What about the model: INSULIN RESISTANCE = Metabo0min + Metabo30min + Metabo120 + Covariates. I run that for every Metabolite, rank them by F statistc/Variance explained, correct for multiple testing and get my top Metabolite list associated with Insulin Resistance.

Dear @gung, sounds interesting although I am not too experienced with cluster analysis. What do you mean by clustering on random effects? The simple model I originally discarded would be METABOLITE = alpha + beta*(Time) + error; where Time would be coded as "0/1/2" or "0/30/120". In step two, I could apply the beta weight to predict INSULIN RESISTANCE = betab1 + covariateb2 + ... Where would I fit the random effects in? How do I cluster in them? Thanks!

Thanks, gung. But what if, e.g., there are 250 subjects whose metabolite level is, on average 10-20-30 at 0/30/120min, and there are 250 for whom it is 20-10-0? Using the mean for each time point would be misleading, wouldn't it? Is there an a priori approach (not based on a posteriori grouping according to similar time trends) that would work?

A related question: If I can apply a mixed effects model or similar that includes more than 1 predictor (e.g, Metabolite intensity at 0, 30, and 120min in an additive model) - which statistics from this model would you recommend I than use to predict my actual outcome-of-interest (IR, insulin resistance)? IR = (metabolite change 0-30-120)*beta. What should I use to quantify (metabolite change 0-30-120)? Thanks again

Hi gung, Unfortunately there is no prior hypothesis as to how the metabolite intentities change over time. I am generally not too confident with OLS regression modeling, as the individual time course vary quite a bit (i.e. comparing Spaghetti vs. mean plots). Does mixed effect modeling account for interindividual variations in time course? By mixed effects models, I have always understood combining random and fixed effects (I worked a lot with meta-analysis). Can you illustrate, what you mean in this context - thanks a lot!

Non-Linear means: E.g., for person 1, the intensity of M over three time points is 11-20-23. For person 2, it may be 22-19-43 etc. By non-linear, I mean "not a straight line over all three time points". But linear model between, 0/30, 30/120 are possible. Thanks!

Hi, M is the intensity of a metabolite (measured by Mass spectrometry originally). IR is insulin resistance (measured as a continuous variable). It is an epidemiolgical study, where 500 subject had blood taken at 0/30/120 min that where each analysed by MS to evaluate the intensity of M. At time 0min, all subject also had the insulin resistance measured (once). There are about 200 metabolites and the question is: Which among them are significantly associated in their time course with insulin resistance?