Table 1 Studies reporting low skeletal muscle mass and/or muscle weakness in adults with obesity and knee or hip osteoarthritis

Batsis et al. [59], 2015

To describe the impact of dynapenic obesity on physical function in knee OA

Longi-tudinal,

North American population from OsteoArthritis Initiative (OAI), age ≥ 60 years, n = 526 in subgroup with knee OA (rKOA),

BMI ≥30 kg/m²

NI

Lowest sex-specific tertile of knee extensor strength (dynapenia)

Secondary analysis of prospective data from longitudinal cohort. Excluded severe knee OA. No assessment of muscle mass or body composition

Prevalence of dynapenic obesity was 16%.

Clemence et al. [69], 2017

To analyze the association between low lean mass and clinical symptoms in knee and hip OA

Cross-sectional

French adults with hip and knee OA (KL grade ≥ 2) from KHOALA study, n = 358, age 63.4 ± 8.4 years

BMI ≥30 kg/m², or sex specific FM or WC cut-offs

DXA

ASM/BMI < 0.789 for men and < 0.512 for women (FNIH cutoffs)

Secondary analysis of prospective data from longitudinal cohort. No information on exclusion criteria. No assessment of muscle strength or function

SO prevalence was 16.2%. Low lean mass was associated with pain and impaired function in subjects with normal BMI, but not with obesity (no significant differences between NSO and SO groups).

Ji et al. [60], 2016

To identify the prevalence of SO in knee and hip orthopedic surgery (OS) patients

Cross-sectional

Korean orthopedic surgery patients (hip or knee TJA or femoral fracture repair) (OS, n = 222) compared to control non-surgical outpatients (non-OS, n = 364)

BMI > 25 kg/m²

DXA

ASM/height², ASM/weight, and ASM/height and fat mass (residuals)

Retrospective analysis of data. No assessment of muscle strength or function

SO prevalence ranged from 1.3–35.4% in TKA and 0–18.4% in THA patients depending on definition used. SO rates were higher in OS patients compared to non-OS patients.

Jin et al. [61], 2017

To examine the associations between obesity, sarcopenia and OA in elderly

Cross-sectional

Korean population (KNHANES) age ≥ 65 years group with knee OA (K/L grade ≥ 2) (n = 1865) compared to lumbar spondylosis group (n = 1709)

BMI ≥25 kg/m²

DXA

ASM/weight, 2SDs below average of sex-matched young reference group

Secondary analysis of population survey data. No assessment of muscle strength or function

Results indicate correlation between SO and NSO with knee OA, but no relationship with lumbar spondylosis. Females with SO had increased OR for knee OA when adjusted for age and waist circumference (OR 1.80, CI 1.03–3.12).

Knoop et al. [62], 2011

To identify distinct clinical phenotypes and their impact in knee OA

Cross-sectional

North American population with knee OA (K/L grade 0–4) from OsteoArthritis Initiative (n = 842, age 63.2 ± 9.1)

BMI ≥30 kg/m²

NI

Low mean score of quadriceps and hamstring isometric strength

Secondary analysis of prospective data from longitudinal cohort. No assessment of muscle mass or body composition. No clear cut-off for defining weakness

Dynapenic obesity group (“obese and weak” phenotype) had higher pain and poorer physical function compared to “minimal joint disease”, “strong muscle”, and “non-obese and weak” phenotypes.

Lee et al. [63], 2016

To investigate association between lower limb muscle mass and knee OA

Cross-sectional

Korean population (KNHANES) age ≥ 50 years, n = 821 with knee OA (K/L grade ≥ 2), (n = 821), and control group without knee OA (n = 4103)

BMI ≥27.5 kg/m²

DXA

ASM/weight, 2SD below the mean in sex-matched young reference group (< 29.5% in men, < 23.2% in women)

Secondary analysis of population survey data. No assessment of muscle strength or function

SO prevalence was 5.2% in knee OA group compared to 1.8% in control group.

Lee et al. [64], 2012

To analyze the association between knee OA, sarcopenia and obesity

Cross-sectional

Korean population (KNHANES) with bilateral knee OA (K/L grade ≥ 2) age ≥ 50 years, n = 2893

BMI ≥27.5 kg/m²

DXA

ASM/weight, 2SD below the mean in sex-matched young reference group (< 26.8% in men, < 21% in women)

Secondary analysis of population survey data. No assessment of muscle strength or function

SO prevalence was 3% overall. When adjusted for age and sex, SO had stronger association with knee OA (OR 3.51, CI 2.15–5.75) compared to NSO (OR 2.38, CI 1.80–3.15).

Manoy et al. [65], 2017

To assess association between leptin, vitamin D, muscle strength and physical performance in knee OA

Cross-sectional

Thailand knee OA patients (K/L grade < 3) (n = 208), age 65 ± 7 years

BMI > 25 kg/m²

BIA

ASM/weight < 30.4% in men and < 25.8% in women, and EWGSOP gait speed and grip strength cutoffs

Unclear if data collected retrospectively or prospectively. No description of sampling methods. Excluded severe knee OA

SO prevalence was 13.9%. Patients with SO had poorer performance on the timed up and go (TUG), sit to stand (STS) and 6 min walk tests (6MWT) compared to those with NSO or NO.

Oosting et al. [66], 2016

To determine the association of obesity and recovery after THA when stratified by muscle strength

Cross-sectional

Netherlands THA patients (n = 297), age 69 ± 11 years

BMI > 30 kg/m²

NI

Maximal handgrip strength (< 20 kg for woman and < 30 kg for men)

Secondary analysis of prospective cohort. No assessment of muscle mass or body composition

Obesity and muscle weakness (dynapenic obesity) was associated with prolonged length of stay > 4 days (OR 3.59, CI 1.09–11.89) and delayed inpatient recovery (> 2 days to walk with gait aid) (OR 6.21, CI 1.64–23.65), but not in those with obesity alone.

Segal et al. [67], 2005

To analyze the impact of low limb lean mass in knee OA distinct from body weight

Cross-sectional

Japanese female orthopedic knee OA (K/L grade ≥ 2) patients age ≥ 45 years (n = 341), compared to control group with fracture, sprains or back pain (n = 604)

BMI > 24.9 kg/m²

BIA

Lower limb LST

Unclear if data collected retrospectively or prospectively. No clear cut-off for defining low LST. No assessment of muscle strength or function

Females with knee OA had 5–15% less lower limb LST compared to control groups across BMI categories, with significant 1.8 kg and 1.5 kg differences in overweight and obesity groups, respectively.

Suh et al. [68], 2016

To analyze the association between obesity, sex, and lower extremity lean mass in knee OA

Cross-sectional

Korean population (KNHANES) age ≥ 50 years with unilateral knee OA (K/L grade ≥ 2) (n = 4246; 1829 men and 2417 women)

BMI ≥27.5 kg/m²

DXA

Lower extremity LST/weight, in lowest quartile

Secondary analysis of population survey data. No assessment of muscle strength or function

In females, obesity and low muscle mass was strongly association with knee OA (OR 2.31, CI 1.35–3.93) compared to obesity and normal muscle mass (OR 1.03, CI 0.26–4.02).