Alzheimer's disease (AD), an incurable chronic neurodegenerative disease first described by Alois Alzheimer in 1906, has become a significant public health priority. After 10-15 years of preclinical stage, the symptoms develop from mild cognitive impairment to impaired memory, language difficulties and, at last, death in an average time of seven to ten years after diagnosis.After more than one hundred years of research, researchers had made tremendous progress. Nevertheless, Alzheimer's disease is still marked histologically by the accumulation of amyloid beta (Aβ), the emergence of neurofibrillary tangles, and loss of synapse. However, recently, with endless failure in clinical trials, some researchers are now turning to other symptoms such as aberrant lipid metabolism (Hamilton et al., 2015), abnormal calcium transfer (Dolev et al., 2013), and autophagic malfunction (Li et al., 2017). Mitochondria-associated endoplasmic reticulum membrane (MAM) is pivotal for the pathways mentioned above, which are disturbed in AD. Thus, Area-Gomez et al. proposed that disturbed MAM function is a "coherent and unified explanation" for both familial and sporadic AD: MAM hypothesis (Area-Gomez et al., 2012).In this review, we update the knowledge about MAM's role in pathologies shared by familial and sporadic forms of AD. We also highlight the importance of AD models used in studies.