Subject-specific observed profiles of change from baseline vs week trt=10000u

Similar documents
Latin Square Design. Design of Experiments - Montgomery Section 4-2

Statistical Consulting Topics. RCBD with a covariate

Modelling Intervention Effects in Clustered Randomized Pretest/Posttest Studies. Ed Stanek

Mixed Effects Models Yan Wang, Bristol-Myers Squibb, Wallingford, CT

Mixed Models Lecture Notes By Dr. Hanford page 151 More Statistics& SAS Tutorial at Type 3 Tests of Fixed Effects

Replicated Latin Square and Crossover Designs

GLM Example: One-Way Analysis of Covariance

Mixed models in R using the lme4 package Part 2: Longitudinal data, modeling interactions

Problem Points Score USE YOUR TIME WISELY USE CLOSEST DF AVAILABLE IN TABLE SHOW YOUR WORK TO RECEIVE PARTIAL CREDIT

K ABC Mplus CFA Model. Syntax file (kabc-mplus.inp) Data file (kabc-mplus.dat)

PROC GLM AND PROC MIXED CODES FOR TREND ANALYSES FOR ROW-COLUMN DESIGNED EXPERIMENTS

Linear mixed models and when implied assumptions not appropriate

1'-tq/? BU-- _-M August 2000 Technical Report Series of the Department of Biometrics, Cornell University, Ithaca, New York 14853

More About Regression

DV: Liking Cartoon Comedy

Resampling Statistics. Conventional Statistics. Resampling Statistics

Box-Jenkins Methodology: Linear Time Series Analysis Using R

Supplementary Figures Supplementary Figure 1 Comparison of among-replicate variance in invasion dynamics

Exercises. ASReml Tutorial: B4 Bivariate Analysis p. 55

For these items, -1=opposed to my values, 0= neutral and 7=of supreme importance.

I. Model. Q29a. I love the options at my fingertips today, watching videos on my phone, texting, and streaming films. Main Effect X1: Gender

Mixed Linear Models. Case studies on speech rate modulations in spontaneous speech. LSA Summer Institute 2009, UC Berkeley

RANDOMIZED COMPLETE BLOCK DESIGN (RCBD) Probably the most used and useful of the experimental designs.

TWO-FACTOR ANOVA Kim Neuendorf 4/9/18 COM 631/731 I. MODEL

Sociology 7704: Regression Models for Categorical Data Instructor: Natasha Sarkisian

8 Nonparametric test. Question 1: Are (expected) value of x and y the same?

Chapter 27. Inferences for Regression. Remembering Regression. An Example: Body Fat and Waist Size. Remembering Regression (cont.)

Bootstrap Methods in Regression Questions Have you had a chance to try any of this? Any of the review questions?

The Time Series Forecasting System Charles Hallahan, Economic Research Service/USDA, Washington, DC

RCBD with Sampling Pooling Experimental and Sampling Error

UPDATED STANDARDIZED CATCH RATES OF BLUEFIN TUNA (THUNNUS THYNNUS) FROM THE TRAP FISHERY IN TUNISIA

STAT 113: Statistics and Society Ellen Gundlach, Purdue University. (Chapters refer to Moore and Notz, Statistics: Concepts and Controversies, 8e)

Discriminant Analysis. DFs

Model II ANOVA: Variance Components

SECTION I. THE MODEL. Discriminant Analysis Presentation~ REVISION Marcy Saxton and Jenn Stoneking DF1 DF2 DF3

Moving on from MSTAT. March The University of Reading Statistical Services Centre Biometrics Advisory and Support Service to DFID

MANOVA COM 631/731 Spring 2017 M. DANIELS. From Jeffres & Neuendorf (2015) Film and TV Usage National Survey

Repeated measures ANOVA

K-Pop Idol Industry Minhyung Lee

Paired plot designs experience and recommendations for in field product evaluation at Syngenta

2. ctifile,s,h, CALDB,,, ACIS CTI ARD file (NONE none CALDB <filename>)

MANOVA/MANCOVA Paul and Kaila

Multiple-point simulation of multiple categories Part 1. Testing against multiple truncation of a Gaussian field

For these exercises, use SAS data sets stored in a permanent SAS data library.

AP Statistics Sampling. Sampling Exercise (adapted from a document from the NCSSM Leadership Institute, July 2000).

Detecting Medicaid Data Anomalies Using Data Mining Techniques Shenjun Zhu, Qiling Shi, Aran Canes, AdvanceMed Corporation, Nashville, TN

Release Year Prediction for Songs

Normalization Methods for Two-Color Microarray Data

Variation in fibre diameter profile characteristics between wool staples in Merino sheep

K3. Why did the certain ethnic mother put her baby in a crib with 20-foot high legs? So she could hear it if it fell out of bed.

Block Block Block

Algebra I Module 2 Lessons 1 19

Visual Encoding Design

Sociology 704: Topics in Multivariate Statistics Instructor: Natasha Sarkisian

ISC0904: 1k x 1k 18µm N-on-P ROIC. Specification January 13, 2012

Predicting the Importance of Current Papers

Analysis of local and global timing and pitch change in ordinary

1. Model. Discriminant Analysis COM 631. Spring Devin Kelly. Dataset: Film and TV Usage National Survey 2015 (Jeffres & Neuendorf) Q23a. Q23b.

Relationships Between Quantitative Variables

Using assessment and research to promote learning. Thakur B. Karkee, Ph. D. Measurement Incorporated. Kevin Fatica CTB/McGraw-Hill

K3. Why did the certain ethnic mother put her baby in a crib with 20-foot high legs? So she could hear it if it fell out of bed.

The Human Features of Music.

On the design of turbo codes with convolutional interleavers

Relationships. Between Quantitative Variables. Chapter 5. Copyright 2006 Brooks/Cole, a division of Thomson Learning, Inc.

A STATISTICAL VIEW ON THE EXPRESSIVE TIMING OF PIANO ROLLED CHORDS

ASReml Tutorial: C1 Variance structures p. 1. ASReml tutorial. C1 Variance structures. Arthur Gilmour

subplots (30-m by 33-m) without space between potential subplots. Depending on the size of the

WEB APPENDIX. Managing Innovation Sequences Over Iterated Offerings: Developing and Testing a Relative Innovation, Comfort, and Stimulation

Connection for filtered air

in the Howard County Public School System and Rocketship Education

Cohere Technologies Performance Evaluation of OTFS Waveform in Multi User Scenarios Agenda item: Document for: Discussion

UC San Diego UC San Diego Previously Published Works

MATH& 146 Lesson 11. Section 1.6 Categorical Data

Supplemental Material: Color Compatibility From Large Datasets

Best Pat-Tricks on Model Diagnostics What are they? Why use them? What good do they do?

Restoration of Hyperspectral Push-Broom Scanner Data

Reproducibility Assessment of Independent Component Analysis of Expression Ratios from DNA microarrays.

Setting Energy Efficiency Requirements Using Multivariate Regression

Improvised Duet Interaction: Learning Improvisation Techniques for Automatic Accompaniment

Implementation of an MPEG Codec on the Tilera TM 64 Processor

Music Genre Classification

Lecture 5: Clustering and Segmentation Part 1

DOES MOVIE SOUNDTRACK MATTER? THE ROLE OF SOUNDTRACK IN PREDICTING MOVIE REVENUE

GENOTYPE AND ENVIRONMENTAL DIFFERENCES IN FIBRE DIAMETER PROFILE CHARACTERISTICS AND THEIR RELATIONSHIP WITH STAPLE STRENGTH IN MERINO SHEEP

Optimum Frame Synchronization for Preamble-less Packet Transmission of Turbo Codes

LCD and Plasma display technologies are promising solutions for large-format

Placement Rent Exponent Calculation Methods, Temporal Behaviour, and FPGA Architecture Evaluation. Joachim Pistorius and Mike Hutton

Package spotsegmentation

a user's guide to Probit Or LOgit analysis

Do Television and Radio Destroy Social Capital? Evidence from Indonesian Villages Online Appendix Benjamin A. Olken February 27, 2009

Hidden Markov Model based dance recognition

Modeling memory for melodies

FPA (Focal Plane Array) Characterization set up (CamIRa) Standard Operating Procedure

Tutorial 0: Uncertainty in Power and Sample Size Estimation. Acknowledgements:

Exploring the Effects of Pitch Layout on Learning a New Musical Instrument

Data Mining. Dr. Raed Ibraheem Hamed. University of Human Development, College of Science and Technology Department of CS

A Statistical Framework to Enlarge the Potential of Digital TV Broadcasting

Elasticity Imaging with Ultrasound JEE 4980 Final Report. George Michaels and Mary Watts

The Great Beauty: Public Subsidies in the Italian Movie Industry

System Identification

Transcription:

Mean of age 1 The MEANS Procedure Analysis Variable : age N Mean Std Dev Minimum Maximum ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ 109 55.5321101 12.1255537 26.0000000 83.0000000 ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ Subject-specific observed profiles of change from baseline vs week trt=10000u diff 20 0-20 -40 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 week Subject-specific observed profiles of change from baseline vs week trt=5000u diff 20 0-20 -40 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 week

Subject-specific observed profiles of change from baseline vs week trt=placebo diff 20 0-20 -40 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 week Raw Data: Change scores vs Age diff 30 20 10 0-10 -20-30 -40-50 20 30 40 50 60 70 80 90 age

diff Boxplots of raw data (change scores) by gender 40 20 0-20 -40-60 F sex M

Model M1: Full model 2 Unspecified var-cov structure Model Information Data Set Dependent Variable Covariance Structures Subject Effects Estimation Method Residual Variance Method Fixed Effects SE Method Degrees of Freedom Method WORK.THREE diff Variance Components, Unstructured site, id(site) REML None Kenward-Roger Kenward-Roger Class Level Information Class Levels Values id 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 trt 3 10000U 5000U Placebo site 9 1 2 3 4 5 6 7 8 9 sex 2 F M week 5 2 4 8 12 16 measocc 5 2 3 4 5 6 Dimensions Covariance Parameters 16 Columns in X 98 Columns in Z Per Subject 1 Subjects 9 Max Obs Per Subject 95 Number of Observations Number of Observations Read 545 Number of Observations Used 522 Number of Observations Not Used 23 Iteration History Iteration Evaluations -2 Res Log Like Criterion 0 1 3600.93973714 1 2 3396.97724037 0.00001347 2 1 3396.95992468 0.00000001 3 1 3396.95990814 0.00000000

Model M1: Full model 3 Unspecified var-cov structure Convergence criteria met. Estimated R Matrix for id(site) 1 1 Row Col1 Col2 Col3 Col4 Col5 1 70.2952 36.4542 31.6631 29.4481 22.2192 2 36.4542 61.6512 46.9627 29.1634 21.9909 3 31.6631 46.9627 77.3702 45.2065 28.7920 4 29.4481 29.1634 45.2065 69.3619 50.6900 5 22.2192 21.9909 28.7920 50.6900 78.4851 Estimated R Correlation Matrix for id(site) 1 1 Row Col1 Col2 Col3 Col4 Col5 1 1.0000 0.5538 0.4293 0.4217 0.2991 2 0.5538 1.0000 0.6800 0.4460 0.3161 3 0.4293 0.6800 1.0000 0.6171 0.3695 4 0.4217 0.4460 0.6171 1.0000 0.6870 5 0.2991 0.3161 0.3695 0.6870 1.0000 Covariance Parameter Estimates Z Cov Parm Subject Estimate Error Value Pr Z Intercept site 0.06721 1.9983 0.03 0.4866 UN(1,1) id(site) 70.2952 10.6244 6.62 <.0001 UN(2,1) id(site) 36.4542 8.3284 4.38 <.0001 UN(2,2) id(site) 61.6512 9.5236 6.47 <.0001 UN(3,1) id(site) 31.6631 8.8164 3.59 0.0003 UN(3,2) id(site) 46.9627 8.7722 5.35 <.0001 UN(3,3) id(site) 77.3702 11.1893 6.91 <.0001 UN(4,1) id(site) 29.4481 8.2834 3.56 0.0004 UN(4,2) id(site) 29.1634 7.7951 3.74 0.0002 UN(4,3) id(site) 45.2065 8.9780 5.04 <.0001 UN(4,4) id(site) 69.3619 10.2367 6.78 <.0001 UN(5,1) id(site) 22.2192 8.1412 2.73 0.0063 UN(5,2) id(site) 21.9909 7.6484 2.88 0.0040 UN(5,3) id(site) 28.7920 8.5322 3.37 0.0007 UN(5,4) id(site) 50.6900 9.1842 5.52 <.0001 UN(5,5) id(site) 78.4851 11.4405 6.86 <.0001

Model M1: Full model 4 Unspecified var-cov structure Fit Statistics -2 Res Log Likelihood 3397.0 AIC (smaller is better) 3429.0 AICC (smaller is better) 3430.1 BIC (smaller is better) 3432.1 agec 1 97 1.72 0.1922 sex 1 95.6 1.09 0.3001 agec*sex 1 93.9 2.53 0.1149 trt 2 93.9 5.06 0.0082 agec*trt 2 94.8 2.19 0.1174 week 4 94.1 19.52 <.0001 agec*week 4 90 2.44 0.0526 trt*week 8 131 2.24 0.0283 agec*trt*week 8 127 1.18 0.3192 trt*sex 2 96.9 0.59 0.5551 sex*week 4 93.9 0.60 0.6642 trt*sex*week 8 131 0.89 0.5293 Least Squares Means Effect trt week Estimate Error DF t Value Pr > t trt 10000U -6.7958 1.3034 51.5-5.21 <.0001 trt 5000U -5.9849 1.1002 41.4-5.44 <.0001 trt Placebo -1.8226 1.1554 49.6-1.58 0.1210 week 2-8.0900 0.9091 21.2-8.90 <.0001 week 4-8.3454 0.8328 16.7-10.02 <.0001 week 8-5.8699 0.9368 20.9-6.27 <.0001 week 12-2.4661 0.9097 19.5-2.71 0.0136 week 16 0.4325 0.9507 18.6 0.45 0.6544 trt*week 10000U 2-11.7219 1.6984 82.3-6.90 <.0001 trt*week 10000U 4-12.5073 1.5443 79.7-8.10 <.0001 trt*week 10000U 8-8.9155 1.7655 77.3-5.05 <.0001 trt*week 10000U 12-2.8369 1.7705 84-1.60 0.1128 trt*week 10000U 16 2.0028 1.7969 63.8 1.11 0.2692 trt*week 5000U 2-9.1533 1.4431 74.1-6.34 <.0001 trt*week 5000U 4-9.0616 1.3388 73.4-6.77 <.0001 trt*week 5000U 8-7.0119 1.4961 68.3-4.69 <.0001 trt*week 5000U 12-3.6045 1.4085 63.9-2.56 0.0129 trt*week 5000U 16-1.0935 1.5078 56.6-0.73 0.4713 trt*week Placebo 2-3.3949 1.5324 82.2-2.22 0.0295

Model M1: Full model 5 Unspecified var-cov structure Least Squares Means Effect trt week Estimate Error DF t Value Pr > t trt*week Placebo 4-3.4672 1.3901 79.8-2.49 0.0147 trt*week Placebo 8-1.6822 1.5529 73.6-1.08 0.2822 trt*week Placebo 12-0.9568 1.4820 71.5-0.65 0.5206 trt*week Placebo 16 0.3881 1.5787 62.3 0.25 0.8066 Group by time profile plots for model M1 Estimate 3.0000 2.0000 1.0000 0-1.0000-2.0000-3.0000-4.0000-5.0000-6.0000-7.0000-8.0000-9.0000-10.0000-11.0000-12.0000-13.0000 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 week trt 10000U 5000U Placebo

AIC values for variants of model M1 with a variety of var-cov structures 6 Obs Descr Value 1 AIC (smaller is better) 3429.0 2 AIC (smaller is better) 3420.8 3 AIC (smaller is better) 3415.2 4 AIC (smaller is better) 3413.8 5 AIC (smaller is better) 3421.3 6 AIC (smaller is better) 3414.0 7 AIC (smaller is better) 3419.3 8 AIC (smaller is better) 3415.8 9 AIC (smaller is better) 3413.9 10 AIC (smaller is better) 3466.5 11 AIC (smaller is better) 3460.6 12 AIC (smaller is better) 3423.6 13 AIC (smaller is better) 3458.0

M4 is best among models considered so far 7 Convergence criteria met. Covariance Parameter Estimates Z Cov Parm Subject Estimate Error Value Pr Z Intercept site 4.53E-17... AR(1) id(site) 0.6404 0.03527 18.16 <.0001 Residual 72.0612 6.3868 11.28 <.0001 Fit Statistics -2 Res Log Likelihood 3409.8 AIC (smaller is better) 3413.8 AICC (smaller is better) 3413.8 BIC (smaller is better) 3414.2 age 1 468 1.75 0.1870 sex 1 468 1.79 0.1814 age*sex 1 468 2.57 0.1096 trt 2 468 1.85 0.1587 age*trt 2 468 2.26 0.1055 week 4 468 1.46 0.2138 age*week 4 468 2.38 0.0505 trt*week 8 468 1.37 0.2080 age*trt*week 8 468 1.17 0.3169 trt*sex 2 468 0.66 0.5190 sex*week 4 468 0.61 0.6558 trt*sex*week 8 468 0.74 0.6546

M4a: same as M4 without site effect 8 Convergence criteria met. Covariance Parameter Estimates Z Cov Parm Subject Estimate Error Value Pr Z AR(1) id(site) 0.6404 0.03527 18.16 <.0001 Residual 72.0613 6.3868 11.28 <.0001 Fit Statistics -2 Res Log Likelihood 3409.8 AIC (smaller is better) 3413.8 AICC (smaller is better) 3413.8 BIC (smaller is better) 3419.1 age 1 98 1.75 0.1894 sex 1 98 1.79 0.1838 age*sex 1 98 2.57 0.1121 trt 2 98 1.85 0.1630 age*trt 2 98 2.26 0.1098 week 4 378 1.46 0.2142 age*week 4 378 2.38 0.0509 trt*week 8 378 1.37 0.2088 age*trt*week 8 378 1.17 0.3175 trt*sex 2 98 0.66 0.5208 sex*week 4 378 0.61 0.6558 trt*sex*week 8 378 0.74 0.6546

Model M4b has ar(1) residual var-cov matrix with trt-specific parameters 9 Convergence criteria met. Estimated R Matrix for id(site) 1 1 Row Col1 Col2 Col3 Col4 Col5 1 59.1924 41.1814 28.6508 19.9330 13.8678 2 41.1814 59.1924 41.1814 28.6508 19.9330 3 28.6508 41.1814 59.1924 41.1814 28.6508 4 19.9330 28.6508 41.1814 59.1924 41.1814 5 13.8678 19.9330 28.6508 41.1814 59.1924 Estimated R Correlation Matrix for id(site) 1 1 Row Col1 Col2 Col3 Col4 Col5 1 1.0000 0.6957 0.4840 0.3367 0.2343 2 0.6957 1.0000 0.6957 0.4840 0.3367 3 0.4840 0.6957 1.0000 0.6957 0.4840 4 0.3367 0.4840 0.6957 1.0000 0.6957 5 0.2343 0.3367 0.4840 0.6957 1.0000 Estimated R Matrix for id(site) 2 1 Row Col1 Col2 Col3 Col4 Col5 1 75.4201 46.6360 28.8373 17.8315 11.0261 2 46.6360 75.4201 46.6360 28.8373 17.8315 3 28.8373 46.6360 75.4201 46.6360 28.8373 4 17.8315 28.8373 46.6360 75.4201 46.6360 5 11.0261 17.8315 28.8373 46.6360 75.4201 Estimated R Correlation Matrix for id(site) 2 1 Row Col1 Col2 Col3 Col4 Col5 1 1.0000 0.6183 0.3824 0.2364 0.1462 2 0.6183 1.0000 0.6183 0.3824 0.2364 3 0.3824 0.6183 1.0000 0.6183 0.3824 4 0.2364 0.3824 0.6183 1.0000 0.6183 5 0.1462 0.2364 0.3824 0.6183 1.0000

Model M4b has ar(1) residual var-cov matrix with trt-specific parameters 10 Estimated R Matrix for id(site) 8 1 Row Col1 Col2 Col3 Col4 Col5 1 82.2313 51.2377 31.9258 19.8927 12.3950 2 51.2377 82.2313 51.2377 31.9258 19.8927 3 31.9258 51.2377 82.2313 51.2377 31.9258 4 19.8927 31.9258 51.2377 82.2313 51.2377 5 12.3950 19.8927 31.9258 51.2377 82.2313 Estimated R Correlation Matrix for id(site) 8 1 Row Col1 Col2 Col3 Col4 Col5 1 1.0000 0.6231 0.3882 0.2419 0.1507 2 0.6231 1.0000 0.6231 0.3882 0.2419 3 0.3882 0.6231 1.0000 0.6231 0.3882 4 0.2419 0.3882 0.6231 1.0000 0.6231 5 0.1507 0.2419 0.3882 0.6231 1.0000 Covariance Parameter Estimates Z Cov Parm Subject Group Estimate Error Value Pr Z Variance id(site) trt 10000U 75.4201 11.3412 6.65 <.0001 AR(1) id(site) trt 10000U 0.6183 0.06258 9.88 <.0001 Variance id(site) trt 5000U 59.1924 9.9515 5.95 <.0001 AR(1) id(site) trt 5000U 0.6957 0.05705 12.20 <.0001 Variance id(site) trt Placebo 82.2313 12.5935 6.53 <.0001 AR(1) id(site) trt Placebo 0.6231 0.06359 9.80 <.0001 Fit Statistics -2 Res Log Likelihood 3399.6 AIC (smaller is better) 3411.6 AICC (smaller is better) 3411.8 BIC (smaller is better) 3427.7 age 1 111 1.62 0.2060 sex 1 102 1.29 0.2592 age*sex 1 101 1.93 0.1681 trt 2 74.8 1.81 0.1708

Model M4b has ar(1) residual var-cov matrix with trt-specific parameters 11 age*trt 2 74.7 2.22 0.1163 week 4 351 1.44 0.2217 age*week 4 349 2.33 0.0554 trt*week 8 299 1.37 0.2069 age*trt*week 8 298 1.18 0.3078 trt*sex 2 74.3 0.64 0.5313 sex*week 4 337 0.58 0.6756 trt*sex*week 8 295 0.76 0.6344

Model M4c has AR(1) residual var-cov matrix with sex-specific parameters 12 Convergence criteria met. Estimated R Matrix for id(site) 1 1 Row Col1 Col2 Col3 Col4 Col5 1 70.4629 46.8258 31.1179 20.6792 13.7423 2 46.8258 70.4629 46.8258 31.1179 20.6792 3 31.1179 46.8258 70.4629 46.8258 31.1179 4 20.6792 31.1179 46.8258 70.4629 46.8258 5 13.7423 20.6792 31.1179 46.8258 70.4629 Estimated R Correlation Matrix for id(site) 1 1 Row Col1 Col2 Col3 Col4 Col5 1 1.0000 0.6645 0.4416 0.2935 0.1950 2 0.6645 1.0000 0.6645 0.4416 0.2935 3 0.4416 0.6645 1.0000 0.6645 0.4416 4 0.2935 0.4416 0.6645 1.0000 0.6645 5 0.1950 0.2935 0.4416 0.6645 1.0000 Estimated R Matrix for id(site) 7 1 Row Col1 Col2 Col3 Col4 Col5 1 74.5208 44.5038 26.5776 15.8721 9.4788 2 44.5038 74.5208 44.5038 26.5776 15.8721 3 26.5776 44.5038 74.5208 44.5038 26.5776 4 15.8721 26.5776 44.5038 74.5208 44.5038 5 9.4788 15.8721 26.5776 44.5038 74.5208 Estimated R Correlation Matrix for id(site) 7 1 Row Col1 Col2 Col3 Col4 Col5 1 1.0000 0.5972 0.3566 0.2130 0.1272 2 0.5972 1.0000 0.5972 0.3566 0.2130 3 0.3566 0.5972 1.0000 0.5972 0.3566 4 0.2130 0.3566 0.5972 1.0000 0.5972 5 0.1272 0.2130 0.3566 0.5972 1.0000

Model M4c has AR(1) residual var-cov matrix with sex-specific parameters 13 Covariance Parameter Estimates Z Cov Parm Subject Group Estimate Error Value Pr Z Variance id(site) sex F 70.4629 8.1026 8.70 <.0001 AR(1) id(site) sex F 0.6645 0.04241 15.67 <.0001 Variance id(site) sex M 74.5208 10.8741 6.85 <.0001 AR(1) id(site) sex M 0.5972 0.06543 9.13 <.0001 Fit Statistics -2 Res Log Likelihood 3407.4 AIC (smaller is better) 3415.4 AICC (smaller is better) 3415.5 BIC (smaller is better) 3426.2 age 1 98 1.77 0.1861 sex 1 98 1.81 0.1818 age*sex 1 98 2.60 0.1104 trt 2 98 1.99 0.1425 age*trt 2 98 2.41 0.0950 week 4 378 1.49 0.2045 age*week 4 378 2.46 0.0452 trt*week 8 378 1.38 0.2051 age*trt*week 8 378 1.17 0.3177 trt*sex 2 98 0.67 0.5162 sex*week 4 378 0.56 0.6927 trt*sex*week 8 378 0.71 0.6832

Model M4b-9. Same as previous model but drop age 14 This is final model when week is a factor Convergence criteria met. Covariance Parameter Estimates Z Cov Parm Subject Group Estimate Error Value Pr Z Variance id(site) trt 10000U 77.0849 11.2200 6.87 <.0001 AR(1) id(site) trt 10000U 0.6404 0.05773 11.09 <.0001 Variance id(site) trt 5000U 60.4942 9.5723 6.32 <.0001 AR(1) id(site) trt 5000U 0.6938 0.05451 12.73 <.0001 Variance id(site) trt Placebo 82.9271 12.2492 6.77 <.0001 AR(1) id(site) trt Placebo 0.6155 0.06273 9.81 <.0001 Fit Statistics -2 Res Log Likelihood 3445.6 AIC (smaller is better) 3457.6 AICC (smaller is better) 3457.8 BIC (smaller is better) 3473.8 trt 2 78.2 3.99 0.0224 week 4 375 18.53 <.0001 trt*week 8 319 1.93 0.0545 Least Squares Means Effect trt week Estimate Error DF t Value Pr > t trt 10000U -6.5438 1.0797 42.3-6.06 <.0001 trt 5000U -6.0313 1.0071 37.8-5.99 <.0001 trt Placebo -2.4877 1.1234 38.6-2.21 0.0328 week 2-7.9426 0.8375 273-9.48 <.0001 week 4-8.5434 0.8283 267-10.31 <.0001 week 8-5.8555 0.8326 271-7.03 <.0001 week 12-2.8514 0.8354 271-3.41 0.0007 week 16 0.08804 0.8350 268 0.11 0.9161 trt*week 10000U 2-10.8019 1.4554 96.2-7.42 <.0001 trt*week 10000U 4-11.8383 1.4519 96.2-8.15 <.0001 trt*week 10000U 8-8.2156 1.4738 100-5.57 <.0001 trt*week 10000U 12-3.4952 1.4786 99.6-2.36 0.0200

Model M4b-9. Same as previous model but drop age 15 This is final model when week is a factor Least Squares Means Effect trt week Estimate Error DF t Value Pr > t trt*week 10000U 16 1.6318 1.4623 95.3 1.12 0.2673 trt*week 5000U 2-9.4078 1.3165 83.1-7.15 <.0001 trt*week 5000U 4-9.3634 1.3029 81.3-7.19 <.0001 trt*week 5000U 8-6.9794 1.3029 81.3-5.36 <.0001 trt*week 5000U 12-3.5000 1.2963 79.9-2.70 0.0085 trt*week 5000U 16-0.9061 1.3061 81.5-0.69 0.4898 trt*week Placebo 2-3.6180 1.5687 95.5-2.31 0.0233 trt*week Placebo 4-4.4286 1.5393 91.7-2.88 0.0050 trt*week Placebo 8-2.3714 1.5393 91.7-1.54 0.1269 trt*week Placebo 12-1.5589 1.5536 93.5-1.00 0.3183 trt*week Placebo 16-0.4615 1.5590 93.1-0.30 0.7678 Group by time profile plots for model M4b-9 Estimate 2.0000 1.0000 0-1.0000-2.0000-3.0000-4.0000-5.0000-6.0000-7.0000-8.0000-9.0000-10.0000-11.0000-12.0000 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 week trt 10000U 5000U Placebo

Model M4b-9(ML). Model M4b-9 fit with ML estimation 16 Convergence criteria met. Covariance Parameter Estimates Z Cov Parm Subject Group Estimate Error Value Pr Z Variance id(site) trt 10000U 74.9452 10.7624 6.96 <.0001 AR(1) id(site) trt 10000U 0.6407 0.05689 11.26 <.0001 Variance id(site) trt 5000U 58.7896 9.1735 6.41 <.0001 AR(1) id(site) trt 5000U 0.6940 0.05372 12.92 <.0001 Variance id(site) trt Placebo 80.5139 11.7204 6.87 <.0001 AR(1) id(site) trt Placebo 0.6155 0.06181 9.96 <.0001 Fit Statistics -2 Log Likelihood 3477.4 AIC (smaller is better) 3519.4 AICC (smaller is better) 3521.3 BIC (smaller is better) 3576.0 trt 2 80.5 4.11 0.0200 week 4 386 19.09 <.0001 trt*week 8 329 1.99 0.0468

Model M5(ML). Same var-cov as m4b-9, but linear effect of time in each trt 17 Convergence criteria met. Covariance Parameter Estimates Z Cov Parm Subject Group Estimate Error Value Pr Z Variance id(site) trt 10000U 76.7667 10.9402 7.02 <.0001 AR(1) id(site) trt 10000U 0.6316 0.05789 10.91 <.0001 Variance id(site) trt 5000U 58.8241 9.1193 6.45 <.0001 AR(1) id(site) trt 5000U 0.6886 0.05426 12.69 <.0001 Variance id(site) trt Placebo 80.6451 11.6939 6.90 <.0001 AR(1) id(site) trt Placebo 0.6117 0.06214 9.84 <.0001 Fit Statistics -2 Log Likelihood 3487.8 AIC (smaller is better) 3511.8 AICC (smaller is better) 3512.4 BIC (smaller is better) 3544.1 Solution for Fixed Effects Effect trt Estimate Error DF t Value Pr > t Intercept -4.5724 1.6330 84.6-2.80 0.0063 trt 10000U -9.4396 2.2488 173-4.20 <.0001 trt 5000U -6.6279 2.1394 163-3.10 0.0023 trt Placebo 0.... week 0.2596 0.1419 156 1.83 0.0692 week*trt 10000U 0.6951 0.1944 317 3.58 0.0004 week*trt 5000U 0.3811 0.1817 303 2.10 0.0368 week*trt Placebo 0.... trt 2 170 9.20 0.0002 week 1 467 67.91 <.0001 week*trt 2 319 6.38 0.0019

Model M5. Refit model M5 with REML and test hypotheses 18 Convergence criteria met. Covariance Parameter Estimates Z Cov Parm Subject Group Estimate Error Value Pr Z Variance id(site) trt 10000U 78.4561 11.3968 6.88 <.0001 AR(1) id(site) trt 10000U 0.6378 0.05787 11.02 <.0001 Variance id(site) trt 5000U 60.2633 9.5375 6.32 <.0001 AR(1) id(site) trt 5000U 0.6946 0.05415 12.83 <.0001 Variance id(site) trt Placebo 82.5252 12.2199 6.75 <.0001 AR(1) id(site) trt Placebo 0.6188 0.06214 9.96 <.0001 Fit Statistics -2 Res Log Likelihood 3488.8 AIC (smaller is better) 3500.8 AICC (smaller is better) 3501.0 BIC (smaller is better) 3516.9 Solution for Fixed Effects Effect trt Estimate Error DF t Value Pr > t Intercept -4.5712 1.6531 81.9-2.77 0.0070 trt 10000U -9.4363 2.2757 168-4.15 <.0001 trt 5000U -6.6261 2.1658 158-3.06 0.0026 trt Placebo 0.... week 0.2598 0.1430 153 1.82 0.0712 week*trt 10000U 0.6953 0.1959 313 3.55 0.0004 week*trt 5000U 0.3810 0.1831 299 2.08 0.0383 week*trt Placebo 0.... trt 2 164 8.98 0.0002 week 1 461 66.92 <.0001 week*trt 2 315 6.28 0.0021

Model M5. Refit model M5 with REML and test hypotheses 19 Convergence criteria met. Covariance Parameter Estimates Z Cov Parm Subject Group Estimate Error Value Pr Z Variance id(site) trt 10000U 78.4561 11.3968 6.88 <.0001 AR(1) id(site) trt 10000U 0.6378 0.05787 11.02 <.0001 Variance id(site) trt 5000U 60.2633 9.5375 6.32 <.0001 AR(1) id(site) trt 5000U 0.6946 0.05415 12.83 <.0001 Variance id(site) trt Placebo 82.5252 12.2199 6.75 <.0001 AR(1) id(site) trt Placebo 0.6188 0.06214 9.96 <.0001 Fit Statistics -2 Res Log Likelihood 3488.8 AIC (smaller is better) 3500.8 AICC (smaller is better) 3501.0 BIC (smaller is better) 3516.9 Solution for Fixed Effects Effect trt Estimate Error DF t Value Pr > t Intercept -4.0515 1.4524 67.5-2.79 0.0069 trt 10000U -8.0458 2.0030 138-4.02 <.0001 trt 5000U -5.8641 1.9143 130-3.06 0.0027 trt Placebo 0.... week2 0.2598 0.1430 153 1.82 0.0712 week2*trt 10000U 0.6953 0.1959 313 3.55 0.0004 week2*trt 5000U 0.3810 0.1831 299 2.08 0.0383 week2*trt Placebo 0.... trt 2 135 8.55 0.0003 week2 1 461 66.92 <.0001 week2*trt 2 315 6.28 0.0021

Model M5. Refit model M5 with REML and test hypotheses 20 Least Squares Means Effect trt week2 Estimate Error DF t Value Pr > t trt 10000U 0.00-12.0973 1.3793 71.9-8.77 <.0001 trt 5000U 0.00-9.9156 1.2470 64.4-7.95 <.0001 trt Placebo 0.00-4.0515 1.4524 67.5-2.79 0.0069 Differences of Least Squares Means Effect trt _trt week2 Estimate Error DF t Value Pr > t Adjustment trt 10000U Placebo 0.00-8.0458 2.0030 138-4.02 <.0001 Dunnett trt 5000U Placebo 0.00-5.8641 1.9143 130-3.06 0.0027 Dunnett Differences of Least Squares Means Effect trt _trt Adj P trt 10000U Placebo 0.0002 trt 5000U Placebo 0.0051

Model M5. Refit model M5 with REML and test hypotheses 21 Convergence criteria met. Covariance Parameter Estimates Z Cov Parm Subject Group Estimate Error Value Pr Z Variance id(site) trt 10000U 78.4561 11.3968 6.88 <.0001 AR(1) id(site) trt 10000U 0.6378 0.05787 11.02 <.0001 Variance id(site) trt 5000U 60.2633 9.5375 6.32 <.0001 AR(1) id(site) trt 5000U 0.6946 0.05415 12.83 <.0001 Variance id(site) trt Placebo 82.5252 12.2199 6.75 <.0001 AR(1) id(site) trt Placebo 0.6188 0.06214 9.96 <.0001 Fit Statistics -2 Res Log Likelihood 3488.8 AIC (smaller is better) 3500.8 AICC (smaller is better) 3501.0 BIC (smaller is better) 3516.9 Solution for Fixed Effects Effect trt Estimate Error DF t Value Pr > t Intercept -0.4137 1.5263 78.9-0.27 0.7871 trt 10000U 1.6878 2.1081 162 0.80 0.4245 trt 5000U -0.5302 1.9984 150-0.27 0.7911 trt Placebo 0.... week16 0.2598 0.1430 153 1.82 0.0712 week16*trt 10000U 0.6953 0.1959 313 3.55 0.0004 week16*trt 5000U 0.3810 0.1831 299 2.08 0.0383 week16*trt Placebo 0.... trt 2 157 0.68 0.5088 week16 1 461 66.92 <.0001 week16*trt 2 315 6.28 0.0021

Model M5. Refit model M5 with REML and test hypotheses 22 Least Squares Means Effect trt week16 Estimate Error DF t Value Pr > t trt 10000U 0.00 1.2741 1.4542 83.8 0.88 0.3835 trt 5000U 0.00-0.9439 1.2900 73.9-0.73 0.4667 trt Placebo 0.00-0.4137 1.5263 78.9-0.27 0.7871 Differences of Least Squares Means Effect trt _trt week16 Estimate Error DF t Value Pr > t Adjustment trt 10000U Placebo 0.00 1.6878 2.1081 162 0.80 0.4245 Dunnett trt 5000U Placebo 0.00-0.5302 1.9984 150-0.27 0.7911 Dunnett Differences of Least Squares Means Effect trt _trt Adj P trt 10000U Placebo 0.6301 trt 5000U Placebo 0.9480

Model M5a. Refit model M5 but add site effect and test 23 Convergence criteria met. Covariance Parameter Estimates Z Cov Parm Subject Group Estimate Error Value Pr Z Intercept site 0... Variance id(site) trt 10000U 78.4561 11.3968 6.88 <.0001 AR(1) id(site) trt 10000U 0.6378 0.05787 11.02 <.0001 Variance id(site) trt 5000U 60.2633 9.5375 6.32 <.0001 AR(1) id(site) trt 5000U 0.6946 0.05415 12.83 <.0001 Variance id(site) trt Placebo 82.5251 12.2199 6.75 <.0001 AR(1) id(site) trt Placebo 0.6188 0.06214 9.96 <.0001 Fit Statistics -2 Res Log Likelihood 3488.8 AIC (smaller is better) 3500.8 AICC (smaller is better) 3501.0 BIC (smaller is better) 3502.0 trt 2 164 8.98 0.0002 week 1 461 66.92 <.0001 week*trt 2 315 6.28 0.0021

2024 © artsdocbox.com Privacy Policy | Terms of Service | Feedback