Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 May 1.
Published in final edited form as: Health Psychol. 2015 Sep 7;35(5):465–473. doi: 10.1037/hea0000276

Sleep Duration and the Cortisol Awakening Response in Dementia Caregivers Utilizing Adult Day Services

Amanda N Leggett 1, Yin Liu 2, Laura Cousino Klein 3, Steven H Zarit 4
PMCID: PMC4781675  NIHMSID: NIHMS719467  PMID: 26348500

Abstract

Objective

Sleep complaints are common among caregivers and are associated with detriments in mental and physical health. Cortisol, a biomarker of the stress process, may link sleep with subsequent health changes in caregivers. The current study examines whether sleep duration is directly associated with the cortisol awakening response (CAR), or if it is moderated by Adult Day Services (ADS) use, an intervention found previously to influence daily CAR by reducing stressor exposure.

Methods

Associations were examined in caregivers (N=158) of individuals with dementia (IWD) on days when IWDs attended ADS and days when IWDs did not attend ADS. Data were gathered over 8 consecutive days. Caregivers were primarily female (87.3%) with a mean age of 61.59. A multi-level growth curve model tested the association of an interaction of today's ADS use and last night's sleep duration with today's CAR as the outcome.

Results

The interaction between ADS use and within-person sleep duration was significant such that when an individual sleeps longer than their average but does not use ADS, they have a smaller or blunted CAR. On the other hand when an individual sleeps longer than their average and uses ADS, they have a higher but nonsignificant CAR. Sleeping shorter than usual was associated with a dynamic rise regardless of ADS use.

Conclusions

Findings indicate that ADS use moderates the association between sleep duration and CAR such that longer than average sleep is associated with blunted, dysregulated cortisol patterns only on non-ADS days.

Keywords: sleep, cortisol, adult day services, caregiving

Sleep problems are common among caregivers for individuals with dementia (IWDs) with as many as two-thirds of caregivers complaining about their sleep (McCurry, Logsdon, Teri & Vitiello, 2007). Caregivers may experience sleep problems due to normal aging changes in the circadian timing system, environmental factors, sleep disorders, or night time disturbances due to the IWD (McCurry, Pike, Vitiello, Logsdon, & Teri, 2008). Problems with sleep have implications for the physical and mental health of caregivers (Baglioni et al., 2011). Both prolonged sleep (9 hours or more a night) and short sleep duration (6.9 hours or less of sleep; Fang et al., 2015) are associated with diabetes, coronary heart disease, stroke, cardiovascular disease, morbidity and risk of mortality (Ayas et al., 2003; Cappuccio, Cooper, D'Elia, Strazzullo, & Miller, 2011; Cappuccio, D'Elia, Strazzullo, & Miller, 2010). Recently researchers have connected sleep with health by examining associations with Hypothalamus-Pituitary-Adrenal Cortex (HPA) axis activity, and in particular a stress hormone cortisol.

Cortisol and sleep duration

Cortisol is a hormonal output that is responsive to stressors. It has anti-inflammatory effects, helps the body respond to stress, mobilizes energy, and communicates with the immune system (Piazza, Almeida, Dmitrieva, & Klein, 2010; Sternberg & Gold, 2002). Cortisol levels vary diurnally in relation to a circadian pattern. Typically, cortisol increases sharply following awakening, declines throughout the day, remains low during the early sleep period, and finally begins to rise shortly before awakening in the morning (Elder, Wetherell, Barclay, & Ellis, 2013; Fries, Dettenborn, & Kirschbaum, 2009; Garde, Karlson, Hansen, Persson, & Åkerstedt, 2012; Wilhelm, Born, Kudielka, Schlotz, & Wüst, 2007). Research has frequently focused on the cortisol awakening response (CAR), that is, the extent of change in cortisol levels from awakening to 30 minutes post awakening (Seltzer et al., 2010). The CAR is believed to play a role in increasing energy for the day, arousal, and anticipation for the day's events (Elder et al., 2013). By contrast, a low or blunted CAR is associated with exhaustion, burnout, and perceived fatigue in adults (Adam, Hawkley, Kudielka, & Cacioppo, 2006; Chida & Steptoe, 2009).

Sleep is related to the CAR through interconnected physiological mechanisms (Chrousos, 2009; Saper, Scammell, & Lu, 2005). As nighttime sleep is ending, HPA activity increases (Buckley & Schatzberg, 2005), interacting with the suprachiasmatic nucleus (biological clock of the brain) (Aubry et al., 2010; van Cauter & Turek, 1995). Across the CAR period, individuals go from sleep inertia, or grogginess following awakening, to growing alertness, with the cortisol acrophase occurring at the time of peak alertness (Ikeda & Hayashi, 2008). However, the growing alertness and rise in CAR are dependent on factors such as sleep duration (Elder et al., 2013; Tassi & Muzet, 2000). Indeed, though some studies report insignificant associations, many have found an association between prolonged sleep duration and a smaller, or blunted, CAR (Kumari et al., 2009; Schlotz, Hellhammer, Schulz, & Stone, 2004; Vreeburg et al., 2009; Wüst et al., 2000).

The differences in reported variations in associations and significance levels between cortisol and sleep are likely due to different definitions or measurements of sleep and cortisol, few measurement occasions, small sample sizes (in particular most studies have an n < 100) and ultimately low statistical power (Garde, Karlson, Hansen, Persson, & Åkerstedt, 2012). This study builds on these limitations by considering these associations in a larger sample of family caregivers and across eight consecutive days. Additionally, while prior work has found great variability in individuals' nightly sleep duration and that this variability may be associated with stress (Mezick et al., 2009); studies examining sleep duration and CAR primarily look at average levels of sleep duration. The current study extends this prior research by considering within-person associations in sleep duration as related to the CAR. In line with most prior research, prolonged sleep is anticipated to be associated with a blunted CAR slope.

Stressor exposure and CAR

In addition to the effects of sleep and awakening, fluctuations in the daily CAR are posited to be related to psychosocial factors such as stress exposure (Elder et al., 2013). Under conditions of chronic stress, CAR sometimes is elevated in response to stressors, but a more common response is a blunted or attenuated morning rise, particularly among older adults (Bremmer et al., 2007; Burke et al., 2005; Heaney et al., 2010; Leggett et al., 2014). Further, arousal or stress may interfere with circadian rhythms that generally regulate sleep and wake cycles. In addition, the CAR is responsive to anticipation of the day's impending stressors (Elder, Wetherell, Barclay, & Ellis, 2013; Fries et al., 2009; Lovell, Moss, & Wetherell, 2011). CAR may, therefore, react adaptively such that processes of arousal and energy are gauged to meet the day's demands. However, much of the cortisol and sleep literature has been conducted in populations experiencing normal, everyday challenges. In the current study, we examine associations of CAR and sleep in people experiencing high levels of chronic stress, family caregivers of individuals with dementia (IWD).

Impact of Adult Day Service use

An additional feature of the current study is the considerable fluctuation of stressor levels across the days of observation due to the introduction on some days of a respite program- adult day services (ADS). While sleep duration is posited to be directly associated with the CAR slope, ADS use may moderate this association with daily cortisol reactivity by impacting caregivers' sleep wake patterns, stressor exposure, and anticipation of the day's events. ADS provide IWDs social/therapeutic activities and medical care, and their caregivers receive respite from care responsibilities. ADS may structure caregiver's days differently from non-ADS days and therefore affect sleep patterns. For example, caregivers may wake up earlier on ADS days in order to help their relative with dementia prepare and travel to the service on time, which may vary from days where they are home and actively caring for their relative.

Prior research has found that ADS use reduces daily exposure to care-related stressors of family caregivers (Zarit et al., 2011). Furthermore previous work using the same data set as in the current study found that the reduced stressor exposure on ADS days leads to better regulation of cortisol, particularly, increased CAR among caregivers who have blunted CAR (no or small morning rise) on days they provide all the care (Klein et al., 2014). In the current paper, we extend this research by taking into account the role that sleep duration may play in CAR for caregivers in relation to the fluctuation in stressor exposure on high stress days where they provide all the care to a relative with dementia compared to low stress days when their relative attends ADS. While prolonged sleep is posited to be associated with a blunted CAR, the impact of ADS use may moderate this effect.

The current study

The current study takes advantage of a natural experiment to examine how utilizing ADS might affect the relation of sleep duration and cortisol. In the Daily Stress and Health (DaSH) study, daily cortisol samples are collected and daily diary information is obtained from family caregivers for IWDs for eight days about their health, well-being, and daily care and non-care related activities. Caregivers' relatives with dementia in DaSH attended adult day services (ADS) at least two days a week, thus creating the opportunity to compare their responses on respite and non-respite days. This study builds on prior work examining associations between sleep, stress, and cortisol in several ways. First the within-person association between daily sleep duration and CAR in the context of an intervention, ADS use, is examined. Second, in contrast to prior studies which use an aggregated measure of CAR, the current study uses a growth curve approach to examine a more dynamic measure of daily CAR. Finally, ADS is uniquely considered as a moderator of the association between daily sleep duration and CAR. Two hypotheses are proposed. First, an association of prolonged sleep duration and a blunted CAR is hypothesized. Second, it is hypothesized that caregivers whose sleep is prolonged will have more dynamic (i.e., less blunted) CAR patterns on days where they anticipate ADS respite.

Methods

Sample

The DaSH study examines the daily experiences and stress of primary, family caregivers for IWDs who attend ADS at least two days a week. Caregivers were recruited through ADS programs in Colorado, New Jersey, Pennsylvania, and Virginia. Eligibility requirements included being an informal caregiver with primary care responsibility for the IWD, living with the IWD, not having an endocrine disorder or problem affecting saliva production, and the IWD had a dementia diagnosis not including mild cognitive impairment and had attended ADS for at least two days a week for at least a month. A total of 200 caregivers were eligible for the study. Sixteen (8%) did not complete an initial interview and another 11 (6%) either completed no daily interviews or their interview days did not include both ADS and non-ADS days. An additional 9 (5%) caregivers were excluded for missing or invalid saliva samples (e.g., mixed up salivettes) and 6 (3%) caregivers did not have valid CAR values for both ADS and non-ADS days. The resulting analytic sample was 158 (79%) caregivers. Table 1 contains caregiver and care receiver demographic information.

Table 1. Sample Characteristics of Caregivers and Individuals with Dementia.

M SD Range
CG's characteristics
 Female, % 87.3
 Age 61.59 10.54 39 – 87
 Race, %
  White 74.1
  African American 24.1
  Other 1.9
 Married, % 69.6
 College graduate, % 53.8
 Employed, % 43.0
 Relation to IWD, %
  Spouse 38.0
  Child 58.2
  Others 3.8
 Number of ADS days 4.11 1.46 1 – 6
 Length of care (in months) 62.05 46.00 3 – 264
 Waking cortisol sample (nmol/L) 9.12 5.89 0.06-57.53
 Second cortisol sample (nmol/L) 12.28 6.95 0.30-58.38
 CAR (nmol/L) 3.09 6.57 -41.44-29.66
IWD's characteristics
 Age 81.82 8.63 57 – 100
 Female, % 60.1
 ADL impairmenta 3.06 0.49 2 – 4
Open in a new tab

Notes. Participant N = 158.

ADS = adult day services. CG = caregiver. IWD = individual with dementia.

a

Mean scores of 13 items rated on a 4-point scale ranging from 1 (does not need help) to 4 (cannot do without help).

Procedure

ADS programs distributed flyers and put announcements regarding the study in their newsletters. Following a telephone interview to determine eligibility, interviewers gave participants the baseline interview in their home or another public place the participant preferred. Participants received training on saliva collection procedures during this in-person interview. Following the baseline interview, caregivers filled out daily diaries over 8 consecutive days, some of which the IWD was attending ADS and some of which the caregiver was actively caring for the IWD. Daily diary responses were given to a telephone interviewer from the Penn State Survey Research Center each evening. The data for this study was taken from the daily diary interviews in conjunction with the saliva samples. Saliva samples were provided five times a day. In this analysis the focus is on wakening (before getting out of bed) and half an hour after getting out of bed, which constitute CAR. Salivettes were color-coded and numbered in correspondence with the day and time they were to be used. Participants were instructed to roll a cotton swab in their mouth for two minutes saturating it, replace it in the salivette and keep it in the refrigerator until the end of the 8 days. They were also instructed not to have caffeinated products or tobacco within 30 minutes of a sample or to eat, drink, or brush their teeth before a sample. Caregivers were given a saliva collection worksheet to track their collection times and report any problems. At the end of the 8 days salivettes were picked up by an express mail service and delivered overnight to the Pennsylvania State University where they were assayed. A valid day for the CAR depended on the first two cortisol samples. Flags, or indications that a saliva sample may not be valid, were created for samples that had extreme values or where the caregivers' sleep-waking pattern did not allow the normal diurnal cortisol curve to take place (e.g., if less than 15 or more than 45 minutes elapsed between the first and second morning samples). Of 1,264 total days, 1,128 days were used (89%) from 158 participants.

Measures

Cortisol levels and Cortisol Awakening Response (CAR)

The CAR is an estimate of the time slope of the morning cortisol samples: the nmol/l waking level and the 30 minute after waking nmol/l level.

Sleep duration

Sleep duration was included as a covariate of next morning CAR. Sleep duration was calculated from self-reports of the time the caregiver went to bed and awoke the next morning. Daily sleep duration was centered at the person-means to represent within-person effects (i.e., daily deviations from an individual's own mean; Hoffman & Stawski, 2009). The person-mean of sleep duration is also included as a between-person covariate (i.e., individual's average across 8 days).

ADS use

To consider anticipatory effects of the ADS intervention on cortisol levels and CAR, same-day ADS use was included as a within-person covariate (ADS day = 1). ADS attendance was determined during the daily telephone interviews.

Overnight care-related stressors

Overnight care-related stressors were assessed using the Daily Record of Behavior (DRB; Fauth, Zarit, Femia, Hofer, & Stephens, 2006). Caregivers reported on the occurrence of behavior problems of the IWD during the overnight period . The DRB includes 19 items from six categories: depressive behaviors, disruptive behaviors, memory-related behaviors, reality problems, restless behaviors, and resistance to help with activities of daily living (ADL), and up to three additional items caregivers identified that were not included in the list. A frequency count was computed by summing the total number of behaviors reported from each time frame for a maximum count of 22 (α = 0.80). Both the person-mean centered score and person-mean score of overnight stressors were included.

Depressive mood

Though the CAR is relatively stable, it may vary in relation to psychosocial factors. As prior work from the DaSH study found a between-person level association between depressive mood and the CAR, this association is controlled in the present analysis (Leggett, Zarit, Kim, Almeida, & Klein, 2014). Depressive symptoms were assessed from an inventory of emotions from the Non-Specific Psychological Distress Scale developed for the original MIDUS survey and adapted for daily use in the National Study of Daily Experiences (NSDE) (Kessler et al., 2002). The scale included seven items (α = 0.77). Participants were asked how frequently they felt each emotion on a 5-point scale from 1 (none of the day) to 5 (all day) with higher scores indicating greater depressive mood. Sample items include feeling worthless, hopeless, and so sad that nothing could cheer you up.

Demographics

Caregivers' gender (1 = female, 0 = male), age and the duration of time providing care in months were included as controls. Steroidal medication use (1 = took a steroid medication) and current smoking status (1 = currently smoking) were considered as controls in analyses but were not significant and thus excluded for the parsimony of the final model.

Analysis

First, a two-level multilevel model is examined for each individual cortisol sample (waking level and 30 minutes post waking level) as the outcome variable (SAS PROC MIXED; Littell, Miliken, Stroup, & Wolfinger, 1996; Raudenbush & Bryk, 2002). These models are compared with a three-level growth-curve model assessing the rate of change in morning cortisol levels to determine whether the effects of sleep and ADS use are persistent across morning cortisol levels, or associated truly with the awakening response (Adam, Hawkley, Kudielka, & Cacioppo, 2006; Raudenbush & Bryk, 2002). To examine the primary research question regarding whether ADS use moderates the association between nightly sleep duration and morning CAR, a main effects model (Model 1) is considered first and next an interaction of within-person fluctuation in sleep duration the previous night and today's ADS use with time (Model 2) is added. The 3-level growth curve model was specified as:

Level 1 (sample level):

Cortisolsdi = π0di + π1di (Timesdi) + εsdi

Level 2 (day level):

π0di = β00i + β01i (Sleep durationdi) + β02i (ADSdi) + β03i (Sleep durationdi*ADSdi) + β04i (Overnight stressorsdi) + υ0di

π1di = β10i + β11i (Sleep durationdi) + β12i (ADSdi) + β13i (Sleep durationdi*ADSdi) + β14i (Overnight stressorsdi)

Level 3 (person level):

β00i = γ000 + γ001 (Average sleep durationi) + γ002 (Average overnight stressorsi) + γ003 (Average depressive moodi) + γ004 (Agei) + γ005 (Genderi) + γ006 (Duration of carei) + u00i

β01i = γ010

β02i = γ020

β03i = γ030

β04i = γ040

β10i = γ100 + γ101 (Average sleep durationi) + γ102 (Average overnight stressorsi) + γ103 (Average depressive moodi) + γ104 (Agei) + γ105 (Genderi) + γ106 (Duration of carei) + u10i

β11i = γ110

β12i = γ120

β13i = γ130

β14i = γ140

At Level 1, change in cortisol is modeled as a function of an intercept (γ000, the mean level of morning cortisol for each individual averaged across days) and time elapsed since wakeup (γ100). At Level 2, sleep durationdi (γ110, slope parameter that represents the effect of last night's sleep duration on today's CAR time slope), ADSdi (γ120, slope parameter that represents the effect of today's ADS use on today's CAR time slope), sleep duration by ADSdi (γ130, slope parameter for the interaction of time, sleep duration and ADS), overnight stressorsdi (γ140, slope parameter that represents the effect of last night's stressors on today's CAR time slope), and the person-specific deviations from the intercept (υ0di). These covariates reflect deviations from an individual's own mean, by centering around the person-mean.

At Level 3, person-mean scores of sleep duration (γ101), overnight stressors (γ102), and depressive mood (γ103) were entered as between-person covariates, each covariate indicating individuals' average level of the covariate across days. Age (γ104), gender (γ105), and duration of care (γ106) were also entered as between-person level controls. γ100 reflects the grand-mean level of individual CAR slopes and u10i reflects individual deviations from that mean. γ110, γ120, γ130 and γ140 represent the average effects of within-person slopes for sleep duration, ADS, sleep duration by ADS and overnight stressors respectively.

Results

Sample characteristics of caregivers and IWDs are presented in Table 1. Intraclass correlations (ICC) were examined for the two waking cortisol samples. The ICC for waking cortisol revealed that 36% of the variance was between-person and 64% of the variance was within person, and the ICC for the 30 minute post-waking cortisol sample revealed that 37% of the variance is between-person and 63% of the variance is within-person. On average caregivers slept for 7.68 hours a night (SD = 0.94; range of individual averages across days: 4.58 to 11.34; within-person, day level variation ranging from -2.97 to 3.54 around individuals' means, SD = 0.97). The ICC for sleep duration revealed that 41% of the variance was between-person and 59% of the variance was within-person. Individuals reported on average one overnight care-related stressor (M = 1.28, SD = 1.79; range: 0 to 16; within-person variation ranging from -7.63 to 7.38, SD = 1.36). Individuals had an average depressive mood score of 2.50 (SD = 2.76; range: 0 to 17.43). A significant difference in wake time (t = -2.03(1242), p <.05) and sleep duration (t = 9.32(1242), p <.001) for ADS versus non-ADS days was found. ADS days were associated with earlier wake times (M of 6.42 military time, SD of 1.11, compared to M of 7.26 military time, SD of 1.35) and shorter sleep durations (M of 7.36 hours, SD of 1.22 compared to M of 8.05 hours, SD of 1.41). However the bedtime the night before ADS did not significantly differ from a night before non-ADS use. On average, participants started use of ADS at 8:52am (SD=1:08; range: 6:30am to 11:55am) and ended ADS at 4:20pm (SD=0:57; range: 1:45pm to 6:00pm).

Two-level models with individual cortisol samples as the outcome showed only person-level sleep duration was associated with waking cortisol (β= -0.96, p < .01). The interaction between sleep duration and ADS use was not significant. In regards to thirty minutes post-waking cortisol, within-person (β = -0.37, p < .05) and between-person (β = -1.20, p < .01) sleep duration and ADS use (β = 1.31, p < .001) were significant covariates. In a second model including an interaction between within-person sleep duration and ADS, the interaction was significant (β = 0.80, p < .05). Specifically, on non-ADS days, higher cortisol levels at 30-min post-waking were associated with shorterwithin-person sleep duration (β = -0.71, p <= 0.01). On ADS days, however, within-person sleep duration did not predict cortisol levels.

To test the primary hypotheses regarding the association of sleep duration, ADS use and the CAR, a three-level growth -curve was modeled where Model 1 considered only main effects, and Model 2 included a three-way time by day-level sleep duration by ADS use interaction. Parameter estimates for Model 1 and 2 are in Table 2. In Model 1, sleeping shorter than one's average (γ110 = -0.88, p < .01), ADS use (γ120 = 1.91, p < .01), and less depressive mood (γ103 = -0.58, p < .01) were associated with a larger CAR. The between-person sleep duration was significantly associated with the cortisol intercept (γ001 = -0.96, p < .01). Given significant main effects of sleep duration and ADS use at Level 2, in Model 2 an interaction between time, within-person sleep duration and ADS use was added to examine whether the association between sleep duration and CAR varied given daily ADS use. The interaction was significant (γ130 = 2.07, p < .01) showing that ADS use moderated the association between within-person sleep duration and CAR.

Table 2. The Effects of Sleep Duration, Depressive Mood, Overnight Stressors, and Adult Day Services (ADS) attendance on the Cortisol Awakening Response (CAR) of Family Caregivers.

Model 1 Model 2
β (s.e.) β (s.e.)
Fixed effects
 Intercept, γ000 9.90 (0.89)*** 9.82 (0.88)***
 Sample-Level Covariates
  Time, γ100 2.39 (1.51) 2.72 (1.51)
 Day-Level Covariates
  Sleep Durationa, γ010 -0.02 (0.20) 0.20 (0.21)
  ADS use, γ020 -0.07 (0.32) -0.07 (0.32)
  Sleep Duration × ADS use, γ030 0.15 (0.28) -0.37 (0.34)
  Care-Related Stressorsa, γ040 0.07 (0.11) 0.06 (0.11)
  Sleep Duration × Time, γ110 -0.88 (0.34)** -1.76 (0.46)***
  ADS use × Time, γ120 1.91 (0.68)** 1.93 (0.67)**
  Sleep Duration × ADS use × Time, γ130 -- 2.07 (0.73)**
  Care-Related Stressors × Time, γ140 0.12 (0.23) 0.15 (0.23)
 Person-Level Covariates
  Sleep Durationb, γ001 -0.96 (0.33)** -0.96 (0.33)**
  Care-Related Stressorsb, γ002 -0.05 (0.18) -0.06 (0.18)
  Depressive Moodb, γ003 -0.16 (0.12) -0.15 (0.12)
  Sleep Duration × Time, γ101 -0.22 (0.57) -0.21 (0.56)
  Care-Related Stressors × Time, γ102 0.27 (0.31) 0.32 (0.31)
  Depressive Mood × Time, γ103 -0.58 (0.20)** -0.60 (0.20)**
 Person-Level Controls
  Ageb, γ004 0.07 (0.03)* 0.07 (0.03)*
  Genderb, γ005 -0.68 (0.93) -0.67 (0.93)
  Duration of Careb, γ006 0.01 (0.01) 0.00 (0.01)
  Age × Time, γ104 0.04 (0.05) 0.04 (0.05)
  Gender × Time, γ105 2.18 (1.57) 2.12 (1.57)
  Duration of Care × Time, γ106 0.01 (0.01) 0.01 (0.01)
 Random effects
  Level3 Intercept VAR, u00i 11.29 (1.71)*** 11.24 (1.70)***
  Level3 Slope VAR, u10i 25.66 (4.83)*** 25.43 (4.78)***
  Level3 Intercept and slope COV -3.16 (2.12) -3.00 (2.10)
  Level2 Intercept VAR, u0di 7.89 (0.86)*** 7.92 (0.86)***
  Residual VAR, εsdi 17.57 (0.80)*** 17.48 (0.80)***
  -2 Log-Likelihood 14521.4 14512.2
  AIC/BIC 14531.4/14546.8 14522.2/14537.5
Open in a new tab

Notes. ADS = adult day services; VAR = variance; COV = covariance; AIC = Akaike information criterion; BIC = Bayesian information criterion. Sleep length, care-related stressors, and depressive mood are taken from the day before to predict next day CAR. ADS use is same day. Participant N = 158; Observation N = 2,335.

a

Person-mean-centered scores (i.e., time-varying).

b

Person-mean scores across days (i.e., time-invariant).

*

p < .05.

**

p < .01.

***

p < .001.

More specifically, on ADS days, controlling for other covariates, the within-person sleep duration and time interaction was nonsignificant (β = 0.32, p = 0.55); there was no association betweenCAR and within-person sleep duration. On non-ADS days, however, the association between CAR and within-person sleep duration was significant (β = -1.76, p = 0.0001); the longer a caregiver slept than his/her average hours, the smaller the CAR.

Simple slopes were also run for examining the three-way interaction (Curran, Bauer, & Willoughby, 2006; Preacher, Curran, & Bauer, 2006). The conditional values of within-person sleep duration were set at one standard deviation above (i.e., when one slept longer) and below (i.e., when one slept shorter) the within-person mean. The conditional values of ADS use were set at either 1 (ADS days) or 0 (non-ADS days). Regardless of ADS use, if an individual slept shorter than his/her average, a significant CAR was observed (simple slope at below average sleep duration on non-ADS days, β = 4.40 (1.62), p < .01; simple slope at below average sleep duration on ADS days, β = 4.34 (1.54), p < .01). Likewise, when an individual slept longer than his/her average and used ADS, they had a significant CAR (β = 4.95 (1.63), p < .01). On the other hand when an individual slept longer than his/her average but did not use ADS, there was a non-significant CAR, indicating blunted or no CAR (β =1.04 (1.53), p = 0.50). Figures 1 and 2 visually display the interaction.

Figure 1.

Figure 1

Open in a new tab

Cortisol levels on ADS days as a function of last night's sleep duration. Oversleep and undersleep were defined as 1 sd above and below the within-person average sleep hours, respectively. These cortisol slopes were significant at p < .05.

Figure 2.

Figure 2

Open in a new tab

Cortisol levels on non-ADS days as a function of last night's sleep duration. Oversleep and undersleep were defined as 1 sd above and below the within-person average sleep hours, respectively. The cortisol slope associated with undersleep was significant at p < .05, whereas the one associated with oversleep was nonsignificant.

Additional analyses

To give full consideration to the question regarding the association of sleep duration and ADS use on the CAR, a “treatment-response” model was considered. Whereby the primary analyses considered the “anticipation” effect of last night's sleep and today's ADS use on today's CAR, the “treatment-response” where last night's sleep and yesterday's ADS use might have an association with next day's CAR was also examined. In the treatment-response model, within-person sleep duration (β = -1.93, p < .001) had a significant association with the CAR, however yesterday's ADS use did not (β = -0.41, p = 0.56). Neither sleep duration nor yesterday's ADS use had an effect on the cortisol intercept. The three-way interaction between CAR, sleep duration and yesterday's ADS use was marginal (β = 1.33, p = 0.067).

Discussion

Consistent with prior literature on the association of sleep duration and CAR, when caregivers slept longer, they had a smaller CAR (Kumari et al., 2009; Schlotz, Hellhammer, Schulz, & Stone, 2004; Vreeburg et al., 2009; Wüst et al., 2000). A significant interaction, however, revealed that this association varied given ADS use. On ADS days, regardless of whether a caregiver slept longer or shorter than his/her average hours, cortisol increases were observed. However, on non-ADS days, the association between cortisol and time varied according to within-person sleep duration. When individuals slept longer than their average, a blunted CAR pattern was found. Yet, a dynamic time slope (CAR) was maintained if they slept shorter than was typical.

While our primary growth curve model predicted rate of change in CAR, our preliminary models predicting the level of individual cortisol samples which make up CAR presented a similar pattern of results, but only for the 30-minute post-waking cortisol levels. On non-ADS days, when a caregiver overslept, his/her 30-minute post-waking cortisol level would decrease. On ADS days, however, within-person sleep duration did not predict cortisol levels at the second sampling occasion. Thus, the significant pattern of associations between within-person sleep and cortisol levels, and within-person sleep, cortisol levels and time, were present on non-ADS days, but not on ADS days. These patterns of results suggest that the CAR may be driven by the post-waking cortisol level rather than the waking level, and may be most critical for the activation of morning cortisol. Consistent with other chronic stress studies, the anticipation of daily stress may affect that activation (Elder et al., 2013; Fries et al., 2009; Lovell, Moss, & Wetherell, 2011). Further, sleep and ADS did have an impact on the CAR, and not just a persistent effect across the morning sampling protocol.

Taken together, these results suggest that while sleep duration is associated with the degree of rise, ADS had a beneficial, anticipatory effect on the degree of rise in the CAR, and thus one's physiological response. This finding is consistent with prior work showing a protective or restorative effect of ADS use on daily cortisol by providing a predictable amount of time away from constant caregiving demands (Klein et al., 2014). Models were also adjusted for employment status and total number of days of ADS use, but these covariates were not significant and thus trimmed from the final models for parsimony. This, however, suggests that the benefit of respite on the awakening response is not explained by employment or a result of dose of the intervention received, but a benefit that is experienced at the daily level. In contrast, a treatment-response effect of ADS use was not found. Thus the anticipatory effects of what is to come appear more salient upon CAR than the events of the prior day, which is consistent with prior work (Fries et al., 2009).

It may be stressful for a caregiver to oversleep and then have to take his or her relative to ADS. For example, oversleeping may result in increased difficulty in getting the IWD to ADS and the caregiver may have to deal with increased time pressure or traffic. Yet, these stressors may be more controllable or manageable as compared to the primary objective stressors, such as behavioral and emotional problems associated with dementia that a caregiver may have to respond to on non-ADS days where actively caring for his or her relative. Types of stressors may have different physiological effects and future research should further consider this question in caregiving samples.

Limitations

The caregivers in the DaSH study volunteered to participate and thus there may be some element of selection bias in the study towards caregivers who self-selected to use ADS and had availability to participate in a week-long study. However, there are practical limitations of random assignment in experimental design to ADS use and the current study provided a unique position to examine within-person associations for individuals who were utilizing ADS several days a week. Further, while longer sleep duration was associated with a smaller or blunted CAR, individuals waking later than usual on non-ADS days may take their first sample at a time where cortisol has already begun to rise leaving less room for a dynamic CAR (Garde, Karlson, Hansen, Persson, & Åkerstedt, 2012). Descriptive results did find some differences in sleep and wake patterns on ADS versus non-ADS days. As it is impossible to measure saliva during sleep, assessing the association between various sleep stages and cortisol is beyond the current study. Night time awakening, however, was considered as a control in addition to overnight stressors. Awakening was not a significant covariate and did not impact the results and was therefore trimmed for parsimony of the final model. Objective (e.g. polysomnography or actigraphy) and self-report measures of sleep duration may differ from one another, with self-reports tending to overestimate total sleep time; however objective measures of sleep were not included in the DaSH study (Lauderdale, Knutson, Yan, Liu, & Rathouz, 2008).

Implications and Future Directions

Future studies may further consider whether ADS use, itself, impacts sleep duration and quality. Descriptive results did show that ADS use was associated with different sleep structures. While next day ADS use was not associated with a difference in bedtime, ADS use may alter other night time behaviors such as alcohol intake which could relate to next morning's CAR. Future studies should further investigate the mechanisms through which ADS can moderate associations between sleep and cortisol regulation.

These results extend the literature on sleep, stress, and cortisol in several ways. First, a moderating effect of ADS use suggests that respite interventions may provide an opportunity for physiological recovery and override the association between prolonged sleep and a maladaptive CAR pattern. Additionally, the importance of anticipating the day's events to waking cortisol is affirmed, as opposed to a carryover effect of respite. Chronic stress and burnout, such as many caregivers face, have been associated with blunted or maladaptive cortisol patterns which may correspond to negative health outcomes over time. Providing caregivers or other individuals experiencing chronic stress with respite or a mechanism to reduce their stressor exposure may have positive effects on starting the day with proper arousal and energy, and may reduce burden, allostatic load, and poor health outcomes. As the CAR is a measure of HPA axis reactivity (Schmidt-Reinwald et al., 1999), the finding that a respite intervention helps to normalize this response in a sample experiencing chronic stress is very promising and may therefore have key public health benefit. Given the importance of sleep duration for health, however, ADS programs may consider offering sleep hygiene programs, such as developed by Susan McCurry and colleagues, which have been found to improve sleep of both people with dementia and their caregivers (McCurry, Gibbons, Logsdon, Vitiello, & Teri, 2003; McCurry, Logsdon, Vitiello, & Teri, 1998). Maintaining a normal sleep schedule and utilizing a respite intervention, both modifiable behaviors, have the potential to greatly improve caregiver's mental, physical, and physiological well-being on a daily basis and over time.

Acknowledgments

This work was supported by grant RO1 AG031758 from the National Institute on Aging. Dr. Leggett is funded by a National Institute of Mental Health (T32 MH073553) Geriatric Mental Health Services Fellowship.

Contributor Information

Amanda N. Leggett, Department of Psychiatry, The University of Michigan, Ann Arbor, MI 48109, USA.

Yin Liu, Department of Human Development and Family Studies, The Pennsylvania State University, University Park, PA 16802, USA

Laura Cousino Klein, Department of Biobehavioral Health and Penn State Institute of the Neurosciences, The Pennsylvania State University, University Park, PA 16802, USA.

Steven H. Zarit, Department of Human Development and Family Studies, The Pennsylvania State University, University Park, PA 16802, USA.

References

  1. Adam EK, Hawkley LC, Kudielka BM, Cacioppo JT. Day-to-day dynamics of experience-cortisol associations in a population-based sample of older adults. Proceedings of the National Academy of Sciences. 2006;103:17058–17063. doi: 10.1073/pnas.0605053103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Aubry JM, Jermann F, Gex-Fabry M, Bockhorn L, Van der Linden M, Gervasoni N, Bondolfi G. The cortisol awakening response in patients remitted from depression. Journal of Psychiatric Research. 2010;44:1199–1204. doi: 10.1016/j.jpsychires.2010.04.015. [DOI] [PubMed] [Google Scholar]
  3. Ayas NT, White DP, Al-Delaimy WK, Manson JE, Stampfer MJ, Speizer FE, Hu FB. A prospective study of self-reported sleep duration and incident diabetes in women. Diabetes Care. 2003;26:380–384. doi: 10.2337/diacare.26.2.380. [DOI] [PubMed] [Google Scholar]
  4. Baglioni C, Battagliese G, Feige B, Spiegelhalder K, Nissen C, Voderholzer U, Riemann D. Insomnia as a predictor of depression: A meta-analytic evaluation of longitudinal epidemiological studies. Journal of Affective Disorders. 2011;135:10–19. doi: 10.1016/j.jad.2011.01.011. [DOI] [PubMed] [Google Scholar]
  5. Bremmer MA, Deeg DJH, Beekman ATF, Pennix BWJH, Lips P, Hoogendijk WJG. Major depression in late life is associated with both hypo- and hypercortisolemia. Biological Psychiatry. 2007;62:479–486. doi: 10.1016/j.biopsych.200611.033. [DOI] [PubMed] [Google Scholar]
  6. Buckley TM, Schatzberg AF. On the Interactions of the Hypothalamic-Pituitary-Adrenal (HPA) Axis and Sleep: Normal HPA Axis Activity and Circadian Rhythm, Exemplary Sleep Disorders. The Journal of Clinical Endocrinology & Metabolism. 2005;90:3106–3114. doi: 10.1210/jc.2004-1056. [DOI] [PubMed] [Google Scholar]
  7. Buijs RM, van Eden CG, Goncharuk VD, Kalsbeek A. The biological clock tunes the organs of the body: Timing by hormones and the autonomic nervous system. Journal of Endocrinology. 2003;177:17–26. doi: 10.1677/joe.0.1770017. [DOI] [PubMed] [Google Scholar]
  8. Burke HM, Davis MC, Otte C, Mohr DC. Depression and cortisol responses to psychological stress: A meta-analysis. Psychoneuroendocrinology. 2005;30:846–856. doi: 10.1016/j.psyneuen2005.02.010. [DOI] [PubMed] [Google Scholar]
  9. Cappuccio FP, Cooper D, D'Elia L, Strazzullo P, Miller MA. Sleep duration predicts cardiovascular outcomes: A systematic review and meta-analysis of prospective studies. European Heart Journal. 2011;32:1484–1492. doi: 10.1093/eurheratj/ehr007. [DOI] [PubMed] [Google Scholar]
  10. Cappuccio FP, D'Elia L, Strazzullo P, Miller MA. Sleep duration and all-cause mortality: A systematic review and meta-analysis of prospective studies. Sleep. 2010;33(5):585–592. doi: 10.1093/sleep/33.5.585. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Chida Y, Steptoe A. Cortisol awakening response and psychosocial factors: A systematic review and meta-analysis. Biological Psychology. 2009;80:265–278. doi: 10.1016/j.biopsycho.2008.10.004. [DOI] [PubMed] [Google Scholar]
  12. Chrousos GP. Stress and disorders of the stress system. Nature Reviews Endocrinology. 2009;5:374–381. doi: 10.1038/nrendo.2009.106. [DOI] [PubMed] [Google Scholar]
  13. Curran PJ, Bauer DJ, Willoughby MT. Testing and probing interactions in hierarchical linear growth models. In: Bergeman CS, Boker SM, editors. Methodological issues in aging research Notre Dame series on quantitative methods. Mahwah, NJ: Lawrence Erlbaum Associates Publishers; 2006. pp. 99–129. [Google Scholar]
  14. Elder GJ, Wetherell MA, Barclay NL, Ellis JG. The cortisol awakening response- Applications and implications for sleep medicine. Sleep Medicine Reviews. 2014;18:215–224. doi: 10.1016/j.smrv.2013.05.001. [DOI] [PubMed] [Google Scholar]
  15. Fang SC, Subramanian SV, Piccolo R, Yang M, Yaggi HK, Bliwise DL, Araujo AB. Geographic variations in sleep duration: A multilevel analysis from the Boston Area Community Health (BACH) Survey. Journal of Epidemiology and Community Health. 2015;69:63–69. doi: 10.1136/jech-2013-203256. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Fauth EB, Zarit SH, Femia EE, Hofer SM, Stephens MAP. Behavioral and psychological symptoms of dementia and caregivers' stress appraisals: Intra-individual stability and change over short-term observations. Aging and Mental Health. 2006;10:563–573. doi: 10.1080/13607860600638107. [DOI] [PubMed] [Google Scholar]
  17. Fries E, Dettenborn L, Kirschbaum C. The cortisol awakening response (CAR): Facts and future directions. International Journal of Psychophysiology. 2009;72:67–73. doi: 10.1016/j.ijpsycho.2008.03.014. [DOI] [PubMed] [Google Scholar]
  18. Garde AH, Karlson B, Hansen AM, Persson R, Åkerstedt T. Sleep and salivary cortisol. In: Kristenson M, Garvin P, Lundberg U, editors. The role of saliva cortisol measurement in health and disease. Sharjah, U.A.E.: Bentham Science Publishers; 2012. pp. 116–128. [DOI] [Google Scholar]
  19. Heaney JLJ, Phillips AC, Carroll D. Ageing, depression, anxiety, social support and the diurnal rhythm and awakening response of salivary cortisol. International Journal of Psychophysiology. 2010;78:201–208. doi: 10.1016/j.ijpsycho.2010.07.009. [DOI] [PubMed] [Google Scholar]
  20. Hoffman L, Stawski RS. Persons as contexts: Evaluating between-person and within-person effects in longitudinal analysis. Research in Human Development. 2009;6:97–120. doi: 10.1080/15427600902911189. [DOI] [Google Scholar]
  21. Ikeda H, Hayashi M. Effect of sleep inertia on switch cost and arousal level immediately after awakening from normal nocturnal sleep. Sleep and Biological Rhythms. 2008;6:120–125. doi: 10.1111/j.1479-8425.2008.00341.x. [DOI] [Google Scholar]
  22. Kessler RC, Andrews G, Colpe LJ, Hiripi E, Mroczek DK, Normand SLT, Zaslavsky AM. Short screening scales to monitor population prevalences and trends in non-specific psychological distress. Psychological Medicine. 2002;32:959–976. doi: 10.1017/S0033291702006074. [DOI] [PubMed] [Google Scholar]
  23. Klein LC, Kim K, Almeida DM, Femia EE, Rovine MJ, Zarit SH. Anticipating an easier day: Effects of Adult Day Services on daily cortisol and stress. The Gerontologist Advance Access published July 4, 2014. 2014 doi: 10.1093/geront/gnu060. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Kumari M, Badrick E, Ferrie J, Perski A, Marmot M, Chandola T. Self-reported sleep duration and sleep disturbance are independently associated with cortisol secretion in the Whitehall II study. The Journal of Clinical Endocrinology & Metabolism. 2009;94:4801–4809. doi: 10.1210/jc.2009-0555. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kunz-Ebrecht SR, Kirschbaum C, Marmot M, Steptoe A. Differences in cortisol awakening response on work days and weekends in women and men from the Whitehall II cohort. Psychoneuroendocrinology. 2004;29:516–528. doi: 10.1016/S0306-4530(03)00072-6. [DOI] [PubMed] [Google Scholar]
  26. Lauderdale DS, Knutson KL, Yan LL, Liu K, Rathouz PJ. Sleep duration: How well do self-reports reflect objective measures? The CARDIA sleep study. Epidemiology. 2008;19:838–845. doi: 10.1097/EDE.0b013e318187a7b0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Leggett AN, Zarit SH, Kim K, Almeida DM, Klein LC. Depressive mood, anger, and daily cortisol of caregivers on high and low stress days. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences. 2014 doi: 10.1093/geronb/gbu070. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Littell RC, Miliken GA, Stroup WW, Wolfinger RD. SAS systems for mixed models. Cary, NC: SAS Institute; 1996. [Google Scholar]
  29. Lovell B, Moss M, Wetherell MA. Perceived stress, common health complaints and diurnal patterns of cortisol secretion in young, otherwise healthy individuals. Hormones and Behavior. 2011;60(3):301–305. doi: 10.1016/j.yhbeh.2011.06.007. doi:http://dx.doi.org/10.1016/j.yhbeh.2011.06.007. [DOI] [PubMed] [Google Scholar]
  30. McCurry S, Gibbons LE, Logsdon RG, Vitiello M, Teri L. Training caregivers to change the sleep hygiene practices of patients with dementia: The NITE-AD Project. Journal of the American Geriatrics Society. 2003;51:1455–1460. doi: 10.1046/j.1532-5415.2003.51466.x. doi: 10.1046/j.1532-5415.200351466.x. [DOI] [PubMed] [Google Scholar]
  31. McCurry SM, Logsdon RG, Teri L, Vitiello MV. Sleep disturbances in caregivers of persons with dementia: Contributing factors and treatment implications. Sleep Medicine Review. 2007;11:143–153. doi: 10.1016/j.smrv.2006.09.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. McCurry SM, Logsdon RG, Vitiello MV, Teri L. Successful behavioral treatment for reported sleep problems in elderly caregivers of dementia patients: A controlled study. Journal of Gerontology: Psychological Sciences. 1998;53B:122–129. doi: 10.1093/geronb/53B.2.P122. [DOI] [PubMed] [Google Scholar]
  33. McCurry SM, Pike KC, Vitiello MV, Logsdon RG, Teri L. Factors associated with concordance and variability of sleep quality in persons with Alzheimer's disease and their caregivers. Sleep. 2008;31:741–748. doi: 10.1093/sleep/31.5.741. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Mezick EJ, Matthew KA, Hall M, Kamarck TW, Buysse DJ, Owens JF, Reis SE. Intra-individual variability in sleep duration and fragmentation: Associations with stress. Psychoneuroendocrinology. 2009;34:1346–1354. doi: 10.1016/j.psyneuen.2009.04.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Piazza JR, Almeida DM, Dmitrieva NO, Klein LC. Frontiers in the use of biomarkers of health in research on stress and aging. Journal of Gerontology: Psychological Sciences. 2010:65B, 513–525. doi: 10.1093/geronb/gbq049. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Preacher KJ, Curran PJ, Bauer DJ. Computational tools for probing interaction effects in multiple linear regression, multilevel modeling, and latent curve analysis. Journal of Educational and Behavioral Statistics. 2006;31:437–448. [Google Scholar]
  37. Raudenbush SW, Bryk AS. Hierarchical linear models: Applications and data analysis methods. 2nd. London, England: Sage; 2002. [Google Scholar]
  38. Saper CB, Scammell TE, Lu J. Hypothalamic regulation of sleep and circadian rhythms. Nature. 2005;437:1257–1263. doi: 10.1038/nature04284. [DOI] [PubMed] [Google Scholar]
  39. Schmidt-Reinwald A, Pruessner JC, Hellhammer DH, Federenko I, Rohleder N, Schurmeyer TH, Kirschbaum C. The cortisol response to awakening in relation to different challenge tests and a 12-hour cortisol rhythm. Life Sciences. 1999;64:1653–1660. doi: 10.1016/s0024-3205(99)00103-4. [DOI] [PubMed] [Google Scholar]
  40. Schlotz W, Hellhammer J, Schulz P, Stone AA. Perceived work overload and chronic worrying predict weekend-weekday differences in the cortisol awakening response. Psychosomatic Medicine. 2004;66:207–214. doi: 10.1097/01.psy.0000116715.78238.56. [DOI] [PubMed] [Google Scholar]
  41. Schoenborn CA, Adams PF, Peregoy JA. Vital Health Statistics. 257. Vol. 10. National Center for Health Statistics; 2013. Health behaviors of adults: United States, 2008-2010. [PubMed] [Google Scholar]
  42. Seltzer MM, Greenberg JS, Hong J, Smith LE, Almeida DM, Coe C, Stawski RS. Maternal cortisol levels and behavior problems in adolescents and adults with ASD. Journal of Autism and Developmental Disorders. 2010;40:457–469. doi: 10.1007/s10803-009-0887-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Sternberg EM, Gold PW. The mind-body interaction in disease. Scientific American (Updated from The Mysteries of the Mind: Special Issue, 1997) 2002:82–89. [Google Scholar]
  44. Tassi P, Muzet A. Sleep inertia. Sleep Medicine Reviews. 2000;4(4):341–353. doi: 10.1053/smrv.2000.0098. doi:http://dx.doi.org/10.1053/smrv.2000.0098. [DOI] [PubMed] [Google Scholar]
  45. Thorn L, Hucklebridge F, Evans P, Clow A. Suspected non-adherence and weekend versus week day differences in the awakening cortisol response. Psychoneuroendocrinology. 2006;31:1009–1018. doi: 10.1016/j.psyneuen.2006.05.012. [DOI] [PubMed] [Google Scholar]
  46. van Cauter E, Turek FW. Endocrine and other biological rhythms. Philadelphia: W.B. Saunders; 1995. [Google Scholar]
  47. Vreeburg SA, Kruijtzer BP, van Pelt J, van Dyck R, DeRijk RH, Hoogendijk WJG, Penninx BWJH. Associations between sociodemographic, sampling and health factors and various salivary cortisol indicators in a large sample without psychopathology. Psychoneuroendocrinology. 2009;34:1109–1120. doi: 10.1016/j.psyneuen.2009.04.024. [DOI] [PubMed] [Google Scholar]
  48. Wilhelm I, Born J, Kudielka BM, Schlotz W, Wüst S. Is the cortisol awakening rise a response to awakening? Psychoneuroendocrinology. 2007;32:358–366. doi: 10.1016/j.psyneuen.2007.01.008. [DOI] [PubMed] [Google Scholar]
  49. Wüst S, Wolf J, Hellhammer D, Federenko I, Schommer N, Kirschbaum C. The cortisol awakening response- Normal values and confounds. Noise & Health. 2000;2:79–88. [PubMed] [Google Scholar]
  50. Zarit SH, Kim K, Femia EE, Almeida DM, Savla J, Molenaar PCM. Effects of adult day care on daily stress of caregivers: A within-person approach. Journal of Gerontology: Psychological Sciences. 2011;66B:538–546. doi: 10.1093/geronb/gbr030. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES