Does PCA preserve the order in data?

So, I have a set of features, and want to perform PCA and use the first principal component for furthur analysis. Currently, all my features come from some physical energy functions, so the domain knowledge tells me that the lower energies are better (although there is so much noise in energy values).

So, my question is whether or not after reducing the dimensions to 1st PCA, the data points still have the same physical meanings?

I mean for example if data points that are lower than mean, correspond to the the same points lower than mean in the original set? My guess is that it probably depends on the sign of eigen-values, but I want to make sure