Depressive symptoms in patients with chronic pain

Views on the association between chronic pain conditions and depression have a long history in both the pain and psychiatric literature. These ideas have included pain being an antecedent to depression, depression being a consequence of chronic pain,1 or a common comorbid complication.2 It has also been argued that chronic pain, especially when it is deemed “non-specific”, could itself be seen as a forme fruste of depression or masked depression.3 Estimates of the proportion of patients with chronic pain and mental disorders vary from almost none (1.5%) to perhaps all (100%),2 with most studies nominating figures in the mid-range. For example, one study reported that 59% of one sample of patients with chronic pain demonstrated current symptoms for at least one psychiatric diagnosis, mainly major depression, substance misuse and anxiety disorders.4 In another study of patients with chronic pain, 30%–40% were reported to be depressed.5

In recent years, research in this field has attempted to examine the nature of depression in populations of patients with chronic pain. In part, this has been prompted by the clinical observation that there is often symptom overlap between clinical depression and the common features of chronic pain conditions.6 Clinically, this is manifest by the limitations found when using somatic symptoms in self-report measures of depression in both populations with depression and those with chronic pain,7 and has generated debate about the utility of categorical diagnostic criteria, as described in the Diagnostic and statistical manual of mental disorders, fourth edition (DSM-IV),8 for understanding and treating depression in people with chronic pain.9

The importance of investigating the putative association between chronic pain and mental disorders is underscored by two principal findings. First, there is evidence that patients with chronic pain who have a comorbid mental disorder have a more impaired quality of life than those without this comorbidity.2,10 Second, treatment outcomes can be compromised when both pain and a mental disorder are present, but only one of them is targeted for treatment. For example, higher levels of depression have been shown to be predictive of poorer treatment outcomes for patients with chronic pain,11 as well as higher health care costs over time.12 Equally, the presence of pain in people being treated for mood disturbance has predicted delayed responses to mood interventions.13 Therefore, improving our understanding of how chronic pain and depression are linked, and treating both components offers the prospect of enhancing treatment effects beyond the benefits of treating either condition alone.14

In this study, we examined three main questions. In a sample of patients with chronic pain:

• What somatic (pain) and psychological variables correlate with depressive symptoms, measured by the depression subscale of the Depression Anxiety Stress Scales (DASS)?15

• After controlling for other variables known to affect physical disability, do depressive symptoms measured by the DASS depression subscale correlate with physical disability, as measured by the Roland Morris Disability Questionnaire (RMDQ)?16

• Which DASS depression subscale items are rated as most frequently experienced?

Data were available from a total of 2419 patients with pain, who were referred to the Pain Management Research Institute at Royal North Shore Hospital, Sydney (a tertiary referral centre), for assessment and treatment between January 2000 and December 2007. Demographic data were collected, and all patients were asked to complete a set of self-report measures. Use of the de-identified dataset for this study was approved by the Human Research Ethics Committee of Royal North Shore Hospital.

Complete datasets were available for 812 of the 2419 referred patients. The mean age of these 812 participants was 49.58 years (SD, 16.25), and most (56.5%) were female. The mean duration of pain was 84.43 months (SD, 117.56).

Box 1 shows mean scores for the study participants on each independent variable. A comparison of the 812 participants who had complete datasets with the 1607 patients who had incomplete datasets showed no differences in terms of sex ratio, but the incomplete dataset cohort were slightly older (mean age, 53.4 years; SD, 17.1) and had had a longer duration of pain (mean, 95 months; SD, 129.1). However, on the assessment measures, there were no significant differences between the two groups. Accordingly, we believe that the results of the study sample are generalisable to the full dataset.

DASS scores for depressive symptoms for our sample of patients with pain (mean, 13.61; SD, 11.54) were higher than those of a healthy community sample (mean, 5.06; SD, 7.57) and lower than those of a sample of (non-pain) patients with diagnosed mood disorders (mean, 17.24; SD, 11.79).7 Only 30.3% of our patients with pain scored higher than the mean score of the mood disorder group.

Our finding that the severity of depressive symptoms in our sample of patients with chronic pain was best correlated with a combination of heightened catastrophising, reduced sense of control over life, increased physical disability, lower pain self-efficacy beliefs, higher use of unhelpful self-management strategies, and lower perceived social support (after controlling for the possible effects of age, sex and duration of pain) is consistent with previous studies of patients with chronic pain.26 Interestingly, and somewhat contrary to clinical expectations, pain severity, pain-related distress, and fear of movement/(re)injury were not significantly associated with depressive symptom severity.

Our finding that the severity of depressive symptoms was a significant but relatively smaller contributor to physical disability in this sample (after controlling for the possible effects of age, sex and duration of pain) is consistent with findings of some previous studies of patients with chronic pain, but not with some treatment studies, which found that depression level contributed to less significant improvement in pain-related disability.11,27 It is not surprising that cognitive, pain and behavioural variables accounted for more physical disability than depressive symptoms but it is notable that social support (as measured by the MPI), sense of control over life, and catastrophising did not significantly contribute to physical disability. Taken together, these findings appear consistent with recent calls to treat both persisting pain and depression in patients presenting with both, rather than expecting treatment aimed at one to affect the other.28

Our finding that patients with pain reported that they more frequently experienced the depression subscale items reflecting sadness, lack of initiative and anhedonia than those reflecting low self-worth and lack of meaning is consistent with some earlier research in similar samples of patients with chronic pain.6,9 It is also interesting to compare these items to those endorsed by patients with both depression and medical illnesses, who invariably favour a cognitive formulation.29

Although it is likely that our findings are generalisable to patients with chronic pain and diagnosed depression, further research is needed to confirm this. A potential limitation of our study is the large number of assessment tools administered to patients who may have incurred some degree of fatigue, and this may have affected the reliability of responses obtained. However, as the measures are completed over the course of 3–4 hours, with opportunity for breaks, and given that our findings are consistent with those reported elsewhere, it seems unlikely this was a major factor. In addition, previous research from the same centre that included the same assessment tools suggests a high degree of reliability in the data obtained.30

Overall, our findings suggest that depressive symptoms, as measured by the DASS, were associated more strongly with cognitive variables than pain experience in our large sample of patients with chronic pain. Similarly, physical disability due to pain appears more related to fear of re-injury and low self-efficacy for activity when in pain than either depressive symptoms or pain intensity. These findings are consistent with previous correlational studies in similar populations,20 but the small effect of depressive symptoms on disability is interesting, given previous findings that patients with both depression and pain are more recalcitrant to rehabilitative treatment than those without depression. Patients’ endorsement of depression subscale items of the DASS may shed some light on this. The relatively high degree of endorsement of mood-related items versus those reflecting low self-worth and lack of a sense of meaning suggests differences in the experience of depressive symptoms in this population relative to those presenting in a mental health setting. This finding might also explain why relatively brief cognitive-behavioural interventions for pain that target unhelpful pain-related cognitions rather than longer-standing concerns about self-worth might also be effective for many patients with both depression and chronic pain.26,31 Ultimately, our study reaffirms the view that depression is multifaceted, and that variations across clinical presentations reflect putative “subtypes” of depression.32