Positive Longitudinal Cortical thickness changes?

[Victor E]

I’m working with the longitudinal MRI dataset and computing within-subject difference scores between consecutive visits. During QC, I noticed a number of positive change values (i.e.  apparent increases in cortical thickness or volume from one MRI visit to the next). Conceptually, these differences should generally be zero or negative, so I wanted to check whether such increases are expected in ADNI?

To better understand this, I compiled a list of all observations > +2 SD for any ROI (CSV available upon request or attached if allowed by the list).

Before interpreting these cases, I wanted to ask:

Are positive longitudinal changes of this magnitude considered normal given scanner variability, site changes, or FreeSurfer longitudinal pipeline behavior?
And, is there recommended guidance on how ADNI researchers typically handle or flag these positive change observations during QC?

Any insights, documentation references, or best practices would be greatly appreciated