Multi-level Supermarket Discount Study: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Multi-level Supermarket Discount Study (NCT04178824), a randomized na study. The primary outcome is Fruit Intake as Assessed by 24-h Dietary Recalls.

2. Clinical Question

3. Trial Design

This was a randomized, single-masked, parallel-group study of a behavioral intervention. The registry lists three groups: a 15% discount intervention, a 30% discount intervention, and a no-intervention control group receiving a 0% discount. The discounts applied to fruits, vegetables, and non-caloric beverages at designated supermarkets. Enrollment was 167, and the study status is completed.

The planned intervention lasted 32 weeks. It followed an 8-week baseline and was followed by a 16-week period without discounts. The primary outcome schedule lists fruit intake at Week 8, 24, 40, and 56. These points span the baseline, intervention, and follow-up periods as described in the registry, although the registry does not specify in the outcome entry which visits correspond to each study period. The scheduled measurement time points should not be mistaken for reported results at those visits.

The registry identifies the allocation as randomized, the masking as single, and the model as parallel. It does not identify who was masked, explain the randomization sequence or allocation concealment, or describe procedures for maintaining masking. The method of assigning participants to groups, group sizes at enrollment, retention over time, and handling of participants who did not follow the assigned discount are not reported in the registry record.

The sponsor was Icahn School of Medicine at Mount Sinai. The record reports a start date of 2016-09-01 and a primary completion date of 2022-11-15. The listed phase is NA. The registry describes additional outcomes including body weight, body mass index, percent body fat, waist circumference, fasting glucose, cholesterol level, blood pressure, and gross weekly purchasing of fruits and vegetables. It does not report numerical results for these outcomes.

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

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 Fruit Intake as Assessed by 24-h Dietary Recalls, measured over Week 8, 24, 40, 56.

The primary endpoint is fruit intake, measured in grams using 24-h dietary recalls. A dietary recall asks a participant to report foods consumed over a defined day; here, the registry identifies the measure as fruit intake assessed by 24-h dietary recalls. The outcome is therefore a continuous measure of reported intake, with the mean and a 95% confidence interval identified as the intended summary format.

As a general statistical rationale, measuring intake can help evaluate whether an intervention aimed at changing food purchasing is accompanied by a change in reported consumption. That makes intake a distinct outcome from purchasing: a discount may alter what is bought, while intake captures reported consumption. The registry also names purchasing outcomes, but the primary endpoint is fruit intake, not purchases or a health biomarker.

The outcome was scheduled for assessment at Week 8, 24, 40, and 56. Repeated measurements can describe how an outcome varies over time, including whether group differences appear during the intervention or persist into follow-up. However, no mean fruit-intake values, confidence intervals, or comparisons between groups are reported in the registry record. The record also does not describe the recall procedures, how multiple recall days were handled, or steps taken to reduce errors in dietary reporting. Consequently, the endpoint’s measurement schedule and unit are clear, but the actual intake levels and their precision are not.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a mean summarize?

The primary outcome is continuous and measured in grams. A mean summarizes the average measured intake within a group or at a particular assessment. It can provide a compact description of the center of a distribution, but does not describe every participant’s intake or the spread and shape of individual values. The registry identifies the mean as the parameter of interest. It does not report the actual means or describe whether the primary analysis used raw values, changes from baseline, or adjusted estimates.

In general, group means are most informative when interpreted alongside a measure of uncertainty and the comparison they are intended to answer. An isolated mean does not tell whether groups differ, and a difference between means does not by itself establish how precisely that difference has been estimated. The registry names a 95% confidence interval as the dispersion measure, but no confidence interval values are reported.

What does a confidence interval convey?

In general, a confidence interval expresses the uncertainty around an estimate under a stated statistical procedure. A narrower interval indicates greater precision than a wider interval, while the interval’s location indicates which values remain compatible with the data and method. A confidence interval is not a guarantee that the true value lies inside a particular computed interval. Its interpretation depends on the estimate, the data, and the assumptions of the analysis.

For this outcome, an interval could describe uncertainty around a mean or a between-group contrast, but those are different quantities. The registry record does not state which quantities would receive intervals or give the intervals themselves. It therefore supports describing the planned summary format, not interpreting a specific estimate.

How can the randomized groups be compared?

Random allocation is intended, in general, to make treatment groups comparable on average at the time of assignment. It does not ensure that every group has the same baseline characteristics, especially in a particular study. Adjustment for baseline outcome values or other covariates can be used in some analyses, but whether adjustment was used, and which variables might have been included, is not reported in the registry record.

The three groups make more than one comparison possible: each discount level could be contrasted with the no-discount control, and the two discount levels could be compared with each other. These comparisons address different questions. The registry does not identify a primary contrast among them or report any contrast estimates. It would be inappropriate to assume that the 15% and 30% groups were compared in a particular way or that a dose-response pattern was formally tested.

Why do repeated measurements matter?

Fruit intake was scheduled for measurement at Week 8, 24, 40, and 56. Repeated observations from the same participant are generally related rather than independent. Statistical methods for repeated measures account for that within-participant relationship when estimating changes or group differences over time. Different methods can make different assumptions about the correlation structure, missing observations, and shape of change over time.

The registry identifies a parallel study and multiple outcome times, but it does not report a repeated-measures model, a model for change, or separate comparisons at each visit. It also does not state how correlations among repeated recalls were handled. The presence of several scheduled assessments alone does not establish that all visits were analyzed together.

What assumptions underlie common comparisons?

In general, methods that compare means may rely on assumptions about the distribution of measurements, the independence of observations across participants, or the adequacy of a model for the observed data. For repeated measures, the relationship among measurements from the same person also matters. If assumptions are not well matched to the data, standard errors and intervals may not accurately represent uncertainty. The registry does not report diagnostic checks, transformations, robust methods, or other details that would allow readers to evaluate these issues for this trial.

Dietary intake data can have variation across people and days. More generally, over-dispersion means that observed variability is greater than a model expects; when relevant, accounting for it can prevent uncertainty from being understated. The registry does not characterize the distribution of fruit-intake measurements or report whether over-dispersion was assessed. This is a general statistical consideration, not a claim that over-dispersion occurred in this study.

What can the posted analysis information establish?

The registry record lists 14 outcome measures and reports zero analyses. It names the primary outcome and its scheduled times, but does not provide a statistical analysis plan, a test statistic, a p-value, or estimates of effect. The exact analysis population, handling of protocol deviations, missing data, and any adjustment for multiple comparisons are not reported in the registry record. No statistical model or analysis procedure should be inferred from the outcome description alone.

The record describes planning illustrations based on a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the effect to detect expressed as a fraction of the observed effect. Such calculations are planning illustrations, not observed power and not evidence about the results. The registry does not report the study’s own sample-size justification or design assumptions. It does not report interim analyses or explain whether any were conducted.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 14 outcome measures, including the primary fruit-intake outcome and secondary measures of body weight, body mass index, percent body fat, waist circumference, fasting glucose, cholesterol level, blood pressure, and gross weekly purchasing of fruits and vegetables. The primary outcome itself has four scheduled time points: Week 8, 24, 40, and 56. Multiple outcomes and time-specific comparisons create multiple opportunities to observe a result that appears notable by chance. In general, statistical plans may address this through a defined hierarchy of outcomes, adjustment procedures, or emphasis on estimation and uncertainty.

The registry record does not report whether any multiplicity adjustment was planned or used, how the primary comparisons were prioritized, or whether analyses at different time points were treated as separate tests. It also does not report a statistical analysis plan. The record lists zero analyses, so it is not possible to determine how the outcomes were evaluated or whether reported results were selected from a broader set of analyses.

Missing-data handling and analysis populations are also not described. The record gives total enrollment of 167 but does not state how many participants contributed measurements at each visit, whether withdrawals occurred, or how incomplete dietary recalls were handled. It does not identify an intention-to-treat population, a per-protocol population, or another analysis set. In general, missing observations can affect estimates if the people with observed data differ from those without observed data. Without participant flow and a stated missing-data approach, completeness and potential bias cannot be assessed.

The randomized design is recorded, but the record does not explain how adherence to the assigned discount was handled in analysis. The posted information is therefore insufficient to assess the analysis population, multiplicity strategy, or the possible effect of missing outcome data on the estimates.

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 lists serious adverse events and other adverse events separately for the three groups. For the 15% discount group, 38 participants were at risk for each category, and zero serious events and zero other events were recorded. For the 30% discount group, 33 participants were at risk for each category, with zero serious events and zero other events recorded. For the no-intervention control group, 36 participants were at risk for each category, with zero serious events and zero other events recorded. Deaths are listed as zero in each group. No serious adverse-event terms are listed.

These figures describe what the registry record reports; they do not establish that no events occurred outside the record’s reporting scope. The record does not describe how adverse events were collected, the duration of safety observation for each participant, or whether all randomized participants contributed to the listed at-risk counts. The at-risk counts should not be interpreted as the total enrollment, since the registry reports enrollment of 167 and gives group-specific counts for the safety entries.

Zero recorded events in the listed categories do not prove that an intervention is risk-free. When event counts are zero, the record offers no event types or comparative event estimates to evaluate. It also does not provide confidence intervals for event rates or explain how differences in exposure or follow-up were handled. These safety entries are limited descriptive information, not a complete safety analysis.

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

The record establishes the study’s broad design and the outcomes it set out to assess. It describes a randomized, single-masked, parallel-group behavioral study with discounts of 15% and 30% compared with a no-discount control. Fruit intake, measured in grams by 24-h dietary recalls, was the primary outcome, scheduled at Week 8, 24, 40, and 56. The record also identifies outcomes related to purchasing, body composition, and cardiovascular measures.

Those design details do not reveal the observed effect. Numerical fruit-intake results, group comparisons, confidence intervals, and statistical tests are not reported in the registry record. The record therefore does not show whether either discount level was associated with higher, lower, or similar fruit intake relative to control. It cannot establish whether an observed difference, if one existed, was precise or compatible with a wide range of values.

The intervention addressed discounts on fruits, vegetables, and non-caloric beverages, while the primary outcome concerned fruit intake. An effect on purchasing would not automatically establish an effect on consumption, and an effect on reported consumption would not by itself establish a change in body composition or cardiovascular measures. Each outcome requires its own reported estimate and uncertainty. The registry does not post numerical results for these secondary outcomes.

Enrollment of 167 and a completed status provide context about the study record, but are not substitutes for participant-level outcome data or analysis results. The numbers available for each assessment, the treatment of missing data, and the analysis population are not reported. The registry’s safety entries list zero events in the stated categories, but do not establish that there was no risk or that safety was assessed comprehensively.

Accordingly, the record supports a description of the research question, intervention structure, and planned measurement schedule. It does not support a conclusion about whether the discounts changed fruit intake, purchasing, or health outcomes. That distinction is important: absence of posted results is not evidence that the groups were equivalent, and it is not evidence that the intervention had a particular effect.

13. Why This Trial Matters Statistically

This study is statistically notable because it places a behavioral and economic intervention alongside measurements at several stages: supermarket purchasing, dietary intake, and health-related outcomes. Those are related but distinct constructs. Separating them helps clarify what an analysis can establish. A change in purchasing is not the same endpoint as a change in reported intake, and neither is interchangeable with a measured change in body composition or a cardiovascular marker.

The three-arm structure also raises a question about the comparison of discount levels. The study includes 15%, 30%, and 0% discounts, but the registry does not report which contrasts were designated as primary or how comparisons across the three groups were handled. Statistical interpretation depends on the comparison being estimated, not simply on the presence of multiple groups.

Measurement over baseline, intervention, and follow-up creates an opportunity to examine timing and persistence. Repeated measurements can be informative, but only when the analysis and its assumptions are clear. The registry gives the schedule and study periods but does not report how repeated observations were modeled or how missing measurements were addressed.

The registry record also illustrates why design information and results information must be distinguished. It reports randomization, enrollment, outcome definitions, and study status, yet lists zero analyses and contains no numerical primary-outcome results. A completed study record can therefore describe the intended question without allowing readers to estimate its answer. Statistical transparency requires estimates, uncertainty, analysis details, and accounting for the participants contributing data, not only a protocol-like description of the intervention.

For biostatistical interpretation, the primary issue is not simply whether fruit intake was measured, but what estimate was produced, for which comparison and time point, using what analysis population and assumptions. Those details are not reported in the registry record, limiting assessment of both the intervention estimate and its precision.

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 main limitation of the available record is that it contains no numerical results for the primary outcome. Means, confidence intervals, between-group estimates, and statistical tests are not reported in the registry record. The same absence applies to the listed secondary outcomes. As a result, the magnitude and precision of any group differences cannot be assessed.

Although the registry identifies randomization and single masking, it does not describe sequence generation, allocation concealment, the identity of the masked party, or how masking was maintained. It also does not provide baseline characteristics or group-specific counts at the primary outcome visits. These omissions prevent assessment of the balance of measured characteristics and the availability of data over time.

The primary endpoint relies on 24-h dietary recalls, but the registry does not describe recall administration, the number of recall days, or the handling of incomplete reports. The accuracy and variability of reported intake therefore cannot be evaluated from this record. The outcome is explicitly fruit intake, while discounts applied to fruits, vegetables, and non-caloric beverages; the registry does not report how exposure to each discounted product category was measured.

The record lists four primary-outcome assessment times and 14 outcome measures, but does not state how multiplicity was managed, which contrast was primary, or whether repeated measurements were analyzed jointly. Missing-data procedures, analysis populations, adherence handling, and sensitivity analyses are not reported. The study’s own sample-size justification and any interim analyses are also not reported in the registry record.

Finally, the safety entries list zero events but do not explain ascertainment or the duration of observation represented by the at-risk counts. These limitations concern the information available in the registry record; they do not establish what methods or results may exist outside it.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Multi-level Supermarket Discount Study — NCT04178824.
• Registry record last updated 2024-10-02; status last verified 2024-09.
• Results first posted 2024-09-19.
• 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

The registry describes a completed randomized, single-masked, parallel-group behavioral study with enrollment of 167 participants and an obesity condition. Participants were assigned to discounts of 15% or 30% on fruits, vegetables, and non-caloric beverages, or a 0% discount control. The intervention period was 32 weeks, preceded by an 8-week baseline and followed by a 16-week period without discounts.

The primary outcome was fruit intake in grams, assessed using 24-h dietary recalls at Week 8, 24, 40, and 56. Secondary outcomes included body composition, fasting glucose, cholesterol, blood pressure, and gross weekly fruit and vegetable purchasing. The registry lists zero analyses and does not report numerical outcome results, analysis methods, missing-data procedures, or the study’s own sample-size justification.

Safety entries list zero serious adverse events, zero other adverse events, and zero deaths in each group. These entries do not constitute a complete safety assessment. The record supports describing the design and planned outcomes, but not estimating the effects of the discounts.