Comparing both HC groups,
bilateral clusters of local GM volume differences in the putamen,
lentiform nucleus and insula,
suggest a larger gray matter volume in the HC-A group as compared to the HC-B group,
during active pill phase.
These clusters are largest in the active pill-phase,
and appear to fade away in the inactive pill-phase,
suggesting a differential short-term effect of the type of hormones contained in the HC.
When comparing active to inactive pill-phases,
our results indicate that some structures in the limbic system show larger local gray matter volume in the active phase than in the inactive phase.
However,
the regions affected are not the same in both groups; in the HC-A group,
the cingulate gyrus is altered,
and in the HC-B group the affected regions are more frontally located,
and more widespread.
To the best of our knowledge,
this is the first time that a VBM analysis has been performed controlling for the type of hormonal contraceptives.
The inactive phase in the HC groups can be considered equivalent to the early follicular phase of the next cycle.
Therefore we believe that the long-term effects of HC use can be evaluated by comparing the NC follicular phase with the scans performed at the end of the inactive pill phase,
when the effect of synthetic hormones should have been removed from the subjects,
as these have half-life times of 15 to 30 hours.
Again our results show that the type of contraception affects GM volume,
but now with respect to the NC follicular phase.
Differences are more widespread in the HC-B group,
with clusters of smaller GM volume bilaterally in the culmen.
In the HC-A group,
only a small cingulate cluster was observed.
This shows that different types of progestins can differentially modulate GM volumes on the long-term.
Unfortunately,
our study is not perfect and contains important limitations that need to be considered.
Firstly,
the HC-B group is composed of a relatively small cohort of subjects (N=10),
because the splitting of the HC subjects into two sub-groups was not anticipated in the setup of the study,
but was retrospectively performed because of new insights which were published recently (Pletzer and Kerschbaum 2014).
Another limitation is that the endogenous concentrations of synthetic hormones,
contained in HC,
cannot be measured with standard lab-techniques available to us.
In conclusion,
we found that VBM results of regional GM volume differences are significantly influenced by the hormonal environment in female subjects.
The disregard hereof introduces heterogeneity in the data,
and consequently causes a loss of sensitivity and decreased precision of results.
We advocate that these factors should be kept in mind in future VBM study designs involving women.
Similarly,
these confounds must be taken into account when interpreting results of behavioral studies.