Steps Independent variables β p  R2 ΔR2
1 Age only – linear effect     .001 .001
  Age – linear -.030 .766    
2 Age only – quadratic effect     .073 .072*
  Age – linear Age – quadratic -2.141 2.128 .007 .007    
3 Add other socio-demographic variables     .096 .023
  Age – linear -2.048 .010    
  Age – quadratic 2.019 .012    
  Sex (reference=male) 0.156 .122    
  Work status (reference=not working) 0.024 .843    
4 Add clinical variables     .457 .361*
  Age – linear -1.770 .008    
  Age – quadratic 1.606 .015    
  Sex (reference=male) 0.144 .081    
  Work status (reference=not working) 0.070 .472    
  Pre-stroke fatigue (reference=no) 0.348 <.001    
  Physical functioning (SF–36) -0.477 <.001    
  ADL (Barthel Index ) 0.174 .092    
  Sleep quality (PSQI total score) 0.004 .964    
  Respiratory disease 0.233 .005    
5 Add depressive symptoms     .491 .034*
  Age – linear -1.295 .053    
  Age – quadratic 1.165 .080    
  Sex (reference=male) 0.127 .115    
  Work status (reference=not working) 0.072 .446    
  Pre-stroke fatigue (reference=no) 0.292 .001    
  Physical functioning (SF–36) -0.454 <.001    
  ADL (Barthel Index) 0.144 .153    
  Sleep quality (PSQI total) -0.058 .505    
  Respiratory disease 0.191 .019    
  Depression (BDI–II score) 0.223 .017    
Note: FSS-7 = Fatigue Severity Scale 7 item version; ADL = Activities of Daily Living; PSQI = Pittsburgh Sleep Quality Index; BDI-II=Beck Depression Inventory II. The model was significant (p<.05) at steps 2 through 5, but not at step 1. Test of final model: F(10,89)=8.58, p<.001), adjusted R2 = .434
Table 3: Multivariate relationships between post-stroke fatigue (FSS-7) and age, controlling for other demographic and clinical variables (n=100).