FRESH Study (Fitness, Relaxation and Eating to Stay Healthy): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for FRESH Study (Fitness, Relaxation and Eating to Stay Healthy) (NCT01636258), a randomized na study. The primary outcome is Effect of "FRESH" Program on Weight Loss.

2. Clinical Question

The study’s brief summary describes a pilot randomized controlled trial intended to examine whether a lifestyle program could result in weight loss and thereby delay or prevent progression from prediabetes to diabetes. However, the primary outcome posted in the registry record is weight change. A result for weight change alone does not answer whether progression to diabetes was delayed or prevented. The registry record also lists blood sugars, cholesterol, waist circumference, blood pressure, diet, physical activity, perceived stress, and class attendance among the proposed outcomes. The posted primary result addressed weight, while four secondary outcomes are listed as posted. Thus the clinical question addressed by the numerical comparison is narrower than the broader motivation described in the study summary.

3. Trial Design

The FRESH Study (Fitness, Relaxation and Eating to Stay Healthy) was registered as a randomized, parallel-group study with no masking. Its status is completed. The registry record reports enrollment of 27 participants and identifies prediabetes and overweight as the conditions under study. The sponsor was The Cleveland Clinic. The study started in 2012-05, and primary completion was recorded in 2012-10.

Participants were assigned to usual care or to a behavioral program combining diet, exercise, stress management, and culinary education. The control arm was described as continuing usual care from each participant’s primary medical care team. The intervention arm was described as receiving the four lifestyle components. The record does not report the randomization sequence generation method, allocation concealment, or whether assignment was stratified by any participant characteristic. With no masking, participants and study personnel were not masked according to the registry classification; details about who measured outcomes and whether assessors knew assignments are not reported in the registry record.

The brief summary describes the proposed study as a pilot trial of 6 weeks duration with 30 participants. The registry record’s reported enrollment is 27, and the primary outcome timeframe is baseline to a final follow-up visit at 8-14 weeks. These entries describe different aspects of the record: a proposed duration and enrollment in the summary, and the posted enrollment and endpoint follow-up timeframe. The record does not explain the timing difference or report the protocol’s original sample-size justification. The posted analysis compares mean weight change between the intervention and control groups. The summary mentions an intention-to-treat analysis of variance, while the reported sponsor analysis is a two-sided t-test. Details about implementation of the proposed analysis are not reported in the registry record.

4. Participant Flow and Accounting

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

Before assignment: Two participants who were enrolled in study were excluded at baseline laboratory check due to glycated hemoglobin (HgbA1c) value below 5.7%.

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 Effect of "FRESH" Program on Weight Loss, measured over Baseline line and final followup visit (at 8-14 weeks).

The primary endpoint is change in weight, measured in kilograms from baseline to the final follow-up visit. The registry describes the outcome as a mean with standard deviation and gives the follow-up timeframe as 8-14 weeks. The endpoint therefore summarizes the average change in weight within each randomized group over the stated observation period. The posted control mean was -0.5 kg, with a standard deviation of 2.0, among 12 participants. The intervention mean was -0.2 kg, with a standard deviation of 2.2, among 14 participants.

Weight is used as a direct, continuous measure related to the study’s stated interest in weight loss. A continuous endpoint retains information about the magnitude of change, rather than reducing the result to whether a participant crossed a single threshold. The group means describe average change, while the standard deviations describe how spread out individual changes were around those means. Neither quantity alone describes every participant’s outcome.

The registry summary also raises the longer-term question of progression from prediabetes to diabetes, but the primary endpoint reported is weight change, not diabetes progression. Consequently, the posted primary comparison evaluates average weight change over the registered follow-up timeframe; it does not directly quantify a change in diabetes incidence or establish whether progression was delayed or prevented. Further details about how weight was measured, including equipment, measurement procedures, and assessor training, are not reported in the registry record.

7. Results Posted in the Registry

• Arm B: Intervention Group: mean -0.20 versus -0.50 on control; difference 0.30 (95% CI -1.32 to 1.92), standardized effect d = 0.14; z-test p = 0.716.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary measure is continuous: weight change in kilograms. For a continuous measure, a group mean provides a concise description of the average observed value, and a standard deviation describes variation among individual values around that average. The control mean was -0.5 kg and the intervention mean was -0.2 kg. The respective standard deviations were 2.0 and 2.2. These summaries do not show the full distribution of changes, whether there were unusually influential observations, or how individual participants’ measurements were related across time.

What does the mean difference represent?

The sponsor reports a mean difference (net) of 0.3 kg. It represents the intervention-versus-control contrast in the average change, using the comparison as reported. A value of zero would correspond to equal group means under that contrast. The estimate alone is not enough to judge how stable the comparison is; the confidence interval adds information about uncertainty. The result is an estimate for these observed groups and follow-up, rather than a statement that every participant experienced the group-average change.

What does the t-test assess?

The sponsor’s reported method is a two-sided t-test. In general, a two-sided test evaluates departures from the null in either direction: it does not restrict the question to whether the intervention mean is greater or less than the control mean. A t-test compares a mean difference with the uncertainty expected under a model in which the group means are equal. Its p-value describes how unusual a test statistic at least as extreme as the observed one would be if that null model and the test assumptions were correct. It is not the probability that the null hypothesis is true.

The sponsor reports p = 0.72. This value is consistent with the broad confidence interval around the estimate: the data do not sharply locate the mean difference relative to zero. The p-value should be interpreted with the estimated effect and its interval, not in isolation. It does not measure the size or clinical importance of the observed difference.

Why is the confidence interval useful?

A confidence interval gives a range of estimates compatible with the observed data under the analysis method and its assumptions. In repeated use of the same interval procedure under its assumptions, the stated confidence level describes the long-run coverage of the procedure. It does not mean there is a 95% probability that a fixed, unknown effect lies inside this particular interval. For this result, the sponsor’s interval ranges from -1.4 to 2.0 kg. Its inclusion of zero indicates that the data are compatible with no mean difference, while its width also leaves room for differences in either direction.

What does the recalculation add?

The unadjusted recalculation uses the posted group sizes, means, and standard deviations to estimate the mean contrast and its uncertainty. It produces a mean difference of 0.3 kg, a 95% confidence interval from -1.315 to 1.915, and p = 0.716. Because it is derived from summary statistics, it cannot reproduce details that would require participant-level information, such as handling of missing measurements, covariate adjustment, or the exact analysis population. No adjustment is represented in this recalculation. The registry does not report whether any such features were part of the sponsor’s analysis beyond its stated t-test.

The summary describes an intention-to-treat analysis of variance. Intention-to-treat generally means that participants are analyzed according to their randomized assignment, regardless of adherence to the assigned program. Analysis of variance is a general method for comparing group means. However, the result reported in the analyses is a two-sided t-test, and the record does not explain whether this test was the final implementation of the stated intention-to-treat plan, whether any participants were excluded, or how missing outcome data were addressed. The analysis population and missing-data methods are not reported in the registry record.

What can pilot estimates tell readers?

A pilot study can describe feasibility and provide preliminary estimates for future planning, but an effect estimate from a small study is often imprecise. The observed mean difference is one estimate among many that could arise from random variation. A wide interval communicates that imprecision more directly than a single p-value. The registry summary says that results would be used to help design a larger randomized trial; it does not report the later design decisions or a formal sample-size justification for this study.

The planning illustrations use the observed effect as a reference and show how participant requirements change when the target effect is set at different stated fractions of that observed effect. Those calculations assume a two-sided significance level of 0.05, equal allocation, and a normal approximation. They are illustrative planning calculations, not a retrospective measure of the chance that this trial detected an effect. Their values should not be interpreted as the number of participants actually needed under every possible future design: different outcomes, assumptions, analysis methods, or target effects would change a planning calculation.

Weight change is a continuous endpoint, but the broader list of outcomes includes measures with different scales and potentially different distributions. General statistical considerations depend on the outcome’s measurement and distribution. For example, over-dispersion—more variability than a chosen statistical model expects—can make standard errors too small if the model does not account for it. That is a general concept, not a reported feature of this trial. Distribution checks, model diagnostics, and any adjustments used for the outcomes are not reported in the registry record.

10. Confidence Intervals and Precision

The sponsor’s 95% confidence interval for the mean difference is -1.4 to 2.0 kg, around a reported estimate of 0.3 kg. This interval expresses substantial uncertainty about the size and direction of the group contrast. Values below zero and above zero are both included, so the result is compatible with different underlying mean differences under the assumptions of the analysis. The interval does not mean that all values within it are equally plausible, and it is not a range for individual participants’ weight changes.

The interval is broad relative to the point estimate, which reflects limited precision in the comparison. The posted group standard deviations, 2.0 kg in control and 2.2 kg in intervention, describe participant-level spread, while the confidence interval reflects uncertainty in the estimated group contrast. These are different quantities. The standard deviations should not be read as confidence limits, and the confidence interval should not be read as the range of changes experienced by most participants.

The recalculated interval, -1.315 to 1.915 kg, is close to the sponsor’s interval of -1.4 to 2.0 kg. The difference in displayed endpoints is consistent with the use of differently rounded reported values and analysis details; the registry record does not give enough information to attribute it to a particular cause. Both intervals include zero and convey considerable uncertainty. The sponsor’s result is the trial’s reported analysis, while the recalculation is a simple unadjusted check based on summary statistics.

Precision is also limited by the number of participants contributing to the posted group summaries: 12 in control and 14 in intervention. The registry does not report an original sample-size justification. The confidence interval is therefore especially important for avoiding an overconfident reading of the point estimate. A result that does not rule out zero is not proof of no effect, just as the positive point estimate is not proof of a meaningful difference.

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 are illustrations based on the observed effect, not the trial’s original design assumptions. The listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect targeted in each planning scenario is a stated fraction of the observed effect. The original sample-size justification and the design assumptions used when the trial was planned are not reported in the registry record.

The 80% power column gives the calculated participants per arm for a design with an 80% chance of detecting the specified effect under the planning assumptions, if that effect and those assumptions were correct. The 90% power column gives the corresponding per-arm requirement for a 90% chance. These are design targets, not guarantees about what a future trial will find. The calculation does not indicate the power achieved by the completed study.

For planning at 100% of the observed effect, the illustration gives 778 participants per arm for 80% power and 1,041 for 90% power. Planning for 75% of the observed effect gives 1,382 participants per arm for 80% power and 1,850 for 90% power. Planning for 50% of the observed effect gives 3,109 participants per arm for 80% power and 4,161 for 90% power. Smaller target effects require more participants because they are harder to distinguish from random variation. The effect fractions refer to planning scenarios, not separate observed results.

These figures are not observed power. Observed power is not a useful substitute for reading the estimated effect and its confidence interval after a study is complete. The calculations are best read as illustrations of how demanding future planning may be if the observed estimate is used as a reference. The registry record does not report a chosen target effect or original sample-size rationale for the pilot.

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 = 0.300; pooled standard deviation σ = 2.111.
2. n = 2 × (1.96 + 0.842)² × 2.111² / 0.300² = 777.0, rounded up to 778 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 <- 0.3                      # observed difference in means
sigma <- 2.11068709192054                     # 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(1636258)
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 = 778 (80% power), 1,041 (90% power)
 75% of observed effect: n per arm = 1,382 (80% power), 1,850 (90% power)
 50% of observed effect: n per arm = 3,109 (80% power), 4,161 (90% power)
power.t.test: 778 
Simulated power at n = 778 per arm: 0.803 (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 sponsor reports a two-sided t-test with a mean difference (net) of 0.3, a 95% confidence interval from -1.4 to 2.0, and p = 0.72. A simple unadjusted recalculation from the posted means, standard deviations, and group sizes gives a mean difference of 0.3, a 95% confidence interval from -1.315 to 1.915, and p = 0.716. The point estimates agree, and the interval endpoints and p-values are close.

The recalculation is a check based on summary statistics, rather than a reconstruction of every possible feature of the sponsor’s analysis. It does not incorporate unreported participant-level data, covariates, missing-data procedures, or other model details. The registry record does not explain the small differences in the displayed interval endpoints and p-values, so no specific explanation can be assigned. The sponsor’s reported analysis is the official trial result; the recalculation is useful as a transparent, unadjusted comparison and does not replace it.

The brief summary states that an intention-to-treat analysis of variance was planned, whereas the posted analysis is identified as a two-sided t-test. The registry record does not explain this difference in method descriptions. The sponsor’s numerical result should therefore be read as the analysis actually posted, while the brief-summary language indicates the proposed analysis without establishing how it was implemented.

14. Multiplicity, Missing Data and Analysis Populations

The registry record lists a primary weight-change outcome and four secondary outcomes as posted. The broader study summary mentions additional intended measurements, including blood sugars, cholesterol, waist circumference, blood pressure, diet, physical activity, perceived stress, and class attendance. The record also reports five outcome measures and four analyses. It does not identify the relationship between each listed outcome and each analysis in enough detail to establish a complete analysis map.

When a study examines multiple outcomes or makes multiple statistical comparisons, the chance of at least one nominally unusual result can increase. General methods for addressing multiplicity include prespecifying a primary comparison, adjusting inference across a defined family of tests, or treating secondary results as exploratory. The registry record does not report whether multiplicity was considered, whether any adjustment was made, or which secondary comparisons were prespecified. It is therefore not possible to determine from the posted information how multiplicity was handled.

The brief summary says analysis would be by intention-to-treat analysis of variance. Intention-to-treat generally retains participants in the groups to which they were randomized for analysis. However, the posted primary comparison reports group sizes of 12 and 14, while the registered enrollment is 27. The record does not explain the difference between enrollment and the group sizes used for the posted weight summaries, nor does it state whether all randomized participants contributed to the primary analysis. The analysis population and the handling of participants without a follow-up weight are not reported in the registry record.

Missing data can affect a comparison if outcome availability is related to assignment, participant characteristics, or the outcome itself. General approaches may include analyses based on observed values, imputation, or models using available repeated measurements, but the registry record does not identify a method used for this endpoint. The amount and pattern of missing outcome data, any imputation, and any sensitivity analyses are not reported. These omissions limit assessment of whether the posted comparison fully reflects the randomized groups.

15. Safety Results

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

Other (non-serious) adverse events

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

The registry reports zero participants affected by serious adverse events in the control group, among 13 at risk, and zero in the intervention group, among 14 at risk. It also reports zero participants affected by other adverse events in the control group, among 13 at risk, and zero in the intervention group, among 14 at risk. These are participant counts as recorded for the adverse-event categories, not rates calculated from follow-up time.

The adverse-event at-risk counts are not the same as the group sizes shown for the primary weight summaries: the latter are 12 in control and 14 in intervention. The registry record does not explain the difference between the control-group counts. It also does not report deaths in the adverse-event record. A blank or unreported death value should not be interpreted as a documented count of zero.

Zero affected participants in the posted categories means no serious or other adverse events were recorded for the participants counted in those fields. It does not demonstrate that no event of any kind occurred, and it does not quantify the ability of this pilot to detect uncommon events. The record does not provide event descriptions or additional safety follow-up details. These figures should be read as the recorded adverse-event counts, with the stated denominators and reporting limits.

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

The primary result is a comparison of mean weight change over the registry’s stated baseline-to-final-follow-up timeframe of 8-14 weeks. The control mean was -0.5 kg and the intervention mean was -0.2 kg. The sponsor’s reported mean difference was 0.3 kg, with a 95% confidence interval from -1.4 to 2.0 and p = 0.72. The point estimate is small relative to the uncertainty expressed by the interval, which includes zero and values in either direction.

The result does not establish that the two groups had exactly the same average change. It also does not establish a precise difference favoring either group. The confidence interval shows why the p-value alone is incomplete: the estimate remains uncertain, and the data are compatible with a range of underlying group contrasts. The study’s enrollment and pilot purpose should also be kept in view when considering how much confidence to place in a single estimate.

The registry summary gives a broader rationale involving possible delay or prevention of progression from prediabetes to diabetes. That is not the primary outcome reported in the posted comparison. The reported weight result therefore cannot establish whether diabetes progression was affected. Nor does the record’s description of a multi-part lifestyle program allow the result to isolate the contribution of diet, exercise, stress management, or culinary education separately.

The result is specific to the participants enrolled, the comparison with usual care, the reported weight endpoint, and the stated follow-up. It does not determine outcomes beyond that observation period. It does not show how the program would perform in a larger or different population, and no such extrapolation is established by the registry result. The study was described as a pilot intended to help design a larger randomized trial; the posted estimate is best viewed as preliminary and imprecise rather than as a definitive answer to the broader clinical motivation.

In short, the posted data give a modest positive estimate for the intervention-versus-control mean contrast, but with substantial uncertainty. The sponsor’s test did not provide strong evidence against a zero mean difference. That is a statement about the statistical evidence in this trial, not a medical recommendation or a conclusion about what an individual should do.

17. Why This Trial Matters Statistically

This trial is statistically informative because it illustrates the distinction between an estimated effect and the precision of that estimate. The reported mean difference is 0.3 kg, but the sponsor’s 95% confidence interval extends from -1.4 to 2.0 kg. A single point estimate can sound definitive when presented alone; the interval shows that the data leave considerable uncertainty about the group contrast.

The study is explicitly described as a pilot, and its results were intended to help design a larger randomized trial. Pilot results can help identify the scale of variability and provide preliminary estimates, but small studies commonly yield wide intervals. Using the observed effect as the basis for planning can also be challenging: the observed estimate may differ from the underlying effect, and the registry’s planning illustrations show that requirements depend strongly on the target effect chosen. Smaller target effects require more participants under the stated assumptions.

The record also shows why a statistical plan should be distinguishable from the analysis ultimately reported. The brief summary describes an intention-to-treat analysis of variance, while the posted analysis is a two-sided t-test. Without more detail, readers cannot establish how the planned approach was implemented or how missing measurements and secondary comparisons were handled. Clear reporting of analysis populations, missing-data procedures, and multiplicity decisions would make the result easier to assess.

Finally, the primary endpoint and the study’s broader motivation are not interchangeable. Weight change is the posted primary outcome, while the summary mentions prevention or delay of diabetes as a motivation. A trial can motivate a future question without answering it through the endpoint reported. Keeping those distinctions clear helps readers avoid making claims beyond the scope of the observed comparison.

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 registry record gives limited detail about eligibility, participant characteristics, recruitment, randomization procedures, and allocation concealment. The study is classified as unmasked, but the record does not clarify whether outcome assessors knew assignment. These omissions restrict assessment of how participants entered the trial and how assignment and measurement were implemented.

The reported enrollment is 27, while the brief summary describes a proposed 30 participants. The posted primary outcome summaries use 12 participants in control and 14 in intervention. The record does not explain the difference among these counts, identify the number with missing follow-up weight, or clarify which participants were included in the analysis. The adverse-event counts also use at-risk denominators that are not identical to the posted primary outcome group sizes. The reasons for these differences are not reported in the registry record.

The summary describes a 6 weeks duration, while the primary outcome timeframe is stated as baseline and a final follow-up visit at 8-14 weeks. The record does not explain how these time descriptions relate to the intervention schedule or the timing of outcome collection. It also does not provide the detailed intervention curriculum, adherence information, or the content of usual care. These details would matter for interpreting how the program was delivered and what comparison was made in practice.

The registry does not report the original sample-size justification, the detailed statistical plan, the missing-data method, or any multiplicity adjustment. Although the brief summary names an intention-to-treat analysis of variance, the posted result is a two-sided t-test; the relationship between the planned and reported methods is not explained. The primary estimate is also imprecise, and the pilot enrollment limits the ability to draw firm conclusions from the group comparison.

Lastly, the primary endpoint is weight change, not diabetes progression. The record’s wider prevention rationale should not be mistaken for an observed result on that outcome. Follow-up is described only for the primary endpoint timeframe, so longer-term outcomes are not established by the posted primary comparison.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: FRESH Study (Fitness, Relaxation and Eating to Stay Healthy) — NCT01636258.
• Registry record last updated 2022-05-10; status last verified 2022-05.
• Results first posted 2014-06-09.
• 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

The FRESH Study was a completed randomized, parallel-group, unmasked pilot study sponsored by The Cleveland Clinic. It enrolled 27 participants with prediabetes and overweight. Participants were assigned to usual care or a program combining diet instruction, exercise, stress management, and culinary education. The study started in 2012-05 and primary completion was recorded in 2012-10.

The primary endpoint was mean change in weight in kilograms from baseline to final follow-up, described as 8-14 weeks. The control mean was -0.5 kg among 12 participants; the intervention mean was -0.2 kg among 14 participants. The sponsor reported a mean difference of 0.3 kg, with a 95% confidence interval from -1.4 to 2.0 and p = 0.72. This estimate is imprecise and does not establish an effect on diabetes progression, which is described as part of the study’s motivation but is not the posted primary endpoint.

The study was intended as a pilot to inform future trial design. Details about the original sample-size rationale, missing-data handling, multiplicity, and the analysis population are not reported in the registry record.