Skip to main content

Table 4 Comparison of observed subgroup characteristics that are in line with proposed underlying phenotype with other subgroups (research question 2; findings resulting in accepted hypotheses in bold)

From: Construct validity of the OCTOPuS stratification algorithm for allocating patients with knee osteoarthritis into subgroups

  Low muscle strength subgroup (L) High muscle strength subgroup (H) Obesity subgroup (O) Difference between L and H Difference between L and O Difference between H and O
Baseline variables
 Gender (female), n (%) 310 (74%) 88 (41%) 388 (67%) P < 0.001 P = 0.03 P < 0.001
 Age (years), mean ± SD 65.2 ± 8.4 62.3 ± 8.6 62.7 ± 8.3 P < 0.001 P < 0.001 P = 0.55
 Radiographic severity:
  K&L grade 0/1, n (%) 95 (24%) 72 (35%) 128 (24%) P = 0.02* P = 0.44 P < 0.001*
  K&L grade 2, n (%) 131 (33%) 64 (31%) 145 (27%)    
  K&L grade 3, n (%) 92 (23%) 48 (23%) 137 (25%)    
  K&L grade 4, n (%) 83 (21%) 42 (12%) 132 (24%)    
 Knee surgery, n (%) 51 (40%) 24 (63%) 65 (42%) P = 0.01 P = 0.80 P = 0.02
 Nr. of comorbidities:
  0, n (%) 142 (62%) 97 (71%) 151 (43%) P = 0.02 P < 0.001 P < 0.001
  1, n (%) 61 (15%) 28 (21%) 105 (30%)    
  2. n (%) 13 (3%) 8 (6%) 52 (15%)    
   ≥ 3, n (%) 14 (3) 3 (2%) 41 (12)    
Outcome variables
 Knee pain (0–10), mean ± SD 5.6 ± 2.0 4.7 ± 2.2 5.8 ± 2.1 P < 0.001 P = 0.31 P < 0.001
 Physical function (WOMAC, 0–68), mean ± SD 43.6 ± 17.3 30.6 ± 18.2 48.7 ± 18.6 P < 0.001 P < 0.001 P < 0.001
Stratification variables
 Body mass index (kg/m2), mean ± SD 26.4 ± 2.5 25.8 ± 2.4 35.5 ± 5.0 P < 0.01 P < 0.001 P < 0.001
 Quad. strength (Nm/kg), mean ± SD 0.71 ± 0.31 1.62 ± 0.32 0.78 ± 0.48 P < 0.001 P = 0.03 P < 0.001
 30s-CST (repetitions), mean ± SD 8.5 ± 2.3 13.2 ± 1.2 9.0 ± 2.7 P < 0.001 P = 0.15 P < 0.001
  1. K/L = Kellgren & Lawrence; CIRS = Cumulative Illness Rating Score; 30s-CST = 30 s chair stand test
  2. * significant finding in the opposite direction as expected, therefore hypothesis not accepted