Validation of a Food Frequency Questionnaire: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Validation of a Food Frequency Questionnaire (NCT03138330), a registered phase not stated study. The primary outcome is Validated Food Frequency Questionnaire.

2. Clinical Question

The study’s stated purpose was to collect nutritional information from young children aged 15-36 months and validate a newly developed food frequency questionnaire. Although the dietary information concerns toddlers, the people reported as completing the primary assessments were their parents. The registry record identifies energy, carbohydrates, fat, protein, fibre, sugars, calcium and iron as the dietary measures compared.

The primary outcome is framed as a Pearson correlation between estimates from the two methods. This asks whether participants with relatively higher estimates from one method also tend to have relatively higher estimates from the other. It does not, by itself, answer whether the methods give similar numerical values for each participant. The numerical correlation results are not reported in the registry record, so the central validation question cannot be answered from the posted results.

3. Trial Design

This was an observational study, not a randomized clinical trial of a treatment. The registry lists an intervention entry stating that the study was observational and not an intervention. Its stated focus was validation of dietary patterns and inflammatory salivary biomarkers among healthy toddlers of 15-36 months old in Singapore. The listed condition is Nutrition.

The record reports an enrollment of 95 and a status of completed. The study start is listed as 2015-12, with primary completion in 2016-11. The sponsor is Danone Asia Pacific Holdings Pte, Ltd. These registry entries describe the study’s administrative record; they do not provide further information about how participants were recruited, how dietary assessments were scheduled, or how the study was conducted beyond the outcome descriptions.

For the primary outcome, 91 parents completed the sqFFQ and a 2-day weighed food record, with the food record designated as the reference method. Intake estimates for energy and the listed nutrients were compared using Pearson correlations. The primary outcome timeframe is described as up to 1 month. The record also lists salivary inflammatory biomarkers as a secondary outcome, but does not give results for that outcome.

There is no treatment group and control group comparison to interpret. The registry’s safety entry uses the label FFQ Group, but that label does not make this an intervention study or establish a randomized allocation. Allocation is not stated, and the record explicitly cautions that the listed arms are not a like-for-like control comparison. Details about the protocol, recruitment procedures, and any further study design features are not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Ninety-one parents of Singaporean toddlers who completed the sqFFQ and a 2-day weighed food record. Here pearson correlation between the two methods (FFQ and food diary) is presented.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Validated Food Frequency Questionnaire, measured over up to 1 month.

The primary endpoint is the Pearson correlation between dietary intake estimates from the sqFFQ and those from a 2-day weighed food record. The record reports that 91 parents completed both methods. Energy, carbohydrates, fat, protein, fibre, sugars, calcium and iron were assessed using each method. The outcome timeframe is listed as up to 1 month.

As a general statistical explanation, Pearson correlation describes the direction and strength of a linear association between paired measurements. A positive correlation means that higher values from one method tend to accompany higher values from the other. Correlation is useful in questionnaire validation because it can indicate whether the questionnaire ranks people similarly to a reference method across the measured dietary components.

Correlation is not the same as agreement. Two methods can be strongly associated while still differing systematically in the values they produce. A correlation alone also does not establish that the questionnaire measures dietary intake without error, nor that it is suitable for every population or purpose. Assessing agreement or systematic bias would require suitable analyses and results, which are not reported in the registry record.

The registry record names the measure and comparator but does not post the coefficient for any nutrient, confidence intervals, or a statement of the results. Consequently, it is possible to describe what the endpoint was intended to measure, but not to quantify the observed association or its precision.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies Pearson correlation as the primary statistical measure. The paired observations are dietary estimates from the sqFFQ and the 2-day weighed food record for the same reported participants. Pairing matters because the question concerns whether the two methods produce related estimates within people, rather than whether separate groups have different averages. The record states that 91 parents completed both methods.

What does Pearson correlation describe?

As a general statistical explanation, Pearson correlation summarizes the direction and strength of a linear relationship between two quantitative measurements. If participants with larger estimates from the food record also tend to have larger estimates from the questionnaire, the association is positive. If high estimates on one method tend to pair with low estimates on the other, the association is negative. If the relationship is not well described by a straight-line pattern, a single Pearson coefficient may not fully represent the association.

Correlation concerns co-movement, not equality. For example, a method could yield values that are consistently higher than those from a reference, yet preserve a similar ranking across participants. That pattern could result in an association without close numerical agreement. Conversely, assessment of agreement asks how closely the paired values match, and whether there is systematic bias or variation in the differences. The registry record does not report such agreement analyses.

Why use a reference method?

A validation comparison needs a method against which the questionnaire can be evaluated. The registry names a 2-day weighed food record as the reference method and describes comparing intake estimates from that record with estimates from the sqFFQ. This defines the direction of the comparison in the study description, but does not establish that the reference method is error-free. As a general point, every dietary measurement approach can have limitations, and the interpretation of a validation result depends on the properties of both methods and the population in which they are used.

The listed dietary components are energy, carbohydrates, fat, protein, fibre, sugars, calcium and iron. A correlation for one component need not match the correlation for another. Each component can have different measurement characteristics, and results should therefore be read separately rather than treated as a single overall validation result. The registry does not post individual coefficients, so no component-specific comparison can be made.

What would uncertainty add?

As a general statistical explanation, a confidence interval expresses uncertainty around an estimated quantity under the assumptions of the analysis. For a correlation, it would help readers judge how precisely the association was estimated in the analyzed sample. The registry record does not report confidence intervals. It also does not report p-values for the correlations. Without these values, readers cannot assess the precision or statistical evidence associated with the observed coefficients. Even if such values were available, they would complement rather than replace consideration of the magnitude and practical meaning of the association.

The record does not describe checks of the assumptions relevant to Pearson correlation, handling of outliers, transformations, or any alternative correlation measure. Such details are not reported in the registry record and should not be inferred from the named endpoint alone. The numerical results and participant-level data are also not posted in the record, limiting independent assessment of the analysis.

How should planning calculations be read?

Any sample-size calculations associated with this record are planning illustrations based on fractions of an observed effect, not observed power and not evidence about the study’s outcome. The stated assumptions for those illustrations are a two-sided significance level of 0.05, equal allocation between two hypothetical arms, and a normal approximation. Those assumptions describe a planning calculation and do not turn this observational method-comparison study into a randomized two-arm trial. The record does not report that these assumptions were part of the original study design or its sample-size justification.

In general, sample-size illustrations depend on their assumptions about the effect to detect and the statistical procedure. They are not a substitute for reported results, and they do not show what the study found. No numerical planning estimates are included in the registry record. The study’s reported enrollment and the number completing the primary assessments should be kept distinct from hypothetical sample sizes in a planning exercise.

Overall, the methods description establishes what was compared and names Pearson correlation as the primary measure. It does not provide the numerical outputs or the analysis details needed to reproduce or fully evaluate the validation results. The distinction between association and agreement is central when interpreting this endpoint.

10. Multiplicity, Missing Data and Analysis Populations

The primary outcome description includes correlations for energy, carbohydrates, fat, protein, fibre, sugars, calcium and iron. The registry record also lists salivary inflammatory biomarkers as a secondary outcome. Considering multiple outcomes raises a general statistical issue: the more results examined, the more important it is to know which analyses were prespecified and how the interpretation accounts for the number of comparisons. The registry record does not report a multiplicity strategy, hierarchy among outcomes, or adjustment for multiple testing.

The record lists two outcome measures and zero analyses, but does not post outcome-level numerical analysis results. It does not report whether analyses were conducted for every listed dietary component, whether any results were omitted, or what statistical methods were used for the secondary biomarker outcome. Those details are not reported in the registry record.

Regarding missing data, 95 participants were enrolled and the primary outcome description says 91 parents completed both dietary methods. The record does not explain why the counts differ, whether any dietary observations were incomplete, or how missing observations were handled. It also does not describe an analysis population beyond the 91 parents identified as completing both methods. No imputation approach, sensitivity analysis, or comparison of participants with and without complete assessments is reported.

These omissions matter because the interpretation of a validation result depends on which participants contributed to each comparison and how incomplete measurements were treated. The registry record supports reporting the stated counts, but not assumptions about why they differ or about the effect of missing data on the results.

11. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

The registry safety entry lists an FFQ Group with 91 participants at risk for serious adverse events and 0 serious adverse events. It also lists 91 at risk for other adverse events and 0 other adverse events, with 0 deaths. No serious adverse event terms are listed.

These figures describe what the registry safety record reports for the listed group. They should not be interpreted as a comparison between randomized treatment and control groups: the study is identified as observational and not an intervention, and the record cautions that the arms are not a like-for-like control comparison. The FFQ Group label does not establish that participants received an intervention.

As a general statistical point, a recorded zero means no event was entered in the relevant category for the reported group. It does not prove that an event was impossible, that every possible event was captured, or that the study can rule out uncommon harms. The registry record does not report further detail about adverse-event definitions, monitoring procedures, or the period over which events were collected. The safety figures should therefore be read as registry counts, not as evidence about the safety of a treatment.

12. What the Result Does, and Does Not, Mean

The record establishes that the study compared a newly developed sqFFQ with a 2-day weighed food record for dietary intake among parents of Singaporean toddlers. It identifies Pearson correlation as the primary measure and lists the dietary components considered. It also reports that 91 parents completed both methods, with total enrollment of 95.

What the record does not establish is the strength of the association between the methods. The Pearson correlation coefficients are not reported in the registry record, nor are confidence intervals or p-values. As a result, the posted results do not show how well the questionnaire tracked the food-record estimates, how precise any estimate was, or which dietary components had stronger or weaker associations. The absence of numerical results is not evidence of either good or poor performance.

Even a reported correlation would answer a limited question. As a general statistical point, correlation describes association and does not establish agreement, absence of bias, or the accuracy of an individual’s reported intake. The registry record does not report agreement analyses, and no conclusion about numerical interchangeability of the methods can be drawn.

The study is observational and does not estimate the effect of a treatment. Its stated purpose is questionnaire validation, alongside collection of dietary information and assessment of salivary inflammatory biomarkers. The biomarker outcome is listed, but its findings are not reported in the registry record. No conclusion about a relationship between dietary patterns and inflammatory biomarkers can be drawn from the information posted.

The most defensible conclusion is that the registry describes the comparison and the intended statistical measure but does not provide the results needed to judge validation performance. The record cannot support a claim that the questionnaire is effective, ineffective, accurate, inaccurate, or interchangeable with the reference method. Those judgments require reported results and, for some questions, analyses beyond correlation.

13. Why This Trial Matters Statistically

Questionnaire validation is a measurement problem: before an instrument is used to characterize dietary intake, researchers need evidence about how its measurements relate to an appropriate comparator. This study is statistically relevant because it identifies a concrete comparison between a food frequency questionnaire and a 2-day weighed food record in parents of Singaporean toddlers. It also specifies Pearson correlation as the primary measure and names the dietary components being compared.

The example highlights why naming a statistical method is not enough to evaluate a study’s findings. A correlation is a measure of association, but it is not a complete assessment of agreement or measurement error. Readers need the numerical results and their uncertainty, as well as clarity about the participant set and analysis methods, to understand what the comparison shows. The registry record gives some of the design and completion information but does not report the correlations or their precision.

It also illustrates the distinction between enrollment and endpoint completion. The record reports enrollment of 95 and states that 91 parents completed both primary dietary assessments. It does not explain the difference or report how incomplete data were handled. Those details are important for assessing what population the estimates represent.

Finally, a registry record can define an outcome without posting its results. Here, the secondary biomarker outcome is named, but no findings are reported. Keeping the distinction between study aims, methods, and numerical evidence prevents an intended validation exercise from being mistaken for demonstrated validation. The case is a useful reminder that statistical interpretation depends on results and design details, not only on the endpoint label.

14. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The largest limitation for interpreting this study from its registry record is the absence of numerical primary outcome results. Pearson correlations are named as the outcome, but coefficients for energy and the listed nutrients are not reported. Confidence intervals, p-values, and any other measures of uncertainty are also not reported in the registry record. Without those values, the central validation result cannot be assessed.

The registry gives limited information about the analysis. It does not report checks of assumptions, handling of outliers, any transformations, or whether agreement was assessed in addition to correlation. Since correlation does not establish agreement, the lack of reported agreement analyses limits conclusions about whether the questionnaire and food record gave similar values. The record also does not describe how the food record was processed or how the questionnaire was scored beyond the outcome description.

The participant counts require caution. Total enrollment is 95, while the primary outcome description identifies 91 parents who completed both methods. The record does not state why these counts differ, how missing information was handled, or whether the reported participants constitute the complete analysis population for each dietary component. Recruitment details and participant characteristics beyond the stated population are not reported in the registry record.

Salivary inflammatory biomarkers are listed as a secondary outcome, but their measures, analysis methods, and results are not described. The registry record also does not report a multiplicity strategy or a statistical analysis plan. These omissions limit assessment of how the outcomes were prioritized and how results across multiple measures should be interpreted. The stated completion status and study purpose do not resolve these uncertainties.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Validation of a Food Frequency Questionnaire — NCT03138330.
• Registry record last updated 2020-04-14; status last verified 2020-04.
• Results first posted 2020-04-02.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

18. Record Summary

This completed observational study was titled Validation of a Food Frequency Questionnaire. Its stated purpose was to collect nutritional information from young children aged 15-36 months to validate a newly developed food frequency questionnaire. Enrollment was 95; the primary outcome description reports 91 parents completing the sqFFQ and a 2-day weighed food record. The reference method was the weighed food record, and Pearson correlation was the stated comparison measure for energy, carbohydrates, fat, protein, fibre, sugars, calcium and iron.

The primary outcome timeframe was up to 1 month. Salivary inflammatory biomarkers were listed as a secondary outcome. The record reports 0 serious adverse events, 0 other adverse events and 0 deaths among the listed FFQ Group, with 91 at risk in each adverse-event category.

The numerical correlation results, confidence intervals, p-values, biomarker findings, and further details of missing-data handling are not reported in the registry record. The posted record therefore describes the validation question and study methods, but does not provide enough results to judge the questionnaire’s measured association with the reference method.