Skip to main content
Tessa NewsLink
Paste a health news link, or browse

Source study found

Story checked

We May Have Been Wrong About Fasting Before Blood Tests, Study Reveals : ScienceAlert (opens in a new tab)

sciencealert.com · 2026-10-04

Short answerEvidenceSource

Short answer

Mixed

Mixed.

One key claim is not backed by the study. 2 other points were not covered by the paper.

  • 3 supported
  • 1 not supported
  • 2 not covered

Checked against the study summary. The full text wasn't available, so some details couldn't be settled either way.

Share this check

Follow the evidence trail
1
2

NewsLink checks it

Mixed

One claim isn't supported by the study. Three of six check out. Two claims the study doesn't address.

  • 3 supported
  • 1 not supported
  • 2 not covered
Open claim evidence
3
Source paper

Source layer

The 2 papers the story cites

Source study separated from background citations.

The research anchor for the report.

Then inspect each claim

Evidence layer

Claim by claim

Each claim gets a verdict. Expand it to see the evidence directly below.

6 claims in this story

Showing all 6 claimsChoose a verdict to focus the list.

Claim 1 of 6Not supportedThe researchers found that fasting duration was linked to APS-exceeding differences for only 10 of 121 analytes, and that fasting beyond 12 hours was not associated with meaningful additional differences for any test examined.View evidenceHide evidence

As stated10 of 121 analytes

Why this verdict

The '10' confirmed analytes and the lack of APS-exceeding differences between 8–12 h and >12 h fasting are supported, but the claim is not supported as stated because the supplied paper profile says the study covered 372 test items and that 10 of 27 initially APS-exceeding analytes were confirmed in paired analyses, not '10 of 121 analytes.' That numerical framing conflicts with the profile.

1

Study evidence

Large outpatient LIS dataset analyzed: 9,755,547 test results from 101,148 patients across 372 analytes.

“Retrospective single-center cross-sectional study using laboratory information system data collected between February 2021 and December 2023”
2

Study evidence

Only 10 of 27 analytes that exceeded APS in between-group comparisons were confirmed as APS-exceeding in paired within-person analyses among repeatedly tested patients, suggesting confounding in the between-group results.10 of 27 analytes

“Paired within-person analyses among repeatedly tested patients were performed to minimize between-patient confounding.”
Claim 2 of 6Not coveredThe study was a retrospective cross-sectional analysis of lab data from a single university hospital, led by Sunghwan Shin, using more than 9.7 million test results from more than 101,000 adult outpatients with documented fasting durations.View evidenceHide evidence

As stated9.7 million test results; more than 101,000 patients

Why this verdict

The study design, single-center outpatient setting, and sample size are supported by the abstract profile: 9,755,547 test results from 101,148 adult outpatients in a retrospective single-center cross-sectional LIS study. However, the supplied scientific profile does not verify the author-lead attribution to Sunghwan Shin, so the full claim is not completely verifiable at the supplied abstract-profile depth.

1

Study evidence

Large outpatient LIS dataset analyzed: 9,755,547 test results from 101,148 patients across 372 analytes.

“Retrospective single-center cross-sectional study using laboratory information system data collected between February 2021 and December 2023”
Claim 3 of 6Not coveredAn accompanying commentary says the study provides strong support for accumulating evidence that fasting blood sampling is not necessary.View evidenceHide evidence

Why this verdict

The supplied PaperScientificProfile covers the study abstract and does not include the accompanying commentary. The paper's own findings are broadly consistent with analyte-specific relaxation of blanket fasting rules, but the quoted commentary judgment that the study provides 'strong support' cannot be verified from the supplied profile.

Claim 4 of 6SupportedA new study suggests that fasting before many routine blood tests may be unnecessary.View evidenceHide evidence

Why this verdict

The paper profile supports the general, hedged headline claim: only a small subset of routine analytes showed APS-exceeding fasting-associated differences after paired within-person confirmation, and the discussion supports analyte-specific rather than uniform fasting recommendations. The headline is broad, but the word 'may' keeps it aligned with the observational evidence and the story's body caveats.

1

Study evidence

Large outpatient LIS dataset analyzed: 9,755,547 test results from 101,148 patients across 372 analytes.

“Retrospective single-center cross-sectional study using laboratory information system data collected between February 2021 and December 2023”
2

Study evidence

Only 10 of 27 analytes that exceeded APS in between-group comparisons were confirmed as APS-exceeding in paired within-person analyses among repeatedly tested patients, suggesting confounding in the between-group results.10 of 27 analytes

“Paired within-person analyses among repeatedly tested patients were performed to minimize between-patient confounding.”
Claim 5 of 6SupportedThe findings suggest many patients could eat and drink in the hours before their tests without throwing off the results, except for a few notable exceptions.View evidenceHide evidence

Why this verdict

The profile supports the associational, hedged idea that many routine tests were minimally affected, with exceptions. Paired analyses confirmed APS-exceeding differences for only 10 analytes, and specific exceptions included glucose, triglycerides, γ-glutamyl transferase, and lipase. The claim is framed as a suggestion rather than direct patient instruction.

1

Study evidence

Large outpatient LIS dataset analyzed: 9,755,547 test results from 101,148 patients across 372 analytes.

“Retrospective single-center cross-sectional study using laboratory information system data collected between February 2021 and December 2023”
2

Study evidence

Only 10 of 27 analytes that exceeded APS in between-group comparisons were confirmed as APS-exceeding in paired within-person analyses among repeatedly tested patients, suggesting confounding in the between-group results.10 of 27 analytes

“Paired within-person analyses among repeatedly tested patients were performed to minimize between-patient confounding.”
Claim 6 of 6SupportedFasting appeared to matter mainly for glucose, triglycerides, gamma-glutamyl transferase, and lipase, which were higher in patients who fasted for less than eight hours.View evidenceHide evidence

As stated10 analytes; 4 named analytes

Why this verdict

The profile reports higher concentrations with <8 h fasting for glucose, triglycerides, γ-glutamyl transferase, and lipase, with the stated pattern fitting the paper's analyte-specific conclusion. The abstract-level profile notes some uncertainty about the full list of all 10 confirmed analytes, but these four named analytes are supported.

1

Study evidence

Large outpatient LIS dataset analyzed: 9,755,547 test results from 101,148 patients across 372 analytes.

“Retrospective single-center cross-sectional study using laboratory information system data collected between February 2021 and December 2023”
2

Study evidence

Only 10 of 27 analytes that exceeded APS in between-group comparisons were confirmed as APS-exceeding in paired within-person analyses among repeatedly tested patients, suggesting confounding in the between-group results.10 of 27 analytes

“Paired within-person analyses among repeatedly tested patients were performed to minimize between-patient confounding.”
Then look for missing context

Context layer

What the story left out

Important study details the story did not include.

  • The paper profile says the broad panel covered 372 test items.

    The story claim instead states '121 analytes,' which is inconsistent with the supplied profile's 372 test items.

    From Retrospective cross-sectional using laboratory information system (LIS) data

  • Between-group comparisons initially identified 27 analytes exceeding APS, but paired within-person analyses confirmed only 10, suggesting between-patient confounding.

    The story captures the headline number of 10 affected analytes, but it does not accurately preserve the 27-to-10 sequence or the implication that many initial between-person findings may have reflected confounding.

    From Retrospective cross-sectional using laboratory information system (LIS) data; Paired within-person repeated-measures ana

  • Within the <8 h subgroup, glucose and lactate decreased with increasing fasting duration, while lipids showed no significant linear postprandial trend.

    The story discusses named analyte exceptions but does not report the 30-minute interval postprandial trend analysis, the lactate finding, or the lipid no-linear-trend result.

    From subgroup time-binned trend analysis

  • Fasting duration depended on recorded time since last meal, so measurement precision and accuracy depended on documentation.

    The story does not mention possible misclassification or documentation error in the fasting-duration exposure.

    From Retrospective cross-sectional using laboratory information system (LIS) data

  • Multiple-analyte screening across hundreds of test items raises potential false-positive concerns, and the abstract profile does not report multiplicity-adjustment details.

    The story does not mention multiple-comparison or false-positive concerns, which are material because the study screened many analytes.

    From Retrospective cross-sectional using laboratory information system (LIS) data

7 things the story did carry across
  • The paper studied associations between fasting-duration categories (<8 h, 8–12 h, >12 h) and routine laboratory analyte results, comparing observed differences with biological-variation-based analytical performance specifications.
  • The abstract profile reports 9,755,547 test results from 101,148 adult outpatients in a retrospective single-center outpatient LIS dataset.
  • Glucose, triglycerides, γ-glutamyl transferase, and lipase were higher with <8 h fasting.
  • No analyte exceeded APS in comparisons between the 8–12 h and >12 h fasting groups.
  • The retrospective cross-sectional observational design limits causal inference.
  • Single-center outpatient central-phlebotomy data may limit generalizability to other settings, hospitals, or inpatient populations.
  • The paper's discussion supports considering analyte-specific fasting recommendations rather than uniform fasting requirements.
Then read the study layer

Study layer

Study at a glance

Scan the study first. Expand only the parts you want to inspect.

Pieces of work

3

Evidence read

study summary

Lead result

secondary data

1Lead resultsecondary dataQuantify associations between fasting duration categories (<8 h, 8–12 h, >12 h) and routine laboratory analyte results in adult outpatients, and judge whether observed differences exceed analytical performance specifications (APS).Retrospective cross-sectional using laboratory information system (LIS) dataExpand

In plain English

Retrospective cross-sectional analysis of 9,755,547 routine laboratory test results from 101,148 adult outpatients (Feb 2021–Dec 2023) evaluating associations between categorized fasting duration (<8 h, 8–12 h, >12 h) and analyte concentrations using linear mixed-effects models with patient-level random intercepts and adjustment for age, sex, and department; observed between-group differences were compared with biological-variation–based total allowable error (APS). Between-group comparisons flagged 27 analytes as exceeding APS, but paired within-person analyses confirmed only 10 analytes. Reported higher concentrations with <8 h fasting included glucose (10.8%), triglycerides (20.1%), γ-glutamyl transferase (14.8%), and lipase (36.4%). No analyte exceeded APS between the 8–12 h and >12 h groups. Within the <8 h subgroup, glucose and lactate decreased with increasing fasting duration while lipids showed no significant linear trend.

Key findings

  • Large outpatient LIS dataset analyzed: 9,755,547 test results from 101,148 patients across 372 analytes.
  • Between-group comparisons across fasting categories identified 27 analytes with differences that exceeded APS.27 analytes (exceeded APS)
“Retrospective single-center cross-sectional study using laboratory information system data collected between February 2021 and December 2023”
What this piece can’t prove
  • Retrospective cross-sectional design limits causal inference.
  • Single-center outpatient central phlebotomy population may limit generalizability to other settings or inpatient populations.
  • Fasting duration derived from recorded time since last meal; measurement precision and accuracy depend on how that time was documented.
  • Multiple-analyte screening across 372 items raises the potential for false-positive findings from multiple comparisons; abstract does not report multiplicity adjustment details.

1 further detail could not be confirmed from the summary.

2secondary dataUse paired within-person analyses among repeatedly tested patients to assess whether fasting-related differences persist after minimizing between-patient confounding, including APS comparison.Paired within-person repeated-measures analysis (restricted to repeat-tested patients)Expand

In plain English

Within-person (paired) analyses restricted to repeatedly tested outpatients were used to assess whether fasting-related differences in routine laboratory tests persisted after minimizing between-patient confounding. Of 27 analytes that exceeded the analytical performance specification (APS) in between-group comparisons, only 10 remained APS-exceeding in paired within-person analyses, consistent with confounding explaining many between-person differences. Specific analytes with higher values after <8 hours fasting included glucose, triglycerides, γ-glutamyl transferase, and lipase. No analyte exceeded APS when comparing 8–12 hours versus >12 hours fasting in the paired analyses.

Key findings

  • Only 10 of 27 analytes that exceeded APS in between-group comparisons were confirmed as APS-exceeding in paired within-person analyses among repeatedly tested patients, suggesting confounding in the between-group results.10 of 27 analytes
  • Glucose values were higher with <8 hours fasting in paired within-person analyses.≈10.8% higher with <8 hours fasting (as reported in abstract)
“Paired within-person analyses among repeatedly tested patients were performed to minimize between-patient confounding.”
What this piece can’t prove
  • Single-center, retrospective outpatient dataset may limit generalizability (explicit in abstract).
  • Paired analysis was restricted to repeatedly tested patients, which may introduce selection bias and affect applicability to all outpatients.

2 further details could not be confirmed from the summary.

3secondary dataCharacterize short-term postprandial time trends within the <8 h fasting subgroup at 30-minute intervals.subgroup time-binned trend analysisExpand

In plain English

In outpatients who reported <8 hours fasting, postprandial time-since-meal trends assessed in 30-minute bins showed significant decreases in glucose and lactate with increasing time since the last meal, while lipid concentrations exhibited no significant linear trend. These analyses used the study's laboratory information system data; the abstract does not report subgroup effect sizes or full modeling details for the 30-minute trend tests.

Key findings

  • Within the <8 h fasting subgroup, glucose decreased significantly with increasing time since last meal (assessed at 30-minute intervals).
  • Within the <8 h fasting subgroup, lactate decreased significantly with increasing time since last meal (assessed at 30-minute intervals).
“Postprandial trends within the subgroup fasting less than 8 hours were assessed at 30-minute intervals.”
What this piece can’t prove
  • Abstract does not report effect sizes, confidence intervals, or exact model specification for the 30-minute interval trend analyses.
  • Single-center, retrospective observational data from outpatient phlebotomy may limit generalizability.
  • Subgroup analyses (restricted to <8 h fasting) may be susceptible to residual confounding or selection effects; abstract does not detail covariate adjustment for these trend tests.
  • The abstract summarizes aggregate results across analytes; specific details (e.g., which lipid fractions) for the no-trend finding are not provided.
Finally, the search trail

Method layer

NewsLink found the paper. Tessa takes you deeper.

NewsLink checks the story. Tessa is where you inspect the paper, authors, evidence, and research context.

Papers considered

The selected paper, plus nearby candidates.

PubMed, Europe PMC, Crossref · 16 candidate papers

And 10 more candidates considered.