Alzheimer's disease, in its early onset familial form, is known to be a heterogeneous disorder. This suggests that the different degenerative mechanisms, initiated by different genetic causes and ending in the shared phenotype of the disease, should intersect at some point in the degenerative cascade to form a 'bottleneck' from which the pathological features that are common to each of the genetic forms emerge. A growing body of evidence suggests that disturbances of energy metabolism may play a fundamental role in the onset and progression of Alzheimer's disease. In light of this, we propose a 'mitochondrial bottleneck hypothesis', which unifies the various forms of the disease in which different causes lead to the disorder via disturbances of mitochondrial function. The characterization of such a bottleneck would present a unique target for therapeutic intervention because it would be the earliest point in a neurodegenerative cascade shared by all forms of Alzheimer's disease, independent of cause.