Biology and Experience of Eating in Women With Obesity: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Biology and Experience of Eating in Women With Obesity (NCT02805972), a randomized phase 2 study. The primary outcome is Number of Participants Who Reported Nausea at 10 Minutes Post Treatment.

2. Clinical Question

The study’s stated purpose was to understand how the opioid system is involved in eating behavior. The posted primary endpoint addresses a specific, short-term symptom after each intervention; it is not a direct measure of eating behavior or of the broader study purpose. The registry reports 17 participants with nausea after naloxone and 18 after placebo, with 38 recorded for each treatment. Those observations are compared using a risk difference and other treatment contrasts. The endpoint is described among participants who completed both intervention visits, so it concerns the crossover participants with both visits completed rather than a general population of people with obesity.

3. Trial Design

The registry describes a randomized, quadruple-masked, phase 2 crossover study. Its title is Biology and Experience of Eating in Women With Obesity, and the listed condition is obesity. The sponsor is the University of California, San Francisco. The registry reports an enrollment of 41 and a completed status. The listed study start is 2017-05-20 and primary completion is 2018-07-25.

The two interventions were naloxone and placebo. One sequence received 4 mg / 0.1 ml naloxone at visit 1 and 0.1 ml saline at visit 2; the other received saline at visit 1 and naloxone at visit 2. Both sequence descriptions specify a washout period between visits, but its length is not reported in the registry record. The registry labels the trial randomized and quadruple masked; it does not identify the masking roles or provide further details about how randomization was conducted.

In a crossover design, participants receive both interventions in an assigned order. That structure means the outcome data relate to intervention visits and treatment order, rather than to two groups that each receive only one intervention. The primary outcome population is described as participants who completed both intervention visits. The posted primary results give treatment-specific counts, but the registry record does not report the statistical method used to account for pairing, visit, sequence, or possible period effects. The result should therefore be read as the registry’s posted comparison, not as evidence that any particular crossover model or adjustment was used.

The registry lists cortisol and the Subjective Opiate Withdrawal Scale, including an abbreviated version, among the secondary outcomes. It lists eight outcome measures and one analysis. Details of the secondary outcome results are not included in the primary nausea comparison.

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.

Intervention Visit 1

Washout Period

Reasons for not completing this period:

Intervention Visit 2

Analysis population for the primary outcome, as defined in the registry: All participants who received Naloxone, then Placebo, or Placebo, then Naloxone, by participating in both study intervention visits.

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 Number of Participants Who Reported Nausea at 10 Minutes Post Treatment, measured over 10 minutes post-treatment.

The primary endpoint counts participants who reported nausea 10 minutes after an intervention. The registry defines the response using the nausea item on the Subjective Opiate Withdrawal Scale: a response of 0 means absence of nausea, and a response of 1 means presence. The endpoint is therefore binary and reported as a count of participants, rather than as a continuous symptom score or a measure of nausea severity.

The specified time point is 10 minutes post-treatment. The outcome population comprises participants who received both interventions by participating in both study visits. The posted counts are 17 after naloxone and 18 after placebo, with 38 participants recorded for each treatment. The registry record does not describe how a missing item response, if any, was handled.

A binary symptom endpoint can provide a straightforward comparison of whether the symptom was reported at a specified time. It does not indicate how intense or long-lasting nausea was, whether it affected eating behavior, or how participants felt at other times. Those are distinct questions. Here, the nausea count is a narrow outcome related to the study’s broader interest in the opioid system and eating behavior; it should not be treated as a direct measurement of that broader interest.

7. Results Posted in the Registry

• Naloxone: 44.7% (17/38) versus 47.4% (18/38) on control; risk difference -2.6 percentage points (95% CI -25.0 to 19.8), risk ratio 0.94 (0.58–1.54); unadjusted two-proportion z-test p = 0.818.

8. Statistical Interpretation

9. Statistical Methods Explained

What kind of outcome was recorded?

The primary outcome is a binary event count: whether a participant reported nausea at a specified assessment time. A binary outcome has two possible recorded states in this definition, presence or absence. The registry reports the number of participants with the event and the number recorded for each treatment. This is different from a symptom scale that retains multiple response levels, a measure of severity, or a continuous outcome such as a laboratory value.

The event was defined from the nausea item on the Subjective Opiate Withdrawal Scale at 10 minutes post-treatment. The counts are 17 after naloxone and 18 after placebo. The recorded treatment denominators are 38 for each. The primary population is described as participants who completed both intervention visits. The record does not explain whether all completed-visit participants had a nonmissing nausea response at each assessment or how missing responses were treated.

What does a risk difference express?

A risk difference compares the event proportions for two treatments by subtracting one from the other. It expresses an absolute contrast on the event-proportion scale. The posted estimate is -0.026. The registry also lists the treatment proportions as 0.447 after naloxone and 0.474 after placebo. The sign indicates the direction of the posted contrast; the magnitude describes the estimated separation on that scale. It is an estimate, not a guarantee that the same difference would occur in another study.

The registry reports a risk-difference interval from -0.250 to 0.198. A confidence interval is a way to express statistical uncertainty around an estimate. In general, it is constructed using an analysis procedure and its assumptions; repeated use of a procedure under those assumptions has a defined coverage property. The exact confidence level and construction method for this interval are not reported in the registry record. Since this interval includes values on both sides of zero, the direction of a difference is not resolved by the interval alone.

How do the relative measures differ?

The risk ratio compares the event proportions as a ratio, while the odds ratio compares the odds of the event. In general, these measures use different scales from the risk difference, so their numerical values should not be read as absolute changes in event proportion. The posted risk ratio is 0.944, with an interval from 0.580 to 1.537. The posted odds ratio is 0.899, with an interval from 0.365 to 2.218. The intervals include the value that represents no relative difference. Neither ratio, by itself, supplies evidence of equivalence or rules out meaningful differences.

What does the p-value contribute?

The unadjusted p-value is 0.818. In general, a p-value describes the compatibility of observed data, or data more extreme under a specified test, with a null model. Its interpretation depends on the test and assumptions. The registry does not state which test produced this p-value, whether it accounts for the crossover structure, or whether it is the result of a prespecified analysis. The p-value does not measure the size of the treatment contrast, and it is not the probability that the null hypothesis is true. It does not replace the estimate and interval.

The record lists one analysis for the primary outcome, but that listing does not identify an analysis method. No adjustment method, covariates, participant-level pairing procedure, or handling of treatment sequence and visit is reported. It would therefore be inappropriate to attribute a particular statistical test or model to the sponsor from these results alone. In a crossover study, a statistical analysis may need to reflect that a participant contributes measurements under both interventions; whether and how that was done is not reported.

Why does the crossover design affect interpretation?

In general, crossover comparisons can use within-participant information, because each participant receives both interventions. Such analyses also need to consider the order of interventions and whether outcome measurements could vary by visit or period. The registry specifies two intervention sequences and a washout period, but does not report its duration or give details of a paired analysis, period adjustment, or carryover assessment. These missing details limit what can be said about the relation between the design and the posted uncertainty.

The primary endpoint population is participants who completed both visits. That definition identifies the intended population for the posted endpoint, but it does not explain whether any participants did not complete both visits, why they might not have completed them, or how their data were treated. The registry enrollment and treatment-specific outcome counts are not interchangeable descriptions of a single denominator: the treatment counts represent recorded outcome data by intervention in a crossover setting. The record does not provide participant-level flow sufficient to reconstruct the analysis set.

Overall, the methods that can be described are the registered binary nausea definition, the time point, the stated population, and the posted effect measures. The inferential procedure, confidence level, and crossover-specific analysis details are not reported in the registry record. Interpretation should stay with the posted estimates and their uncertainty rather than assume a method that is not documented.

10. Confidence Intervals and Precision

The risk-difference estimate is -0.026, with an interval from -0.250 to 0.198. The interval is wide enough to include values indicating a lower event proportion after naloxone, values indicating a higher event proportion, and values near no difference. It therefore conveys substantial uncertainty about the size and direction of the contrast. The registry record does not report the confidence level or the procedure used to form the interval, so those details cannot be established from the posted results.

The relative estimates have similarly broad intervals across the no-difference value. The risk ratio is 0.944, with an interval from 0.580 to 1.537; the odds ratio is 0.899, with an interval from 0.365 to 2.218. These intervals reinforce that the point estimates alone should not be treated as precise descriptions of the underlying contrast. Different measures use different scales, but each conveys uncertainty in its own terms.

Precision is distinct from whether a p-value meets a threshold. The unadjusted p-value is 0.818, but the interval gives more direct context about the range of contrasts compatible with the reported analysis. A result that does not establish a difference can still be imprecise; it does not demonstrate that the treatments are equivalent. In general, more information can narrow uncertainty, while a small event count or limited outcome data can leave estimates less precise. The registry record does not report a precision target or the original sample-size justification, so the adequacy of precision relative to a prespecified target cannot be assessed.

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 calculated from the observed effect under the stated planning assumptions: a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. The planning scenarios use 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect. The trial’s own original design assumptions and sample-size justification are not reported in the registry record.

The 80% power column means the illustrated number of participants per arm for a design intended to have 80% power to detect the effect specified in that row, under those assumptions. The 90% power column has the same interpretation for a design intended to have 90% power. Power is a design property under a specified effect and analysis; it does not describe the probability that the observed result is correct. The figures are planning illustrations, not observed power.

Read across a row to compare the illustrated per-arm requirements for the two power targets. Read down the scenarios to see how the planning requirement changes when the target effect is reduced: smaller effects need more participants to detect, all else equal. The displayed values should not be treated as a recommendation for a future study without considering its design, outcome, analysis, and target effect. In particular, the registry record does not report the assumptions used for the original trial’s design, and the planning calculations based on the observed result do not establish what enrollment the study should have had.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.474; treatment proportion p₁ = 0.447; average p̄ = 0.461.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7049 = 1.3816.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.7047 = 0.5931.
4. n = (1.3816 + 0.5931)² / (-0.0263)² = 5,630.4, rounded up to 5,631 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
p0 <- 0.473684210526316                          # control proportion
p1 <- 0.447368421052632                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(2805972)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 5,631 (80% power), 7,537 (90% power)
 75% of observed effect: n per arm = 10,021 (80% power), 13,415 (90% power)
 50% of observed effect: n per arm = 22,569 (80% power), 30,213 (90% power)
power.prop.test: 5630.4 
Simulated power at n = 5,631 per arm: 0.801 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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 lists eight outcome measures, one analysis, and four posted secondary outcome names. The secondary names include cortisol, the Subjective Opiate Withdrawal Scale, and the abbreviated Subjective Opiate Withdrawal Scale; cortisol and the full scale name each appear more than once in the listed names. The primary endpoint is the nausea response at 10 minutes. The record does not describe a hierarchy of outcomes or explain how the primary and secondary outcomes were related in the analysis.

Multiplicity arises when several outcomes, time points, comparisons, or analyses are examined. In general, conducting multiple tests can increase the chance of at least one apparently unusual result under a collection of null hypotheses. Researchers may prespecify an outcome hierarchy or use an adjustment strategy, but neither an adjustment strategy nor a statement that no adjustment was used is reported in the registry record. The single listed analysis does not establish that only one inferential test was performed or that multiplicity was addressed.

The primary outcome population is described as all participants who received both interventions by participating in both visits. The registry does not report a participant flow for this endpoint, the number with missing nausea responses, reasons for missing data, or the method used to handle missing observations. It also does not state whether the posted analysis followed an intention-to-treat principle, a per-protocol principle, or another analysis-population definition. These details matter because missingness and population definitions can affect the interpretation of an estimate.

For a crossover analysis, participant pairing and treatment sequence may also be relevant. The record gives the intervention sequences and specifies a washout period, but does not report how the analysis handled pairing, visit, sequence, or potential carryover. It does not report an interim analysis or the statistical analysis plan. These omissions do not show that such procedures were absent; they mean that their use cannot be confirmed from the registry record.

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 serious adverse events, other adverse events, and deaths by intervention sequence. For Naloxone, Then Placebo, the listed at-risk count is 23 for serious adverse events and 23 for other adverse events; each corresponding affected count is 0, and deaths are 0. For Placebo, Then Naloxone, the listed at-risk count is 18 for serious adverse events and 18 for other adverse events; each corresponding affected count is 0, and deaths are 0. No serious adverse-event terms are listed.

These figures are presented by sequence, not as separate safety counts for the naloxone and placebo exposures. Because participants in each sequence received both interventions, the sequence-level figures do not identify which intervention was associated with an event, and the registry record does not give further exposure-specific safety details. The at-risk counts for these safety listings also differ from the treatment-specific counts in the primary nausea outcome; they should not be substituted for those outcome denominators.

Zero recorded events in these registry categories means no events are reported for the listed groups in this record. It does not establish that no symptoms occurred or that there were no unlisted events. The record does not provide details on adverse-event collection, severity grading beyond the serious and other categories, or the period during which events were monitored.

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

The primary result describes nausea reported at a single short-term assessment after naloxone or placebo among participants who completed both intervention visits. The posted counts are close, and the risk-difference estimate is -0.026. Its interval, from -0.250 to 0.198, spans both directions of effect. The unadjusted p-value is 0.818. Taken together, these statistics do not establish a clear difference in this recorded outcome, and the interval leaves considerable uncertainty about the possible size and direction of a difference.

This result does not show that naloxone and placebo are equivalent. Equivalence requires a defined margin and an analysis designed to assess whether the uncertainty excludes differences beyond that margin; neither is reported in the registry record. Nor does a non-significant result show that there is no effect. It means that the posted analysis did not establish a difference, while the estimated contrast remains imprecise.

The outcome is not a direct measure of eating behavior, food intake, or the biological mechanisms involved in eating. The study’s stated purpose concerns the opioid system and eating behavior, but this primary endpoint measures presence or absence of nausea 10 minutes after treatment. It does not indicate symptom severity, later symptoms, or effects on eating. Results for the listed secondary outcomes are not part of the primary comparison.

The crossover design also places limits on what can be inferred from the registry summary. Participants were assigned to sequences that each included both interventions, and a washout period was specified. The registry does not report whether the analysis accounted for within-participant pairing, treatment order, visit, or possible carryover. Therefore, the numerical contrast can be described as posted, but the record does not support conclusions about the precise model behind it.

These findings describe the registry’s recorded endpoint and analysis, not a recommendation about treatment. They should not be generalized beyond the studied population, outcome definition, and time point without additional evidence. The record does not provide the detail needed to evaluate the broader biological question or the full analysis plan.

16. Why This Trial Matters Statistically

This study illustrates why the statistical meaning of a result depends on more than the direction of a point estimate or the size of a p-value. The primary comparison is a binary symptom outcome in a crossover design. Its event counts are close, while its risk-difference interval extends to both sides of zero. Looking at the estimate together with its uncertainty makes clear that the posted result does not identify the direction or size of a stable difference.

It also highlights the distinction between an outcome that is easy to define and a question that is broader than that outcome. Nausea at 10 minutes is a specific binary measure. The study’s stated aim concerns the opioid system and eating behavior, which cannot be answered directly by this endpoint alone. Statistical interpretation must keep the endpoint’s scope aligned with the claim being considered.

The crossover format can be efficient in some settings because each participant receives both interventions, but it requires analysis details that reflect treatment order and repeated observations. The registry gives the intervention sequences and a washout period, yet does not describe whether the posted analysis accounted for pairing, visit, or sequence. That gap is an important reporting limitation: without those details, readers cannot fully connect the design to the reported uncertainty.

Finally, the planning illustrations show how sample-size requirements depend on the effect a future study aims to detect and the desired power. They are based on the observed effect and specified calculation assumptions, not on the original trial design. Distinguishing such planning quantities from observed results helps prevent a retrospective sample-size calculation from being misread as evidence about the quality or success of the trial.

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.

The primary result is limited to a binary report of nausea 10 minutes post-treatment. It does not record nausea severity, duration, symptoms at other times, or eating behavior. The registry record does not report how missing item responses were handled, whether any outcome observations were missing, or why participants may not have completed both intervention visits. The enrollment figure and treatment-specific outcome counts do not provide enough participant-flow detail to reconstruct the analysis population.

Although the study is described as randomized, quadruple masked, and crossover, the registry does not report the randomization procedure, the masking roles, or the washout duration. It also does not state whether the statistical analysis used the paired structure or adjusted for sequence or visit. The confidence level and method for calculating the reported intervals are not given, and the test associated with the unadjusted p-value is not identified.

The record does not report an original sample-size justification, a prespecified analysis plan, multiplicity handling, interim analyses, or a definition of a clinically meaningful difference. The planning illustrations use the observed effect and stated assumptions; they cannot establish what assumptions guided the original design. Secondary outcomes are listed, but their results and analysis details are not part of the primary comparison.

Safety information is reported by sequence, without exposure-specific attribution to naloxone or placebo. The registry does not give further detail on event collection periods or the handling of events beyond the listed categories. These limits leave important design, analysis, and reporting questions unresolved; they should not be interpreted as evidence that the unreported procedures or events did or did not occur.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Biology and Experience of Eating in Women With Obesity — NCT02805972.
• Registry record last updated 2021-11-03; status last verified 2021-10.
• Results first posted 2021-11-03.
• 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.

21. Record Summary

The registry record describes a completed, randomized, quadruple-masked, phase 2 crossover study titled Biology and Experience of Eating in Women With Obesity. The sponsor is the University of California, San Francisco. It lists an enrollment of 41, a start date of 2017-05-20, and primary completion on 2018-07-25.

The primary outcome is the number of participants reporting nausea 10 minutes after naloxone or placebo, among those who completed both intervention visits. The recorded counts are 17 after naloxone and 18 after placebo, with 38 recorded for each treatment. The risk difference is -0.026, with an interval from -0.250 to 0.198; the unadjusted p-value is 0.818. The interval includes values in both directions, so the posted estimate is imprecise.

The registry lists cortisol and two versions of the Subjective Opiate Withdrawal Scale among the secondary outcomes. It does not report the primary analysis method, crossover-specific modeling details, missing-data handling, multiplicity procedure, or original sample-size assumptions. The safety listings report zero serious adverse events, other adverse events, and deaths in both sequence groups.