Pilot Trial of the Nutrition-Supported Diabetes Education Program: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Pilot Trial of the Nutrition-Supported Diabetes Education Program (NCT05067452), a randomized na study. The primary outcome is Change in Hemoglobin A1c (HbA1c) From Baseline to Twelve Weeks by Study Arm.

2. Clinical Question

3. Trial Design

This was a pragmatic pilot randomized controlled trial with a parallel-group design. The registry lists randomized allocation and double masking. It does not report which people were masked, how randomization was generated, whether allocation was concealed, or how masking was maintained. The trial is listed as phase NA. The sponsor was the University of California, San Francisco.

The study addressed a combined program rather than a single isolated component: diabetes-tailored food support was paired with diabetes self-management education, and the intervention description also includes case-management sessions. The comparator was standard care, including diabetes self-management education and access to referrals described in the record. Consequently, an arm comparison concerns the combined intervention package against the described standard-care approach; it cannot separate the contribution of food support from education or case management.

The registry describes the study as a pilot intended to assess feasibility and preliminary impact. The stated areas of interest included glycemic control, food security, diet, mental health, and health care behaviors among Medicaid-enrolled patients with type 2 diabetes in a safety-net county health system. The enrollment listed in the record is 36. The primary outcome analysis reports 13 participants per arm; other outcomes have different reported analysis counts. The registry record does not explain those differences or give a participant-flow account.

The recorded study start was 2021-11-08, primary completion was 2023-05-18, and status is completed. The original design assumptions, including any formal sample-size justification, are not reported in the registry record. Interim analyses, randomization details, and the prespecified statistical analysis plan are also not reported.

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

Reasons for not completing this period:

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 Change in Hemoglobin A1c (HbA1c) From Baseline to Twelve Weeks by Study Arm, measured over Baseline and twelve weeks.

The primary endpoint measures change in hemoglobin A1c (HbA1c) from baseline to twelve weeks, expressed as a percentage of glycated hemoglobin. The registry describes the endpoint as a mean change by study arm, with standard deviations used to describe variation among participants. HbA1c is used in clinical research as a measure of glycemic control, making it relevant to a program intended to support adults with type 2 diabetes. That general rationale does not establish that a change in this pilot demonstrates a clinical benefit.

The key comparison is the difference between the arms’ average changes, rather than either arm’s change considered alone. A difference near zero would mean that the observed average changes were similar; a positive or negative difference indicates which arm had the more positive average change, according to the reported sign convention. The reported sponsor analysis estimates an adjusted contrast using ANCOVA, while the unadjusted recalculation compares posted summary statistics. Those methods do not answer the question in exactly the same way.

The measurement window is baseline and twelve weeks. Although the food-support description extends from baseline to 24 weeks, the primary endpoint’s stated timeframe is twelve weeks. The registry record does not report a threshold for a clinically important HbA1c change, so statistical estimates alone cannot establish whether a particular difference would be meaningful to patients or clinicians.

7. Results Posted in the Registry

• Diabetes-tailored Food Support Plus Diabetes Self-management Education: mean 0.13 versus -0.02 on control; difference 0.15 (95% CI -1.31 to 1.62), standardized effect d = 0.08; z-test p = 0.836.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Change in Health-related Quality of Life (Healthy Days) From Baseline to Twelve Weeks by Study Arm

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Diabetes-tailored Food Support Plus Diabetes Self-management Education: mean 0.80 versus -1.38 on control; difference 2.18 (95% CI -7.48 to 11.84), standardized effect d = 0.16; z-test p = 0.658.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Diabetes-tailored Food Support Plus Diabetes Self-management Education and the control. The interval gives the range of effects that remain compatible with the data.

Change in Consumption of Fruits and Vegetables From Baseline to Twelve Weeks by Study Arm

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Diabetes-tailored Food Support Plus Diabetes Self-management Education: mean 0.32 versus 0.29 on control; difference 0.02 (95% CI -0.70 to 0.74), standardized effect d = 0.02; z-test p = 0.950.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Diabetes-tailored Food Support Plus Diabetes Self-management Education and the control. The interval gives the range of effects that remain compatible with the data.

Change in Consumption of Added Sugars From Baseline to Twelve Weeks by Study Arm

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Diabetes-tailored Food Support Plus Diabetes Self-management Education: mean -0.94 versus -0.41 on control; difference -0.54 (95% CI -1.53 to 0.46), standardized effect d = -0.38; z-test p = 0.289.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Diabetes-tailored Food Support Plus Diabetes Self-management Education and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists secondary outcomes spanning glucose control categories, food security, healthy days, fruit and vegetable intake, added-sugar intake, depressive symptoms, and diabetes self-efficacy. The detailed secondary results include continuous comparisons for health-related quality of life, fruit and vegetable consumption, and added-sugar consumption. Taken together, these results do not form a pattern of precise, conclusive differences across the measured domains. Their confidence intervals are broad enough to leave substantial uncertainty, and the corresponding sponsor analyses do not supply a clear resolution of that uncertainty.

The secondary outcomes complement the primary HbA1c result by describing other dimensions relevant to the program’s stated aims. They are not substitutes for the primary endpoint, and a result in one domain would not by itself establish a change in another. The registry record does not report numerical details for all the listed secondary outcomes in the material available for this interpretation, so no further outcome-specific conclusion is warranted.

Multiple outcomes create multiple opportunities for apparently notable results to arise by chance. Unless a multiplicity approach is prespecified and applied, unadjusted secondary p-values generally carry less confirmatory weight than a single primary test. The registry record does not report a multiplicity adjustment or a hierarchy for secondary testing. These outcomes are therefore best read as exploratory context for a pilot, rather than as independent confirmatory evidence. This caution concerns the strength of inference, not whether the outcomes are relevant or worth studying.

The overall secondary pattern should also be considered alongside the study’s pilot purpose and the uncertainty in the primary comparison. Follow-up research could use these measured domains to refine questions and measurement priorities, but the present estimates do not establish that the combined program changed them.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Change in the Percentage of Participants With Glucose Control From Baseline to Twelve Weeks by Study Arm
• Change in the Percentage of Participants With Low and Very Low Food Security From Baseline to Twelve Weeks by Study Arm
• Change in Health-related Quality of Life (Healthy Days) From Baseline to Twelve Weeks by Study Arm
• Change in Consumption of Fruits and Vegetables From Baseline to Twelve Weeks by Study Arm
• Change in Consumption of Added Sugars From Baseline to Twelve Weeks by Study Arm
• Change in Depressive Symptoms From Baseline to Twelve Weeks by Study Arm
• Change in Diabetes Self-efficacy From Baseline to Twelve Weeks by Study Arm

9. Statistical Interpretation

10. Statistical Methods Explained

What is the primary comparison?

The primary endpoint is a continuous outcome: change in HbA1c from baseline to twelve weeks. A mean summarizes the central value of a continuous measure, while a standard deviation describes how much individual observations vary around that mean. The comparison of interest is the difference between the two arm means. The recalculated mean difference is 0.154, calculated from the posted arm summaries without adjustment. Its sign reflects the ordering of intervention minus standard care; it is not, by itself, a judgment about clinical benefit.

The arm summaries report a mean change of -0.023 with standard deviation 1.63 for standard care, and 0.131 with standard deviation 2.14 for the intervention. The reported analysis counts are 13 in each arm. The standard deviations are substantial relative to the mean changes, which is one reason the comparison is uncertain. The record does not report a transformation of HbA1c change or an alternative estimand.

What does the sponsor analysis do?

The registry reports ANCOVA for the primary endpoint, with the parameter labeled slope. In general, analysis of covariance estimates a group contrast while accounting for a covariate or covariates. The registry record does not specify which covariates were included, how they were selected, or how the model was fitted. The reported slope is -0.092, with a 95% confidence interval from -1.59 to 1.40 and p = 0.900. Because the adjustment details are not reported, the estimate should be described as the sponsor’s posted ANCOVA result rather than reconstructed or attributed to a particular adjustment.

The recalculation is a simple, unadjusted comparison of the posted means and standard deviations. It does not reproduce the ANCOVA model, and it should not be presented as a substitute for the sponsor’s analysis. The recalculated estimate and the sponsor estimate are different summaries based on different methods; the registry marks no discrepancy. The sponsor’s analysis is the relevant reported model-based result, while the recalculation provides a transparent comparison using the aggregate arm summaries.

How should p-values be understood?

A p-value describes how compatible the observed data are with a specified null model, conditional on the test assumptions. It is not the probability that the null hypothesis is true, and it does not measure the size or importance of an effect. For the primary endpoint, the unadjusted recalculated p-value is 0.836 and the sponsor-reported ANCOVA p-value is 0.900. The distinct values reflect the distinct analyses. Neither should be interpreted as proof of no difference.

For a small pilot, estimates and confidence intervals are generally more informative than a binary significant-or-not label. A nonsignificant result may reflect a small underlying contrast, substantial variability, limited information, or some combination. The record does not report a prespecified threshold for statistical significance or a detailed analysis plan, so interpretation should not assume unreported decision rules.

Why does variability matter?

Standard deviations describe individual-level variation, whereas standard errors describe uncertainty in estimated means. A confidence interval for a contrast combines information about variability and the amount of information contributing to the analysis. Large variability relative to the observed contrast, together with limited analysis counts, tends to yield a wide interval. That is visible in the primary interval, which includes values on both sides of zero.

Over-dispersion is a general concern when observed variability is greater than a model assumes. If a model understates variability, its standard errors and confidence intervals can be too small, creating an appearance of greater precision than the data justify. The primary summaries report standard deviations, but the registry record does not describe a dispersion diagnostic or a special over-dispersion adjustment. It would therefore be inappropriate to claim that over-dispersion was tested or corrected.

What is not specified?

The registry record does not report the covariates in the ANCOVA, the handling of missing outcomes, the analysis population, or the approach to deviations from assigned treatment. It also does not report assumptions about outcome distributions, model diagnostics, interim analyses, or whether standard errors were adjusted for any design features. These details can affect estimation and inference. Without them, readers can describe the posted estimates and their stated methods, but cannot fully reproduce or audit the analysis.

General statistical principles help explain the reported methods, but they do not fill gaps in the trial record. In particular, no unreported protocol choice or model feature should be inferred from the fact that the study was randomized, double-masked, or described as pragmatic.

11. Confidence Intervals and Precision

The unadjusted primary contrast is a mean difference of 0.154, with a 95% confidence interval from -1.308 to 1.616. A confidence interval gives a range of values reasonably compatible with the data and the assumptions of the analysis. In repeated use of the same method under its assumptions, intervals at this confidence level would cover the true parameter at the stated rate. It is not a range that guarantees where the true effect lies for this particular trial.

Because the primary interval includes zero, the unadjusted data are compatible with no mean difference as well as with contrasts in either direction. Its endpoints are far from the point estimate, so the estimate is not precise. The sponsor ANCOVA interval, from -1.59 to 1.40, also includes zero and is broad. The intervals are based on different methods and should not be combined or treated as interchangeable.

Precision is distinct from whether a result crosses a conventional significance threshold. The size and breadth of the interval convey the range of uncertainty; a p-value does not show that range. The pilot’s limited outcome analysis counts and the variability in HbA1c changes help explain why the primary contrast remains uncertain, though the registry record does not provide a formal precision target.

The interval does not establish equivalence. That would require a clinically justified margin and an appropriate design and analysis, neither of which is reported in the registry record. Nor does a wide interval establish that every value inside it is equally plausible. It indicates that the available information does not sharply distinguish among a range of values under the stated approach.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning illustrations express the required sample size as participants per arm, not as total participants or total events. They use a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each row plans for a stated fraction of the observed effect: 100%, 75%, or 50%. These are planning assumptions for an illustration based on the observed estimate, not the original design assumptions for this pilot. The trial’s own sample-size justification and original design assumptions are not reported in the registry record.

The 80% power column describes the number of participants per arm that, under the listed assumptions and target effect, would give an 80% chance of rejecting a null hypothesis of no difference if that effect were the true one. The 90% power column describes the corresponding planning target for a 90% chance. These are prospective design concepts; they do not say what the trial’s actual power was after the outcomes were observed.

For the primary endpoint, the illustration gives 2,395 participants per arm for planning around 100% of the observed effect, 4,258 per arm for 75%, and 9,580 per arm for 50% at the 80% power target. At the 90% target, the corresponding figures are 3,207, 5,700, and 12,825 participants per arm. Smaller effects require more participants because they are harder to distinguish from random variation. The figures are especially sensitive to the small observed contrast and should not be read as a recommendation for a future trial without clinical and design justification.

This is not observed power. It is a planning illustration computed from the observed effect and stated assumptions, not a retrospective measure of how likely this completed trial was to detect an effect. Time-to-event illustrations, if present for another outcome, use total events; these continuous-outcome figures are participants per arm.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 0.154; pooled standard deviation σ = 1.902.
2. n = 2 × (1.96 + 0.842)² × 1.902² / 0.154² = 2,394.9, rounded up to 2,395 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- 0.154                      # observed difference in means
sigma <- 1.90216981366018                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(5067452)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 2,395 (80% power), 3,207 (90% power)
 75% of observed effect: n per arm = 4,258 (80% power), 5,700 (90% power)
 50% of observed effect: n per arm = 9,580 (80% power), 12,825 (90% power)
power.t.test: 2395.9 
Simulated power at n = 2,395 per arm: 0.795 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The sponsor-reported primary analysis is ANCOVA, with a slope of -0.092, a 95% confidence interval from -1.59 to 1.40, and p = 0.900. This should be distinguished from the unadjusted recalculation from arm-level means and standard deviations: mean difference 0.154, 95% confidence interval from -1.308 to 1.616, and p = 0.836. The recalculation does not adjust for covariates and is not a reproduction of the sponsor’s model.

Both intervals include zero and span contrasts in either direction. Their estimates are not identical because the analyses differ; the registry does not report the covariates or other model details needed to explain the exact difference. The registry indicates no discrepancy, so the sponsor analysis remains the reported trial analysis, while the unadjusted calculation is supplementary context derived from aggregate summaries.

The sponsor analysis should take precedence as the trial’s own reported result. The recalculation can help readers understand what the arm summaries imply under a simple unadjusted comparison, but it cannot override the sponsor’s fitted model or supply missing model specifications. Neither result establishes equivalence, and neither provides a precise estimate of the treatment contrast.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when several outcomes or comparisons are examined: the more tests performed, the greater the chance that at least one result may appear unusual under a collection of null effects. This is a general statistical issue, especially when a study has a primary endpoint and a range of secondary endpoints. The registry lists multiple secondary domains in addition to the primary HbA1c endpoint.

The registry record does not report a multiplicity adjustment, a hierarchy of secondary hypotheses, or whether secondary analyses were intended to be confirmatory or exploratory. Without that information, the secondary p-values should not be treated as if they were each the sole prespecified test. Unadjusted secondary results carry less weight because repeated testing can make isolated findings less persuasive. This does not make the measurements useless; it limits the strength of claims based on individual secondary tests.

The record also does not report how missing observations were handled, whether any imputation was used, or whether the analysis included all randomized participants. The primary summaries give 13 participants per arm, while some secondary summaries give different arm-specific analysis counts. These counts show that the reported denominators vary by outcome, but they do not explain why. The registry record does not provide a flow of participants through enrollment, follow-up, and analysis or a description of reasons for missingness.

Missing-data methods matter because an analysis can be biased if the observed participants differ systematically from those with missing outcomes. General methods address that issue by making assumptions about how missingness relates to observed or unobserved outcomes. No particular missing-data mechanism or method can be attributed to this trial from the registry record.

The analysis population is likewise not reported in sufficient detail. Randomized allocation is listed, but the record does not state whether analyses followed intention-to-treat, per-protocol, or another population definition. The covariates in the primary ANCOVA and the handling of protocol deviations are also not reported. These are limitations on reproducibility and interpretation; they do not establish that a particular approach was or was not used.

16. 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 adverse-event record reports zero serious adverse events among 15 participants at risk in standard care and zero among 21 participants at risk in the intervention arm. It also reports zero other adverse events in each of those stated risk sets, and zero deaths in both arms. The risk-set counts are those reported for the adverse-event summaries; they should not be assumed to match the primary endpoint analysis counts.

Zero reported events means none were recorded in the listed categories and risk sets in the registry results. It does not prove that an event is impossible, that risk is absent, or that the intervention is safer than standard care. With a pilot-scale record, an absence of observed events can coexist with uncertainty about uncommon harms. The registry record does not give further event descriptions or a detailed safety-monitoring approach, and no serious adverse-event terms are listed.

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

The primary result is an uncertain estimate of the difference in mean HbA1c change between the intervention package and standard care. The sponsor’s ANCOVA estimate and the unadjusted recalculation both have confidence intervals that include zero and extend in either direction. The data therefore do not precisely determine the direction or size of the contrast. A large p-value is not proof that the arms are the same, and an interval containing zero is not an equivalence result.

The results do not show that the intervention works or does not work. They describe what was observed in a pilot comparison, with limited outcome analysis counts and considerable uncertainty. The intervention was a package of food support, diabetes self-management education, and described case-management sessions, so the arm contrast cannot identify which component might account for any observed difference. The registry record does not report a clinically important difference threshold for HbA1c.

The secondary outcomes broaden the description beyond glycemic control, covering diet, health-related quality of life, food security, mental health, and self-efficacy. Their role is useful for understanding the range of outcomes considered, but multiplicity and missing analysis details limit strong conclusions from secondary tests. Nor do the posted results establish that outcomes outside the reported measures changed.

Because this was a pilot, its results are most appropriately treated as preliminary statistical evidence. The record identifies feasibility as a study aim, but does not report quantitative feasibility findings in the information summarized. It also does not report the original sample-size rationale, missing-data methods, or analysis population. These limitations make the findings informative about the uncertainty and scope of the evaluation, rather than definitive about a clinical effect or the feasibility of implementing the program more broadly.

18. Why This Trial Matters Statistically

This trial is statistically interesting because it evaluates a bundled, pragmatic program in a population and care setting explicitly described in the registry: Medicaid-enrolled adults with type 2 diabetes in a safety-net county health system. Its aims span glycemic control and several intermediate outcomes, so the analysis illustrates how a pilot can address a broad set of questions while still leaving considerable uncertainty around any one contrast.

The primary result is a useful example of why point estimates, confidence intervals, and p-values answer different questions. The point estimate describes the observed contrast, the interval conveys its imprecision, and the p-value describes compatibility with a null model under test assumptions. None alone establishes clinical importance. The contrast is small relative to the reported variability, and the confidence interval allows a wide range of possible values.

The trial also highlights the distinction between a sponsor’s adjusted analysis and a simple recalculation from aggregate summaries. Both are informative in different ways, but the unadjusted comparison cannot reproduce the ANCOVA result when model specifications are not reported. Reproducibility depends not only on numerical summaries but also on knowing the model, population, covariates, and missing-data procedures.

Finally, planning calculations based on an observed pilot effect can be highly sensitive to that estimate. Their role is illustrative rather than a substitute for a future design rationale grounded in scientific and clinical priorities. Taken together, the registry results make the uncertainty visible and help define the statistical questions that would need clearer reporting in a larger or confirmatory evaluation.

19. 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 analysis counts are limited, and the primary means are accompanied by substantial standard deviations. This produces broad confidence intervals and weak precision for the primary contrast. The registry lists enrollment of 36, but the primary outcome summaries report 13 participants in each arm; it does not explain the difference or provide a complete participant-flow account. Secondary outcome denominators also vary, with no reported explanation for those variations.

The trial evaluates a combined program, including diabetes-tailored food support and diabetes self-management education, with case-management sessions described for the intervention. The design therefore cannot isolate the effect of any one component from the information reported. The control description includes standard care and referrals, but the record does not quantify what participants actually received in either arm.

Important analytic details are not reported in the registry record: covariates in the ANCOVA, missing-data methods, analysis population, model diagnostics, and the handling of protocol deviations. The record also does not specify which people were masked, how allocation was concealed, or how randomization was implemented. These gaps limit reproducibility and make it difficult to assess potential sources of bias.

The study is described as a pilot, but its original sample-size justification and design assumptions are not reported. The registry lists multiple secondary outcomes, yet does not report a multiplicity strategy. That makes isolated secondary results less persuasive as confirmatory evidence. The registry’s stated feasibility aim is not accompanied in the available record by quantitative feasibility results. These limitations call for restraint in interpreting both the apparent effect and the broader applicability of the findings.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Pilot Trial of the Nutrition-Supported Diabetes Education Program — NCT05067452.
• Registry record last updated 2025-07-29; status last verified 2025-07.
• Results first posted 2025-07-29.
• 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.

23. Record Summary

The completed randomized, double-masked, parallel-group pilot compared diabetes-tailored food support plus diabetes self-management education with standard care and diabetes self-management education among low-income adults with type 2 diabetes. The registry lists enrollment of 36 and identifies the University of California, San Francisco as sponsor. The study start was 2021-11-08, with primary completion on 2023-05-18.

The primary outcome was mean change in HbA1c from baseline to twelve weeks. The sponsor-reported ANCOVA estimate was -0.092, with a 95% confidence interval from -1.59 to 1.40 and p = 0.900. The unadjusted recalculation estimated a mean difference of 0.154, with a 95% confidence interval from -1.308 to 1.616 and p = 0.836. Both estimates are imprecise and their intervals include zero.

Secondary measures covered diet, health-related quality of life, food security, mental health, and diabetes self-efficacy. The registry record reports zero serious adverse events and zero deaths in both listed risk groups. Missing-data methods, analysis-population details, multiplicity handling, and the original sample-size rationale are not reported in the registry record.