Abbreviations save space in a manuscript, but they cost the reader something every time one is used: a small act of memory or, if the definition wasn’t clear, a trip back up the page to find it. Style guides used across scholarly publishing — the ICMJE Recommendations, the AMA Manual of Style, and the APA Publication Manual — converge on the same basic trade: define an abbreviation once, clearly, at first use, and only introduce it at all if it earns its keep. This guide covers when to abbreviate, how to define an abbreviation correctly on first use, where the rules differ from journal to journal, and the mistakes that most often trip up authors and get flagged in copyediting or peer review.
What counts as a “research abbreviation”
In manuscript style guidance, three overlapping categories get lumped together as “abbreviations,” and it helps to keep them distinct:
- Acronyms — pronounced as a word (e.g., ELISA, PCR is technically an initialism but often treated as one, NASA).
- Initialisms — pronounced letter by letter (e.g., DNA, IRB, CI for confidence interval).
- Shortened forms/contractions — a truncated word (e.g., vs., approx., Fig.).
Style guides generally apply the same first-use-definition logic to all three, though units of measurement (SI units like kg, mL, °C) are treated as a separate, exempt category — they don’t need to be spelled out at all under AMA, APA, or ICMJE-aligned house styles.
The first-use rule
The convention shared across AMA, APA, and journals following ICMJE Recommendations is the same basic pattern: write the full term out, followed by the abbreviation in parentheses, the first time the term appears in the running text — then use the abbreviation consistently afterward.
Randomized controlled trial (RCT) designs remain the reference standard for causal inference in clinical research. This RCT enrolled 240 participants…
Two qualifications matter in practice:
- “First use” resets in each stand-alone section. The AMA Manual of Style treats the abstract, the main text, and each table and figure as separate units for this purpose — an abbreviation defined in the body text must be redefined (or defined independently) if it also appears in the abstract, a table, or a figure caption, since readers may see any of these in isolation (e.g., in a database record or a slide reused from the figure).
- A handful of very common abbreviations don’t need spelling out even on first use because they function as words in their own right — AMA style gives examples like DNA, AIDS, and UV. This list is narrow and journal-dependent; when in doubt, define it.
Where the major style guides agree — and where house style still varies
| Convention | AMA Manual of Style | ICMJE-aligned journals | APA Publication Manual |
|---|---|---|---|
| Define at first use | Yes, expanded term + “(ABBR)” | Yes; nonstandard abbreviations discouraged as confusing to readers | Yes, same parenthetical pattern |
| Abbreviations in the title | Discouraged; avoid unless essential | Explicitly recommends avoiding abbreviations in the title | Discouraged for readability and indexing |
| Abbreviations in the abstract | Must be redefined even if already defined in text | Generally discouraged or must be self-contained, since abstracts circulate independently (e.g., in PubMed) | Discouraged; APA recommends minimizing abbreviations that require the reader to hold extra information in mind |
| Minimum-use threshold | Rule of thumb: only introduce an abbreviation if it recurs at least 2–3 times | Not specified numerically, but same underlying logic (don’t abbreviate a term used once) | Not specified numerically; same general guidance to avoid one-off abbreviations |
| Units of measurement | Exempt from spell-out | Exempt (SI units) | Exempt |
Because exact wording and edge cases (plural forms, whether to italicize, how many uses justify introducing an abbreviation) vary by journal, the operative source of truth for any given submission is that journal’s own instructions for authors, not the general style manual in isolation — most journals adopt AMA, APA, or Vancouver/ICMJE style as a base and then layer house-specific exceptions on top.
Common mistakes reviewers and copyeditors flag
- Introducing an abbreviation used only once. If a term appears a single time, spell it out and skip the parenthetical — the abbreviation adds a lookup cost with no payoff.
- Starting a sentence with an abbreviation. Most style guides either prohibit this outright or discourage it strongly; rephrase rather than open a sentence with an acronym.
- Forgetting to redefine in the abstract, tables, and figures. Because abstracts and figures are frequently viewed independently of the full text (indexed separately in databases, reused in slide decks), an abbreviation defined only in the body text will be undefined for a reader who only sees the abstract.
- Inconsistent abbreviation once introduced. Switching back and forth between the spelled-out term and the abbreviation after the first-use definition (with no clear reason) reads as inconsistent and is a common copyediting mark-up.
- Overusing “alphabet soup.” A paragraph dense with unfamiliar initialisms is harder to parse even when every one was technically defined correctly — readability, not just correctness, is part of the standard.
- Using nonstandard, invented, or locally-coined abbreviations that only make sense within one lab or one paper. ICMJE guidance specifically calls out that nonstandard abbreviations can confuse readers; standard, field-recognized abbreviations are preferred over ad hoc shorthand.
- Abbreviating in the title. Titles are indexed and retrieved independently of the full article; an abbreviation there can hurt discoverability as well as readability.
A practical checklist before submission
- List every abbreviation used in the manuscript, including the abstract and each table/figure caption separately.
- For each one, confirm it appears at least 2–3 times in that section — if not, spell it out instead.
- Confirm each abbreviation is defined at its true first appearance in each independently-viewable section (abstract, body, individual tables, individual figures).
- Remove any abbreviation from the title.
- Check the target journal’s instructions for authors for a required style (AMA, APA, Vancouver) and any house-specific exceptions or an approved abbreviation list.
- Read the abbreviation-heavy sections aloud, or have a colleague outside the immediate subfield skim them — a section that’s technically compliant but hard to follow is still worth simplifying.
Frequently asked questions
Do I need to define a standard unit like mL or kg?
No. SI units and other standard units of measurement are treated as an exception across AMA, APA, and ICMJE-aligned styles and don’t need to be spelled out.
Can I use an abbreviation that’s common in my subfield but not to a general audience?
Define it anyway on first use. What counts as “standard” and exempt from definition is a short, conservative list (examples like DNA or UV under AMA style) — a field-specific term that reads as self-evident to specialists is exactly the case most likely to need spelling out for a broader journal readership, editors, and reviewers from adjacent fields.
Does ICMJE maintain an approved list of abbreviations?
No. The ICMJE Recommendations set the underlying principle — avoid nonstandard abbreviations, avoid them in titles — but individual journals, not ICMJE itself, publish their own instructions for authors and any journal-specific abbreviation conventions.
Should abbreviations be italicized or capitalized differently from the source style?
This is exactly the kind of detail that varies by journal house style rather than being fixed at the AMA/APA/ICMJE level — check the target journal’s instructions for authors rather than assuming one convention applies universally.
For related manuscript-mechanics guidance, see CASRAI’s guide to drafting, style, and revision and the Et Al. citation rules across APA, MLA, Chicago, and Vancouver/AMA comparison guide, which covers a parallel style-convergence-and-divergence question for citation practice.
