Mixed Effects Model for log(FEV1) with Random Intercept and Slopes for Age and Log Height

Six Cities Study

 

The Mixed Procedure

 

Model Information

Data Set

WORK.FEV

Dependent Variable

logfev1

Covariance Structure

Unstructured

Subject Effect

id

Estimation Method

REML

Residual Variance Method

Profile

Fixed Effects SE Method

Model-Based

Degrees of Freedom Method

Containment

 

Class Level Information

Class

Levels

Values

id

299

not printed

 

Dimensions

Covariance Parameters

7

Columns in X

5

Columns in Z Per Subject

3

Subjects

299

Max Obs Per Subject

12

Observations Used

1993

Observations Not Used

0

Total Observations

1993

 

Iteration History

Iteration

Evaluations

-2 Res Log Like

Criterion

0

1

-2961.45030802

 

1

2

-4525.34587509

1.08900123

2

1

-4526.52099519

0.21154590

3

1

-4534.30743803

7.32006321

4

1

-4544.85801410

219.20495583

5

1

-4548.65364068

5.27040271

6

1

-4550.87935831

0.00485639

7

3

-4560.41529524

.

8

2

-4580.64140277

.

9

2

-4588.90623048

.

10

1

-4589.86323392

0.00000850

11

1

-4589.89930825

0.00000002

12

1

-4589.89938271

0.00000000

 

Convergence criteria met.

 

Estimated G Matrix

Row

Effect

id

Col1

Col2

Col3

1

Intercept

1

0.01337

-0.00007

-0.01663

2

age

1

-0.00007

0.000012

-0.00036

3

loght

1

-0.01663

-0.00036

0.07984

 

Estimated G Correlation Matrix

Row

Effect

id

Col1

Col2

Col3

1

Intercept

1

1.0000

-0.1648

-0.5091

2

age

1

-0.1648

1.0000

-0.3731

3

loght

1

-0.5091

-0.3731

1.0000

 

 

Estimated V Correlation Matrix for id 35

Row

Col1

Col2

Col3

Col4

Col5

Col6

Col7

Col8

Col9

Col10

Col11

Col12

1

1.0000

0.7180

0.7003

0.6652

0.6274

0.5704

0.5109

0.4887

0.4811

0.4803

0.4793

0.4707

2

0.7180

1.0000

0.7042

0.6792

0.6494

0.6020

0.5505

0.5310

0.5242

0.5234

0.5223

0.5145

3

0.7003

0.7042

1.0000

0.6913

0.6720

0.6371

0.5961

0.5800

0.5743

0.5733

0.5721

0.5652

4

0.6652

0.6792

0.6913

1.0000

0.6929

0.6760

0.6504

0.6391

0.6346

0.6331

0.6315

0.6257

5

0.6274

0.6494

0.6720

0.6929

1.0000

0.6983

0.6855

0.6784

0.6752

0.6735

0.6717

0.6670

6

0.5704

0.6020

0.6371

0.6760

0.6983

1.0000

0.7180

0.7159

0.7141

0.7120

0.7097

0.7063

7

0.5109

0.5505

0.5961

0.6504

0.6855

0.7180

1.0000

0.7381

0.7376

0.7350

0.7324

0.7302

8

0.4887

0.5310

0.5800

0.6391

0.6784

0.7159

0.7381

1.0000

0.7430

0.7410

0.7388

0.7375

9

0.4811

0.5242

0.5743

0.6346

0.6752

0.7141

0.7376

0.7430

1.0000

0.7427

0.7410

0.7405

10

0.4803

0.5234

0.5733

0.6331

0.6735

0.7120

0.7350

0.7410

0.7427

1.0000

0.7415

0.7416

11

0.4793

0.5223

0.5721

0.6315

0.6717

0.7097

0.7324

0.7388

0.7410

0.7415

1.0000

0.7421

12

0.4707

0.5145

0.5652

0.6257

0.6670

0.7063

0.7302

0.7375

0.7405

0.7416

0.7421

1.0000

 

 

Fit Statistics

-2 Res Log Likelihood

-4589.9

AIC (smaller is better)

-4575.9

AICC (smaller is better)

-4575.8

BIC (smaller is better)

-4550.0

 

 

Solution for Fixed Effects

Effect

Estimate

Standard Error

DF

t Value

Pr > |t|

Intercept

-0.2857

0.03903

297

-7.32

<.0001

age

0.02344

0.001278

251

18.34

<.0001

loght

2.2476

0.04692

227

47.90

<.0001

baseage

-0.01610

0.007429

1213

-2.17

0.0304

logbht

0.1774

0.1454

1213

1.22

0.2225

 

Type 3 Tests of Fixed Effects

Effect

Num DF

Den DF

Chi-Square

F Value

Pr > ChiSq

Pr > F

age

1

251

336.50

336.50

<.0001

<.0001

loght

1

227

2294.67

2294.67

<.0001

<.0001

baseage

1

1213

4.70

4.70

0.0302

0.0304

logbht

1

1213

1.49

1.49

0.2222

0.2225