Testing Intervention Strategies for Addressing Obesity and Binge Eating: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Testing Intervention Strategies for Addressing Obesity and Binge Eating (NCT04771455), a randomized na study. The primary outcome is Weight.

2. Clinical Question

The registry identifies the study as a randomized pilot trial of a mobile intervention targeting obesity and binge eating. Its factorial structure combines behavioral weight loss with different sets of additional behavioral components. The posted outcome comparisons are organized by whether each named component is present or absent, rather than only by the full condition descriptions.

The results therefore address component-level contrasts for weight change. They do not, on their own, establish which complete package of components is preferable, nor do the posted summaries describe a direct comparison of every condition against every other condition. Whether the results support estimates of component interactions or other factorial effects depends on the analysis model, which is not reported in the registry record.

The study status is completed, and the registry lists enrollment as 81. The posted primary outcome describes mean weight-change values, but the convention for the sign of change—whether a positive value represents gain or loss—is not reported in the registry record. The clinical question can be understood as a comparison of recorded mean changes, while keeping that direction-of-change limitation in view.

3. Trial Design

The study is listed as randomized, single masked, and factorial. Its stated purpose is to conduct a randomized pilot trial of a mobile intervention targeting obesity and binge eating. The registry lists Northwestern University as sponsor, gives the phase as NA, and records the study as completed.

Behavioral weight loss is the common intervention foundation across the eight condition descriptions. Additional components address overvaluation of weight and shape, unhealthy weight control practices, and negative affect. The conditions vary in which of those components accompany behavioral weight loss, from behavioral weight loss alone to combinations that include all three. The registry calls the design factorial, but does not report the allocation procedure, the masking role, or detailed implementation of the mobile intervention.

The registry lists enrollment as 81. The posted outcome summaries report group counts that vary across component comparisons. These are summary counts for the comparisons as recorded; the registry does not explain the handling of participants without an outcome measurement or whether all randomized participants were included in each analysis.

The primary outcome is weight, described as changes in weight and measured at baseline and post-intervention (16-weeks). The results include mean values and standard deviations for component-present and component-absent groups, alongside unadjusted mean-difference estimates, confidence intervals, and p-values. The registry record does not describe a prespecified analysis model, an adjustment strategy, or a primary analysis population. It also does not report a rationale for the enrollment target or the original design assumptions used to plan the trial.

The registry lists the study start as 2021-02-16 and primary completion as 2022-11-30. These dates and the completed status establish the record’s timeline, but do not provide information about recruitment pace, follow-up completion, or the timing of outcome collection beyond the stated post-intervention assessment.

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.

Only the first 6 of 8 groups are shown.

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.

Only the first 6 of 9 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Weight, measured over Measured at baseline, post-intervention (16-weeks).

The primary endpoint is weight, reported as mean change in pounds. The registry describes measurement at baseline and post-intervention (16-weeks), which makes change over the intervention period the outcome of interest rather than a single weight measurement at one time point. The reported group values are 4.2 pounds and 6.9 pounds for the overvaluation-of-weight-and-shape component contrast; 4.7 pounds and 6.3 pounds for the unhealthy-weight-control-practices contrast; and 8.5 pounds and 2.7 pounds for the negative-affect contrast. Standard deviations are also reported for each group.

Weight is a continuous measure, so reporting a mean gives a summary of the group’s recorded change, while the standard deviation describes variation among individual measurements around that mean. A mean difference then summarizes the contrast between the two groups in a component comparison. The registry does not describe how the change score was calculated from baseline and post-intervention measurements, or how missing measurements were handled.

Change in weight is used as the primary outcome because the intervention targets obesity and includes behavioral weight loss. The registry also concerns binge eating, but the posted primary endpoint is weight; no secondary outcomes are posted in the registry record. The sign convention for weight change is not reported in the registry record. Consequently, the numerical direction of a mean value cannot be translated confidently into weight loss or gain from this record alone.

7. Results Posted in the Registry

• Overvaluation of Weight/Shape (Yes): mean 4.20 versus 4.70 on control; difference -0.50 (95% CI -10.19 to 9.19), standardized effect d = -0.02; z-test p = 0.919.
• Overvaluation of Weight/Shape (no): mean 6.90 versus 4.70 on control; difference 2.20 (95% CI -7.42 to 11.82), standardized effect d = 0.10; z-test p = 0.654.
• Unhealthy Weight Control Practices (no): mean 6.30 versus 4.70 on control; difference 1.60 (95% CI -7.84 to 11.04), standardized effect d = 0.07; z-test p = 0.740.
• Negative Affect (Yes): mean 8.50 versus 4.70 on control; difference 3.80 (95% CI -7.28 to 14.88), standardized effect d = 0.15; z-test p = 0.501.

8. Statistical Interpretation

9. Statistical Methods Explained

What does a mean summarize?

In general, the mean is the arithmetic average of measured values in a group. For a continuous endpoint such as weight change, it gives a central summary on the original measurement scale. The posted group means are in pounds. Their interpretation depends on how individual change was defined, and the registry record does not report whether positive values indicate weight gain or loss.

The standard deviation describes how dispersed the observations are around the group mean. It is not the uncertainty in the mean itself. A group can have a mean that is estimated with limited precision when its observations vary widely, particularly when the group count is modest. In this trial, the posted standard deviations are much larger than the absolute mean differences. That context helps explain why the confidence intervals are broad, although a standard deviation alone does not determine the precision of a comparison.

What is a mean difference?

In general, a mean difference subtracts one group mean from another according to a specified contrast. The posted estimates are consistent with the listed group means and with the comparison order reflected by the reported values: for example, the overvaluation-present mean is 4.2 pounds and the overvaluation-absent mean is 6.9 pounds, with an estimated difference of -0.5 pounds. The record does not explain the outcome’s sign convention, so the sign of a contrast indicates numerical ordering, not necessarily whether one group lost more weight.

The comparisons are component-present versus component-absent summaries. Because the study is identified as factorial, component contrasts may be relevant to the design. However, the registry does not describe an analysis that separates component main effects from interactions, or an adjustment for the other components. The posted simple contrasts should not be mistaken for a fully specified factorial analysis.

How should the confidence intervals be read?

In general, a confidence interval gives a range of values compatible with the observed estimate under the assumptions of the method used to construct it. A narrower interval indicates greater precision; a wider interval reflects greater uncertainty. For the overvaluation-of-weight-and-shape contrast, the interval is -10.19 to 9.195 pounds. For unhealthy weight control practices, it is -7.844 to 11.04 pounds. For the negative-affect contrast with an estimate of 3.8 pounds, it is -7.279 to 14.88 pounds. These intervals all include zero and span values on both sides of zero.

The registry does not report the confidence level or the method used to construct these intervals. It also does not report whether any covariates or design features were included in an interval calculation. The posted comparison statistics are unadjusted re-calculations from summary data, so they should not be presented as results from a documented adjusted model.

What do the p-values say?

In general, a p-value measures how unusual data at least as extreme as the observed result would be under a specified null hypothesis and analysis model. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. The posted unadjusted p-values are 0.919, 0.654, 0.74, 0.501, and 0.624 for the listed contrasts. The registry does not report an adjustment for multiple comparisons or a prespecified significance threshold for these outcome analyses.

Because the p-values come from unadjusted recalculations, they are descriptive of those contrasts under the calculation assumptions, not evidence of a prespecified sponsor analysis. None of the p-values should be interpreted without the corresponding estimate and confidence interval. Here, the intervals are broad and include zero, so the evidence does not precisely distinguish a zero difference from a range of positive and negative differences.

Why does variability matter?

In general, over-dispersion means that observations vary more than a statistical model or simple planning calculation expects. Greater variability makes it harder to estimate a group mean or difference precisely, unless offset by more observations or a more informative design. The standard deviations in the posted weight summaries indicate considerable variation relative to the reported mean differences. The record does not report a formal assessment of over-dispersion or a model designed to account for it.

These methods explain the quantities reported, but they cannot fill gaps in the registry. The original statistical analysis plan, model specification, missing-data method, analysis population, confidence-interval level, and multiplicity strategy are not reported in the registry record. Conclusions should therefore remain tied to the posted group summaries and unadjusted comparisons.

10. Confidence Intervals and Precision

The mean-difference estimates are not very precise. For overvaluation of weight and shape, the estimated difference is -0.5 pounds, and its interval ranges from -10.19 to 9.195 pounds. For unhealthy weight control practices, the estimate is 1.6 pounds, with an interval from -7.844 to 11.04 pounds. The negative-affect comparison with an estimate of 3.8 pounds has an interval from -7.279 to 14.88 pounds; the other negative-affect contrast is -2.0 pounds, with an interval from -9.995 to 5.995 pounds.

In general, an interval describes uncertainty around an estimate, under the calculation method and assumptions. These intervals include zero and also include values in both directions. Thus, the posted data do not pin down whether the corresponding average differences are positive or negative, nor do they rule out differences of a range of sizes. The estimates should not be described as showing that groups are equivalent: equivalence requires a defined margin and an analysis designed to assess it, neither of which is reported in the registry record.

The reported standard deviations are 21.9 and 21.3 pounds for the overvaluation-present and -absent groups; 22.6 and 20.7 pounds for the unhealthy-weight-control-practices-present and -absent groups; and 27.7 and 12.6 pounds for the negative-affect-present and -absent groups. These values show substantial spread in the outcome summaries, but they do not by themselves establish why the intervals are broad.

The registry does not report the confidence level or the procedure used to obtain the intervals. Precision should therefore be judged from the displayed ranges without assigning an unreported confidence level. The sign convention for the change score is also absent, limiting clinical interpretation of the direction of estimates.

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 sample-size display gives planning illustrations in participants per arm, not total event counts. The registry’s stated planning assumptions are a two-sided significance level 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect. These are planning calculations based on the observed contrast, not the original trial design assumptions; the trial’s own design assumptions are not reported in the registry record.

The 80% power column gives the illustrative number of participants per arm under which the specified calculation would have an 80% chance of detecting the assumed effect, if the planning assumptions hold. The 90% power column gives the corresponding illustration for 90% power. For example, planning for 100% of the observed overvaluation-present contrast gives 31,081 participants per arm for 80% power and 41,609 participants per arm for 90% power. These are hypothetical planning quantities, not numbers recruited for this pilot.

Rows that plan for 75% or 50% of an observed effect assume a smaller effect to detect. Smaller effects are harder to distinguish from random variation, so the calculations require more participants per arm. For example, the overvaluation-present planning figures increase from 31,081 participants per arm for 100% of the observed effect to 55,255 for 75% and 124,324 for 50%, for 80% power.

The calculations are not observed power. They are forward-looking illustrations derived from the observed estimates and stated assumptions. Since the observed contrasts are uncertain and the original design rationale is not reported, these figures should not be read as recommendations or as evidence about the adequacy of the pilot’s enrollment.

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 = -0.500; pooled standard deviation σ = 22.248.
2. n = 2 × (1.96 + 0.842)² × 22.248² / -0.500² = 31,080.8, rounded up to 31,081 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 <- -0.5                      # observed difference in means
sigma <- 22.2483223659957                     # 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(4771455)
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 = 31,081 (80% power), 41,609 (90% power)
 75% of observed effect: n per arm = 55,255 (80% power), 73,971 (90% power)
 50% of observed effect: n per arm = 124,324 (80% power), 166,434 (90% power)
power.t.test: 31081.8 
Simulated power at n = 31,081 per arm: 0.799 (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. Multiplicity, Missing Data and Analysis Populations

The registry identifies a factorial randomized design and posts several component-level comparisons for the primary weight outcome. It does not report a plan for multiplicity, such as which comparisons were prespecified, how a primary contrast was selected, or whether p-values were adjusted to account for multiple testing. The listed p-values are unadjusted. Their interpretation should therefore remain tied to each individual contrast, without treating them as a familywise-confirmatory assessment.

The registry lists no analyses and no posted secondary outcomes. This means that the record does not provide a detailed inferential analysis plan or additional outcome results; it does not establish that no analyses were performed outside the registry. The analysis model, handling of factorial interactions, covariate adjustment, and any hierarchy among hypotheses are not reported in the registry record.

Missing-data handling is also not reported. The record does not say how many participants had a post-intervention weight measurement, whether missing values were imputed, or whether analyses used complete cases or another approach. The reported group counts differ across comparison rows, but the registry does not explain the reason. Without a stated missing-data strategy, it is not possible to assess whether the posted summaries represent all randomized participants or a subset with available measurements.

The analysis population is not reported in the registry record. There is no stated intention-to-treat, per-protocol, or other population definition, and no information about exclusions after randomization. The posted re-calculations are based on summary statistics and should not be mistaken for a documented participant-level analysis. These gaps limit assessment of possible bias, the handling of the factorial design, and the robustness of the comparisons.

14. 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 registry reports zero serious adverse events in each of the eight condition groups, and zero other adverse events in each group. It also reports zero deaths in every group. The listed numbers at risk for serious and other adverse events range from 10 to 11 per condition. No serious adverse event terms are listed.

These counts describe what the registry record reports for the groups and outcomes. A zero count means no event was recorded in the relevant category in the posted results; it does not establish that the risk is absent. The group counts at risk are small, so the safety summaries cannot precisely characterize uncommon events. The record does not report event definitions, monitoring procedures, the period over which events were collected, or whether all randomized participants were included in safety reporting.

Safety results should be read separately from the weight outcome. The registry’s zero event counts do not determine whether weight changed, and the weight comparisons do not establish safety. The posted safety information is limited to the serious-event, other-event, and death counts; further details are not reported in the registry record.

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

The results provide preliminary estimates of how mean weight-change values differed across groups defined by intervention components. The estimates are close to zero for some contrasts and farther from zero for others, but each confidence interval includes zero and covers values in both directions. The unadjusted comparisons therefore leave considerable uncertainty about the size and direction of the underlying group differences.

This does not establish that the intervention components have no effect, nor does it show that any component produces a benefit. The intervals are broad, and the pilot’s posted summaries are not a substitute for a prespecified, adequately detailed factorial analysis. The sign convention for weight change is not reported, so even a numerical difference cannot be translated into a direction of weight loss or gain with confidence from the registry record alone.

The trial evaluates a mobile behavioral intervention intended to address obesity and binge eating. However, the posted primary outcome is weight change. The registry does not post secondary outcome results, so the record does not support conclusions about binge-eating outcomes or other effects of the intervention. It also does not provide an adjusted comparison of complete intervention packages or an interaction analysis.

The planning calculations illustrate how large participant counts can become when planning around small observed differences, especially when the target effect is reduced. Those values are not observed power and do not establish that the pilot was adequately or inadequately designed. The original design assumptions are not reported.

Overall, the posted data are best read as uncertain pilot estimates rather than definitive evidence for or against a component or package. The completed status and recorded outcome values do not resolve the methodological gaps around missing data, analysis populations, multiplicity, and the statistical model. Those limitations matter when considering how much weight to place on the numerical contrasts.

16. Why This Trial Matters Statistically

This trial is statistically interesting because its factorial design varies several behavioral components around a common behavioral weight loss intervention. Such a design can, in principle, help distinguish the contribution of individual components and their combinations. The registry identifies that structure, and the posted comparisons are organized around component presence and absence.

That structure also makes the analysis plan important. A factorial study may require explicit decisions about component main effects, interactions, multiple comparisons, and the population included in each analysis. The registry record does not report those decisions. The posted simple contrasts therefore offer useful descriptive information but cannot be assumed to represent a complete factorial analysis.

The estimates also illustrate the distinction between effect size and precision. Some mean differences are numerically small, while their confidence intervals are wide enough to include both negative and positive values. This means the data do not precisely identify the group contrasts. It is not sound to equate a non-small p-value with proof of no difference, or to read an observed estimate as a reliable prediction of a future trial.

The planning illustrations further show why pilot estimates should be handled cautiously in sample-size planning. Calculations based on observed effects can produce very large participant requirements when the estimated contrast is small, and planning for smaller effects increases those requirements. Such calculations are only as informative as the assumptions and effect estimates used. Here, the original design rationale and analysis plan are not reported.

The record’s strengths for statistical interpretation are the explicit outcome scale, group means, standard deviations, comparison estimates, confidence intervals, and event counts. The limits are equally important: no analysis model, missing-data approach, multiplicity plan, or analysis population is documented. Together, these details make the trial a useful example of the need to pair numerical estimates with precision and transparent methods.

17. 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 trial-specific limitations constrain interpretation. The registry describes the study as a pilot and lists enrollment as 81. The posted group summaries are small, and their standard deviations are large relative to the mean differences. The resulting intervals are wide and include zero, so the comparisons have limited precision.

The outcome’s change-score definition and sign convention are not reported in the registry record. Although the endpoint is described as changes in weight measured at baseline and post-intervention (16-weeks), the record does not explain the calculation of change or whether the reported group means are based on paired measurements. This prevents a more detailed assessment of the endpoint construction and the meaning of positive values.

The registry does not report the eligibility criteria, participant characteristics, intervention implementation details, or the role of masking. It also does not explain why comparison counts vary or how missing outcome measurements were handled. These omissions make it difficult to assess representativeness, adherence, follow-up, and potential selection related to available outcome data.

Statistical details are limited: the factorial model, interaction analyses, covariate adjustment, multiplicity procedure, confidence-interval level, and analysis population are not reported. The posted comparisons are unadjusted recalculations from summary statistics, rather than a documented adjusted analysis. The registry also posts no secondary outcomes, so effects on binge eating cannot be evaluated from these results. Finally, the sign convention limits clinical interpretation of the numerical direction of weight change.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Testing Intervention Strategies for Addressing Obesity and Binge Eating — NCT04771455.
• Registry record last updated 2024-09-19; status last verified 2024-09.
• Results first posted 2024-09-19.
• 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.

21. Record Summary

The registry describes a completed, randomized, single-masked factorial pilot trial of a mobile behavioral intervention targeting obesity and binge eating. Behavioral weight loss is combined with components addressing overvaluation of weight and shape, unhealthy weight control practices, and negative affect. Enrollment is listed as 81.

The primary outcome is mean change in weight in pounds, measured at baseline and post-intervention (16-weeks). Posted component comparisons report group means, standard deviations, mean differences, confidence intervals, and unadjusted p-values. The estimates are imprecise, and their confidence intervals include zero.

The registry reports zero serious adverse events, zero other adverse events, and zero deaths in each condition group. No secondary outcomes are posted. Details of the analysis model, missing-data handling, multiplicity, analysis population, and sign convention for weight change are not reported in the registry record.