S9 File. SAS output Glimmix outcomes men and women under 75 years of age The SAS System 10:18 Tuesday, October 13, 2015 49 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable OUT_smoking Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 109770 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 3 418948.02088 0.53117371 0.000102 1 0 2 418268.40998 0.02006907 0.000567 2 0 2 418383.77398 0.00018002 0.021735 3 0 1 418383.49592 0.00000454 8.298E-6 4 0 1 418383.49062 0.00000131 0.005294 5 0 1 418383.48967 0.00000133 0.000127 6 0 1 418383.49048 0.00000028 0.001009 7 0 1 418383.49031 0.00000022 0.000109 8 0 1 418383.49045 0.00000024 0.000864 The SAS System 10:18 Tuesday, October 13, 2015 50 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 418383.49059 0.00000019 0.000072 10 0 1 418383.49047 0.00000157 0.006337 11 0 1 418383.48951 0.00000153 0.000096 12 0 1 418383.49045 0.00000000 0.000095 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 418383.5 Generalized Chi-Square 19639.78 Gener. Chi-Square / DF 0.18 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.8189 0.007155 Residual 0.1789 0.000906 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 0.8549 0.06912 31716 12.37 <.0001 Age -0.03561 0.001288 78051 -27.66 <.0001 Year -0.01309 0.002029 78051 -6.45 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 78051 764.82 <.0001 Year 1 78051 41.61 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 51 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable PRO_Smoking Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 115521 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 3 498884.46067 0.91580803 0.012823 1 0 4 587833.4036 0.40344473 0.000568 2 0 4 619039.92533 0.11688472 0.004778 3 0 2 619981.63992 0.01971206 0.028206 4 0 2 619575.86179 0.00175998 0.010995 5 0 2 619537.53549 0.00011462 2.089E-8 6 0 2 619534.95982 0.00000741 0.000105 7 0 1 619534.79324 0.00000299 0.015822 8 0 2 619534.73737 0.00000249 0.000375 The SAS System 10:18 Tuesday, October 13, 2015 52 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 619534.78188 0.00000001 0.000041 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 619534.8 Generalized Chi-Square 54922.47 Gener. Chi-Square / DF 0.48 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.4873 0.007096 Residual 0.4754 0.002622 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -0.5139 0.1019 33287 -5.04 <.0001 Age 0.01267 0.001683 82231 7.53 <.0001 Year 0.1930 0.003190 82231 60.50 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 82231 56.66 <.0001 Year 1 82231 3660.64 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 53 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable OUT_macroalbuminuria Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 83428 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 2 242293.50359 1.08196299 0.000153 1 0 2 371366.27876 0.78814951 0.00046 2 0 2 463273.77069 0.46186421 0.000217 3 0 2 506158.0941 0.11768619 0.000452 4 0 2 515800.13303 0.00786259 0.000282 5 0 2 516385.02992 0.00062824 0.000719 6 0 2 516405.4014 0.00007323 0.000016 7 0 1 516407.71122 0.00000840 0.000022 8 0 1 516407.97594 0.00000098 0.000777 The SAS System 10:18 Tuesday, October 13, 2015 56 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 516408.00705 0.00000023 0.000129 10 0 1 516408.0101 0.00000003 0.000019 11 0 1 516408.0102 0.00000005 0.000141 12 0 1 516408.01003 0.00000004 0.000021 13 0 1 516408.0101 0.00000000 0.00002 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 516408.0 Generalized Chi-Square 16309.32 Gener. Chi-Square / DF 0.20 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.7549 0.007633 Residual 0.1955 0.001199 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -5.6052 0.2951 27745 -18.99 <.0001 Age 0.03674 0.005049 55680 7.28 <.0001 Year -0.06557 0.007485 55680 -8.76 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 55680 52.95 <.0001 Year 1 55680 76.73 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 57 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable PRO_AlbKreat Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 115521 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 3 533352.38168 0.35223939 0.14908 1 0 3 534043.44337 0.02148036 0.000735 2 0 2 533931.99724 0.00060096 2.172E-6 3 0 2 533923.51005 0.00000692 2.803E-8 4 0 1 533923.40351 0.00000052 0.005168 5 0 1 533923.40875 0.00000180 0.017834 6 0 1 533923.39286 0.00000200 0.007776 7 0 1 533923.4105 0.00000271 0.026786 8 0 1 533923.38665 0.00000206 0.00034 The SAS System 10:18 Tuesday, October 13, 2015 58 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 533923.40478 0.00000018 0.001781 10 0 1 533923.40647 0.00000063 0.006227 11 0 1 533923.40091 0.00000033 0.002011 12 0 1 533923.40378 0.00000007 0.001117 13 0 1 533923.40444 0.00000040 0.003948 14 0 1 533923.40797 0.00000139 0.013776 15 0 1 533923.39567 0.00000486 0.048068 16 0 2 533923.43858 0.00000370 0.000681 17 0 1 533923.40608 0.00000002 0.000935 18 0 1 533923.40588 0.00000000 0.000879 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 533923.4 Generalized Chi-Square 86309.92 Gener. Chi-Square / DF 0.75 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.2101 0.004428 Residual 0.7472 0.003834 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -1.4332 0.06633 33287 -21.61 <.0001 Age 0.01063 0.000987 82231 10.76 <.0001 Year 0.1723 0.002434 82231 70.79 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 82231 115.85 <.0001 Year 1 82231 5011.06 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 59 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable OUT_HbA1c Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 110314 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 2 482169.56102 0.12799230 0.000296 1 0 2 468690.5412 0.00671302 0.001974 2 0 2 469098.5891 0.00039683 0.000147 3 0 1 469096.33897 0.00002199 0.020756 4 0 2 469096.4653 0.00000308 5.229E-8 5 0 1 469096.45656 0.00000043 0.002948 6 0 1 469096.45773 0.00000007 0.002406 7 0 1 469096.45755 0.00000043 0.002166 8 0 1 469096.45715 0.00000042 0.001974 The SAS System 10:18 Tuesday, October 13, 2015 60 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 469096.45766 0.00000038 0.0018 10 0 1 469096.45719 0.00000035 0.001641 11 0 1 469096.45761 0.00000032 0.001496 12 0 1 469096.45723 0.00000029 0.001364 13 0 1 469096.45758 0.00000001 0.001352 14 0 1 469096.4575 0.00000002 0.001094 15 0 1 469096.45748 0.00000000 0.001072 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 469096.5 Generalized Chi-Square 65342.28 Gener. Chi-Square / DF 0.59 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.4252 0.005256 Residual 0.5923 0.002980 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 0.9484 0.06697 32113 14.16 <.0001 Age 0.003882 0.001033 78198 3.76 0.0002 Year -0.01834 0.002428 78198 -7.56 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 78198 14.12 0.0002 Year 1 78198 57.09 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 61 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable PRO_HbA1c Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 115521 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 4 510958.72068 0.93725466 0.00125 1 0 4 621230.79403 0.42791648 0.216117 2 0 4 668405.40305 0.23013623 0.00244 3 0 3 671298.44285 0.03880810 0.000828 4 0 2 670457.41526 0.00074819 1.088E-6 5 1 2 670440.65833 0.00000580 0.001844 6 0 1 670440.51028 0.00000114 0.003474 7 0 1 670440.53397 0.00000158 0.004813 8 0 1 670440.50147 0.00000220 0.006736 The SAS System 10:18 Tuesday, October 13, 2015 62 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 670440.54698 0.00000308 0.009424 10 0 1 670440.48331 0.00000012 0.009009 11 0 1 670440.48583 0.00000071 0.005252 12 0 1 670440.50049 0.00000000 0.005199 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 670440.5 Generalized Chi-Square 84156.30 Gener. Chi-Square / DF 0.73 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.1629 0.005981 Residual 0.7285 0.004535 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -0.01637 0.1049 33287 -0.16 0.8760 Age 0.01611 0.001603 82231 10.05 <.0001 Year 0.1607 0.003598 82231 44.67 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 82231 101.03 <.0001 Year 1 82231 1995.20 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 63 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable OUT_SBP Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 112182 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 3 472149.93205 0.12627392 0.00102 1 0 2 468341.04704 0.00557191 0.001314 2 0 2 468439.2511 0.00013985 2.362E-7 3 0 1 468439.19437 0.00000101 0.016539 4 0 2 468439.19487 0.00000141 0.007303 5 0 1 468439.1943 0.00000166 0.02073 6 0 2 468439.19497 0.00000128 0.00094 7 0 1 468439.19445 0.00000022 0.002711 8 0 1 468439.19454 0.00000062 0.007683 The SAS System 10:18 Tuesday, October 13, 2015 64 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 468439.19429 0.00000026 0.003243 10 0 1 468439.19439 0.00000001 0.003116 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 468439.2 Generalized Chi-Square 77847.91 Gener. Chi-Square / DF 0.69 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.3058 0.004421 Residual 0.6940 0.003430 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 1.8706 0.06307 32608 29.66 <.0001 Age -0.04830 0.000968 79571 -49.90 <.0001 Year 0.1005 0.002303 79571 43.63 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 79571 2490.19 <.0001 Year 1 79571 1903.89 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 65 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable PRO_SBP Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 115521 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 4 480585.51037 1.01166562 0.000634 1 0 4 623722.59376 0.59034229 0.210378 2 0 4 705185.8496 0.20684553 0.041745 3 0 3 726591.52964 0.06151359 0.00602 4 0 2 727291.15019 0.00443875 0.001044 5 0 2 727250.73743 0.00015115 0.000063 6 0 1 727249.20305 0.00000446 0.000794 7 0 1 727249.15722 0.00000116 0.009533 8 0 1 727249.13846 0.00000000 0.009677 The SAS System 10:18 Tuesday, October 13, 2015 66 The GLIMMIX Procedure Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 727249.1 Generalized Chi-Square 88278.55 Gener. Chi-Square / DF 0.76 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.1798 0.005094 Residual 0.7642 0.004239 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 0.7213 0.1349 33287 5.35 <.0001 Age 0.01583 0.002069 82231 7.65 <.0001 Year 0.1418 0.004622 82231 30.68 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 82231 58.55 <.0001 Year 1 82231 941.05 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 67 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable Out_Obesity Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 106284 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 2 359857.38245 0.17559464 0.000912 1 0 2 358544.45021 0.00413846 0.000158 2 0 1 358560.58883 0.00001715 8.009E-6 3 0 0 358560.58803 0.00000001 6.244E-6 Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 10:18 Tuesday, October 13, 2015 68 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 358560.6 Generalized Chi-Square 21658.70 Gener. Chi-Square / DF 0.20 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.8010 0.007143 Residual 0.2038 0.001051 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 1.4777 0.06090 31241 24.27 <.0001 Age -0.02609 0.001089 75040 -23.96 <.0001 Year 0.01607 0.001742 75040 9.22 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 75040 573.97 <.0001 Year 1 75040 85.09 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 69 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable Out_Overweight Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 106284 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 2 414407.19171 0.15300488 0.002555 1 0 2 406160.38571 0.00531637 5.033E-6 2 0 2 406326.10246 0.00002762 0.000045 3 0 1 406326.21846 0.00000030 0.001772 4 0 1 406326.21852 0.00000093 0.00548 5 0 1 406326.21847 0.00000076 0.00051 6 0 1 406326.21851 0.00000027 0.001567 7 0 1 406326.2185 0.00000082 0.004842 8 0 1 406326.21855 0.00000100 0.00301 The SAS System 10:18 Tuesday, October 13, 2015 70 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 406326.21849 0.00000041 0.000216 10 0 1 406326.21851 0.00000011 0.000663 11 0 1 406326.21851 0.00000035 0.002048 12 0 0 406326.21853 . 0.002047 13 0 6 406326.21853 2.00000000 0.000025 14 0 1 406326.21851 0.00000000 0.000023 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 406326.2 Generalized Chi-Square 36754.05 Gener. Chi-Square / DF 0.35 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.6616 0.006484 Residual 0.3458 0.001780 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -1.7236 0.06700 31241 -25.72 <.0001 Age 0.02020 0.001124 75040 17.98 <.0001 Year -0.01577 0.002107 75040 -7.49 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 75040 323.27 <.0001 Year 1 75040 56.04 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 71 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable OUT_BMI Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 106284 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 2 437990.37063 0.60895435 0.004123 1 0 2 446316.83786 0.02787386 4.54E-7 2 0 1 446820.24977 0.00243276 0.005302 3 0 2 446824.40425 0.00006932 0.000762 4 0 1 446824.44873 0.00000157 0.00018 5 0 1 446824.44967 0.00000055 0.002934 6 0 1 446824.44966 0.00000000 0.002924 The SAS System 10:18 Tuesday, October 13, 2015 72 The GLIMMIX Procedure Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 446824.4 Generalized Chi-Square 26554.16 Gener. Chi-Square / DF 0.25 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.7546 0.006929 Residual 0.2498 0.001288 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -2.8374 0.09076 31241 -31.26 <.0001 Age 0.01746 0.001567 75040 11.14 <.0001 Year -0.00249 0.002618 75040 -0.95 0.3413 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 75040 124.20 <.0001 Year 1 75040 0.91 0.3413 The SAS System 10:18 Tuesday, October 13, 2015 73 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable PRO_BMI Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 115521 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 4 515789.87164 0.94899986 0.005372 1 0 4 576386.27167 0.33904201 0.004715 2 0 4 589735.91746 0.05164616 0.0002 3 0 3 588938.33896 0.00398301 0.000279 4 0 2 588809.90949 0.00006618 1.146E-6 5 0 2 588806.77681 0.00000659 1.224E-9 6 0 1 588806.95927 0.00000047 0.00054 7 0 1 588806.97482 0.00000073 0.005685 8 0 1 588806.96228 0.00000001 0.005664 The SAS System 10:18 Tuesday, October 13, 2015 74 The GLIMMIX Procedure Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 588807.0 Generalized Chi-Square 61078.57 Gener. Chi-Square / DF 0.53 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.3875 0.005878 Residual 0.5287 0.002826 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -0.8533 0.08789 33287 -9.71 <.0001 Age 0.006688 0.001413 82231 4.73 <.0001 Year 0.2190 0.002867 82231 76.39 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 82231 22.39 <.0001 Year 1 82231 5835.70 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 75 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable OUT_CholHDL Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 108682 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 2 445215.85309 0.15120561 0.082691 1 0 2 437300.16051 0.00434382 0.021271 2 0 2 437456.39533 0.00011298 7.013E-9 3 0 2 437456.36648 0.00000121 1.264E-8 4 0 1 437456.38492 0.00000013 0.001262 5 0 1 437456.38488 0.00000118 0.011833 6 0 2 437456.38622 0.00000111 0.000468 7 0 1 437456.38495 0.00000044 0.004407 8 0 1 437456.38546 0.00000041 0.000175 The SAS System 10:18 Tuesday, October 13, 2015 76 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 437456.38499 0.00000016 0.001643 10 0 1 437456.38518 0.00000154 0.015437 11 0 1 437456.38342 0.00000000 0.015438 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 437456.4 Generalized Chi-Square 52662.05 Gener. Chi-Square / DF 0.48 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.5167 0.005742 Residual 0.4846 0.002471 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -1.5999 0.06274 31993 -25.50 <.0001 Age 0.02433 0.001005 76686 24.20 <.0001 Year 0.05576 0.002165 76686 25.75 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 76686 585.76 <.0001 Year 1 76686 663.30 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 77 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable PRO_CholHDL Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 115521 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 2 523007.80241 0.88290508 0.015843 1 0 3 617730.62801 0.33479821 0.13452 2 0 3 657555.83593 0.05207408 0.008882 3 0 2 661864.13066 0.00712135 0.003358 4 0 2 661819.72178 0.00022681 0.000201 5 0 1 661816.44355 0.00000433 0.001089 6 0 1 661816.37802 0.00000048 0.003198 7 0 1 661816.38429 0.00000253 0.016967 8 0 1 661816.35036 0.00000205 0.000644 The SAS System 10:18 Tuesday, October 13, 2015 78 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 661816.3778 0.00000000 0.000638 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 661816.4 Generalized Chi-Square 95817.04 Gener. Chi-Square / DF 0.83 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.1050 0.004325 Residual 0.8295 0.004543 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -0.4313 0.09586 33287 -4.50 <.0001 Age 0.01336 0.001438 82231 9.29 <.0001 Year 0.1930 0.003300 82231 58.48 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 82231 86.25 <.0001 Year 1 82231 3420.08 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 79 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable OUT_FeetExamination Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 59707 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 2 278502.79456 0.69728077 0.00005 1 0 2 295946.20914 0.12399558 0.000077 2 0 2 299442.52383 0.00456159 0.000027 3 0 1 299570.82318 0.00011736 0.000187 4 0 2 299570.12324 0.00000143 6.56E-10 5 0 1 299570.11668 0.00000069 0.004848 6 0 1 299570.11698 0.00000072 0.000209 7 0 1 299570.11668 0.00000034 0.002141 8 0 1 299570.11684 0.00000003 0.001975 The SAS System 10:18 Tuesday, October 13, 2015 84 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 299570.11681 0.00000000 0.001972 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 299570.1 Generalized Chi-Square 33557.06 Gener. Chi-Square / DF 0.56 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.4443 0.006796 Residual 0.5621 0.004199 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -4.6421 0.1464 24352 -31.71 <.0001 Age 0.04330 0.002136 35352 20.27 <.0001 Year -0.01163 0.004782 35352 -2.43 0.0150 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 35352 410.91 <.0001 Year 1 35352 5.92 0.0150 The SAS System 10:18 Tuesday, October 13, 2015 85 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable OUT_Diet Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 115521 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 3 466806.79229 0.89772338 0.022721 1 0 2 454379.06694 0.01806321 0.00296 2 0 2 454247.92817 0.00111487 0.000058 3 0 2 454253.8113 0.00003637 0.000025 4 0 1 454254.19005 0.00000067 0.00111 5 0 1 454254.20265 0.00000012 0.000421 6 0 1 454254.20302 0.00000018 0.000527 7 0 1 454254.20336 0.00000001 0.000505 The SAS System 10:18 Tuesday, October 13, 2015 90 The GLIMMIX Procedure Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 454254.2 Generalized Chi-Square 31138.15 Gener. Chi-Square / DF 0.27 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.7484 0.006816 Residual 0.2696 0.001334 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -0.1576 0.06863 33287 -2.30 0.0217 Age -0.00202 0.001205 82231 -1.67 0.0943 Year -0.07857 0.001977 82231 -39.74 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 82231 2.80 0.0943 Year 1 82231 1579.26 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 91 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable OUT_DMoralMed Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 115521 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 2 432025.38439 0.18305405 9.277E-6 1 0 2 427874.5306 0.00669534 1.435E-6 2 0 2 427908.94152 0.00001368 5.453E-9 3 0 1 427908.95293 0.00000127 0.0028 4 0 1 427908.95339 0.00000082 0.000926 5 0 1 427908.9531 0.00000151 0.002517 6 0 1 427908.95257 0.00000005 0.002393 7 0 1 427908.95258 0.00000389 0.006495 8 0 1 427908.95396 0.00000345 0.001393 The SAS System 10:18 Tuesday, October 13, 2015 92 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 427908.95274 0.00000000 0.001389 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 427909.0 Generalized Chi-Square 37780.24 Gener. Chi-Square / DF 0.33 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.6827 0.006438 Residual 0.3271 0.001614 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 0.5678 0.05923 33287 9.59 <.0001 Age -0.00269 0.001008 82231 -2.67 0.0077 Year 0.000374 0.001832 82231 0.20 0.8383 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 82231 7.11 0.0077 Year 1 82231 0.04 0.8383 The SAS System 10:18 Tuesday, October 13, 2015 93 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable OUT_AntihypertensiveMed Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 115521 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 3 442340.04144 0.13460690 0.013869 1 0 2 432904.35773 0.01570213 0.002691 2 0 1 433339.59468 0.00092423 0.008975 3 0 2 433337.05102 0.00001572 5.651E-7 4 0 0 433336.78551 . 0.002304 5 0 6 433336.78551 2.00000000 0.001114 6 0 1 433336.79311 0.00000044 0.001738 7 0 1 433336.79375 0.00000067 0.00266 8 0 1 433336.79282 0.00000032 0.000587 The SAS System 10:18 Tuesday, October 13, 2015 94 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 433336.79323 0.00000000 0.000591 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 433336.8 Generalized Chi-Square 33082.25 Gener. Chi-Square / DF 0.29 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.7360 0.006795 Residual 0.2864 0.001419 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -3.8145 0.06336 33287 -60.20 <.0001 Age 0.05796 0.001099 82231 52.73 <.0001 Year 0.06672 0.001814 82231 36.78 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 82231 2780.59 <.0001 Year 1 82231 1352.84 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 95 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable OUT_LipidLoweringMed Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 115521 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 4 449704.64042 0.10664399 0.000064 1 0 2 443846.24579 0.00884382 0.011139 2 0 2 443754.25721 0.00015555 0.000191 3 0 2 443751.05174 0.00000399 2.457E-9 4 0 1 443750.92367 0.00000107 0.006569 5 0 1 443750.91957 0.00000051 0.002468 6 0 1 443750.91932 0.00000025 0.000473 7 0 1 443750.9193 0.00000000 0.00047 The SAS System 10:18 Tuesday, October 13, 2015 96 The GLIMMIX Procedure Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 443750.9 Generalized Chi-Square 40852.99 Gener. Chi-Square / DF 0.35 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.6722 0.006521 Residual 0.3537 0.001752 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -3.2070 0.06188 33287 -51.83 <.0001 Age 0.02827 0.001033 82231 27.37 <.0001 Year 0.1525 0.001971 82231 77.37 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 82231 749.34 <.0001 Year 1 82231 5986.84 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 1 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 571299.5 Generalized Chi-Square 130366.7 Gener. Chi-Square / DF 1.00 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0018 0.003928 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -2.9313 0.04915 37765 -59.64 <.0001 Age 0.05260 0.000689 92364 76.31 <.0001 Year 0.02911 0.002011 92364 14.47 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 92364 5823.52 <.0001 Year 1 92364 209.48 <.0001 The SAS System 09:52 Thursday, October 15, 2015 39 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable OUT_microalbuminuria Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 93292 Dimensions R-side Cov. Parameters 1 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique None Parameters 0 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 0 453375.94877 0.64247236 . 1 0 0 476892.68277 0.07971855 . 2 0 0 479199.65296 0.00118349 . 3 0 0 479223.49531 0.00000020 . 4 0 0 479223.49858 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 09:52 Thursday, October 15, 2015 40 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 479223.5 Generalized Chi-Square 93510.00 Gener. Chi-Square / DF 1.00 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0024 0.004641 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -2.2691 0.08901 29287 -25.49 <.0001 Age 0.01548 0.001216 64002 12.73 <.0001 Year -0.05755 0.003461 64002 -16.63 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 64002 161.98 <.0001 Year 1 64002 276.51 <.0001 The SAS System 09:52 Thursday, October 15, 2015 41 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable OUT_EyeExamination Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 80223 Dimensions R-side Cov. Parameters 1 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 14 Optimization Information Optimization Technique None Parameters 0 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 0 370676.94188 0.93058785 . 1 0 0 441359.98123 0.41461011 . 2 0 0 466383.07122 0.07389941 . 3 0 0 469479.26562 0.00174834 . 4 0 0 469537.21169 0.00000083 . 5 0 0 469537.23731 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 09:52 Thursday, October 15, 2015 42 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 469537.2 Generalized Chi-Square 80460.88 Gener. Chi-Square / DF 1.00 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0030 0.005008 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -4.0141 0.1485 26606 -27.03 <.0001 Age 0.03132 0.002005 53614 15.62 <.0001 Year -0.06211 0.005493 53614 -11.31 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 53614 244.09 <.0001 Year 1 53614 127.86 <.0001 The SAS System 09:52 Thursday, October 15, 2015 43 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable PRO_EyeExamination Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 115521 Dimensions R-side Cov. Parameters 1 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique None Parameters 0 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 0 538054.52858 0.22057665 . 1 0 0 545893.60209 0.00360250 . 2 0 0 546190.86715 0.00000299 . 3 0 0 546191.18044 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 09:52 Thursday, October 15, 2015 44 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 546191.2 Generalized Chi-Square 117862.6 Gener. Chi-Square / DF 1.02 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0203 0.004245 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -3.2540 0.06000 33287 -54.23 <.0001 Age 0.01179 0.000804 82231 14.68 <.0001 Year 0.2926 0.002583 82231 113.26 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 82231 215.47 <.0001 Year 1 82231 12828.2 <.0001 The SAS System 09:52 Thursday, October 15, 2015 45 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable PRO_FeetExamination Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 115521 Dimensions R-side Cov. Parameters 1 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique None Parameters 0 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 0 557546.07596 0.51959132 . 1 0 0 580380.17958 0.03518125 . 2 0 0 582623.85507 0.00027188 . 3 0 0 582637.13827 0.00000001 . 4 0 0 582637.13876 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 09:52 Thursday, October 15, 2015 46 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 582637.1 Generalized Chi-Square 116421.3 Gener. Chi-Square / DF 1.01 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0078 0.004193 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -0.3853 0.06658 33287 -5.79 <.0001 Age 0.009746 0.000942 82231 10.34 <.0001 Year 0.1324 0.002531 82231 52.31 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 82231 106.94 <.0001 Year 1 82231 2736.58 <.0001 The SAS System 09:52 Thursday, October 15, 2015 48 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable OUT_Insulin Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 115521 Dimensions R-side Cov. Parameters 1 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique None Parameters 0 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 0 561873.24768 0.54960194 . 1 0 0 575232.50155 0.03409065 . 2 0 0 575808.71206 0.00012414 . 3 0 0 575810.35628 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 09:52 Thursday, October 15, 2015 49 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable HbA1c Response Distribution Gaussian Link Function Identity Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Restricted Maximum Likelihood Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Gender 1 1 Number of Observations Read 115521 Number of Observations Used 110314 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 0 0 4 784318.0187 . 5679.288 1 0 4 783290.6071 1027.4115928 1400.029 2 0 4 783051.64203 238.96507585 171.6098 3 0 2 783049.4022 2.23982383 50.15209 4 0 2 783049.18242 0.21978022 1.320952 5 0 2 783049.18227 0.00015098 0.010505 The SAS System 10:00 Thursday, October 22, 2015 12 The GLIMMIX Procedure Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 783049.2 AIC (smaller is better) 783053.2 AICC (smaller is better) 783053.2 BIC (smaller is better) 783070.0 CAIC (smaller is better) 783072.0 HQIC (smaller is better) 783058.6 Generalized Chi-Square 4828266 Gener. Chi-Square / DF 43.77 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 64.9911 0.6879 Residual 43.7696 0.2240 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 50.6475 0.3112 32113 162.77 <.0001 Age -0.02989 0.005123 78198 -5.83 <.0001 Year 0.1087 0.01020 78198 10.65 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 78198 34.03 <.0001 Year 1 78198 113.47 <.0001 The SAS System 10:00 Thursday, October 22, 2015 13 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable HbA1c Response Distribution Gaussian Link Function Identity Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Restricted Maximum Likelihood Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Gender 1 1 Number of Observations Read 115521 Number of Observations Used 110314 Dimensions R-side Cov. Parameters 2 Columns in X 4 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 0 0 4 783499.35074 . 5972.62 1 0 4 782449.52099 1049.8297426 1541.485 2 0 4 782152.42152 297.09946861 234.326 3 0 2 782148.45839 3.96313799 76.75306 4 0 2 782147.94736 0.51102660 2.727433 5 0 2 782147.94672 0.00063604 0.033185 The SAS System 10:00 Thursday, October 22, 2015 14 The GLIMMIX Procedure Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 782147.9 AIC (smaller is better) 782151.9 AICC (smaller is better) 782151.9 BIC (smaller is better) 782168.8 CAIC (smaller is better) 782170.8 HQIC (smaller is better) 782157.3 Generalized Chi-Square 4797158 Gener. Chi-Square / DF 43.49 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 64.0606 0.6793 Residual 43.4880 0.2226 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 57.3991 0.3812 32113 150.56 <.0001 Age -0.03503 0.005094 78197 -6.88 <.0001 Year -1.1669 0.04333 78197 -26.93 <.0001 YearSQ 0.05709 0.001886 78197 30.27 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 78197 47.27 <.0001 Year 1 78197 725.33 <.0001 YearSQ 1 78197 916.12 <.0001 The SAS System 10:00 Thursday, October 22, 2015 15 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable SBP Response Distribution Gaussian Link Function Identity Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Restricted Maximum Likelihood Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Gender 1 1 Number of Observations Read 115521 Number of Observations Used 112182 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 0 0 4 920893.46634 . 1698.098 1 0 4 920820.43546 73.03088770 157.5789 2 0 2 920819.65963 0.77582566 22.05577 3 0 2 920819.64441 0.01522123 0.344028 4 0 2 920819.64441 0.00000370 0.000737 Convergence criterion (GCONV=1E-8) satisfied. The SAS System 10:00 Thursday, October 22, 2015 16 The GLIMMIX Procedure Fit Statistics -2 Res Log Likelihood 920819.6 AIC (smaller is better) 920823.6 AICC (smaller is better) 920823.6 BIC (smaller is better) 920840.5 CAIC (smaller is better) 920842.5 HQIC (smaller is better) 920829.0 Generalized Chi-Square 16913723 Gener. Chi-Square / DF 150.77 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 123.25 1.4527 Residual 150.77 0.7527 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 122.08 0.4907 32608 248.77 <.0001 Age 0.4478 0.007673 79571 58.36 <.0001 Year -0.9421 0.01722 79571 -54.71 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 79571 3405.69 <.0001 Year 1 79571 2993.28 <.0001 The SAS System 10:00 Thursday, October 22, 2015 17 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable BMI Response Distribution Gaussian Link Function Identity Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Restricted Maximum Likelihood Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Gender 1 1 Number of Observations Read 115521 Number of Observations Used 106284 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 0 0 4 518304.452 . 1809.746 1 0 4 514686.03256 3618.4194438 293.6135 2 0 4 514429.70561 256.32695223 22.29593 3 0 2 514427.7722 1.93340571 4.930689 4 0 2 514427.6703 0.10190371 0.112194 5 0 2 514427.67025 0.00005320 0.000581 The SAS System 10:00 Thursday, October 22, 2015 18 The GLIMMIX Procedure Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 514427.7 AIC (smaller is better) 514431.7 AICC (smaller is better) 514431.7 BIC (smaller is better) 514448.5 CAIC (smaller is better) 514450.5 HQIC (smaller is better) 514437.0 Generalized Chi-Square 273217.1 Gener. Chi-Square / DF 2.57 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 33.8374 0.2807 Residual 2.5707 0.01328 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 36.1491 0.1688 31241 214.13 <.0001 Age -0.09950 0.003223 75040 -30.87 <.0001 Year 0.07895 0.004062 75040 19.44 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 75040 952.84 <.0001 Year 1 75040 377.74 <.0001 The SAS System 10:00 Thursday, October 22, 2015 19 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable CholHDL Response Distribution Gaussian Link Function Identity Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Restricted Maximum Likelihood Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Gender 1 1 Number of Observations Read 115521 Number of Observations Used 108682 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 0 0 4 288958.93772 . 4032.219 1 0 4 287899.55533 1059.3823915 387.8939 2 0 4 287877.62986 21.92546666 72.17573 3 0 2 287876.94252 0.68734241 5.833715 4 0 2 287876.93791 0.00461022 0.079817 5 0 2 287876.93791 0.00000087 0.00009 The SAS System 10:00 Thursday, October 22, 2015 20 The GLIMMIX Procedure Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 287876.9 AIC (smaller is better) 287880.9 AICC (smaller is better) 287880.9 BIC (smaller is better) 287897.8 CAIC (smaller is better) 287899.8 HQIC (smaller is better) 287886.3 Generalized Chi-Square 51609.91 Gener. Chi-Square / DF 0.47 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.9538 0.009322 Residual 0.4749 0.002433 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 5.2790 0.03568 31993 147.95 <.0001 Age -0.01721 0.000603 76686 -28.52 <.0001 Year -0.03732 0.001136 76686 -32.84 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 76686 813.16 <.0001 Year 1 76686 1078.42 <.0001 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable Age Response Distribution Gaussian Link Function Identity Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Restricted Maximum Likelihood Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 115521 Dimensions R-side Cov. Parameters 2 Columns in X 2 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 0 0 4 640220.58102 . 19103.79 1 0 2 507193.00241 133027.57862 3162.423 2 0 6 418318.04097 88874.961432 630.2859 3 0 6 372339.14513 45978.895838 116.1032 4 0 4 362167.72665 10171.418486 43.45087 5 0 4 357559.9244 4607.8022519 10.7761 6 0 3 356898.14231 661.78208670 2.875585 7 0 6 356772.01153 126.13078255 0.096646 8 0 3 356771.90276 0.10876754 0.019108 The SAS System 12:36 Sunday, October 25, 2015 16 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 9 0 3 356771.89837 0.00438597 0.000113 10 0 3 356771.89837 0.00000017 1.182E-7 Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 356771.9 AIC (smaller is better) 356775.9 AICC (smaller is better) 356775.9 BIC (smaller is better) 356792.7 CAIC (smaller is better) 356794.7 HQIC (smaller is better) 356781.3 Generalized Chi-Square 16484.18 Gener. Chi-Square / DF 0.14 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 107.64 0.8351 Residual 0.1427 0.000704 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 49.3243 0.05745 33287 858.60 <.0001 Year 0.9855 0.000616 82232 1600.86 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Year 1 82232 2562750 <.0001 The SAS System 12:36 Sunday, October 25, 2015 17 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable Age Response Distribution Gaussian Link Function Identity Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Restricted Maximum Likelihood Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 115521 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 0 0 4 639923.93413 . 19042.27 1 0 2 507302.73723 132621.19689 3169.979 2 0 6 418294.66687 89008.070366 631.448 3 0 6 372216.94976 46077.717107 116.3231 4 0 4 362002.76695 10214.182812 43.46927 5 0 4 357383.33938 4619.4275673 10.78777 6 0 3 356718.1935 665.14588327 2.873888 7 0 6 356591.66936 126.52414162 0.096623 8 0 3 356591.56023 0.10913034 0.019137 The SAS System 12:36 Sunday, October 25, 2015 18 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 9 0 3 356591.55581 0.00441655 0.000114 10 0 70 356591.55581 0.00000190 0.000114 Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 356591.6 AIC (smaller is better) 356595.6 AICC (smaller is better) 356595.6 BIC (smaller is better) 356612.4 CAIC (smaller is better) 356614.4 HQIC (smaller is better) 356600.9 Generalized Chi-Square 16446.86 Gener. Chi-Square / DF 0.14 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 107.60 0.8348 Residual 0.1424 0.000702 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 49.5085 0.05892 33287 840.27 <.0001 Year 0.9494 0.002651 82231 358.08 <.0001 YearSQ 0.001605 0.000114 82231 14.03 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Year 1 82231 128225 <.0001 YearSQ 1 82231 196.87 <.0001 The SAS System 12:36 Sunday, October 25, 2015 19 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable Diabetesduration Response Distribution Gaussian Link Function Identity Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Restricted Maximum Likelihood Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 111986 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 0 0 4 489363.85706 . 6474.905 1 0 2 466679.25073 22684.606326 1794.573 2 0 4 458556.9534 8122.2973306 463.2536 3 0 4 456738.54178 1818.4116246 76.60501 4 0 4 456634.85872 103.68305704 3.120915 5 0 2 456634.64749 0.21122807 0.468726 6 0 3 456634.64266 0.00483228 0.003511 7 0 3 456634.64266 0.00000032 7.81E-6 The SAS System 12:36 Sunday, October 25, 2015 20 The GLIMMIX Procedure Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 456634.6 AIC (smaller is better) 456638.6 AICC (smaller is better) 456638.6 BIC (smaller is better) 456655.5 CAIC (smaller is better) 456657.5 HQIC (smaller is better) 456644.0 Generalized Chi-Square 107230.1 Gener. Chi-Square / DF 0.96 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 29.6730 0.2388 Residual 0.9576 0.004811 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -18.0065 0.1451 32299 -124.10 <.0001 Age 0.2140 0.002820 79684 75.88 <.0001 Year 0.7806 0.003137 79684 248.81 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 79684 5757.04 <.0001 Year 1 79684 61907.5 <.0001 The SAS System 12:36 Sunday, October 25, 2015 21 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_FEMALE_UNDER_75 Response Variable Diabetesduration Response Distribution Gaussian Link Function Identity Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Restricted Maximum Likelihood Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 33288 not printed Number of Observations Read 115521 Number of Observations Used 111986 Dimensions R-side Cov. Parameters 2 Columns in X 4 Columns in Z per Subject 0 Subjects (Blocks in V) 33288 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 0 0 4 490036.28397 . 6775.944 1 0 2 465906.45875 24129.825216 1764.121 2 0 4 457931.10913 7975.3496273 462.6741 3 0 4 456105.48512 1825.6240078 76.08935 4 0 4 456002.37339 103.11172687 3.034729 5 0 2 456002.17224 0.20115294 0.454438 6 0 3 456002.16766 0.00457507 0.003321 7 0 3 456002.16766 0.00000023 3.601E-6 The SAS System 12:36 Sunday, October 25, 2015 22 The GLIMMIX Procedure Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 456002.2 AIC (smaller is better) 456006.2 AICC (smaller is better) 456006.2 BIC (smaller is better) 456023.0 CAIC (smaller is better) 456025.0 HQIC (smaller is better) 456011.5 Generalized Chi-Square 106498.1 Gener. Chi-Square / DF 0.95 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 29.5822 0.2380 Residual 0.9510 0.004778 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -17.0557 0.1496 32299 -114.00 <.0001 Age 0.2124 0.002816 79683 75.43 <.0001 Year 0.6107 0.007370 79683 82.86 <.0001 YearSQ 0.007607 0.000299 79683 25.47 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 79683 5689.53 <.0001 Year 1 79683 6865.54 <.0001 YearSQ 1 79683 648.89 <.0001 The SAS System 12:36 Sunday, October 25, 2015 23 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable OUT_smoking Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 123556 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 3 470463.70743 0.20337199 0.000981 1 0 2 459035.48988 0.00288280 0.000324 2 0 1 459250.45752 0.00008665 9.955E-6 3 0 1 459251.50697 0.00000187 0.001925 4 0 1 459251.51824 0.00000048 0.000709 5 0 1 459251.5178 0.00000000 0.000702 Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 10:18 Tuesday, October 13, 2015 2 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 459251.5 Generalized Chi-Square 25575.49 Gener. Chi-Square / DF 0.21 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.7934 0.006630 Residual 0.2070 0.000989 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 1.0668 0.06273 35933 17.01 <.0001 Age -0.03097 0.001171 87620 -26.45 <.0001 Year -0.02772 0.001836 87620 -15.10 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 87620 699.68 <.0001 Year 1 87620 227.91 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 3 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable PRO_Smoking Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 130132 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 3 563641.42337 0.86780857 0.017544 1 0 4 656850.29057 0.31396290 0.000535 2 0 4 684207.07266 0.10432669 0.003701 3 0 2 683730.50095 0.01691122 0.02775 4 0 2 683241.49987 0.00154816 0.000026 5 0 2 683195.04618 0.00011796 3.06E-8 6 0 2 683191.42089 0.00000888 0.000038 7 0 1 683191.14759 0.00000059 0.000467 8 0 1 683191.12879 0.00000008 0.000261 The SAS System 10:18 Tuesday, October 13, 2015 4 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 683191.12668 0.00000027 0.001682 10 0 1 683191.12064 0.00000155 0.009583 11 0 1 683191.1548 0.00000129 0.000079 12 0 1 683191.12655 0.00000000 0.000213 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 683191.1 Generalized Chi-Square 58826.43 Gener. Chi-Square / DF 0.45 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.4811 0.006417 Residual 0.4521 0.002337 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -0.4734 0.09358 37765 -5.06 <.0001 Age 0.01543 0.001598 92364 9.66 <.0001 Year 0.1693 0.002935 92364 57.69 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 92364 93.29 <.0001 Year 1 92364 3327.69 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 5 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable OUT_microalbuminuria Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 105466 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 2 471064.8269 0.39623062 0.015893 1 0 2 468354.8797 0.00619694 0.016168 2 0 2 468717.30311 0.00004768 2.276E-8 3 0 1 468716.426 0.00000039 0.000032 4 0 1 468716.42207 0.00000019 0.001989 5 0 1 468716.42198 0.00000019 0.000085 6 0 1 468716.42199 0.00000006 0.000587 7 0 1 468716.42204 0.00000056 0.005076 8 0 1 468716.42191 0.00000049 0.000216 The SAS System 10:18 Tuesday, October 13, 2015 6 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 468716.42193 0.00000016 0.001498 10 0 1 468716.42206 0.00000142 0.012966 11 0 2 468716.42174 0.00000129 0.000098 12 0 0 468716.42178 . 0.000013 13 0 6 468716.42178 2.00000000 0.000014 14 0 1 468716.42203 0.00000121 0.01231 15 0 2 468716.42173 0.00000125 0.000157 16 0 1 468716.4218 0.00000026 0.002259 17 0 1 468716.42198 0.00000000 0.002259 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 468716.4 Generalized Chi-Square 50611.75 Gener. Chi-Square / DF 0.48 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.5291 0.005848 Residual 0.4799 0.002524 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -3.3844 0.08803 33190 -38.45 <.0001 Age 0.03375 0.001418 72273 23.79 <.0001 Year -0.01438 0.002938 72273 -4.89 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 72273 566.18 <.0001 Year 1 72273 23.94 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 7 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable OUT_macroalbuminuria Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 88530 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 3 286523.00466 0.99903830 0.000138 1 0 3 384741.56283 0.56420887 0.000159 2 0 3 434378.11792 0.17735993 0.000026 3 0 3 448522.22431 0.01477451 9.589E-6 4 0 2 449630.14617 0.00229603 0.000159 5 0 2 449611.53569 0.00049651 6.93E-10 6 0 2 449606.16759 0.00010169 8.9E-10 7 0 1 449605.06752 0.00002097 0.001633 8 0 1 449604.84027 0.00000415 0.000231 The SAS System 10:18 Tuesday, October 13, 2015 8 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 449604.79559 0.00000077 0.000584 10 0 1 449604.78741 0.00000016 0.000556 11 0 1 449604.78567 0.00000074 0.001055 12 0 1 449604.78329 0.00000142 0.00202 13 0 1 449604.78673 0.00000116 0.000484 14 0 1 449604.78416 0.00000004 0.000448 15 0 1 449604.78379 0.00000002 0.000399 16 0 1 449604.78377 0.00000001 0.000387 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 449604.8 Generalized Chi-Square 10353.33 Gener. Chi-Square / DF 0.12 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.8721 0.007795 Residual 0.1170 0.000694 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -6.0803 0.1979 30252 -30.73 <.0001 Age 0.05550 0.003529 58275 15.73 <.0001 Year -0.04868 0.004609 58275 -10.56 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 58275 247.34 <.0001 Year 1 58275 111.55 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 9 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable PRO_AlbKreat Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 130132 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 4 601057.10417 0.36248477 0.022038 1 0 3 606547.16957 0.03915227 0.002301 2 0 2 606407.46461 0.00171101 0.000267 3 0 2 606378.21338 0.00001760 0.000061 4 0 2 606377.80025 0.00000100 7.806E-9 5 0 1 606377.81672 0.00000001 0.000394 Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 10:18 Tuesday, October 13, 2015 10 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 606377.8 Generalized Chi-Square 93722.71 Gener. Chi-Square / DF 0.72 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.2350 0.004472 Residual 0.7202 0.003534 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -1.7720 0.06460 37765 -27.43 <.0001 Age 0.01411 0.000995 92364 14.17 <.0001 Year 0.1870 0.002379 92364 78.61 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 92364 200.87 <.0001 Year 1 92364 6180.14 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 11 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable OUT_HbA1c Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 123896 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 2 540696.42412 0.13332629 0.004705 1 0 2 526381.34164 0.00678661 0.000091 2 0 2 526797.16912 0.00028908 0.00001 3 0 2 526794.78471 0.00001240 5.671E-9 4 0 1 526794.89963 0.00000065 0.002137 5 0 0 526794.89456 . 0.002212 6 0 5 526794.89456 1.99999947 0.001424 7 0 1 526794.89506 0.00000003 0.001092 8 0 1 526794.89496 0.00000002 0.000822 The SAS System 10:18 Tuesday, October 13, 2015 12 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 526794.89495 0.00000127 0.019311 10 0 1 526794.89359 0.00000300 0.028234 11 0 1 526794.89718 0.00000135 0.006981 12 0 1 526794.89484 0.00000024 0.003037 13 0 1 526794.89501 0.00000005 0.002279 14 0 1 526794.89501 0.00000000 0.00227 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 526794.9 Generalized Chi-Square 76239.72 Gener. Chi-Square / DF 0.62 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.3989 0.004817 Residual 0.6154 0.002925 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 0.6485 0.06242 36369 10.39 <.0001 Age 0.009562 0.000981 87524 9.75 <.0001 Year -0.02590 0.002342 87524 -11.06 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 87524 95.09 <.0001 Year 1 87524 122.30 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 13 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable PRO_HbA1c Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 130132 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 4 579221.39551 0.91244895 0.005369 1 0 5 697856.80911 0.38242081 0.002751 2 0 4 745270.02567 0.24166291 0.002531 3 0 2 746219.37424 0.03327232 0.272271 4 0 2 745341.251 0.00050702 5.706E-7 5 0 1 745327.32885 0.00000784 0.025773 6 0 2 745327.51276 0.00000421 0.000538 7 0 1 745327.41355 0.00000001 0.001115 The SAS System 10:18 Tuesday, October 13, 2015 14 The GLIMMIX Procedure Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 745327.4 Generalized Chi-Square 92555.92 Gener. Chi-Square / DF 0.71 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.1781 0.005863 Residual 0.7113 0.004221 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -0.4255 0.09577 37765 -4.44 <.0001 Age 0.02111 0.001516 92364 13.93 <.0001 Year 0.1620 0.003342 92364 48.47 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 92364 193.93 <.0001 Year 1 92364 2349.60 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 15 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable OUT_SBP Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 126342 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 3 524032.36799 0.16318067 0.000278 1 0 2 520967.14882 0.00413643 0.001091 2 0 2 521024.1441 0.00007462 0.000042 3 0 1 521024.13547 0.00000147 0.02632 4 0 2 521024.13535 0.00000122 0.003973 5 0 1 521024.1356 0.00000022 0.000166 6 0 1 521024.13565 0.00000034 0.006102 7 0 1 521024.13558 0.00000124 0.016614 8 0 1 521024.13583 0.00000096 0.000949 The SAS System 10:18 Tuesday, October 13, 2015 16 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 521024.13563 0.00000019 0.002606 10 0 1 521024.13567 0.00000000 0.002598 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 521024.1 Generalized Chi-Square 84881.46 Gener. Chi-Square / DF 0.67 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.3313 0.004322 Residual 0.6719 0.003136 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 0.9765 0.05757 36892 16.96 <.0001 Age -0.03106 0.000901 89447 -34.48 <.0001 Year 0.08163 0.002148 89447 38.00 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 89447 1188.58 <.0001 Year 1 89447 1443.83 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 17 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable Out_Obesity Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 120003 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 2 420634.75842 0.16518294 0.000041 1 0 2 414726.23373 0.00491821 0.000042 2 0 2 414831.87524 0.00000418 6.618E-8 3 0 1 414831.95014 0.00000028 0.001393 4 0 1 414831.95001 0.00000038 0.000642 5 0 1 414831.95007 0.00000000 0.000641 Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 10:18 Tuesday, October 13, 2015 20 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 414832.0 Generalized Chi-Square 25589.38 Gener. Chi-Square / DF 0.21 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.7883 0.006644 Residual 0.2132 0.001036 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 1.1246 0.05800 35492 19.39 <.0001 Age -0.03052 0.001068 84508 -28.58 <.0001 Year 0.02421 0.001738 84508 13.94 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 84508 816.97 <.0001 Year 1 84508 194.22 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 21 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable Out_Overweight Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 120003 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 2 444860.97405 0.18394245 0.004357 1 0 2 443428.42773 0.00255749 0.004717 2 0 2 443442.15453 0.00001317 8.139E-9 3 0 1 443442.15611 0.00000019 0.001543 4 0 1 443442.15611 0.00000021 0.000272 5 0 1 443442.15611 0.00000000 0.00026 Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 10:18 Tuesday, October 13, 2015 22 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 443442.2 Generalized Chi-Square 39889.54 Gener. Chi-Square / DF 0.33 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.6727 0.006139 Residual 0.3324 0.001613 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -1.0919 0.05900 35492 -18.51 <.0001 Age 0.01941 0.001025 84508 18.94 <.0001 Year -0.02039 0.001907 84508 -10.69 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 84508 358.80 <.0001 Year 1 84508 114.35 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 23 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable OUT_BMI Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 120003 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 2 501608.0618 0.59740259 0.000815 1 0 2 510959.32465 0.04927920 0.001313 2 0 2 511638.51503 0.00042242 1.481E-6 3 0 2 511641.63986 0.00000301 1.824E-9 4 0 1 511641.63833 0.00000002 0.000113 5 0 1 511641.63831 0.00000000 0.000113 Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 10:18 Tuesday, October 13, 2015 24 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 511641.6 Generalized Chi-Square 32921.18 Gener. Chi-Square / DF 0.27 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.7330 0.006437 Residual 0.2743 0.001334 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -3.0569 0.08732 35492 -35.01 <.0001 Age 0.02206 0.001527 84508 14.45 <.0001 Year -0.00595 0.002577 84508 -2.31 0.0209 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 84508 208.74 <.0001 Year 1 84508 5.33 0.0209 The SAS System 10:18 Tuesday, October 13, 2015 25 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable PRO_BMI Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 130132 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 4 581229.80237 0.82513109 0.034981 1 0 4 650814.93035 0.24906967 0.026753 2 0 4 665675.71318 0.06183727 0.000497 3 0 3 664570.69894 0.00475278 0.000171 4 0 2 664408.83078 0.00005460 8.234E-6 5 0 2 664405.83342 0.00001155 6.904E-6 6 0 1 664406.19208 0.00000008 0.005047 7 0 1 664406.20243 0.00000000 0.004954 The SAS System 10:18 Tuesday, October 13, 2015 26 The GLIMMIX Procedure Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 664406.2 Generalized Chi-Square 68919.70 Gener. Chi-Square / DF 0.53 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.3869 0.005728 Residual 0.5296 0.002713 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -1.0666 0.08333 37765 -12.80 <.0001 Age 0.009350 0.001380 92364 6.78 <.0001 Year 0.2218 0.002740 92364 80.94 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 92364 45.92 <.0001 Year 1 92364 6552.02 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 27 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable OUT_CholHDL Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 122348 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 2 491571.82718 0.17019870 0.041207 1 0 2 487695.10877 0.00350798 0.000139 2 0 2 487741.65304 0.00003754 2.072E-9 3 0 1 487741.80425 0.00000073 0.001434 4 0 1 487741.80349 0.00000153 0.019089 5 0 1 487741.80374 0.00000198 0.007533 6 0 1 487741.80345 0.00000078 0.002976 7 0 1 487741.80357 0.00000031 0.001176 8 0 1 487741.80352 0.00000012 0.000464 The SAS System 10:18 Tuesday, October 13, 2015 28 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 487741.80354 0.00000005 0.000184 10 0 1 487741.80353 0.00000000 0.000158 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 487741.8 Generalized Chi-Square 62585.95 Gener. Chi-Square / DF 0.51 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.4946 0.005250 Residual 0.5116 0.002452 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -2.7026 0.05994 36335 -45.09 <.0001 Age 0.03900 0.000972 86010 40.13 <.0001 Year 0.03774 0.002068 86010 18.25 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 86010 1610.52 <.0001 Year 1 86010 333.06 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 29 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable PRO_CholHDL Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 130132 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 3 591230.49064 0.86587510 0.268367 1 0 4 695997.05455 0.31064692 0.017606 2 0 3 736817.13041 0.08851699 0.068831 3 0 2 740015.89664 0.01098572 0.005142 4 0 2 739857.64578 0.00035188 9.246E-7 5 0 2 739851.35708 0.00000701 0.001487 6 0 1 739851.2274 0.00000014 0.000432 7 0 1 739851.2294 0.00000000 0.000415 The SAS System 10:18 Tuesday, October 13, 2015 30 The GLIMMIX Procedure Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 739851.2 Generalized Chi-Square 105666.1 Gener. Chi-Square / DF 0.81 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.1227 0.004435 Residual 0.8120 0.004290 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -0.4799 0.09054 37765 -5.30 <.0001 Age 0.01446 0.001405 92364 10.29 <.0001 Year 0.1871 0.003158 92364 59.23 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 92364 105.97 <.0001 Year 1 92364 3508.33 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 31 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable OUT_EyeExamination Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 91398 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 15 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 2 404140.71559 0.89056200 0.000015 1 0 2 478275.52365 0.33062221 0.000255 2 0 2 499198.56191 0.03544392 0.000037 3 0 2 500170.05168 0.00137099 0.303655 4 0 2 500189.84572 0.00020631 6.106E-6 5 0 1 500187.79887 0.00003298 0.00013 6 0 1 500188.09465 0.00000471 0.00423 7 0 1 500188.0495 0.00000044 0.000191 8 0 1 500188.05443 0.00000018 0.001575 The SAS System 10:18 Tuesday, October 13, 2015 32 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 500188.0548 0.00000004 0.001425 10 0 1 500188.05451 0.00000001 0.001294 11 0 0 500188.05452 . 0.001294 12 0 5 500188.05452 2.00000000 0.033295 13 0 2 500188.04043 0.00000462 8.905E-9 14 0 1 500188.05965 0.00000061 0.006098 15 0 1 500188.0558 0.00000068 0.00043 16 0 1 500188.0535 0.00000006 0.000034 17 0 1 500188.05426 0.00000001 0.000034 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 500188.1 Generalized Chi-Square 45172.04 Gener. Chi-Square / DF 0.49 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.5400 0.006089 Residual 0.4943 0.002807 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -4.2849 0.1597 30711 -26.83 <.0001 Age 0.01900 0.002421 60684 7.85 <.0001 Year 0.02529 0.006177 60684 4.09 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 60684 61.63 <.0001 Year 1 60684 16.76 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 33 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable PRO_FeetExamination Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 130132 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 3 603713.09857 0.43898134 0.000732 1 0 3 616343.88168 0.02433023 0.000518 2 0 2 616907.99541 0.00172921 0.000235 3 0 2 616873.393 0.00008578 0.000033 4 0 1 616871.71003 0.00000141 0.020765 5 0 1 616871.67164 0.00000508 0.043706 6 0 2 616871.60872 0.00000336 1.454E-8 7 0 1 616871.64991 0.00000031 0.003819 8 0 1 616871.65365 0.00000088 0.007582 The SAS System 10:18 Tuesday, October 13, 2015 38 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 9 0 1 616871.64285 0.00000175 0.015022 10 0 2 616871.66425 0.00000122 0.000768 11 0 1 616871.64934 0.00000000 0.000717 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 616871.6 Generalized Chi-Square 86594.62 Gener. Chi-Square / DF 0.67 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.2875 0.004735 Residual 0.6655 0.003258 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -1.0319 0.06847 37765 -15.07 <.0001 Age 0.01698 0.001082 92364 15.69 <.0001 Year 0.1327 0.002433 92364 54.55 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 92364 246.29 <.0001 Year 1 92364 2975.16 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 39 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable OUT_Diet Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 130132 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 3 533929.61146 0.27348344 0.029547 1 0 3 524032.32433 0.07525468 0.003978 2 0 2 523580.03239 0.00015851 0.00005 3 0 2 523572.39732 0.00001657 0.000053 4 0 1 523572.16976 0.00000077 0.001771 5 0 1 523572.16179 0.00000064 0.002957 6 0 1 523572.163 0.00000003 0.002912 7 0 1 523572.16329 0.00000000 0.002892 The SAS System 10:18 Tuesday, October 13, 2015 42 The GLIMMIX Procedure Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 523572.2 Generalized Chi-Square 35890.94 Gener. Chi-Square / DF 0.28 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.7342 0.006310 Residual 0.2758 0.001287 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -0.5317 0.06843 37765 -7.77 <.0001 Age 0.001977 0.001227 92364 1.61 0.1073 Year -0.07826 0.001969 92364 -39.75 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 92364 2.59 0.1073 Year 1 92364 1579.83 <.0001 The SAS System 10:18 Tuesday, October 13, 2015 43 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable OUT_DMoralMed Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 130132 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 2 493616.06926 0.17588439 0.000012 1 0 2 486588.85471 0.01106857 1.764E-6 2 0 2 486676.23445 0.00006638 1.741E-9 3 0 1 486676.2711 0.00000502 0.013768 4 0 2 486676.27399 0.00000462 0.000454 5 0 1 486676.27117 0.00000061 0.001486 6 0 1 486676.27155 0.00000000 0.001484 The SAS System 10:18 Tuesday, October 13, 2015 44 The GLIMMIX Procedure Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 486676.3 Generalized Chi-Square 43238.26 Gener. Chi-Square / DF 0.33 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.6759 0.006031 Residual 0.3323 0.001548 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 0.6924 0.05707 37765 12.13 <.0001 Age -0.00305 0.000994 92364 -3.07 0.0022 Year 0.000588 0.001786 92364 0.33 0.7420 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 92364 9.39 0.0022 Year 1 92364 0.11 0.7420 The SAS System 10:18 Tuesday, October 13, 2015 45 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable OUT_LipidLoweringMed Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 130132 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 4 510217.00146 0.11537516 0.002108 1 0 4 502099.50521 0.01384436 0.000839 2 0 2 502117.42901 0.00023322 0.000048 3 0 2 502110.24019 0.00001319 0.000048 4 0 1 502110.46071 0.00000066 0.003223 5 0 1 502110.44225 0.00000004 0.002959 6 0 1 502110.44303 0.00000147 0.009508 7 0 2 502110.44187 0.00000112 8.366E-9 8 0 1 502110.44282 0.00000001 0.000065 The SAS System 10:18 Tuesday, October 13, 2015 48 The GLIMMIX Procedure Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 502110.4 Generalized Chi-Square 44203.21 Gener. Chi-Square / DF 0.34 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.6880 0.006253 Residual 0.3397 0.001593 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -2.7060 0.05931 37765 -45.63 <.0001 Age 0.02306 0.001029 92364 22.42 <.0001 Year 0.1522 0.001875 92364 81.15 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 92364 502.48 <.0001 Year 1 92364 6585.23 <.0001 The SAS System 09:52 Thursday, October 15, 2015 29 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable PRO_SBP Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 130132 Dimensions R-side Cov. Parameters 1 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique None Parameters 0 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 0 546798.41397 1.00921002 . 1 0 0 714855.57138 0.58669545 . 2 0 0 816810.69819 0.19513989 . 3 0 0 849349.73925 0.01704682 . 4 0 0 851963.52322 0.00011173 . 5 0 0 851978.6443 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 09:52 Thursday, October 15, 2015 30 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 851978.6 Generalized Chi-Square 134566.6 Gener. Chi-Square / DF 1.03 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0341 0.004054 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 0.5812 0.1200 37765 4.84 <.0001 Age 0.02045 0.001806 92364 11.32 <.0001 Year 0.1389 0.004514 92364 30.78 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 92364 128.20 <.0001 Year 1 92364 947.28 <.0001 The SAS System 09:52 Thursday, October 15, 2015 31 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable PRO_EyeExamination Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 130132 Dimensions R-side Cov. Parameters 1 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique None Parameters 0 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 0 607311.37394 0.22176020 . 1 0 0 617466.56856 0.00452696 . 2 0 0 617890.24409 0.00000483 . 3 0 0 617890.7884 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 09:52 Thursday, October 15, 2015 32 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 617890.8 Generalized Chi-Square 133644.5 Gener. Chi-Square / DF 1.03 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0270 0.004026 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -3.4894 0.05757 37765 -60.61 <.0001 Age 0.01466 0.000786 92364 18.65 <.0001 Year 0.2971 0.002505 92364 118.64 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 92364 347.97 <.0001 Year 1 92364 14074.9 <.0001 The SAS System 09:52 Thursday, October 15, 2015 33 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable OUT_FeetExamination Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 68145 Dimensions R-side Cov. Parameters 1 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique None Parameters 0 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 0 329398.91077 0.59505372 . 1 0 0 340165.07415 0.06315681 . 2 0 0 342027.12827 0.00113477 . 3 0 0 342063.05256 0.00000039 . 4 0 0 342063.06493 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 09:52 Thursday, October 15, 2015 34 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 342063.1 Generalized Chi-Square 67965.07 Gener. Chi-Square / DF 1.00 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 0.9974 0.005404 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -4.2962 0.1055 27925 -40.70 <.0001 Age 0.04339 0.001480 40217 29.32 <.0001 Year -0.02014 0.003581 40217 -5.62 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 40217 859.62 <.0001 Year 1 40217 31.63 <.0001 The SAS System 09:52 Thursday, October 15, 2015 35 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable OUT_Insulin Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 130132 Dimensions R-side Cov. Parameters 1 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique None Parameters 0 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 0 633282.0715 0.60693219 . 1 0 0 658231.36219 0.05874963 . 2 0 0 659979.00588 0.00050477 . 3 0 0 659990.6805 0.00000003 . 4 0 0 659990.68119 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 09:52 Thursday, October 15, 2015 36 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 659990.7 Generalized Chi-Square 130150.4 Gener. Chi-Square / DF 1.00 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0002 0.003921 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -3.1651 0.07340 37765 -43.12 <.0001 Age 0.01485 0.001010 92364 14.71 <.0001 Year 0.02194 0.002967 92364 7.39 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 92364 216.28 <.0001 Year 1 92364 54.67 <.0001 The SAS System 09:52 Thursday, October 15, 2015 37 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable OUT_AntihypertensiveMed Response Distribution Binomial Link Function Logit Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Residual PL Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 130132 Dimensions R-side Cov. Parameters 1 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique None Parameters 0 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 0 0 0 581288.26112 0.12089492 . 1 0 0 571023.46129 0.00375901 . 2 0 0 571299.2963 0.00000433 . 3 0 0 571299.528 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 09:52 Thursday, October 15, 2015 38 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 575810.4 Generalized Chi-Square 115528.5 Gener. Chi-Square / DF 1.00 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0001 0.004161 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -2.5890 0.07289 33287 -35.52 <.0001 Age 0.01219 0.000995 82231 12.25 <.0001 Year -0.00185 0.002838 82231 -0.65 0.5134 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 82231 149.99 <.0001 Year 1 82231 0.43 0.5134 The SAS System 10:00 Thursday, October 22, 2015 1 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable HbA1c Response Distribution Gaussian Link Function Identity Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Restricted Maximum Likelihood Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Gender 1 0 Number of Observations Read 130132 Number of Observations Used 123896 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 0 0 4 900541.14152 . 7432.178 1 0 4 899745.6325 795.50901742 2546.011 2 0 4 898999.92071 745.71179364 899.1843 3 0 2 898966.79932 33.12139145 430.8915 4 0 2 898953.61842 13.18089222 47.21969 5 0 2 898953.46967 0.14874971 2.870118 6 0 3 898953.46913 0.00054675 0.017789 The SAS System 10:00 Thursday, October 22, 2015 2 The GLIMMIX Procedure Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 898953.5 AIC (smaller is better) 898957.5 AICC (smaller is better) 898957.5 BIC (smaller is better) 898974.5 CAIC (smaller is better) 898976.5 HQIC (smaller is better) 898962.9 Generalized Chi-Square 6479593 Gener. Chi-Square / DF 52.30 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 70.2585 0.7197 Residual 52.2999 0.2534 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 52.5306 0.3152 36369 166.67 <.0001 Age -0.07112 0.005280 87524 -13.47 <.0001 Year 0.1737 0.01058 87524 16.41 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 87524 181.47 <.0001 Year 1 87524 269.21 <.0001 The SAS System 10:00 Thursday, October 22, 2015 3 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable HbA1c Response Distribution Gaussian Link Function Identity Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Restricted Maximum Likelihood Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Gender 1 0 Number of Observations Read 130132 Number of Observations Used 123896 Dimensions R-side Cov. Parameters 2 Columns in X 4 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 0 0 4 900138.47817 . 7553.692 1 0 5 898537.80428 1600.6738932 479.4391 2 0 4 898521.31277 16.49151082 30.80959 3 0 2 898521.25072 0.06205060 2.068153 4 0 2 898521.25043 0.00028244 0.008246 Convergence criterion (GCONV=1E-8) satisfied. The SAS System 10:00 Thursday, October 22, 2015 4 The GLIMMIX Procedure Fit Statistics -2 Res Log Likelihood 898521.3 AIC (smaller is better) 898525.3 AICC (smaller is better) 898525.3 BIC (smaller is better) 898542.3 CAIC (smaller is better) 898544.3 HQIC (smaller is better) 898530.7 Generalized Chi-Square 6462372 Gener. Chi-Square / DF 52.16 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 69.7863 0.7157 Residual 52.1613 0.2528 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 57.4653 0.3919 36369 146.64 <.0001 Age -0.07326 0.005266 87523 -13.91 <.0001 Year -0.7670 0.04587 87523 -16.72 <.0001 YearSQ 0.04178 0.001983 87523 21.06 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 87523 193.55 <.0001 Year 1 87523 279.57 <.0001 YearSQ 1 87523 443.73 <.0001 The SAS System 10:00 Thursday, October 22, 2015 5 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable SBP Response Distribution Gaussian Link Function Identity Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Restricted Maximum Likelihood Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Gender 1 0 Number of Observations Read 130132 Number of Observations Used 126342 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 0 0 4 1028381.6561 . 2590.417 1 0 4 1028215.438 166.21806747 453.0199 2 0 2 1028208.249 7.18901077 89.93031 3 0 2 1028207.9689 0.28013841 4.106232 4 0 2 1028207.9683 0.00057830 0.035248 Convergence criterion (GCONV=1E-8) satisfied. The SAS System 10:00 Thursday, October 22, 2015 6 The GLIMMIX Procedure Fit Statistics -2 Res Log Likelihood 1028208 AIC (smaller is better) 1028212 AICC (smaller is better) 1028212 BIC (smaller is better) 1028229 CAIC (smaller is better) 1028231 HQIC (smaller is better) 1028217 Generalized Chi-Square 17281010 Gener. Chi-Square / DF 136.78 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 128.06 1.3749 Residual 136.78 0.6458 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 129.88 0.4586 36892 283.20 <.0001 Age 0.2784 0.007452 89447 37.35 <.0001 Year -0.6999 0.01606 89447 -43.59 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 89447 1395.16 <.0001 Year 1 89447 1900.37 <.0001 The SAS System 10:00 Thursday, October 22, 2015 7 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable BMI Response Distribution Gaussian Link Function Identity Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Restricted Maximum Likelihood Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Gender 1 0 Number of Observations Read 130132 Number of Observations Used 120003 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 0 0 4 534175.16414 . 2398.678 1 0 2 530608.74646 3566.4176739 806.4725 2 0 4 529617.85252 990.89394298 83.17306 3 0 2 529601.86953 15.98298398 29.51903 4 0 2 529599.42487 2.44466847 1.813237 5 0 2 529599.41533 0.00953549 0.042706 6 0 3 529599.41533 0.00000533 0.000046 The SAS System 10:00 Thursday, October 22, 2015 8 The GLIMMIX Procedure Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 529599.4 AIC (smaller is better) 529603.4 AICC (smaller is better) 529603.4 BIC (smaller is better) 529620.5 CAIC (smaller is better) 529622.5 HQIC (smaller is better) 529608.8 Generalized Chi-Square 207965.3 Gener. Chi-Square / DF 1.73 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 20.0317 0.1568 Residual 1.7330 0.008437 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 33.9003 0.1258 35492 269.52 <.0001 Age -0.08639 0.002442 84508 -35.37 <.0001 Year 0.06820 0.003111 84508 21.92 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 84508 1251.25 <.0001 Year 1 84508 480.62 <.0001 The SAS System 10:00 Thursday, October 22, 2015 9 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable CholHDL Response Distribution Gaussian Link Function Identity Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Restricted Maximum Likelihood Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Gender 1 0 Number of Observations Read 130132 Number of Observations Used 122348 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 0 0 4 346358.70008 . 5562.606 1 0 4 345002.04894 1356.6511309 1026.712 2 0 4 344858.69001 143.35893439 112.7345 3 0 2 344857.4232 1.26680691 22.94712 4 0 2 344857.36679 0.05641222 0.409571 5 0 2 344857.36677 0.00001785 0.001521 The SAS System 10:00 Thursday, October 22, 2015 10 The GLIMMIX Procedure Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 344857.4 AIC (smaller is better) 344861.4 AICC (smaller is better) 344861.4 BIC (smaller is better) 344878.4 CAIC (smaller is better) 344880.4 HQIC (smaller is better) 344866.8 Generalized Chi-Square 69997.83 Gener. Chi-Square / DF 0.57 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 1.0633 0.009953 Residual 0.5721 0.002772 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 6.2763 0.03637 36335 172.56 <.0001 Age -0.03133 0.000626 86010 -50.03 <.0001 Year -0.02364 0.001182 86010 -20.00 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 86010 2502.55 <.0001 Year 1 86010 400.07 <.0001 The SAS System 10:00 Thursday, October 22, 2015 11 Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable Age Response Distribution Gaussian Link Function Identity Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Restricted Maximum Likelihood Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 130132 Dimensions R-side Cov. Parameters 2 Columns in X 2 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 0 0 4 707819.12185 . 20552.22 1 0 2 557124.04922 150695.07263 3265.842 2 0 6 462093.35901 95030.690208 645.5925 3 0 4 431037.56512 31055.793897 257.5794 4 0 4 411057.51229 19980.052828 82.10189 5 0 4 402789.27944 8268.2328453 19.48232 6 0 3 401361.31359 1427.9658472 3.560433 7 0 5 401274.80848 86.50511837 2.673624 8 0 3 401220.58312 54.22535860 0.065299 The SAS System 12:36 Sunday, October 25, 2015 24 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 9 0 3 401220.5456 0.03751875 0.000934 10 0 3 401220.54559 0.00000750 2.676E-6 Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 401220.5 AIC (smaller is better) 401224.5 AICC (smaller is better) 401224.5 BIC (smaller is better) 401241.6 CAIC (smaller is better) 401243.6 HQIC (smaller is better) 401230.0 Generalized Chi-Square 18745.65 Gener. Chi-Square / DF 0.14 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 98.6594 0.7186 Residual 0.1441 0.000670 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 48.4125 0.05170 37765 936.36 <.0001 Year 0.9870 0.000579 92365 1704.94 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Year 1 92365 2906827 <.0001 The SAS System 12:36 Sunday, October 25, 2015 25 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable Age Response Distribution Gaussian Link Function Identity Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Restricted Maximum Likelihood Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 130132 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 0 0 4 707705.14832 . 20526.77 1 0 2 557182.73674 150522.41158 3268.985 2 0 6 462093.48299 95089.253744 646.0852 3 0 4 431011.47402 31082.008974 257.766 4 0 4 411008.7942 20002.679817 82.15083 5 0 4 402730.31416 8278.4800468 19.50098 6 0 3 401298.56413 1431.7500257 3.558663 7 0 5 401211.80486 86.75927051 2.666841 8 0 3 401157.76618 54.03867854 0.067024 The SAS System 12:36 Sunday, October 25, 2015 26 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 9 0 3 401157.7266 0.03958654 0.000985 10 0 3 401157.72659 0.00000864 2.9E-6 Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 401157.7 AIC (smaller is better) 401161.7 AICC (smaller is better) 401161.7 BIC (smaller is better) 401178.8 CAIC (smaller is better) 401180.8 HQIC (smaller is better) 401167.1 Generalized Chi-Square 18730.37 Gener. Chi-Square / DF 0.14 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 98.6500 0.7186 Residual 0.1439 0.000670 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 48.5246 0.05321 37765 911.94 <.0001 Year 0.9650 0.002532 92364 381.08 <.0001 YearSQ 0.000972 0.000109 92364 8.90 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Year 1 92364 145223 <.0001 YearSQ 1 92364 79.26 <.0001 The SAS System 12:36 Sunday, October 25, 2015 27 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable Diabetesduration Response Distribution Gaussian Link Function Identity Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Restricted Maximum Likelihood Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 126158 Dimensions R-side Cov. Parameters 2 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 0 0 4 512396.61897 . 3360.138 1 0 4 495320.59355 17076.025415 840.0998 2 0 4 490815.8176 4504.7759557 185.831 3 0 4 490231.20476 584.61283857 17.7678 4 0 2 490223.67165 7.53310898 5.130721 5 0 2 490222.95658 0.71507339 0.213202 6 0 3 490222.95532 0.00125720 0.000635 The SAS System 12:36 Sunday, October 25, 2015 28 The GLIMMIX Procedure Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 490223.0 AIC (smaller is better) 490227.0 AICC (smaller is better) 490227.0 BIC (smaller is better) 490244.0 CAIC (smaller is better) 490246.0 HQIC (smaller is better) 490232.4 Generalized Chi-Square 93763.53 Gener. Chi-Square / DF 0.74 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 27.6075 0.2076 Residual 0.7432 0.003525 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -18.3596 0.1316 36754 -139.56 <.0001 Age 0.2169 0.002616 89401 82.91 <.0001 Year 0.7757 0.002856 89401 271.66 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 89401 6874.87 <.0001 Year 1 89401 73797.9 <.0001 The SAS System 12:36 Sunday, October 25, 2015 29 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_MALE_UNDER_75 Response Variable Diabetesduration Response Distribution Gaussian Link Function Identity Variance Function Default Variance Matrix Blocked By BSISID Estimation Technique Restricted Maximum Likelihood Degrees of Freedom Method Between-Within Class Level Information Class Levels Values BSISID 37766 not printed Number of Observations Read 130132 Number of Observations Used 126158 Dimensions R-side Cov. Parameters 2 Columns in X 4 Columns in Z per Subject 0 Subjects (Blocks in V) 37766 Max Obs per Subject 16 Optimization Information Optimization Technique Dual Quasi-Newton Parameters in Optimization 1 Lower Boundaries 0 Upper Boundaries 0 Fixed Effects Profiled Residual Variance Profiled Starting From Data Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 0 0 4 514078.44377 . 3720.356 1 0 4 495106.59292 18971.850846 881.0707 2 0 4 490355.78164 4750.8112825 201.7542 3 0 4 489684.93744 670.84419693 20.46968 4 0 2 489675.04427 9.89316831 6.221733 5 0 2 489673.98772 1.05655903 0.295797 6 0 3 489673.98527 0.00244114 0.001223 The SAS System 12:36 Sunday, October 25, 2015 30 The GLIMMIX Procedure Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 489674.0 AIC (smaller is better) 489678.0 AICC (smaller is better) 489678.0 BIC (smaller is better) 489695.1 CAIC (smaller is better) 489697.1 HQIC (smaller is better) 489683.4 Generalized Chi-Square 93220.00 Gener. Chi-Square / DF 0.74 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 27.5764 0.2074 Residual 0.7389 0.003505 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -17.6359 0.1350 36754 -130.66 <.0001 Age 0.2162 0.002614 89400 82.70 <.0001 Year 0.6417 0.006316 89400 101.59 <.0001 YearSQ 0.005964 0.000251 89400 23.77 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 89400 6839.34 <.0001 Year 1 89400 10320.3 <.0001 YearSQ 1 89400 565.24 <.0001 The SAS System 12:36 Sunday, October 25, 2015 37