Table 1 Factors to be considered for the validation and application of biomarkers suggested in previous literature

1. Plausibility

• Biomarkers should be specific to the food (having the ability to distinguish the food or food component of interest from other foods or food components).

• There should be a food chemistry or experimentally based explanation for why the food intake should increase the biomarker, e.g., the biomarker should be a metabolite of a food component.

[71,72,73,74,75,76,77,78,79,80,81]

2. Dose-response

• Evaluation of the dose-response relationship should be performed to assess the suitability of the biomarker over a range of intakes.

• Limit of detection should be evaluated to provide the information about how responsive (sensitive) the biomarker is.

• Baseline habitual level needs to be established.

• Bioavailability of (the precursor of) the biomarker should be evaluated to provide the information about its sensitivity to intake.

• Detailed information about saturation effects of the biomarker should be known.

[9, 10, 12, 13, 71, 73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88]

3. Time-response

• The half-life of the biomarker should be evaluated to specify the degree to which a biomarker reflects exposure, e.g., days, weeks, months or years.

• Kinetics (comprises “formation, distribution, metabolism and excretion”) should be known to make choices, such as on sampling time and matrices.

• Timing of measurement in relation to bioavailability and bioefficacy must be considered.

• Temporal relation of the biomarker with dietary intake should be considered to provide information for choosing types of specimen.

• Repeated measures of the biomarker over time should be evaluated to provide insight into the reproducibility of biomarker concentrations, and thus, the likelihood that the biomarker is a stable estimate of long-term intake.

[9, 10, 12, 71, 74, 76, 77, 79, 82, 85,86,87, 89]

4. Robustness

• Suitability of the biomarker in a free-living population should be investigated using a (controlled) habitual diet to provide information such as its interactions with other foods and its applicability to a certain group of population.

• The biomarker should be validated in a controlled dietary intervention studies as well as in cross-sectional studies.

• Validation of the biomarker in different subjects and study settings is needed.

• Information such as interactions with other food components and influence of food matrix should be excluded or known to be manageable.

[12, 76, 80, 85]

5. Reliability

• Comparison of the biomarker and a gold standard or reference method that provides a good measure of the true exposure is necessary.

• Biomarkers identified using samples from cohort studies should ideally be combined with intervention studies to demonstrate their direct relationships with intake.

• Comparison between the biomarker and an appropriate dietary assessment method should be performed.

• A biomarker should be confirmed in accordance with other biomarkers for the same food or foods.

• Validation of a biomarker can be attempted by measuring it in animals fed different nutrient intake under tightly controlled conditions.

[9, 11, 13, 71, 72, 74, 75, 77, 84, 87, 90]

6. Stability

• Suitable protocols for sample collection, processing, and storage are needed to keep the sample quality for several years.

• Trials should be carried out to determine whether analytes undergo decomposition during storage.

[76]

7. Analytical performance

• Precision, accuracy, and detection limits of the method should be evaluated.

• Comparison against validated methodology or references or references materials is needed.

• The calculation of inter- and intra-batch variation should be performed.

• Statistical quality control procedures (coefficient of variance, standard deviation and inaccuracy limits for data) should be established.

[10,11,12,13, 71,72,73,74, 76, 82, 84, 91]

8. Reproducibility

• There is the need to develop and use accurate and validated analytical methods to adequately compare the data obtained in different laboratories.

[12, 71, 78]