Recruitment took place at the Northumbria Centre for Sleep Research via local advertising which asked for volunteers over the age of 18 who had “trouble sleeping” for 3 nights a week or more, for a period of 2 weeks or longer. An information sheet outlining participation requirements and demands was made available for potential participants. Written consent was obtained from all participants and those not meeting inclusion criteria were provided with sleep information and advised to seek support from general practioner regarding ongoing sleep concerns. This project received a favourable ethical opinion from Newcastle University Ethics Committee, who also provided indemnity cover.

Of the 105 people who responded to the advertisement, 44 (41.91%) made no further contact after the information sheet was sent out. During a second round of recruitment for people with AI, 24 people reporting CI were excluded. 38 people were invited to complete further screening, of which 8 were excluded for the following reasons: travel across two time zones in the preceding weeks (n = 1), cancelling the appointment (n = 4) and possible caseness for current or recent psychiatric disorder (n = 3). The latter was assessed via the Hospital Anxiety and Depression Scale11 and clinical interview.

Participants who expressed an interest in the study were invited to a screening appointment and were told that the aim of the study was to investigate different types of worries and thoughts in people with insomnia. Based on initial screening interview with a trained psychologist, participants who met Research Diagnostic Criteria for Primary Insomnia12 were selected for inclusion in this study. This criterion was that individuals reported either; difficulty initiating sleep, maintaining sleep or waking up too early or feeling unrefreshed despite adequate opportunity to sleep. Furthermore participants should report day-time impairment related to poor sleep e.g. daytime sleepiness, fatigue, attention, concentration difficulty. Inclusion was subject to successful screening assessment that there was no current psychiatric or mood disorder or other sleep disorder or be the result of substance abuse. Participants were allocated to the AI group if they were reporting sleep difficulties for 2 weeks to 3 months, and to the CI group if they were reporting sleep difficulties for 6 months or longer. Temporal distinctions for those with chronic insomnia was based on the International Classification of Sleep Disorders which stipulates duration criteria as 6 months or longer for CI. Although AI is considered in a number of theoretical models it has not yet been defined in any nosolgies. The cut-off of two weeks to 3 months was used based on a previously stipulated cut off for transient insomnia which affects most people at one time or another, and which would normally resolve when the stressor recedes.3,13 Once all screening criteria were satisfied participants then completed a battery of four standardized questionnaires. Participants were given a diary to complete once each evening, for seven days and an actigraph watch to wear each night in bed for seven nights. Following the seven-day data collection period participants returned the diary and completed three of the standardized questionnaires again. Participants were provided debrief information, given an opportunity to ask questions and received a sleep information package.

Screening questionnaire to obtain details of sleep problems and screen out for other sleep disorders; further details of this can be found in a separate paper.2

The Hospital Anxiety and Depression Scale (HADS)11 is a validated and widely used 14-item self-report measure for symptom severity and ‘caseness’ of anxiety and depression.

Insomnia Severity Index (ISI)13 is a measure which has been shown to be a reliable and valid method of quantifying perceived insomnia severity across a range of population samples.14

Two types of primary measures were used to collect data; standardized questionnaires and semi-idiosyncratic diaries.

Sleep Associated Monitoring Index (SAMI)15 is a self-report measure of sleep preoccupation. Respondents are asked to rate on a five-point scale (ranging from 1 = Not at all to 5 = All the time) items which include “before or as you go to bed how often do you calculate the number of hours sleep you hope to get?” and “throughout the day how often are you aware of your concentration being affected by your sleep (or lack of)?”

Sleep Preoccupation Scale (SPS)10 was used to measure frequency of daytime sleep related preoccupation based on a 7-point scale (ranging from 0 = never to 7 = all the time). Example items include “I feel anxious about what will happen when I try to sleep tonight” and “I have a lie-in after a bad night’s sleep”.

Worry Domains Questionnaire (WDQ)16 was used as a content-based measure of non-pathological worry across five domains; relationships, lack of confidence, aimless future, work and finances. Respondents were asked to rate on a 5-point scale (ranging from ‘not at all’ to ‘extremely’) how much they worried about each of the items. Example items include “I worry that my money will run out” and “I worry that I feel insecure”. It is a scale that assesses both the negative and positive aspects of worry.

Perceived Stress Scale (PSS)17,18 was used to obtain a measure of the degree to which individuals appraised situations in their life as stressful. This measure was only obtained on Day 1 and was not repeated with the other standardized measures.

A semi-idiosyncratic diary was developed for this study, and comprises a 16 item self-report scale which asked respondents to rate frequency of thoughts over the previous 24 hour period. The scale is made up of four subscales relating to work, interpersonal relationships, sleep and finances, these correspond to the domains in the WDQ16 and daytime thoughts of sleep which the literature posits as relevant.10,19 Respondents are asked to rate on a five-point scale (ranging from ‘Not at all’ to ‘Constantly’). The scale takes approximately 3–4 minutes to complete. The diary did not ask about actual sleep.

Actigraphy (AW4: Cambridge Neurotechnology) was used as an objective measure of sleep. Epoch length was set at one minute. Participants were asked to wear the actiwatch at night time for 7 nights, setting markers on ‘lights out’ and ‘lights on’ in order to set markers for time spent in bed.

An apriori power calculation was completed using previous data10 with the closest approximation of sleeper groups (“average” and “poor”). This was chosen in the absence of a comparable study with AI and CI groups. An apriori, repeated measures within-between interaction compared average sleepers (M = 45.39; SD = 17.23) and poor sleepers (M = 57.23; SD = 22.61) scores on the SPS yielding a medium effect size (ES) of d = 0.59. Based on α = 0.05, and desired power at 0.8 the total sample size needed would be 74; 37 in each group. However, as the main hypothesis is based on an interactional hypothesis postulating a medium ES (in the same range of effect as previously published ES10), a desired power of 0.8 and based on two measures with a medium-large correlation of 0.4; the total sample size required would be 40, 20 per group.

The final sample consisted of 31 participants; ten in the AI group (10 females) and twenty in the CI group (11 males and 9 females). Average age in CI group was 35 years (SD = 10.5, range 24–62 years) with average length of sleep difficulties of 74 months (range: 5 months–20 years) and average reported sleep length of 5 hours (range: 3.5–6.5 hours). Ninety percent of participants were of white ethnic origin with 80% in full-time work. In the AI group the average age of participants was 42 years (SD = 11.6, range: 20–55 years) with average length of sleep difficulties of 2.5 months (ranging from 4 weeks to 3 months) and an average reported sleep length of 5 hours (range: 4–6 hours). 100% of participants were of white ethnic origin and 81% were in full-time work.

Table 1 shows the means, standard deviations, range and maximum possible scores on questionnaire items for participant demographics, depression and anxiety scores, insomnia severity and perceived stress. Independent-samples t-tests revealed the groups did not differ on perception of sleep quantity, anxiety, depression or severity of insomnia reported. There was a small effect size on the HADS-Anxiety (r = 0.15), HADS-Depression (r = 0.14), and ISI (r = 0.11), indicating higher levels of anxiety, depression and insomnia severity in the CI group. The CI group was significantly higher on the PSS with a medium to large effect size,20 t (28) = 2.46, p = 0.2, r = 0.4.

Actigraphy data was extracted using the Actiwatch Activity and Sleep Analysis 7 programme (Cambridge Neurotechnology Version 7.23). Consistent with guidelines where a minimum of 5 nights provides aggregate measures which more reliably characterize participants,21 those who recorded markers (identified time to bed and time out of bed) on 5 nights or more were included in the actigraphy analysis. 4 participants were excluded due to no markers at all (n = 3) and only 3 nights (n = 1). Median number of actigraphy nights recorded was 7 for both CI group (n = 17) and AI group (n = 9). Table 2 shows the means and range of recorded sleep parameters. Independent-samples t-test was conducted to compare AI and CI groups and there were no significant differences. Effect size (r) was trivial for all sleep parameters, suggesting any differences, which may be there but are undetected, are unlikely to be great clinical interest.

Preliminary analyses of the daily diary data were carried out using repeated measures analysis of variance (ANOVA). All of the interactions were non-significant, indicating that any differences did not vary as a function of group: Day × Group (F = 1.46, 0 = 0.22), Diary type × Group (F = 0.16, p = 0.69), Diary type × day (F = 0.538, p = 0.78) and Diary type × Day × Group (F = 0.946, p = 0.46). A small to medium ES (η² = 0.04) on the Diary type × day × group interaction suggests that any differences are unlikely to be of great importance. Therefore the mean scores for Preoccupation and Sleep data across the seven days were collapsed (averaged) in order to run the main analyses.

A mixed-design ANOVA was carried out with Group × measure type × preoccupation type, for comparison raw scores were converted to z-scores for analyses, however raw scores are reported (Table 3). Summary output data for interaction effects is shown in Table 4. The main hypothesis of an interaction between preoccupation type and group was not supported [F(1, 28) = 0.033, p = 0.57]; given the small effect size for this interaction (η² = 0.01) and with 15% power and α = 0.1, the effect is unlikely to be of clinical significance. All other main effects and interactions returned non-significant results (F ≤ 0.059, p ≥ 0.78). Effect sizes were all effectively zero (η² ≤ 0.003).

In summary, there was no interaction of preoccupation × group as hypothesised. That is, relative to each other the groups do not show significantly different types of preoccupation, nor do they report significantly different amounts of preoccupation. There was also no significant main effect of measure type or preoccupation, suggesting that how we measure preoccupation (one-off standardized questionnaires vs. daily diary) may be inconsequential and produced similar results in this context.

Due to unbalanced gender representation (AI = 100% female), the main analyses described above were repeated with only female CI participants, however the patterns of results did not alter. As there was no interaction effect of diary type × day × group [F(3.97, 59.48) = 0.574, p = 0.681] the mean daily diary scores across all seven days were collapsed. There was no difference in the patterns of results by gender, that is, there were no interaction effects of measure type × group (F = 0.000, p = 0.99), preoccupation type × group (F = 0.076, p = 0.79), measure type × preoccupation type (F = 0.000, p = 0.99) or measure type × preoccupation type × group (F = 0.256, p = 0.620).

The objectives of this study were to explore the level and type of preoccupation (sleep vs. life event) of AI relative to people with CI. Findings did not support the hypothesis, suggesting there is no difference between people with AI and CI in level and type of preoccupation and that people with AI are as preoccupied during the day by both sleep and life events, as people with CI. Whilst further research is necessary to determine potential differences between AI and CI other than duration of insomnia, these results raise potential implications in relation to the criteria that distinguish them. Determining the point, if at all, at which a shift in attention and cognition occurs in people with AI, would provide valuable information to extend our understanding of the development of insomnia and inform future guidance of when early intervention may be indicated.