Hotelling's T-Square test

Dear all,

using a Two-Sample Hotelling's T-Square test, I have tried to determine
whether population mean vectors (including 4 variables each) were
significantly different from each other. However, given the small size
of samples and the fact that they are unbalanced (12 individuals versus
6), I have doubts about the relevance of this test for my dataset.

Could someone tell me whether the Hotelling's T-square test is adequate
in my case or whether I should just use multiple t-tests with a
correction (e.g. Bonferroni) for multiple testing?

Thanking you in advance for your answer.

Best regards,

Romain