Oxytocin Suppresses Substance Use Disorders Associated With Chronic Stress: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Oxytocin Suppresses Substance Use Disorders Associated With Chronic Stress (NCT02058251), a randomized phase 1 study. The primary outcome is Alcohol Craving.

2. Clinical Question

The clinical question is therefore specific: among the male participants included in the primary analyses, how did the recorded mean alcohol-craving scores compare between the oxytocin and saline groups? The study description places that question in the context of stress-related vulnerability among veterans with co-occurring alcohol use disorder and PTSD. The posted results address a laboratory setting and a short timeframe; they do not, by themselves, answer whether any difference would persist outside that setting or affect longer-term substance use.

3. Trial Design

The registry describes this as a randomized, double-masked, factorial Phase 1 trial. It lists an enrollment of 73 and a completed status. The sponsor was the Medical University of South Carolina. The record lists the study start as 2014-02 and primary completion as 2017-01. The factorial label is recorded, but the factors, their levels, and any interaction analyses are not reported in the registry record.

The registered intervention groups were oxytocin and control. The arm descriptions specify a self-administered 40 IU intranasal dose of oxytocin or a 40 IU intranasal saline dose. The exact masking procedures, including which study roles were masked and how masking was maintained, are not reported in the registry record. The method used to generate or conceal the randomized allocation is also not reported in the registry record.

The primary outcome was alcohol craving, recorded as a mean score with standard deviation. The outcome description says participants rated subjective craving on a scale from 1 to 10, where one was the lowest/better score and 10 was the highest/worst outcome. Its timeframe was 2 hours. The primary-outcome description states that the main outcome analyses were limited to male participants. The posted group sizes for that outcome are 32 in the oxytocin arm and 35 in control. The registry lists stress reactivity as a secondary outcome, also with a 2 hours timeframe.

The study description says participants were exposed to a laboratory-induced stress task, identified as the Trier Social Stress Task. The registry record does not report the full visit sequence, timing of treatment relative to the task, or a detailed statistical analysis plan. Those details should not be inferred from the factorial label or from the posted group summaries.

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.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Main outcomes analyses were limited to male participants.

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 Alcohol Craving, measured over 2 hours.

The primary endpoint is alcohol craving, measured using self-reported ratings on the Visual Analogue Scale. Participants rated subjective craving on a scale of 1-10; one represented the lowest/better score and 10 the highest/worst outcome. The registry describes the outcome as a mean with standard deviation and gives a 2 hours timeframe. The primary-outcome population note says the main outcome analyses were limited to male participants.

A mean summarizes the average score within a group. The standard deviation describes how dispersed individual scores are around that group mean; it does not describe the precision of the mean difference itself. The posted results show these summaries for the oxytocin and control groups, alongside a comparison of their means. This makes the endpoint interpretable as a group-level difference in reported craving scores, rather than a count of participants who crossed a particular threshold.

Self-reported craving is relevant to the study's stated question because the study description asks whether intranasal oxytocin changes alcohol craving following laboratory-induced stress. A rating scale can capture participants' subjective responses at the specified assessment, but it is not the same as observing alcohol consumption, later relapse, or a longer-term clinical outcome. Those outcomes are not the primary endpoint. The registry record also does not report details such as the scale's validation in this particular study, how scores were handled if a participant did not answer, or whether multiple ratings were combined before analysis.

7. Results Posted in the Registry

• Oxytocin: mean 2.89 versus 3.08 on control; difference -0.19 (95% CI -1.47 to 1.09), standardized effect d = -0.07; z-test p = 0.771.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Stress Reactivity

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Oxytocin: mean 0.14 versus 0.10 on control; difference 0.04 (95% CI -0.01 to 0.09), standardized effect d = 0.40; z-test p = 0.102.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Oxytocin and the control. The interval gives the range of effects that remain compatible with the data.

The registry record lists stress reactivity as a secondary endpoint alongside the primary alcohol-craving endpoint. Both are continuous outcomes assessed at the stated 2 hours timeframe, but they represent different aspects of response. The primary result is a comparison of self-reported craving; the secondary result concerns stress reactivity. The overall direction of the posted secondary mean comparison is not the same as the direction of the primary point estimate. The table should be consulted for its exact estimate and interval rather than treating this difference in direction as evidence of a reliable contrast.

The secondary outcome can provide context for the study's broader question about responses after laboratory-induced stress, but it should not be used to replace or reinterpret the primary endpoint. A difference between the directions of two point estimates can occur because of sampling variability, distinct measurement scales, or genuine differences between outcomes; the posted record alone does not identify which explanation applies. The registry does not report a formal joint analysis linking the endpoints.

The record presents an unadjusted comparison for the secondary endpoint. When several outcomes or analyses are considered, each creates another opportunity for a seemingly notable result to arise by chance. An unadjusted result does not account for that broader set of comparisons. The registry record does not report whether outcomes were prioritized in a hierarchy, whether any multiplicity adjustment was planned, or whether additional analyses were conducted but not posted. This limits the weight that can be placed on a secondary result in isolation.

Interpretation should therefore take account of the primary result, the uncertainty around each posted estimate, and the absence of a reported multiplicity strategy. The secondary outcome is informative as part of the registered outcome set, but it is not independent confirmation of the primary finding. The registry record's secondary analysis field does not list reported analyses, so details beyond the posted summary comparison are not available.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Stress Reactivity

9. Statistical Interpretation

10. Statistical Methods Explained

The posted primary outcome is continuous: it is a numeric craving score summarized by a mean and standard deviation in each group. A mean difference expresses the contrast between the two group means in the original score units. Here, the registered comparison is oxytocin relative to control, and the estimated mean difference is -0.19. The negative sign indicates that the oxytocin group mean was lower than the control group mean in this posted comparison. This is a description of the estimate, not a conclusion about an underlying treatment effect.

The group standard deviations describe variation among participants' scores. They are not confidence limits and should not be read as ranges that contain a particular proportion of possible treatment effects. The posted group summaries are 2.89 with standard deviation 2.63 for oxytocin and 3.08 with standard deviation 2.71 for control. The registry lists 32 and 35 participants, respectively, for these outcome summaries. In general, a group mean can be estimated more precisely when there is less variation and more information, but the precision of a comparison also depends on the analysis and the study data.

The comparison includes a confidence interval for the mean difference. In general, a confidence interval is a range produced by a method designed to have a specified coverage rate over repeated studies under its assumptions. It communicates uncertainty about the estimated contrast, not the range of individual participants' scores. The posted interval, -1.469 to 1.089, includes no difference and covers values in both directions. The registry record does not describe the procedure used to construct this interval, so readers should not infer details about adjustments, distributional checks, or model structure beyond what is posted.

The reported p-value is unadjusted. In general, a p-value measures how unusual data at least as extreme as those observed would be under a null model and specified analysis assumptions. It does not measure the probability that a finding is true, and it does not quantify clinical importance. A p-value should be read with the estimate and confidence interval, rather than as a stand-alone decision rule. The registry record does not report the significance threshold or a plan for handling multiple outcomes and tests.

The posted comparison also gives a pooled standard deviation and a standardized effect size. In general, standardizing a mean difference expresses it relative to a measure of variability, which can aid comparisons across scales. A standardized effect size does not replace the original score difference: it depends on the variability measure used and can be less intuitive than the scale's units. The posted standardized effect-size estimate is -0.071. The registry record does not report whether this standardized quantity was part of the original analysis plan.

The trial is labelled randomized, double-masked, and factorial. Randomization is intended in general to make treatment groups comparable on average, while masking can reduce the influence of knowledge of treatment assignment on care or assessment. These design labels do not, on their own, establish how allocation was generated or concealed, which people were masked, or whether masking was successful. Those implementation details are not reported in the registry record. Likewise, the factorial label does not reveal what factors were tested or whether the primary comparison was conditional on other factors.

For the primary endpoint, the outcome description says analyses were limited to male participants. The registry record does not report how that restriction was determined, whether it was specified before data analysis, or the analysis population definition beyond that statement. It also does not report the method for handling missing outcome values, whether the analysis followed assigned treatment, or whether any covariates were included. The arm-level summary comparison should therefore be interpreted as the comparison shown, without assuming an unreported adjusted model or missing-data procedure.

The record lists stress reactivity as a secondary continuous outcome, with its own group summaries and unadjusted comparison. In general, a secondary outcome can add information about a related response, but it is not a substitute for the primary endpoint. If multiple endpoints or analyses are examined, the chance of finding an apparently notable result can increase; methods such as prespecifying a hierarchy or adjusting for multiplicity can address that issue. The registry record does not report a multiplicity strategy. It lists no reported analyses in the secondary outcome analysis field, so details of any formal analysis plan for that outcome are not available in the record.

The sample-size illustrations are separate from the trial's original design. Their listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation, with the effect to detect set to a stated fraction of the observed effect. These illustrations describe how planning quantities change under those assumptions. They do not reveal the assumptions used when the trial was designed, which are not reported in the registry record.

11. Confidence Intervals and Precision

The primary estimate is a mean difference of -0.19, with a confidence interval from -1.469 to 1.089. The interval is the most direct posted summary of uncertainty around the contrast. Because it includes no difference and extends in both directions, the result does not precisely determine whether the population contrast is lower or higher for oxytocin relative to control. The point estimate alone gives a single value, but the interval shows that the data permit a broader range of possibilities.

In general, a confidence interval is not a statement that there is a specified probability that the true effect lies in this particular realized interval. Rather, it is constructed using a method with a long-run coverage property when the method's assumptions hold. It is also not a range expected to contain individual participants' scores. Those scores vary within each group, as reflected by the posted standard deviations, while the confidence interval concerns the estimated difference between group means.

The interval's width indicates that the comparison is not highly precise. The small point estimate should therefore not be interpreted as proof that any underlying difference is necessarily small. Equally, the interval does not establish a meaningful difference in either direction. The registry record does not report a prespecified margin for judging what size of score difference would be important, so the interval cannot be classified against such a threshold from this record alone.

The unadjusted p-value is 0.771. It should be considered alongside the interval and estimate, not as a substitute for them. Since the registry record does not report the full analysis plan or multiplicity approach, the posted interval and p-value do not communicate all possible sources of uncertainty, including uncertainty arising from model choices or selection among analyses.

12. 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 table and curve are planning illustrations based on the observed effect, not a record of the trial's original design calculation. The registry record does not report the original sample-size justification or its assumptions. These illustrations assume a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each row sets the effect to detect as a stated fraction of the observed effect: the full observed effect, 75% of it, or 50% of it.

The figures are participants per arm for 80% power and participants per arm for 90% power. They are not total sample sizes. In plain language, the 80% power column shows the modeled number of participants per group for a study with an 80% chance of detecting the specified effect under the listed assumptions, if that effect is real and the model applies. The 90% column makes the analogous illustration for a 90% chance. Higher power generally requires more participants under otherwise unchanged assumptions.

Each row represents a different effect size to detect. A smaller effect is harder to distinguish from random variation, so the illustrations require more participants per arm for the smaller-effect rows. The curve visualizes the same planning relationship across the plotted assumptions; it does not show observed trial outcomes or establish what a future study will achieve. These quantities are not observed power. They should be treated as conditional planning references rather than as a retrospective judgment of whether this trial was adequately designed.

The calculations use the observed effect as their starting point, so the illustrations are sensitive to the uncertainty of that estimate. Here, the primary comparison's confidence interval is broad. A planning exercise based on a single observed estimate cannot remove that uncertainty, and another plausible effect value could imply different planning requirements. The table and curve are therefore most useful for seeing how required sample size changes with the assumed target effect, not for asserting a definitive sample size for a future trial.

13. 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 = -0.190; pooled standard deviation σ = 2.672.
2. n = 2 × (1.96 + 0.842)² × 2.672² / -0.190² = 3,104.9, rounded up to 3,105 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 <- -0.19                      # observed difference in means
sigma <- 2.67214491776521                     # 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(2058251)
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 = 3,105 (80% power), 4,157 (90% power)
 75% of observed effect: n per arm = 5,520 (80% power), 7,390 (90% power)
 50% of observed effect: n per arm = 12,420 (80% power), 16,627 (90% power)
power.t.test: 3105.9 
Simulated power at n = 3,105 per arm: 0.797 (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.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists alcohol craving as the primary outcome and stress reactivity as a secondary outcome. Considering more than one outcome, and potentially more than one analysis of an outcome, raises a multiplicity question: an unadjusted statistical result does not account for the full number of comparisons considered. The registry record reports an unadjusted p-value for the primary comparison, but does not report a plan for controlling multiplicity, such as a testing hierarchy or an adjustment procedure. The secondary result is also presented as unadjusted.

The study is labelled factorial, but the registry record does not identify the factors or report whether interaction tests were planned or conducted. It also does not report whether there were interim analyses, subgroup analyses beyond the male-participant restriction for the main outcome analyses, or other planned comparisons. These details should be treated as not reported in the registry record rather than inferred from the design label.

The outcome description states that main outcome analyses were limited to male participants. The reason for this restriction, its timing relative to analysis, and its relation to the enrolled population are not reported in the registry record. The primary outcome summaries list 32 participants in the oxytocin arm and 35 in control, while enrollment is listed as 73. The registry does not explain how participants moved from enrollment to the outcome summaries. It does not state whether records were absent, participants were excluded, or another analysis-population rule was applied.

Missing-data methods are not reported in the registry record. There is no posted description of how missing outcome measurements were handled, whether any values were imputed, or whether sensitivity analyses were conducted. The record also does not describe an intention-to-treat or per-protocol analysis population. Without those details, readers cannot assess whether the posted comparison includes everyone assigned to a group or how missing observations might affect it.

The safety summaries use group counts that differ from the primary outcome counts, but the registry does not explain the relationship between those populations. These reporting gaps do not prove that the analyses were inappropriate; they mean the registry record does not permit a detailed assessment of multiplicity, missingness, or analysis-population choices.

15. Safety Results

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

Deaths

Other (non-serious) adverse events

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

The registry's safety table lists serious adverse events by group. For the oxytocin group, 35 participants were at risk and 0 were affected by a serious adverse event. For control, 38 participants were at risk and 0 were affected. The listed deaths were 0 in each group. These are the counts recorded in the registry; they are not a comparison adjusted for differences in exposure or follow-up.

A count of zero affected participants means no serious adverse events were recorded in these safety summaries. It does not establish that no adverse events occurred outside the categories or reporting captured here, nor does it establish that an intervention is free of risk. The registry lists no top serious adverse event terms. The record does not report further detail about the event-collection process, follow-up duration for safety, or how safety populations were defined.

The safety group counts differ from the primary outcome group counts. The reason is not reported in the registry record, so the two sets of denominators should not be assumed to describe identical analysis populations. These figures are descriptive registry entries, not evidence by themselves of a causal safety difference between groups.

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

The result is an estimate of the difference in mean self-reported alcohol-craving score between the oxytocin and saline groups in the participants represented in the posted primary outcome summaries. The point estimate favors a lower score in the oxytocin group, but the confidence interval is broad and includes no difference as well as values in the opposite direction. The result therefore offers limited precision about the size and direction of the underlying group contrast.

This does not mean the study found proof of no difference. A result whose interval includes no difference can reflect an effect near none, or uncertainty that is too substantial to distinguish among several possibilities. Nor does it show that a difference exists simply because the point estimate is not exactly no difference. The interval and p-value should be interpreted together, with attention to the analysis details that are not reported.

The endpoint is a short-term self-report measure at the stated 2 hours timeframe. The posted primary outcome does not measure alcohol consumption, continued abstinence, relapse, or longer-term functioning. The study description sets the question in the context of veterans with alcohol use disorder and PTSD and laboratory-induced stress, but the posted primary analyses were limited to male participants. The result should not be extended automatically to other populations, settings, or outcomes.

The registry identifies the trial as Phase 1 and lists enrollment of 73, but does not provide a complete account of the primary analysis population or missing-data procedures. It also does not report the original sample-size assumptions or a multiplicity strategy. Those omissions constrain how confidently the posted estimates can be generalized or evaluated as a complete test of the stated research question.

Accordingly, the result is best read as a descriptive, uncertain comparison in this trial, not as a recommendation about treatment and not as a definitive answer about longer-term substance-use outcomes. It can inform the statistical discussion of the measured endpoint and the precision of its estimate, while leaving broader clinical questions unresolved.

17. Why This Trial Matters Statistically

This trial illustrates why a point estimate should not be read in isolation. The posted primary mean difference is close to no difference, but its confidence interval covers a broad range of possible contrasts. That distinction matters: a small observed difference is not the same as a precise estimate showing that meaningful differences are absent. Reading the estimate together with the interval gives a more complete account of what the data can and cannot say.

The record also highlights the difference between a planned primary outcome and a related secondary outcome. The study description links both alcohol craving and stress reactivity to the response after laboratory-induced stress. Yet the outcomes measure different things, and their results cannot be treated as interchangeable. A secondary result does not independently confirm the primary result, particularly when its comparison is unadjusted and the registry does not report a multiplicity plan.

The sample-size illustrations show another important statistical distinction. They are calculations based on the observed effect under stated assumptions, not evidence about the trial's original design or its observed probability of success. Their main value is to demonstrate how a smaller target effect requires more participants under otherwise unchanged planning assumptions. The primary estimate's uncertainty also cautions against treating one observed effect as a stable planning target.

Finally, design labels and summary numbers do not answer every methodological question. Randomized, double-masked, and factorial describe features of the registered design, but implementation details, analysis populations, missing-data methods, and the original design assumptions are not reported. Recognizing those boundaries is part of sound statistical interpretation: it separates what the posted record shows from what would require information that is not available.

18. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The primary outcome analysis was limited to male participants, according to the endpoint description. The registry record does not explain why, or report enough detail to judge how this restriction relates to the full enrolled population. The finding therefore should not be assumed to represent participants outside the posted analysis population.

The primary outcome summaries include 32 participants in the oxytocin group and 35 in control, compared with a registered enrollment of 73. The registry record does not explain the difference. It also does not report missing-data methods, exclusions, or a fuller analysis-population definition. Safety summaries use 35 and 38 participants at risk, and the relationship between those counts and the endpoint populations is not reported.

The endpoint is a self-reported craving score with a 2 hours timeframe. The registry record does not describe longer-term alcohol use outcomes or establish whether the short-term rating corresponds to later behavior. The study description refers to a laboratory stress task, so the posted result is also tied to that research setting rather than to all circumstances in which craving might occur.

The record identifies the study as factorial but does not report the factors or any interaction analyses. Details of randomization and masking procedures are also not reported. The statistical analysis plan, original sample-size justification, multiplicity handling, and missing-data approach are not reported in the registry record. These gaps limit assessment of how the posted analysis relates to the intended design.

Finally, the primary confidence interval is broad, and the posted p-value is unadjusted. The registry record does not provide a prespecified threshold for a meaningful score difference. Together, these features make the result an uncertain estimate rather than a precise basis for broad conclusions.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Oxytocin Suppresses Substance Use Disorders Associated With Chronic Stress — NCT02058251.
• Registry record last updated 2019-03-14; status last verified 2018-12.
• Results first posted 2019-03-14.
• 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 registry describes NCT02058251, “Oxytocin Suppresses Substance Use Disorders Associated With Chronic Stress,” as a completed, randomized, double-masked, factorial Phase 1 trial sponsored by the Medical University of South Carolina. Enrollment was 73. The study description focuses on veterans with alcohol use disorder and PTSD and asks whether intranasal oxytocin affects alcohol craving and stress reactivity after laboratory-induced stress.

The primary outcome was alcohol craving, measured as a Visual Analogue Scale score with a 2 hours timeframe. The main outcome analyses were limited to male participants. The posted comparison estimates a mean difference of -0.19, with a confidence interval from -1.469 to 1.089 and an unadjusted p-value of 0.771. Stress reactivity is listed as a secondary outcome. The record does not report the original sample-size rationale, multiplicity plan, missing-data methods, or detailed analysis populations.