Effects of Stress on Sleep: The Moderating Role of Coping Style
Avi Sadeh, Giora Keinan, and Keren Daon Tel Aviv University
The aim of this prospective quasi-experimental study was to assess the role of coping style as a factor moderating the relationship between stress and sleep. Sleep of 36 students was assessed by means of actigraphy and daily logs during low-stress and high-stress periods. The high-stress period was the week that the students were evaluated for acceptance to graduate programs in clinical psychology. The low-stress period was a regular academic week. The students’ ways of coping were assessed during the baseline low-stress period using the COPE inventory. Data analysis revealed that a high emotion-focused coping score was significantly predictive of reduction in sleep time from the low- to the high-stress period. These results suggest that coping style is a key factor in assessing the relationship between stress and sleep.
Key words: sleep, stress, actigraphy, coping
Research on the effects of stress on sleep has yielded confusing and inconsistent results. Many studies have associated stress with sleep difficulties in infants, children, and adults (Cartwright & Wood, 1991; Healey et al., 1981; Lavie, 2001; Lundh & Broman, 2000; Pillar, Malhotra, & Lavie, 2000; Ross, Ball, Sullivan, & Caroff, 1989; Sadeh, 1996; Van Reeth et al., 2000; Waters, Ad- ams, Binks, & Varnado, 1993). These sleep problems included difficulty falling asleep, fragmented sleep, and recurrent and fre- quent nightmares. A few studies have also shown that induced stress can increase arousal and delay sleep onset (Gross & Bork- ovec, 1982; Van Egeren, Haynes, Franzen, & Hamilton, 1983). However, other studies have failed to detect these disruptions or have documented more resilient sleep in association with stressful conditions or posttraumatic stress disorder (Dagan, Zinger, & Lavie, 1997; Lavie, Carmeli, Mevorach, & Liberman, 1991; Pillar et al., 2000).
Stress theories offer some rationale for these seemingly incon- sistent findings. Analysis of the relevant sleep and stress literature (Sadeh, 1996; Sadeh & Gruber, 2002) has led to the identification of two distinct modes of relevant response of the sleep–wake system: (a) the “turn on” response of the “alarm phase” (Selye, 1983), which is compatible with hypervigilance and incompatible with sleep, and (b) the “shut off” response of the stage of exhaus- tion (Selye, 1983), which leads to reduced activity, preservation of energy, and is compatible with extended sleep. The shut off response is also compatible with the conservation-withdrawal hy- pothesis (Engel & Schmale, 1972), which suggests that under conditions of severe and uncontrollable stress the organism tends to withdraw from activity and preserve energy until the circum- stances change.
These two response modes or biobehavioral systems appear to relate to key factors or principal ways of coping in the coping literature (Folkman & Lazarus, 1980; Skinner, Edge, Altman, & Sherwood, 2003). The conceptualization of ways of coping relies on three different aims for the individual’s behavioral and cogni- tive efforts: (a) problem-focused coping (PFC) focuses on manag- ing or altering the problem causing the distress, (b) emotion- focused coping (EFC) is aimed at regulating emotional responses to the problem, and (c) disengagement (DE) is aimed at disengag- ing from the threatening stimuli and related thoughts and emotions.
Analysis of the sleep and stress literature (Sadeh, 1996; Sadeh & Gruber, 2002) suggests that a coping style that favors dissociation and disengagement would be more likely to lead to “escape to sleep.” Increased sleep is likely to reduce exposure to the stressor and the associated adverse feelings and cognitions. Thus, sleep in itself has a role in regulating the exposure to stress and its effects. However, sleep is incompatible with emotional coping, and sleep itself represents loss of control. Therefore, high EFC is more likely to increase alertness and thus produce physiological arousal (Se- lye, 1983) and disrupt or reduce sleep.
The individual’s coping style has rarely been addressed in sleep and stress studies. The first study that has explored this issue was a retrospective study of sleep and coping style conducted with college students (Hicks, Marical, & Conti, 1991). In this study, the authors examined the coping style with the Ways of Coping assessment questionnaire (Folkman & Lazarus, 1980). Their find- ings indicated that short sleepers used more EFC than long sleep- ers. Recently, Morin, Rodrigue, and Ivers (2003) reported that, in comparison with controls, insomnia patients relied more on emotion-oriented coping strategies and perceived their lives as more stressful. Taken together, these studies supported the idea that EFC is associated with compromised sleep.
Our quasi-experimental study was based on exposure to a highly stressful period in the lives of young undergraduate psychology students who apply for graduate programs in clinical psychology. These are highly competitive programs with a 1:10 acceptance ratio. The most intense period is the week in which the students are being interviewed and assessed for the program. Many of them consider this a critical period in their career that would determine
Avi Sadeh, Giora Keinan, and Keren Daon, Department of Psychology, Tel Aviv University, Ramat Aviv, Israel.
We thank Helene and Woolf Marmot for their support and Ornit Arbel and Jeri Hahn-Markowitz for their help in preparing this article.
Correspondence concerning this article should be addressed to Avi Sadeh, Department of Psychology, Tel Aviv University, Ramat Aviv 69978, Israel. E-mail: sadeh@post.tau.ac.il
Health Psychology Copyright 2004 by the American Psychological Association 2004, Vol. 23, No. 5, 542–545 0278-6133/04/$12.00 DOI: 10.1037/0278-6133.23.5.542
542
Th is
d oc
um en
t i s c
op yr
ig ht
ed b
y th
e A
m er
ic an
P sy
ch ol
og ic
al A
ss oc
ia tio
n or
o ne
o f i
ts a
lli ed
p ub
lis he
rs .
Th is
a rti
cl e
is in
te nd
ed so
le ly
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
their ability to pursue their most desirable career track. Previous observations and subjective reports indicated that the students indeed consider this period as one of the most stressful periods of their lives (see also Keinan & Sivan, 2001).
In the present study, we assessed the role of the individual’s coping style in moderating the link between stress and sleep. However, unlike previous studies, our study included a quasi- experimental design and objective sleep measures. On the basis of our theoretical perspective and previous findings (Hicks et al., 1991; Morin et al., 2003), we hypothesized that (a) the sleep of individuals with high EFC would be shortened and/or become more disrupted during the high-stress period than during the low- stress period and (b) the sleep of individuals with high DE would be extended and/or less disrupted during the high-stress period than during the low-stress period. No specific hypothesis was made with regard to PFC. However, because PFC plays a major role in the conceptualization of coping and adjustment (Folkman & Lazarus, 1980; Skinner et al., 2003), we have decided to include this measure in our study to evaluate its contribution.
Method
Participants
Thirty-six undergraduate students (28 women and 8 men) majoring in psychology participated in the study. Their ages ranged between 22 and 32 years (M � 24.75, SD � 2.17), and they were all healthy and medication free. These students were chosen because they all applied to graduate programs in clinical psychology, they agreed to participate in the study on a voluntary basis, and they all signed informed consent forms. Twelve students declined participation because of overload or technical reasons.
Measures
Assessment of ways of coping. Each participant completed the COPE inventory (Carver, Scheier, & Weintraub, 1989) during the low-stress week. The COPE inventory is based on the model and earlier question- naires developed by Folkman and Lazarus (1980, 1985). This questionnaire consists of 52 items that describe different behaviors and mental activities associated with coping with stress. The participants were asked to rate the level of their own use of each behavior described in each item (e.g., “I learn to live with it” “I think hard about what steps to take” “I let my feelings out”). In the present study, we used three global factors: PFC, EFC, and DE. Acceptable Cronbach’s alpha reliability coefficients were reported by the authors (Carver et al., 1989), and in the present study, coefficients of .82, .80, and .59 were obtained for PFC, EFC, and DE, respectively.
Assessment of sleep: Actigraphy. Each participant was monitored with an actigraph (AMA-32, Ambulatory Monitoring Inc., Ardsley, NY) for 5 consecutive nights in each monitoring week (low-stress and high-stress weeks). Participants were asked to attach the actigraph to their nondomi- nant wrist when they went to bed and to remove it after waking up in the morning. The actigraphs were initialized to collect data in Mode 18, with 1-min data-collection intervals. Actigraphic raw data were translated into sleep measures using the Actigraphic Scoring Analysis program for an IBM-compatible PC (Sadeh, Sharkey, & Carskadon, 1994). These sleep– wake measures have been validated against polysomnography with agree- ment rates for sleep–wake identification higher than 90% (Sadeh et al., 1994).
Actigraphic sleep measures included the following: (a) sleep onset time (the first minute after reported bedtime that was identified as sleep by the Actigraphic Scoring Analysis program sleep–wake algorithm, which was followed by at least 15 min of uninterrupted sleep); (b) sleep period (from sleep onset time to morning awakening); (c) true sleep time (sleep time excluding all periods of wakefulness); (d) sleep percentage (percentage of
true sleep time from total sleep period); (e) night wakings (number of wakings that lasted 5 min or longer and that were preceded and followed by at least 15 min of uninterrupted sleep); (f) morning rise time (the first minute of the first episode of wakefulness following the last episode of sleep).
Assessment of sleep and stress: Daily logs. In addition to the objective sleep measures, daily logs were used to obtain subjective information from the participants. The subjective information was used for assessing a subjective perception of sleep quality and a self-reported stress level. The daily logs included the following scales: (a) Morning Stress Level, a 9-point scale ranging from 1 (not stressed at all) to 9 (very stressed); (b) Evening Stress Level, identical to the previous scale; (c) Perceived Sleep Quality, a 3-point scale ranging from 1 (very poor sleep) to 3 (very good sleep); and (d) Difficulty Falling Asleep, a 3-point scale ranging from 1 (very easy) to 3 (very difficult). We collected multiple stress ratings (morning and evening each day) to increase the reliability of the measure.
Procedures
The study was reviewed and approved by the Ethics Committee of the Department of Psychology at Tel Aviv University. Three to 4 months prior to the scheduled admission evaluation process, the students were contacted and asked to participate in the study. The students were recruited by a peer research assistant who administered all of the procedures and reassured them that their identity would remain anonymous and they would not be linked in any way to the information collected in the study. All data were handled with the study’s participant number, and the identity of the participants remained sealed. The participants were informed that the general aim of the study was to explore the relationship between sleep and stress. During a low-stress regular academic week, 2–3 months before the admission evaluation period, each participant completed the COPE inven- tory to assess ways of coping. For a period of 5 days, the participants were asked to wear an actigraph and complete daily logs assessing their per- ceived sleep and stress. During the high-stress period, the participants repeated the sleep and stress assessment using actigraphy and daily logs. A research assistant contacted the participants during the study to verify compliance with the procedures. The students received charts illustrating their sleep patterns following the completion of the study.
Results
Our data analyses included the following: (a) analysis of changes in subjective perception of stress from low-stress to high- stress monitoring period; (b) analysis of changes in sleep patterns between these two periods, considering the ways of coping as moderating factors; and (c) analysis of additional links among the sleep and stress variables.
Changes in Perceived Stress Level
To check if the applicants’ evaluation week was perceived as more stressful than the regular week, we calculated the stress scores for each individual for each period (low and high stress). The participants completed ratings of stress level twice each day (in the morning when awakening and in the evening before bed- time), and we used two statistics, the averaged and maximum ratings, to characterize each period for each individual. The max- imum ratings were used to capture the highest stress levels asso- ciated with specific events (such as the day of interview). The results indicate that stress levels during the interview week were significantly higher in comparison with the baseline week on three of the four measures. Morning maximum stress level during the high-stress week (M � 6.31, SD � 1.47) was significantly higher in comparison with the low-stress week (M � 4.89, SD � 1.82),
543BRIEF REPORTS
Th is
d oc
um en
t i s c
op yr
ig ht
ed b
y th
e A
m er
ic an
P sy
ch ol
og ic
al A
ss oc
ia tio
n or
o ne
o f i
ts a
lli ed
p ub
lis he
rs .
Th is
a rti
cl e
is in
te nd
ed so
le ly
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
