Does PubMed Central--a government-run digital archive of biomedical articles--compete with scientific society journals? A longitudinal, retrospective cohort analysis of 13,223 articles (5999 treatment, 7224 control) published in 14 society-run biomedical research journals in nutrition, experimental biology, physiology, and radiology between February 2008 and January 2011 reveals a 21.4% reduction in full-text hypertext markup language (HTML) article downloads and a 13.8% reduction in portable document format (PDF) article downloads from the journals' websites when U.S. National Institutes of Health-sponsored articles (treatment) become freely available from the PubMed Central repository. In addition, the effect of PubMed Central on reducing PDF article downloads is increasing over time, growing at a rate of 1.6% per year. There was no longitudinal effect for full-text HTML downloads. While PubMed Central may be providing complementary access to readers traditionally underserved by scientific journals, the loss of article readership from the journal website may weaken the ability of the journal to build communities of interest around research papers, impede the communication of news and events to scientific society members and journal readers, and reduce the perceived value of the journal to institutional subscribers.
reveals a 21.4% reduction in full-text hypertext markup language (HTML) article downloads and a 13.8% reduction in portable document format (PDF) article downloads from the journals' websites
large adjusted cohort supports the publisher-site difference, though nonrandom treatment limits causal attribution
the effect of PubMed Central on reducing PDF article downloads is increasing over time, growing at a rate of 1.6% per year
the age interaction supports the reported trend, though causal interpretation remains limited
There was no longitudinal effect for full-text HTML downloads.
the HTML age interaction supports no discernable trend, but complete test statistics are not reported
There is strong evidence to suggest that mandated deposit of published articles in a public repository is responsible for drawing significant numbers of readers away from journal websites
causal responsibility is not established by the nonrandomized NIH-funded versus non-NIH-funded comparison
Derived from the full evaluation — not a separate score.
Strengths
The analysis extends a preliminary single-publisher study to 13,223 articles across 14 journals in four biomedical fields and adds PDF downloads as an outcome.
↳ Introduction, paragraph 3; Materials and Methods, paragraph 1
The model identifies first-year downloads as the baseline covariate, second-year downloads as outcomes, and includes article age plus journal and nested-section effects. The log transformation and exponentiation are explicitly explained.
↳ Statistical Methods
The Discussion connects publisher-site usage displacement to scientific societies, journal editors, librarians, and institutional subscription decisions rather than invoking policy relevance only in general terms.
↳ Discussion, paragraphs 4–5
Limitations
NIH funding determines treatment status, and treatment and control articles already show different usage profiles. Baseline adjustment and broad random effects do not fully establish that PMC caused the second-year divergence.
↳ Results, Figure 1; Limitations, second paragraph; Discussion, first paragraph
The paper acknowledges nonidentical cohort profiles and the absence of a pre-deposit comparison cohort, then dismisses a differential trajectory as highly unlikely without carrying that uncertainty into the conclusion.
↳ Limitations, second paragraph; Discussion, first paragraph
Publisher-site downloads do not measure total readership, subscription cancellations, or loss of scholarly community engagement. The article recognizes that additional PMC readership and referrals back to journals cannot be estimated.
↳ Discussion, paragraphs 4–5; Limitations, third paragraph
The large multi-journal cohort and explicitly specified pretest-posttest model support a useful estimate of publisher-site download differences after PMC availability. The main methodological dispute turns on identification: first-year downloads and journal/section effects mitigate confounding, but NIH funding is nonrandom and Figure 1 shows pre-existing cohort differences. Accordingly, the reported associations remain informative while the Discussion's assertion that PMC is “responsible” for the decline warrants revision. The study has clear policy relevance, although its evidence does not directly measure total readership, subscription behavior, or downstream community effects.
Nabu’s assessment, alongside the field’s view.
Are you an author of this paper?
Sound3.1
Confidence mediumThe study meaningfully expands a preliminary single-publisher analysis to 13,223 articles across 14 journals and adds PDF downloads and a longitudinal estimate. The advance is useful but primarily an incremental expansion of scope.
“expand on those preliminary results by enlarging our study”
The pretest-posttest model uses first-year downloads, article age, and journal and section effects, with consistent platform-based measurement. NIH funding is not randomized, and the model does not establish that PMC rather than differential article trajectories caused the second-year difference.
“The usage profile of NIH-sponsored (treatment) articles is similar with, but not identical to, control articles”
The question, model, transformed estimates, and HTML/PDF results are presented in a traceable sequence. The phrases “reduces” and “is responsible” communicate stronger causal certainty than the design supports.
“is responsible for drawing significant numbers of readers away”
The paper connects its findings to prior access studies and identifies several concrete limitations and future-research boundaries. Its treatment-control comparability concern is dismissed as unlikely rather than reflected in a more qualified conclusion.
“although this is highly unlikely”
Caveats4 of 4 checks
Reported sample totals, model structure, and headline estimates are mutually consistent. The causal attribution in the Discussion extends beyond the adjusted observational comparison, although it does not invalidate the reported download associations.
The article-level download analysis does not require human-subject consent or ethics approval, and no text-visible conduct problem is established. No actual conflict, suspicious availability claim, image-integrity issue, or protocol discrepancy was identified.
Flags: 0 declared / 5 total
8 of 8 checkable references verified
13 references in manuscript 5 have no canonical index record — counted, but not index-checkable 1 reference confirmed by manual review
References not verified automatically are resolved manually within 24 hours.
No retraction or expression of concern found in Crossref (includes Retraction Watch).
Sources: Crossref (incl. Retraction Watch) ✓
Where this paper’s evidence sits on the path from initial observation to real-world use.
The analysis uses actual operational download data from deployed journal and repository systems. It does not test policy alternatives or directly measure total readership, subscription decisions, or community effects.
AI-generated, human-governed. Something look off? Contact us to request a review.
Keep reading
Other papers Nabu has evaluated blind, starting with the same field.
Baptiste Scancar, Jennifer Anne Byrne, David Causeur, et al. · 2026 · Media and communications
These findings suggest that paper mills inflate the citation metrics of supported publications and affected journals at scale.
TentativeChao Ma, Yiwei Li, Feng Guo, et al. · 2019 · Economics
papers published between October and December on average get as much as 18.5% fewer citations than those published in the other months in the year
TentativeMasashi Kawai, Koichi Ojiro, Makoto Asaoka, et al. · 2022 · Other social sciences
日本の機関リポジトリによるオープンアクセス論文の数は限られるものの，その被引用数と様々なタイプのオープンアクセス論文の被引用数の比較からは，日本の機関リポジトリによるオープンアクセスのメリットは確認できなかった．
Tentative