I neeed advice on how to analyze cumulative measures

[marcel]

Hello,

I need to test if a cumulative measure over time is different in an experimental group from a control group. It is a measure of children over time. These are animals, so the children leave the parents and the parents go on to have a new litter. I have 100 subjects followed over 5 time points. Below is an example of 2 subjects.

Times 1 to 5 are the sampling points that can be analyzed by using repeated measures (not interested in this right now). Cumultime1 to Cumultime5 are the cumulative number of children during the study that I want to analyze. What type of analysis can be used for cumulative number of children in the example below? What good information source on the analysis can you recommend?

            Time1          Time 2        Time3        Time4         Time5           Cumultime1   Cumultime2     Cumultime3   Cumultime4   Cumultime5

Sub1    5 children    6 children   4 children   4 children    0 children     5 children      11 children       15 children    19 children     19 children

Sub2    7 children    7 children   2 children   0 children    2 children     7 children      14 children       16 children    16 children     18 children

Thank you for your time.

Regards,

Marcel

[SteveDenham]

One analysis would just be to compare the cumulative values for the fifth period. All that is needed there is a t test.  If you want to compare the cumulative profiles, then a MANOVA using PROC GLM could be used. Once you do the repeated measures analysis (PROC GLIMMIX with a Poisson distribution), you could look at partial areas under the curve using LSMESTIMATE statements.

SteveDenham

[marcel]

Dear Steve,

Thank you for your answer. The final goal of this work is to find out at what early time (from CumulTime1 to CumulTime5 out of 15 CumulTime points) the cumulative number of children is significantly higher or lower in the experimental group compared to the control group. (I abbreviated my data for simplicity in my first message).

If I understood well your message, two types of analyses are applicable to cumulative number of children:

1. a MANOVA using PROC GLM

2. a repeated measures analysis (PROC GLIMMIX with a Poisson distribution).

These two analyses would take care of the adjustment for multiple comparisons when this adjustment is entered in the respective sas code.

Is that right?

Regards,

Marcel

[SteveDenham]

An adjustment for multiple comparisons can be added as an option to the LSMEANS statement in GLIMMIX, which would be my choice. @StatDave may recommend GENMOD or GEE, as there doesn't appear to be additional random effects in play, only repeated measures.

SteveDenham