Fatigue beyond inflammation in inflammatory rheumatic diseases: a narrative review and treatable-traits framework

Abstract

Fatigue frequently persists after clinically visible inflammation has improved in inflammatory rheumatic diseases (IRDs). This narrative review integrates evidence updated through 21 July 2026 on mechanisms, assessment, and management of fatigue in rheumatoid arthritis (RA), systemic lupus erythematosus (SLE), and Sjögren’s disease (SjD). Across these conditions, associations between fatigue and conventional inflammatory measures are inconsistent and usually weaker than associations with pain, sleep disturbance, mood symptoms, disability, central pain sensitivity, comorbidity, medication effects, and reduced activity tolerance. Recent quantitative findings reinforce this mismatch. In a 2026 RA cohort of 253 patients, 80% reported fatigue and 10% had severe fatigue; fatigue remained present in some patients in remission. In an international survey of 1,155 people with SjD, physical fatigue was among the most frequent symptoms, while mean work and activity impairment reached 46.6% and 48.4%, respectively. Longitudinal RA data showed that improvement in self-reported central pain sensitivity tracked improvement in fatigue, whereas inflammatory markers and imaging were not consistently associated with fatigue. Recent SLE imaging findings also linked region-specific cerebellar alterations with fatigue and cognitive dysfunction, although no central or metabolic biomarker is ready for routine use. Immune-targeted therapies produce variable, generally small-to-moderate fatigue improvements, whereas personalized exercise, rehabilitation, cognitive-behavioural or self-management interventions, and treatment of sleep disorders have supportive clinical evidence. The mismatch between inflammatory remission and continuing fatigue is framed as a fatigue remission gap. This is a practical signal, not a validated diagnosis, and should prompt direct fatigue measurement and a treatable-traits assessment rather than automatic escalation of immunosuppression. The proposed framework sequentially evaluates disease activity, fatigue severity, pain and fibromyalgia overlap, sleep, mood, comorbidities, medication effects, and deconditioning, then selects a shared, modifiable treatment target for reassessment.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00296-026-06266-2.

Introduction

Fatigue is one of the most prevalent and disabling symptoms reported by patients with inflammatory rheumatic diseases (IRDs). In rheumatoid arthritis (RA), systemic lupus erythematosus (SLE), Sjögren’s disease (SjD; historically referred to as Sjögren’s syndrome) and related inflammatory arthritides, fatigue affects physical activity, cognitive performance, social participation, work capacity and quality of life [1–4]. Patients frequently describe fatigue as qualitatively different from ordinary tiredness: it may be unpredictable, disproportionate to exertion and only partially relieved by rest.

The traditional explanatory model has linked fatigue to systemic inflammation. This model is biologically plausible because cytokines can induce sickness behaviour, alter neurotransmitter function and reduce motivation during acute immune activation [5–8]. However, routine clinical experience and longitudinal research show that fatigue often persists despite normalized C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), low composite disease activity scores or clinical remission [2–4, 13, 15, 16]. Conversely, some patients with objectively active disease report only modest fatigue. This imperfect coupling between inflammation and symptom burden is a major reason why fatigue remains an unmet need in rheumatology.

The clinical importance of fatigue extends beyond symptom reporting. Fatigue contributes to impaired physical health, reduced work participation, disability, psychological distress and patient dissatisfaction with care [2–4]. It also influences how patients interpret disease activity and treatment success. In SLE, for example, fatigue may mediate a substantial proportion of the effect of flares and depression on physical health impairment [40, 41]. Therefore, fatigue is not merely a secondary complaint but a central patient-reported outcome (PRO) that requires direct assessment and targeted management.

Several studies illustrate individual components of this framework. Davies et al. showed that severe fatigue in primary Sjögren’s syndrome was associated with lower rather than higher levels of selected proinflammatory cytokines, directly challenging a simple inflammation-fatigue model [67]. In RA, Løppenthin et al. linked physical activity with fatigue and sleep [68], while Walter et al. demonstrated that fatigue remained frequent and persistent during early intensive, tightly controlled treatment [69]. Harrold et al. found that coexisting Sjögren’s syndrome attenuated improvement in pain and fatigue despite comparable improvement in RA disease activity [70], and van Lint et al. identified patient-specific fatigue patterns around biologic injections [71]. More recent RHEI studies describe adaptation to systemic autoimmune rheumatic disease as a non-linear process [72], show that baseline fatigue and sleep quality carry prognostic information in inflammatory arthritis [73], document the clinical and relational burden of fibromyalgia overlap [74], and the under-recognition of depressive and anxiety symptoms without validated screening instruments [77]. Together, these findings support the need for multidimensional fatigue assessment but do not themselves validate the framework proposed in this review.

This narrative review aims to integrate current evidence on the mechanisms, assessment and management of fatigue in IRDs, with emphasis on clinical practice. The review focuses on rheumatology populations while using selected mechanistic evidence from experimental inflammation and non-rheumatic immune-mediated disorders only when it helps explain pathways relevant to IRDs. Its practical contribution is a treatable-traits framework that links fatigue phenotypes to assessment and management choices rather than treating fatigue as a single undifferentiated outcome.

Methods

This article is a narrative review designed to integrate mechanistic and clinical evidence on fatigue in inflammatory rheumatic diseases. The search and synthesis followed recommendations for transparent narrative biomedical reviews and comprehensive literature searching [65, 66]. Searches were conducted in PubMed, Scopus, Web of Science Core Collection, Embase, the Cochrane Library. Directory of Open Access Journals (DOAJ) was searched as a supplementary source of open-access literature. The final update covered records indexed or published through 21 July 2026. Database-specific search strings and the date of the final search are provided in Online Resource 1.

Searches combined controlled vocabulary, where available, with free-text terms for fatigue and IRDs. Core concepts included fatigue, central fatigue, patient-reported fatigue, inflammatory rheumatic disease, rheumatoid arthritis, systemic lupus erythematosus, Sjögren, sicca, cytokines, type I interferon, neuroinflammation, microglia, mitochondrial dysfunction, immunometabolism, pain, central sensitization, fibromyalgia, sleep, insomnia, obstructive sleep apnea, depression, anxiety, obesity, cardiovascular disease, anaemia, endocrine disorders, medication adverse effects, exercise, rehabilitation, pacing, self-management, and cognitive behavioural therapy. Disease-specific and intervention-specific searches were run separately and reference lists of relevant reviews, recommendations, and included articles were checked manually.

No lower publication-date restriction was applied because foundational mechanistic and patient-reported outcome literature remained relevant. English-language full texts were prioritized; non-English records with an English abstract were considered when directly relevant and when no equivalent English-language evidence was available. Eligible evidence included systematic reviews, randomized and non-randomized trials, prospective and retrospective cohorts, registries, clinical recommendations, validated patient-reported outcome studies, and mechanistic studies with clear relevance to rheumatology. Case reports, letters, conference abstracts, and non-rheumatology mechanistic studies were retained only when they clarified an otherwise underdeveloped pathway and were labelled as indirect evidence.

The author screened titles and abstracts, reviewed potentially relevant full texts, and selected evidence purposively for narrative synthesis. When findings conflicted, greater weight was given to rheumatology-specific populations, longitudinal or interventional designs, validated fatigue instruments, transparent adjustment for disease activity and comorbidities, larger samples, and clinically interpretable outcomes. Evidence from neurological disease, experimental inflammation, or myalgic encephalomyelitis/chronic fatigue syndrome was treated as mechanistic context rather than direct proof in RA, SLE, or SjD.

This review was not designed to provide pooled effect estimates, a PRISMA flow diagram, or a formal risk-of-bias assessment. Because the diseases, designs, interventions, and fatigue instruments were heterogeneous, findings were synthesized narratively.

Fatigue and inflammation: limitations of the classical model

Inflammation remains an important contributor to fatigue. IL-1beta, IL-6, TNF-alpha and type I interferon pathways can trigger sickness behaviour characterized by reduced activity, impaired concentration, anhedonia, sleep changes and increased perceived effort [5–8, 17–20]. In chronic rheumatic disease, however, the same pathways can become maladaptive when they interact with pain, insomnia, psychological distress and reduced physical capacity. However, conventional measures of inflammatory activity do not consistently explain fatigue severity. In RA, fatigue correlates only modestly with ESR, CRP, and disease activity indices, and systematic reviews indicate that pain and patient global assessment often explain more variance in fatigue than objective inflammatory markers [2, 5, 12]. A 2026 cross-sectional cohort of 253 patients found fatigue in 80% and severe fatigue in 10%; FACIT-F scores worsened across CDAI categories, but fatigue was still present among patients in remission [78]. Similar patterns occur in SLE, where fatigue is frequently more closely associated with depression, anxiety, pain, and sleep disturbance than with inflammatory disease activity [3, 34, 38, 41, 42].

In early RA, improvements in fatigue can lag behind disease remission, and residual symptoms may remain clinically important even when inflammatory targets are achieved [13, 15, 69]. In the CAP-RA study, self-reported central pain sensitivity was strongly associated with fatigue at baseline and three months, and reduction in central pain sensitivity was associated with reduction in fatigue; quantitative sensory tests and inflammatory markers or imaging were not consistently related to fatigue [80]. Data from the German National Database also showed that, despite substantial improvements in inflammatory outcomes between 2007 and 2023, mean fatigue trajectories did not improve in parallel for many patients with IRDs [16]. Thus, achieving inflammatory treatment targets may not resolve fatigue in all patients.

Biologic and targeted synthetic disease-modifying antirheumatic drugs (DMARDs) can reduce fatigue, but their effects are generally small to moderate and may be mediated partly by improvements in pain and inflammation rather than direct effects on fatigue pathways [14]. Ecological momentary assessment data also suggest that some patients experience distinct fatigue fluctuations around biologic injection cycles, reinforcing the need to ask about timing, medication burden and symptom context rather than assuming a single inflammatory driver [71].

Physical fatigue, cognitive fatigue, emotional exhaustion, susceptibility to fatigue and post-activity worsening may have different dominant drivers. A patient with active synovitis, anaemia and nocturnal pain requires a different approach from a patient in remission whose fatigue is sustained by insomnia, depression and deconditioning. These differences support a multidimensional clinical assessment rather than a single-mechanism explanation (Table 1) [2–5].

Table 1.

From inflammation-centred assumptions to treatable traits in persistent fatigue

Clinical problem	Why inflammation alone is insufficient	Treatable traits to screen	Practical implication
Normal CRP/ESR but severe fatigue	Fatigue often correlates more strongly with pain, patient global assessment, sleep and mood than with CRP/ESR	Pain phenotype, fibromyalgia overlap, insomnia/sleep apnea risk, depression/anxiety, medication effects	Do not dismiss fatigue because inflammatory markers are normal; measure fatigue directly and screen drivers
Fatigue persisting in remission	Fatigue can lag behind disease control and may persist despite treat-to-target improvement	Residual pain, sleep disturbance, deconditioning, obesity, anaemia, endocrine disease, cardiovascular disease	Treat active disease, then avoid reflexive immunosuppression escalation when objective inflammation is controlled
Disease-specific mechanisms	Shared fatigue pathways coexist with disease-specific contributors such as type I interferon in SLE and sicca-related sleep disruption in SjD	SLE interferon/organ activity, RA chronic pain/nociplastic mechanisms, SjD dryness, dysautonomia-like symptoms and pain	Disease-specific assessment should accompany the cross-cutting treatable-traits screen
Variable response to biologics/tsDMARDs	Immune-targeted therapy improves fatigue inconsistently and may not address pain, sleep, mood or activity tolerance	Medication timing, post-dose fatigue, glucocorticoid sleep effects, sedating analgesics, polypharmacy	Discuss realistic fatigue expectations and review drugs as potential contributors
Multidimensional fatigue	Physical, cognitive, emotional and post-activity fatigue may be maintained by different mechanisms	Activity pattern, pacing, fear of harm, coping, work demands and cognitive symptoms	Select an initial target with the patient and reassess the phenotype over time

Mechanistic pathways of fatigue

Cytokine signalling and sickness behaviour

Cytokine signalling is one of the best-characterized biological pathways linking inflammation with fatigue. Peripheral cytokines can influence the brain via humoral pathways, endothelial activation, circumventricular organs and afferent neural routes such as the vagus nerve [5–8, 17–20]. These signals alter arousal, reward, motivation, attention and effort valuation. Clinically, this helps explain fatigue accompanying inflammatory flares, infection-like symptoms and interferon-dominant states.

IL-1beta is particularly relevant to central fatigue because it can be produced within the central nervous system during peripheral inflammation and is closely linked to fever, sickness behaviour and reduced activity [17–19]. IL-6 may connect peripheral inflammatory activity, stress responses and sleep regulation [20]. TNF-alpha and type I interferon pathways can modify monoaminergic and glutamatergic neurotransmission, contributing to low motivation, cognitive slowing and altered mood [5–7].

At the neurotransmitter level, cytokines can reduce release of serotonin, dopamine and noradrenaline, increase monoamine uptake and degradation, activate indoleamine-2,3-dioxygenase and redirect tryptophan metabolism toward kynurenine metabolites. Cytokine-driven oxidative stress may also impair tetrahydrobiopterin-dependent monoamine synthesis and alter glutamatergic signalling [5–7, 21]. These pathways explain why fatigue, depression, pain sensitivity and cognitive symptoms often cluster together.

Neuroimmune mechanisms and central fatigue

Central fatigue refers to reduced capacity or willingness to initiate and sustain mental or physical activity because of altered central nervous system processing. In IRDs, central fatigue may arise from repeated peripheral immune activation, microglial priming, altered astrocyte function, disrupted neurotransmission and changes in functional brain networks [6, 22–24]. These mechanisms are most clinically relevant when patients describe cognitive fatigue, poor concentration or disproportionate effort during routine tasks.

Microglial activation can amplify neuroimmune signalling by releasing cytokines, chemokines, and reactive oxygen species, potentially affecting synaptic function, neuronal excitability, and network efficiency. Recent SLE imaging adds disease-specific support to a central component: in 72 women with SLE and 25 matched healthy individuals, region-specific cerebellar volume alterations, particularly involving lobule VIIB, were associated with cognitive dysfunction and fatigue [81]. This observation is hypothesis-generating rather than diagnostic. Overall, the evidence remains uneven and much of the broader mechanistic literature comes from experimental inflammation, chronic fatigue, or neurological disease models. Central mechanisms therefore help explain why serum markers may not reflect symptom generation, but they do not yet provide a routine clinical test for central fatigue [6, 22–24].

Central mechanisms also help explain why fatigue can respond to interventions that do not directly suppress inflammation, such as exercise, sleep treatment and cognitive-behavioural approaches. These interventions may improve autonomic regulation, sleep architecture, mood, coping, activity pacing and perceived exertion.

Mitochondrial dysfunction and immunometabolism

Mitochondria regulate ATP production, redox homeostasis, apoptosis, innate immune signalling and immunometabolic programming. Mitochondrial dysfunction may therefore contribute to fatigue both by impairing cellular energy metabolism and by amplifying inflammatory pathways [25–31]. In rheumatic disease, the key issue is not simply ‘low energy’, but a self-reinforcing interaction between immune activation, oxidative stress, damage-associated molecular patterns and inefficient energy handling.

In RA, mitochondrial alterations in immune and synovial cells may act as amplifiers of inflammation. Reactive oxygen species and metabolic intermediates such as succinate can activate the NLRP3 inflammasome and promote IL-1beta and IL-18 production [26, 27]. In SLE, mitochondrial dysfunction is linked to oxidative stress, neutrophil extracellular traps, mitochondrial DNA release and type I interferon activation [28–31]. These disease-specific pathways are biologically plausible explanations for persistent fatigue, but they should be interpreted alongside pain, sleep disturbance, mood symptoms and physical inactivity.

Evidence from multiple sclerosis and other neuroinflammatory diseases shows that mitochondrial failure in high-demand neural systems can produce fatigue through an energy supply–demand mismatch [61, 62]. Rheumatic diseases differ in tissue targets, inflammatory timing, treatment exposure and comorbidity patterns (Table 2)

Table 2.

Disease-specific and comparative mitochondrial/immunometabolic patterns relevant to fatigue

Condition	Dominant pathway	Inflammatory link	Fatigue relevance
RA	Metabolic reprogramming in synovial immune cells; reactive oxygen species and inflammatory metabolites	Inflammasome activation, IL-1beta/IL-18 signalling, pain and systemic symptoms	Chronic cytokine burden plus chronic pain and reduced physical capacity can increase perceived effort and reduce activity tolerance
SLE	Mitochondrial stress in immune cells, mitochondrial DNA/NET-associated damage-associated molecular pattern signalling, type I interferon activation	Oxidative stress, immune activation and organ/systemic symptom burden	Energetic inefficiency plus interferon-driven symptoms may contribute to severe persistent fatigue
Sjögren’s disease	Immune activation, interferon signatures, sicca-related symptom burden and possible autonomic dysfunction; direct mitochondrial data are less developed	Fatigue frequently coexists with dryness, pain, poor sleep and dysautonomia-like symptoms	Requires clinical phenotyping rather than reliance on glandular or serological measures alone
Comparative multiple sclerosis evidence	Axonal mitochondrial injury and respiratory-chain dysfunction in neuroinflammatory disease	Energy supply–demand mismatch in demyelinated axons	Useful mechanistic comparator, but should be cited cautiously outside rheumatology scope

Physical deconditioning and activity dysregulation

Physical deconditioning is a major amplifier of fatigue but should not be treated as the sole cause. Pain, flare-related rest, fear of damage, low mood, glucocorticoid effects and reduced confidence can lead to reduced activity. In turn, inactivity reduces aerobic capacity, muscle strength and metabolic efficiency, making everyday tasks more effortful and reinforcing the fatigue-inactivity cycle [2, 32, 33, 68].

Behavioural patterns such as overactivity followed by symptom worsening (‘boom and bust’) may also maintain fatigue. Some patients push through fatigue on better days and then require prolonged recovery, whereas others avoid activity because of fear of pain or flare. Both patterns can reduce confidence and worsen disability. Rehabilitation therefore needs to be gradual, individualized and linked to education about pacing, energy conservation and safe physical activity [9, 33, 51–56].

Pain, sleep, mood and behavioural factors

Psychological and behavioural factors are among the strongest and most consistent correlates of fatigue in IRDs. Depression, anxiety, stress, coping style, illness beliefs, and self-efficacy can influence fatigue directly and indirectly through pain, sleep, and activity patterns [34–39, 59, 76]. These associations do not imply that fatigue is ‘psychological’; they reflect bidirectional brain-body interactions. Their clinical importance is reinforced by a 2026 RHEI study of 255 people with inflammatory arthritis: validated PHQ-9 and GAD-7 screening identified substantially more depressive and anxiety symptoms than routine clinical records, including depressive symptoms in 24.8% and anxiety symptoms in 20.0% of the RA subgroup [77].

In SLE, fatigue is often more strongly associated with depression, anxiety and pain than with inflammatory disease activity [34, 38, 41, 42]. In RA, path analysis and biopsychosocial models indicate that depression, sleep disturbance, disability and pain explain a substantial portion of fatigue variance [5, 36, 37, 39]. Across immune-mediated inflammatory disease populations, anxiety and depression may remain significant predictors even after accounting for disease activity [35, 76].

Sleep disturbance is both a cause and consequence of fatigue. Pain, nocturnal stiffness, sicca symptoms, mood symptoms, medication effects and inflammatory activity can disrupt sleep, while poor sleep worsens pain sensitivity, cognitive performance and perceived exertion [57, 58, 60, 73, 75]. Sleep should therefore be assessed systematically rather than treated as a secondary issue.

Disease-specific contributors and common comorbidities

In RA, chronic synovitis, pain sensitization, functional disability, fibromyalgia overlap, and medication-related fatigue may dominate even when objective inflammation is low. Coexisting Sjögren’s syndrome in RA can reduce improvement in pain and fatigue despite similar improvement in clinical disease activity [70], while recent CAP-RA data support central pain sensitivity as a fatigue driver beyond pain intensity and inflammation [80]. In SLE, type I interferon activation, anaemia, renal involvement, glucocorticoid exposure, depression, sleep disturbance, and possible central network changes may interact to sustain fatigue [28–31, 34, 38, 40–42, 81]. In SjD, fatigue is commonly linked to dryness-related sleep disruption, pain, autonomic symptoms, mood symptoms, and poor activity tolerance; glandular or serological measures alone may underestimate burden [67, 75]. In a 2026 international survey of 1,155 people with SjD, physical fatigue was among the most frequently reported symptoms, and mean work and activity impairment were 46.6% and 48.4%, respectively [79].

Common comorbidities should be assessed because they can mimic or amplify rheumatic fatigue. Obesity can increase inflammation, pain, sleep apnea risk and exertional intolerance. Obstructive sleep apnea causes non-restorative sleep and daytime sleepiness and is easily missed if fatigue is attributed to rheumatic disease. Cardiovascular disease and reduced cardiorespiratory fitness reduce activity tolerance. Anaemia, thyroid disease, diabetes, renal or hepatic dysfunction, vitamin deficiency, chronic infection and medication adverse effects are common, modifiable contributors that should be addressed before concluding that fatigue reflects occult inflammation (Table 3) [2, 3, 32, 50].

Table 3.

Main mechanisms contributing to fatigue and their clinical translation

Mechanism	Clinical clues	Assessment focus	Management implication
Cytokine signalling	Flare-related malaise, low motivation, feverish symptoms, cognitive slowing	Disease activity, CRP/ESR, organ activity, infection	Control active inflammation, then reassess residual fatigue separately
Central neuroimmune mechanisms	Cognitive fatigue, disproportionate effort, poor concentration	Sleep, mood, pain, autonomic symptoms, cognitive impact	Consider sleep interventions, exercise, CBT-based approaches and symptom phenotyping
Mitochondrial/immunometabolic dysfunction	Low endurance, post-activity worsening, systemic exhaustion	Disease-specific inflammation, physical capacity, comorbidities	Combine disease control with graded rehabilitation and pacing
Pain and nociplastic mechanisms	Widespread pain, tender points, poor recovery, high symptom burden	Pain phenotype, fibromyalgia overlap, medication burden	Treat pain mechanisms; avoid reflexive immunosuppression when inflammation is absent
Sleep disturbance	Non-restorative sleep, morning fatigue, daytime sleepiness, snoring or nocturnal pain/dryness	Insomnia, sleep apnea risk, sicca symptoms, medication timing	CBT-I, sleep hygiene, sleep disorder referral, pain/sicca management
Mood, coping and behaviour	Persistent fatigue with low mood, anxiety, stress, low self-efficacy or avoidance	PHQ-9/GAD-7 or equivalent; coping, work and social stressors	Psychological support, CBT, self-management and shared goal setting
Deconditioning/activity dysregulation	Effort intolerance, inactivity, boom-and-bust activity pattern	Physical capacity, daily activity pattern, fear of harm	Tailored exercise, rehabilitation, pacing and education

Integrated clinical model and fatigue phenotyping

The evidence supports an integrated model in which inflammation may initiate fatigue but is rarely the only maintaining factor. Acute cytokine signalling can generate sickness behaviour. Recurrent immune activation may sensitize central neuroimmune pathways. Mitochondrial and immunometabolic changes may amplify inflammatory signalling and reduce energy efficiency. Pain, sleep disruption, mood symptoms and deconditioning then interact with these biological pathways to sustain fatigue [2–7, 63].

For clinical practice, the most useful question is not whether fatigue is inflammatory or non-inflammatory, but which contributors are dominant and modifiable in a given patient. A pragmatic phenotype-based approach can distinguish: (1) inflammation-dominant fatigue, (2) pain/sleep-dominant fatigue, (3) mood/coping-dominant fatigue, (4) deconditioning/activity-dysregulation fatigue and (5) complex overlapping fatigue with fibromyalgia or multimorbidity. These categories are not mutually exclusive and should be revisited over time.

These phenotypes are proposed as a pragmatic clinical heuristic rather than validated biological subtypes. They were derived by synthesizing existing fatigue literature, PRO studies, treatment trials, treatable-traits concepts and clinical patterns described across RA, SLE, SjD and broader systemic autoimmune rheumatic diseases [2–5, 10, 11, 34–42, 50, 64, 67–76]. They should therefore be used to structure assessment and communication, not to replace disease-specific diagnosis or evidence-based treatment. Prospective validation is needed to test whether phenotype-guided fatigue care improves outcomes compared with usual care.

This phenotype-based model also improves communication. Patients often experience frustration when fatigue is dismissed because inflammatory markers are normal. Explaining fatigue as a legitimate multidimensional symptom can validate the patient experience while guiding targeted management. The suggested clinical phenotypes are summarized in Table 4.

Table 4.

Practical clinical phenotypes of fatigue in inflammatory rheumatic diseases

Phenotype	Typical clues	Dominant assessment questions	Management focus
Inflammation-dominant	Fatigue fluctuates with flare, feverish malaise, raised CRP/ESR or active joints/organs	Is there active inflammatory disease, infection, anaemia or poor medication adherence?	Optimize DMARD/biologic strategy; reassess fatigue after control
Pain/sleep-dominant	Non-restorative sleep, nocturnal pain/dryness, morning exhaustion, daytime sleepiness	What is the pain phenotype? Is insomnia, OSA risk or sicca-related awakening present?	Pain management, CBT-I, sleep assessment, physiotherapy and sicca management
Mood/coping-dominant	Persistent fatigue with low mood, anxiety, stress, low self-efficacy or illness-related distress	Are depression, anxiety, stress, maladaptive coping or work/social stressors present?	Psychological support, CBT/self-management, treatment of depression/anxiety
Deconditioning/activity dysregulation	Effort intolerance, avoidance or boom-and-bust activity pattern, low confidence with activity	What is physical capacity, activity pattern and fear of harm?	Tailored exercise, pacing, rehabilitation and education
Complex overlap/fibromyalgia	Severe fatigue with widespread pain, cognitive symptoms, autonomic/sensory complaints or multimorbidity	Is fibromyalgia, medication burden or comorbidity driving symptoms?	Multimodal plan; avoid immunosuppression escalation without evidence of inflammation

Clinical implications

The 2023 EULAR recommendations for fatigue management in inflammatory rheumatic and musculoskeletal diseases provide a useful clinical anchor for this review. They emphasize that fatigue should be monitored, discussed and managed as part of routine care, with shared decisions that take account of disease activity, comorbidities, psychological and social factors, and patient preferences [50]. This guidance supports a practical distinction between fatigue that improves with control of active inflammation and fatigue that persists despite low disease activity, where non-inflammatory contributors need direct assessment.

Routine fatigue assessment

Fatigue should be assessed directly and longitudinally. A single question such as ‘Are you tired?’ is insufficient because it does not capture severity, impact, cognitive dimensions, susceptibility to fatigue or coping. At the same time, assessment must be feasible in busy rheumatology clinics. A tiered approach is useful: a short numerical rating scale or visual analogue scale can be used for screening and follow-up, while multidimensional instruments can be used for complex or persistent fatigue [43–49].

PROMIS Fatigue offers standardized T-score reporting and cross-disease comparison and has been validated in RA and SLE [43, 44]. FACIT-F remains widely used in clinical trials and observational studies. Disease-specific tools can add depth: BRAF instruments are RA-specific, FATIGUE-PRO was developed for SLE, and FAS provides a brief option for SLE or general clinical use [45–49]. These instruments are not interchangeable. They differ in recall period, scoring, dimensionality and sensitivity to change, which complicates comparison across trials and may partly explain inconsistent estimates of treatment effects on fatigue (Table 5).

Table 5.

Fatigue instruments relevant to rheumatology practice and research

Instrument	Main use/context	Strengths	Limitations
Visual analogue/numerical rating scale fatigue	General IRD clinic use	Very brief; useful for repeated monitoring	Limited dimensionality; cannot identify drivers
FACIT-F	RA, SLE and chronic illness studies	Widely used and interpretable in trials	Not rheumatology-specific; less detailed on cognitive fatigue
PROMIS Fatigue	RA, SLE and cross-disease assessment	Standardized T-score reporting; allows cross-condition comparison	Requires familiarity with PROMIS scoring or CAT platform
FATIGUE-PRO	SLE	SLE-specific; captures physical, mental/cognitive and susceptibility domains	Longer; less practical for every visit
BRAF-NRS/BRAF-MDQ	RA	RA-specific; captures severity, impact, coping and burden	Less generalizable outside RA
CAT Fatigue RA	RA	High measurement precision through adaptive testing	Requires electronic implementation
FAS	SLE/general fatigue	Simple and reliable; low respondent burden	Less multidimensional

From fatigue score to treatable traits

A more actionable way to use fatigue assessment is to translate the score into treatable traits. The concept originated in chronic airway disease as a precision-medicine strategy linking multidimensional assessment to targeted management; here it is adapted pragmatically to fatigue in IRDs [64]. In clinic this can be framed as a fatigue vital sign followed by a brief driver screen: active inflammation, pain or fibromyalgia overlap, sleep disturbance, mood and coping, anaemia or endocrine/metabolic causes, medication burden, cardiovascular risk, obesity, sleep apnea risk and activity tolerance [2, 3, 50]. The purpose is to choose the first modifiable target with the patient and define a review interval.

This approach also guards against two common errors: escalating immunosuppression in a patient whose disease is controlled, or psychologising fatigue without checking inflammatory activity and medical causes. The same patient may move between treatable traits over time, so reassessment is part of management rather than an administrative exercise.

Clinical evaluation of persistent fatigue

Persistent fatigue should trigger structured evaluation rather than automatic escalation of immunosuppression. Clinicians should first check inflammatory activity with the appropriate disease-specific index, examination, organ assessment and routine inflammatory markers. If disease activity is low, the evaluation should shift toward common modifiable contributors: pain or fibromyalgia overlap, sleep disturbance, depression, anxiety, anaemia, thyroid disease, diabetes, vitamin deficiency, renal or hepatic dysfunction, infection, glucocorticoid exposure, medication adverse effects, reduced physical capacity, obesity, cardiovascular disease and barriers to activity.

In practice, the fatigue score should start a clinical conversation rather than be treated as a diagnosis. The priority is to identify the first realistic treatment target and to agree when fatigue, function and patient goals will be reviewed.

Pharmacological implications

Pharmacological therapy remains essential for controlling inflammatory disease activity and preventing organ damage. However, the effect of immune-targeted therapy on fatigue is variable and should be discussed realistically with patients. In active disease, improved inflammatory control may reduce fatigue, pain and disability. In remission or low disease activity, persistent fatigue should not be used alone as evidence of occult inflammation without corroborating clinical findings [13–15].

Medication review is also important. Glucocorticoids may disturb sleep and mood, while some patients report post-dose or post-methotrexate fatigue. Analgesics, sedating medications and polypharmacy may contribute to daytime sleepiness or reduced activity. Treatable comorbidities such as anaemia, hypothyroidism and sleep apnoea should be addressed because they can mimic or amplify rheumatic fatigue.

Exercise, rehabilitation and activity pacing

Exercise is one of the most consistently supported non-pharmacological interventions for fatigue in IRDs. The 2022 American College of Rheumatology guideline conditionally or strongly supports physical activity, aerobic exercise, resistance training, aquatic exercise and mind–body approaches in RA according to patient context and preferences [51]. Systematic reviews and trials indicate that combined aerobic and strengthening interventions can improve fatigue, physical function and quality of life in RA [52, 53, 68]. Evidence in SLE is summarized separately below [54].

The LIFT trial demonstrated that remotely delivered personalized exercise and cognitive-behavioural interventions can reduce fatigue severity and impact across inflammatory rheumatic diseases [10]. A recent randomized study of high-intensity exercise in RA reported sustained improvements in multidimensional fatigue and health-related quality of life [53]. In SLE, exercise appears to reduce fatigue modestly, although certainty of evidence is lower and programmes must be adapted to disease activity, organ involvement and comorbidities [54].

The largest barrier is implementation. Patients may fear joint damage, flares or post-exertional worsening, and many do not receive a clear clinician recommendation for physical activity [55, 56]. Rehabilitation should include education, goal setting, pacing, graded progression and symptom-contingent adjustment. For severe fatigue, starting below the threshold of exacerbation may be safer than prescribing standard exercise targets immediately.

Psychological interventions and self-management

Cognitive-behavioural and self-management interventions are clinically important because fatigue is shaped by coping, self-efficacy, mood, sleep and activity patterns. The RAFT trial showed that cognitive-behavioural approaches delivered by rheumatology teams can reduce the impact of RA fatigue with benefits persisting at two years [11]. The LIFT trial extended the relevance of remotely delivered cognitive-behavioural and exercise interventions across inflammatory rheumatic diseases [10].

Psychological support should not be framed as implying that fatigue is imaginary. Rather, it targets modifiable mechanisms that influence fatigue intensity and impact: catastrophizing, avoidance, low self-efficacy, depressive rumination, anxiety-driven hypervigilance, poor sleep habits and maladaptive activity cycles. Screening for depression and anxiety is important because these factors are consistently associated with fatigue across RA, SLE and broader immune-mediated inflammatory populations [34–39, 59, 76].

Sleep management

Sleep disturbance should be evaluated routinely in patients with persistent fatigue. In RA, axial spondyloarthritis and psoriatic arthritis, sleep problems are common and interact with pain, mood and disease activity [60, 73]. Simple screening questions should address sleep duration, sleep quality, nocturnal pain, restless legs, snoring, witnessed apnoea, medication timing, caffeine/alcohol use, daytime sleepiness and sicca-related awakenings.

Cognitive behavioural therapy for insomnia (CBT-I) has a growing evidence base in chronic disease populations and in RA. The Sleep-RA trial showed that nurse-led group CBT-I improved insomnia and broader patient-reported outcomes in RA [57]. A recent systematic review and meta-analysis reported efficacy of CBT-I across chronic disease populations [58]. In rheumatology practice, referral pathways for CBT-I, digital CBT-I or trained nurse-delivered sleep interventions could become an important component of fatigue care.

A practical management algorithm

Table 6 translates the treatable-traits approach into a stepwise clinic algorithm. It is intended as a pragmatic framework, not as a formal guideline.

Table 6.

Treatable-traits algorithm for persistent fatigue in inflammatory rheumatic diseases

Step	Assessment	Action	Key point
1. Confirm disease status	Clinical examination, disease-specific index, CRP/ESR, organ assessment	Treat active inflammation according to guidelines	Do not attribute all fatigue to inflammation if disease is controlled
2. Measure fatigue directly	NRS/VAS, FACIT-F, PROMIS Fatigue, BRAF, FATIGUE-PRO or FAS	Record baseline and follow-up	Fatigue should be a measured outcome, not a vague complaint
3. Identify treatable traits	Pain/fibromyalgia, sleep, mood, medication effects, anaemia, thyroid disease, diabetes, obesity, CVD, OSA risk	Treat specific contributors; coordinate primary care or specialist input	Start with the first realistic modifiable driver
4. Phenotype fatigue pattern	Inflammatory, pain/sleep, mood/coping, deconditioning/activity dysregulation or complex overlap	Select first management target with the patient	Revisit phenotype over time
5. Multimodal plan	Exercise/rehabilitation, pacing, CBT/self-management, CBT-I, pain management	Set measurable goals and review at 8–12 weeks	Match intervention to phenotype and patient preference
6. Reassess and adapt	Fatigue score, function, work participation, adverse effects and patient goals	Escalate, de-escalate or redirect care according to response	Fatigue management is iterative

The fatigue remission gap: a fatigue-informed treat-to-target perspective

One way to make persistent fatigue more visible is to describe it as a fatigue remission gap: the mismatch between inflammatory remission and the patient’s remaining fatigue burden. This term is not intended to create a new diagnosis, biomarker category or separate guideline. It is a practical label for a common clinical situation in which conventional disease activity targets are met but the patient continues to report clinically important fatigue, non-restorative sleep, pain amplification or poor activity tolerance [13, 15, 16, 50, 69].

The proposed concept has not been prospectively validated. The gap may reflect multiple contributors, including residual inflammation not captured by routine measures, central sensitization, comorbidity, medication effects or social and occupational demands. For this reason, the term should prompt structured reassessment rather than imply that inflammation is absent or that symptoms are non-organic.

A fatigue-informed treat-to-target perspective would therefore ask two parallel questions: is inflammatory disease controlled, and which treatable traits are maintaining fatigue? Tracking both over time turns fatigue into a measurable outcome rather than a vague complaint. This approach provides a clear research agenda: define clinically meaningful fatigue remission, test phenotype-specific interventions and evaluate whether reducing residual fatigue improves work participation, adherence and quality of life (Fig. 1) [2, 3, 10, 11, 41, 50].

Fig. 1.

Graphical summary of the treatable-traits fatigue framework. Inflammatory, disease-specific, neuroimmune, metabolic, pain, sleep, mood, medication, comorbidity and activity pathways converge on persistent fatigue. The fatigue remission gap highlights residual fatigue after inflammatory control and leads to pragmatic phenotyping and treatable-traits management

Future directions

Future research should move beyond broad fatigue scores and toward more clinically meaningful, mechanism-informed phenotyping. Rather than treating fatigue as a single construct, studies should integrate patient-reported outcomes with disease activity, pain, sleep disturbance, mood, physical performance, actigraphy, and selected inflammatory or metabolic biomarkers. It is unlikely that one universal biomarker of fatigue will be identified. A more realistic aim is to define reproducible fatigue patterns that help predict which patients are most likely to benefit from intensified disease control, sleep-focused treatment, psychological support, rehabilitation, or multimodal care.

Future studies should also test whether phenotype-guided fatigue management leads to better patient-centred outcomes than conventional approaches. Trials should define clinically meaningful fatigue remission, examine how the fatigue remission gap varies across measurement instruments, and determine whether improvement in fatigue translates into better work participation, treatment adherence, daily functioning, and quality of life.

Implementation research is equally important. Fatigue interventions will have limited impact unless feasible in routine rheumatology services. Remote delivery, nurse-led programmes, digital fatigue assessment and integrated rehabilitation pathways deserve particular study, as do effects on work participation, adherence and patient satisfaction.

Mechanistic research should remain clinically anchored. Studies of cytokines, microglia, mitochondrial function and post-exertional malaise are useful only if linked to human phenotypes, PROs and treatment decisions. Evidence borrowed from multiple sclerosis, ME/CFS or post-viral syndromes can generate hypotheses for rheumatology, but should not be treated as direct proof in RA, SLE or SjD.

Conclusion

Fatigue in inflammatory rheumatic diseases cannot be reduced to a raised CRP, swollen joint count or disease activity score. Inflammation can clearly make people tired, especially during active disease, but the clinical problem is often what remains after inflammation improves. Pain, poor sleep, mood symptoms, medication effects, loss of physical capacity and central neuroimmune or metabolic changes may all sustain fatigue, with different weights in different patients.

There is still no simple biomarker, and available scales measure different aspects of fatigue rather than one biological process. This limitation is also a reminder that fatigue must be asked about directly. The practical task is to control active disease while looking for modifiable non-inflammatory contributors. Careful measurement, honest discussion and tailored combinations of exercise, rehabilitation, sleep treatment, psychological support and self-management are more likely to help than assuming that remission alone will solve the problem.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

BRAF

Bristol Rheumatoid Arthritis Fatigue scale

CAT

Computerized adaptive test

CBT

Cognitive behavioural therapy

CBT-I

Cognitive behavioural therapy for insomnia

CRP

C-reactive protein

DMARD

Disease-modifying antirheumatic drug

ESR

Erythrocyte sedimentation rate

FACIT-F

Functional Assessment of Chronic Illness Therapy-Fatigue

FAS

Fatigue Assessment Scale

IRD

Inflammatory rheumatic disease

ME/CFS

Myalgic encephalomyelitis/chronic fatigue syndrome

NRS

Numerical rating scale

OSA

Obstructive sleep apnea

PRO

Patient-reported outcome

PROMIS

Patient-Reported Outcomes Measurement Information System

RA

Rheumatoid arthritis

SjD

Sjögren’s disease

SLE

Systemic lupus erythematosus

tsDMARD

Targeted synthetic disease-modifying antirheumatic drug

VAS

Visual analogue scale

Author contributions

Anna Gwóźdź-Broczkowska conceptualized the review, designed and performed the literature search, selected and interpreted the evidence, drafted and critically revised the manuscript, and approved the final version. She confirms that she meets all four ICMJE authorship criteria. The author takes full responsibility for the integrity and accuracy of every aspect of this review and for resolving any questions related to its accuracy.

Funding

No funding was received for this work.

Data availability

Not applicable; no new data were generated or analysed for this article.

Declarations

Conflict of interest

The author declares no competing interests.

Ethical approval and consent to participate

Not applicable; this is a narrative review of published literature.

Ethical use of artificial intelligence

No artificial intelligence tools were used in the preparation, writing, editing, revision, or data analysis of this manuscript.