Interplay among pain intensity, sleep disturbance and emotion in patients with non-specific low back pain

Background Low back pain (LBP) is the most common problem worldwide. There are several negative consequences of LBP, such as sleep disorders, work leave, disability, depression, anxiety, and poor quality of life. In this study, we designed to evaluate the prevalence of sleep disturbance in patients with non-specific LBP(NSLBP), and cross-correlation among sleep disorder, anxiety, depression and pain intensity in patients with NSLBP. Aim In this study, we designed to evaluate the prevalence of sleep disturbance in patients with NSLBP, and cross-correlation among sleep disorder, anxiety, depression and pain intensity in patients with NSLBP. Methods A cross-sectional self-assessment questionnaire survey was carried out in an outpatient clinic. Anonymous assessments were used to characterize the presence of NSLBP, PSQI, VAS, SF-36 form, ODI, BAI and BDI. Cross-correlation among the severity of NSLBP and sleep disturbance, anxiety, depression and life quality were evaluated. Results Patients with NSLBP have a higher incidence of sleep disorder, anxiety and depression, and higher ODI scores than healthy people without LPB (P < 0.01). NSLBP patients with sleep disorders have more severe anxiety, depression, an increased VAS score and poor daily living (P < 0.05). NSLBP patients with anxiety have declined sleep quality, poor daily living, decreased work and social skills, and increased LBP severity (P < 0.05). NSLBP patients with depression have declined sleep quality, poor daily living, decreased work and social skills (P < 0.05). Significant associations were found between the severity of NSLBP and sleep disorders, anxiety and ODI scores. Conclusion Psychological and social factors play an important role in the development of NSLBP. NSLBP leads to sleep disorders, which decrease the sleep quality and increase the unpleasant emotions and memories in return; these can exacerbate the severity of LBP, with the cycle repeating to form a vicious circle.


INTRODUCTION
Low back pain (LBP) is defined as mild to severe pain in the area of the lumbar, lumbosacral or sacroiliac regions (Vrbanic, 2011). It is a major global health and socio-economic problem and the most common musculoskeletal disease, with the reported lifetime prevalence ranging from 40% to 85% (Fujii & Matsudaira, 2013), annual prevalence from 25% to 60% (Livshits et al., 2011), and point prevalence 11.9% ± 2.0% (Hoy et al., 2012). The economic burden of LBP is huge, with direct healthcare expenditures of over $90 billion annually in the United States (Borczuk, 2013). A total of 85% of the LBP cases are described as nonspecific (Maatta et al., 2015). The most troubling aspect of LBP is functional disability (Hush et al., 2009), associated with increased health care spending (Ferreira et al., 2010), decreased daily physical function, impaired psychological well-being, and poor quality of life (Sezgin et al., 2014).
Sleep is essential to keep the normal status of emotional, mental and physical health. Sleep disturbance lead to decreased work ability, increased sick leave, and a higher injury rate (Hillman, Murphy & Pezzullo, 2006). Several studies have confirmed that patients with pain have sleep disturbance (Azevedo et al., 2011;Van de Water, Eadie & Hurley, 2011). The prevalence of sleep disturbance in patients with chronic LBP is more than 50% (Bahouq et al., 2013). The pain has been reported to have a bidirectional relationship with sleep; pain hinders sleep while sleep disturbance may decrease pain thresholds and the mental capacity to manage pain (Schuh-Hofer et al., 2013). However, the reason is still unclear on whether poor sleep quality makes an individual vulnerable to LBP, or if sleep disturbance is caused by LBP. Therefore, an effort should be made to detect the association of pain and sleep disturbance in order to improve the management of LBP.
Pain and sleep disturbance can impair patients' life quality, including physical activity, social integration and emotions (Lavigne et al., 2011). Depression has been reported to occur in patients with chronic pain and sleep disturbance (Campbell et al., 2013), which can also influence sleep quality (Spira et al., 2008), but their relationship is poorly understood.
In this study, we designed to evaluate the prevalence of sleep disturbance in patients with NSLBP compared with 112 age-and sex-matched healthy people without pain; then, cross-correlation among NSLBP, sleep disorder, anxiety and depression was analyzed. The results can provide a theoretical basis for the prevention and treatment of NSLBP. The questionnaires in this study are PSQI, SF-36, ODI, BAI and BDI.

MATERIAL AND METHODS
Patients with non-specific LBP were enrolled from January 2014 to January 2016 in an orthopedic clinic of the First Affiliated Hospital of Sun Yat-sen University and Wenming Hospital; no limitation was set in the aspects of gender, age, height, weight, occupation, educational level and birth place. The chief compliance was LBP, which lasts more than one month. All patients signed informed consent forms after obtaining approval from the Ethics Committee of the First Affiliated Hospital of Sun Yat-sen University. Patients had a certain education level and can understand questionnaire survey. All patients completed questionnaires and lumbar MRI were taken to find out organic diseases (Norton et al., 2016). The questionnaire focuses on patient's general information, body weight, education, occupation, exercise, smoking and drinking habit.
The exclusion criteria were: individuals whose reported age was <18 years or >65 years; LBP attributed to spine fracture, spine inflammation, spinal tumor, spinal tuberculosis, disc herniation, spinal stenosis, spondylolisthesis, aneurysm or lithiasis; individuals who have mental disorders, history of cancer or severe chronic physical disorders (e.g., hypertension, diabetes, coronary heart disease, chronic kidney disease, bronchitis, asthma, etc.) (Norton et al., 2016).
For the control group without LBP, the control age-and gender-matched asymptomatic subjects was enrolled from healthy volunteers without LBP, with the exclusion criteria of history of malignant tumors, psychiatric diseases, chronic systemic diseases.
PSQI is the 24-item survey that evaluates sleep quality, including subjective sleep quality, sleep latency, sleep duration, sleep efficiency, sleep disturbance, use of sleep medication and daytime dysfunction (Alsaadi et al., 2014). One of seven gradual statements (0-3 points) per item to be selected to sum up total score (0-21). Participants completed the PSQI with regard to sleep in the past month.
Pain intensity of patients with NSLBP were assessed by VAS score, with 0 = no pain and 10 = extreme pain, which is widely used to measure pain with good validity and reliability (Eichen et al., 2014).
SF-36 is the 36-item generic quality of life questionnaire. It has an eight scale profile of functional health and well being, including physical function, role function physical, body pain, general health perceptions, vitality, social functioning, role function emotional and mental health. The total score is 0 to 100, with 100 indicating the best possible score. The SF-36 has been recommended for comprehensively measuring the health and quality of life in participants with or without LBP (Hirsch et al., 2014).
BAI is a 21-item self report questionnaire designed to evaluate anxiety symptoms (Bryan et al., 2015). One of 21 gradual statements (0-3 points) per item is selected to sum up the total score (0-63), with 0 = no anxiety and 63 = severe anxiety.
BDI is the standard questionnaire for evaluation of cognitive, affective and somatic symptoms of depression (Ellegaard & Pedersen, 2012). BDI was validated in several countries, and composed of 21 items that address the cognitive, affective, behavioral and somatic components of depression.

Statistical analysis
Statistical analysis were conducted using SPSS version 18 and SAS V9.3. The demographic data including age, sex, education, occupation, exercise, smoking, and drinking were evaluated between the two groups. Characteristics of study participants were summarized by means and SDs for continuous variables, by frequencies and percentages for categorical variables. The pain intensity, quality of daily life, dysfunctional and psychosocial factors were compared and analyzed between these two groups. For continuous variable, if the variable was normally distributed and had constant variances, t -test ANOVA would be conducted for comparisons tests; for categorical variable, the difference would be tested with the chi-square test. Sleep disturbance, anxiety, depression and other items of psychosocial and emotional relevance were compared with the severity of LBP. Spearman correlation analysis was done to identify the correlation among the PSQI, VAS, SF-36, ODI, BAI and BDI. Factors found to be significant on univariate analysis were included in a multivariate binary logistic regression analysis.
There was no significant difference in age, sex, height, weight, smoking and drinking habits between the two groups (p > 0.05). NSLBP patients have higher ODI scores and more incidences of anxiety and depression than healthy people (P < 0.01). NSLBP patients were more likely to have sleep disorders compared with healthy people (P < 0.01). Patients were further divided into mild, moderate and severe LBP groups; statistical analysis showed that the level of anxiety increased with the severity of LBP (P < 0.05) ( Table 2). Patients were divided into two groups according to the presence of sleep disorders, anxiety or depressive symptoms. Statistical analysis showed that NSLBP patients with sleep disorders have more severe anxiety, depression, increased VAS score and poor daily living (P < 0.05). NSLBP patients with anxiety have declined sleep quality, poor daily living, decreased work and social skills, increased LBP severity (P < 0.05). NSLBP patients with depression have declined sleep quality, poor daily living, decreased work and social skills (P < 0.05) ( Table 2).
We then analyzed the relationship of smoking, drinking habits and body mass index on anxiety, depression, quality of sleep and severity of LBP. Results confirmed that smoking habits can affect a patient's anxiety and depression mood in NSLBP patients, resulting in the increased incidence of anxiety and depression. Drinking habits can lead to increased incidence of anxiety and depression, decreased quality of sleep, poor daily lives, declined work and social skills (P < 0.05). Body mass index in NSLBP patients can affect daily lives, work and social skills, but there was no significant correlation between the degree of obesity and patients' anxiety, depression, daily lives and VAS score (Table 2).
In the multiple logistic regression model, all predictive variables were significant after controlling for age, gender, and other variables. The effect of NSLBP VAS score, BAI and BDI on sleep quality PSQI was analyzed, results showed that patients with VAS > 7 were associated with higher PSQI score(OR: 2.248, 95% CI [1.235-4.093]) in comparison to patients with VAS 0-3. Patients with moderate anxiety (BAI 26-35) were associated with higher PSQI score (OR: 2.253, 95% CI [1.357-3.741]) compared with patients without anxiety (Table 4). Next, we evaluated the effects of PSQI, BAI and BDI on VAS score. Results showed sleep quality PSQI scores were strongly correlated with the severity of NSLBP  (Table 6). Finally, we evaluted the effect of PSQI, VAS score and BAI on depression BDI scores; results showed no significant correlation (Table 7).  Notes.

DISCUSSION
In this study we investigated the association between pain intensity and sleep quality in non-specific LBP patients, together with the relationship between pain intensity, patients' life quality, ODI score and emotions. Most of the previous studies on LBP and sleep disturbance focused on patients with persistent pain (Marty et al., 2008). In this study, we included patients with acute and chronic LBP, and we demonstrated that 29.0% of NSLBP patients and 12.5% included healthy people have sleep disturbance, healthy people usually have slight sleep disturbance in older age, while NSLBP patients have more severe sleep disturbance of all age. The prevalence of sleep disturbance in our study is lower than previous study (50-55%) (O'Donoghue et al., 2009).

Sleep and pain intensity in NSLBP patients
In this study, we found pain intensity is associated with the influence of sleep disturbance, woth severe NSLBP patients having more sleep disturbance. This is similar to previous study, which demonstrated that pain intensity was the most likely influence factor in LBP patients to be associated with sleep disturbance (Alsaadi et al., 2011). Several studies also confirmed significant relationship between sleep disturbance and pain intensity in patients with musculoskeletal pain (Vitiello et al., 2009). However, Lewandowski et al. (2010) reported day-time pain doesn't associated with objectively and subjectively assessed disturbed sleep. Tang et al. (2012) determined that pain didn't affect the subsequent night's sleep in patients with chronic pain and insomnia. The relationship of pain and sleep disturbance is conflicting, which may depend on the study design. It was said that some factors may confuse the results, such as chronic patients and sleep disturbance patients have higher BMI compared with control (Vorona et al., 2005).
Smoking also has been described strongly associated with sleep disturbance and LBP ( Van de Water, Eadie & Hurley, 2011). Therefore, we used multivariate regression to rule out the influence of these baseline factors, such as BMI, drinking and smoking. Results showed pain intensity and negative moods such as anxiety and depression have been strongly associated with sleep disturbance and pain (Wang et al., 2010). Pain has been reported to have bidirectional relationship with sleep; pain hinders sleep, while sleep disturbance also decreases pain thresholds and the mental capacity to manage pain (Schuh-Hofer et al., 2013). Austad et al. (2016) demonstrated sleep disturbance was moderately correlated to pain, fatigue, physical function, and utility in patients with rheumatoid arthritis. Therefore, we designed to evaluate the effect of sleep disturbance on the severity of pain intensity in NSLBP patients. Results showed patients with sleep disturbance have higher pain intensity compared to LBP patients without sleep disturbance, which is similar with previous study (O'Donoghue et al., 2009). The presence of disturbed sleep increase fatigue, daytime sleepiness and low mood, which may lead to negative impact individual such as weakening memory, work barrier, stress in daily life, poor life quality and more severe pain (Sezgin et al., 2014).

Sleep and depression in NSLBP patients
Studies have indicated that LBP was associated with patients' emotional status (Ramond et al., 2011). Polatin et al. (1993) reported 59% of the investigated LBP patients has psychiatric illnesses, the most frequently mentioned psychiatric illness are depression and anxiety. The prevalence of anxiety and depression were reported 48.57% and 55% respectively (Sagheer, Khan & Sharif, 2013) in chronic LBP patients. The current study also demonstrated significantly higher incidence of depression (68.3% vs 11.6%) and anxiety (50.1 vs 16.1%) in NSLBP patients compared with healthy people. Arnstein et al. (1999) demonstrated that severe pain is an important factor for depression and disability. Depression also has a bidirectional relationship with pain. Bair et al. (2003) demonstrated that depression is associated with more pain sites, greater pain intensity, longer duration of pain, and increased risk of poor treatment response. Depression produces substantial disability and decrements in health-related quality of life (Bair et al., 2008). Hung, Liu & Fu (2015) found depression was the most powerful factor associated with disability in LBP patients.
In relationship of sleep quality and emotional status, significant correlation between depression, anxiety, and sleep disturbance was well established (Guo et al., 2014), and insufficient sleep may result in depression (Xu et al., 2012). Longitudinal studies have identified insomnia as a risk factor for new-onset or recurrent depression, and this association has been identified in young, middle-aged and older adults (Franzen & Buysse, 2008). In clinical practice, sleep disorder and prolonged use of sleep medication may be early indicators or reversible risk factor for depression (Jaussent et al., 2011). Sleep disturbance was determined to be more prevalent among Chinese adolescents with depressive symptoms (Guo et al., 2014). Gregory et al. (2011) demonstrated sleep deprivation exacerbates fatigue, depression and pain, and sleep disturbance was moderately correlated with anxiety (r = 0.39) and depression (r = 0.50). This may create a perpetual cycle, where sleep deprivation leads to fatigue and apathy at work, impaired academic and social functioning, reduced levels of motivation, and impaired ability to regulate mood and emotional responses. In addition, depression is identified as the most frequent cause of chronic insomnia in both clinical and epidemiological samples (Tetsunaga et al., 2013). Emery, Wilson & Kowal (2014) found participants with depressive disorder had higher self-reports of pain and disability.

Sleep and anxiety in NSLBP patients
Anxiety disorders has been reported to be present in up to 60% of patients with chronic pain (Fishbain et al., 1986). Varni et al. (1996) found that the pain intensity has been associated with the severity of anxiety. Bair et al. (2008) found that musculoskeletal pain is much more disabling when depression and anxiety were both present. In addition, anxiety and sleep disturbance frequently co-occur, and this association persists across the lifespan (Spira et al., 2009). Sleep disturbance is commonly observed in individuals with anxiety and related disorders, and recent research suggests that sleep disturbance may predict the development of an anxiety disorder (Batterham, Glozier & Christensen, 2012). Stress is known to be the most common cause of transient insomnia. Anxiety and depression preceded the development of sleep disturbance among young women, whereas anxiety and sleep had a bidirectional association in men and older women (Spira et al., 2009). The bidirectional effects of sleep, anxiety or depression can be seen in many different hypothetical scenarios.

Effect on life quality
LBP is a disabling disease which restricts life quality, and psychological factors may have a larger impact on disability and quality of life than pain itself (Scholich et al., 2012). Previous studies have investigated the quality of life in LBP patients, and SF-36 is the most widely used HRQL forms to assess health concept in diseased groups as well as general population (Demiral et al., 2006). Soysal, Kara & Arda (2013) suggested the quality of life evaluated by SF-36 has significant impairment among LBP patients. Sleep disturbance, depression and anxiety are major factors affecting the life quality in patients with LBP. Granja et al. (2005) demonstrated that LBP and sleep disturbances were significantly associated with worse HRQoL in all dimensions. The incidence of disability in LBP patients has been reported to be 65%, and the disability assessed by ODI score was higher in LBP patients than healthy people. ODI is a validated tool to assess symptoms severity and functional change in patients with chronic LBP (Herndon, Zoberi & Gardner, 2015). Altug et al. (2015) demonstrated that LBP affects mobility of the patients, LBP patients feel their-selves more depressive. Houde, Cabana & Leonard (2015) reported that significant positive association was found between pain intensity and disability for both young and older individuals, with stronger correlation in the young group. In this study we evaluate the effect of LBP characteristic on patient's life quality, disability score and emotional property. We found that the disability and life quality decreased with the increased pain, and relationship of disability and life quality was inverse. This is similar to a previous study (Breivik et al., 2006;Harker et al., 2012).

Strengths and weaknesses of the study
The strength of the present study is that the results are based on the large prospective cohort study. A medical assistant in our research group distributed and helped patients to fill the questionnaire, which can improve the reliability of results. The response rate of our participants was more than 98%. We used a standard definition of non-specific LBP by experts in the field (Norton et al., 2016), MRI and medical history was taken to exclude organic diseases (e.g., spinal tumor, inflammatory, etc.) that could have strong effect on results.
The current study have some limitations. In this cohort study, the accuracy and completeness of the data within the database were dependent on the self-reported questionnaire, which may affect the internal validity of our study. The cross-sectional nature of the analyses limits causal inferences regarding the relationship between depression, anxiety and pain outcomes. Depression and anxiety were assessed with self-report measures rather than in standardized clinical interviews to diagnose specific mental disorders. Findings may not generalize to all racial and ethnic groups, as our samples were mostly southern Chinese. Age distribution may influence the depression scores in low back pain (Calvo-Lobo et al., 2017); further study may make full consideration of these limitations.

Implications of the study
The identification of interplay among pain intensity, sleep disturbance and emotion in patients with NSLBP has several potential uses. First, there are cross-relationship among pain intensity, sleep disturbance, anxiety and depression; Second, LBP also influenced by the social and psychological status of the patients and their quality of life (Nguyen & Randolph, 2007). Third, we should evaluate the sleep quality and emotion status of the LBP patients during clinical visit, and drug may be needed to treat the sleep disorder, anxiety and depression along with LBP. Fourth, though the importance of sleep and its association with anxiety, depression and pain intensity in LBP patients is gaining attention, much remains to be learned. To more fully elucidate the interplay of sleep, pain intensity and expression of psychological symptoms, cross-sectional and longitudinal research is needed.

CONCLUSION
In this study, we evaluate the cross-correlation among sleep disorder, anxiety, depression and pain intensity in patients with NSLBP. Psychological and social factors play an important role in the development of NSLBP. NSLBP lead to sleep disorder, which decrease the sleep quality and increase the unpleasant emotions and memories in return; these can exacerbate the severity of LBP, the cycle can repeating to form a vicious circle.