Background: Therapeutic research into Alzheimer’s disease (AD) has been dominated by the amyloid cascade hypothesis (ACH) since the 1990s. However, targeting amyloid in AD patients has not yet resulted in highly significant disease-modifying effects. Furthermore, other promising theories of AD etiology exist. Objective: We sought to directly investigate whether the ACH still dominates the opinions of researchers working on AD and explore the implications of this question for future directions of research. Methods: During 2019, we undertook an international survey promoted with the help of the Alzheimer’s Association with questions on theories and treatments of AD. Further efforts to promote a similar study in 2021 did not recruit a significant number of participants. Results: 173 researchers took part in the 2019 survey, 22% of which held “pro-ACH” opinions, tended to have more publications, were more likely to be male, and over 60. Thus, pro-ACH may now be a minority opinion in the field but is nevertheless the hypothesis on which the most clinical trials are based, suggestive of a representation bias. Popular vote of all 173 participants suggested that lifestyle treatments and anti-tau drugs were a source of more therapeutic optimism than anti-amyloid treatments. Conclusion: We propose a more democratic research structure which increases the likelihood that promising theories are published and funded fairly, promotes a broader scientific view of AD, and reduces the larger community’s dependence on a fragile economic model.

The amyloid cascade hypothesis (ACH) has dominated contemporary biomedical research into Alzheimer's disease (AD) since the 1990s but lacks confirmation by successful clinical trials of anti-amyloid medicines in human AD. In this uncertain period regarding the centrality of betaamyloid (Aβ) in the AD disease process, and with the community apparently divided about the ACH's validity, we used citation practices as a proxy for measuring how researchers have invested their belief in the hypothesis between 1992 and 2019. We sampled 445 articles citing Hardy & Higgins ("HH92") and classified the polarity of their HH92 citation according to Greenberg (2009)'s citation taxonomy of positive, neutral, and negative citations, and then tested four hypotheses. We identified two major attitudes towards HH92: a majority (62.7%) of neutral attitudes with consistent properties across the time period, and a positive attitude (35.0%), tending to cite HH92 earlier on within the bibliography as time went by, tending to take HH92 as