S5 File. SAS output Glimmix outcomes under 75 years of age The SAS System 15:14 Monday, October 12, 2015 3 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 233326 Dimensions R-side Cov. Parameters 2 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 890087.07537 0.69679044 0.00183 1 0 2 877576.87985 0.01544435 0.001601 2 0 2 877780.79902 0.00020021 0.000042 3 0 1 877782.03915 0.00000302 0.005288 4 0 1 877782.05047 0.00000126 0.006603 5 0 1 877782.04819 0.00000043 0.002543 6 0 1 877782.04902 0.00000061 0.003212 7 0 1 877782.05017 0.00000000 0.0032 The SAS System 15:14 Monday, October 12, 2015 4 The GLIMMIX Procedure Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 877782.1 Generalized Chi-Square 45199.53 Gener. Chi-Square / DF 0.19 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.8049 0.004861 Residual 0.1937 0.000673 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept 0.7258 0.05083 67649 14.28 <.0001 Age -0.03303 0.000866 166E3 -38.13 <.0001 Gender 0 0.4425 0.03480 67649 12.72 <.0001 Gender 1 0 . . . . Year -0.01509 0.001937 166E3 -7.79 <.0001 *Dit is de lineare trend* Year*Gender 0 -0.01090 0.002386 166E3 -4.57 <.0001 *Dit is de interactieterm* Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 166E3 1453.64 <.0001 Gender 1 67649 161.70 <.0001 Year 1 166E3 223.77 <.0001 Year*Gender 1 166E3 20.85 <.0001 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 -1.2107 0.01198 67649 -101.05 <.0001 0.05 -1.2342 -1.1872 1 -1.5137 0.01388 67649 -109.03 <.0001 0.05 -1.5409 -1.4865 The SAS System 15:14 Monday, October 12, 2015 5 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 245653 Dimensions R-side Cov. Parameters 2 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1062548.248 0.88774149 0.000942 1 0 4 1244654.4705 0.35185294 0.060361 2 0 3 1303253.2197 0.10952646 0.008924 3 0 2 1303787.2992 0.01813576 0.056065 4 0 2 1302898.7884 0.00164611 0.000052 5 0 2 1302814.0619 0.00011754 5.474E-9 6 0 2 1302807.8138 0.00000827 3.372E-8 7 0 1 1302807.3705 0.00000065 0.007744 The SAS System 15:14 Monday, October 12, 2015 6 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 8 0 1 1302807.3621 0.00000078 0.002429 9 0 1 1302807.3303 0.00000214 0.026726 10 1 1 1302807.4156 0.00000251 0.007806 11 0 1 1302807.3154 0.00000070 0.00216 12 0 1 1302807.3428 0.00000191 0.02395 13 0 1 1302807.2664 0.00000225 0.006995 14 0 1 1302807.3561 0.00000001 0.006516 15 0 1 1302807.3565 0.00000065 0.00229 16 0 1 1302807.3305 0.00000206 0.02579 17 0 1 1302807.4128 0.00000073 0.015424 18 0 1 1302807.3839 0.00000000 0.015561 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 1302807 Generalized Chi-Square 113773.3 Gener. Chi-Square / DF 0.46 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.4840 0.004763 Residual 0.4632 0.001747 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -0.6005 0.07446 71052 -8.07 <.0001 Age 0.01410 0.001158 175E3 12.17 <.0001 Gender 0 0.2053 0.04915 71052 4.18 <.0001 Gender 1 0 . . . . Year 0.1928 0.003124 175E3 61.70 <.0001 Year*Gender 0 -0.02306 0.004222 175E3 -5.46 <.0001 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 175E3 148.14 <.0001 Gender 1 71052 17.45 <.0001 Year 1 175E3 7007.94 <.0001 The SAS System 15:14 Monday, October 12, 2015 7 The GLIMMIX Procedure Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Year*Gender 1 175E3 29.82 <.0001 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 2.6354 0.01737 71052 151.76 <.0001 0.05 2.6014 2.6695 1 2.7233 0.01930 71052 141.08 <.0001 0.05 2.6855 2.7612 The SAS System 15:14 Monday, October 12, 2015 8 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 198758 Dimensions R-side Cov. Parameters 2 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 906447.65274 0.54009191 0.234425 1 0 2 925228.7325 0.04234770 0.060432 2 0 2 928086.85238 0.00029963 0.000725 3 0 2 928109.72546 0.00000373 1.89E-7 4 0 1 928109.75959 0.00000002 0.000165 5 0 1 928109.76109 0.00000176 0.038447 6 0 1 928109.759 0.00000433 0.0563 7 0 2 928109.76328 0.00000259 9.304E-9 The SAS System 15:14 Monday, October 12, 2015 9 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 8 0 1 928109.76225 0.00000005 0.0006 9 0 1 928109.76108 0.00000069 0.014474 10 0 1 928109.7603 0.00000001 0.014514 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 928109.8 Generalized Chi-Square 105545.5 Gener. Chi-Square / DF 0.53 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.4753 0.004067 Residual 0.5310 0.002032 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -3.4252 0.07931 62477 -43.18 <.0001 Age 0.02801 0.001073 136E3 26.11 <.0001 Gender 0 0.4015 0.06380 62477 6.29 <.0001 Gender 1 0 . . . . Year -0.02647 0.003611 136E3 -7.33 <.0001 Year*Gender 0 0.01205 0.004601 136E3 2.62 0.0088 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 136E3 681.54 <.0001 Gender 1 62477 39.61 <.0001 Year 1 136E3 74.75 <.0001 Year*Gender 1 136E3 6.86 0.0088 The SAS System 15:14 Monday, October 12, 2015 10 The GLIMMIX Procedure Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 -1.4746 0.01196 62477 -123.32 <.0001 0.05 -1.4980 -1.4511 1 -2.0330 0.01536 62477 -132.38 <.0001 0.05 -2.0631 -2.0029 The SAS System 15:14 Monday, October 12, 2015 11 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 245653 Dimensions R-side Cov. Parameters 2 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1134484.1271 1.30732398 0.061356 1 0 3 1140671.1715 0.10301368 0.010977 2 0 2 1140422.1971 0.00232205 0.012357 3 0 2 1140385.3168 0.00002578 1.13E-6 4 0 1 1140384.8206 0.00000135 0.012504 5 0 1 1140384.8254 0.00000042 0.001928 6 0 1 1140384.834 0.00000065 0.014858 7 0 2 1140384.8471 0.00000041 0.004256 The SAS System 15:14 Monday, October 12, 2015 14 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 8 0 1 1140384.8391 0.00000001 0.00431 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 1140385 Generalized Chi-Square 180111.9 Gener. Chi-Square / DF 0.73 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.2229 0.003151 Residual 0.7332 0.002601 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -1.5549 0.05167 71052 -30.10 <.0001 Age 0.01238 0.000701 175E3 17.67 <.0001 Gender 0 -0.1015 0.04274 71052 -2.37 0.0176 Gender 1 0 . . . . Year 0.1728 0.002429 175E3 71.16 <.0001 Year*Gender 0 0.01386 0.003391 175E3 4.09 <.0001 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 175E3 312.18 <.0001 Gender 1 71052 5.64 0.0176 Year 1 175E3 11156.8 <.0001 Year*Gender 1 175E3 16.71 <.0001 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 1.4842 0.009590 71052 154.77 <.0001 0.05 1.4654 1.5030 1 1.4095 0.009927 71052 141.98 <.0001 0.05 1.3900 1.4289 The SAS System 15:14 Monday, October 12, 2015 15 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 234210 Dimensions R-side Cov. Parameters 2 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1022787.3294 0.13085207 0.001282 1 0 2 994998.5348 0.01987094 0.002435 2 0 2 995819.9071 0.00059857 0.000041 3 0 2 995815.423 0.00001836 0.002244 4 0 1 995815.65348 0.00000176 0.015921 5 0 2 995815.64124 0.00000081 1.762E-8 6 0 1 995815.64341 0.00000007 0.000091 7 0 1 995815.64302 0.00000017 0.003716 The SAS System 15:14 Monday, October 12, 2015 16 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 8 0 1 995815.64273 0.00000075 0.012482 9 0 1 995815.64409 0.00000018 0.009174 10 0 1 995815.64347 0.00000001 0.009114 11 0 1 995815.64355 0.00000026 0.003475 12 0 1 995815.64311 0.00000067 0.011515 13 0 1 995815.64205 0.00000232 0.038729 14 0 2 995815.64634 0.00000187 2.754E-8 15 0 1 995815.64214 0.00000049 0.008964 16 0 1 995815.64378 0.00000006 0.008628 17 0 1 995815.6435 0.00000000 0.008495 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 995815.6 Generalized Chi-Square 141531.6 Gener. Chi-Square / DF 0.60 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.4115 0.003553 Residual 0.6043 0.002088 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept 0.7663 0.05197 68482 14.75 <.0001 Age 0.006831 0.000711 166E3 9.61 <.0001 Gender 0 0.04514 0.04585 68482 0.98 0.3249 Gender 1 0 . . . . Year -0.01838 0.002415 166E3 -7.61 <.0001 Year*Gender 0 -0.00720 0.003309 166E3 -2.18 0.0296 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 166E3 92.37 <.0001 Gender 1 68482 0.97 0.3249 Year 1 166E3 169.82 <.0001 Year*Gender 1 166E3 4.73 0.0296 The SAS System 15:14 Monday, October 12, 2015 17 The GLIMMIX Procedure Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 0.9073 0.009421 68482 96.30 <.0001 0.05 0.8888 0.9258 1 0.9543 0.01007 68482 94.80 <.0001 0.05 0.9346 0.9740 The SAS System 15:14 Monday, October 12, 2015 18 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 245653 Dimensions R-side Cov. Parameters 2 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1090248.2407 0.92360593 0.065481 1 0 4 1319107.2617 0.61223426 0.004524 2 0 4 1413607.0914 0.38319991 0.073401 3 0 3 1417361.0639 0.03572710 0.001093 4 0 2 1415641.6887 0.00062460 1.493E-6 5 0 1 1415610.7143 0.00002887 0.003227 6 0 1 1415610.6781 0.00000153 0.016827 7 0 1 1415610.7482 0.00000002 0.016983 The SAS System 15:14 Monday, October 12, 2015 19 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 8 0 1 1415610.7483 0.00000175 0.002296 9 0 1 1415610.6702 0.00000111 0.009574 10 0 1 1415610.7197 0.00000410 0.035494 11 0 1 1415610.5365 0.00000024 0.033644 12 0 1 1415610.5473 0.00000006 0.033015 13 0 1 1415610.5498 0.00000137 0.017906 14 0 2 1415610.6109 0.00000130 0.003298 15 0 1 1415610.6689 0.00000127 0.01101 16 0 1 1415610.7256 0.00000001 0.011265 17 0 1 1415610.7256 0.00000000 0.011278 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 1415611 Generalized Chi-Square 176623.9 Gener. Chi-Square / DF 0.72 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.1714 0.004202 Residual 0.7190 0.003092 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -0.1807 0.07816 71052 -2.31 0.0208 Age 0.01875 0.001101 175E3 17.03 <.0001 Gender 0 -0.1033 0.05844 71052 -1.77 0.0772 Gender 1 0 . . . . Year 0.1605 0.003593 175E3 44.66 <.0001 Year*Gender 0 0.002020 0.004900 175E3 0.41 0.6802 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 175E3 290.17 <.0001 Gender 1 71052 3.12 0.0772 Year 1 175E3 4326.16 <.0001 Year*Gender 1 175E3 0.17 0.6802 The SAS System 15:14 Monday, October 12, 2015 20 The GLIMMIX Procedure Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 2.9427 0.01579 71052 186.36 <.0001 0.05 2.9118 2.9737 1 3.0203 0.01741 71052 173.49 <.0001 0.05 2.9862 3.0544 The SAS System 15:14 Monday, October 12, 2015 21 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 238524 Dimensions R-side Cov. Parameters 2 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 995846.08576 0.14288353 0.003152 1 0 2 988644.37591 0.00706451 0.002343 2 0 2 988811.62803 0.00026643 2.348E-7 3 0 1 988811.58699 0.00000083 0.000526 4 0 1 988811.58757 0.00000001 0.000134 5 0 1 988811.58756 0.00000029 0.009912 6 0 1 988811.58739 0.00000211 0.064321 7 0 1 988811.58868 0.00000178 0.001771 The SAS System 15:14 Monday, October 12, 2015 22 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 8 0 1 988811.5876 0.00000037 0.011232 9 0 1 988811.58737 0.00000046 0.004805 10 0 1 988811.58765 0.00000102 0.031227 11 0 1 988811.58702 0.00000077 0.004326 12 0 1 988811.58749 0.00000001 0.003987 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 988811.6 Generalized Chi-Square 162470.6 Gener. Chi-Square / DF 0.68 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.3202 0.003096 Residual 0.6812 0.002312 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept 1.3482 0.04820 69500 27.97 <.0001 Age -0.03913 0.000658 169E3 -59.49 <.0001 Gender 0 0.07634 0.04196 69500 1.82 0.0689 Gender 1 0 . . . . Year 0.09632 0.002255 169E3 42.72 <.0001 Year*Gender 0 -0.01071 0.003079 169E3 -3.48 0.0005 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 169E3 3538.65 <.0001 Gender 1 69500 3.31 0.0689 Year 1 169E3 3351.86 <.0001 Year*Gender 1 169E3 12.09 0.0005 The SAS System 15:14 Monday, October 12, 2015 23 The GLIMMIX Procedure Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 0.09307 0.008163 69500 11.40 <.0001 0.05 0.07707 0.1091 1 0.1534 0.008694 69500 17.65 <.0001 0.05 0.1364 0.1704 The SAS System 15:14 Monday, October 12, 2015 24 Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 226287 Dimensions R-side Cov. Parameters 2 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 780468.54634 0.19862114 0.000457 1 0 2 773169.29439 0.00650458 1.903E-6 2 0 2 773289.57148 0.00005572 6.054E-7 3 0 1 773289.63226 0.00000019 0.000091 4 0 1 773289.63221 0.00000002 0.000148 5 0 1 773289.6322 0.00000039 0.003713 6 0 1 773289.6323 0.00000098 0.005988 7 0 1 773289.63205 0.00000157 0.009656 The SAS System 15:14 Monday, October 12, 2015 27 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 8 0 2 773289.63246 0.00000078 0.001896 9 0 1 773289.63226 0.00000000 0.001895 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 773289.6 Generalized Chi-Square 47230.05 Gener. Chi-Square / DF 0.21 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.7943 0.004865 Residual 0.2087 0.000738 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept 1.5955 0.04579 66733 34.84 <.0001 Age -0.02835 0.000763 16E4 -37.15 <.0001 Gender 0 -0.5797 0.03206 66733 -18.08 <.0001 Gender 1 0 . . . . Year 0.01782 0.001643 16E4 10.85 <.0001 Year*Gender 0 0.004600 0.002127 16E4 2.16 0.0306 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 16E4 1380.39 <.0001 Gender 1 66733 327.00 <.0001 Year 1 16E4 267.35 <.0001 Year*Gender 1 16E4 4.68 0.0306 The SAS System 15:14 Monday, October 12, 2015 28 The GLIMMIX Procedure Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 -0.4508 0.01039 66733 -43.37 <.0001 0.05 -0.4712 -0.4304 1 0.06956 0.01078 66733 6.45 <.0001 0.05 0.04842 0.09070 The SAS System 15:14 Monday, October 12, 2015 29 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 226287 Dimensions R-side Cov. Parameters 2 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 859328.40702 0.20725434 0.000037 1 0 2 849577.69595 0.00491618 0.011612 2 0 2 849775.24171 0.00002781 2.468E-8 3 0 1 849775.36358 0.00000022 0.000431 4 0 1 849775.36364 0.00000022 0.003127 5 0 1 849775.36365 0.00000000 0.0031 The SAS System 15:14 Monday, October 12, 2015 30 The GLIMMIX Procedure Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 849775.4 Generalized Chi-Square 76643.53 Gener. Chi-Square / DF 0.34 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.6675 0.004458 Residual 0.3387 0.001196 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -1.6996 0.04952 66733 -34.32 <.0001 Age 0.01977 0.000757 16E4 26.12 <.0001 Gender 0 0.5881 0.03785 66733 15.54 <.0001 Gender 1 0 . . . . Year -0.01553 0.002018 16E4 -7.70 <.0001 Year*Gender 0 -0.00506 0.002634 16E4 -1.92 0.0548 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 16E4 682.31 <.0001 Gender 1 66733 241.45 <.0001 Year 1 16E4 161.95 <.0001 Year*Gender 1 16E4 3.69 0.0548 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 -0.1523 0.009707 66733 -15.69 <.0001 0.05 -0.1713 -0.1333 1 -0.6752 0.01086 66733 -62.17 <.0001 0.05 -0.6965 -0.6539 The SAS System 15:14 Monday, October 12, 2015 31 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 127852 Dimensions R-side Cov. Parameters 2 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 593577.09108 0.82057359 0.041269 1 0 2 622251.23359 0.12262392 0.023915 2 0 2 627964.04735 0.00438068 0.00004 3 0 2 628119.27407 0.00041256 0.000282 4 0 1 628118.66432 0.00001020 0.019059 5 0 2 628118.65592 0.00000127 1.09E-8 6 0 1 628118.65765 0.00000028 0.003924 7 0 1 628118.65721 0.00000005 0.00329 The SAS System 15:14 Monday, October 12, 2015 44 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 8 0 1 628118.65728 0.00000038 0.002348 9 0 1 628118.65777 0.00000012 0.000659 10 0 1 628118.65759 0.00000007 0.000457 11 0 1 628118.6575 0.00000054 0.007606 12 0 1 628118.6582 0.00000089 0.005511 13 0 1 628118.65703 0.00000007 0.0047 14 0 1 628118.65717 0.00000054 0.003355 15 0 1 628118.65786 0.00000002 0.003448 16 0 1 628118.65782 0.00000040 0.002468 17 0 1 628118.65731 0.00000003 0.002105 18 0 1 628118.65738 0.00000024 0.001503 19 0 1 628118.65769 0.00000001 0.001547 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 628118.7 Generalized Chi-Square 68867.73 Gener. Chi-Square / DF 0.54 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.4665 0.004689 Residual 0.5387 0.002750 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -4.7736 0.1099 52277 -43.43 <.0001 Age 0.04524 0.001421 75570 31.84 <.0001 Gender 0 0.1015 0.09376 52277 1.08 0.2788 Gender 1 0 . . . . Year -0.01121 0.004742 75570 -2.36 0.0181 Year*Gender 0 0.01064 0.006397 75570 1.66 0.0962 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 75570 1013.66 <.0001 Gender 1 52277 1.17 0.2788 The SAS System 15:14 Monday, October 12, 2015 45 The GLIMMIX Procedure Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Year 1 75570 3.29 0.0697 Year*Gender 1 75570 2.77 0.0962 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 -1.8692 0.01545 52277 -120.96 <.0001 0.05 -1.8995 -1.8389 1 -2.1164 0.01794 52277 -117.95 <.0001 0.05 -2.1516 -2.0812 The SAS System 15:14 Monday, October 12, 2015 46 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 245653 Dimensions R-side Cov. Parameters 2 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1141175.6877 1.39527304 0.001283 1 0 2 1169068.1611 0.11531269 0.380893 2 0 2 1170770.0715 0.00758706 0.002077 3 0 2 1170724.6032 0.00019832 0.00021 4 0 1 1170722.2032 0.00000850 0.051881 5 0 2 1170722.1686 0.00000215 0.000128 6 0 1 1170722.1193 0.00000002 0.000446 7 0 1 1170722.1188 0.00000107 0.026297 The SAS System 15:14 Monday, October 12, 2015 47 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Subiterations Function Change Gradient 8 0 1 1170722.1429 0.00000024 0.020205 9 0 2 1170722.1375 0.00000161 0.019944 10 0 1 1170722.1013 0.00000017 0.015661 11 0 1 1170722.1051 0.00000050 0.003173 12 0 1 1170722.1163 0.00000077 0.016087 13 0 1 1170722.1337 0.00000000 0.016027 Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 1170722 Generalized Chi-Square 165809.2 Gener. Chi-Square / DF 0.67 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.2801 0.003397 Residual 0.6750 0.002397 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -0.9058 0.05585 71052 -16.22 <.0001 Age 0.01442 0.000780 175E3 18.49 <.0001 Gender 0 0.02592 0.04455 71052 0.58 0.5606 Gender 1 0 . . . . Year 0.1399 0.002568 175E3 54.48 <.0001 Year*Gender 0 -0.00666 0.003521 175E3 -1.89 0.0586 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 175E3 341.75 <.0001 Gender 1 71052 0.34 0.5606 Year 1 175E3 5931.26 <.0001 Year*Gender 1 175E3 3.58 0.0586 The SAS System 15:14 Monday, October 12, 2015 48 The GLIMMIX Procedure Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 1.7066 0.01063 71052 160.57 <.0001 0.05 1.6858 1.7274 1 1.7653 0.01156 71052 152.74 <.0001 0.05 1.7427 1.7880 The SAS System 15:14 Monday, October 12, 2015 49 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 245653 Dimensions R-side Cov. Parameters 2 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 959944.13357 1.00637241 0.014625 1 0 2 946035.48832 0.11641522 0.024331 2 0 2 945958.46625 0.00849884 2.326E-6 3 0 2 945948.48281 0.00031139 0.00038 4 0 1 945948.54257 0.00002101 0.018809 5 0 2 945948.52112 0.00000119 4.844E-8 6 0 1 945948.52268 0.00000028 0.004536 7 0 1 945948.52229 0.00000001 0.004432 The SAS System 15:14 Monday, October 12, 2015 58 The GLIMMIX Procedure Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 945948.5 Generalized Chi-Square 85073.46 Gener. Chi-Square / DF 0.35 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 0.6806 0.004514 Residual 0.3463 0.001180 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -3.0673 0.04716 71052 -65.04 <.0001 Age 0.02564 0.000729 175E3 35.19 <.0001 Gender 0 0.2288 0.03551 71052 6.44 <.0001 Gender 1 0 . . . . Year 0.1542 0.001916 175E3 80.48 <.0001 Year*Gender 0 -0.00391 0.002573 175E3 -1.52 0.1291 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 175E3 1238.56 <.0001 Gender 1 71052 41.51 <.0001 Year 1 175E3 12543.1 <.0001 Year*Gender 1 175E3 2.30 0.1291 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 0.6603 0.01019 71052 64.82 <.0001 0.05 0.6403 0.6803 1 0.4812 0.01057 71052 45.53 <.0001 0.05 0.4605 0.5019 The SAS System 13:41 Wednesday, October 14, 2015 45 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 198758 Dimensions R-side Cov. Parameters 1 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 958340.48258 2.00000000 . 1 0 0 979957.43832 0.65053327 . 2 0 0 983099.35225 0.00397864 . 3 0 0 983125.43687 0.00000029 . 4 0 0 983125.43876 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 13:41 Wednesday, October 14, 2015 46 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 983125.4 Generalized Chi-Square 199858.5 Gener. Chi-Square / DF 1.01 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0056 0.003190 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -2.8324 0.06686 62477 -42.37 <.0001 Age 0.02436 0.000778 136E3 31.31 <.0001 Gender 0 0.4944 0.05884 62477 8.40 <.0001 Gender 1 0 . . . . Year -0.05757 0.003469 136E3 -16.60 <.0001 Year*Gender 0 0.005383 0.004514 136E3 1.19 0.2330 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 136E3 980.20 <.0001 Gender 1 62477 70.59 <.0001 Year 1 136E3 591.21 <.0001 Year*Gender 1 136E3 1.42 0.2330 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 -1.5066 0.008062 62477 -186.87 <.0001 0.05 -1.5224 -1.4908 1 -2.0711 0.01048 62477 -197.64 <.0001 0.05 -2.0916 -2.0505 The SAS System 13:41 Wednesday, October 14, 2015 49 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 245653 Dimensions R-side Cov. Parameters 1 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1031221.0089 1.01204213 . 1 0 0 1348872.2531 0.59363084 . 2 0 0 1542715.3783 0.20153728 . 3 0 0 1605183.6173 0.02418877 . 4 0 0 1610271.6349 0.00025989 . 5 0 0 1610301.8931 0.00000002 . 6 0 0 1610301.8942 0.00000000 . The SAS System 13:41 Wednesday, October 14, 2015 50 The GLIMMIX Procedure Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 1610302 Generalized Chi-Square 254440.7 Gener. Chi-Square / DF 1.04 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0358 0.002956 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept 0.6724 0.09751 71052 6.90 <.0001 Age 0.01761 0.001307 175E3 13.47 <.0001 Gender 0 0.07722 0.07691 71052 1.00 0.3154 Gender 1 0 . . . . Year 0.1475 0.004671 175E3 31.58 <.0001 Year*Gender 0 -0.00843 0.006498 175E3 -1.30 0.1945 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 175E3 181.52 <.0001 Gender 1 71052 1.01 0.3154 Year 1 175E3 1947.30 <.0001 Year*Gender 1 175E3 1.68 0.1945 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 3.6089 0.01808 71052 199.57 <.0001 0.05 3.5735 3.6444 1 3.6389 0.01960 71052 185.67 <.0001 0.05 3.6005 3.6773 The SAS System 13:41 Wednesday, October 14, 2015 53 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 226287 Dimensions R-side Cov. Parameters 1 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1097999.1944 0.52465378 . 1 0 0 1116752.3471 0.02733728 . 2 0 0 1117894.6757 0.00010185 . 3 0 0 1117898.9984 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 13:41 Wednesday, October 14, 2015 54 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 1117899 Generalized Chi-Square 226525.3 Gener. Chi-Square / DF 1.00 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0011 0.002976 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -3.0960 0.06152 66733 -50.33 <.0001 Age 0.01954 0.000720 16E4 27.12 <.0001 Gender 0 0.3019 0.05673 66733 5.32 <.0001 Gender 1 0 . . . . Year 0.006853 0.003126 16E4 2.19 0.0283 Year*Gender 0 -0.02198 0.004303 16E4 -5.11 <.0001 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 16E4 735.50 <.0001 Gender 1 66733 28.33 <.0001 Year 1 16E4 3.69 0.0546 Year*Gender 1 16E4 26.08 <.0001 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 -1.7788 0.008234 66733 -216.02 <.0001 0.05 -1.7949 -1.7626 1 -1.7972 0.008808 66733 -204.05 <.0001 0.05 -1.8144 -1.7799 The SAS System 13:41 Wednesday, October 14, 2015 55 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 245653 Dimensions R-side Cov. Parameters 1 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1149020.3691 1.18253692 . 1 0 0 1323397.5154 0.37252355 . 2 0 0 1384392.4636 0.04249240 . 3 0 0 1389281.9714 0.00043477 . 4 0 0 1389309.7632 0.00000003 . 5 0 0 1389309.7641 0.00000000 . The SAS System 13:41 Wednesday, October 14, 2015 56 The GLIMMIX Procedure Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 1389310 Generalized Chi-Square 248953.7 Gener. Chi-Square / DF 1.01 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0135 0.002892 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -0.3296 0.06390 71052 -5.16 <.0001 Age 0.006158 0.000857 175E3 7.18 <.0001 Gender 0 -0.04908 0.04838 71052 -1.01 0.3104 Gender 1 0 . . . . Year 0.2027 0.002988 175E3 67.84 <.0001 Year*Gender 0 0.004640 0.004165 175E3 1.11 0.2652 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 175E3 51.58 <.0001 Gender 1 71052 1.03 0.3104 Year 1 175E3 9696.35 <.0001 Year*Gender 1 175E3 1.24 0.2652 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 2.6391 0.01173 71052 224.97 <.0001 0.05 2.6161 2.6621 1 2.6291 0.01243 71052 211.59 <.0001 0.05 2.6048 2.6535 The SAS System 13:41 Wednesday, October 14, 2015 57 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 231030 Dimensions R-side Cov. Parameters 1 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1009062.9274 0.29942438 . 1 0 0 996146.16584 0.01088239 . 2 0 0 996439.79475 0.00001594 . 3 0 0 996439.99076 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 13:41 Wednesday, October 14, 2015 58 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 996440.0 Generalized Chi-Square 231262.8 Gener. Chi-Square / DF 1.00 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0010 0.002945 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -1.8377 0.04209 68328 -43.66 <.0001 Age 0.03000 0.000492 163E3 60.96 <.0001 Gender 0 -0.2294 0.03953 68328 -5.80 <.0001 Gender 1 0 . . . . Year 0.05338 0.002209 163E3 24.17 <.0001 Year*Gender 0 -0.01686 0.003022 163E3 -5.58 <.0001 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 163E3 3715.83 <.0001 Gender 1 68328 33.67 <.0001 Year 1 163E3 885.55 <.0001 Year*Gender 1 163E3 31.15 <.0001 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 0.2614 0.005827 68328 44.86 <.0001 0.05 0.2499 0.2728 1 0.7074 0.006532 68328 108.30 <.0001 0.05 0.6946 0.7202 The SAS System 13:41 Wednesday, October 14, 2015 59 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 245653 Dimensions R-side Cov. Parameters 1 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1126156.121 0.87602882 . 1 0 0 1340244.4299 0.35993138 . 2 0 0 1434603.7533 0.08150231 . 3 0 0 1447419.7748 0.00229880 . 4 0 0 1447615.8151 0.00000100 . 5 0 0 1447615.8607 0.00000000 . The SAS System 13:41 Wednesday, October 14, 2015 60 The GLIMMIX Procedure Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 1447616 Generalized Chi-Square 247252.8 Gener. Chi-Square / DF 1.01 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0065 0.002872 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -0.3266 0.07004 71052 -4.66 <.0001 Age 0.01384 0.000942 175E3 14.69 <.0001 Gender 0 0.04186 0.05330 71052 0.79 0.4322 Gender 1 0 . . . . Year 0.1900 0.003309 175E3 57.41 <.0001 Year*Gender 0 -0.00665 0.004588 175E3 -1.45 0.1470 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 175E3 215.78 <.0001 Gender 1 71052 0.62 0.4322 Year 1 175E3 6623.02 <.0001 Year*Gender 1 175E3 2.10 0.1470 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 2.9032 0.01304 71052 222.69 <.0001 0.05 2.8776 2.9287 1 2.9459 0.01419 71052 207.62 <.0001 0.05 2.9181 2.9737 The SAS System 13:41 Wednesday, October 14, 2015 61 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 171621 Dimensions R-side Cov. Parameters 1 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 Max Obs per Subject 15 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 800778.40102 0.96357120 . 1 0 0 941511.55956 0.40436441 . 2 0 0 984730.40017 0.05022622 . 3 0 0 988493.21938 0.00063711 . 4 0 0 988527.3088 0.00000008 . 5 0 0 988527.31251 0.00000000 . The SAS System 13:41 Wednesday, October 14, 2015 62 The GLIMMIX Procedure Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 988527.3 Generalized Chi-Square 171796.5 Gener. Chi-Square / DF 1.00 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0011 0.003417 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -3.3034 0.1098 57317 -30.09 <.0001 Age 0.02022 0.001276 114E3 15.84 <.0001 Gender 0 -0.08887 0.1018 57317 -0.87 0.3826 Gender 1 0 . . . . Year -0.06208 0.005485 114E3 -11.32 <.0001 Year*Gender 0 0.01881 0.007524 114E3 2.50 0.0124 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 114E3 250.98 <.0001 Gender 1 57317 0.76 0.3826 Year 1 114E3 196.10 <.0001 Year*Gender 1 114E3 6.25 0.0124 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 -2.7225 0.01380 57317 -197.23 <.0001 0.05 -2.7496 -2.6954 1 -2.8886 0.01589 57317 -181.76 <.0001 0.05 -2.9197 -2.8574 The SAS System 13:41 Wednesday, October 14, 2015 63 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 245653 Dimensions R-side Cov. Parameters 1 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1145363.7408 0.58231222 . 1 0 0 1163336.9871 0.02621272 . 2 0 0 1164054.6955 0.00005258 . 3 0 0 1164055.5402 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 13:41 Wednesday, October 14, 2015 64 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 1164056 Generalized Chi-Square 251506.2 Gener. Chi-Square / DF 1.02 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0238 0.002921 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -3.3471 0.04778 71052 -70.05 <.0001 Age 0.01326 0.000562 175E3 23.61 <.0001 Gender 0 -0.05610 0.04440 71052 -1.26 0.2065 Gender 1 0 . . . . Year 0.2928 0.002587 175E3 113.16 <.0001 Year*Gender 0 0.004314 0.003597 175E3 1.20 0.2304 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 175E3 557.27 <.0001 Gender 1 71052 1.60 0.2065 Year 1 175E3 26872.4 <.0001 Year*Gender 1 175E3 1.44 0.2304 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 1.1956 0.007171 71052 166.72 <.0001 0.05 1.1815 1.2096 1 1.1968 0.007624 71052 156.97 <.0001 0.05 1.1818 1.2117 The SAS System 13:41 Wednesday, October 14, 2015 65 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 245653 Dimensions R-side Cov. Parameters 1 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1195204.9176 0.59600399 . 1 0 0 1233447.0479 0.05382613 . 2 0 0 1235737.6433 0.00040858 . 3 0 0 1235749.325 0.00000002 . 4 0 0 1235749.3254 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 13:41 Wednesday, October 14, 2015 66 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 1235749 Generalized Chi-Square 245665.1 Gener. Chi-Square / DF 1.00 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0001 0.002854 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -2.6722 0.05805 71052 -46.03 <.0001 Age 0.01351 0.000709 175E3 19.05 <.0001 Gender 0 -0.4106 0.05386 71052 -7.62 <.0001 Gender 1 0 . . . . Year -0.00183 0.002838 175E3 -0.64 0.5190 Year*Gender 0 0.02387 0.004105 175E3 5.81 <.0001 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 175E3 362.91 <.0001 Gender 1 71052 58.11 <.0001 Year 1 175E3 24.24 <.0001 Year*Gender 1 175E3 33.81 <.0001 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 -1.9664 0.008470 71052 -232.16 <.0001 0.05 -1.9830 -1.9498 1 -1.8594 0.008630 71052 -215.45 <.0001 0.05 -1.8763 -1.8425 The SAS System 13:41 Wednesday, October 14, 2015 67 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 245653 Dimensions R-side Cov. Parameters 1 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1176003.5754 0.37124382 . 1 0 0 1160514.8382 0.02750258 . 2 0 0 1160807.2485 0.00005565 . 3 0 0 1160807.3815 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 13:41 Wednesday, October 14, 2015 68 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 1160807 Generalized Chi-Square 245607.3 Gener. Chi-Square / DF 1.00 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 0.9998 0.002853 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -1.8264 0.04909 71052 -37.21 <.0001 Age 0.007785 0.000597 175E3 13.04 <.0001 Gender 0 -0.01012 0.04477 71052 -0.23 0.8211 Gender 1 0 . . . . Year -0.00538 0.002440 175E3 -2.21 0.0274 Year*Gender 0 -0.00817 0.003439 175E3 -2.38 0.0175 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 175E3 170.10 <.0001 Gender 1 71052 0.05 0.8211 Year 1 175E3 30.30 <.0001 Year*Gender 1 175E3 5.64 0.0175 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 -1.5269 0.007245 71052 -210.74 <.0001 0.05 -1.5411 -1.5127 1 -1.4129 0.007423 71052 -190.35 <.0001 0.05 -1.4274 -1.3983 The SAS System 13:41 Wednesday, October 14, 2015 69 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 245653 Dimensions R-side Cov. Parameters 1 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1074479.0826 0.16957880 . 1 0 0 1058199.6015 0.00683057 . 2 0 0 1058457.2876 0.00000470 . 3 0 0 1058457.3638 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 13:41 Wednesday, October 14, 2015 70 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 1058457 Generalized Chi-Square 245653.2 Gener. Chi-Square / DF 1.00 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0000 0.002853 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept 0.7784 0.03979 71052 19.56 <.0001 Age -0.00434 0.000478 175E3 -9.08 <.0001 Gender 0 0.1318 0.03692 71052 3.57 0.0004 Gender 1 0 . . . . Year 0.001098 0.002017 175E3 0.54 0.5861 Year*Gender 0 -0.00178 0.002828 175E3 -0.63 0.5294 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 175E3 82.46 <.0001 Gender 1 71052 12.74 0.0004 Year 1 175E3 0.02 0.8823 Year*Gender 1 175E3 0.40 0.5294 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 0.6327 0.005828 71052 108.56 <.0001 0.05 0.6213 0.6441 1 0.5235 0.006092 71052 85.94 <.0001 0.05 0.5116 0.5355 The SAS System 13:41 Wednesday, October 14, 2015 71 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 245653 Dimensions R-side Cov. Parameters 1 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1104162.9123 0.16476053 . 1 0 0 1084279.4611 0.00708708 . 2 0 0 1084809.3147 0.00000537 . 3 0 0 1084809.7526 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 13:41 Wednesday, October 14, 2015 72 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 1084810 Generalized Chi-Square 246061.6 Gener. Chi-Square / DF 1.00 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0017 0.002858 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -2.7539 0.04144 71052 -66.46 <.0001 Age 0.05350 0.000498 175E3 107.35 <.0001 Gender 0 -0.2315 0.03822 71052 -6.06 <.0001 Gender 1 0 . . . . Year 0.01983 0.002144 175E3 9.25 <.0001 Year*Gender 0 0.009251 0.002941 175E3 3.15 0.0017 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 175E3 11524.9 <.0001 Gender 1 71052 36.70 <.0001 Year 1 175E3 276.70 <.0001 Year*Gender 1 175E3 9.89 0.0017 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 0.6966 0.006034 71052 115.44 <.0001 0.05 0.6847 0.7084 1 0.8105 0.006547 71052 123.79 <.0001 0.05 0.7976 0.8233 The GLIMMIX Procedure Model Information Data Set SYS.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 171958 Dimensions R-side Cov. Parameters 1 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 664966.17344 1.07856422 . 1 0 0 915647.35176 0.77798261 . 2 0 0 1093097.0481 2.00000000 . 3 0 0 1187268.0505 0.90779391 . 4 0 0 1215358.3674 0.05146135 . 5 0 0 1217509.0119 0.00029744 . 6 0 0 1217520.6925 0.00000001 . The SAS System 12:18 Sunday, October 25, 2015 4 The GLIMMIX Procedure Convergence criterion (PCONV=1.11022E-8) satisfied. Fit Statistics -2 Res Log Pseudo-Likelihood 1217521 Generalized Chi-Square 178211.9 Gener. Chi-Square / DF 1.04 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0364 0.003535 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -5.2494 0.1932 57997 -27.17 <.0001 Age 0.04276 0.002400 114E3 17.81 <.0001 Gender 0 0.6191 0.1414 57997 4.38 <.0001 Gender 1 0 . . . . Year -0.1581 0.009386 114E3 -16.84 <.0001 Year*Gender 0 0.02704 0.01156 114E3 2.34 0.0194 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 114E3 317.27 <.0001 Gender 1 57997 19.18 <.0001 Year 1 114E3 625.23 <.0001 Year*Gender 1 114E3 5.47 0.0194 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 -3.6970 0.02289 57997 -161.50 <.0001 0.05 -3.7419 -3.6521 1 -4.6689 0.03725 57997 -125.35 <.0001 0.05 -4.7419 -4.5959 The SAS System 12:18 Sunday, October 25, 2015 5 The SAS System 13:41 Wednesday, October 14, 2015 1 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 234210 Dimensions R-side Cov. Parameters 2 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1685638.8361 . 13176.45 1 0 5 1683109.0692 2529.7669276 2634.67 2 0 4 1682762.5187 346.55048408 251.8678 3 0 2 1682760.3283 2.19036722 60.12117 4 0 2 1682760.1918 0.13648552 1.060011 5 0 2 1682760.1918 0.00004224 0.004559 The SAS System 13:41 Wednesday, October 14, 2015 2 The GLIMMIX Procedure Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 1682760 AIC (smaller is better) 1682764 AICC (smaller is better) 1682764 BIC (smaller is better) 1682783 CAIC (smaller is better) 1682785 HQIC (smaller is better) 1682770 Generalized Chi-Square 11305741 Gener. Chi-Square / DF 48.27 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 67.8434 0.4997 Residual 48.2728 0.1699 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept 51.8679 0.2471 68482 209.89 <.0001 Age -0.05103 0.003687 166E3 -13.84 <.0001 Gender 0 -0.4584 0.1983 68482 -2.31 0.0208 Gender 1 0 . . . . Year 0.1159 0.01040 166E3 11.15 <.0001 Year*Gender 0 0.04907 0.01405 166E3 3.49 0.0005 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 166E3 191.62 <.0001 Gender 1 68482 5.34 0.0208 Year 1 166E3 362.56 <.0001 Year*Gender 1 166E3 12.19 0.0005 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 50.3590 0.04952 68482 1017.04 <.0001 0.05 50.2619 50.4560 1 50.1892 0.05247 68482 956.61 <.0001 0.05 50.0863 50.2920 The SAS System 13:41 Wednesday, October 14, 2015 3 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 238524 Dimensions R-side Cov. Parameters 2 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1949706.0114 . 4253.095 1 0 4 1949514.7294 191.28191816 1543.761 2 0 2 1949472.2989 42.43055615 274.988 3 0 2 1949471.0358 1.26309237 22.80376 4 0 2 1949471.0272 0.00857699 0.30556 Convergence criterion (GCONV=1E-8) satisfied. The SAS System 13:41 Wednesday, October 14, 2015 4 The GLIMMIX Procedure Fit Statistics -2 Res Log Likelihood 1949471 AIC (smaller is better) 1949475 AICC (smaller is better) 1949475 BIC (smaller is better) 1949493 CAIC (smaller is better) 1949495 HQIC (smaller is better) 1949481 Generalized Chi-Square 34200705 Gener. Chi-Square / DF 143.39 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 126.38 1.0023 Residual 143.39 0.4919 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept 126.95 0.3775 69500 336.33 <.0001 Age 0.3609 0.005354 169E3 67.41 <.0001 Gender 0 -1.5515 0.3171 69500 -4.89 <.0001 Gender 1 0 . . . . Year -0.9002 0.01677 169E3 -53.68 <.0001 Year*Gender 0 0.1546 0.02289 169E3 6.75 <.0001 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 169E3 4544.66 <.0001 Gender 1 69500 23.94 <.0001 Year 1 169E3 4892.48 <.0001 Year*Gender 1 169E3 45.58 <.0001 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 138.24 0.07114 69500 1943.25 <.0001 0.05 138.10 138.38 1 137.82 0.07532 69500 1829.68 <.0001 0.05 137.67 137.96 The SAS System 13:41 Wednesday, October 14, 2015 5 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 226287 Dimensions R-side Cov. Parameters 2 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1057758.1356 . 4117.357 1 0 2 1049758.5597 7999.5758931 543.8745 2 0 4 1049366.8489 391.71086726 37.45321 3 0 2 1049364.4983 2.35055275 7.191245 4 0 2 1049364.4066 0.09170830 0.123352 5 0 2 1049364.4066 0.00002742 0.000415 The SAS System 13:41 Wednesday, October 14, 2015 6 The GLIMMIX Procedure Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 1049364 AIC (smaller is better) 1049368 AICC (smaller is better) 1049368 BIC (smaller is better) 1049387 CAIC (smaller is better) 1049389 HQIC (smaller is better) 1049374 Generalized Chi-Square 481296.5 Gener. Chi-Square / DF 2.13 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 26.4996 0.1508 Residual 2.1270 0.007534 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept 35.8136 0.1097 66733 326.41 <.0001 Age -0.09277 0.002000 16E4 -46.37 <.0001 Gender 0 -1.6009 0.06157 66733 -26.00 <.0001 Gender 1 0 . . . . Year 0.07264 0.003119 16E4 23.29 <.0001 Year*Gender 0 0.001554 0.003464 16E4 0.45 0.6538 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 16E4 2150.58 <.0001 Gender 1 66733 676.10 <.0001 Year 1 16E4 841.50 <.0001 Year*Gender 1 16E4 0.20 0.6538 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 29.4234 0.02792 66733 1054.02 <.0001 0.05 29.3686 29.4781 1 31.0042 0.02968 66733 1044.74 <.0001 0.05 30.9461 31.0624 The SAS System 13:41 Wednesday, October 14, 2015 7 The GLIMMIX Procedure Model Information Data Set WORK.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 231030 Dimensions R-side Cov. Parameters 2 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 636476.7204 . 9757.327 1 0 4 635142.61477 1334.1056308 3551.906 2 0 4 633862.43373 1280.1810393 957.6371 3 0 2 633825.13606 37.29767312 439.0092 4 0 2 633813.01589 12.12016533 35.09591 5 0 2 633812.9418 0.07409722 1.428286 6 0 3 633812.94167 0.00012233 0.004424 The SAS System 13:41 Wednesday, October 14, 2015 8 The GLIMMIX Procedure Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 633812.9 AIC (smaller is better) 633816.9 AICC (smaller is better) 633816.9 BIC (smaller is better) 633835.3 CAIC (smaller is better) 633837.3 HQIC (smaller is better) 633822.6 Generalized Chi-Square 121590.0 Gener. Chi-Square / DF 0.53 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 1.0169 0.006873 Residual 0.5263 0.001853 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept 5.6664 0.02831 68328 200.14 <.0001 Age -0.02445 0.000437 163E3 -55.99 <.0001 Gender 0 0.2515 0.02203 68328 11.42 <.0001 Gender 1 0 . . . . Year -0.03249 0.001152 163E3 -28.21 <.0001 Year*Gender 0 0.003955 0.001536 163E3 2.58 0.0100 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 163E3 3134.61 <.0001 Gender 1 68328 130.36 <.0001 Year 1 163E3 1372.28 <.0001 Year*Gender 1 163E3 6.63 0.0100 Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 4.0360 0.005888 68328 685.47 <.0001 0.05 4.0244 4.0475 1 3.7337 0.006249 68328 597.50 <.0001 0.05 3.7214 3.7459 The SAS System 13:41 Wednesday, October 14, 2015 9 The SAS System 12:36 Sunday, October 25, 2015 1 The GLIMMIX Procedure Model Information Data Set SYS.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 245653 Dimensions R-side Cov. Parameters 2 Columns in X 6 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1348357.0299 . 39680.91 1 0 2 975065.68653 373291.34334 3231.071 2 0 6 840058.38594 135007.30059 737.1448 3 0 6 776678.97397 63379.411973 148.3715 4 0 4 762927.82446 13751.149509 49.80366 5 0 4 758421.71214 4506.1123197 9.610959 6 0 3 758117.48564 304.22649873 3.139451 7 0 3 758066.60859 50.87705335 0.989569 The SAS System 12:36 Sunday, October 25, 2015 2 The GLIMMIX Procedure Iteration History Objective Max Iteration Restarts Evaluations Function Change Gradient 8 0 3 758061.48286 5.12573275 0.086392 9 0 3 758061.44489 0.03796156 0.002138 10 0 3 758061.44487 0.00002362 3.242E-6 Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 758061.4 AIC (smaller is better) 758065.4 AICC (smaller is better) 758065.4 BIC (smaller is better) 758083.8 CAIC (smaller is better) 758085.8 HQIC (smaller is better) 758071.1 Generalized Chi-Square 35229.61 Gener. Chi-Square / DF 0.14 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 102.86 0.5462 Residual 0.1434 0.000485 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept 49.3252 0.05619 71052 877.85 <.0001 Gender 0 -0.9134 0.07708 71052 -11.85 <.0001 Gender 1 0 . . . . Year 0.9855 0.000617 175E3 1596.77 <.0001 Year*Gender 0 0.001549 0.000845 175E3 1.83 0.0668 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Gender 1 71052 140.43 <.0001 Year 1 175E3 5445178 <.0001 Year*Gender 1 175E3 3.36 0.0668 The SAS System 12:36 Sunday, October 25, 2015 3 The GLIMMIX Procedure Gender Least Squares Means Standard Gender Estimate Error DF t Value Pr > |t| Alpha Lower Upper 0 60.9628 0.05221 71052 1167.70 <.0001 0.05 60.8605 61.0651 1 61.8565 0.05561 71052 1112.37 <.0001 0.05 61.7476 61.9655 The SAS System 12:36 Sunday, October 25, 2015 6 The GLIMMIX Procedure Model Information Data Set SYS.ZODIAC_45_UNDER_75 Response Variable Gender 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 71054 not printed Number of Observations Read 245653 Number of Observations Used 245653 Dimensions R-side Cov. Parameters 1 Columns in X 2 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1039600.6966 0.19363314 . 1 0 0 1038676.0096 0.00041825 . 2 0 0 1038677.0283 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 12:36 Sunday, October 25, 2015 7 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 1038677 Generalized Chi-Square 245653.0 Gener. Chi-Square / DF 1.00 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0000 0.002853 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept 0.05762 0.01771 71053 3.25 0.0011 Year -0.01390 0.001357 175E3 -10.24 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Year 1 175E3 104.96 <.0001 The SAS System 12:36 Sunday, October 25, 2015 8 The GLIMMIX Procedure Model Information Data Set SYS.ZODIAC_45_UNDER_75 Response Variable Gender 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 71054 not printed Number of Observations Read 245653 Number of Observations Used 245653 Dimensions R-side Cov. Parameters 1 Columns in X 3 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1040202.1756 0.19242169 . 1 0 0 1038999.9846 0.00106666 . 2 0 0 1039003.1046 0.00000001 . 3 0 0 1039003.1046 0.00000000 . Convergence criterion (PCONV=1.11022E-8) satisfied. The SAS System 12:36 Sunday, October 25, 2015 9 The GLIMMIX Procedure Fit Statistics -2 Res Log Pseudo-Likelihood 1039003 Generalized Chi-Square 245659.9 Gener. Chi-Square / DF 1.00 Covariance Parameter Estimates Standard Cov Parm Estimate Error Residual (VC) 1.0000 0.002853 Solutions for Fixed Effects Standard Effect Estimate Error DF t Value Pr > |t| Intercept -0.4316 0.03330 71053 -12.96 <.0001 Age 0.007959 0.000459 175E3 17.35 <.0001 Year -0.01419 0.001358 175E3 -10.45 <.0001 Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 175E3 301.04 <.0001 Year 1 175E3 109.16 <.0001 The SAS System 12:36 Sunday, October 25, 2015 10 The GLIMMIX Procedure Model Information Data Set SYS.ZODIAC_45_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 71054 not printed Gender 2 0 1 Number of Observations Read 245653 Number of Observations Used 238144 Dimensions R-side Cov. Parameters 2 Columns in X 7 Columns in Z per Subject 0 Subjects (Blocks in V) 71054 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 1004749.4046 . 9787.97 1 0 2 959298.06153 45451.343097 1947.215 2 0 4 949839.50705 9458.5544746 478.9688 3 0 4 948279.03796 1560.4690897 52.53826 4 0 2 948251.43295 27.60501775 17.2442 5 0 2 948247.90416 3.52878827 0.99522 6 0 3 948247.89205 0.01210264 0.006636 7 0 2 948247.89205 -0.00000001 0.006636 The SAS System 12:36 Sunday, October 25, 2015 11 The GLIMMIX Procedure Convergence criterion (GCONV=1E-8) satisfied. Fit Statistics -2 Res Log Likelihood 948247.9 AIC (smaller is better) 948251.9 AICC (smaller is better) 948251.9 BIC (smaller is better) 948270.2 CAIC (smaller is better) 948272.2 HQIC (smaller is better) 948257.5 Generalized Chi-Square 201043.9 Gener. Chi-Square / DF 0.84 Covariance Parameter Estimates Standard Cov Parm Subject Estimate Error CS BSISID 28.5721 0.1570 Residual 0.8442 0.002912 Solutions for Fixed Effects Standard Effect Gender Estimate Error DF t Value Pr > |t| Intercept -18.0901 0.1019 69053 -177.45 <.0001 Age 0.2152 0.001919 169E3 112.12 <.0001 Gender 0 -0.1674 0.04936 69053 -3.39 0.0007 Gender 1 0 . . . . Year 0.7810 0.002380 169E3 328.09 <.0001 Year*Gender 0 -0.00494 0.002073 169E3 -2.38 0.0172 Year*Gender 1 0 . . . . Type III Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Age 1 169E3 12571.6 <.0001 Gender 1 69053 11.50 0.0007 Year 1 169E3 135469 <.0001 Year*Gender 1 169E3 5.67 0.0172 The SAS System 12:36 Sunday, October 25, 2015 12