Analysis of linear mixed effect model

I have a hard time analyze my results from using the nlme package in R.

This is the scenario:

I got a repeated measures analysis (7 timepoints within each of the 20 subjects). I got three response variables, which might affect each other: Having a higher responsevariable1 (the length of a protein) might be correlated to responsevariable2 (the size of the gene expressing this protein). I have three factors (one is age, one is genotype and the third one is the application of a drug), which might interact (the drug for example might only work in old subjects and thus only there the protein and gene size increase).

As a summary:
factor of 2: factor1 (age: young or old)
factor of 2: factor2 (genotype: genotype1 or genotype2)
factor of 2: factor3 (drug: yes or no)
factor of 7: time (timepoint 1, 2, 3, 4, 5, 6 or 7)
subject: 1 to 20
formula: rv1*rv2*rv3 ~ time*f1*f2*f3 random = ~1|subject method = "ML"
rv = response variable and f = factor

Now i wanted to describe this interaction, which you can see in the picture:

How can i interpret the p-value 
time:factor1(6m) (p = 0.0437), 
time:factor3nvns (p = 0.0012) and 
time:factor16m:factor3nVNS (p = 0.0236)??

Does it mean that for example the three response variables change over time is dependent on factor 1 (= age)? Or what would be a better interpretation?

And how to interpret the time:factor1:factor3 positive p-value?

Thanks a lot in advance!


1 Answer 1


I don't know what you expect but I don't think lme does what you expect. These are the same models:

lme(Sepal.Length * Sepal.Width ~ Petal.Length, random = ~1|Species, data = iris)

iris <- within(iris, y <- Sepal.Length * Sepal.Width)
lme(y ~ Petal.Length, random = ~1|Species, data = iris)

If you need a multivariate mixed model, you need different software.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.