- Email: [email protected]
Attempts to control unwanted thoughts in the night: development of the thought control questionnaire-insomnia revised (TCQI-R)
ARTICLE IN PRESS
Behaviour Research and Therapy 43 (2005) 985–998 www.elsevier.com/locate/brat
Attempts to control unwanted thoughts in the night: development of the thought control questionnaire-insomnia revised (TCQI-R) Melissa J. Reea, Allison G. Harveya,c,, Rachel Blakeb, Nicole K.Y. Tangc, Metka Shawe-Taylord a
b
Department of Psychiatry, University of Oxford, Warneford Hospital, Oxford OX3 7JX, UK Department of Psychology, University of Surrey, Guildford and South London and Maudsley NHS Trust, UK c Department of Experimental Psychology, University of Oxford, UK d Surrey Oaklands NHS Trust, UK Received 14 January 2004; received in revised form 2 July 2004; accepted 7 July 2004
Abstract The attempted control of intrusive, uncontrollable thoughts has been implicated in the maintenance of a range of psychological disorders. The current paper describes the refinement of the Thought Control Questionnaire Insomnia (TCQI; Behav. Cogn. Psychoth. 29 (2001)) through its administration to a sample (n=385) including good sleepers and individuals with insomnia. Several items with poor psychometric properties were discarded, resulting in a 35-item revised TCQI. Factor analysis revealed six factors; aggressive suppression, cognitive distraction, reappraisal, social avoidance, behavioural distraction, and worry. The attempted management of unwanted thoughts was compared across individuals with insomnia and good sleepers, and the impact of these strategies on sleep quality, anxiety and depression was investigated. With the exception of cognitive distraction, individuals with insomnia, relative to good sleepers, more frequently used every thought control strategy. The strategies of aggressive suppression and worry, in particular, appeared to be unhelpful, with the use of these strategies predicting sleep impairment,
Corresponding author. Department of Psychiatry, University of Oxford, Warneford Hospital, Headington, Oxford
OX3 7JX, UK. Tel: +44-1865 223 912; fax: +44-1865-793-101. E-mail address: [email protected] (A.G. Harvey). 0005-7967/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.brat.2004.07.003
ARTICLE IN PRESS 986
M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
anxiety and depression. The strategy of cognitive distraction appeared to be helpful, with the use of this strategy predicting better sleep quality. r 2004 Elsevier Ltd. All rights reserved. Keywords: Insomnia; Suppression; Worry; Distraction; Avoidance; Reappraisal
1. Introduction The attempted control of intrusive, uncontrollable thoughts has been identified as an important feature in a range of psychological disorders including obsessive compulsive disorder, post traumatic stress disorder, social phobia, insomnia, and depression (e.g., Abramowitz, Whiteside, Kalsy, & Tolin, 2003; Harvey, 2002a; Harvey & Bryant, 1998; Reynolds & Wells, 1999). A body of research is emerging to endorse the proposal that some thought control strategies are helpful but that others are unhelpful because they actually perpetuate unwanted thoughts rather than put an end to them (e.g., Abramowitz et al., 2003; Amir, Cashman, & Foa, 1997; Harvey, 2001; Salkovskis & Campbell, 1994). In order to aid the systematic investigation of strategies used in an attempt to control unwanted thoughts, Wells and Davies (1994) developed the Thought Control Questionnaire (TCQ) which asks respondents to indicate the frequency with which they employ each of thirty different thought control strategies. The development of the TCQ led to the identification of five dimensions of thought control; namely, punishment, worry, distraction, social control, and reappraisal. A factor analysis of a preliminary version of the TCQ suggested that it contained six factors (including two distraction factors: cognitive distraction and behavioural distraction), while a subsequent analysis revealed only five factors because the distraction factors collapsed into one (Wells & Davies, 1994). Interestingly, in a subsequent investigation of the psychometric properties of the TCQ in a clinical sample (Reynolds & Wells, 1999), support for the distraction scale being split into behavioural and cognitive dimensions was once again obtained. Research employing the TCQ to date has revealed an acceptably robust factor structure, but whether the distraction scale comprises one or two factors is unclear. Since the development of the TCQ, research has been broadly consistent in suggesting that certain thought control strategies are associated with psychopathology. Strategies of punishment (e.g., ‘I punish myself for having the thought’) and worry (e.g., ‘I focus on different negative thoughts’) have been identified to be associated with more severe symptoms across a number of psychological disorders (e.g., Amir et al., 1997; Andrews, Troop, Joseph, Hiskey, & Coyne, 2002; Reynolds & Wells, 1999). Conversely, the strategy of social control (e.g., ‘I ask my friends if they have similar thoughts’) may be an adaptive strategy that may provide a protective buffer against psychological distress (Blake, Shawe-Taylor, & Murray, 2003; Harvey, 2001; Wells & Davies, 1994). The TCQ, then, has proved to be a useful assessment tool in the investigation of thought control strategies across a range of disorders. It can be concluded that thought control is a multifaceted construct, with some thought control strategies being unhelpful and others being more effective in the management of unwanted thoughts and images. A ‘racing mind’ while lying in bed is as a key symptom of insomnia (e.g., Espie, 1991; Harvey, 2000; Lichstein & Rosenthal, 1980) and hence difficulty with thought control has been implicated
ARTICLE IN PRESS M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
987
in the maintenance of insomnia and potentially its treatment (e.g., Harvey, 2002a, Harvey, in press). Accordingly, Harvey (2001) developed an adaptation of the TCQ specifically for use in insomnia (see method section for a description of this measure). Employing the TCQI the frequency with which thoughts interfered with sleep was associated with the strategies of suppression, punishment, reappraisal, and worry (Harvey, 2001). The data also revealed that individuals with insomnia used the thought control strategies of suppression, reappraisal, and worry more frequently than good-sleeper controls. Social control and replacement strategies, however, were associated with better sleep. While the data produced by the TCQI is promising, its psychometric properties need to be investigated. Accordingly, the primary aim of the current study was to assess the psychometric properties of the TCQI and revise it as necessary in order to enhance its psychometric properties. A secondary aim was to replicate and extend Harvey (2001) results by investigating thought control strategies employed by good sleepers and patients with insomnia with a larger sample of participants than employed by Harvey (2001); n ¼ 60; and to examine the relationship between thought control strategies and measures of insomnia severity, anxiety, and depression.
2. Method 2.1. Participants In total, 385 participants completed the TCQI for this study. One hundred and ninety one individuals met criteria for primary insomnia (Diagnostic and Statistical Manual for Mental Disorders-IV criteria: American Psychiatric Association, 1994), 64 of whom where treatmentseeking and 127 who were not treatment-seeking but responded to recruitment advertisements. Participants seeking treatment for insomnia were recruited primarily (n=58) via advertisements at general practitioner surgeries, while 6 were recruited from an Internet advertisement for a research study offering a psychological treatment for insomnia. Of the treatment-seeking participants, 10 reported occasional use of sleeping tablets (twice per week or less). Good sleepers were 96 university students (n=65) and members of the general public (n=31) who responded to recruitment advertisements. Finally, 98 non-treatment-seeking individuals participated. These were self-reported poor sleepers but did not meet diagnostic criteria. They were regarded as a subthreshold group. Information regarding use of sleeping tablets was not sought from the nontreatment seeking component of the sample. The sample had a mean number of years education of 13.2 (SD=2.14). The mean age of respondents was 27.41 years (SD=11.42) and the range was 18 to 63 years. Females constituted 63% (n=241) of the sample, males constituted 36% (n=139), and 5 respondents did not indicate their sex. 2.2. Measures For individuals seeking treatment for insomnia, all measures were administered in an assessment phase prior to commencement of treatment. For non-treatment seeking participants,
ARTICLE IN PRESS 988
M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
measures were completed either in a laboratory testing session, or in the participant’s home, in which case they posted completed questionnaires to the experimenter. TCQI: The 43-item TCQI (Harvey, 2001) was administered to all participants. Items were rated on a four-point scale where 1=almost never and 4=almost always, yielding a minimum score of 43 and a maximum score of 172. The Thought Control Questionnaire-Insomnia (TCQI) differs from the TCQ in several ways: Firstly, the instructions of the TCQI ask respondents to indicate the frequency (never, sometimes, often, almost always) with which they employ each thought control strategy while being kept awake by thoughts. Secondly, the TCQI contains thirteen additional items relevant to insomnia (e.g., ‘I count sheep’, ‘I decide to put them ‘‘on hold’’ until the morning’). Thirdly, the TCQI contains a suppression and a replacement scale in place of the original distraction scale of the TCQ. The decision to divide the distraction scale in this way was based on evidence suggesting that it is important to differentiate between thought suppression and replacement (Salkovskis & Campbell, 1994). The TCQI suppression scale was comprised of items describing what Salkovskis and Campbell (1994) refer to as ‘simple distraction’. Simple distraction comprises strategies that divert attention away from the unwanted thought without using a specific task to replace it (e.g., ‘I try to push the thoughts out of my head’). The TCQI replacement scale contained items describing strategies that replace the unwanted thought with another thought (e.g., ‘I call to mind positive images instead’). A final difference between the TCQ and TCQI is that the TCQI has an additional question at the beginning of the questionnaire that asks respondents to indicate the frequency with which thoughts keep them awake at night (where 0=never, 10=every night). Insomnia Diagnostic Interview (IDI): In the absence of a psychometrically validated interview corresponding to the DSM-IV criteria for primary insomnia, a structured clinical interview, the Insomnia Diagnostic Interview (IDI: Harvey, Nelson, Neitzert-Semler, Tang, Sharpley, & Ree, 2003), was used to determine the presence of insomnia. This interview schedule has recently been validated against clinician-based diagnoses of insomnia on 55 individuals (Harvey et al., 2003). The results showed good internal consistency (a=0.87), sensitivity (92%), specificity (89%), and test–retest reliability (r=0.90), as well as high diagnostic agreement for the presence (90%) and absence of insomnia (92%). Participants were classified as good sleepers on the basis of their selfreport that they slept ‘very well’ or ‘reasonably well’ on a screening question. These participants were not administered the IDI. Pittsburgh Sleep Quality Index (PSQI; Buysse, Reynolds, Monk, Berman, & Kupfer, 1989): In order that the association between the TCQI and insomnia severity could be investigated, 190 of the participants also completed the PSQI. The PSQI contains 19 items that sum to yield a global score (higher scores indicate more severe impairment) and has sound psychometric properties (Buysse et al., 1989; Carpenter & Andrykowski, 1998). The PSQI has been found to differentiate those with and without insomnia with a 90% sensitivity and 87% specificity. In addition, it’s internal consistency (a=0.83) and test-retest reliability (r=0.85) are acceptable (Buysse et al., 1989). Beck Anxiety Inventory (BAI; Beck, Brown, Epstein, & Steer, 1988): In order that the association between the TCQI and anxiety symptoms could be investigated 134 of the participants completed the BAI. Beck Depression Inventory (BDI; Beck, Ward, Mendelson, Mock, & Erbaugh, 1961): In order that the association between the TCQI and depression could be investigated 184 of the participants completed the BDI.
ARTICLE IN PRESS M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
989
3. Results 3.1. Overview In line with the first aim of the current paper, to investigate the psychometric properties of the TCQI and to make revisions on the basis of these, the first analyses conducted were to select the psychometrically adequate items in the TCQI, and to assess the factor structure of the questionnaire. The second aim of the present paper was to assess the thought control strategies employed by good sleepers and those with insomnia, and to investigate the relationship between thought control strategies and measures of insomnia severity, anxiety, and depression. Correlational analyses were employed to achieve this second aim. Since the stability of the TCQ-I-Revised factor structure has not yet been cross-validated with independent samples, the TCQ-I-Revised total score is also included in the analyses. It is important to note, however, that this total score may not be readily interpretable as some of the scales are proposed to be helpful while others are proposed to be unhelpful. 3.2. Selection of items for factor analysis 1. Items were deleted if they were not easily interpretable in the context of insomnia. Specifically, the items ‘I find out how my friends deal with these thoughts’, ‘I ask my friends if they have similar thoughts’, and ‘I talk to a friend about the thought’ were removed as they are not strategies that could easily be employed while in bed trying to sleep. 2. Items which did not employ the full response range of the scale were removed as these items did not have a high capacity to discriminate between participants. Specifically, the items ‘I slap or pinch myself to stop the thought’ and ‘I think more about the minor problems I have’ were removed as no participant had responded to these items with a ‘4’ (almost always). 3. The percentage of missing values for each item was examined. No item had a missing value rate of above 5% so no items were discarded due to a high rate of missing values. 4. In order to assess for item redundancy, the correlation matrix of the remaining 38 items was inspected. Items which showed a correlation of above 0.60 with other items were examined to determine if this could be explained by highly similar item content (as in Rapee, Craske, Brown, & Barlow, 1996). For example, the item ‘I don’t talk about the thought to anyone’ correlated highly with the items ‘I keep the thought to myself’, (r=0.65) and ‘I avoid discussing the thought’(r=0.60) and so it was considered redundant and discarded. Therefore, 37 items remained for factor analysis. 3.3. Factor analysis Since the TCQI differs quite substantially from the TCQ, and since a factor analysis has not yet been performed on the TCQI, it was considered appropriate to perform an exploratory factor analysis (EFA) rather than a confirmatory factor analysis on the data. A principal components analysis with oblique (Direct Quartimin) rotation and Kaiser Normalisation was performed on the data (37 items) as the factors were expected to be correlated. Missing values were replaced with the mean value for that item. Inspection of the break in slope
ARTICLE IN PRESS 990
M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
on the scree plot indicated that six factors should be retained in the final solution. These six retained factors accounted for 47.03% of the total variance. The six factors were labelled as follows: aggressive suppression, cognitive distraction, reappraisal, social avoidance, behavioural distraction, and worry. Factor loadings are listed in Table 1. As can be seen from Table 1, the factor structure is mostly in line with that of the TCQ (Wells & Davies, 1994) and that proposed by Harvey (2001) for the TCQI. The main difference between the factor structure suggested by Harvey and the factor structure obtained here is that the suppression and the punishment factors were combined into an ‘aggressive suppression’ factor. Also, the existence of a cognitive distraction factor and a behavioural distraction factor was supported (in line with Wells & Davies, 1994). Items 3 and 9 did not load highly on any factor and so were excluded from subsequent analyses. 3.4. Scale reliabilities In order to estimate the internal consistency of the scales, Cronbach alpha coefficients were calculated as follows: whole scale=0.83, aggressive suppression=0.79, cognitive distraction=0.64, reappraisal=0.76, social avoidance=0.69, behavioural distraction=0.66, and worry=0.78. These results were similar to those found by Wells and Davies (1994) and Reynolds and Wells (1999). 3.5. Scale intercorrelations Table 2 presents the Pearson correlations between the TCQI-Revised scales and the total score. The initial question ‘how often does thinking too much keep you awake’ was also included in this analysis. This question is answered on an 11-point scale where 0=Never and 10=Every night. The scale intercorrelations on the TCQI-Revised were weak to moderate. 3.6. TCQI-revised discrimination of good and poor sleepers In order to assess whether the factors of the TCQI-Revised are able to discriminate good sleepers and those with insomnia, mean scale scores were compared across these two groups. Table 3 indicates that all but the cognitive distraction scale discriminated good sleepers and those with insomnia. The ability of the TCQ-I-Revised total score to discriminate good and poor sleepers was also assessed. The ability of the initial question on the TCQI-Revised (‘how often does thinking too much keep you awake?’) to discriminate between good sleepers and those with insomnia was also assessed. The mean for the good sleeper sample was 4.81 (SD=2.73) while the mean for the insomnia sample was 7.26 (SD=1.94). This difference was significant at the 0.01 level [t(259)=8.49]. 3.7. Correlations between TCQI-revised and the PSQI, BAI, and BDI Correlations were inspected in order to investigate the association between thought control strategies and self-report psychopathology measures. Table 4 displays correlations between the
ARTICLE IN PRESS M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
991
Table 1 Items and factor loadings of the TCQI-revised items Scale/Item
Factor loadings 1
Aggressive suppression (20) I get angry at myself for having the thought (1) I tell myself not to think about them now (11) I tell myself not to be so stupid (2) I try to push the thoughts out of my head (42) I say ‘stop’ to myself (22) I shout at myself for having the thought (15) I punish myself for having the thought (37) I tell myself that something bad will happen if I think the thought
Behavioural distraction/suppression (30) I occupy myself with work instead (39) I keep myself busy (28) I do something I enjoy (6) I try to block them out by reading, watching TV, or listening to the radio (43) I do something physical Social avoidance (17) I keep the thought to myself (21) I avoid discussing the thought (41) I seek reassurance form others (e.g., my bed partner or a friend the next day) Worry (35) I focus on different negative thoughts (27) I worry about more minor things (13) I replace the thought with a more trivial bad thought (16) I dwell on other worries (33) I think about past worries instead
3
4
5
6
0.71 0.68 0.67 0.64 0.63 0.50 0.45 0.36
Cognitive distraction/suppression (25) I think pleasant thoughts instead (18) I think about something else instead (4) I call to mind positive images instead (8) I decide to put them ‘on hold’ until the morning (10) I let my mind go blank Reappraisal (23) I analyse the thought rationally (32) I try a different way of thinking about it (40) I prefer to think things through rather than distract from them (29) I try to reinterpret the thought (19) I challenge the thought’s validity (36) I question the reasons for having the thought (5) If the thoughts relate to a problem I make a decision about it in order to solve the problem
2
0.71 0.69 0.61 0.53 0.42 0.76 0.68 0.67
0.37
0.64 0.54 0.50 0.54
0.76 0.75 0.53 0.48 0.38 0.82 0.78 0.66
0.78 0.73 0.68 0.66 0.60
ARTICLE IN PRESS 992
M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
Table 1 (continued ) Scale/Item
Factor loadings 1
(12) I focus on the thought (7) I ruminate about them (3) I count sheep (9) I get out of bed and write about them
2
3
4
5
6 0.42 0.35
0.35
Note: Loadings below .35 not shown. Table 2 Scale intercorrelations N=385
TCQ-IRevised total
TCQ-I-Revised total Aggressive suppression Cognitive distraction Reappraisal Social avoidance Behavioural distraction Worry How often does thinking too much keep you awake?
1 0.77**
1
0.34** 0.67** 0.35** 0.62** 0.65** 0.35**
Aggressive suppression
Cognitive distraction
Reappraisal Social avoidance
Behavioural Worry distraction
0.22** 0.29** 0.24** 0.33**
1 0.13** 0.04 0.019**
1 0.07 0.29**
1 0.07
1
0.46** 0.33**
0.09 0.12*
0.32** 0.17**
0.18** 0.19**
0.30** 0.14**
1 0.45**
Note: * Significant at the 0.05 level (2-tailed); ** Significant at the 0.01 level (2-tailed).
Table 3 Comparison of TCQI-revised scale scores between good sleepers and individuals with insomnia Scale
Good sleeper mean (SD) n=96
Insomnia mean (SD) n=191
t(df)
Aggressive suppression Cognitive distraction Reappraisal Social avoidance Behavioural distraction Worry TCQIR total score
1.74 2.19 2.10 2.34 1.79 1.57 71.70
2.01** 2.11 2.33** 2.52* 2.05** 1.93** 80.08
4.23 1.31 3.36 2.11 4.25 5.99 5.64
Note: ** significant at the 0.01 level.
(285) (285) (285) (285) (285) (285) (285)
ARTICLE IN PRESS M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
993
Table 4 Pearson correlations between scales of the TCQI-revised and measures of sleep quality, anxiety and depression
PSQIa (n=192) BDI (n=195) BAI (n=136)
Aggressive suppression
Cognitive distraction
Reappraisal
Social avoidance
Behavioural distraction
Worry
TCQIR total score
0.32**
0.14*
0.12
0.22**
0.21**
0.46**
0.34**
0.33**
0.01
0.11
0.22**
0.25**
0.37**
0.36**
0.40**
0.11
0.24**
0.19**
0.22**
0.32**
0.43**
Note: * Significant at the 0.05 level (2-tailed); ** Significant at the 0.01 level (2-tailed). a The mean PSQI score for participants with insomnia was 10.03 (SD=2.93) and the mean for good sleepers was 6.58 (SD=2.78).
TCQI-Revised and the PSQI, BDI and BAI. Higher aggressive suppression, social avoidance, behavioural distraction, and worry scale scores were associated with worse sleep quality, and higher depression and anxiety scores. The reappraisal scale was only associated with anxiety, and the cognitive distraction scale was associated negatively with sleep quality. Pearson correlations between the initial question on the TCQI-Revised and measures of sleep quality, anxiety, and depression were also calculated, and are as follows: PSQI=0.72, po0.01; BAI=0.24, po0.01; BDI=0.34, po0.01. 3.8. Multiple regression analyses In order to investigate which thought control strategies predict poor sleep quality, BAI and BDI scores, three multiple regression analyses were performed. The six TCQI-Revised scales were entered as predictors, with all scales being entered simultaneously. The PSQI, BDI, and BAI scores were entered as dependent variables. The results of these analyses are presented in Table 5. Together, the scales of the TCQI-Revised predicted 29% of the variation in PSQI scores, 22% of the variation in BDI scores, and 21% of the variation in BAI scores. PSQI scores were predicted by more frequent use of aggressive suppression and worry, and by less frequent use of cognitive distraction. Depression and anxiety scores were also predicted by the aggressive suppression and worry scales (although the worry scale was only a marginally significant predictor of depression). Finally, the behavioural distraction subscale was a significant predictor of depression scores.
4. Discussion The primary aim of the present study was to revise the TCQI to assess and, if necessary, enhance its psychometric properties. With respect to this aim, TCQI items with poor psychometric properties were removed and the psychometric properties of the 35-item TCQI-Revised (TCQI-R) were examined. The TCQI-R yielded six factors (aggressive suppression, cognitive distraction,
ARTICLE IN PRESS 994
M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
Table 5 Summary of multiple regression analyses for TCQI-R factors predicting scores on measures of sleep quality, anxiety and depression Variable
Beta
T
p
Dependent variable: PSQI (n=192) Aggressive suppression Cognitive distraction Reappraisal Social avoidance Behavioural distraction Worry
0.20 0.18 0.06 0.16 0.10 0.35
2.74 2.69 0.91 1.79 1.36 4.79
0.007 0.008 0.365 0.074 0.172 0.000
Dependent variable: BDI (n=195) Aggressive suppression Cognitive distraction Reappraisal Social avoidance Behavioural distraction Worry
0.19 0.08 0.08 0.10 0.16 0.27
2.44 1.20 1.04 1.53 2.22 3.74
0.016 0.232 0.302 0.127 0.027 0.000
Dependent variable: BAI (n=136) Aggressive suppression Cognitive distraction Reappraisal Social avoidance Behavioural distraction Worry
0.27 0.04 0.10 0.05 0.05 0.17
2.93 0.46 1.11 0.61 0.60 1.91
0.004 0.641 0.271 0.545 0.547 0.058
reappraisal, social avoidance, behavioural distraction, worry) with a structure largely consistent with previous research on the TCQ (Wells & Davies, 1994). The internal consistency of the questionnaire and its scales were acceptable and also in line with previous research (Wells & Davies, 1994; Reynolds & Wells, 1999). The factor structure of the TCQI-R did, however, differ from that proposed for the TCQI in two ways. Firstly, an ‘aggressive suppression’ factor emerged that consisted of items from the punishment and suppression scales of the TCQI. Participants employing this strategy appeared to be using suppression or self-critical strategies in an attempt to control their thoughts. Secondly, behavioural and cognitive aspects of distraction were extracted as separate factors. This is consistent with Wells and Davies’ (1994) original proposal that cognitive and behavioural distraction are distinct thought control strategies. We note that this distinction has not consistently been found in past research. It is possible that behavioural and cognitive distraction were more distinct in the context of insomnia than they are for other psychological disorders. This suggestion is supported by the finding that cognitive distraction predicts better sleep quality and that behavioural distraction predicts worse sleep quality. Perhaps the deployment of cognitive distraction strategies is helpful because these strategies are compatible with sleep (e.g., ‘I think pleasant thoughts instead’), whereas the behavioural distraction strategies are less compatible with sleep (e.g., ‘I occupy myself with work instead’). In addition to displaying a readily interpretable factor structure, the TCQI-R discriminated between people with insomnia and good sleeper controls on five of its six scales (all scales other than
ARTICLE IN PRESS M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
995
cognitive distraction). This is consistent with previous research employing the TCQI (Harvey, 2001). Together, the psychometric properties of the individual items, the internal consistency of the questionnaire, its factor structure, and the ability of the TCQI-R to discriminate good sleepers and individuals with insomnia suggests that this is measure holds promise for future research investigating thought control strategies in insomnia. Turning now to the second aim of the paper, which was to examine the relationships between the TCQI-R and measures of insomnia severity, anxiety, and depression. Firstly, there was a positive correlation between severity of insomnia (measured by the PSQI) and self-reported frequency with which thoughts interfered with sleep (as assessed by the first question of the TCQIR). This finding is consistent with previous research suggesting that unwanted nocturnal cognitive activity is important in the maintenance of insomnia (e.g., Espie, 1991; Harvey, 2000; Lichstein & Rosenthal, 1980). Self-reported frequency of thoughts interfering with sleep was associated with the aggressive suppression and worry scales of the TCQI-R. Secondly, the relationship between the TCQI-R scales and measures of anxiety, depression, and severity of insomnia was examined. The more frequent use of aggressive suppression, behavioural distraction, social avoidance, and worry were each associated with higher anxiety, higher depression, and poorer sleep quality. The regression analyses also echoed the emerging pattern of aggressive suppression and worry being unhelpful thought control strategies, with poor sleep quality, anxiety, and depression being predicted by both of these. The suggestion that worry is an unhelpful thought control strategy is consistent with previous research using the TCQ (Wells & Davies, 1994) and investigating insomnia (Harvey, 2002; in press). The aggressive suppression scale of the TCQI-R contains items relating to punishment (e.g., ‘I get angry at myself for having the thought’) in addition to items relating to suppression (e.g., ‘I try to push the thoughts out of my head’). The finding that aggressive suppression predicted poor sleep, anxiety, and depression is consistent with previous experimental studies reporting adverse consequences following the deployment of thought suppression (Harvey, 2003; Wegner, 1989) and punishment (e.g., Reynolds & Wells, 1999). In addition to findings suggesting that some thought control strategies are unhelpful, the present study provided evidence to suggest that some thought control strategies may be helpful. The use of cognitive distraction as a thought control strategy predicted better sleep quality. This finding is consistent with previous experimental manipulations (Harvey & Payne, 2002). These findings need to be interpreted in the light of several limitations. First, participants were classified as good sleepers on the basis of their self-report of sleeping well or reasonably well. Future research should employ a more stringent method for classifying people as good sleepers including setting criteria for sleep onset latency, wake after sleep onset, total sleep time, and daytime fatigue and functioning. Second, the present study did not collect test-retest data for the TCQI-R. This will be an important direction in order to fully understand the psychometric properties of this questionnaire. Another avenue for future investigations is to further assess the inter-scale reliabilities of the TCQI-R. As with the TCQ, the scales with fewer items have more modest reliabilities, and the stability of these will need to be assessed with independent samples. In summary, the present study has refined the TCQI by eliminating items with poor psychometric properties. The resulting 35-item questionnaire, the TCQI-R, demonstrated a clear factor structure, good internal consistency, and an ability to discriminate between good sleepers and individuals with insomnia. This investigation has also suggested that individuals with
ARTICLE IN PRESS 996
M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
insomnia use worry, aggressive suppression, reappraisal, and behavioural distraction more than good sleepers. Worry and aggressive suppression were noted as unhelpful strategies that were associated with being kept awake by thoughts, and were predictors of poor sleep, depression, and anxiety scores. Cognitive distraction, however, may be a helpful thought control strategy. Future research employing the TCQI-R and thought control manipulation experiments will help to delineate the causal pathways between insomnia and strategies of thought control, and may constitute a process measure in treatment outcome research. Acknowledgment The authors would like to gratefully acknowledge Allison Bugg for her assistance in data collection. This work was supported by the Wellcome Trust (Grant no. 065913).
Appendix Thought control questionnaire insomnia-revised Many people find that as they are trying to get to sleep at night thoughts relating to the day they have just had or thoughts about tomorrow come to mind. Other times thoughts relating to ongoing problems or stressors at work or at home come to mind. Sometimes these thoughts make it hard to fall asleep. How often does thinking too much keep you awake? 0 1 2 3 4 5 6 Never
7
8
9 10 Every night
Below are a number of things that people do to control these thoughts. Please read each statement carefully and indicate how often you use each technique to control the thoughts that run through your mind as you are trying to get to sleep at night by circling the appropriate number. There are no right or wrong answers. Do not spend too much time thinking about each one. Almost never
Sometimes
When thoughts running through my mind keep me awake at night . . . 1. I tell myself not to think about them now . . . . . 1 2 2. I try to push the thoughts out of my head. . . . . 1 2 3. I call to mind positive images instead . . . . . . 1 2 4. If the thoughts relate to a problem I make a 1 2 decision about it in order to solve the problem . .....
Often
Almost always
3 3 3 3
4 4 4 4
ARTICLE IN PRESS M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
5. I try to block them out by reading a book, watching TV or listening to the radio . . . . . . . . ... 6. I ruminate about them . . . . . . . . . . . 7. I decide to put them ‘‘on hold’’ until the morning. . . 8. I let my mind go blank. . . . . . . . . . . 9. I tell myself not to be so stupid . . . . . . . . 10. I focus on the thought . . . . . . . . . . . 11. I replace the thought with a more trivial bad thought . 12. I punish myself for thinking the thought . . . . . 13. I dwell on other worries . . . . . . . . . . 14. I keep the thought to myself . . . . . . . . . 15. I think about something else instead . . . . . . . 16. I challenge the thoughts validity . . . . . . . . 17. I get angry at myself for having the thought . . . . 18. I avoid discussing the thought . . . . . . . . . 19. I shout at myself for having the thought. . . . . 20. I analyse the thought rationally . . . . . . . . 21. I think pleasant thoughts instead. . . . . . . . 22. I worry about more minor things instead. . . . . 23. I do something that I enjoy . . . . . . . . . 24. I try to reinterpret the thought . . . . . . . . 25. I occupy myself with work instead. . . . . . . 26. I try a different way of thinking about it . . . . . 27. I think about past worries instead . . . . . . . 28. I focus on different negative thoughts . . . . . . 29. I question the reasons for having the thought . . 30. I tell myself that something bad will happen if I think the thought . . . . . . . . . . . . . . . . 31. I keep myself busy . . . . . . . . . . . . 32. I prefer to think things through than distract from them 33. I seek reassurance from others (e.g. my bed partner or a friend on the following day). . . . . . ... 34. I say ‘‘stop’’ to myself . . . . . . . . . . . 35. I do something physical to block them (e.g. turn over, get out of bed) . . . . . . . . . . . . . .
997
Almost never
Sometimes
Often
Almost always
1
2
3
4
1 1
2 2
3 3
4 4
1 1 1 1
2 2 2 2
3 3 3 3
4 4 4 4
1 1 1 1 1 1
2 2 2 2 2 2
3 3 3 3 3 3
4 4 4 4 4 4
1 1 1 1 1 1 1 1 1 1 1 1 1
2 2 2 2 2 2 2 2 2 2 2 2 2
3 3 3 3 3 3 3 3 3 3 3 3 3
4 4 4 4 4 4 4 4 4 4 4 4 4
1 1
2 2
3 3
4 4
1
2
3
4
1 1
2 2
3 3
4 4
ARTICLE IN PRESS 998
M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
References Abramowitz, J. S., Whiteside, S., Kalsy, S., & Tolin, D. F. (2003). Thought control strategies in obsessive-compulsive disorder: A replication and extension. Behaviour Research and Therapy, 41, 529–540. American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. Amir, N., Cashman, L., & Foa, E. B. (1997). Strategies of thought control in obsessive-compulsive disorder. Behaviour Research and Therapy, 35, 775–777. Andrews, L., Troop, N., Joseph, S., Hiskey, S., & Coyne, I. (2002). Attempted versus successful avoidance: Associations with distress, symptoms, and strategies for mental control. Personality and Individual Differences, 33, 897–907. Beck, A. T., Brown, G., Epstein, N., & Steer, R. A. (1988). An inventory for measuring clinical anxiety: Psychometric properties. Journal of Consulting and Clinical Psychology, 56, 893–897. Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for measuring depression. Archives of General Psychiatry, 4, 561–571. Blake, R., Shawe-Taylor, M., & Murray, J. (2003). Cognitive processes involved in the maintenance of insomnia in a population with co-morbid anxiety. Manuscript submitted. Buysse, D. J., Reynolds, C. F., Monk, T. H., Berman, S. R., & Kupfer, D. J. (1989). The Pittsburgh Sleep Quality Index: A new instrument for psychiatric practice and research. Psychiatry Research, 28, 193–213. Carpenter, J. S., & Andrykowski, M. A. (1998). Psychometric evaluation of the Pittsburgh Sleep Quality Index. Journal of Psychosomatic Research, 45, 5–13. Espie, C. A. (1991). The psychological treatment of insomnia. England: Wiley. Harvey, A. G. (2000). Pre-sleep cognitive activity in insomnia: A comparison of sleep-onset insomniacs and good sleepers. British Journal of Clinical Psychology, 39, 275–286. Harvey, A. G. (2001). I can’t sleep, my mind is racing! An investigation of strategies of thought control in insomnia. Behavioural and Cognitive Psychotherapy, 29, 3–12. Harvey, A. G. (2002). A cognitive model of insomnia. Behaviour Research and Therapy, 40, 869–893. Harvey, A. G. (2003). The attempted suppression of pre-sleep cognitive activity in insomnia. Cognitive Therapy and Research, 27, 593–602. Harvey, A. G. (in press). Unwanted intrusive thought in insomnia. In: Clark, D. A. (Ed.), The Nature and Treatment of Unwanted Intrusive Thoughts in Clinical Disorders. Guilford Press. Harvey, A. G., & Bryant, R. A. (1998). The effect of attempted thought suppression. In acute stress disorder. Behaviour Research and Therapy, 36, 583–590. Harvey, A. G., Nelson, J., Neitzert-Semler, C., Tang, N. K. Y., Sharpley, A., & Ree, M. (2003). Validation of the Insomnia Diagnostic Interview (IDI). Manuscript in preparation. Harvey, A. G., & Payne, S. (2002). The management of unwanted pre-sleep thoughts in insomnia: Distraction with imagery versus general distraction. Behaviour Research and Therapy, 36, 757–763. Lichstein, K. L., & Rosenthal, T. L. (1980). Insomniacs’ perceptions of cognitive versus somatic determinants of sleep disturbance. Journal of Abnormal Psychology, 89, 105–107. Rapee, R. M., Craske, M. G., Brown, T. A., & Barlow, D. H. (1996). Measurement of perceived control over anxietyrelated events. Behavior Therapy, 27, 279–293. Reynolds, M., & Wells, A. (1999). The Thought Control Questionnaire: Psychometric properties in a clinical sample, and relationships with PTSD and depression. Psychological Medicine, 29, 1089–1099. Salkovskis, P. M., & Campbell, P. (1994). Thought suppression induces intrusion in naturally occurring negative intrusive thoughts. Behaviour Research and Therapy, 27, 677–682. Wegner, D. M. (1989). White bears and other unwanted thoughts: Suppression, obsession and the psychology of mental control. New York: Viking. Wells, A., & Davies, M. I. (1994). The thought control questionnaire: A measure of individual differences in the control of unwanted thoughts. Behaviour Research and Therapy, 32, 871–878.
Behaviour Research and Therapy 43 (2005) 985–998 www.elsevier.com/locate/brat
Attempts to control unwanted thoughts in the night: development of the thought control questionnaire-insomnia revised (TCQI-R) Melissa J. Reea, Allison G. Harveya,c,, Rachel Blakeb, Nicole K.Y. Tangc, Metka Shawe-Taylord a
b
Department of Psychiatry, University of Oxford, Warneford Hospital, Oxford OX3 7JX, UK Department of Psychology, University of Surrey, Guildford and South London and Maudsley NHS Trust, UK c Department of Experimental Psychology, University of Oxford, UK d Surrey Oaklands NHS Trust, UK Received 14 January 2004; received in revised form 2 July 2004; accepted 7 July 2004
Abstract The attempted control of intrusive, uncontrollable thoughts has been implicated in the maintenance of a range of psychological disorders. The current paper describes the refinement of the Thought Control Questionnaire Insomnia (TCQI; Behav. Cogn. Psychoth. 29 (2001)) through its administration to a sample (n=385) including good sleepers and individuals with insomnia. Several items with poor psychometric properties were discarded, resulting in a 35-item revised TCQI. Factor analysis revealed six factors; aggressive suppression, cognitive distraction, reappraisal, social avoidance, behavioural distraction, and worry. The attempted management of unwanted thoughts was compared across individuals with insomnia and good sleepers, and the impact of these strategies on sleep quality, anxiety and depression was investigated. With the exception of cognitive distraction, individuals with insomnia, relative to good sleepers, more frequently used every thought control strategy. The strategies of aggressive suppression and worry, in particular, appeared to be unhelpful, with the use of these strategies predicting sleep impairment,
Corresponding author. Department of Psychiatry, University of Oxford, Warneford Hospital, Headington, Oxford
OX3 7JX, UK. Tel: +44-1865 223 912; fax: +44-1865-793-101. E-mail address: [email protected] (A.G. Harvey). 0005-7967/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.brat.2004.07.003
ARTICLE IN PRESS 986
M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
anxiety and depression. The strategy of cognitive distraction appeared to be helpful, with the use of this strategy predicting better sleep quality. r 2004 Elsevier Ltd. All rights reserved. Keywords: Insomnia; Suppression; Worry; Distraction; Avoidance; Reappraisal
1. Introduction The attempted control of intrusive, uncontrollable thoughts has been identified as an important feature in a range of psychological disorders including obsessive compulsive disorder, post traumatic stress disorder, social phobia, insomnia, and depression (e.g., Abramowitz, Whiteside, Kalsy, & Tolin, 2003; Harvey, 2002a; Harvey & Bryant, 1998; Reynolds & Wells, 1999). A body of research is emerging to endorse the proposal that some thought control strategies are helpful but that others are unhelpful because they actually perpetuate unwanted thoughts rather than put an end to them (e.g., Abramowitz et al., 2003; Amir, Cashman, & Foa, 1997; Harvey, 2001; Salkovskis & Campbell, 1994). In order to aid the systematic investigation of strategies used in an attempt to control unwanted thoughts, Wells and Davies (1994) developed the Thought Control Questionnaire (TCQ) which asks respondents to indicate the frequency with which they employ each of thirty different thought control strategies. The development of the TCQ led to the identification of five dimensions of thought control; namely, punishment, worry, distraction, social control, and reappraisal. A factor analysis of a preliminary version of the TCQ suggested that it contained six factors (including two distraction factors: cognitive distraction and behavioural distraction), while a subsequent analysis revealed only five factors because the distraction factors collapsed into one (Wells & Davies, 1994). Interestingly, in a subsequent investigation of the psychometric properties of the TCQ in a clinical sample (Reynolds & Wells, 1999), support for the distraction scale being split into behavioural and cognitive dimensions was once again obtained. Research employing the TCQ to date has revealed an acceptably robust factor structure, but whether the distraction scale comprises one or two factors is unclear. Since the development of the TCQ, research has been broadly consistent in suggesting that certain thought control strategies are associated with psychopathology. Strategies of punishment (e.g., ‘I punish myself for having the thought’) and worry (e.g., ‘I focus on different negative thoughts’) have been identified to be associated with more severe symptoms across a number of psychological disorders (e.g., Amir et al., 1997; Andrews, Troop, Joseph, Hiskey, & Coyne, 2002; Reynolds & Wells, 1999). Conversely, the strategy of social control (e.g., ‘I ask my friends if they have similar thoughts’) may be an adaptive strategy that may provide a protective buffer against psychological distress (Blake, Shawe-Taylor, & Murray, 2003; Harvey, 2001; Wells & Davies, 1994). The TCQ, then, has proved to be a useful assessment tool in the investigation of thought control strategies across a range of disorders. It can be concluded that thought control is a multifaceted construct, with some thought control strategies being unhelpful and others being more effective in the management of unwanted thoughts and images. A ‘racing mind’ while lying in bed is as a key symptom of insomnia (e.g., Espie, 1991; Harvey, 2000; Lichstein & Rosenthal, 1980) and hence difficulty with thought control has been implicated
ARTICLE IN PRESS M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
987
in the maintenance of insomnia and potentially its treatment (e.g., Harvey, 2002a, Harvey, in press). Accordingly, Harvey (2001) developed an adaptation of the TCQ specifically for use in insomnia (see method section for a description of this measure). Employing the TCQI the frequency with which thoughts interfered with sleep was associated with the strategies of suppression, punishment, reappraisal, and worry (Harvey, 2001). The data also revealed that individuals with insomnia used the thought control strategies of suppression, reappraisal, and worry more frequently than good-sleeper controls. Social control and replacement strategies, however, were associated with better sleep. While the data produced by the TCQI is promising, its psychometric properties need to be investigated. Accordingly, the primary aim of the current study was to assess the psychometric properties of the TCQI and revise it as necessary in order to enhance its psychometric properties. A secondary aim was to replicate and extend Harvey (2001) results by investigating thought control strategies employed by good sleepers and patients with insomnia with a larger sample of participants than employed by Harvey (2001); n ¼ 60; and to examine the relationship between thought control strategies and measures of insomnia severity, anxiety, and depression.
2. Method 2.1. Participants In total, 385 participants completed the TCQI for this study. One hundred and ninety one individuals met criteria for primary insomnia (Diagnostic and Statistical Manual for Mental Disorders-IV criteria: American Psychiatric Association, 1994), 64 of whom where treatmentseeking and 127 who were not treatment-seeking but responded to recruitment advertisements. Participants seeking treatment for insomnia were recruited primarily (n=58) via advertisements at general practitioner surgeries, while 6 were recruited from an Internet advertisement for a research study offering a psychological treatment for insomnia. Of the treatment-seeking participants, 10 reported occasional use of sleeping tablets (twice per week or less). Good sleepers were 96 university students (n=65) and members of the general public (n=31) who responded to recruitment advertisements. Finally, 98 non-treatment-seeking individuals participated. These were self-reported poor sleepers but did not meet diagnostic criteria. They were regarded as a subthreshold group. Information regarding use of sleeping tablets was not sought from the nontreatment seeking component of the sample. The sample had a mean number of years education of 13.2 (SD=2.14). The mean age of respondents was 27.41 years (SD=11.42) and the range was 18 to 63 years. Females constituted 63% (n=241) of the sample, males constituted 36% (n=139), and 5 respondents did not indicate their sex. 2.2. Measures For individuals seeking treatment for insomnia, all measures were administered in an assessment phase prior to commencement of treatment. For non-treatment seeking participants,
ARTICLE IN PRESS 988
M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
measures were completed either in a laboratory testing session, or in the participant’s home, in which case they posted completed questionnaires to the experimenter. TCQI: The 43-item TCQI (Harvey, 2001) was administered to all participants. Items were rated on a four-point scale where 1=almost never and 4=almost always, yielding a minimum score of 43 and a maximum score of 172. The Thought Control Questionnaire-Insomnia (TCQI) differs from the TCQ in several ways: Firstly, the instructions of the TCQI ask respondents to indicate the frequency (never, sometimes, often, almost always) with which they employ each thought control strategy while being kept awake by thoughts. Secondly, the TCQI contains thirteen additional items relevant to insomnia (e.g., ‘I count sheep’, ‘I decide to put them ‘‘on hold’’ until the morning’). Thirdly, the TCQI contains a suppression and a replacement scale in place of the original distraction scale of the TCQ. The decision to divide the distraction scale in this way was based on evidence suggesting that it is important to differentiate between thought suppression and replacement (Salkovskis & Campbell, 1994). The TCQI suppression scale was comprised of items describing what Salkovskis and Campbell (1994) refer to as ‘simple distraction’. Simple distraction comprises strategies that divert attention away from the unwanted thought without using a specific task to replace it (e.g., ‘I try to push the thoughts out of my head’). The TCQI replacement scale contained items describing strategies that replace the unwanted thought with another thought (e.g., ‘I call to mind positive images instead’). A final difference between the TCQ and TCQI is that the TCQI has an additional question at the beginning of the questionnaire that asks respondents to indicate the frequency with which thoughts keep them awake at night (where 0=never, 10=every night). Insomnia Diagnostic Interview (IDI): In the absence of a psychometrically validated interview corresponding to the DSM-IV criteria for primary insomnia, a structured clinical interview, the Insomnia Diagnostic Interview (IDI: Harvey, Nelson, Neitzert-Semler, Tang, Sharpley, & Ree, 2003), was used to determine the presence of insomnia. This interview schedule has recently been validated against clinician-based diagnoses of insomnia on 55 individuals (Harvey et al., 2003). The results showed good internal consistency (a=0.87), sensitivity (92%), specificity (89%), and test–retest reliability (r=0.90), as well as high diagnostic agreement for the presence (90%) and absence of insomnia (92%). Participants were classified as good sleepers on the basis of their selfreport that they slept ‘very well’ or ‘reasonably well’ on a screening question. These participants were not administered the IDI. Pittsburgh Sleep Quality Index (PSQI; Buysse, Reynolds, Monk, Berman, & Kupfer, 1989): In order that the association between the TCQI and insomnia severity could be investigated, 190 of the participants also completed the PSQI. The PSQI contains 19 items that sum to yield a global score (higher scores indicate more severe impairment) and has sound psychometric properties (Buysse et al., 1989; Carpenter & Andrykowski, 1998). The PSQI has been found to differentiate those with and without insomnia with a 90% sensitivity and 87% specificity. In addition, it’s internal consistency (a=0.83) and test-retest reliability (r=0.85) are acceptable (Buysse et al., 1989). Beck Anxiety Inventory (BAI; Beck, Brown, Epstein, & Steer, 1988): In order that the association between the TCQI and anxiety symptoms could be investigated 134 of the participants completed the BAI. Beck Depression Inventory (BDI; Beck, Ward, Mendelson, Mock, & Erbaugh, 1961): In order that the association between the TCQI and depression could be investigated 184 of the participants completed the BDI.
ARTICLE IN PRESS M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
989
3. Results 3.1. Overview In line with the first aim of the current paper, to investigate the psychometric properties of the TCQI and to make revisions on the basis of these, the first analyses conducted were to select the psychometrically adequate items in the TCQI, and to assess the factor structure of the questionnaire. The second aim of the present paper was to assess the thought control strategies employed by good sleepers and those with insomnia, and to investigate the relationship between thought control strategies and measures of insomnia severity, anxiety, and depression. Correlational analyses were employed to achieve this second aim. Since the stability of the TCQ-I-Revised factor structure has not yet been cross-validated with independent samples, the TCQ-I-Revised total score is also included in the analyses. It is important to note, however, that this total score may not be readily interpretable as some of the scales are proposed to be helpful while others are proposed to be unhelpful. 3.2. Selection of items for factor analysis 1. Items were deleted if they were not easily interpretable in the context of insomnia. Specifically, the items ‘I find out how my friends deal with these thoughts’, ‘I ask my friends if they have similar thoughts’, and ‘I talk to a friend about the thought’ were removed as they are not strategies that could easily be employed while in bed trying to sleep. 2. Items which did not employ the full response range of the scale were removed as these items did not have a high capacity to discriminate between participants. Specifically, the items ‘I slap or pinch myself to stop the thought’ and ‘I think more about the minor problems I have’ were removed as no participant had responded to these items with a ‘4’ (almost always). 3. The percentage of missing values for each item was examined. No item had a missing value rate of above 5% so no items were discarded due to a high rate of missing values. 4. In order to assess for item redundancy, the correlation matrix of the remaining 38 items was inspected. Items which showed a correlation of above 0.60 with other items were examined to determine if this could be explained by highly similar item content (as in Rapee, Craske, Brown, & Barlow, 1996). For example, the item ‘I don’t talk about the thought to anyone’ correlated highly with the items ‘I keep the thought to myself’, (r=0.65) and ‘I avoid discussing the thought’(r=0.60) and so it was considered redundant and discarded. Therefore, 37 items remained for factor analysis. 3.3. Factor analysis Since the TCQI differs quite substantially from the TCQ, and since a factor analysis has not yet been performed on the TCQI, it was considered appropriate to perform an exploratory factor analysis (EFA) rather than a confirmatory factor analysis on the data. A principal components analysis with oblique (Direct Quartimin) rotation and Kaiser Normalisation was performed on the data (37 items) as the factors were expected to be correlated. Missing values were replaced with the mean value for that item. Inspection of the break in slope
ARTICLE IN PRESS 990
M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
on the scree plot indicated that six factors should be retained in the final solution. These six retained factors accounted for 47.03% of the total variance. The six factors were labelled as follows: aggressive suppression, cognitive distraction, reappraisal, social avoidance, behavioural distraction, and worry. Factor loadings are listed in Table 1. As can be seen from Table 1, the factor structure is mostly in line with that of the TCQ (Wells & Davies, 1994) and that proposed by Harvey (2001) for the TCQI. The main difference between the factor structure suggested by Harvey and the factor structure obtained here is that the suppression and the punishment factors were combined into an ‘aggressive suppression’ factor. Also, the existence of a cognitive distraction factor and a behavioural distraction factor was supported (in line with Wells & Davies, 1994). Items 3 and 9 did not load highly on any factor and so were excluded from subsequent analyses. 3.4. Scale reliabilities In order to estimate the internal consistency of the scales, Cronbach alpha coefficients were calculated as follows: whole scale=0.83, aggressive suppression=0.79, cognitive distraction=0.64, reappraisal=0.76, social avoidance=0.69, behavioural distraction=0.66, and worry=0.78. These results were similar to those found by Wells and Davies (1994) and Reynolds and Wells (1999). 3.5. Scale intercorrelations Table 2 presents the Pearson correlations between the TCQI-Revised scales and the total score. The initial question ‘how often does thinking too much keep you awake’ was also included in this analysis. This question is answered on an 11-point scale where 0=Never and 10=Every night. The scale intercorrelations on the TCQI-Revised were weak to moderate. 3.6. TCQI-revised discrimination of good and poor sleepers In order to assess whether the factors of the TCQI-Revised are able to discriminate good sleepers and those with insomnia, mean scale scores were compared across these two groups. Table 3 indicates that all but the cognitive distraction scale discriminated good sleepers and those with insomnia. The ability of the TCQ-I-Revised total score to discriminate good and poor sleepers was also assessed. The ability of the initial question on the TCQI-Revised (‘how often does thinking too much keep you awake?’) to discriminate between good sleepers and those with insomnia was also assessed. The mean for the good sleeper sample was 4.81 (SD=2.73) while the mean for the insomnia sample was 7.26 (SD=1.94). This difference was significant at the 0.01 level [t(259)=8.49]. 3.7. Correlations between TCQI-revised and the PSQI, BAI, and BDI Correlations were inspected in order to investigate the association between thought control strategies and self-report psychopathology measures. Table 4 displays correlations between the
ARTICLE IN PRESS M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
991
Table 1 Items and factor loadings of the TCQI-revised items Scale/Item
Factor loadings 1
Aggressive suppression (20) I get angry at myself for having the thought (1) I tell myself not to think about them now (11) I tell myself not to be so stupid (2) I try to push the thoughts out of my head (42) I say ‘stop’ to myself (22) I shout at myself for having the thought (15) I punish myself for having the thought (37) I tell myself that something bad will happen if I think the thought
Behavioural distraction/suppression (30) I occupy myself with work instead (39) I keep myself busy (28) I do something I enjoy (6) I try to block them out by reading, watching TV, or listening to the radio (43) I do something physical Social avoidance (17) I keep the thought to myself (21) I avoid discussing the thought (41) I seek reassurance form others (e.g., my bed partner or a friend the next day) Worry (35) I focus on different negative thoughts (27) I worry about more minor things (13) I replace the thought with a more trivial bad thought (16) I dwell on other worries (33) I think about past worries instead
3
4
5
6
0.71 0.68 0.67 0.64 0.63 0.50 0.45 0.36
Cognitive distraction/suppression (25) I think pleasant thoughts instead (18) I think about something else instead (4) I call to mind positive images instead (8) I decide to put them ‘on hold’ until the morning (10) I let my mind go blank Reappraisal (23) I analyse the thought rationally (32) I try a different way of thinking about it (40) I prefer to think things through rather than distract from them (29) I try to reinterpret the thought (19) I challenge the thought’s validity (36) I question the reasons for having the thought (5) If the thoughts relate to a problem I make a decision about it in order to solve the problem
2
0.71 0.69 0.61 0.53 0.42 0.76 0.68 0.67
0.37
0.64 0.54 0.50 0.54
0.76 0.75 0.53 0.48 0.38 0.82 0.78 0.66
0.78 0.73 0.68 0.66 0.60
ARTICLE IN PRESS 992
M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
Table 1 (continued ) Scale/Item
Factor loadings 1
(12) I focus on the thought (7) I ruminate about them (3) I count sheep (9) I get out of bed and write about them
2
3
4
5
6 0.42 0.35
0.35
Note: Loadings below .35 not shown. Table 2 Scale intercorrelations N=385
TCQ-IRevised total
TCQ-I-Revised total Aggressive suppression Cognitive distraction Reappraisal Social avoidance Behavioural distraction Worry How often does thinking too much keep you awake?
1 0.77**
1
0.34** 0.67** 0.35** 0.62** 0.65** 0.35**
Aggressive suppression
Cognitive distraction
Reappraisal Social avoidance
Behavioural Worry distraction
0.22** 0.29** 0.24** 0.33**
1 0.13** 0.04 0.019**
1 0.07 0.29**
1 0.07
1
0.46** 0.33**
0.09 0.12*
0.32** 0.17**
0.18** 0.19**
0.30** 0.14**
1 0.45**
Note: * Significant at the 0.05 level (2-tailed); ** Significant at the 0.01 level (2-tailed).
Table 3 Comparison of TCQI-revised scale scores between good sleepers and individuals with insomnia Scale
Good sleeper mean (SD) n=96
Insomnia mean (SD) n=191
t(df)
Aggressive suppression Cognitive distraction Reappraisal Social avoidance Behavioural distraction Worry TCQIR total score
1.74 2.19 2.10 2.34 1.79 1.57 71.70
2.01** 2.11 2.33** 2.52* 2.05** 1.93** 80.08
4.23 1.31 3.36 2.11 4.25 5.99 5.64
Note: ** significant at the 0.01 level.
(285) (285) (285) (285) (285) (285) (285)
ARTICLE IN PRESS M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
993
Table 4 Pearson correlations between scales of the TCQI-revised and measures of sleep quality, anxiety and depression
PSQIa (n=192) BDI (n=195) BAI (n=136)
Aggressive suppression
Cognitive distraction
Reappraisal
Social avoidance
Behavioural distraction
Worry
TCQIR total score
0.32**
0.14*
0.12
0.22**
0.21**
0.46**
0.34**
0.33**
0.01
0.11
0.22**
0.25**
0.37**
0.36**
0.40**
0.11
0.24**
0.19**
0.22**
0.32**
0.43**
Note: * Significant at the 0.05 level (2-tailed); ** Significant at the 0.01 level (2-tailed). a The mean PSQI score for participants with insomnia was 10.03 (SD=2.93) and the mean for good sleepers was 6.58 (SD=2.78).
TCQI-Revised and the PSQI, BDI and BAI. Higher aggressive suppression, social avoidance, behavioural distraction, and worry scale scores were associated with worse sleep quality, and higher depression and anxiety scores. The reappraisal scale was only associated with anxiety, and the cognitive distraction scale was associated negatively with sleep quality. Pearson correlations between the initial question on the TCQI-Revised and measures of sleep quality, anxiety, and depression were also calculated, and are as follows: PSQI=0.72, po0.01; BAI=0.24, po0.01; BDI=0.34, po0.01. 3.8. Multiple regression analyses In order to investigate which thought control strategies predict poor sleep quality, BAI and BDI scores, three multiple regression analyses were performed. The six TCQI-Revised scales were entered as predictors, with all scales being entered simultaneously. The PSQI, BDI, and BAI scores were entered as dependent variables. The results of these analyses are presented in Table 5. Together, the scales of the TCQI-Revised predicted 29% of the variation in PSQI scores, 22% of the variation in BDI scores, and 21% of the variation in BAI scores. PSQI scores were predicted by more frequent use of aggressive suppression and worry, and by less frequent use of cognitive distraction. Depression and anxiety scores were also predicted by the aggressive suppression and worry scales (although the worry scale was only a marginally significant predictor of depression). Finally, the behavioural distraction subscale was a significant predictor of depression scores.
4. Discussion The primary aim of the present study was to revise the TCQI to assess and, if necessary, enhance its psychometric properties. With respect to this aim, TCQI items with poor psychometric properties were removed and the psychometric properties of the 35-item TCQI-Revised (TCQI-R) were examined. The TCQI-R yielded six factors (aggressive suppression, cognitive distraction,
ARTICLE IN PRESS 994
M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
Table 5 Summary of multiple regression analyses for TCQI-R factors predicting scores on measures of sleep quality, anxiety and depression Variable
Beta
T
p
Dependent variable: PSQI (n=192) Aggressive suppression Cognitive distraction Reappraisal Social avoidance Behavioural distraction Worry
0.20 0.18 0.06 0.16 0.10 0.35
2.74 2.69 0.91 1.79 1.36 4.79
0.007 0.008 0.365 0.074 0.172 0.000
Dependent variable: BDI (n=195) Aggressive suppression Cognitive distraction Reappraisal Social avoidance Behavioural distraction Worry
0.19 0.08 0.08 0.10 0.16 0.27
2.44 1.20 1.04 1.53 2.22 3.74
0.016 0.232 0.302 0.127 0.027 0.000
Dependent variable: BAI (n=136) Aggressive suppression Cognitive distraction Reappraisal Social avoidance Behavioural distraction Worry
0.27 0.04 0.10 0.05 0.05 0.17
2.93 0.46 1.11 0.61 0.60 1.91
0.004 0.641 0.271 0.545 0.547 0.058
reappraisal, social avoidance, behavioural distraction, worry) with a structure largely consistent with previous research on the TCQ (Wells & Davies, 1994). The internal consistency of the questionnaire and its scales were acceptable and also in line with previous research (Wells & Davies, 1994; Reynolds & Wells, 1999). The factor structure of the TCQI-R did, however, differ from that proposed for the TCQI in two ways. Firstly, an ‘aggressive suppression’ factor emerged that consisted of items from the punishment and suppression scales of the TCQI. Participants employing this strategy appeared to be using suppression or self-critical strategies in an attempt to control their thoughts. Secondly, behavioural and cognitive aspects of distraction were extracted as separate factors. This is consistent with Wells and Davies’ (1994) original proposal that cognitive and behavioural distraction are distinct thought control strategies. We note that this distinction has not consistently been found in past research. It is possible that behavioural and cognitive distraction were more distinct in the context of insomnia than they are for other psychological disorders. This suggestion is supported by the finding that cognitive distraction predicts better sleep quality and that behavioural distraction predicts worse sleep quality. Perhaps the deployment of cognitive distraction strategies is helpful because these strategies are compatible with sleep (e.g., ‘I think pleasant thoughts instead’), whereas the behavioural distraction strategies are less compatible with sleep (e.g., ‘I occupy myself with work instead’). In addition to displaying a readily interpretable factor structure, the TCQI-R discriminated between people with insomnia and good sleeper controls on five of its six scales (all scales other than
ARTICLE IN PRESS M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
995
cognitive distraction). This is consistent with previous research employing the TCQI (Harvey, 2001). Together, the psychometric properties of the individual items, the internal consistency of the questionnaire, its factor structure, and the ability of the TCQI-R to discriminate good sleepers and individuals with insomnia suggests that this is measure holds promise for future research investigating thought control strategies in insomnia. Turning now to the second aim of the paper, which was to examine the relationships between the TCQI-R and measures of insomnia severity, anxiety, and depression. Firstly, there was a positive correlation between severity of insomnia (measured by the PSQI) and self-reported frequency with which thoughts interfered with sleep (as assessed by the first question of the TCQIR). This finding is consistent with previous research suggesting that unwanted nocturnal cognitive activity is important in the maintenance of insomnia (e.g., Espie, 1991; Harvey, 2000; Lichstein & Rosenthal, 1980). Self-reported frequency of thoughts interfering with sleep was associated with the aggressive suppression and worry scales of the TCQI-R. Secondly, the relationship between the TCQI-R scales and measures of anxiety, depression, and severity of insomnia was examined. The more frequent use of aggressive suppression, behavioural distraction, social avoidance, and worry were each associated with higher anxiety, higher depression, and poorer sleep quality. The regression analyses also echoed the emerging pattern of aggressive suppression and worry being unhelpful thought control strategies, with poor sleep quality, anxiety, and depression being predicted by both of these. The suggestion that worry is an unhelpful thought control strategy is consistent with previous research using the TCQ (Wells & Davies, 1994) and investigating insomnia (Harvey, 2002; in press). The aggressive suppression scale of the TCQI-R contains items relating to punishment (e.g., ‘I get angry at myself for having the thought’) in addition to items relating to suppression (e.g., ‘I try to push the thoughts out of my head’). The finding that aggressive suppression predicted poor sleep, anxiety, and depression is consistent with previous experimental studies reporting adverse consequences following the deployment of thought suppression (Harvey, 2003; Wegner, 1989) and punishment (e.g., Reynolds & Wells, 1999). In addition to findings suggesting that some thought control strategies are unhelpful, the present study provided evidence to suggest that some thought control strategies may be helpful. The use of cognitive distraction as a thought control strategy predicted better sleep quality. This finding is consistent with previous experimental manipulations (Harvey & Payne, 2002). These findings need to be interpreted in the light of several limitations. First, participants were classified as good sleepers on the basis of their self-report of sleeping well or reasonably well. Future research should employ a more stringent method for classifying people as good sleepers including setting criteria for sleep onset latency, wake after sleep onset, total sleep time, and daytime fatigue and functioning. Second, the present study did not collect test-retest data for the TCQI-R. This will be an important direction in order to fully understand the psychometric properties of this questionnaire. Another avenue for future investigations is to further assess the inter-scale reliabilities of the TCQI-R. As with the TCQ, the scales with fewer items have more modest reliabilities, and the stability of these will need to be assessed with independent samples. In summary, the present study has refined the TCQI by eliminating items with poor psychometric properties. The resulting 35-item questionnaire, the TCQI-R, demonstrated a clear factor structure, good internal consistency, and an ability to discriminate between good sleepers and individuals with insomnia. This investigation has also suggested that individuals with
ARTICLE IN PRESS 996
M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
insomnia use worry, aggressive suppression, reappraisal, and behavioural distraction more than good sleepers. Worry and aggressive suppression were noted as unhelpful strategies that were associated with being kept awake by thoughts, and were predictors of poor sleep, depression, and anxiety scores. Cognitive distraction, however, may be a helpful thought control strategy. Future research employing the TCQI-R and thought control manipulation experiments will help to delineate the causal pathways between insomnia and strategies of thought control, and may constitute a process measure in treatment outcome research. Acknowledgment The authors would like to gratefully acknowledge Allison Bugg for her assistance in data collection. This work was supported by the Wellcome Trust (Grant no. 065913).
Appendix Thought control questionnaire insomnia-revised Many people find that as they are trying to get to sleep at night thoughts relating to the day they have just had or thoughts about tomorrow come to mind. Other times thoughts relating to ongoing problems or stressors at work or at home come to mind. Sometimes these thoughts make it hard to fall asleep. How often does thinking too much keep you awake? 0 1 2 3 4 5 6 Never
7
8
9 10 Every night
Below are a number of things that people do to control these thoughts. Please read each statement carefully and indicate how often you use each technique to control the thoughts that run through your mind as you are trying to get to sleep at night by circling the appropriate number. There are no right or wrong answers. Do not spend too much time thinking about each one. Almost never
Sometimes
When thoughts running through my mind keep me awake at night . . . 1. I tell myself not to think about them now . . . . . 1 2 2. I try to push the thoughts out of my head. . . . . 1 2 3. I call to mind positive images instead . . . . . . 1 2 4. If the thoughts relate to a problem I make a 1 2 decision about it in order to solve the problem . .....
Often
Almost always
3 3 3 3
4 4 4 4
ARTICLE IN PRESS M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
5. I try to block them out by reading a book, watching TV or listening to the radio . . . . . . . . ... 6. I ruminate about them . . . . . . . . . . . 7. I decide to put them ‘‘on hold’’ until the morning. . . 8. I let my mind go blank. . . . . . . . . . . 9. I tell myself not to be so stupid . . . . . . . . 10. I focus on the thought . . . . . . . . . . . 11. I replace the thought with a more trivial bad thought . 12. I punish myself for thinking the thought . . . . . 13. I dwell on other worries . . . . . . . . . . 14. I keep the thought to myself . . . . . . . . . 15. I think about something else instead . . . . . . . 16. I challenge the thoughts validity . . . . . . . . 17. I get angry at myself for having the thought . . . . 18. I avoid discussing the thought . . . . . . . . . 19. I shout at myself for having the thought. . . . . 20. I analyse the thought rationally . . . . . . . . 21. I think pleasant thoughts instead. . . . . . . . 22. I worry about more minor things instead. . . . . 23. I do something that I enjoy . . . . . . . . . 24. I try to reinterpret the thought . . . . . . . . 25. I occupy myself with work instead. . . . . . . 26. I try a different way of thinking about it . . . . . 27. I think about past worries instead . . . . . . . 28. I focus on different negative thoughts . . . . . . 29. I question the reasons for having the thought . . 30. I tell myself that something bad will happen if I think the thought . . . . . . . . . . . . . . . . 31. I keep myself busy . . . . . . . . . . . . 32. I prefer to think things through than distract from them 33. I seek reassurance from others (e.g. my bed partner or a friend on the following day). . . . . . ... 34. I say ‘‘stop’’ to myself . . . . . . . . . . . 35. I do something physical to block them (e.g. turn over, get out of bed) . . . . . . . . . . . . . .
997
Almost never
Sometimes
Often
Almost always
1
2
3
4
1 1
2 2
3 3
4 4
1 1 1 1
2 2 2 2
3 3 3 3
4 4 4 4
1 1 1 1 1 1
2 2 2 2 2 2
3 3 3 3 3 3
4 4 4 4 4 4
1 1 1 1 1 1 1 1 1 1 1 1 1
2 2 2 2 2 2 2 2 2 2 2 2 2
3 3 3 3 3 3 3 3 3 3 3 3 3
4 4 4 4 4 4 4 4 4 4 4 4 4
1 1
2 2
3 3
4 4
1
2
3
4
1 1
2 2
3 3
4 4
ARTICLE IN PRESS 998
M.J. Ree et al. / Behaviour Research and Therapy 43 (2005) 985–998
References Abramowitz, J. S., Whiteside, S., Kalsy, S., & Tolin, D. F. (2003). Thought control strategies in obsessive-compulsive disorder: A replication and extension. Behaviour Research and Therapy, 41, 529–540. American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. Amir, N., Cashman, L., & Foa, E. B. (1997). Strategies of thought control in obsessive-compulsive disorder. Behaviour Research and Therapy, 35, 775–777. Andrews, L., Troop, N., Joseph, S., Hiskey, S., & Coyne, I. (2002). Attempted versus successful avoidance: Associations with distress, symptoms, and strategies for mental control. Personality and Individual Differences, 33, 897–907. Beck, A. T., Brown, G., Epstein, N., & Steer, R. A. (1988). An inventory for measuring clinical anxiety: Psychometric properties. Journal of Consulting and Clinical Psychology, 56, 893–897. Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for measuring depression. Archives of General Psychiatry, 4, 561–571. Blake, R., Shawe-Taylor, M., & Murray, J. (2003). Cognitive processes involved in the maintenance of insomnia in a population with co-morbid anxiety. Manuscript submitted. Buysse, D. J., Reynolds, C. F., Monk, T. H., Berman, S. R., & Kupfer, D. J. (1989). The Pittsburgh Sleep Quality Index: A new instrument for psychiatric practice and research. Psychiatry Research, 28, 193–213. Carpenter, J. S., & Andrykowski, M. A. (1998). Psychometric evaluation of the Pittsburgh Sleep Quality Index. Journal of Psychosomatic Research, 45, 5–13. Espie, C. A. (1991). The psychological treatment of insomnia. England: Wiley. Harvey, A. G. (2000). Pre-sleep cognitive activity in insomnia: A comparison of sleep-onset insomniacs and good sleepers. British Journal of Clinical Psychology, 39, 275–286. Harvey, A. G. (2001). I can’t sleep, my mind is racing! An investigation of strategies of thought control in insomnia. Behavioural and Cognitive Psychotherapy, 29, 3–12. Harvey, A. G. (2002). A cognitive model of insomnia. Behaviour Research and Therapy, 40, 869–893. Harvey, A. G. (2003). The attempted suppression of pre-sleep cognitive activity in insomnia. Cognitive Therapy and Research, 27, 593–602. Harvey, A. G. (in press). Unwanted intrusive thought in insomnia. In: Clark, D. A. (Ed.), The Nature and Treatment of Unwanted Intrusive Thoughts in Clinical Disorders. Guilford Press. Harvey, A. G., & Bryant, R. A. (1998). The effect of attempted thought suppression. In acute stress disorder. Behaviour Research and Therapy, 36, 583–590. Harvey, A. G., Nelson, J., Neitzert-Semler, C., Tang, N. K. Y., Sharpley, A., & Ree, M. (2003). Validation of the Insomnia Diagnostic Interview (IDI). Manuscript in preparation. Harvey, A. G., & Payne, S. (2002). The management of unwanted pre-sleep thoughts in insomnia: Distraction with imagery versus general distraction. Behaviour Research and Therapy, 36, 757–763. Lichstein, K. L., & Rosenthal, T. L. (1980). Insomniacs’ perceptions of cognitive versus somatic determinants of sleep disturbance. Journal of Abnormal Psychology, 89, 105–107. Rapee, R. M., Craske, M. G., Brown, T. A., & Barlow, D. H. (1996). Measurement of perceived control over anxietyrelated events. Behavior Therapy, 27, 279–293. Reynolds, M., & Wells, A. (1999). The Thought Control Questionnaire: Psychometric properties in a clinical sample, and relationships with PTSD and depression. Psychological Medicine, 29, 1089–1099. Salkovskis, P. M., & Campbell, P. (1994). Thought suppression induces intrusion in naturally occurring negative intrusive thoughts. Behaviour Research and Therapy, 27, 677–682. Wegner, D. M. (1989). White bears and other unwanted thoughts: Suppression, obsession and the psychology of mental control. New York: Viking. Wells, A., & Davies, M. I. (1994). The thought control questionnaire: A measure of individual differences in the control of unwanted thoughts. Behaviour Research and Therapy, 32, 871–878.