Below the Quantification Limit Is Not Zero: A Biomarker Data Audit
November 2, 2026 · 5 min read · Aqrab editorial team
A biomarker table lists several values as “below the lower limit of quantification” (LLOQ). The analysis file turns each one into half that limit. The new numbers look exact, but the assay did not measure them. Before interpreting a group difference or a regression coefficient, ask what was actually observed.
What “Below LLOQ” Does—and Does Not—Say
The LLOQ is the lowest concentration the assay can quantify with its stated performance. A result below it is not automatically zero, and it is not a precise measurement at LLOQ/2. The limit of detection and the limit of quantification are related but distinct assay properties; use the term the laboratory actually validated. FDA biomarker guidance asks whether an assay measures its intended analyte, how variable its measurements are, and what range provides reliable data.
For a simple hypothetical example, suppose an assay can reliably quantify concentrations from 2 units upward. One group has four reported results: <2, <2, 4, and 6. Another has <2, 3, 4, and 6. Replacing every <2 with 1 produces group means of 3 and 3.5 units. Those are means of a constructed dataset, not fully measured group means. The two unknown values in each group might differ; the report alone cannot tell us by how much.
The practical takeaway
Keep the below-limit flag and the assay threshold in the dataset. Report how many values fall below the limit in each study group before presenting a modeled comparison.
Why a Convenient Replacement Needs a Sensitivity Check
Replacing all below-limit results with one number assumes information the assay did not provide. The direction and size of any resulting bias depend on the distribution, the proportion below the limit, the estimand, and the analysis model. Methodological studies have found that simple fixed substitution can perform differently from approaches that account for censoring; no single alternative is guaranteed to fit every dataset.
A defensible analysis should state its target quantity, record each assay’s validated range, prespecify how below-limit values enter the model, and test whether reasonable alternative handling changes the conclusion. If the conclusion depends heavily on the chosen replacement value, show that fragility rather than hiding it behind extra decimal places. If limits differ by batch, site, or assay version, report those limits alongside the below-limit counts.
Four Questions for a Methods Review
- Assay: What were the validated LLOQ, units, specimen type, and analytical range?
- Accounting: How many results were below the limit in each group and at each time point?
- Analysis: Were below-limit results treated as censored observations, replaced with a fixed value, or excluded—and why?
- Robustness: Does the scientific conclusion survive plausible alternative assumptions about the unmeasured values?
Where Aqrab Fits
Use Aqrab to review whether a biomarker study separates measured concentrations from below-limit observations and whether its analysis matches the assay’s validated range. A methods critique can reveal a hidden substitution assumption; it cannot recover the unmeasured concentrations or establish a biological effect from the flag alone.
Source Anchors
The FDA’s 2026 biomarker assay-validation guidance describes fit-for-purpose validation and the need to establish a reliable measurement range, including the LLOQ. A methodological comparison of substitution and likelihood methods illustrates why fixed replacements should not be treated as universally accurate. The numerical example here is hypothetical and is not taken from either source.
Keep reading
Don't stop at one method.
Good methods judgment comes from contrast. Read the neighboring guides, see where the assumptions diverge, and avoid treating every observational problem like it needs the same hammer.
Batch Effects: When a Biomarker Follows the Plate, Not the Patient
A practical group-by-batch audit shows why a biomarker difference cannot be separated from technical conditions when cases and controls occupy different batches.
Only the Disagreements Were Checked: The Discrepant-Resolution Trap
When a third test checks only disagreements, agreement is not proof that both tests were right. Use this three-question audit before accepting a revised diagnostic-accuracy table.
The Result Arrived After the Decision: Audit Diagnostic Turnaround
A practical timeline shows why diagnostic accuracy alone cannot establish that a result can reach the clinical decision it is meant to support.