Preschooler Emotion Regulation in the Context of Maternal Borderline Personality Disorder: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Preschooler Emotion Regulation in the Context of Maternal Borderline Personality Disorder (NCT03060902), a randomized na study. The primary outcome is Maternal Emotion Dysregulation.

2. Clinical Question

This question is about a group-level score comparison at the stated follow-up, not about whether every mother changed in the same way. A difference in group means describes the average separation between groups on the measure; it does not, on its own, establish why that separation occurred. Interpretation also depends on the analysis used, the precision of the estimate, and which participants contributed endpoint data. The registry reports a sponsor analysis and permits an unadjusted recalculation from the posted group summaries. Those approaches yield different statistical conclusions, so the reported analysis must be kept distinct from the recalculation.

3. Trial Design

The registry describes a completed randomized, parallel-group study with no masking. The study title is “Preschooler Emotion Regulation in the Context of Maternal Borderline Personality Disorder,” and the sponsor is the University of Oregon. The registered condition entries are Borderline Personality Disorder and Emotional Problem. The trial phase is listed as NA.

The registry records enrollment of 93. The study began on 2017-10-23, and primary completion is recorded as 2022-05-04. These dates describe the registry record’s study milestones; they do not establish when any individual participant entered or completed the study. The randomized design assigns participants to a group for comparison, while the parallel model means the groups are followed as separate arms rather than participants crossing between interventions. The registry does not report the allocation ratio or the method used to generate or conceal the allocation sequence.

One arm received Dialectical Behavior Therapy Skills, described as a weekly 2.5 hour group session for 1 year. The other arm was Family Services as Usual, with mothers continuing the services they were receiving in the community. The record lists masking as none. Further information about intervention delivery, attendance, adherence, or fidelity is not reported in the registry record.

The posted primary outcome is Maternal Emotion Dysregulation, measured at 1 year. Its endpoint population description says 37 Dialectical Behavior Therapy Skills participants and 40 Family Services as Usual participants completed the final time point. These endpoint counts are not the same as a complete account of who was randomized, excluded from analysis, or included in each analysis. Those details are not reported in the registry record. Although the rationale discusses children’s emotion regulation, the registry lists one outcome measure and one analysis, and the posted primary outcome is maternal rather than child-focused.

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: 37 DBT randomized participants completed the final time point and 40 Family Services as Usual participants completed the final time point.

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 Maternal Emotion Dysregulation, measured over 1 year.

The primary endpoint is Maternal Emotion Dysregulation at 1 year, recorded as a mean score on a scale. The measure is the Difficulties with Emotion Regulation Scale, with a minimum score of 36 and a maximum score of 180. Higher scores indicate greater difficulties with emotion regulation. The group summaries are reported as means with standard deviations.

A continuous score retains gradations across the measure’s range rather than reducing participants to a binary category. Comparing group means can therefore describe the average level of reported difficulty in each arm at the endpoint. The standard deviation describes the spread of scores within each group; it is not the uncertainty around the mean difference.

The measure is relevant to the intervention’s stated rationale because the study proposes examining whether changes in mothers’ ability to regulate emotions are associated with children’s emotion regulation. A maternal endpoint addresses the maternal part of that rationale directly. It does not, by itself, establish a change in children’s emotion regulation or demonstrate that a maternal score change is transmitted to children. The registry’s listed outcome is a measure of maternal emotion dysregulation, and further details about validation, scoring procedures, or clinically meaningful thresholds are not reported in the registry record.

7. Results Posted in the Registry

• Dialectical Behavior Therapy Skills: mean 82.35 versus 91.80 on control; difference -9.45 (95% CI -19.92 to 1.02), standardized effect d = -0.40; z-test p = 0.077.

8. Statistical Interpretation

9. Statistical Methods Explained

What does a mean difference represent?

General statistical explanation: A mean difference subtracts one group’s average score from the other group’s average score, using the stated group order. It expresses the result in the original scale’s units, so readers can relate the estimate to the outcome measure. A negative estimate does not inherently mean a better or worse result; its interpretation depends on which arm is subtracted from which and what higher scores mean. In this registry comparison, higher scores indicate greater difficulties with emotion regulation, and the recalculated difference is reported as -9.45 for Dialectical Behavior Therapy Skills relative to Family Services as Usual.

A mean difference is a summary of group averages, not a description of every participant. Participants in a group can have scores above or below that group’s mean. The standard deviations, 21.37 for Dialectical Behavior Therapy Skills and 25.45 for Family Services as Usual, describe that within-group spread. They do not show the standard error or confidence interval for the difference by themselves.

What does the confidence interval add?

General statistical explanation: A confidence interval conveys uncertainty around an estimated effect. Wider intervals indicate less precision, while narrower intervals indicate greater precision, given the method used. A 95% confidence interval is often described through the repeated-sampling behavior of the method: across repeated studies under its assumptions, intervals constructed in this way would contain the target value at the stated rate. It is not a probability statement that the particular interval contains the true value.

The unadjusted recalculated mean difference has a 95% confidence interval from -19.92 to 1.02. Since zero lies within this interval, the recalculated result does not exclude no mean difference at that confidence level. The interval also conveys that the summary estimates do not pin down a single precise value. The sponsor analysis does not have a confidence interval posted in the registry record, so the precision of that analysis cannot be assessed from a corresponding reported interval.

Why can p-values differ?

General statistical explanation: A p-value is calculated relative to a specified statistical model and null hypothesis. Different models, covariate adjustments, assumptions about variance, analysis populations, or handling of missing outcomes can produce different standard errors and p-values, even when they address a related group comparison. A p-value is not the probability that the null hypothesis is true, and it does not measure the probability that a result arose “by chance.” It also does not state the size or practical importance of a difference.

The registry reports an ANOVA result of F = 10.16 with p = .002. Separately, the recalculation based on the posted group summaries reports p = 0.077 and a confidence interval for the unadjusted mean difference. These are not interchangeable outputs. The registry does not provide the sponsor model’s full specification, so the reason for the difference is not reported in the registry record. The record marks a discrepancy; accordingly, the sponsor’s own analysis takes precedence, while the recalculation remains a descriptive check.

What is ANOVA?

General statistical explanation: Analysis of variance, or ANOVA, is a method for testing whether group means differ under a specified model. Its F statistic compares variation attributed to the group effect with variation used to represent residual differences under that model. The value of the F statistic is interpreted together with the model and its reference distribution. A reported F statistic and p-value alone do not reveal all model choices or establish how missing observations were handled.

For this study, the registry labels the reported analysis as ANOVA and gives F = 10.16 and p = .002. It does not report additional details such as covariates, model diagnostics, degrees of freedom, or the precise analysis population. Those details should not be inferred from the method label alone.

Why do assumptions matter?

General statistical explanation: Statistical estimates and tests rely on assumptions about the data and the model. For continuous outcomes, relevant considerations can include the distribution of scores, the relationship between variability and group membership, independence of observations, and whether the analysis appropriately represents repeated observations or other design features. A method can be useful while still depending on assumptions that need to be assessed. The registry does not report diagnostics or enough detail to judge these assumptions for the sponsor’s ANOVA.

The recalculation is explicitly unadjusted and uses posted summary statistics. It does not incorporate participant-level information or an adjusted model. Consequently, it should be understood as a simple comparison reconstructed from the available summaries, not as a reproduction of the sponsor’s analysis. The registry’s discrepancy flag is important because the sponsor and recalculation do not give the same statistical conclusion.

What does a standardized effect mean?

General statistical explanation: A standardized effect expresses a group difference relative to a measure of outcome variability, rather than in the original scale’s units. Standardization can help compare effects measured on different scales, but it can be less intuitive than the mean difference and depends on how variability is defined. The registry record includes a standardized-effect calculation, but it is not necessary to interpret the primary result in original scale units. No additional standardized value is needed to describe the posted mean difference, its confidence interval, or the sponsor analysis.

Overall, the statistical methods should be read as distinct layers: the endpoint is a scale score, group summaries describe the observed endpoint data, the unadjusted recalculation estimates a mean difference with uncertainty, and the sponsor reports an ANOVA analysis. The registry record does not report enough detail to make the recalculation equivalent to the sponsor’s model or to resolve why their results differ.

10. Confidence Intervals and Precision

Precision concerns how tightly an estimate is determined, rather than whether the estimate is desirable or important. The unadjusted recalculated mean difference is -9.45, and its 95% confidence interval extends from -19.92 to 1.02. That interval is fairly broad relative to the point estimate and includes zero. It therefore leaves uncertainty about the magnitude and direction of the underlying mean difference under the recalculation’s assumptions.

The interval is on the original scale-score metric, which makes it interpretable in relation to the Difficulties with Emotion Regulation Scale. Its lower and upper limits represent a range of values compatible with the unadjusted analysis at the stated confidence level. The point estimate is not the only plausible value indicated by the analysis. Because the interval includes values on either side of zero, the recalculation does not precisely distinguish a negative mean difference from no difference or a positive one.

The standard deviations are 21.37 for the Dialectical Behavior Therapy Skills group and 25.45 for the Family Services as Usual group. These describe variability among endpoint scores, not the uncertainty in the sponsor’s estimated effect. The registry does not provide a confidence interval for the sponsor’s ANOVA result. Thus, the sponsor’s reported p-value cannot be paired with a reported sponsor confidence interval to assess the precision of that analysis from the registry record.

Precision is also limited by what is reported about the endpoint population and analysis. The registry states that 37 and 40 participants, respectively, completed the final time point, but it does not report a full participant flow or missing-data strategy. The recalculation uses the posted summaries and should not be treated as adjusted for unreported features. The disagreement between the recalculation and the sponsor analysis is an additional reason to keep the inferential conclusions separate. It does not erase either result; it means the model and analysis details are needed to understand their difference.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm, not total events. They are illustrations based on the observed effect, not a statement of the trial’s original enrollment target or design rationale. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The target effect is set to the stated fraction of the observed effect.

The 80% power column gives the number of participants per arm that the planning calculation indicates would be needed to have an 80% chance of detecting the assumed effect under those assumptions, if the effect and assumptions were correct. The 90% power column gives the corresponding per-arm planning figure for a higher target chance of detection. These are design calculations, not guarantees about a future study’s results.

The planning illustration lists 98 participants per arm for 80% power and 131 participants per arm for 90% power when planning for 100% of the observed effect. For 75% of the observed effect, it lists 174 participants per arm for 80% power and 233 participants per arm for 90% power. For 50% of the observed effect, it lists 391 participants per arm for 80% power and 524 participants per arm for 90% power. Smaller assumed effects require more participants because they are harder to distinguish from ordinary variation under the same planning framework.

These figures are not observed power. They do not say what power the completed study had, and they should not be interpreted as evidence that a result is or is not reliable. The registry’s original sample-size justification and design assumptions are not reported in the registry record. The planning illustration is a retrospective calculation based on the observed effect and its stated assumptions, not a reconstruction of the sponsor’s original planning.

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 = -9.450; pooled standard deviation σ = 23.580.
2. n = 2 × (1.96 + 0.842)² × 23.580² / -9.450² = 97.7, rounded up to 98 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 <- -9.45                      # observed difference in means
sigma <- 23.5798687867426                     # 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(3060902)
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 = 98 (80% power), 131 (90% power)
 75% of observed effect: n per arm = 174 (80% power), 233 (90% power)
 50% of observed effect: n per arm = 391 (80% power), 524 (90% power)
power.t.test: 98.7 
Simulated power at n = 98 per arm: 0.792 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

13. Analyses Reported by the Sponsor

The registry’s sponsor-reported analysis is an ANOVA with F = 10.16 and p = .002. The separate recalculation uses the posted group means, standard deviations, and endpoint counts to estimate an unadjusted mean difference of -9.45, with a 95% confidence interval from -19.92 to 1.02 and p = 0.077. These results do not align statistically: the recalculated interval includes zero and its p-value differs from the sponsor’s reported p-value.

The record marks this as a discrepancy. The sponsor’s own analysis takes precedence over the recalculation. The recalculation is not an alternative sponsor analysis, nor does it explain why the two results differ. It is an unadjusted reconstruction from summary information, while the registry does not report the sponsor’s full model specification, analysis population, or missing-data handling. Without those details, the sponsor analysis cannot be reproduced from the registry record.

The two results should therefore be reported with their methods attached. The ANOVA result is the sponsor’s reported analysis; the mean difference and interval are an unadjusted recalculation. Neither should be silently substituted for the other, and the recalculation should not be used to override the sponsor result.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists one outcome measure and one analysis. It lists no secondary outcomes as posted. With only one listed outcome measure, the record does not show a broader set of registered endpoint tests for which multiplicity adjustment can be assessed. However, the registry does not report whether additional outcomes or analyses were considered during study conduct, whether any interim analyses occurred, or whether the analysis was changed from a prespecified plan. Those details are not reported in the registry record.

Multiplicity is the issue that arises when multiple hypotheses are tested: conducting many tests can increase the chance of at least one apparently notable result. Whether an adjustment is appropriate depends on the questions and analysis plan. Since the record does not describe interim monitoring or a complete prespecified analysis plan, it is not possible to evaluate those aspects from the posted record alone. No interim analysis should be assumed.

The endpoint population description reports final-time-point completion counts of 37 for Dialectical Behavior Therapy Skills and 40 for Family Services as Usual. These counts indicate who completed the stated time point according to the record. They do not establish that every randomized participant was included in the sponsor analysis, nor do they identify a complete analysis population. The registry does not report a participant flow, reasons for missing endpoint data, or whether missing outcomes were imputed.

Missing-data handling can affect estimates when the observed outcomes do not adequately represent outcomes that are missing. In general, statistical approaches make different assumptions about the relationship between missingness and observed or unobserved outcomes. The registry does not state which assumptions or method, if any, were used. It also does not report whether analyses followed intention-to-treat principles or used another analysis population. These omissions limit assessment of how the reported findings relate to all randomized participants.

The discrepancy between the sponsor analysis and the unadjusted recalculation adds a separate interpretive limitation. The registry identifies the sponsor method as ANOVA but does not provide enough detail to determine whether differences in model specification, data inclusion, or missing-data handling account for the results. It would be inappropriate to select one explanation without additional information.

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 reports serious adverse event figures for each arm. In the Dialectical Behavior Therapy Skills group, 46 participants were listed as at risk for serious adverse events, with 0 serious adverse events reported and 0 deaths. In the Family Services as Usual group, 47 participants were listed as at risk, with 0 serious adverse events reported and 0 deaths.

These entries describe the registry’s posted serious adverse event counts. They should not be interpreted as proof that no adverse experiences of any kind occurred. The registry lists no serious adverse event terms, but the meaning of the at-risk population, event collection procedures, follow-up coverage, and definitions used are not further detailed in the registry record. The serious-event figures also do not provide a complete description of non-serious events or other safety outcomes.

When reading these figures, distinguish the event count from the number listed at risk. The record reports zero serious events and zero deaths in each group; it does not establish that the groups had identical safety experiences in every respect. No medical advice or treatment conclusion follows from these registry entries alone.

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

The posted results describe a randomized comparison of Dialectical Behavior Therapy Skills with Family Services as Usual on maternal emotion dysregulation at 1 year. The group means were 82.35 and 91.8, respectively, and higher scale scores indicate greater difficulties with emotion regulation. The unadjusted recalculation reports a mean difference of -9.45, with a 95% confidence interval from -19.92 to 1.02. The sponsor reports an ANOVA result of F = 10.16 and p = .002.

The two statistical accounts should not be merged into a single conclusion. The unadjusted recalculation does not exclude zero at its reported confidence level, while the sponsor’s reported p-value is different. The registry flags a discrepancy, and the sponsor’s own analysis takes precedence. The sponsor’s model details and confidence interval are not reported in the registry record, which limits independent assessment of that analysis. The recalculation offers a summary-based estimate but is not a substitute for the sponsor analysis.

These results do not, by themselves, establish the size or practical importance of any effect. A p-value does not measure how much an outcome changed, and the unadjusted interval includes a range of possible values. The scale measures maternal emotion dysregulation; interpretation should remain tied to that endpoint rather than be extended to broader mental health outcomes that are not reported.

The study rationale emphasizes children’s emotion regulation and the possibility of a pathway from maternal regulation to child outcomes. The registry’s posted primary outcome is maternal emotion dysregulation, and it lists no secondary outcomes. The reported material therefore does not establish whether preschool children’s emotion regulation changed, whether maternal change caused any child change, or how much improvement in children could be attributed to treating mothers.

Finally, the endpoint completion counts do not describe all aspects of participation or analysis. The registry does not report the full analysis population, missing-data methods, or reasons participants may not have completed the final time point. Those gaps, together with the lack of detail on the sponsor’s model, should shape how confidently the reported results are generalized beyond the posted endpoint comparison.

17. Why This Trial Matters Statistically

This trial raises a useful statistical distinction between a study’s motivating question and its registered primary endpoint. The rationale concerns whether maternal emotion regulation might relate to the development of emotion regulation in preschool children, while the posted primary outcome is a maternal scale score. Keeping those questions separate prevents a result on the mother’s score from being treated as a direct result on a child outcome.

It also illustrates why an estimate, its uncertainty, and the analysis method all matter. The unadjusted recalculation supplies a mean difference on the scale’s original metric and a confidence interval. The sponsor reports an ANOVA with a different p-value. A single p-value cannot explain the size or precision of a result, and a recalculated summary comparison cannot stand in for an analysis whose model details are not reported. The registry’s discrepancy flag makes that distinction especially important.

The trial also demonstrates that design and reporting information are part of statistical interpretation. Randomization and a parallel-group model describe the broad comparison, but the registry does not provide details on allocation procedures, adherence, analysis populations, missing-data methods, or the sponsor’s full model. Without such information, readers cannot fully judge how the analysis corresponds to the randomized groups or why the two statistical results differ.

Finally, the planning illustration is useful only when understood as a design exercise based on the observed effect and explicit assumptions. It is not observed power and does not reveal the original rationale for enrollment. Separating planning calculations from analysis of completed data helps keep the questions of study design and result interpretation distinct.

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.

Several trial-specific limitations constrain interpretation of the registry record. The primary endpoint is a maternal emotion dysregulation score, while the study rationale emphasizes preschool children’s emotion regulation. No child outcome is listed among the posted results, so the maternal endpoint cannot answer the child-focused question by itself.

The sponsor analysis and the unadjusted recalculation differ. The registry reports the sponsor’s ANOVA result but not its complete model specification or a confidence interval. The recalculation is based on posted summary statistics and endpoint counts, so it cannot establish the reason for the discrepancy. The sponsor analysis takes precedence, but its reported detail is insufficient to reproduce the analysis from the registry record.

The record says 37 and 40 participants completed the final time point, but it does not give a complete participant flow, reasons for non-completion, the exact analysis population, or missing-data procedures. It also does not report intervention attendance, adherence, or fidelity, or detailed characteristics of the participants and community services. The absence of these details limits assessment of how the assigned intervention was delivered and how well endpoint data represent the randomized groups.

Additional protocol and analysis information is not reported in the registry record. This includes the original sample-size justification, allocation procedure, interim analysis plans, and a detailed prespecified statistical plan. The registry lists no secondary outcomes, which limits the outcome picture available from the posted results. These reporting gaps should not be filled by assumptions; they define what can and cannot be concluded from the record.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Preschooler Emotion Regulation in the Context of Maternal Borderline Personality Disorder — NCT03060902.
• Registry record last updated 2024-01-12; status last verified 2024-01.
• Results first posted 2024-01-12.
• 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

NCT03060902 is a completed randomized, unmasked, parallel-group study sponsored by the University of Oregon. It compares Dialectical Behavior Therapy Skills with Family Services as Usual in a study concerning maternal Borderline Personality Disorder and preschooler emotion regulation. The intervention is described as a 2.5 hour group session/week for 1 year; the comparison group continued community services.

The posted primary outcome is maternal emotion dysregulation at 1 year, measured with the Difficulties with Emotion Regulation Scale. The sponsor reports an ANOVA result of F = 10.16 and p = .002. An unadjusted recalculation from group summaries estimates a mean difference of -9.45, with a 95% confidence interval from -19.92 to 1.02 and p = 0.077. Because the record flags a discrepancy, the sponsor’s analysis takes precedence. The registry does not report a child outcome among the posted results or enough model and missing-data detail to resolve the difference.