Calculating Incidence Rate

[joeception]

I have been looking through several of the previously asked questions on the topic and I am still having a bit of trouble putting together a code that would work with my data set.

I am working on calculating the incidence rate for obesity and have 6 separate time points spread out over about 60 years.

I know for example it should be set up in such a way that it is written BMI_Timepoint1 > 30 = 1 and so on. (defining obesity as over 30).

Any tips on where to get started? I am a new learner to SAS some every little bit of help is appreciated.

Variables I have data for: BMI, age, ID#, birthdays, etc

[SASKiwi]

Start by defining what your incident rate is in plain English. Is it a BMI greater than 30 on 1 or more of the observation points for a subject? Or do they have to have a BMI greater than 30 for 2 or more observation points? Until you nail your definition you can't code it.

[StatDave]

If your data are clusters of correlated observations (such as subjects repeatedly measured over time) on a binary response, there are several modeling methods you can use. One of them is a Generalized Estimating Equations (GEE) model which can be fit in PROC GEE or PROC GENMOD using their REPEATED statement. See the example titled "Generalized Estimating Equations" in the Overview section of the PROC GENMOD documentation. You can add an OUTPUT statement with the PRED= option to save the predicted probabilities for each of the observations in your data. This will give you the predicted incidence rate in each population of your study, where a population is defined by one setting of the predictors in your model.

[joeception]

Thank you both, I will take some time to try and work it out and come back with the code I generate.