Four Decades of Research in Alzheimer's Disease (1975-2014): A Bibliometric and Scientometric Analysis

Abstract

Background: Bibliometric and scientometric methods can be applied to the study of a research field.

Objective: We hypothesized that a bibliometric and scientometric analysis of the Alzheimer's disease (AD) research field could render trends that provide researchers and funding agencies valuable insight into the history of the field, current tendencies, and potential future directions.

Methods: We performed searches in publicly available databases including PubMed, Scopus, Web of Science, and Alzheimer's Funding Analyzer for the period 1975-2014, and conducted a curve fitting analysis with non-linear regression.

Results: While the rate and impact of publications continue to increase, the number of patents per year is currently declining after peaking in the late 2000s, and the funding budget has plateaued in the last 5-10 years analyzed. Genetics is the area growing at a fastest pace, whereas pathophysiology and therapy have not grown further in the last decade. Among the targets of pathophysiology research, amyloid-β continues to be the focus of greatest interest, with tau and apolipoprotein E stagnant after a surge in the 1990s. The role of inflammation, microglia, and the synapse are other research topics with growing interest. Regarding preventative strategies, education attainment, diet, and exercise are recently gaining some momentum, whereas NSAIDs and statins have lost the spotlight they once had.

Conclusion: Our bibliometric and scientometric analysis provides distinct trends in AD research in the last four decades, including publication and patent output, funding, impact, and topics. Our findings could inform the decision-making of research funding agencies in the near future.

Keywords: Alzheimer’s disease; amyloid-β peptides; bibliometrics; h-index; neurofibrillary tangle; scientometrics; tau proteins.

Fig. 1

Evolution of AD research output and funding. A) Total number (left y axis) and relative frequency (right y axis) of AD publications per year between 1975 and 2014, based on PubMed database. B) Number of patents on AD per year based on Scopus database. C) Aggregated funding amount per year in millions of US dollars, based on Alzheimer’s Funding Analyzer database (www.j-alz/uberresearch.com). D) Number of starting projects funded per year, based on Alzheimer’s Funding Analyzer database.

Fig. 2

Evolution of AD research impact. A) h-index of the AD field per year based on Web of Science database. B) Yearly h-index of the AD field corrected for the number of papers published each year. C) Yearly h-index of the AD field corrected for the year (age) of publication. D) Yearly h-index of the AD field corrected by the number of papers and the year (age) of publication.

Fig. 3

Trends in AD publication subtypes. Graphs represent the relative frequency of publications of each subtype per year with respect to the total number of papers published each year. Classification of articles as journal articles (A), reviews (B), systematic reviews and meta-analyses (C), clinical trials (D), case reports (E), and editorial notes, news, comments, letters, and introductory journal articles (F) was achieved using available PubMed filters. Note that the sum of the frequencies for different publication subtypes in the same year can exceed 100% because some publications are categorized in more than one subtype.

Fig. 4

Trends in human versus animal studies and of human studies by age group. Graphs represent the relative frequency of papers within each target group per year with respect to the total number of papers published each year. Classification of studies in human versus animal (A) and of human studies in different age group (B-F) was achieved using available PubMed filters. Note that the sum of the relative frequencies of human and animal publications and of the different human age groups for any given year exceeds 100% because PubMed may classify some publications as both animal and human studies, and because some human studies may include more than one age group.

Fig. 5

Trends in AD research themes. Graphs represent the relative frequency of papers about each research theme per year with respect to the total number of papers published each year. Classification of papers under diagnosis (A), pathology (B), pathophysiology (C), therapy (D), genetics (E), epidemiology (F), prevention (G), and economics (H) was based on the MeSH subheadings available under the MeSH term “Alzheimer disease” in PubMed database, as explained in the methods section. Note that the sum of the frequencies of different research themes for the same year exceeds 100% in many years because the same article may be categorized under more than one theme.

Fig. 6

Research trends in AD diagnosis. Graphs represent the relative frequency of each research theme per year with respect to the total number of papers published each year. Classification of papers under psychology (A), blood (C), CSF (D), and radionuclide (E) was based on the MeSH subheadings available under the MeSH term “Alzheimer disease” in PubMed database, as explained in the methods section. MRI papers (B) were obtained by a search combining the MeSH terms “Alzheimer disease” and “Magnetic resonance imaging” because the “radiography” MeSH subheading yielded few papers.

Fig. 7

Research trends in AD pathophysiology. Graphs represent the relative frequency of papers on each research topic per year with respect to the total number of papers published each year. Searches were conducted in PubMed database by combining the “Alzheimer disease” MeSH term with MeSH terms of interest. NFTs, neurofibrillary tangles.

Fig. 8

Research trends in AD prevention. Graphs represent the relative frequency of literature on each prevention strategy or factor per year with respect to the total number of papers published each year. Searches were conducted in PubMed database by combining the “Alzheimer disease” MeSH term with MeSH terms “Educational Status” (A), “Diet” (B), “Exercise” (C), “Anti-Inflammatory Agents, Non-Steroidal” (NSAIDs) (D), and “Hydroxymethylglutaryl-CoA Reductase Inhibitors” (statins) (E), as explained in the methods section.