INTRODUCTION
Chronic pain encompasses a broad spectrum of significant discomfort that persists for more than 3 months and often originates from an injury, illness, or unidentified cause. It significantly impacts an individual’s quality of life, potentially leading to physical limitations, emotional distress, and social isolation [1]. Chronic pain is distinct from acute pain; it does not serve a protective nociceptive function and should not be considered merely a prolonged version of acute pain. Instead, it is sustained by factors that are physiologically and mechanistically distant from the original cause, such as altered pain modulation, glial activation, central sensitization, and neuroimmune signaling [2]. Chronic pain not only contributes to insomnia but is also exacerbated by it, creating a bidirectional relationship where pain can interrupt sleep and disrupted sleep can lower the pain threshold. Epidemiological research has shown that insufficient sleep duration and poor sleep quality are risk factors for the development of chronic pain. Moreover, there is robust evidence suggesting that limited sleep or sleep disturbances may exacerbate or trigger spontaneous pain symptoms, such as headaches and muscle soreness, and can also lead to hyperalgesia, an increased sensitivity to painful stimuli [3].
Insomnia is highly prevalent, particularly among individuals with psychiatric conditions such as depression, dysthymia, mania, and others. The role of dysfunctional sleep beliefs in perpetuating insomnia has been established by numerous studies. These studies indicate that individuals with insomnia have less realistic expectations about necessary sleep compared to those who sleep well, strongly agree with assertions about the harmful effects of insomnia, and are more likely to attribute their sleeplessness to stable, external factors. Moreover, cognitive models suggest that individuals with insomnia often excessively worry about their sleep and its daytime consequences [4].
Individuals experiencing persistent pain often harbor dysfunctional sleep beliefs that can exacerbate both their pain and sleep issues. Such beliefs may worsen their condition by promoting maladaptive behaviors and increasing anxiety before bedtime. Patients with chronic pain frequently endorse problematic sleep notions; for example, they catastrophize sleep loss—believing that insufficient sleep will severely affect their pain levels— and maintain unrealistic sleep expectations, believing that adhering to specific sleep hours is necessary to manage their pain effectively. Addressing these dysfunctional beliefs is crucial for improving outcomes related to both sleep and pain [5].
The Pain and Sleep Questionnaire (PSQ) is a diagnostic tool employed to assess the relationship between a patient’s pain and their sleep patterns. This tool typically requires individuals to rate their pain levels, describe their pain’s characteristics, and explain how pain impacts their ability to fall asleep, remain asleep, and the overall quality of their rest. By collecting this data, healthcare providers can gain a deeper understanding of how pain influences sleep and vice versa, thus enabling them to devise more specific and effective treatment strategies. This questionnaire is especially valuable in the management of chronic pain conditions, where disrupted sleep frequently serves as a complicating factor [6].
Two widely used versions are the PSQ-5 [7] and PSQ-3 [8] scales. Both are abbreviated versions of the original PSQ [7], developed for rapid assessment of pain-related sleep quality while maintaining reliability. This study focused on evaluating the reliability and validity of the Korean versions of the PSQ-5 and PSQ-3 among patients experiencing chronic pain.
METHODS
Participants and Procedure
We conducted an anonymous online survey among patients who visited the Pain Clinic at Asan Medical Center, Seoul, Korea, from August 14th to November 20th, 2023. Inclusion criteria included patients aged 18 to 79 years experiencing chronic pain lasting over three months. We distributed advertisements detailing the study’s objectives, the survey instructions, and the rewards for participation at the Pain Clinic. Participants could engage in the survey by selecting “agree” to the participation consent question. We requested participants to report the duration of pain they experienced and excluded those whose pain lasted less than three months from the statistical analysis. We collected participants’ demographic information, including age, sex, marital status, past psychiatric history, and current insomnia symptoms. Additionally, participants were asked to identify their type of pain according to the International Classification of Diseases, 11th edition (ICD-11); options included chronic primary pain, chronic cancer-related pain, chronic surgical or post-traumatic pain, chronic secondary musculoskeletal pain, chronic secondary visceral pain, chronic neuropathic pain, chronic secondary headache or orofacial pain, and unknown. The study protocol was approved by the Institutional Review Board (IRB) of the Asan Medical Center (2023-0753).
Measures
PSQ-3 and PSQ-5
The PSQ-3 and PSQ-5 are rating scales developed to assess the impact of pain on an individual’s sleep. The PSQ-5 incorporates the first 5 items from the original scale [7]. The previous study clustered 5 items of the PSQ-5 into two factors: factor I (item 2 and 3, assessing the need for medication) and factor II (items 1, 4, and 5, evaluating sleep problems due to pain). Factor II of the PSQ-5 was referred to as the PSQ-3 [8]. Initially, items were scored on a 100-mm Visual Analog Scale (VAS) from 0 (never) to 100 (always), but we modified this scale to a range from 0 (never) to 10 (always). The Korean version of the PSQ-3 and PSQ-5 was created through translation and back-translation processes. In this study, we implemented the Korean version of the PSQ-3 and PSQ-5.
Pain-related Beliefs and Attitudes about Sleep (PBAS) Scale
The Pain-related Beliefs and Attitudes about Sleep (PBAS) is a self-report rating scale developed to assess one’s pain-related dysfunctional beliefs about sleep [9]. It consists of 10 items rated from 0 (strongly disagree) to 10 (strongly agree), whose average score indicates the level of pain-related dysfunctional beliefs about sleep. The Korean version of the PBAS, developed and validated through translation and back-translation methods, yielded a Cronbach’s alpha of 0.947 in this sample.
Dysfunctional Beliefs and Attitudes about Sleep-6 items (DBAS-6)
The Dysfunctional Beliefs and Attitudes about Sleep-6 items (DBAS-6), a condensed version [10] of the DBAS-16, assesses one’s dysfunctional beliefs about sleep [11]. It includes 6 items, rated between 0 (strongly disagree) and 10 (strongly agree), with an average score reflecting the level of such beliefs. The Korean version of this scale [10] was applied, yielding a Cronbach’s alpha of 0.772 in this sample.
Dysfunction Beliefs about Sleep-2 items (DBS-2)
The Dysfunction Beliefs about Sleep-2 items (DBS-2), an ultra-brief rating scale, was developed to assess one’s dysfunctional beliefs about sleep [12]. It includes 2 items rated from 0 (strongly disagree) to 10 (strongly agree), and the averaged score reflects the level of such beliefs. The original Korean version was used, with a Cronbach’s alpha of 0.881.
Pain severity (numeric rating scale)
Pain severity was assessed by asking patients, “How would you rate your pain on a scale of 0 to 10, with 0 being no pain at all and 10 being the worst pain imaginable?” Responses ranged from 0 (no pain) to 10 (the worst pain imaginable) [15].
Discrepancy between desired time in bed and desired total sleep time (DBST) index
The discrepancy between desired time in bed and desired total sleep time (DBST) index quantifies the discrepancy between one’s desired time in bed (TIB) and one’s desired total sleep time (TST) [16]. This index is calculated by subtracting the desired TST from the desired TIB. Desired TIB is assessed based on responses to the question, “How many hours do you want to sleep per day?,” whereas desired TST is determined from answers to, “From what time to what time do you want to sleep?”
Statistical Analysis
In this study, demographic variables and rating scale scores were presented as mean±standard deviation. Clinical significance was set at p=0.05 for two-tailed tests. Statistical analyses were conducted using JASP v0.14.1 (JASP team, Amsterdam, The Netherlands).
Initially, the structural validity of the Korean version of the PSQ-5 and PSQ-3 was evaluated. The normality of items was confirmed based on skewness and kurtosis values within ±2 [17]. The Kaiser-Meyer-Olkin (KMO) measure and Bartlett’s test of sphericity were employed to check for sampling adequacy and data suitability. Exploratory factor analysis (EFA) was conducted to determine if the PSQ-5 supports a two-factor structure. Confirmatory factor analysis for PSQ-5 and PSQ-3 was performed to validate model fit using the Comparative Fit Index (CFI), Tucker-Lewis index (TLI), Root-Mean-Square-Error of Approximation (RMSEA), and Standardized Root Mean Square Residual (SRMR) [18,19]. The internal consistency reliability of PSQ-5 and PSQ-3 was assessed using Cronbach’s alphas.
Second, the convergent validities of the PSQ-5 and PSQ-3 were evaluated using Pearson’s correlation analysis with established rating scales such as PBAS, DBAS-6, DBS-2, ISI, DBST index, and pain severity. Linear regression analysis with enter methods was conducted to identify predictive variables for the PSQ-5 and PSQ-3. Mediation analysis was performed to determine whether sleep-related cognition rating scales (PBAS, DBAS-6, and DBS-2) could mediate the impact of pain severity on the PSQ-5 and PSQ-3.
RESULTS
We collected 100 samples from patients with chronic pain. The mean age and duration of pain were 44.9±12.8 years and 50.4±66.9 months, respectively (Table 1).
Reliability and Validity of the Korean Version of the PSQ-3 and PSQ-5
We verified the normality assumption for all items of the PSQ-5 and PSQ-3, which showed skewness and kurtosis values within ±2 (Table 2). Prior to factor analysis, sampling adequacy and data suitability were confirmed, indicated by a KMO value of 0.841 and significance in Bartlett’s test of sphericity (p<0.001). In the EFA, a single-factor model was proposed for the PSQ-5. Consequently, we performed CFA to assess whether the single-factor models of the PSQ-5 and PSQ-3 demonstrated good fit. The CFA results indicated excellent model fits for the Korean versions of the PSQ-5 (CFI=0.966, TLI=0.933, RMSEA=0.162, SRMR=0.003) and PSQ-3 (CFI=1.000, TLI=1.000, RMSEA=0.000, SRMR<0.001). Both the PSQ-5 and PSQ-3 exhibited high internal consistency, as reflected by Cronbach’s alphas of 0.904 and 0.903, respectively.
Evidence Based on Relationships with Other Variables
The PSQ-5 and PSQ-3 demonstrated strong convergent validity with other pre-established rating scales (Table 3). The PSQ-5 was significantly correlated with pain severity (r=0.58, p<0.001), PSQ-3 (r=0.96, p<0.001), PBAS (r=0.86, p<0.001), DBAS-6 (r=0.50, p<0.001), DBS-2 (r=0.42, p<0.001), and ISI (r=0.67, p<0.001). The PSQ-3 also displayed strong convergent validity with pain severity (r=0.61, p<0.001), PBAS (r=0.86, p<0.001), DBAS-6 (r=0.42, p<0.001), DBS-2 (r=0.43, p<0.001), and ISI (r=0.62, p<0.001).
Linear regression analysis was conducted to determine the extent to which the PSQ-5 and PSQ-3 predicted pain severity and pain-related dysfunctional beliefs about sleep. Since the PSQ scales are directly associated with insomnia severity, we excluded the ISI score during the analysis. The results indicated that the PSQ-5 was predicted by pain severity (β=0.14, p=0.044) and PBAS (β=0.74, p<0.001) (Table 4). Similarly, the PSQ-3 was predicted by pain severity (β=0.15, p=0.023) and PBAS (β=0.78, p<0.001). In the mediation analysis, pain severity directly affected the PSQ-5 and PBAS solely mediated this association (Table 5 and Fig. 1A). Similarly, for the PSQ-3, pain severity directly impacted PSQ-3, with PBAS alone mediating this association (Table 6 and Fig. 1B).
DISCUSSION
In this study, we observed that the single-factor models of the PSQ-5 and PSQ-3 exhibited good fit. The PSQ-5 and PSQ-3 also demonstrated reliable internal consistency among patients with chronic pain. Both scales showed strong convergent validity with pain severity, PBAS, DBAS-6, DBS-2, and ISI. The PSQ-5 and PSQ-3 were predicted by pain severity and pain-related dysfunctional beliefs about sleep. The mediation analysis showed that pain severity directly impacted the PSQ-5 and PSQ-3, with PBAS exclusively mediating the association.
In this study, the PSQ-5 was proposed as a single-factor model based on parallel analysis, diverging from previous findings. In prior research, the PSQ-5 was categorized into two factors; factor 1 (item 2 and 3) assessing the need for medication and factor II (item 1, 4, and 5) focusing on sleep disturbance due to pain [8]. However, no similar results were found in this study. This discrepancy may be due to the inclusion of participants from a pain clinic in a tertiary-level hospital where pain medications are prescribed, potentially leading them to respond similarly to the impact of medication on sleep as they did to other items regarding the influence of pain on their sleep disturbances.
Both the PSQ-5 and PSQ-3 were associated with pain severity, PBAS, DBAS-6, DBS-2, and ISI. The significant association of PSQ scales with pain severity or insomnia severity indicates that these scales can measure an individual’s sleep states concerning pain severity. Both scales correlated with dysfunctional beliefs about sleep (DBAS-6 and DBS-2), including those related to pain dysfunctional beliefs about sleep (PBAS). Both scales can be usefully applied to patients with chronic pain who also experience sleep disturbances. Furthermore, mediation analysis showed that pain severity directly influenced the PSQ-5 and PSQ-3, and the PBAS, but not the DBAS-6 and BDS-2, mediated the association. It demonstrates that although pain severity directly affects the PSQ scores, the relationship is primarily mediated by the PBAS, highlighting the significance of pain-related beliefs in the sleep-pain dynamic. In clinical settings, utilizing PSQ scales can enhance the comprehensive assessment of a patient’s condition, enabling targeted interventions that concurrently address both pain and sleep disturbances.
In this study, we did not observe a significant association between the DBST index and the PSQ-5/PSQ-3 scales. Previous studies indicated that the DBST index was significantly associated with insomnia severity [16,20,21]. Therefore, we expected that the PSQ-5 and PSQ-3 might be significantly associated with the DBST index, as these are scales designed to measure sleep disturbance. However, we found that the ISI was significantly associated with the DBST index among patients with chronic pain. We speculate that the responses of participants who suffer from sleep disturbance in relation to pain differ from those to questions of usual sleep disturbance. Further research is needed to explore the meaning of this lack of association.
There are several limitations in this study. First, the small sample size of 100 participant responses may limit the robustness of the results. Despite this, 100 samples appear sufficient to validate the small item questionnaires according to the rule of thumb; 10 samples for 1 item [22]. Second, conducting an anonymous online survey might bias the results. The survey was conducted in the pain clinic by posting the enrollment poster on the pain clinic’s board, but we are unsure whether the responses were from patients with chronic pain. To address this, we included a question regarding the duration of pain experienced by participants. Additionally, participants were enrolled at a tertiary-level hospital, potentially involving more severe cases. Third, the PSQ-5 and PSQ-3 were originally scored on the continuous 100-mm VAS; however, we used an 11-point Likert-type scale ranging from 0 to 10 in this study. Previous studies comparing the 100-mm VAS and the 11-point Likert scale for PSQ scales are lacking. Nonetheless, the DBAS-16, a widely used sleep-related cognition scale, was initially scored on a 100-mm VAS but has been validated using a transformation to an 11-point Likert scale [11].
In conclusion, the Korean versions of the PSQ-3 and PSQ-5 are reliable and valid tools for assessing sleep quality in patients with chronic pain. Both scales demonstrated strong internal consistency, and confirmatory factor analysis supported their single-factor structures. Utilizing the PSQ scales to address both pain and insomnia may be advantageous for understanding patients suffering from chronic pain and sleep disturbances.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
Print 
