Evaluation of Diet and Sleep in Vascular Health: A Pilot Study: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Evaluation of Diet and Sleep in Vascular Health: A Pilot Study (NCT04990973), a randomized na study. The primary outcome is Score on the Pittsburgh Sleep Quality Index (PSQI).

2. Clinical Question

The study’s broad aim was to test the impact of diet on sleep and cardiovascular disease risk factors. Its listed secondary outcomes include urinary 6-sulfatoxymelatonin level, tryptophan, kynurenine, and serotonin. The record does not report results for those outcomes in the information available for this analysis.

3. Trial Design

This was a randomized, single-masked, crossover pilot study sponsored by Columbia University. The registry lists enrollment of 11 and describes two diet conditions: the Mediterranean Diet and the Average American Diet (AAD). In a crossover design, participants receive more than one condition in sequence, allowing observations from the same participant under different conditions. The record describes two sequences: Mediterranean Diet followed by a 4-week washout and AAD, or AAD followed by a 4-week washout and the Mediterranean Diet.

The registry lists the study as starting on 2021-07-02, with primary completion on 2022-10-28, and status as terminated. It identifies the study phase as NA. The meaning of the single-masked designation for participants, staff, or outcome assessors is not reported in the registry record. The record also does not report the method used to generate the randomization sequence or conceal treatment assignment.

The primary outcome was the score on the Pittsburgh Sleep Quality Index, measured over a stated timeframe of 4 weeks. The posted arm-level estimates are a mean of 0.3 with a standard deviation of 1.5 for the Mediterranean Diet condition, based on 3 participants, and a mean of -0.5 with a standard deviation of 2.1 for the AAD condition, based on 4 participants. These figures are not an estimable two-arm comparison. The record does not describe the statistical approach used to account for the crossover sequence, repeated measurements, or period effects.

The registry lists 7 outcome measures and 4 secondary outcomes, and reports 0 analyses. The reason for termination, the protocol’s planned sample size, and any sample-size justification are not reported in the registry record. No specific conclusions about the planned analysis can be drawn from the design description alone.

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.

Before assignment: 11 individuals signed a consent form prior to screening assessments. Of those, 4 participants passed the eligibility criteria and were randomized to a study condition and participated in the study activities.

First Intervention (4 Weeks)

Washout (4 Weeks)

Second Intervention (4 Weeks)

Analysis population for the primary outcome, as defined in the registry: 1 participant in the Mediterranean Diet condition did not complete the questionnaire.

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.

Population: Baseline characteristics were only collected from participants who completed the study.

6. Primary Endpoint

The registry lists the primary outcome as Score on the Pittsburgh Sleep Quality Index (PSQI), measured over 4 weeks.

The primary endpoint is the Pittsburgh Sleep Quality Index (PSQI), a questionnaire intended to assess sleep quality and related experiences. The registry says its 19 questions cover sleep quality, latency, duration, habitual sleep efficiency, sleep disturbances, medication use, and daytime dysfunction. It describes a score range from 0 (minimum) to 21 (maximum), with scores greater than 5 indicating poor sleep quality. On that scale, a lower score represents a better reported outcome.

The PSQI provides a structured way to measure several dimensions of reported sleep using a single score. In a diet study, such a measure can capture participants’ perceptions of sleep during the specified assessment period. It is a questionnaire outcome, not a direct measure of every aspect of sleep or cardiovascular risk. The registry does not report whether questionnaire completion or scoring was verified by an assessor, nor does it describe how repeated PSQI measurements were handled in the crossover design.

The record labels the primary outcome as a score and gives a timeframe of 4 weeks. However, the posted arm-level mean for AAD is -0.5, which cannot be a raw score on the stated 0-to-21 scale. The record does not clarify whether the arm-level values represent changes from another measurement or another derived quantity. Their interpretation as raw PSQI scores is therefore not warranted. One participant in the Mediterranean Diet condition did not complete the questionnaire.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What do the reported summaries describe? The primary outcome is continuous: it is a numerical score rather than a count of events or a category. For a continuous outcome, the mean is the arithmetic average of the observed values, and the standard deviation describes how dispersed those values are around the mean. The registry reports the mean and standard deviation for each diet condition, along with the number of participants contributing to each summary. These descriptive quantities do not, by themselves, quantify uncertainty about a difference between conditions.

The endpoint description defines PSQI scores from 0 to 21, with lower scores representing better sleep quality and scores greater than 5 indicating poor sleep quality. Yet the reported AAD mean is -0.5. Since a raw score on a scale beginning at 0 cannot be negative, the arm-level figures should not be treated as absolute PSQI scores. They might represent some other summary, but what that summary is cannot be established from the registry record. The distinction matters: a mean change, for example, has a different interpretation from a mean score at a particular assessment.

Why does the crossover design matter for analysis?

General statistical explanation: a crossover design measures participants under more than one condition, so observations from the same participant are often correlated. A comparison that treats every condition measurement as independent may not reflect that correlation. Analyses of crossover data can use the within-participant pairing and may also consider treatment sequence, period, and possible carryover from an earlier condition. Which features are appropriate depends on the design and available data. The registry describes a 4-week washout between diets but does not report the analysis model or whether it assessed period or carryover effects. It is therefore not possible to say that any particular crossover method was used.

Randomization to treatment sequence can help balance factors across sequences, but it does not remove the need to describe how data were analyzed. The record does not give the randomization procedure, the number assigned to each sequence, or participant-level outcomes. It also does not state whether every participant completed both conditions. Those omissions prevent reconstruction of a paired comparison from the arm-level means and standard deviations.

What would uncertainty measures add?

General statistical explanation: a confidence interval gives a range of values compatible with the observed data and an analysis model, conveying uncertainty around an estimate. Its width depends on factors such as sample size, variability, and the analysis design. A p-value, when calculated under a stated model, describes how unusual the data would be under a specified null hypothesis; it does not measure the size or importance of an effect. These measures answer different questions and are most useful when the estimand and method are clear.

No confidence interval or p-value for a two-arm primary-outcome comparison is reported in the registry record. The record explicitly indicates that values are missing or not estimable for that comparison. It would be inappropriate to infer a comparative estimate or its precision from the separate arm summaries, particularly in a crossover study where the participant-level pairing is unavailable.

How should the small samples be understood?

General statistical explanation: estimates from small samples can be imprecise, and a few observations may strongly influence the mean and standard deviation. The standard deviation is a measure of observed spread, not a confidence interval and not a measure of the uncertainty in a treatment contrast. A confidence interval for a contrast would require a specified estimand and suitable analysis of the study’s data. The small arm-level counts make caution important, but they do not allow a numerical precision calculation without further information.

The registry includes a general description of planning illustrations based on a fraction of an observed effect, with assumptions about significance level, allocation, and a normal approximation. Such an illustration is not observed power, and it does not establish that the original study was designed around those assumptions. The protocol’s own sample-size justification is not reported in the registry record. No numerical planning result should be interpreted as a trial finding.

What is not established by the posted analysis?

The record reports 0 analyses and does not give the model, estimand, hypothesis test, confidence interval, or adjustment strategy for the primary comparison. It also does not report how the repeated-condition structure was incorporated. General statistical practice would call for an analysis aligned with the randomized crossover design and a clear statement of which participants and measurements are included. Those general considerations do not establish what the investigators planned or performed. The registry record is insufficient to reconstruct a comparative statistical analysis.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Mediterranean Diet Condition", "Average American Diet (AAD) Condition")
n   <- c(3, 4)
m   <- c(0.3, -0.5)
sp  <- c(1.5, 2.1)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Mediterranean Diet Condition  3  0.300  1.500  -3.426  4.026
CARM  Average American Diet (AAD) Condition  4  -0.500  2.100  -3.842  2.842

11. Multiplicity, Missing Data and Analysis Populations

The registry identifies the PSQI as the primary outcome and lists 4 secondary outcomes: change in urinary 6-sulfatoxymelatonin level, change in tryptophan, change in kynurenine, and change in serotonin. It reports 7 outcome measures overall and 0 analyses. The record does not describe how many hypotheses were tested, whether any secondary results were analyzed, or whether adjustments for multiple comparisons were planned or used. A multiplicity strategy is not reported in the registry record.

General statistical explanation: when several outcomes or hypotheses are examined, the chance of at least one apparently unusual result can increase if each is considered separately. Prespecified outcome priorities and a stated multiplicity approach help readers distinguish confirmatory analyses from exploratory ones. No particular approach can be attributed to this study from the registry information.

For the primary questionnaire, the registry notes that 1 participant in the Mediterranean Diet condition did not complete the questionnaire. The posted primary summaries include 3 participants for the Mediterranean Diet condition and 4 for AAD. The record does not state how missing questionnaire data were handled, whether participants with incomplete measurements were excluded from other analyses, or whether any imputation was used. It does not identify an intention-to-treat, per-protocol, or other analysis population. The planned population and the analysis population are therefore not reported in the registry record.

Because the study is a crossover, missing data in one condition may affect whether a participant can contribute to a paired comparison. The record does not say whether the same participants contributed to both arm-level summaries or explain the missingness pattern. These details are important for interpreting any contrast, but no comparative analysis is reported.

12. 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 safety record lists 4 participants at risk in each diet group. It reports 0 serious adverse events and 0 deaths in the Mediterranean Diet group, and 0 serious adverse events and 0 deaths in the AAD group. It also reports 0 other adverse events in each group. No serious adverse event terms are listed.

These figures describe what the registry records for the groups and outcomes specified; they are not proof that no event could occur. The numbers at risk are small, and the registry record does not provide details about how adverse events were defined, collected, monitored, or attributed to a condition. The safety entries also do not describe timing, severity beyond the serious-event classification, or participant-level exposure across the crossover sequences.

Safety counts should not be confused with the primary sleep-outcome sample sizes. The registry gives 3 participants for the Mediterranean Diet PSQI summary and 4 for the AAD summary, while the safety record lists 4 at risk in each group. The available information does not explain the difference. These summaries are descriptive and do not establish comparative safety.

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

The posted results describe a small pilot study with arm-level PSQI summaries, but do not provide an estimable two-arm comparison. The registry reports a mean of 0.3 with standard deviation 1.5 for 3 participants in the Mediterranean Diet condition, and a mean of -0.5 with standard deviation 2.1 for 4 participants in the AAD condition. The negative mean is inconsistent with the stated range for a raw PSQI score, and the registry does not clarify the scale or derivation of the posted values. Their meaning should therefore be treated cautiously.

The figures do not establish whether one diet is associated with better sleep quality than the other. No comparative estimate, confidence interval, or p-value is reported, and the registry does not describe a statistical analysis that accounts for the crossover structure. Separate arm summaries do not reveal the within-participant pattern that a crossover comparison would need.

These results also do not show whether diet changes cardiovascular risk factors. Although the study’s broad aim includes cardiovascular risk and the registry lists secondary measures related to melatonin and several metabolites, results for those outcomes are not reported in the information available for this analysis. Nor do the recorded zero adverse events establish that the diets are free of risk; they are limited observations in a small study.

General statistical explanation: a pilot study can provide experience with recruitment, measurements, and study procedures, but a small sample generally offers limited precision for estimating effects. The registry does not describe whether feasibility outcomes were planned or reported. The available record is best read as a limited description of a terminated crossover study, not as a definitive estimate of a diet effect or as a basis for medical decisions.

14. Why This Trial Matters Statistically

This study raises several statistical issues that are especially visible in a small randomized crossover trial. The design has participants receive both diet conditions in different sequences, so the key comparison depends on how measurements within the same person relate to one another. Arm-level means and standard deviations do not preserve that pairing. Without a reported paired or otherwise appropriate crossover analysis, readers cannot determine the estimated contrast or its precision from the summaries alone.

The endpoint reporting also illustrates why outcome definitions and numerical summaries need to align. The PSQI is described as a score that cannot be below 0, yet the AAD mean is negative. A clear explanation of whether the value is a change, adjusted estimate, or another quantity would be needed to interpret it. The registry does not provide that clarification, and the ambiguity prevents a straightforward reading of the scale’s clinical meaning.

Missing questionnaire data are consequential in a crossover setting because a participant may contribute information under one condition but not another. The record notes one incomplete questionnaire in the Mediterranean Diet condition, but does not explain how missingness affected the analysis population or whether a paired outcome was available. Reporting those choices would help readers assess potential selection and precision concerns.

Finally, the record lists multiple outcome measures but no analyses. A transparent account of which outcomes were tested, how uncertainty was quantified, and how multiplicity was handled would distinguish descriptive findings from inferential claims. These are general reporting considerations; they do not establish what the investigators planned. The registry’s limited reporting makes those questions particularly relevant to understanding what the study can and cannot answer.

15. 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 primary interpretation is constrained by the absence of an estimable two-arm comparison. The registry reports arm-level means and standard deviations but no contrast, confidence interval, or p-value. It also reports 0 analyses, so the analysis method cannot be assessed or reconstructed.

The posted AAD mean is negative, whereas the PSQI is defined as a score from 0 to 21. The record does not explain whether the values are changes or another derived quantity. This ambiguity prevents interpreting the arm-level means as absolute sleep scores and complicates any comparison with the stated threshold for poor sleep quality.

Only 3 participants contribute to the Mediterranean Diet primary-outcome summary and 4 to the AAD summary. The registry notes one incomplete questionnaire in the Mediterranean Diet condition but does not report the missing-data method, analysis population, or whether participants contributed measurements under both conditions. It also does not report individual paired observations, sequence group sizes, or the number completing both diet periods.

The crossover design includes a 4-week washout, but the registry does not report whether the analysis accounted for period, sequence, or carryover effects. The reason for termination and the original sample-size justification are not reported in the registry record. No secondary-outcome results are reported in the information available for this analysis, and the approach to multiple outcomes is not described. Together, these limits restrict interpretation to the posted descriptive summaries and safety counts.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Evaluation of Diet and Sleep in Vascular Health: A Pilot Study — NCT04990973.
• Registry record last updated 2024-10-17; status last verified 2024-10.
• Results first posted 2024-10-17.
• 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.

19. Record Summary

The registry record describes a randomized, single-masked crossover pilot study of the Mediterranean Diet and the Average American Diet, with enrollment of 11. The study is listed as terminated. Its primary outcome is the PSQI score over a timeframe of 4 weeks. The registry reports means and standard deviations for 3 participants in the Mediterranean Diet condition and 4 in the AAD condition, but no estimable two-arm comparison and 0 analyses.

The record lists 4 secondary outcomes and 7 outcome measures overall. It notes one incomplete Mediterranean Diet questionnaire and reports 0 serious adverse events, 0 other adverse events, and 0 deaths in each diet group. The meaning of the negative AAD mean, the analysis method, missing-data handling, and sample-size justification are not reported in the registry record. The available information supports descriptive reporting only, not a precise comparative estimate.