Three Decades of Research on the Development of Ibogaine Treatment of Substance Use Disorders: A Scientometric Analysis

The investigators conducted a scientometric analysis using SCOPUS as the data source; the search was executed on 28 March 2023. A two‑phase search strategy combined Medical Subject Headings (MeSH) and Boolean operators to capture terms related to ibogaine and substance use disorders, with a temporal filter restricting results to 1993–2022. No language or publication‑type restrictions were applied. The complete search syntax is reported in an appendix (not reproduced here). Data processing removed duplicates in Mendeley before title and abstract screening in Rayyan. Bibliometric and network analyses were performed with VOSviewer and CiteSpace; geographic visualisation used QGIS. For quantitative trend testing the team used R: linear regression models estimated yearly publication slopes within each decade, an analysis of covariance compared slopes across decades, and Tukey’s test investigated inter‑decadal differences. Statistical significance was set at p < .05. Network metrics from CiteSpace included citation burst (a rapid increase in citations to a publication), degree (number of connections an item has) and sigma (a composite measure incorporating structural centrality and burstiness), and the authors used betweenness centrality to identify bridging nodes. VOSviewer produced cluster maps for keywords, authors, countries and organisations; for the keyword‑by‑decade analyses the inclusion threshold was at least 10 co‑occurrences per term. A complementary screening sought human randomised clinical trials and systematic reviews/meta‑analyses in the dataset. Selected full texts were read by one author (MHFN), who extracted study characteristics (title, authors, year, country, document type, intervention, follow‑up time and purpose). Methodological quality of identified systematic reviews was appraised with AMSTAR 2 (A MeaSurement Tool to Assess Systematic Reviews).

The search initially returned 1,590 references; after applying the 1993–2022 filter and removing 67 duplicates, 1,523 references remained for analysis. Publication counts by decade were 367 (1993–2002), 541 (2003–2012) and 615 (2013–2022). Document types were dominated by original articles (916; 60%) and reviews (379; 25%), with smaller numbers of book chapters, books, conference papers and other formats. English accounted for 1,484 publications (97%). Regression analyses showed an increasing publication trend in each decade. The estimated slopes were 4.05 additional publications per year in 1993–2002 (p = .004), 2.13 per year in 2003–2012 (p = .075) and 2.95 per year in 2013–2022 (p = .008). The authors report no statistically significant differences in slopes between decades. Mean publication output was lower in the first decade than in the subsequent two decades, while the difference between the latter two was not significant. Geographic coverage spanned 70 countries across five continents. Europe contributed 28 countries, the Americas 16 (North America 6, Central America 1, South America 9), Asia 15, Africa 9 and Oceania 2. The top 10 collaborating countries together produced 1,493 publications, reportedly representing 77% of publications from all countries involved; the United States had the largest scientific contribution, followed by other high‑income nations. Co‑citation and authorship networks highlighted a small group of highly influential researchers. From 4,438 distinct authors initially identified, 22 authors met the threshold of at least 10 publications (432 documents). Glick SD emerged as the most prominent node by citation size; other frequently co‑cited or network‑central authors included Maisonneuve IM, Mash DC, Popik P and Sershen H. Metrics reported for selected nodes included degree and sigma (for example, Maisonneuve degree 4.8, sigma 1.003; Popik had burst 6.94, degree 12.75, sigma 1.04), although some extracted values were incomplete in the text. Nine publications exceeded 100 citations over the three decades and centred on ibogaine’s central nervous system effects, receptor interactions and preclinical addiction models. Keyword mapping revealed temporal shifts in research focus. In 1993–2002 keywords clustered around ibogaine, cocaine, morphine and dopamine, with links to nicotine addiction and withdrawal. The 2003–2012 period retained addiction, nicotine, ibogaine and cocaine as salient terms and showed weaker mentions of 18‑MC and alcohol and depression. In 2013–2022 the network emphasised addiction with alcohol, nicotine and cocaine, and included opioids and noribogaine (ibogaine’s main active metabolite). Notably, the third decade also featured a cluster linking ibogaine research to classic psychedelics—LSD, ayahuasca, psilocybin—plus MDMA and ketamine; morphine and 18‑MC were less prominent. Organisational analysis (minimum five publications and one citation) identified 12 institutions meeting the criterion; the USA contributed five organisations and Canada two. Albany Medical College (Albany, New York, USA) was noted as the most productive institution across the whole period. The complementary screening found no human meta‑analyses and only a limited body of clinical research. Extracted records included review articles, observational studies, two Phase I studies related to mood disorders and one randomised double‑blind, placebo‑controlled study; however, that trial did not test therapeutic efficacy of noribogaine for SUDs and the authors of that trial described effects on opioid withdrawal as inconclusive. Only one systematic review with meta‑analysis was identified and it concerned preclinical (animal) studies: this meta‑analysis pooled 28 animal experiments and reported that ibogaine reduced drug self‑administration particularly in the first 24 hours after dosing, had no reliable effect on drug‑induced conditioned place preference, and was associated with motor impairment and brain cell loss in animal models. AMSTAR 2 appraisal rated that preclinical review as low in reliability. The extracted text also notes two ongoing clinical trials registered on ClinicalTrials.gov: NCT04003948 (a Phase II study in Spain recruiting 20 patients on methadone maintenance) and NCT05029401 (a two‑stage Phase I/Phase II study in the UK to determine maximum tolerated and target treatment doses).

Fernandes‑Nascimento and colleagues interpret their findings as evidence of sustained and geographically widening scholarly interest in ibogaine for SUDs across 1993–2022, with a persistent upward publication trend and concentration of productivity in developed countries, especially the USA. They highlight that a relatively small number of preclinical researchers—Glick SD, Maisonneuve IM, Mash DC, Popik P, Sershen H and others—have been especially influential in shaping the field, largely through animal studies on ibogaine and the 18‑MC congener. The authors argue that keyword dynamics reflect changing emphases: early work centred on cocaine, morphine and nicotine and on dopaminergic mechanisms, whereas the most recent decade shows increased attention to opioids, noribogaine and links to research on classic psychedelics. They note that the rise of opioid‑related keywords aligns with broader increases in opioid use and that some clinical observations suggest ibogaine may reduce cravings or provide rapid symptomatic improvement during detoxification, but emphasise that the published human evidence is limited and inconclusive. On efficacy and safety, the discussion stresses that robust human data are lacking. The only meta‑analysis identified concerned animal studies and reported both short‑term reductions in self‑administration and concerning signals for motor and neural toxicity. The investigators therefore call for Phase I and Phase II human trials and double‑blind randomised controlled trials to establish safety, tolerability and efficacy. They mention two registered clinical trials (a Phase II study in Spain and a Phase I/II study in the UK) as steps in this direction. The authors also conjecture—based on some preclinical and clinical observations—that ibogaine’s subjective, hallucinogenic effects could contribute to therapeutic benefit, perhaps via psychological mechanisms and neural plasticity, rather than by simple pharmacological suppression of withdrawal. Key limitations acknowledged by the authors include reliance on a single database (SCOPUS), which may limit representativeness; the fact that scientometric inclusion did not pre‑screen studies for methodological quality; potential publication bias favouring positive or successful studies; and that the screening to prepare a detailed evidence table was conducted by a single author. They conclude that, while scientometrics provides a broad panorama of research activity and trends, rigorous clinical research is needed to resolve questions about ibogaine’s efficacy and safety before therapeutic recommendations can be made.

Over 1993–2022 ibogaine research publications showed overall growth with fluctuations between decades, and the United States emerged as the predominant contributor. The co‑occurrence of ibogaine with research on classic psychedelics in the most recent decade suggests a shifting conceptual context for its potential therapeutic use. Importantly, no controlled double‑blind randomised clinical trials in humans were identified that establish efficacy for SUDs, and human safety data remain insufficient. The authors recommend continued scientometric monitoring and, crucially, well‑designed Phase I/II trials and double‑blind RCTs to evaluate whether ibogaine can be used safely and effectively in SUD treatment.