How do I interpret subsets in Tukey's HSD output?

When performing Tukey's HSD test in SPSS gives subsets of the groups being compared for alpha=0.05.

This table is a summary of all the pairwise group comparisons made in the previous table. The means of the groups in each subset do not differ from each other. This is shown by the p-value for each subset being above 0.05.

For example if there are three groups (Controls, Alzheimer's and Frontal Temporal Development) Tukey's HSD first performs all three pairwise group mean comparisons. Suppose the control mean differs from both the other two groups but the two patient group means do not differ using pairwise Tukey's HSD then there are two subsets consisting of {Controls} and {Alzheimer's, Frontal Temporal Development}.