LISTSERV at the University of Georgia
Menubar Imagemap
Home Browse Manage Request Manuals Register
Previous messageNext messagePrevious in topicNext in topicPrevious by same authorNext by same authorPrevious page (July 2006, week 2)Back to main SAS-L pageJoin or leave SAS-L (or change settings)ReplyPost a new messageSearchProportional fontNon-proportional font
Date:         Mon, 10 Jul 2006 17:22:12 -0400
Reply-To:     Chiao-Wen Hsiao <chsiao@USUHS.MIL>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Chiao-Wen Hsiao <chsiao@USUHS.MIL>
Subject:      piecewise regression in mixed model (random intercept random
              coeff)
Content-Type: text/plain; charset=US-ASCII

Hi all,

I am trying to set up a piecewise reg in mixed model. Can someone help to verify my codes? Thanks so much!!

The goal of the analysis is to model weight changes over time for the three treatment groups in the study. I identified two knots where weight changes may be observed: week 1 of the treatment and the beginning of follow-up (3 months f/u).

-group2: treatment type variable (3 groups) -idnumber: identifying number for the individuals enrolled in the study -twk2: time variable, number of weeks (0=BL, 2=WK1, 3=WK2, 4=WK3, 5=WK4, 6=WK5, 7=WK6, 8=WK7, 9=WK8, 10=WK9, 11=WK10, 12=WK11, 13=WK12, 14=WK13, 26=3 months f/u, 38=6 months f/u, 72=12 months f/u, 96=18 months f/u, 120=24 months f/u) -group1: rounds of the group (10 rounds) -wtbase: baseline weight

knot at week 1 notbase: 0 = not week1; 1 = week1 and on week1 (continuous var): 0 = baseline; TWK2(time variable) - 2 knot at 3 months f/u nottx: 0 = not 3 months f/u; 1 = 3 months f/u and on followup (continuous var): 0 = not follow up; TWK2(time variable) - 26

twk2 0 2 3 4 5 6 7 8 9 10 11 12 13 14 26 38 72 96 120 notbase 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 week1 0 0 1 2 3 4 5 6 7 8 9 10 11 12 24 36 70 94 118 nottx 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 followup 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 14 66 70 94

proc mixed data=bct.t_aa; class group2 idnumber twk2 group1 notbase nottx; model wt = wtbase group2 notbase notbase*group2 week1 week1*group2 nottx nottx*group2 followup followup*group2/s ddfm=res outpredm=predicted; random intercept notbase week1 nottx followup/sub=idnumber; repeated twk2/type=AR(1) sub=idnumber group=group2*group1; run;

Thanks so much in advance,

- Joyce Hsiao

------------------------------------- Chiaowen Joyce Hsiao Research Statistician USU Center for Health Disparities & Department of Medical and Clinical Psychology Uniformed Services University of the Health Sciences 4301 Jones Bridge Road Bethesda, Maryland 20814 (w) 301-2951530 (fax) 301-2951650 USUCHD: www.usuchd.org USUMPS:http://www.usuhs.mil/mps/Psychology/


Back to: Top of message | Previous page | Main SAS-L page