Table 3 Univariate and multivariate logistic regression analyses of training cohort.

From: Development of a nomogram prediction model for gait speed trajectories in persons with knee osteoarthritis

variables

Univariate model

Multivariable model

OR

95%CI

P value

OR

95%CI

P value

Age, year

 < 60

 ≥ 60

2.188

1.577–3.037

 < 0.001

2.873

1.983–4.215

 < 0.001

Gender

 Female

 Male

0.405

0.289–0.568

 < 0.001

0.578

0.397–0.835

0.004

Race

 Other

 White or Caucasian

0.323

0.236–0.443

 < 0.001

0.459

0.319–0.660

 < 0.001

Marital status

 Other

 Married

0.419

0.307–0.571

 < 0.001

0.827

0.570–1.204

0.318

Education

 None/primary

 Secondary

0.738

0.541–1.005

0.052

   

 Tertiary

0.647

0.456–0.918

0.012

   

Income, $

 < 50,000

 ≥ 50,000

0.330

0.240–0.455

 < 0.001

0.644

0.443–0.933

0.020

Smoking history

 No

 Yes

1.140

0.837–1.554

0.404

   

Drinking history

 No

 Yes

0.532

0.381–0.744

 < 0.001

   

Obesity

 No

 Yes

2.377

1.725–3.275

 < 0.001

1.874

1.320–2.674

 < 0.001

Depressive symptoms

 No

 Yes

2.442

1.666–3.579

 < 0.001

1.906

1.220–2.959

0.004

Comorbidity

 No

 Yes

2.112

1.535–2.906

 < 0.001

1.438

1.008–2.043

0.044

History of knee injury

 No

 Yes

1.264

0.681–2.347

0.466

   

History of knee surgery

 No

 Yes

0.725

0.523–1.005

0.050

   

KLG

 2

 3

1.259

0.923–1.717

0.147

   

 4

0.982

0.645–1.495

0.932

   

WOMAC pain

 < 5

 ≥ 5

2.511

1.780–3.504

 < 0.001

1.761

1.227–2.546

0.002

  1. Obesity was defined as BMI ≥ 30 kg/m2, depressive symptoms were defined as CES-D (catchment-area epidemiology survey-depression) score ≥ 16.
  2. KLG Kellgren–Lawrence grade, WOMAC Western Ontario & McMaster Universities Osteoarthritis Index.