedly been shown (Pinquart & Sorensen, 2003) and was also reflected in the current data showing lower physical and mental well-being in caregivers. Given that the two groups were comparable on other examined variables, it is most likely that increased HCC in caregivers were mainly attributable to repeated cortisol reactions to stressful experiences in caregivers’ daily routines. This interpretation was further supported by our tentative dimensional finding suggesting that caregivers who reported greater caregiving burden also tended to exhibit higher HCC. Besides increased stress exposure in caregivers, the fact that the HCC group effect missed significance after controlling for BDI-II or ESSI scores further suggests that higher depressiveness and lower perceived social support in this group also partly contributed to the observed long-term endocrine aberrations. A potential role of depressiveness was further highlighted by our finding of a 6
Page 6 of 14
positive association of HCC and depressiveness within caregivers, although this relationship was not seen across the whole sample. The finding within caregivers could be seen as suggestive of increased long-term cortisol secretion in more depressed individuals which has also been indicated by some but not all previous HCC research (review: Staufenbiel et al., 2013). However, as this association was only seen within the relatively small group of caregivers, further investigation is needed until firm
ip t
conclusions can be drawn.
The latter fact that associations between HCC and self-report measures were only seen within caregivers but not across the whole sample is an interesting aspect of the current findings. It can only
cr
be speculated what may be responsible for this result. We have previously suggested that, amongst other things, individual differences in self-awareness, social desirability and retrospective bias may
us
have had a negative effect on the quality of self-report data in previous HCC research (Stalder & Kirschbaum, 2012). It is conceivable that this has applied less to the current caregiver sample who
an
may have been particularly sensitized to observing their stress-related living situation through frequent participation in caregiver self-help groups. This could have resulted in more accurate self-reports and thus increased the likelihood of detecting psychoendocrine covariance in this group. Still, other
M
explanations are clearly possible and future corroboration of this effect in a larger sample is expedient. Such testing may also benefit from the use of more elaborate self-assessment methods, such as ecological momentary assessments or the experience sampling method, which may further help to
Ac ce p
te
d
raise ecological validity of self-report data across the whole sample.
7
Page 7 of 14
References: Braun, M., Scholz, U., Hornung, R., Martin, M., 2010. Die subjektive Belastung pflegender Ehepartner von Demenzkranken. Hinweise zur Validität der deutschen Version des Zarit Burden Interviews. Z. Gerontol. Geriatr., 43, 111-119.
am Main: Pearson Assessment & Information GmbH.
ip t
Hautzinger, M., Keller, F., Kühner, C., 2009. BDI-II. Beck-Depressions-Inventar. Revision. Frankfurt
Henley, P., Jahedmotlagh, Z., Thomson, S., Hill, J., Darnell, R., Jacobs, D., Johnson, J., Williams, N.C., Williams, R.M., Van Uum, S., Bend, J.R., Koren, G., 2013. Hair Cortisol as a Biomarker of
cr
Stress Among a First Nation in Canada. Ther. Drug. Monit., 35, 595-599.
Jenkins, C. D., Stanton, B. A., Niemcryk, S. J., Rose, R. M., 1988. A scale for the estimation of sleep
us
problems in clinical research. J. Clin. Epidemiol., 41, 313-321.
Irwin, M., Hauger, R., Patterson, T. L., Semple, S., Ziegler, M., Grant, I., 1997. Alzheimer caregiver
an
stress: basal natural killer cell activity, pituitary-adrenal cortical function, and sympathetic tone. Ann. Behav. Med., 19, 83-90.
Katschnig, H., 1980. Sozialer Stress und psychische Erkrankung. München: Urban & Schwarzenberg.
M
Kirschbaum, C., Tietze, A., Skoluda, N., Dettenborn, L., 2009. Hair as a retrospective calendar of cortisol production – Increased cortisol incorporation into hair in the third trimester of pregnancy. Psychoneuroendocrinology, 34, 32-37.
d
Pinquart, M., Sorensen, S., 2003. Differences between caregivers and noncaregivers in psychological health and physical health: a meta-analysis. Psychol. Aging, 18, 250-267.
te
Russell, E., Koren, G., Rieder, M., Van Uum, S., 2012. Hair cortisol as a biological marker of chronic stress: current status, future directions and unanswered questions. Psychoneuroendocrinology, 37,
Ac ce p
589-601.
Schoenmakers, B., Buntinx, F., Delepeleire, J., 2010. Factors determining the impact of care-giving on caregivers of elderly patients with dementia. A systematic literature review. Maturitas, 66, 191200.
Schreiner, A. S., Morimoto, T., Arai, Y., Zarit, S., 2006. Assessing family caregiver's mental health using a statistically derived cut-off score for the Zarit Burden Interview. Aging Ment. Health, 10, 107-111.
Spaderna, H., Weidner, G., Zahn, D., Smits, J., 2009. Psychological Characteristics and Social Integration of Patients with Ischemic and Non-Ischemic Heart Failure Newly Listed for Heart Transplantation: The Waiting for a New Heart Study. App. Psychol. Health Well-Being, 1, 188210. Stalder, T., Kirschbaum, C., 2012. Analysis of cortisol in hair – State of the art and future directions. Invited Minireview. Brain Behav. Immun., 26, 1019-1029.
8
Page 8 of 14
Staufenbiel, S. M., Penninx, B. W., Spijker, A. T., Elzinga, B. M., van Rossum, E. F., 2013. Hair cortisol,
stress
exposure,
and
mental
health
in
humans:
a
systematic
review.
Psychoneuroendocrinology, 38, 1220-1235. Vedhara, K., Cox, N. K., Wilcock, G. K., Perks, P., Hunt, M., Anderson, S., Lightman, S.L., Shanks, N.M., 1999. Chronic stress in elderly carers of dementia patients and antibody response to
Ac ce p
te
d
M
an
us
cr
ip t
influenza vaccination. Lancet, 353, 627-631.
9
Page 9 of 14
Conflicts of interest
Ac ce p
te
d
M
an
us
cr
ip t
The authors have no conflicts of interest to declare.
10
Page 10 of 14
Contributors
Antje Tietze, Tobias Stalder, Lucia Dettenborn and Clemens Kirschbaum designed the study. Tobias Stalder chiefly wrote the manuscript. Susann Steudte and Nina Alexander also participated in writing
ip t
the manuscript as well as in analysing the data. All authors contributed and have approved the final
Ac ce p
te
d
M
an
us
cr
manuscript.
11
Page 11 of 14
Role of the funding source
The research described here was supported by a grant from the German Research Foundation (DFG;
Ac ce p
te
d
M
an
us
cr
ip t
KI 537/31-1).
12
Page 12 of 14
Table 1 Comparison of demographic, hair-related and self-report measures of dementia caregivers and noncaregiver controls Non-caregiver controls
(n = 20)
(n = 20)
71.2 (6.1) 19 (95) 20 (100) 26.66 (3.75) 2 (10) 12 (60) 16 (80)
72.2 (6.4) 17 (85) 13 (68.4) 25.13 (3.94) 0 (0) 16 (80) 17 (85)
1.59 (0.72) 10 (50)
1.68 (0.88) 4 (20)
t(35) = -.34 Fisher’s exact
.74 .10
4.8 (2.0) 4.9 (1.8) 111.5 (72.5) 16.3 (5.7) 11.6 (5.7) 10.5 (5.9)
6.7 (1.4) 6.9 (1.5) 84.8 (66.5) 9.8 (7.1) 15.4 (3.9) 8.6 (5.4)
t(38) = -3.41 t(38) = -3.77 t(34) = 1.15 t(32) = 2.95 t(38) = -2.48 t(36) = 1.06
.002 .001 .26 .006 .018 .29
p
t(38) = - 0.50 Fisher’s exact Fisher’s exact t(36) = 1.23 Fisher’s exact Fisher’s exact Fisher’s exact
.62 .61 .008 .23 .49 .30 1.0
cr
ip t
Test statistic
us
an
M
Demographics and health-related variables Age, year (M, SD) Female (n, %) Married (n, %) Body mass index (M, SD) Daily alcohol consumption (n, %) Chronic diseases (n, %) Regular medication (n, %) Hair-related variables Washes per week (M, SD) Hair treatment (n, %) Self-report measures Physical well-being (M, SD) Psychological well-being (M, SD) SRRS (M, SD) BDI-II (M, SD) ESSI (M, SD) JSPQ (M, SD)
Dementia caregivers
Ac ce p
te
d
SRRS, Social Readjustment Rating Scale; BDI-II, Beck Depression Inventory; ESSI, ENRICHD Social Support Inventory; JSPQ, Jenkins Sleep Problem Questionnaire
13
Page 13 of 14
p = .04
25
20
b)
50 40 30 20 10
r = .43, p = .058 0 0
15
10
20
30
40
50
60
Caregiving burden (ZBI)
5
0
Caregivers
Controls
c)
50
cr
10
60
40 30
us
Hair cortisol concentration (pg/mg)
Hair cortisol concentration (pg/mg)
30
60
ip t
a)
Hair cortisol concentration (pg/mg)
35
20 10
r = .48, p = .045
an
0
5
10
15
20
25
30
35
Depressiveness (BDI-II)
Ac ce p
te
d
M
Figure 1 Mean (± SEM) HCC of dementia caregivers and non-caregiver controls (a) as well as associations of HCC with perceived caregiving burden (b) and depressiveness (c) within dementia caregivers.
14
Page 14 of 14
