Associations between multisite pain, depressive symptoms and disturbed sleep with work ability and health-related quality of life in female health care workers with recurrent nonspecific low back pain: a cross-sectional study

: Background : The purpose of this study is to investigate how multisite pain, depressive symptoms and disturbed sleep are associated with health-related quality of life (HRQoL) and work ability index (WAI) in health care workers with recurrent non-specific low back pain (LBP). Methods: 219 female health care workers suffering from recurrent non-specific LBP were recruited for the study. Multisite pain (three or more pain sites with pain intensity of 4 or more on the numeric rating scale), depressive symptoms [modified Finnish version of the nine-item Patient Health Questionnaire (PHQ─9-mFIN)], disturbed sleep, HRQoL (RAND-36) and WAI short form were assessed by validated questionnaires. A generalized linear mixed model with and without covariates was used for statistical analysis. Statistical significance was set at p< 0.05. Results : Depressive symptoms and disturbed sleep were significantly associated with mental HRQoL and WAI (p<0.001). Multisite pain was only significantly associated with physical HRQoL. Conclusions : In female health care workers with recurrent non-specific LBP and currently able to work, depressive symptoms and disturbed sleep were associated with decreased self-reported work ability. Actions to identify these symptoms may decrease work absenteeism due to LBP.

physical and psychosocial risk factors for musculoskeletal pain that include constrained postures, forceful movements, high emotional strain, and pressure caused by staff shortages (11). Health care workers often have to work irregular shifts and, as a result, frequently experience sleep disturbances (12). Work-related psychosocial factors, such as high job demand, low job control and low social support, play an important role in the prevalence and incidence of low back pain in health care workers. Organizational factors also play an important role in the occurrence of MSD. (13). Among health care workers, musculoskeletal disorders are an important risk factor for them to consider changing job or even leaving the nursing profession (14).
LBP is often concurrent with pain from other body sites (2). Multi-site musculoskeletal pain has been associated with a greater negative impact on patients` physical functioning and disability, leading to an increased risk for depressive disorders. Moreover, the reporting of multi-site pain appears to worsen the prognosis -there is an increased likelihood of the problem becoming chronic. Multi-site musculoskeletal pain also increases the risk for poor future self-perceived work ability (15). There is a strong association between number of pain sites and sleep quality as well as psychological distress and overall health (16). The afore mentioned relationships are complex and interactive and might therefore be components of a larger, multi-symptom syndrome. Widespread pain together with insomnia and symptoms of depression pose a significant threat to work ability (17).
Psychosocial factors play an important role in the development of disabling persistent LBP, and depression might have an adverse effect on the prognosis of LBP (18). Patients with chronic LBP have significant impairment of the psychological status and reduced HRQoL (19).
Sleep disturbance is highly prevalent in the general population, especially among women (20).
Poor sleep quality is known to cause a wide range of physiological and psychological symptoms. Furthermore, sleep disturbance is a common feature of LBP, and the estimated prevalence of sleep disturbance among patients with LBP is about 59% (21). Insomnia forms a health risk and has been associated with a significant risk for lower QoL (22).
Work ability is defined as how physically and mentally able a worker is to cope with the demands of their work (23). The Work Ability Index (WAI) is an instrument that has been widely used to assess work ability (24). The WAI predicts both register-based disability pension (DP) and longterm sickness absence. The individual's own evaluation of work ability compared with lifetime best work ability (WAS) and workers belief of future work ability (FWA) predict both disability pension and long-term sick leave (>10 days) (25). A previous study has indicated that those female workers who have high WAI scores also have higher quality of life, even after retirement (26). Another study suggests that the beliefs of returning to work, pain intensity and work strain are predictive of work ability among women on sick leave due to long-term LBP (27). In addition, individual factors, such as older age, have been associated with poor work ability (28). In the general working population, poor work ability combined with one or more chronic disease is associated with high risk for long-term sickness absence (29).
The causes of the high prevalence, incidence and recurrence of LBP are multifactorial, and therefore LBP should be understood within a biopsychosocial framework.

Aim of the study
The aim of this study is to investigate the associations of multisite pain, depressive symptoms and disturbed sleep with quality of life and future work ability in female health care workers with recurrent nonspecific LBP, and to identify those symptoms that are the strongest risk factors for reduced QoL and work ability.

Study design and participants
This cross-sectional study was part of a randomized controlled trial (the NURSE RCT, clinical trial registration NCT01465698) aimed at reducing pain, movement-control impairment and fearavoidance beliefs in working female health care workers with recurrent non-specific LBP (NSLBP) (n=219) (45). The NURSE-RCT (30) was conducted in the form of 3 identical, consecutive substudies. More precise information on recruitment is available in the protocol article on the NURSE RCT (30,31). The study was approved by the Ethics Committee of Pirkanmaa Hospital District, Finland (ETL code R08157).
The inclusion criteria were women aged 30-55 years; worked at current job for at least 12 months; intensity of LBP of at least 2 on the Numeric Rating Scale (scale 0-10) during the past 4 weeks (30. The exclusion criteria were serious former back injury (fracture, surgery, disc protrusion); chronic LBP defined by a physician or self-report of continuous LBP for 7 months or more (30); disease or symptoms that limit participation in moderate intensity neuromuscular exercise; regular engagement in neuromuscular-type exercise more than once a week; pregnant or recently delivered.
Independent variables were set approximately three same size of groups. Group 1 was multisite pain (NRS 4 or more, pain daily or nearly daily and at least 3 pain sites, n=50). Group 2 was depressive symptom (PHQ-9-mFIN sum score from 9 questions; points 10 or more, which is categories at least moderate depression, n= 61). Group 3 was disturbed sleep (sum score from 3 questions developed by the Finnish Institute of Occupational Health, points 10 or more, n= 55).

Measurements
Questionnaire data was collected during the baseline measurements. Before the baseline measurements informed consent was obtained in writing from all participants.

Dependent variables
The Work ability was assessed using the short form of WAI, which is the sum score (range 3-27) from four question items: (1) Current work ability (0-10), where 0=unable to work and 10=the best possible), (2) Work ability in relation to physical work demands (1-5), (3) Work ability in relation to mental work demands (1-5) (1=very poor, 5=very good), and (4) Personal prognosis for work ability in 2 years' time (1=hardly able to work, 4=not sure, 7=almost certain work ability) (33).

Independent variables
The criteria for multisite pain during the past 4 weeks were as follows: number of pain sites ≥3 (lower back, upper back and neck, shoulders and upper limbs, hips, knees and lower limbs) with daily or nearly daily pain and intensity of pain of at least 4 on the Numeric Rating Scale (0-10) (34). (3=no sleepiness or tiredness, 13=daily long-term sleepiness or tiredness). Disturbed sleep contained the following items: 1) Feeling of unsatisfactory awakening lasting for at least one month (score range 1-3), 2) Feeling of tiredness in the daytime (score range 1-5) and 3) Feeling of sleepiness in the daytime (score range 1-5).

Statistical analysis
The associations between the main independent variable and the outcomes were estimated and tested with generalized linear models with gamma distribution due to the skewed distribution of the dependent variables. First, all models included only one independent variable at a time. In the second stage, multiple covariates were included in the models. All non-significant covariates (p>0.20) were dropped from the models one at a time. For the parameter estimates and p-values, 95% confidence intervals were reported for each model. Independent variables with p<0.05 were considered statistically significant. All the analyses were performed using IBM SPSS software (version 24, IBM Corp, Armonk, NY, USA).

Covariates
The covariates included background variables, such as age, civil status, level of education and smoking, work-related factors covering shift work, health-related factors, such as perceived health, cardiovascular and respiratory diseases, blood pressure disease and the use of medication, and fitness including a modified push-up test and six-minute walk test (6MWT) (37), and amount of exercise.

Descriptive data of the study participants
Baseline characteristics of the study participants are presented in Table 1. The mean age of the participants was 46 years. In total, 41% of participants had normal body weight (BMI under 25 kg/m2) and 18% were obese (BMI 30kg/m2 or over). Most of participants (62%) perceived their health to be good or very good, and 70% of participants were working shift work. More than one third (36%) of the study participants experienced LBP daily or on most days of the week. The mean LBP intensity was 36, measured on the 100 mm visual analog scale (VAS). The duration of the LBP symptoms were less than three months for 65%, 3-6 months for 15% and more than 6 months for 21% for of the study population. In total, 41% of participants reported musculoskeletal pain at 4-6 sites. Depressive symptoms and sleep disturbance indices, mean (SD) Depressive symptoms** 7,4 (3,77) 2 Disturbed sleep*** 7,53 (2,78) ______________________________________________________________________________________________________________________ * Multi-site pain (NRS >=4, pain daily or nearly daily and at least 3 pain sites, n=50) ** Depressive symptoms (sum score from 9 questions; points 0-27, highest quarter of all participants, n=61) *** Disturbed sleep (sum score from 3 questions developed by the Finnish Institute of Occupational Health; from 3=no tiredness or sleepiness to 13 = long-term, daily tiredness and sleepiness, sum score 10 or over, highest quarter of all participants, n=55)

Health Related Quality of Life and Work Ability
Descriptive data on the outcome measures (HRQoL, work ability) of the study are presented in Table 2. The mean physical and mental scores of the summary scales of RAND-36 were 73 and 77, respectively (0-100 scale). When compared with the normative data of the female Finnish population (50), these health care workers had lower values for the physical components of RAND-36, especially bodily pain, but the values of the mental components of RAND-36 were higher. The majority of the study participants reported their work ability to be good. However, 12% of the women believed their future work ability to be poor.  WAS (current work ability compared with lifetime best, 0= completely unable to work, 10= work ability at its best) FWA (individual`s own prognosis of his/ her future work ability, poor (1,4) and good (7)) The associations of the physical components of RAND-36 with multisite pain, depressive symptoms and disturbed sleep are presented in Figures 1-4. The bodily pain subscale of RAND-36 was associated with multisite pain when adjusted (p< 0.001), but not with depressive symptoms (p=0.426) or disturbed sleep (p= 0.282) (Fig. 1). The physical functioning subscale of RAND-36 was also associated with multisite pain (p= 0.003), but not with depressive symptoms (p=0,614) or disturbed sleep (p =0,614) (Fig.2). The physical role functioning subscale of RAND-36 was not associated with multisite pain (p=0.104), depressive symptoms (p=0.620) or disturbed sleep (p=0.127) (Fig. 3). However, the general health subscale of RAND-36 was associated with multisite pain (p=0.008) and depressive symptoms (p =0.006), but not with disturbed sleep (p=0.079) (Fig.   4).
The associations of the mental component of the RAND-36 subscale with multisite pain, depressive symptoms and disturbed sleep are presented in Figures 5-8. The social functioning subscale of RAND-36 was associated with depressive symptoms (p<0.001) and disturbed sleep (p < 0.001), but not with multisite pain (p=0.077). The vitality subscale of RAND-36 was associated with depressive symptoms (p<0.001) and disturbed sleep (p<0.001), but not with multisite pain when adjusted (p=0.193) (Fig. 6). The mental health subscale of RAND-36 was associated with depressive symptoms (p<0.001) and disturbed sleep (p<0.001), but not with multisite pain after adjustments (p=0.142) (Fig.7). The emotional role functioning subscale of RAND-36 was associated with depressive symptoms and with disturbed sleep (p<0.005), but not with multisite pain (p=0.783), and adjustments did not affect the results (Fig.8).
Work ability index (WAI) was associated with depressive symptoms (p=0.007) and disturbed sleep (p<0.001). WAI was not, however, associated with multisite pain when adjusted with age, perceived health and modified push-up test (p=0.155) (Fig.9).
Of the physical components of RAND-36, physical functioning was associated (p< 0.001) with poor future work ability (FWA) after adjustments. The physical role functioning of RAND-36 was also associated with poor FWA (p=0.031) ( Table 3). Of the mental components, social functioning was associated with poor FWA (p 0.009). In addition, LBP intensity was associated with poor future work ability (p 0.001) ( Table 3).

Discussion
The most important findings of this study indicate that recurrent nonspecific LBP combined with depressive symptoms and disturbed sleep have a negative impact on mental HRQoL and work ability in female health care workers. Furthermore, a clinically meaningful intensity of LBP (i.e., VAS ≥ 40 mm) was strongly associated with self-reported future work ability.
In this study, the participants were female health care workers with recurring nonspecific LBP who were still able to work. Among female health care workers, LBP is a common health problem that is accompanied by several individual, physical and psychosocial risk factors, such as high workrelated emotional strain (8). The mean intensity of LBP among the female health care workers in the VAS scale was 36 mm, which is close to the clinically meaningful criterion of 40 mm. Most of the women had pain symptoms that had lasted for less than three months and 72% had multisite pain in three or more sites. Previous studies have concluded that multisite musculoskeletal pain is a threat to work ability (15) and that widespread pain, insomnia and symptoms of depression together are a serious threat to workability (17). The findings of this present study, however, show that depressive symptoms and disturbed sleep were more significantly associated with work ability than multisite pain.
Moreover, the majority of the female health care workers reported their work ability and their self-perceived assessment of future work ability (FWA) to be good. However, 12% of the participants were uncertain of their future work ability. Previous studies have shown that poor future work ability predicts long-term sickness absence, disability pension and long-term unemployment (25,38). Furthermore, poor work ability combined with chronic disease is also associated with high risk for long-term sickness absence (29). Patients with LBP are also at risk for poor health-related quality of life (HRQoL) (39). In this study, those participants with poor FWA also had lower HRQoL. In addition, there was a strong association between LBP intensity and poor FWA.
The limitations of the present study include the cross-sectional design and the limited questionnaire used for assessing the quality of sleep. The strengths of the study include the unique study population in terms of non-chronic LBP (i.e., sub-acute or recurrent LBP) with physically strenuous work. Another strength of the study is the use of the modified Finnish version of the nine-item Patient Health Questionnaire, which was recently validated among the present study population (35).

Conculsions
In conclusion, identifying those people with a good or unfavorable prognosis among patients with LBP is an important goal (40). Moreover, recommendations have been made for the use of screening methods in health care to identify those patients in the early stages of LBP with the aim of guiding them to the appropriate treatment to support their staying at work (25). Therefore, the questions relating to depressive symptoms, sleep quality and self-reported future work ability used in this study might be appropriate tools for identifying patients with poor HRQoL and/or work ability outcome among health care workers with recurrent non-specific LBP. Based on the current results, it would be interesting to define the cut-off points for the psychosocial symptoms of depression and sleep disturbance and to ascertain when they begin to significantly affect HRQoL and work ability. Furthermore, future studies are needed to evaluate the effectiveness of interventions designed to modify psychosocial aspects in LBP patients.

Declarations
Ethics approval and consent to participate: The Ethics Committee of Pirkanmaa Hospital District (ETL code R08157) approved the study protocol. The participants recruited to the study received a written information letter clarifying the aims of the study, including risks and benefits. The participants provided their written consent to participate at the first study appointment. We also obtained consent for data sharing and the risk of identification is low.

Consent for publication: Not applicable.
Availability of data and material: The datasets used and analyzed during the current study are available from the corresponding author upon reasonable request.

Competing interests:
The authors declare that they have no competing interest.