The impact of depression among chronic low back pain patients in Japan

Low back pain (LBP) is a common health issue affecting at least 80 % of individuals during their lifetime [1] and poses a severe economic burden on individuals and their communities [2‐5]. The Global Burden of Disease Study 2013 found that globally, back pain was one of the leading cause of years lived with disability (YLDs) [6]. In Japan, back pain is the top cause of YLD and the 2^nd and 4^th most frequent reason for outpatient visits for women and men, respectively [6, 7].

One of the main characteristics of LBP is recurrence, and a number of patients develop chronic LBP (CLBP). In Japan CLBP is the most prevalent type of chronic pain [8], with a prevalence estimated at 23 %, and 11–12 % of the population is disabled by it [9]. Though considerable research has been directed at understanding back pain, most Japanese epidemiological studies examine LBP in general, with few focused on CLBP [10‐12].

While burdensome in its own right, pain is also risk factor for depression, and many studies have examined the co-occurrence of pain and depression [13‐16]. The comorbidity is clinically well established but the underlying mechanisms are not well understood, though a potential explanation is disruption of the mesolimbic dopamine system [17, 18]. Recent data from animal models indicate that regulation of dopamine activity in the ventral tegmental area (VTA) mediates depressive and anxiogenic responses [19] suggesting a neurological link between depression and chronic pain.

CLBP in particular is often co-morbid with depression [20], a main cause of disability worldwide [6]. Depression increases the risk of developing LBP [21], and CLBP is affected by the patient’s mental state [22]. In spite of that, the mental state of most CLBP patients is not routinely assessed. Thus, in chronic pain, psychosocial risk factors become relevant, and are important to explain how individuals respond to back pain. Recent studies have demonstrated that psychosocial factors are important risk factors for LBP among Japanese workers [22, 23]; however, data examining the role of depression in CLBP patients in Japan is lacking.

The objective of this study was to investigate the impact of depression on health-related quality of life (HRQoL) in CLBP, as well as to assess the relationship between depression and work impairment and healthcare use among CLBP patients in Japan.

Of the participants surveyed, 425 were identified as having CLBP. The average age of a respondent with CLBP was 54 years old, and 44 % were female (Table 1). When assessed according to depression status, CLBP patients with depression (PHQ-9 ≥ 10; N = 70) were younger than CLBP patients without depression (PHQ-9 < 10; N = 355) by approximately 9 years on average, but did not differ in terms of average CCI score, gender, or employment status. Patients with depression were less likely to be married or live with a partner (Table 1). Patients indicated their LBP was either mild (47 %) or moderate (44 %) rather than severe (9 %). Both overall severity of pain and current level of pain were near the midpoint of the NRS, and almost half reported daily problems with pain. Depression was significantly associated with more severe pain and higher levels of pain, current and in the prior week (Table 1).

CLBP patients with depression had worse HRQoL than CLBP patients without depression (Table 2). Depression was also associated with more impairment while at work (presenteeism). Overall work impairment, which is largely driven by presenteeism, was also significantly higher among CLBP patients with depression. There was no significant difference in absenteeism or rate of labor force participation between CLBP patients with and without depression. Depressed CLBP patients reported more activity impairment than those without depression. Depression was also associated with approximately two more healthcare provider visits among CLBP patients in the 6 month recall period (Table 2).

The pattern of results was consistent after covariates were incorporated into the regression analysis. Adjusted HRQoL scores were lower on all of the eight Japanese norm-based scores. Adjusted mean MCS and PCS using international norms were also lower (Fig. 1).

Regression-adjusted presenteeism and overall work impairment were 1.8 and 1.9 times as high, respectively, among those with depression relative to those without depression (Fig. 2). Activity impairment was 1.7 times as high in patients with depression compared with patients without depression after adjustment for covariates. HCP visits were almost twice as frequent in patients with depression compared with patients without depression. Likewise, work impairment was greater in patients with depression compared with patients without depression.

Analysis of depression based on PHQ-9 scores as a continuous variable also demonstrated the association between depression and pain among CLBP patients. Greater depression was significantly associated with more frequent problems with pain, greater current and past-week severity of pain (based on NRS scores), pain at more sites in addition to LBP, and more presenteeism and overall work impairment (P < 0.001, Table 3). Moreover, additional regression analysis conducted using PHQ-9 scores as a continuous variable corroborated the findings, indicating lower HRQoL scores with higher PHQ-9 scores, with the exception of the Japanese PCS score. Pain was likewise worse with greater depression as was presenteeism, overall work impairment, and activity impairment. Consistent with the results shown in Fig. 2, HCP visits were more frequent with greater depression scores, but there was no significant association with ER visits or hospitalizations (data not shown).

When assessing the relationship between work impairment and other characteristic and outcomes, presenteeism was very closely related to overall work impairment (rho = 0.99). Greater presenteeism was associated with more-severe LBP, more-severe pain in the prior week and currently based on the NRS. Although, there was a trend for greater presenteeism being associated with more frequent problems with pain, it did not reach statistical significance (P = 0.08). Presenteeism was moderately related to the severity of depression according to the PHQ-9 score (Table 4).

Our results demonstrated that CLBP patients with depression had significantly more severe and higher levels of pain, as well as significantly worse HRQoL compared with CLBP patients without depression. These observations are consistent with those recently published by Hiyama et al., which showed that depressed patients and those with neuropathic LBP had a higher level of pain and poorer quality of life compared with non-depressed patients [16]. The majority of patients had mild (47 %) or moderate (44 %) LBP. Current and prior week pain severity scores were similar (4.6/10 vs 4.5/10) and almost half of all patients reported daily pain problems. Overall sociodemographic patient characteristics were similar between the two groups of CLBP patients with the exception of age, marital status and sleeping difficulties. CLBP patients with depression were significantly younger, on average 9 years, compared with CLBP patients without depression. These observations tend to be consistent with observations for major depressive disorder where estimates in the general population are 15–17 %, while the 1-year prevalence rate in individuals ≥ 65 years is lower, at 1–4 % [31]. Significantly more CLBP patients with depression were single/divorced compared with CLBP patients without depression (44.3 % vs 30.1 %). However, differences in marital status and sleeping difficulties were consistent with differences observed in major depression disorder [27].

Epidemiological, cross-sectional, and prospective studies suggest that insomnia, chronic pain and depression are a cluster of symptoms that are mutually interactive. Studies using a variety of methods, including neuroimaging, suggest the mesolimbic dopamine system has been proposed as a key factor in promoting the comorbidity of this cluster of symptoms, [32] and our observations of both higher ratings of pain severity as well as greater prevalence of sleep difficulties among CLBP patients with depression are additional supportive evidence to this body of data.

The adjusted mean HRQoL scores in the CLBP depression group were lower than in the CLBP group without depression. The health status using both Japanese norm-based as well was international scores, indicated significantly poorer outcomes for CLBP patients with depression compared with CLBP patients without depression. Lower PCS scores in CLBP patients with depression are indicative that a decline of mental health could have an effect on physical health in CLBP patients. A similar relationship has been reported among CLBP patients in the United Kingdom, in whom depression as measured by the Hospital Anxiety and Depression Scale (HADS) was correlated with PCS scores [33]. Labor force and absenteeism did not differ by depression status, potentially because of Japanese working habits, where there is a tendency for less sick leave claims compared to other countries [34]. However, presenteeism, overall work and activity impairment were lower in CLBP patients with depression, demonstrating that, even though employees are present at work, they are less productive than those CLBP patients without depression. Additional analyses indicated that presenteeism was closely related to overall work impairment. The current study also demonstrates more frequent use of healthcare among CLBP patients who have depression, consistent with the relationship between depression and healthcare visits recently demonstrated in the US using National Health and Nutrition Examination Survey data [35].

Treatment approaches, especially for Japanese workers, have focused on ergonomic approaches in the management of LBP. Consistent with a focus on musculoskeletal symptoms the majority of patients surveyed in our study were diagnosed with LBP by an orthopedist. However, recent studies highlight the importance of psychosocial risk factors in the development of CLBP [22, 23] and our data further highlights the need for mental health evaluation and treatment in addition to physical assessment and therapy.

One limitation of our study is that the analysis was cross-sectional. Therefore our results cannot indicate whether increased pain leads to depression, or whether depression leads to increased pain. Another limitation is selection bias that may not result in an all-encompassing representation of all patients with CLBP. The data were derived from opt-in surveys completed over the Internet. Compared to the general population our study population could be over-representative of individuals who live in urban environments and are technology literate.

Nakamura et al. has shown that chronic musculoskeletal pain does not necessarily improve with treatment and that patients have a high degree of dissatisfaction with it [11, 12]. Ineffective treatment may lead to “doctor shopping”. In our study, a significantly higher number of CLBP patients with depression than those without depression were using prescription pain medication (55.7 % vs 35.8 %, P = 0.002) indicating that depressed CLBP patients not only suffer more but may also find treatment less effective. Moreover, increased mental and physical suffering often require assistance. All these factors pose undue strain and increase societal cost.

We have demonstrated that depression among CLBP patients is associated with higher pain scores and lower HRQoL scores, as well as lower labor productivity and increased healthcare use. Our results underscore the need to screen for depression in CLBP patients as an essential part of CLBP patient care.