It is easy to think about rheumatologic illnesses, including chronic inflammatory and arthritic conditions such as rheumatoid arthritis (RA), as disorders that primarily affect the musculoskeletal and immune systems. But these rheumatologic disorders have prominent neuropsychiatric symptoms and many other confounding and complicating factors. Pain, systemic inflammation, disability, fatigue, sleep disruption, and treatment effects—all of these accompany rheumatologic conditions, and are associated with depression in and of themselves.

Although some rheumatologic illnesses—most notably systemic lupus erythematosus—directly cause neuropsychiatric symptoms, the depression occurring in most of these patients is multifactorial. This chapter reviews the diagnosis and treatment of depression when it occurs in the context of a rheumatologic disease. It includes an initial discussion of the role of two salient factors common to rheumatologic disease in general—systemic inflammation and corticosteroid treatment—and then turns to the specific diseases whose association with depression has been studied: systemic lupus, RA, scleroderma, and fibromyalgia.

The relationship of depression to inflammatory processes has long been a subject of investigation.¹ Significant associations were discovered between inflammation and depression, including epidemiologic observations (the high female to male ratio, and lower rates of depression in cultures with diets rich in fish) and biological associations (e.g., between symptoms of depression and pro-inflammatory cytokines). Although there is strong evidence of a link between depression and inflammation, the precise mechanisms remain elusive.

Depression is substantially more prevalent in patients with rheumatologic illnesses than in the general population,² and depression is highly correlated with disease-related morbidity within this population. Compared to other major medical comorbidities, patients with rheumatologic illness exhibit the strongest correlation between depression and lower quality of life measurements.³ Depression also appears to predict work disability more strongly in this population than among patients without a rheumatic illness.⁴

The link between depression and rheumatologic illnesses likely includes both psychological and biological components. Restricted independence, higher medical burden, and patients’ concerns about appearance (e.g., skin lesions or musculoskeletal deformities) have all been associated with depression in these patients.^5,6

What is known about the biological link between depression and systemic inflammation remains limited, but some associations have been demonstrated in research studies (Chapters 1 and 2). Healthy test subjects given an endotoxin (which promotes systemic inflammation as manifested by a rise in body temperature and in pro-inflammatory cytokines) (Table 10-1) develop depressed mood, anxiety, and memory impairment.⁷ Patients with existing major depression have higher concentrations of TNFa and IL-6 even without stimulation by an endotoxin.⁸

Furthermore, antidepressants may reduce at least some markers of systemic inflammation. A meta-analysis showed an inconsistent association between treatment with SSRIs, SNRIs, and TCAs and a reduction in TNFa and IL-6 levels and a consistent association with reduction in IL-1-beta (IL1B).⁹ The association between antidepressants and reduced IL-6 levels was stronger for SSRIs than for other antidepressant classes.

Nonpharmacologic treatment modalities for depression have been examined to a lesser extent in this regard. The evidence supporting these treatments for patients with specific rheumatologic disorders is discussed later in this chapter under the disorder subheadings. It is additionally worth noting that one study demonstrated that cognitive-behavioral therapy (CBT) correlated with lower IL-6 levels in women with depression who did not have a rheumatologic illness.¹⁰ Electroconvulsive therapy (ECT), by contrast, was shown to increase proinflammatory markers, including TNFa and IL-6, after a single session, although repeat sessions did not further affect these parameters.¹¹ Studies examining the efficacy of repetitive transcranial magnetic stimulation (rTMS) in treating rheumatic pain have been conducted but have not consistently shown a benefit.¹²

Conversely, some immunosuppressive and anti-inflammatory agents have been associated with antidepressant effects in studies, although because these agents have an established, powerful effect in lowering disease activity (and thereby reducing pain and disability) these findings must be interpreted with caution. Etanercept was associated with a reduction in depressive symptoms when administered to patients with psoriasis.¹³ Celecoxib was associated with greater reductions depressive symptoms and in IL-6 levels when provided as an adjunct to sertraline, compared to sertraline with a placebo adjunct.¹⁴

THE ROLE OF PRESCRIBED GLUCOCORTICOIDS (BOX 10-8)

Glucocorticoids (GCs), which are frequently prescribed in rheumatologic disorders, affect mood and other psychiatric symptoms. In addition to the systemic risks attributed to GC therapy, which include osteoporosis, cardiovascular effects, diabetes mellitus, and weight gain,¹⁵ associated neuropsychiatric side effects include depression, mania, psychosis, delirium, cognitive impairment and memory dysfunction, suicide,^16,17 and sleep.¹⁸ Historically, the risk of neuropsychiatric side effects from GCs has been considered to be very low at doses below 40 mg per day of prednisone (or equivalent), and higher at doses greater than 80 mg per day.¹⁹ However, more recent evidence suggests that lower doses of prednisone (7.5 mg per day or greater) also appear to increase the risk of depression when prescribed for at least 6 months.²⁰

Among affective symptoms, depression appears to be most closely associated with chronic GC therapy,¹⁶ while short courses of treatment are more commonly associated with mania.^21,22 Paradoxically, however, some patients with preexisting depression often experience an improvement in their moods with GC administration.²¹ Some authors have argued that a steroid-induced mood disorder (either depression or mania) may more closely resemble bipolar disorder than major depression—not only in symptomatology but possibly also in its response to treatment.¹⁶

The evidence regarding the choice of the agent to treat steroid-induced mood symptoms is limited.¹⁶ Studies of conventional antidepressants have demonstrated improvement in depression^16,23 in some and adverse outcomes such as agitation and psychosis in others.²⁴ Lithium provides effective prophylaxis against steroid-induced psychosis,²⁵ and lamotrigine reduces negative effects of prescribed GCs on memory function.²⁶ Evidence supporting the use of antipsychotics in treating steroid-induced mood symptoms is positive but remains primarily based on uncontrolled trials²⁷ and case reports,¹⁶ and meanwhile use of these treatments necessitates considering their potentially unfavorable side-effect profiles. Steroid-induced mood disorders are typically reversible with discontinuation of the GC.¹⁶

Systemic lupus erythematosus (SLE) is an autoimmune disease that causes inflammation and tissue destruction in multiple organ systems. The prevalence of lupus varies between 0.03% and 0.14% worldwide²⁸; it is more common among women and among non-white ethnic groups.²⁹

EPIDEMIOLOGY

The reported prevalence of depressive symptoms in patients with SLE ranges between 17% and 71% (Fig 10-1). Despite this wide range, it is clear that major depressive disorder is more prevalent among patients with SLE than in the general population,^30,31 and suicide is more prevalent in this group as well.³² Depression is associated with poorer quality of life in patients with SLE—perhaps even more strongly than for indicators of disease activity³³—poorer cognitive functioning,³⁴ and medication nonadherence.³⁵

Some studies (Box 10-1) have found an association with depression severity and SLE disease activity^36,37 as well; however, this finding has not shown consistency across all populations.³⁸ Among patients with SLE, higher depression severity has additionally been associated with pain, arthritis, and prednisone doses greater than 7.5 mg.³¹

PATHOPHYSIOLOGY

The American College of Rheumatology has formally defined 19 syndromes affecting the central and peripheral nervous systems that fall under the heading of “neuropsychiatric lupus,” or NPSLE (Box 10-2). These are listed in Table 10-2. The five psychiatric NPSLE syndromes include delirium, anxiety disorder, mood disorder, cognitive dysfunction, and psychosis. Distinguishing between NPSLE (a.k.a. a psychiatric disorder “due to a general medical condition”) and other depressive disorders that occur in the context of SLE has proved notoriously difficult,³⁹ particularly given that neuropsychiatric lupus can by definition occur at any time during the course of SLE or even predate its onset.³⁹ In truth, trying to make this distinction might mislead us in the same way that trying to understand depression as either a purely organic or purely psychological phenomenon misleads us. Several neurobiological mechanisms for NPSLE have been proposed; four are described below.

CEREBROVASCULAR DYSFUNCTION

SLE is highly associated with vascular disease,⁴⁰ and both inflammation and thrombotic phenomena contribute to its vascular manifestations.⁴¹ It has been proposed that cerebrovascular occlusion and/or hemorrhage may underlie or contribute to NPSLE,⁴² although rigorous studies to investigate this mechanism have not been conducted. Cerebrovascular pathology consistent with this model has been demonstrated in postmortem studies of patients with SLE,⁴³ and a difference in cerebral perfusion between SLE patients and controls has been observed with magnetic resonance imaging (MRI).⁴⁴

ANTIBODY-MEDIATED

Higher levels of circulating serum antibodies to the human N-methyl-D-aspartate (NMDA) NR2 receptor have been associated with increased depression as well as hypomania and some cognitive measures⁴⁵ in SLE patients, although in at least one study this association was found only with CSF NR2 antibodies and not with serum antibodies.⁴⁶ Anti-ribosomal P (anti-P) antibody has also been implicated in the pathogenesis of NPSLE, but the association between serum anti-P levels was not supported by a meta-analysis of the existing literature.⁴⁷ Thus the role of antibodies in NPSLE remains to be further clarified.

INFLAMMATORY CYTOKINES

Consistent with studies of inflammatory markers in non-SLE depression, significantly higher levels of CSF IL-6 are found in patients with NPSLE compared to other patients with SLE and to healthy controls.⁴⁸ Murine research has implicated a TNF family ligand (TWEAK) as another potential mediator between SLE, NPSLE, and depression.⁴⁹

BLOOD–BRAIN BARRIER DISRUPTION

There is an absence of definitive studies on how closely blood–brain barrier (BBB) disruption and NPSLE are linked, although this association has been supported by case reports.⁵⁰ The difficulties in measuring BBB disruption probably contribute to this. The degree of BBB disruption in patients with SLE is typically measured indirectly, by the presence or quantity of proteins (e.g., albumin or immunoglobulins) or serum molecules in the CSF. These markers require lumbar puncture to obtain, are often transient, and can be reduced or reversed by treatment with corticosteroids.

CLINICAL PRESENTATION, COURSE, AND NATURAL HISTORY

Patients with SLE commonly present with symptoms that overlap with DSM IV criteria for major depressive disorder, including fatigue, psychomotor retardation, and sleep disturbance (Box 10-3).⁵¹ Cognitive problems are commonly reported as well.³⁴ Therefore, the somatic and cognitive symptoms of SLE may lead to false positive symptoms of depression screening tools.⁵¹

ASSESSMENT AND DIFFERENTIAL DIAGNOSIS

When depression is identified in the context of SLE, assessment of cognitive functioning, energy, and sleep should be part of the initial workup (Box 10-4 and Box 10-5). NPSLE is difficult to rule out, and therefore a rheumatologist should be consulted to assist in investigating other disease parameters if this is possible. Because further treatment might include use of GC, immunomodulators, antimalarial agents, and/or anticoagulants depending upon the specific NPSLE syndrome, consideration to the effects these treatments may have on neuropsychiatric symptoms should be given as well.

TREATMENT

Controlled studies establishing the optimal use of antidepressants in patients with depression and SLE are lacking, although clinical experience suggests that the SSRIs are the first-line agents of choice (Box 10-6 to Box 10-8). However, there are case reports of an association between bupropion⁵² and sertraline⁵³ with cutaneous and systemic lupus symptoms, respectively. There are reports of successful treatment of NPSLE depression with ECT.⁵⁴

There are currently no studies establishing the efficacy of psychotherapy in treating major depressive disorder when it coexists with SLE. CBT has, however, been shown to improve depressive symptoms, stress, and other quality-of-life measures in patients with SLE.⁵⁵ Negative cognitions related to physical appearance are commonly observed and are predictive of depression among SLE patients and therefore might be anticipated in psychotherapy.⁶

Rheumatoid arthritis is an idiopathic, chronic, multisystem disease that occurs with a prevalence of approximately 0.5% to 1.0%.⁵⁶ It affects women three times more often than men. RA presents with symmetrical, peripheral, inflammatory synovitis. Patients with RA additionally have laboratory evidence of both autoantibodies and systemic inflammation.⁵⁷ Destruction and erosion of cartilage and bone from the synovitis is the hallmark of advanced disease. These destructive changes can compromise functional status and lead to loss of valued activities, including employment, recreational activities, and higher levels of independence.