Match your attribution to what you actually read
AMA Secondary Citations When You Have Not Read the Original
Trace a review to an original exercise study, check what changes after reading the paper, and handle an unavailable original without implying direct access.
Before formatting an AMA secondary citation, check what you actually read. You may have consulted the original study or learned about it through another author’s account. We traced a Jackson and colleagues review to an Erickson and colleagues trial, then checked the original Methods and Results. That comparison shows why a title, a DOI, and a review’s summary each leave different work unfinished. ICMJE recommends direct references to original research whenever possible. When access remains unavailable, your wording needs to identify the account you actually used.
Identify the source you actually read
A review can support a sentence about its authors’ interpretation. A claim about a particular experiment requires a closer check of that experiment. Before copying anything from the review’s bibliography, save the passage you are using and identify which part is the review authors’ assessment. Record any original study they cite beside it.
Keep a short reading note for each item. Mark whether you inspected the full text, the abstract alone, or only a bibliographic record. Add the section or page supporting your proposed sentence. A downloaded PDF you have not examined does not count as having read its contents.
The same decision applies when a clinical database topic points to a trial. If your sentence reports the topic authors’ recommendation, identify that topic. If you need the trial’s methods or outcome, follow its reference and inspect the relevant passage.
Follow a real review to its original study
Our starting source is Jackson and colleagues’ 2016 review, Promoting brain health through exercise and diet in older adults. On printed page 4487, its physical activity discussion describes a year of walking compared with stretching and toning and cites Erickson and colleagues, 2011. The reference list supplies the original article’s title, journal, volume, and pages.
That entry leads to Exercise training increases size of hippocampus and improves memory. Searching the title in PubMed identifies PMID 21282661 and a free PMC full-text link. Match the authors and publication details before using that link. Similar titles and shared authors can otherwise send a reference check toward a different paper.
| Checkpoint | Source or locator | What you can establish |
|---|---|---|
| Review consulted | Jackson et al, J Physiol. 2016;594(16):4485-4498 | The review’s account and its attribution |
| Passage and reference list | Printed page 4487; Erickson et al, 2011 | Which original paper the review points to |
| Original record | PMID 21282661; DOI 10.1073/pnas.1015950108 | The paper’s bibliographic identity |
| Original full text | PMC3041121, Methods and Results | The details behind the proposed statement |
Try to locate the original before giving up
PubMed’s help distinguishes citation records and abstracts from full-text articles. On a record, inspect Full Text Links for the publisher or PMC. A paywall at one publisher link does not establish that every legitimate copy is unavailable. PubMed also describes library access and interlibrary loan options.
Search the exact title first, then use author, journal, and year details if the result is uncertain. Compare the DOI with the candidate article. Keep a note of the version you open, particularly if it is an author manuscript. Read enough of the relevant sections to check the claim, including any referenced table or supplement.
If you have only an abstract, limit your assertion to information it contains and disclose that reading limitation when relevant. Do not fill in missing eligibility criteria or numerical details from the review and attribute them to your own inspection of the original.
Separate a review’s interpretation from an original finding
The original paper’s Methods distinguishes 179 enrolled participants from 120 with complete MRI data included in the analyses. Its Results also says both groups improved on the spatial memory task, without a significant advantage for aerobic exercise over the stretching control on that comparison. Those details matter when choosing a denominator or describing a between-group result.
A sentence saying that exercise improved memory more than the control would need support beyond the title’s wording. Check the comparison your sentence actually makes. Within-group change, a correlation, and a difference between treatment groups answer different questions. Preserve that distinction when shortening a scientific result.
Once you have inspected the original passage, cite the original for that passage. Keep the review as a separate source if you also discuss its synthesis or interpretation. There is no need to remove a review that still supports a distinct part of your argument.
Original-study reference after checking the relevant full text
Erickson KI, Voss MW, Prakash RS, et al. Exercise training increases size of hippocampus and improves memory. Proc Natl Acad Sci U S A. 2011;108(7):3017-3022. doi:10.1073/pnas.1015950108
This entry identifies the original paper. Its presence alone cannot show which parts you read.
When the original remains unavailable
The original in this worked chain is openly accessible. To illustrate the unavailable-source decision, suppose instead that you had read only the Jackson review. The sentence below deliberately reports the review’s description and states that the original was not consulted. It is an attribution example, not a universal AMA secondary-reference template.
Illustrative wording when only the review was read
Jackson and colleagues describe an exercise trial by Erickson and colleagues comparing walking with stretching and toning; the original report was not consulted.
Attach the reference number for the review to this account. Check the destination’s instructions for any additional indirect-source notation.
Check attribution against the target policy
For an AMA cited by citation, check the journal or course instructions for indirect sources. The public guidance checked here supports seeking the original and verifying references, but does not establish one current, universal AMA format for every unread original. Do not present an APA-style “as cited in” construction as an official AMA rule without checking the required policy.
If you need to cite a source cited in another source in AMA, identify the account you consulted and explain the reading limit. Follow any required local notation. Ask the editor or instructor about unresolved entry details before adding an unread paper as though you had verified its findings.
If a precise claim depends on information you cannot check, narrow the sentence to the review’s supported account or leave that claim out. Enter the verified source into the generator and compare the finished reference with its record. Formatting software cannot certify that you read the paper.