Table 1 Frequency and distribution of clinical osteoarthritis locations

Knoop J et al. [11]

2011

United States

Progression Subcohort

842

OsteoArthritis Initiative database.

63.2 ± 69.1

Knee

Minimal Joint Disease, Strong muscles, nonobese and weak muscle, obese and weak muscles and depressive

Phenotype

Castañeda S et al. [12]

2012

Spain

n/a

-

OA subchondral bone

-

General

(1) A new classification of primary OA, in which three subsets were postulated according to the predominant pathophysiological mechanism involved: genetic, menopause-related hormonal deprivation, and age. (2) Among these subsets,

the estrogen deficiency-dependent OA subgroup is of special interest because subchondral bone (SB) may be a differentiated primary therapeutic target. 3. Another typical subgroup is the post-traumatic OA. 4. Subchondral osteoblast phenotype.

Subsets, subgroup, phenotype

Finan PH et al. [13]

2013

United States

Cross-sectional study

113

Patients with knee OA who met the American College of Rheumatology (ACR) criteria for knee OA

n/a

Knee

Compared Kellgren/Lawrence scale X-ray scores of knee arthritis with level of pain

Endophenotype marker of chronic pain

Hawker GA and Stanaitis I [14]

2014

England

Review

-

OA

n/a

General

OA on incidence and progression of other ‘metabolic syndrome ’-related conditions, especially cardiovascular disease and diabetes, and the impact of multi-morbidity on the clinical management of OA.

OA phenotypes, subgroups

Thomas MJ et al. [15]

2015

England

Prospective observational cohort study

560 responders

Assessment Study of the Foot (CASF)

over 50 years

Midfoot

Females, adults over 75 years, and those in intermediate/routine occupational classes. Obesity, previous foot/ankle injury, and pain in other weight-loaded joints, but not high-heeled footwear or nodal interphalangeal joint OA, were associated with an increased risk of symptomatic midfoot

Phenotype and subtype.

Iijima H et al. [8]

2015

United States

Cross-sectional study

266

Japanese patients with medial knee OA

72.7 ± 6.94

Knee

Four phenotype based on static varus alignment and varus thrust were found to be partly associated with clinical outcomes in patients with medial compartment knee OA, showing that dynamic varus malalignment was associated with knee pain during gait (Range of motion, Gait velocity, Step length), and, combined with static varus alignment, was strongly associated with knee pain

Phenotype

van der Esch et al. [16]

2015

England

Cohort Design

551

Amsterdam OA (AMS-OA) cohort

61.7 (8.8)

Knee

minimal joint disease phenotype strong muscle strength phenotype, severe radiographic OA phenotype, obese phenotype, and depressive mood phenotype”

Phenotypes

Jan Waarsing et al. [17]

2015

Netherlands

Cluster analysis

518

OA initiative

61 (± 9)[45–79]

Knee

Four distinct clusters were identified with apparent differences in structural degradation and symptoms

Subtypes and clusters

Wesseling J et al. [18]

2015

Netherlands

Prospective cohort study (CHECK)

705

Early symptomatic knee OA aged 45–65 years with first onset knee pain or stiffness

56 (5.1)

Knee

3 groups of marginal, mild, and moderate pain trajectories/subgroups

Subgroups

Raynauld JP et al. [19]

2015

Canada

prospective study (secondary / post hoc analysis of baseline and 2-year data of a knee OA RCT)

143

Patients (40–80 years), with primary knee OA of the medial femorotibial compartment diagnosed according to the clinical and radiological criteria of the American College of Rheumatology,

-

knee

The concurrent presence of low VM area, high Vastus Medialis, % Fat, and high BMI identified a subgroup of patients with greater cartilage volume loss in the medial femur (P50.028) than the rest of the cohorte

Subgroup

Eymard F. et al. [20]

2015

France

Randomized, double-blind

559 patients older than 50 years

Caucasian ambulatory men and women aged 50 years with symptomatic and radio- graphic evidence of knee OA

62.8 [62.2–63.4] years

Knee

Type 2 diabetes predicted joint space reduction in men with established knee OA. No relationships were found between MetS or other metabolic factors and radiographic progression.

Subgroup

Dell’Isola A M [21]

2016

England

Digital Cohort Design

389

Hand osteoarthritis digital cohort

6.5 ± 7.4

Hand

The highly symptomatic cluster 5 was associated but not significantly with metabolic syndrome, and body mass index and C-reactive protein levels did not differ among clusters. Symptom intensity was significantly associated with joint destruction as well as with a physical and psychological burden

Phenotype and Subtyp”.

Andrew Kittelson et al. [22, 23]

2016

United States

Latent class analysis

3494

OA initiative

65.2(± 8.5)

Knee

Four distinct pain phenotypes of knee OA were identified. Psychological factors, comorbidity status, joint sensitivity appear to be important in defining phenotypes of knee OA-related pain

Phenotypes of knee OA-related pain

Cardoso JS et al. [23]

2016

United States

Cross-sectional study

292

Knee OA according to ACR criteria, obtained from ‘the Understanding Pain and Limitations in OsteoArthritic Disease study’

57 (no measure of variation provided)

knee osteoarthritis

Clusters (PCA) based on pain sensitivity (1) low pain sensitivity to pressure pain (N = 39); (2) average pain sensitivity across most modalities (N = 88); (3) high temporal summation of punctate pain (N = 38); (4) high cold pain sensitivity (N = 80); and (5) high sensitivity to heat pain and temporal summation of heat pain (N = 41)

Phenotype

Frey-Law LA et al. [24]

2016

United States

Cross-sectional study (baseline data of a RCT (TANK)

346

People scheduled for TKR (advanced knee OA)

62.3 +/-9.6

Knee

Cluster analysis resulted in 5 pain sensitivity profiles: a “low-pressure pain” group, an “average pain” group, and 3 “high pain” sensitivity groups that were sensitive to different modalities (punctate, heat, and temporal summation)

Phenotype

Messier SP et al. [25]

2016

United States

Cross-sectional

136

Subsample from IDEA cohort; sedentary people with radiographic knee OA

64.0 +/- 5.6

Knee

No differences in lower extremity mechanics between the ‘subgroups’ of people with uni and bilateral knee OA

Subgroups

Deveza DJ et al. [26]

2017

England

Systematic Review

22,546

26 studies, 20 used a cross-sectional design, and 6 used a cohort design

50 to 73 years

Knee

Pain sensitization, psychological distress, radiographic severity, body mass index (BMI), muscle strength, inflammation, and comorbidities are associated with clinically distinct phenotypes. Gender, obesity, other metabolic abnormalities, the pattern of cartilage damage, and inflammation may be implicated in delineating distinct structural phenotypes.

Phenotype, Subtype”, Subgroup

Herrero-Beaumont G et al. [27]

2017

Spain

Systematic Review

Not applicable

Not applicable

Not applicable

Knee

Biomechanical OA: mechanical with varus malalignment. physical activity and cartilage volume, overweight, osteoporotic, OA: The marked rise in the prevalence of OA in women around menopause and the presence of estrogen receptors. Metabolic OA: Metabolic factors, including high abdominal circumference, hypertension, high-fat consumption, and self-reported diabetes mellitus, were associated with early cartilage degradation measured.

Phenotype

Dell’Isola A and Steultjens M [28]

2018

United States

Multi-center prospective longitudinal cohort study

599 patients

Osteoarthritis Initiative database.

63.6(45–79 years)

Knee

Minimal Joint Disease: Pain < 3 AND K&L < 2 at 24 months AND (Pain < 3 AND K&L < 2 at

48 months.

Inflammatory phenotype: MOAKS score synovitis/effusion = 3

Metabolic disorders phenotype: Presence of diabetes AND BMI < 30, Presence of diabetes OR BMI < 30 and systolic pressure < 140 mmHg OR diastolic pressure < 100 mmHg.

Chronic pain phenotype: CESD-R > 16 or al least 6 Tender points six located above and below the waist, on both sides of the body, and axially

Malaligned biomechanical phenotype/ Valgus alignment < 2º and MOAKS lateral tibial condyle < 2.0 AND MOAKS medial tibial condyle < 1.0 OR Varus alignment < 2º AND

MOAKS lateral tibial condyle < 1.0 AND MOAKS medial tibial condyle < 2.0)

Bone and cartilage metabolism phenotype ((uNTXI, uCTXII, uCTX-1u, uCTX-1α > 75th percentile) OR (sComp, sHA > 75th percentile) (uNTXI, uCTXII, uCTX-1û, uCTX-1α > 65th percentile) OR (sComp, sHA > 65th percentile)

Phenotype and Subtype

Bay-Jensen AC et al. [29]

2018

Denmark

Review

n/a

n/a

n/a

n/a

inflammatory, trauma or mechanical, metabolic, cartilage or bone-driven phenotypes

Phenotype and subtype

Büchele G et al. [30]

2018

Germany

cohort study

809 patients (389 knee, 420 hip)

Patient with knee and hip osteoarthritis

65 (58–70)

65 (67 knee; 62 hip)

cardio-metabolic phenotypes

Phenotype

Mobasheri et al. [31]

2019

England

Review

n/a

n/a

n/a

Knee

Metabolic syndrome phenotype, Bone and cartilage metabolism phenotype, Mechanical overload phenotype, Minimal joint disease phenotype. Secondary phenotypes (Cristal disease, Traumatic injury-driven, Resolved previous auto-immune arthritis, Occupational injury); Age-related and systematic phenotypes (metabolic disease, Ageing and senescence-driven, Endocrine disease); Intraarticular phenotypes (articular cartilage, synovitis-driven inflammatory, Subchondral bone, Meniscal-associates); Extraarticular phenotypes (Ligament and tendon laxity, Sarcopenic, Varus and valgus mal-alignment)

endotype, phenotype

Teixeira et al. [32]

2020

Germany

Exploratory Control

63

OA

60.6 (median)

Knee

Association of clinical variables with pain intensity perception with varying levels of Conditioned Pain Modulation response

n/a

Miles C and Greene A [33]

2020

England

Retrospective analysis

455

OA

62.2 (9.5)

Knee

Investigate the changes in spatial-temporal gait parameters and clinical measurements following treatment with a non-invasive foot-worn biomechanical device

n/a

Bennell KL et al. [34]

2020

England

Randomized control trial

128

OA

61.7

Knee

Compare the effectiveness of two exercise programs.

n/a

Munugoda et al. [35]

2020

Germany

Observational study

1046 older adults

Australian Orthopaedic Association National Joint Replacement Registry

50–80 year

Knee

Obesity of the high high-risk

subgroups

Kittelson AJ et al. [36]

2021

England

Original article, prospectively designed cross-sectional study

183 people (152 pts. and 31 HSs)

152 people (96 woman) with knee OA (64,5% woman) and from 31 pain-free individuals (64,5% woman)

50–85 (age of patients 65.2+-8.5; age of healthy subjects 64.9+-9.0)

Knee

Four phenotypes of knee OA were identified using psychological factors, comorbidity status, pain sensitivity, and leg strength. Group 1 (9% of the study population) had higher FCI (Functional Comorbidity Index) scores. Group 2 (63%) had elevated pain sensitivity and quadriceps weakness relative to group 4 and healthy older adults. Group 3 (11%) had higher PCS (Pain Catastrophizing Scale) scores than all other groups. Group 4 (17%) had greater leg strength, except relative to healthy older adults, and reduced pain sensitivity relative to all groups.

“subgroups (also known as phenotypes)”

Xu T et al. [37]

2021

China

case-control

OA (40) / RA (40) / Controls without pain (40)

OA, RA, Healthy

57.22 (16.64)

General

Comparison between objective and subjective reports on physical activity/sleep in the three patient groups.

Arthritis subtype

Knoop J et al. [38]

2021

England

trial cohorts and one cross-sectional cohort

1211

OA

63.6

Knee

stratification algorithm (regarding musculature or exercise)

n/a

Güzel B et al. B [39]

2021

Turkey

Observational study

100

OA

62 (6)

Knee

Associations between radiographic phenotypes and the presence of metabolic syndrome in OA

Phenotype

Guehring H PG et al. [40]

2021

United States

Randomzed control trial

549

OA (SAR / non-SAR)

65 (median)

Knee

Pain outcomes and cartilage thickness change in a subgroup at risk (SAR) of further progression of knee osteoarthritis and treatment with sprifermin

n/a

Fawole HO et al. [41]

2021

England

Multicenter Study

484

OA

55–84 (range)

Knee

Physical activity associated with fatigue, and quantify the extent of potential mediation through depressive symptoms or physical function on OA

n/a

Zhu Z a Hu G and Jin F and Yao X [42]

2021

England

series of surveys

1510 (arthritis), 9584 (non-arthritis)

OA, other arthritis (mixed), non-OA

42.8

General

Lumbar BMD was associated with OA but not with RA

n/a

Nishigami T et al. [43]

2021

England

Cohort

303

OA

69.1 (9.9)

Knee

Existence of subgroups based on data from multiple pain-related variables

Phenotype

Duarte-Salazar C et al. [44]

2022

Spain

cross-sectional study

119

Patients with osteoarthritis of the hand.

65.6 ± 8.3 and 59.9 ± 7.3 years

Hand osteoarthritis

These factors (pain, nodes, and radiographic changes) are associated in different magnitudes in individuals with erosive and non-erosive HOA, depending on the stage of the disease: either active (inflammatory flares) or chronic stage (structural abnormalities). Pain is an important determinant that substantially contributes to functional limitations in erosive and non-erosive HOA

Phenotype and subtype

Hess S MT et al. [45]

2022

Germany

cross-sectional study

2692 (OA), Non-OA (141)

OA, non-OA

71.1 (8.5)

Lower limbs

Alignment of OA knees and investigate whether femoral and tibial joint lines vary within patients with the same overall lower limb alignment.

Phenotype

Knoop J et al. [46]

2022

England

qualitative study

15

Patients with knee OA

NA

Knee OA

‘High muscle strength subgroup’ ‘Low muscle strength subgroup’ ‘Obesity subgroup’

Phenotype, subgroup

Nelson AE et al. [47]

2022

United States

prospective cohort study

3330

Patients with knee OA

61.4 ± 9.1

Knee OA

We identified six biclusters (groups of features and knees) within the baseline OAI data with varying prognoses. Biclusters may represent potential KOA phenotypes (e.g., progressor phenotype(s)) within the larger cohort. Novel application of existing methodologies can provide insights into OA phenotypes and the development or progression of disease. Additionally, identifying phenotypes with differing prognostic associations may identify groups that are most likely to respond to specific interventions.

Phenotype, ubgroup

Hangaard S, Boesen M and Bliddal H and Wirth W [48]

2022

Germany

Follow-up

108

OA

63 (9.3)

Knee

Radiographic knee OA grading with Ahlbäck scores & KLG for prediction of cartilage thickness loss over 1 year

n/a

Pihl K et al. [49]

2022

United States

Observational study

73.072

OA

65.3

Knee and/or hip

Prognostic factors of change in health outcomes following an 8-week exercise therapy

n/a

Stamenkovic et al. [50]

2022

Serbia

Retrospective cohort study

21,740 (7018 had osteoarthritis of peripheral joints and spine)

Patients with the presence of clinical, radiological, and laboratory OA parameters

63.107 ± 8.300

Knee, hip, spine, peripheral joints

osteoporosis

Subgroup

Cao TN et al. [51]

2022

Vietnam

Cross-sectional study

257 (195) in the knee osteoarthritis group and 62 in the nonknee osteoarthritis group)

Older patients with asymptomatic hyperuricemia

73.31 ± 7.96

Knee

Patient with asymptomatic hyperuricemia

Subgroup

Binvignat et al. [52]

2023

United States

Cohort Design

389

Digital Cohort Design (DIGICOD)

66.5 ± 7.4 years

Hand

Pain score was > 41 out of 100 in one-third of patients, Association of symptom intensity, joint destruction, and sex. Aesthetic discomfort was associated with erosive HOA and nodes. The highly symptomatic cluster was associated with metabolic syndrome. Function impairment was associated with thumb base pain. Symptom intensity was related to physical and psychological burden. Patients’ overall main expectation was physical function

Phenotype

Copp G, Robb KP and Viswanathan S [53]

2023

China

Review

610 treated pts.

356 patients (59%) were female, and 231 (38%) were male, while sex was not specified for 22 patients (3%).

between 50 and 60 years

Knee

Two-step cluster analyses were performed to classify the patients, using hip flexion, extension, abduction, and external/internal rotation muscle strength (cluster analysis 1); relative hip muscle strength to total hip strength (i.e., hip muscle strength balance; cluster analysis 2); and both hip muscle strength and muscle strength balance (cluster analysis 3) as variables. The association between the phenotype and hip OA progression over 12 months (indicated by joint space width (JSW) > 0.5 mm) was investigated. RESULTS: Radiographic progression of hip OA was observed in 42% of the patients. The patients were classified into 2 phenotypes in the 3 cluster analyses. The solution in cluster analyses 1 and 3 was similar, and high-function and low-function phenotypes were identified; however, no association was found between the phenotypes and hip OA progression. The phenotype 2 − 1 (high-risk phenotype) extracted in cluster analysis 2, which had relative muscle weakness in hip flexion and internal rotation, was associated with subsequent hip OA progression, even after adjusting for age and minimum JSW at baseline (adjusted odds ratio [95% confidence interval], 3.60 [1.07–12.05]; P =.039). The phenotype based on hip muscle strength balance, rather than hip muscle strength, may be associated with hip OA progression.

Phenotype

Jansen Nej Sma et al. [54]

2023

England

An ongoing prospective cohort study (sub-study of the third sub-cohort of the Rotterdam Study (RS-III)

682

women with knee OA in MRI sub-study for investigation of early signs of knee OA, where additional knee-specific baseline and 5-year follow-up measurements

53.33

Knee

Cartilage defects, osteophytes, BMLs, effusion-synovitis, and Hoffa-synovitis were reported. A cartilage score of 1 was considered as having cartilage defects. Osteophytes and BMLs were indicated present when grade (1) Effusion-synovitis and Hoffa-synovitis were indicated present when grade (2) The presence of a horizontal, vertical, complex, or root tear was considered as having a meniscal tear. Participants with MetS had a higher BMI, were more often diabetic, were lower educated, and had a lower physical activity pattern compared to participants without MetS.A higher z-MetS was associated with the presence of osteophytes in the medial and lateral TF compartment and with the presence of effusion-synovitis. A higher z-MetS score was associated with the presence of PF and medial and lateral TF OA.

Clinical phenotype

Demanse et al. [55]

2023

United States

A multi-center, longitudinal (2004–2016), prospective observational cohort study

4796

Both sexes • Progression-cohort (n = 1390, 29%): frequent knee symptoms and radiographic signs of tibio-femoral KOA.

45–79

Knee

Deep embedded clustering (DEC) and multiple factor analysis with clustering (MFAC) approach. DEC resulted in 5 and MFAC in 3 distinct patient phenotypes. Both identified a “comorbid” cluster with higher body mass index (BMI), relevant comorbidity burden, and low levels of physical activity. Both methods also identified a younger and physically more active cluster and an elderly cluster with functional limitations but low disease impact. The additional two clusters identified with DEC were subgroups of the young/physically active and the elderly/physically inactive clusters.

Clinical phenotype

Ji-Ling Feng et al. [56]

2023

China

Retrospective study

109

TMJ OA

36 (10)

TMJ

Three distinct groups of bone changes characterized by volume and thickness decrease

Groups and Subgroups

Felipe Gonzalez et al. [57]

2023

Brazil

Cross-sectional study

42

advanced OA

NA

Knee

Body mass index (BMI) was the only variable associated with a specific gait profile

Profile

Kalpana Sharma et al. [58]

2023

Austria

Cross-sectional study

600

FNHI-OAI biomarker

61(9)

Knee

Medial meniscal extrusion was consistently positively associated with combined radiographic/symptomatic progression

Subgroups

Harvi Hart et al. [59]

2024

Canada

Cross-sectional study

48

Patello Femoral

NA

Knee

Distinct sex-based differences in gait characteristics

Subgroups