Late / elderly-onset rheumatoid arthritis — how it differs from young-onset disease, and the relevant investigations

Definitions and the age threshold. Late-onset rheumatoid arthritis (LORA), also called elderly-onset RA (EORA), is rheumatoid arthritis whose onset occurs at or after 60 years (the WHO definition of "elderly" being 60– 74), although many studies use 65 years; young-onset RA (YORA) is onset before this. There is no agreed international threshold — the Japan College of Rheumatology, which has issued the only national guideline on the subject, recommends the term LORA. Roughly one-third of all new RA now begins after the age of 60, and this proportion is rising as the population ages.

Epidemiology and sex ratio. YORA is strongly female-predominant, about 3:1, with peak onset between 25 and 45 years. In LORA the female preponderance is much less marked and approaches 1:1. The incidence of RA continues to rise with age into the seventh and eighth decades, so RA is not a disease that "burns out" at the door of old age.

Pathogenesis — why LORA is different. LORA arises on a background of immunosenescence and inflammaging: thymic involution, expansion of CD4+CD28− senescent T cells, shortened lymphocyte telomeres, and a high baseline IL-6 level. Importantly, the HLA-DRB1 shared-epitope association is weaker in LORA than in YORA — which is consistent with the higher proportion of seronegative disease and with the clinical overlap with polymyalgia rheumatica.

Mode of onset and joint pattern — the core clinical difference. YORA characteristically has an insidious onset over weeks to months, producing a symmetrical small-joint polyarthritis of the MCP, PIP, wrist and MTP joints, sparing the DIP joints. LORA tends to have an abrupt, almost "infection-like" acute onset, is more often oligoarticular at presentation, and shows more frequent involvement of large, proximal joints — especially the shoulders — producing a girdle distribution that closely mimics polymyalgia rheumatica. Pitting oedema of the hands and feet is characteristic of LORA and rare in YORA.

Systemic and constitutional features. Fever, weight loss, profound fatigue, anorexia and a markedly raised ESR at onset are all more prominent in LORA. This may relate to differences in cytokine profile and to the higher IL-6 tone of ageing. Classical severe extra-articular rheumatoid disease (vasculitis, Felty's syndrome, rheumatoid lung) is not more frequent in LORA — the systemic picture is one of constitutional inflammation rather than organ-based rheumatoid complications. Because the presentation is dominated by weight loss, weakness, immobility and low mood, LORA is frequently mislabelled as "failure to thrive" or as an occult malignancy.

Serology — and two crucial caveats. LORA is more often seronegative for rheumatoid factor and anti-CCP than YORA. Caveat one: rheumatoid factor positivity increases with age in the healthy population as an expression of age-related immune dysregulation, and therefore loses diagnostic specificity in older adults — anti-CCP is the more useful test in this age group. Caveat two: seronegative RA is not a mild form of the disease — it may carry more inflammatory activity and, in some series, a worse prognosis, and it requires the same intensive treat-to-target strategy as seropositive disease.

The three phenotypes of LORA, and the differential diagnosis. Three recognised subsets: (i) classical RA-like — usually seropositive, erosive, and behaves much like YORA; (ii) PMR-like LORA — girdle pain and stiffness with a very high ESR, usually seronegative, occurring in about 25% of LORA (PMR-like YORA is exceptional); (iii) RS3PE-like — remitting seronegative symmetrical synovitis with pitting oedema. The critical differential is therefore polymyalgia rheumatica: in follow-up studies more than 20% of patients initially diagnosed as PMR are reclassified as seronegative LORA, and some authorities regard PMR, seronegative LORA and RS3PE as parts of a single spectrum of inflammatory disease of old age. Other differentials at this age: polyarticular gout and calcium pyrophosphate disease, nodal and erosive osteoarthritis of the hands, paraneoplastic polyarthritis, viral arthritis, and immune checkpoint inhibitor-induced arthritis. Kane's comparison of osteoarthritis versus inflammatory arthritis is the bedside discriminator to quote: Clinical feature Osteoarthritis Inflammatory arthritis Duration of stiffness Minutes Hours Pain Usually with activity Occurs even at rest and at night Fatigue Unusual Common Swelling Common but little synovial reaction Very common, with synovial proliferation and thickening Erythema and warmth Unusual Common

Disease course and outcome. LORA presents with higher disease activity, more extra-articular manifestations, greater disability, more radiographic damage and a higher ESR at onset than YORA. Yet with adequate treatment the probability of achieving remission or low disease activity is similar to — and in some series better than — YORA, partly because of the higher proportion of seronegative patients. Where outcomes are worse in practice, the reason is largely under-treatment: older patients receive higher glucocorticoid and NSAID doses, fewer DMARDs and lower DMARD doses, and discontinue treatment more often for adverse events. The correct conclusion is that LORA is under-treated rather than intrinsically treatment-resistant.

Investigations to send — the second half of the question. (a) To confirm the diagnosis: rheumatoid factor and anti-CCP antibody (anti-CCP being the more specific in this age group); ANA; ESR and CRP; complete blood count (normocytic anaemia of chronic disease, thrombocytosis; leucopenia would suggest Felty's syndrome); plain radiographs of hands, wrists and feet for periarticular osteopenia, marginal erosions, joint space narrowing and deformity — remembering that coexisting osteoarthritic change is almost universal at this age and must not be over-interpreted; and musculoskeletal ultrasound with power Doppler, or MRI of hands and shoulders, which detects synovitis and erosions earlier and helps distinguish the synovitis of LORA from the bursitis and tenosynovitis pattern of PMR. Classification uses the ACR/EULAR 2010 criteria (score ≥6 of 10), which are less sensitive in seronegative large-joint LORA and therefore support but never replace clinical judgement. (b) To exclude the differential diagnoses: serum uric acid and, decisively, joint aspiration with polarised light microscopy and culture for crystals and sepsis; TSH; calcium, phosphate, alkaline phosphatase and 25(OH) vitamin D; serum and urine protein electrophoresis with free light chains to exclude myeloma; creatine kinase; chest radiograph; temporal artery ultrasound if there are any giant cell arteritis features; and age-appropriate malignancy screening where a paraneoplastic arthritis is suspected. (c) Baseline safety and pre-treatment work-up: renal function with eGFR, liver function tests, hepatitis B and C and HIV serology, latent tuberculosis screening with a chest radiograph and Mantoux or IGRA — mandatory in India before any biologic or JAK inhibitor, lipid profile with cardiovascular risk assessment, DEXA before or early in glucocorticoid therapy, and a baseline ophthalmic assessment before hydroxychloroquine.

How the differences change management. The treat-to-target principle and the target itself (remission or low disease activity) are the same as in YORA — age alone is never a reason to withhold a DMARD. Methotrexate remains anchor therapy, but must be dose-reduced for the age-related fall in GFR, given with folic acid, avoided if eGFR is below 30, and monitored closely for cytopenias, hepatotoxicity and pneumonitis. Minimise glucocorticoids — long-term low-dose steroid used instead of a DMARD is the single commonest and most damaging error in LORA (infection, osteoporosis, diabetes, cataract, skin fragility, delirium). Use NSAIDs sparingly: Kane's notes that older patients are particularly susceptible to gastrointestinal, CNS and renal adverse effects, and they are flagged in the Beers criteria. Biologics are as effective as in younger patients but carry a higher absolute risk of serious infection. Screen and vaccinate before immunosuppression (influenza, pneumococcal, recombinant zoster vaccine, COVID-19), and provide bone protection with any steroid course. Assess comorbidity, polypharmacy, cognition, frailty and adherence before choosing an agent — a comprehensive geriatric assessment is part of the rheumatology decision here.

A 2024 systematic review and meta-analysis (Acuña-Rocha et al., Cureus 2024, searching to November 2023) confirmed that EORA presents with more intense disease activity, a higher prevalence of extra-articular manifestations, greater disability and more pronounced radiographic change at onset, but that the probability of achieving remission or low disease activity over one to two years is similar to YORA. It also formally documented the treatment disparity — EORA patients receive higher glucocorticoid and NSAID doses and fewer or lower-dose DMARDs — establishing under-treatment, rather than biology, as the main driver of poorer real-world outcomes.

Guidelines and drug safety. The Japan College of Rheumatology has issued the first and currently only national guideline specific to LORA, recommending the term LORA and a treat-to-target approach explicitly adapted for comorbidity and frailty; dedicated ACR and EULAR guidance for LORA is still awaited and is recognised as an urgent gap (ACR Open Rheumatology, 2025). On drug safety, the ORAL Surveillance trial has led regulators and current EULAR recommendations to advise that JAK inhibitors should not be used first-line in patients over 65, or in those with cardiovascular risk factors, a smoking history or malignancy risk, because of increased major adverse cardiovascular events, venous thromboembolism, malignancy and herpes zoster — a restriction that falls disproportionately on exactly the LORA population.