Time-Restricted Eating to Address Persistent Cancer-Related Fatigue: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Time-Restricted Eating to Address Persistent Cancer-Related Fatigue (NCT05256888), a randomized na study. The primary outcome is Fatigue, as Assessed Using the Functional Assessment of Chronic Illness Therapy-Fatigue (FACIT-F) Questionnaire.

2. Clinical Question

The trial also aimed to assess whether the intervention could be delivered and whether participants would adhere to it and complete study activities. It proposed exploring relationships between fatigue, circadian rhythm, and glucose metabolism. The posted outcome information does not report results for those feasibility or exploratory aims, so the week-12 fatigue data should not be treated as answering them.

3. Trial Design

FREDA was a randomized, parallel-group behavioral trial with no masking. The registry lists an enrollment of 50 and a 1:1 allocation plan between time-restricted eating and control. The study is marked completed. Its start date was 2023-01-26 and its primary completion date was 2024-06-01. The phase is listed as NA, which does not provide a phase classification.

The intervention lasted 12 weeks. Participants assigned to time-restricted eating selected a 10-hour window for consuming food and beverages, with water allowed at any time and black coffee and unsweetened tea allowed in the morning. The control group did not have that eating-window instruction. Both groups received weekly healthy lifestyle tips relevant to cancer survivorship. The trial is sponsored by the University of Maryland, Baltimore.

The primary outcome was the FACIT-F fatigue subscale score at 12 weeks, with baseline score specified as a covariate. The registry describes a comparison of group scores controlling for baseline levels. This distinction matters when reading the posted estimates: the displayed group means and the simple recalculation from them are unadjusted summaries, whereas the registry's reported analysis is a linear regression.

Outcome summaries are posted for 11 participants in the time-restricted eating group and 14 in the control group. The registry does not explain in the outcome information why these outcome group sizes differ from the planned enrollment or allocation. The reasons for any unavailable outcome data are not reported in the registry record. The record lists two outcome measures but posts no secondary outcome analyses.

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: Participants must have completed baseline assessments, including the FACIT-F questionnaire, in order to be randomized. There were 20 people who consented to the study but did not complete baseline assessments and therefore were not randomized to an arm.

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 Fatigue, as Assessed Using the Functional Assessment of Chronic Illness Therapy-Fatigue (FACIT-F) Questionnaire, measured over 12 weeks.

The primary endpoint is fatigue measured with the Functional Assessment of Chronic Illness Therapy-Fatigue questionnaire, abbreviated FACIT-F. The questionnaire has 13 items, and the fatigue subscale ranges from 0 to 52. Higher scores indicate less fatigue. Thus, an increase in score represents a direction associated with less fatigue on this measure; it should not be described as more fatigue merely because the numerical score is higher.

The registry specifies the group comparison at 12 weeks and says baseline fatigue levels will be controlled for. Taking baseline severity into account can help distinguish the week-12 group contrast from pre-existing differences in fatigue. The reported endpoint is a continuous score, summarized for each group by a mean and standard deviation. A mean describes the group's average score, while the standard deviation describes variation among individual scores around that average.

FACIT-F is used in this trial because the research question concerns persistent cancer-related fatigue, and the chosen scale provides a numerical measure that can be compared between randomized groups. However, the posted summary statistics alone do not describe every participant's experience, the course of fatigue during the 12 weeks, or how scores relate to other clinical or daily-life outcomes. The registry record does not report a threshold for interpreting a difference as clinically important. Statistical evidence about a group contrast therefore needs to be distinguished from a judgment about its practical importance.

7. Results Posted in the Registry

• Time-restricted Eating: mean 38.70 versus 37.20 on control; difference 1.50 (95% CI -5.73 to 8.73), standardized effect d = 0.16; z-test p = 0.684.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous, so its group summaries are means and standard deviations. A mean locates the center of the observed scores for a group. A standard deviation conveys how dispersed those scores are around the mean. These summaries are descriptive: they show the observed endpoint distributions at a broad level, but do not by themselves adjust for baseline scores or other variables.

What does the unadjusted mean difference represent?

The recalculated mean difference of 1.5 is the time-restricted eating mean minus the control mean, using the posted week-12 group means. It is unadjusted: it does not control for baseline fatigue. As higher FACIT-F scores indicate less fatigue, a positive difference means the intervention group's observed average score was higher. This calculation is useful as a transparent summary of the posted means, but it is not the same estimand as the registry's baseline-controlled primary comparison.

The recalculation also uses summary statistics rather than participant-level observations. It cannot account for individual baseline scores, the relation between baseline and week-12 scores, or any covariates included in a fitted regression unless those details are available. The registry record reports a linear regression slope for the primary analysis, but its model specification and confidence interval are not reported. Consequently, the regression estimate cannot be independently reproduced from the group means and standard deviations alone.

What does a confidence interval tell the reader?

A confidence interval describes uncertainty around an estimated quantity under the assumptions of the method used to calculate it. For the unadjusted recalculation, the interval from -5.731 to 8.731 includes both negative and positive differences. Because the interval includes zero, the data do not rule out no unadjusted difference at the conventional two-sided significance level used in the planning illustration. More importantly, the width indicates that the estimate is not very precise: a variety of underlying differences remain compatible with the summary data and calculation assumptions.

A confidence interval is not a range that guarantees the true effect lies within its endpoints, nor does it express the probability that a particular value is correct. It is best read as a range of values reasonably compatible with the data under the statistical procedure. The registry's reported regression result does not include a confidence interval, so its precision cannot be assessed from the posted regression output as fully as the recalculated unadjusted contrast.

Why can baseline adjustment change the estimate?

Baseline adjustment compares outcomes while accounting for starting fatigue levels. If baseline scores are related to week-12 scores, adjustment can change the estimated group contrast and its precision. It can also address chance baseline imbalances in a randomized trial, though the posted record does not report the baseline group summaries needed to assess that issue. The sponsor-reported slope of -1.77 and p = 0.11 come from a linear regression, while the recalculated difference of 1.5 and p = 0.684 come from an unadjusted comparison of posted summary statistics. They answer related but not identical statistical questions. Their differing estimates should not be treated as a direct contradiction.

What assumptions matter for a continuous outcome?

Methods for means and linear regression generally rely on assumptions about how observations are related, how residual variation behaves, and whether the model represents the outcome relationship adequately. General statistical explanation: when variability is substantially greater than a model assumes, standard errors can be too small and intervals too narrow; this is one reason analysts consider the distribution and variance structure of the data. The registry record does not report diagnostic checks, robust variance methods, transformations, or other details that would show how these assumptions were assessed.

Randomization supports a comparison between assigned groups, but the posted endpoint summaries do not reveal adherence, departures from assignment, or the handling of missing endpoint values. The analysis population and missing-data method are not reported in the registry record. These details affect how broadly the estimate can be interpreted and how closely it reflects the planned randomized comparison.

How should the p-values be read?

The recalculated p-value of 0.684 relates to the unadjusted group contrast, while the reported p-value of 0.11 relates to the registry's linear regression slope. A p-value measures how incompatible the observed result is with a specified null model, under the analysis assumptions; it is not the probability that the null hypothesis is true. Neither value describes the size or practical importance of the difference. Interpretation should therefore consider the estimate, its uncertainty, the endpoint scale, and the analysis method together. The registry record does not report a multiplicity adjustment or a prespecified strategy for interpreting several outcomes.

10. Confidence Intervals and Precision

The unadjusted recalculated mean difference is 1.5 FACIT-F points. Its 95% confidence interval runs from -5.731 to 8.731. This range is wide enough to include a difference favoring either group, as well as zero. The posted data therefore leave considerable uncertainty about the size and direction of the unadjusted contrast. Since higher scores indicate less fatigue, positive values correspond to a higher average score in the time-restricted eating group, while negative values correspond to a higher average score in control.

Precision depends on both the estimated variability and the amount of information available for the comparison. The group standard deviations are 9.2 and 9.1, and the outcome summaries are based on 11 participants in the time-restricted eating group and 14 in control. The standard errors reported for the individual group means are 2.774 and 2.432, respectively. These measures help convey why the group means should not be read as exact estimates of the underlying group averages.

The interval pertains to an unadjusted recalculation from posted summary statistics. It is not the confidence interval for the trial's baseline-adjusted regression, for which the registry record does not report interval limits. The registry's regression slope of -1.77 with p = 0.11 is therefore not accompanied by the same direct precision information. Nor does the unadjusted interval answer questions about adherence, feasibility, or outcomes not posted in the registry record.

Statistical uncertainty should not be mistaken for proof of no difference. A broad interval can be consistent with multiple possibilities, and the posted data do not identify one precise effect. The registry record does not report a threshold for a clinically important FACIT-F difference, so the interval cannot be classified against such a threshold using the posted information. The result is best described as an imprecise estimate, with uncertainty that should remain central to its interpretation.

11. 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 figures are participants per arm, not total events. They illustrate how many participants in each arm would be required under the stated planning assumptions if the target were to detect a specified fraction of the observed effect. The assumptions are a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. The trial's own original design assumptions are not reported in the registry record.

For planning around 100% of the observed effect, the illustration gives 584 participants per arm for 80% power and 781 participants per arm for 90% power. For 75% of the observed effect, it gives 1,037 participants per arm for 80% power and 1,389 participants per arm for 90% power. For 50% of the observed effect, it gives 2,334 participants per arm for 80% power and 3,124 participants per arm for 90% power.

The 80% power column describes the planning sample size associated with an 80% chance of detecting the assumed effect under the calculation assumptions, if that effect were the true effect and the model assumptions held. The 90% column makes the analogous illustration for a 90% chance. These are design-planning concepts, not probabilities that the observed result is correct. The figures are not observed power.

Smaller target effects require more participants because they are harder to distinguish from random variation. The observed effect used as the planning reference comes from the posted group summaries, so these calculations inherit uncertainty in that estimate. They are illustrative rather than a reconstruction of the trial's original planning. The registry record does not report the trial's sample-size justification or original power assumptions.

12. 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 = 1.500; pooled standard deviation σ = 9.144.
2. n = 2 × (1.96 + 0.842)² × 9.144² / 1.500² = 583.3, rounded up to 584 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 <- 1.5                      # observed difference in means
sigma <- 9.14361264347484                     # 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(5256888)
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 = 584 (80% power), 781 (90% power)
 75% of observed effect: n per arm = 1,037 (80% power), 1,389 (90% power)
 50% of observed effect: n per arm = 2,334 (80% power), 3,124 (90% power)
power.t.test: 584.3 
Simulated power at n = 584 per arm: 0.801 (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.

13. Analyses Reported by the Sponsor

The registry reports a linear regression analysis for the primary comparison, with a slope of -1.77 and p = 0.11. The endpoint description says that the week-12 FACIT-F comparison was intended to control for baseline fatigue. No confidence interval for the regression slope is reported, and the registry record does not give enough model detail to reproduce this analysis from the posted group summaries.

The recalculated mean difference of 1.5, its 95% confidence interval from -5.731 to 8.731, and p = 0.684 are simple unadjusted calculations from the posted week-12 means and variability. They do not control for baseline. Thus, the recalculation is not a replacement for the trial's reported regression analysis and should not be read as if both estimates came from the same model.

The slope and the unadjusted difference have different analysis bases, and the record does not specify the regression coding sufficiently to interpret the slope's direction independently. They should be presented as separate results: the reported regression is the trial's stated primary analysis, while the recalculation is an unadjusted check based on summary statistics. The registry record marks the discrepancy indicator as false, but the distinction between adjusted and unadjusted analyses remains important for reading the estimates.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists two outcome measures, but no secondary outcome analyses are posted. The primary endpoint is identified as week-12 FACIT-F fatigue, and the registry description specifies controlling for baseline levels. The record does not report whether the trial used a formal multiplicity strategy to account for testing more than one outcome, nor does it describe any hierarchy for interpreting outcomes. Without that information, the posted primary result should be considered on its own rather than treated as part of a fully documented family of tests.

The registry record does not report the analysis population used for the primary result. It does not state whether participants were analyzed according to randomized assignment, whether any per-protocol or other populations were used, or how participants with unavailable week-12 scores were handled. Reasons for the difference between the planned enrollment and the outcome group sizes are not reported in the registry record. These omissions make it difficult to determine whether the posted summaries represent all randomized participants or a subset defined by data availability or another analysis rule.

Missing-data methods can matter because the estimate may be affected when outcome availability differs across participants or relates to fatigue. General statistical explanation: complete-case analysis uses participants with observed outcomes, whereas imputation or model-based methods use assumptions to address missing values. The registry record does not say which, if any, approach was used. It also does not report sensitivity analyses examining how conclusions might change under alternative assumptions about missing outcomes.

The registry does not report interim analyses, stopping rules, or a detailed statistical analysis plan. Those items should not be inferred from the posted p-values or from the study's completed status. As a result, the available record supports a description of the reported endpoint result, but not a full evaluation of the trial's multiplicity control, analysis population, or missing-data robustness.

15. 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.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: shortness of breath (6.7%), pneumonia (6.7%).

The serious adverse event summaries report 0 affected participants in the time-restricted eating group and 2 in the control group. The corresponding numbers at risk are 15 in each group. The registry lists shortness of breath and pneumonia among the top serious event terms, each at 6.7%. No deaths are listed in either group.

For other adverse events, the record reports 1 affected participant in the time-restricted eating group and 0 in control, with 15 at risk in each group. These figures are descriptive counts in the registry record. They do not by themselves establish whether an event was caused by an intervention, nor do they provide a statistical comparison of safety between groups.

The serious adverse event summaries and the primary fatigue outcome have different group counts: the safety record lists 15 at risk per group, while the posted fatigue summaries are based on 11 participants in the intervention group and 14 in control. The registry record does not explain this difference. The event descriptions do not provide timing, severity details beyond the serious classification, exposure duration, or causal assessments. Therefore, the counts should be interpreted as the posted safety reporting, not as a complete causal assessment or a basis for comparing risk precisely.

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

The posted results show mean FACIT-F scores of 38.7 for time-restricted eating and 37.2 for control at 12 weeks. On this scale, higher scores indicate less fatigue. The unadjusted recalculated contrast is 1.5 points, with a 95% confidence interval from -5.731 to 8.731. The interval spans values in both directions and includes zero, so the posted summary data do not provide a precise estimate of the unadjusted group difference.

The registry's reported linear regression analysis gives a slope of -1.77 and p = 0.11. The outcome description says the intended comparison controls for baseline fatigue, but the record does not give the regression confidence interval or enough detail to reproduce the model. The unadjusted calculation and the reported regression should not be treated as identical analyses.

These results do not establish that time-restricted eating reduces fatigue, and they do not establish that it has no effect. A non-small p-value is not evidence that groups are equivalent. Equivalence would require an appropriate design and a prespecified margin, neither of which is reported in the registry record. The record also does not report a threshold for a clinically important FACIT-F difference, so the numerical contrast cannot be labeled clinically important or unimportant from these data alone.

The stated feasibility aims include recruitment, adherence, and completion of study activities, but results for those aims are not reported in the registry record. The planned exploration of fatigue, circadian rhythm, and glucose metabolism is also not represented by posted secondary analyses. Consequently, the primary endpoint summaries cannot resolve those questions. They describe a limited week-12 comparison, with uncertainty, rather than a comprehensive account of intervention feasibility, longer-term outcomes, or underlying mechanisms.

17. Why This Trial Matters Statistically

This trial illustrates why the analysis method needs to be read alongside the endpoint description. The primary question specifies a baseline-controlled comparison, while a simple calculation from the posted group means is unadjusted. The resulting estimates are not interchangeable: one is the reported regression slope, and the other is a transparent contrast of the observed week-12 averages. Keeping those estimands distinct prevents a summary calculation from being mistaken for the trial's primary model.

It also shows why uncertainty matters when interpreting a continuous clinical outcome. The unadjusted interval extends from -5.731 to 8.731 around a mean difference of 1.5. That range includes different directions of effect and does not locate a narrow set of plausible values. Reporting only the group means or a p-value would conceal this imprecision. The FACIT-F direction also matters: higher scores indicate less fatigue, so statistical signs require interpretation in the context of the scale.

The planning illustration reinforces a broader design principle. When a target effect is smaller, more participants are needed to detect it under otherwise stated assumptions. However, a planning calculation based on an observed effect is inherently uncertain, especially when the outcome summaries are limited. It should not be treated as evidence that the completed trial had a particular probability of detecting its true effect.

Finally, a registry record can leave important interpretive questions open. Here, details on missing outcomes, analysis populations, multiplicity, and feasibility results are not reported. Recognizing those gaps is part of statistical interpretation: it limits claims to what the posted endpoint and safety summaries can support, without filling in unreported methods or outcomes.

18. 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 endpoint group summaries are based on 11 participants in the time-restricted eating group and 14 in control, while the registry lists enrollment of 50. The record does not explain this difference, and the reasons for unavailable endpoint data are not reported in the registry record. The analysis population and missing-data procedures are also not reported, limiting assessment of whether the endpoint results reflect the randomized groups as planned.

The primary outcome is measured at 12 weeks, and the posted information does not provide a longer-term fatigue comparison. Although the registry describes a feasibility objective and planned exploration of circadian rhythm and glucose metabolism, corresponding results are not reported in the registry record. Nor does the record provide adherence outcomes that would indicate how closely participants followed the selected eating window.

The reported primary regression has a slope of -1.77 and p = 0.11 but no confidence interval. Its model specification is not reported, which prevents independent reconstruction and limits assessment of precision. The recalculated unadjusted contrast has an interval, but it does not answer the baseline-adjusted question specified for the primary endpoint.

The registry does not report a clinically important difference threshold for FACIT-F, multiplicity procedures, or detailed analysis assumptions. The safety summaries are counts and event terms, without enough information to establish causality or compare risks precisely. These limitations mean the posted results are best understood as a preliminary, incomplete statistical record of the week-12 outcome rather than a full account of efficacy, feasibility, or safety.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Time-Restricted Eating to Address Persistent Cancer-Related Fatigue — NCT05256888.
• Registry record last updated 2026-07-20; status last verified 2026-07.
• Results first posted 2025-05-07.
• 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.

22. Record Summary

NCT05256888, titled the Fatigue REDuction After Cancer (FREDA) Trial, is a completed randomized, parallel-group, unmasked behavioral study sponsored by the University of Maryland, Baltimore. The registry lists an enrollment of 50, a start date of 2023-01-26, and a primary completion date of 2024-06-01.

The study compared a 12-week time-restricted eating intervention with weekly healthy lifestyle education. Its primary endpoint was FACIT-F fatigue score at 12 weeks, with baseline levels specified for the comparison. Higher FACIT-F scores indicate less fatigue. The posted group means were 38.7 for intervention and 37.2 for control. The unadjusted recalculation gives a mean difference of 1.5, with a 95% confidence interval from -5.731 to 8.731 and p = 0.684. The reported linear regression slope was -1.77 with p = 0.11.

Reasons for outcome group sizes, missing-data handling, analysis populations, and feasibility results are not reported in the registry record. Interpretation should retain the distinction between the unadjusted recalculation and the reported regression.