I am measuring variable X at several timepoints for a set of replicates. The 0h timepoint values actually come from separate control replicates and are therefore independent from the replicates being measured.

To do a repeated measures ANOVA, I averaged 0h timepoint measurements and entered the mean as the 0h measurement for each replicate but I feel like this is incorrect.

Am I better keeping the distribution of 0h measurements and not doing doing an indepednent rather than repeated measures analysis? Is it possible to merge repeated measures and independent ANOVA somehow?