Exercise After Clinically Significant Weight Loss: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Exercise After Clinically Significant Weight Loss (NCT03685123), a randomized na study. The primary outcome is Change in Weight.

2. Clinical Question

3. Trial Design

PREVAIL-P, titled “Exercise After Clinically Significant Weight Loss,” is a randomized, parallel-group pilot study sponsored by East Carolina University. Its registry status is completed, and its recorded enrollment is 39. The record lists no applicable phase. Masking is listed as none, so the study is described as unmasked.

The planned sequence began with a weight-loss phase for all participants before randomization. The registry describes this phase as including OPTIFAST and supervised aerobic exercise training two to three times per week, with a goal of losing 7% of body weight. After that phase, participants were to be randomized to PA-REC or WM-REC. PA-REC involved exercise at the minimum physical activity recommendations; WM-REC involved exercise at the weight maintenance recommendations. The posted primary outcome values are for those two post-weight-loss groups.

The primary outcome was change in weight from week 10 to week 28, measured in kilograms. The primary endpoint description identifies weight measurement in a hospital gown on a balance beam scale. The registered timeframe is the end of the weight-loss phase through follow-up. The record also lists secondary outcomes concerning cardiorespiratory fitness, body fat percentage, waist circumference, blood lipids, blood pressure, fasting glucose, fasting insulin, and systemic inflammation.

The arm-level primary outcome results include 12 participants in PA-REC and 9 in WM-REC. The relationship between those outcome counts and the total enrollment of 39 is not reported in the registry record. Details such as the randomization procedure, allocation concealment, adherence to the assigned activity recommendations, and the timing and completeness of follow-up are also not reported in the registry record. These omissions limit how fully the design and the analysis population can be reconstructed from the posted information.

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 have to make a weight loss goal in order to be assigned to be randomized in the present study. All participants completed an Optifast weight loss program after completing screening for inclusion/exclusion criteria. Participants that were able to obtain at least 7% weight loss were available for randomization in the present study

Weight Loss Phase

Reasons for not completing this period:

Weight Maintenance Phase

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.

Population: This population is the baseline values for participants after the completion of the 10 week weight loss phase. Participants who met the weight loss goal of 7% were randomized to study groups.

6. Primary Endpoint

The registry lists the primary outcome as Change in Weight, measured over End of weight loss phase (week 10) to follow-up (week 28).

The primary endpoint is change in weight, measured in kilograms, from the end of the weight-loss phase at week 10 to follow-up at week 28. Weight is a continuous measure: each participant’s outcome can take a range of values rather than being limited to categories. The registry says weight was measured in a hospital gown on a balance beam scale and that the change was calculated using the two time points.

Weight change is used to describe whether weight was maintained or changed during the follow-up period after the initial weight-loss phase. That focus matches the study’s stated aim of evaluating exercise training amount in relation to weight maintenance following clinically significant weight loss. The endpoint therefore concerns the follow-up period, not the amount of weight lost during the initial phase.

There is a sign ambiguity in the registry’s endpoint description. It says the change is calculated with week 28 weight subtracted from week 10 weight, but also says a positive value indicates an increase in weight and a negative value indicates a reduction. Those statements do not give a consistent interpretation of the sign. The reported group values should therefore be read with caution as to whether positive values indicate gain or loss; the registry record does not resolve this wording conflict.

The registry reports means and standard deviations for each group, which describe the center and spread of the posted continuous outcomes. It does not report a confidence interval for the between-group contrast or the model-based contrast estimate. Consequently, the endpoint values provide a descriptive account, while the precision and direction of the adjusted comparison cannot be fully assessed from the reported model result alone.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous, and the registry summarizes each arm with a mean and standard deviation. In general, a mean describes the center of a group’s observed values. A standard deviation describes how dispersed those values are around the mean. These summaries are useful for describing each arm, but they do not by themselves provide the model-adjusted contrast between arms or its uncertainty.

The registry identifies the comparison method as mixed models analysis. In general, mixed models can account for correlated observations, such as repeated measurements from the same participant, and can represent variation at more than one level. Their suitability and interpretation depend on the actual model specification, the data included, and the assumptions used. The registry record does not report the model formula, which measurements entered the model, the covariance structure, any adjustment variables, or the assumptions assessed. It is therefore not possible to reconstruct the analysis from the method label alone.

A superiority analysis evaluates whether the data provide evidence of a difference between groups under a specified statistical model. The registry reports p = 0.256 for this analysis. In general, a p-value describes how compatible the observed data are with a null hypothesis under the model assumptions; it is not the probability that the null hypothesis is true, and it is not a measure of clinical importance. It also does not report the estimated size of the difference. Because no model-based contrast or confidence interval is reported, the p-value cannot be translated into a precise range of plausible between-group effects.

A confidence interval, in general, gives a range of estimates compatible with the data and the method used, while reflecting sampling uncertainty. It helps distinguish a small, precisely estimated contrast from an uncertain estimate that could include materially different effects. No confidence interval for the primary between-group contrast is reported in the registry record. That absence limits assessment of precision, independently of the p-value.

The reported arm counts are 12 for PA-REC and 9 for WM-REC. The overall enrollment was 39, but the registry does not explain the relationship between enrollment and these outcome counts. It does not report whether participants were omitted because of missing measurements, withdrawal, ineligibility for the follow-up analysis, or another reason. The primary analysis population and the approach to missing outcomes are not reported in the registry record. General statistical methods for missing data can rely on different assumptions, so the missing-data approach matters to interpretation; it cannot be inferred from the arm summaries.

The posted means and standard deviations are unadjusted descriptive summaries. A recalculation based only on summary statistics would not necessarily reproduce a mixed-model analysis, which may use information and assumptions beyond those summaries. No recalculated two-arm comparison is available in the registry record for this continuous endpoint. The record indicates that values were missing or not estimable for that type of comparison. Accordingly, the sponsor’s reported model test is the available inferential result, and the descriptive arm summaries should not be mistaken for a substitute model estimate.

General statistical considerations also include the distribution of the outcome and the influence of unusually high or low observations. With modest arm counts, those features can affect means and model estimates. The registry provides no participant-level distribution, diagnostic checks, or sensitivity analyses. It does not report whether the model accounted for baseline weight or other prognostic factors. These details are relevant to understanding the analysis but should not be guessed.

Finally, the registry lists 32 outcome measures and 60 analyses, while 15 secondary outcomes are posted. In general, evaluating multiple outcomes or analyses can increase the chance of observing apparently notable results by chance if no multiplicity strategy is used. The registry record does not report a multiplicity adjustment, a hierarchy of outcomes, or a prespecified approach to interpreting the full set of tests. The primary p-value should therefore be interpreted as the result reported for the primary comparison, without assuming that it was adjusted for all other analyses.

What the p-value can say

In general, a p-value is conditional on the null hypothesis and the analysis assumptions. It is not the probability that an observed difference is due to chance, nor does it quantify how likely the treatment strategies are to have the same effect. Here the reported p-value belongs to the mixed-model superiority analysis; the underlying estimate and its uncertainty are not stated in the registry record.

Why model details matter

A method name alone does not establish which observations were analyzed or how repeated measures and missing outcomes were handled. Those choices can affect both the estimate and its uncertainty. The registry record does not include enough detail to assess these aspects of the mixed-model analysis or to reproduce it.

10. Confidence Intervals for Each Arm

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

R code and output

arm <- c("PA-REC", "WM-REC")
n   <- c(12, 9)
m   <- c(0.7, -1.5)
sp  <- c(4.13, 4.65)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  PA-REC  12  0.700  4.130  -1.924  3.324
CARM  WM-REC  9  -1.500  4.650  -5.074  2.074

11. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary comparison as a mixed-model superiority analysis, with p = 0.256. It does not give the model-based contrast estimate or a confidence interval. The posted arm summaries are means of 0.7 kg for PA-REC and -1.5 kg for WM-REC, with standard deviations of 4.13 and 4.65, respectively. Those summaries are descriptive and do not identify the mixed model’s adjusted estimate.

No recalculated two-arm comparison is available for this continuous endpoint in the registry record; the record notes that values were missing or not estimable for such a comparison. Therefore, there is no numerical recalculation to compare with the sponsor’s model result. The record does not flag a discrepancy between a recalculation and the sponsor analysis. In any event, an unadjusted calculation from arm summaries would not necessarily match a mixed model, which may use different information and assumptions. The sponsor’s reported analysis is the available inferential result, but the absence of its effect estimate and confidence interval prevents a direct assessment of magnitude and precision.

12. Multiplicity, Missing Data and Analysis Populations

The registry lists 32 outcome measures and 60 analyses, and 15 secondary outcomes are posted. These counts indicate that the primary weight comparison sits within a broader set of outcomes and analyses. The secondary outcomes include cardiorespiratory fitness, body fat percentage, waist circumference, blood lipids, blood pressure, fasting glucose, fasting insulin, and systemic inflammation. The record does not identify which analyses correspond to every listed measure or explain how the analysis count relates to the posted outcomes.

In general, when several outcomes or tests are examined, a multiplicity strategy can help control the chance of false-positive findings across a family of comparisons. Possible strategies include defining a hierarchy or adjusting inference across selected tests. The registry record does not report a multiplicity adjustment, a testing hierarchy, or whether the primary comparison was designated as the sole confirmatory analysis. No claim about adjustment should therefore be inferred from the reported p-value.

The record also does not report the primary analysis population, the number of participants with missing primary outcome data, reasons for missingness, or the method used to handle missing data. The outcome counts are 12 in PA-REC and 9 in WM-REC, while enrollment is 39; the record does not explain the relationship among these counts. It does not state whether the primary analysis followed an intention-to-treat principle, included only participants with observed follow-up, or used another population definition.

These omissions matter because different missing-data assumptions and analysis populations can yield different estimates. In general, complete-case analyses and model-based approaches do not answer identical questions when outcomes are missing. Without the analysis population and missing-data methods, the robustness of the reported result cannot be evaluated from the registry record. The model specification and sensitivity analyses are also not reported in the registry record.

13. 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 non-serious adverse event terms

The adverse-event record lists zero serious adverse events and zero deaths in each of the stated groups: the weight-loss phase, PA-REC, and WM-REC. For other, non-serious adverse events, the record lists one affected participant in PA-REC and none in WM-REC. It lists 39 participants at risk during the weight-loss phase, 16 in PA-REC, and 17 in WM-REC for the adverse-event entries.

These are counts for the groups and at-risk populations as recorded, not estimates of comparative risk. The registry does not report event details for the one non-serious event, its timing, severity, relationship to the intervention, or the duration of observation. The relationship between the adverse-event at-risk counts and the primary outcome counts is not reported in the registry record. The event counts should therefore be interpreted as a limited safety description, not as evidence that the groups have identical safety profiles.

The absence of reported serious events in these entries does not establish that an event is impossible or that the study can rule out uncommon harms. General interpretation of safety counts depends on the number observed, the length and completeness of follow-up, and consistent event collection. Those details are not reported in the registry record. The figures describe what was entered for this study and should not be generalized beyond the recorded participants and observation periods.

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

The result concerns weight change from the end of the weight-loss phase at week 10 to follow-up at week 28 among participants assigned to two exercise recommendation groups. It does not evaluate the initial weight-loss phase against a control condition. Both groups were described as entering randomization after that shared phase, and the comparison is between PA-REC and WM-REC.

The posted group means are 0.7 kg for PA-REC and -1.5 kg for WM-REC. The reported mixed-model superiority test has p = 0.256. These values describe the registry’s results, but they do not establish equivalence between the groups or show that their effects are the same. A non-significant superiority test cannot rule out meaningful differences. Since the model-based contrast and confidence interval are not reported in the registry record, the size and precision of the comparison remain unclear.

The endpoint’s sign convention is internally inconsistent: the stated subtraction order does not align with the stated interpretation of positive and negative changes. It is consequently uncertain whether the posted positive and negative means should be described as gain or loss. The numerical summaries can be quoted, but their directional meaning should not be asserted without clarification from the registry record.

The findings also do not establish whether participants adhered to the assigned activity recommendations, whether the two prescriptions were delivered as planned, or whether differences in activity explain the observed summaries. Adherence, activity measurements, and implementation details are not reported in the registry record. Nor does the result establish effects on the listed secondary outcomes; outcome names alone are not results.

In general, a pilot study can provide experience with study procedures and preliminary estimates, but its results may be uncertain and should not be treated as a definitive comparison. The recorded outcome counts are 12 and 9, and the registry does not explain their relationship to the enrollment of 39. Missing-data methods and the analysis population are also not reported. These features limit conclusions about the size, precision, and generalizability of the primary comparison.

15. Why This Trial Matters Statistically

PREVAIL-P illustrates a common statistical challenge in post-weight-loss research: the question concerns maintenance after an initial phase, so the follow-up comparison is conditional on the study sequence that preceded it. The primary outcome spans week 10 to week 28, while all participants were described as entering a weight-loss phase before randomization. Keeping that structure clear is important when interpreting what the randomized comparison can address.

The registry also demonstrates why an endpoint definition needs an unambiguous direction. The stated subtraction order and the explanation of positive and negative values conflict. That ambiguity affects the plain-language meaning of the arm means even though the numerical summaries and the reported test remain available. Clear conventions help readers distinguish gain from loss and avoid reversing an interpretation.

A second statistical lesson is that a p-value alone is not a complete account of a group comparison. The registry reports p = 0.256 but does not report the model-based contrast or a confidence interval. Without an estimate and its uncertainty, readers cannot judge the plausible magnitude of the difference. The group means and standard deviations add descriptive context but do not replace the inferential estimate.

The record’s counts of 32 outcome measures, 60 analyses, and 15 posted secondary outcomes also make transparent reporting of multiplicity important. The registry does not state how the set of analyses was handled. Finally, the relationship between total enrollment and the smaller primary outcome counts is unexplained. Reporting participant flow, missing-data methods, model details, and outcome direction would make the statistical evidence easier to interpret and evaluate.

16. 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.

Several limitations specific to the posted record constrain interpretation. First, the primary endpoint description has a sign inconsistency: the specified subtraction order conflicts with the stated meanings of positive and negative changes. The record does not clarify which convention should govern interpretation, so the direction of the reported means cannot be confidently translated into gain or loss.

Second, the arm-level primary outcome counts are 12 for PA-REC and 9 for WM-REC, compared with enrollment of 39. The registry record does not explain how those counts arose, describe participant flow, or state why other enrolled participants do not appear in the arm summaries. The analysis population and missing-outcome handling are not reported. This prevents assessment of selection and missing-data effects.

Third, the model result is reported only as a mixed-model superiority test with p = 0.256. The model-based contrast, confidence interval, covariates, repeated-measure structure, and diagnostic checks are not reported in the registry record. The arm summaries are informative descriptively, but they do not reveal the adjusted estimate or its precision.

Fourth, the study is unmasked and described as a pilot, and the posted outcome counts are modest. The registry does not report adherence to the assigned activity recommendations, the degree of separation in actual activity, or details of intervention delivery after randomization. Those missing details limit interpretation of how the assigned strategies were implemented and how broadly the results may apply.

Finally, the registry lists multiple outcomes and analyses but does not report a multiplicity strategy. The secondary outcome names do not provide their results. The adverse-event record gives limited counts without event descriptions, observation details, or attribution. Together, these gaps mean that the record supports a cautious description of the posted summaries and test, rather than a complete appraisal of efficacy, safety, or the wider analysis plan.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Exercise After Clinically Significant Weight Loss — NCT03685123.
• Registry record last updated 2024-07-03; status last verified 2024-06.
• Results first posted 2024-07-03.
• 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.

20. Record Summary

NCT03685123 is the completed, randomized, parallel-group PREVAIL-P pilot sponsored by East Carolina University. It enrolled 39 participants and compared two physical-activity recommendations after a shared weight-loss phase. The primary endpoint was change in weight from week 10 to week 28.

The posted arm summaries were 0.7 kg (standard deviation 4.13; n = 12) for PA-REC and -1.5 kg (standard deviation 4.65; n = 9) for WM-REC. The registry reports a mixed-model superiority test with p = 0.256, but no model-based contrast or confidence interval. The endpoint’s sign description is internally inconsistent, and the registry does not explain the relationship between enrollment and the outcome counts.

The record lists secondary outcomes spanning fitness, body composition, cardiometabolic measures, and inflammation. It reports no serious adverse events or deaths in the listed groups, and one non-serious event in PA-REC. Model details, missing-data methods, multiplicity handling, adherence, and the broader analysis population are not reported in the registry record.