Postpartum Weight Loss for Women at Elevated Cardiovascular Risk: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Postpartum Weight Loss for Women at Elevated Cardiovascular Risk (NCT04914819), a randomized na study. The primary outcome is Change in Weight in Pounds From Baseline to 16-week Follow-up.

2. Clinical Question

The central statistical question is whether the posted results estimate a different mean weight change between the online-program and usual-care arms, and how much uncertainty accompanies that estimate. The registry reports a mixed models analysis of a mean difference in final values, with a 95% confidence interval and p-value. The result should be read as an estimate from this study, not as a guarantee about what would occur in another group of postpartum women. The registry record does not report the detailed eligibility criteria, recruitment-strategy results, or the full analysis plan.

3. Trial Design

The registry identifies this as a randomized, parallel-group study with no masking. It lists an enrollment of 89 and a completed status. The recorded start date is 2021-06-18, and the primary completion date is 2021-12-31. The study is named the EmPOWER-Mom Study and was sponsored by the University of Pennsylvania. Its phase is listed as NA.

Participants assigned to online behavioral weight loss were to attend a virtual introduction and take part in a 16-week online program. They also completed a virtual weigh-in at the beginning and end of the study. Participants assigned to usual care completed the same virtual weigh-ins but received no additional intervention. The registry identifies the primary measurement period as baseline and 16 weeks.

The stated study aim included testing feasibility and effectiveness of the online program compared with usual care among postpartum women with cardiovascular risk factors. The brief summary also describes testing behavioral approaches to recruitment: two email strategies and two mailer strategies, informed by behavioral design. The registry record does not report results for those recruitment comparisons in the registry record described for the primary endpoint.

Three secondary outcomes are named: study completion rate between arms, number of weeks with one or more logins in the intervention arm, and number of weeks with daily data reporting for at least five of seven days in the intervention arm. The record lists seven outcome measures and four analyses overall. It does not report the allocation procedure, concealment methods, detailed visit schedule, or the number randomized to each arm as a design target. Arm counts attached to the posted endpoint are 40 for online behavioral weight loss and 20 for usual care; these are not the same as the total enrollment figure, and the record does not explain the difference.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: 190 individuals signed up for an info session. Of the 190, 115 attended the info session. Excluded: 75 individuals did not attend info session. Of the 115, 89 provided consent and enrolled in run-in period. Excluded: 26 individuals who did not provide consent. Of the 89 enrolled, 60 completed run-in period and were randomized to one of two arms. Note: 27 people signed up for an info session after receiving a recruitment mailer but were excluded due to recruitment goals being met.

Overall Study

Reasons for not completing this period:

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Change in Weight in Pounds From Baseline to 16-week Follow-up, measured over Baseline and 16 weeks.

The primary endpoint is mean change in weight, measured in pounds, from baseline to 16-week follow-up. It is a continuous measure: each participant’s weight can take a range of values rather than being limited to a yes-or-no outcome. A mean summarizes the average change across the participants contributing data to an arm. Negative changes indicate weight values lower at follow-up than at baseline, while positive changes would indicate higher values.

Comparing mean change between randomized arms is relevant to the study’s stated question about a behavioral weight-loss program versus usual care. It expresses the average difference in the measured outcome rather than the number of participants who crossed a threshold. The registry’s primary endpoint description specifies a comparison between arms, while the reported analysis is described as a mixed models analysis of mean difference in final values. Those descriptions are related but not identical formulations; the record does not provide enough detail to reconstruct the model or determine precisely how baseline measurements entered it.

Weight change was assessed at baseline and 16 weeks, matching the program’s stated duration. A single continuous endpoint can convey the direction and estimated size of average change, but it does not show the full distribution of individual changes, variation in program engagement, or whether the change was sustained after follow-up. Those outcomes are not established by this endpoint.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

General statistical context: A randomized comparison is designed to make the treatment groups comparable on average at the start of a study, so differences in outcomes can be interpreted in relation to assignment. Randomization does not ensure that every possible source of imbalance disappears, particularly in a study with limited enrollment. The registry says allocation was randomized and the model was parallel, but does not report the randomization method or allocation concealment.

What does a mean change describe?

General statistical context: For a continuous endpoint, a mean is the arithmetic average of observed values. Mean change uses the difference between a participant’s follow-up and baseline measurements and then summarizes those changes across participants. It gives a group-level description; it does not imply that every participant changed by the mean amount. The registry reports arm means of -8.6 pounds for online behavioral weight loss and -3.3 pounds for usual care. Their confidence intervals describe uncertainty around those arm-level estimates as reported in the record.

What is the between-arm estimate?

The registry reports a mixed models analysis with a mean difference in final values of 5.3. It gives a 95% confidence interval from 0.04 to 10.8 and p = 0.048. General statistical context: A model-based comparison can account for the structure of the measurements or other specified features, depending on the model. The model description alone does not reveal which features were included. The registry record does not report the full equation, covariates, correlation structure, estimation method, or assumptions used in this mixed model. Consequently, the reported estimate can be interpreted as the sponsor’s analysis, but its detailed construction cannot be evaluated from the record alone.

How should the confidence interval be read?

General statistical context: A confidence interval expresses uncertainty in an estimate under a specified statistical procedure. A 95% confidence interval is not a range containing 95% of individual outcomes, and it is not a statement that there is a 95% probability that the particular true effect lies inside the observed interval. Rather, the procedure is designed to produce intervals that cover the underlying value at the stated rate over repeated studies when its assumptions hold. For this result, the reported interval extends from 0.04 to 10.8. Its width reflects uncertainty about the size of the between-arm estimate. It is important to retain the endpoints as reported and not substitute an arm-level interval for the model-based comparison interval.

What does the p-value say?

General statistical context: A p-value measures how unusual data at least as incompatible with a specified null hypothesis would be if that hypothesis and the analysis assumptions were true. It is not the probability that the null hypothesis is true and does not measure clinical importance. The registry’s reported p-value is 0.048. This is near the commonly used 0.05 threshold, but a threshold does not transform the estimate into a certain or clinically meaningful result. The estimated size and confidence interval remain essential to interpretation.

Why can a mixed model be used?

General statistical context: Mixed models can represent outcomes measured repeatedly or data with grouping, incorporating fixed effects and, depending on the specification, random effects. Such models can make use of available measurements under particular assumptions about the data and missingness. The registry names a mixed models analysis but does not report whether it used repeated measures, what random effects were included, how missing observations were handled, or what assumptions were adopted. Those details cannot be inferred from the method label.

What is over-dispersion?

General statistical context: Over-dispersion occurs when observed variability is greater than a model assumes. It is especially relevant to certain count-data models, where underestimated variability can make standard errors too small and confidence intervals too narrow. The primary endpoint in this study is continuous weight change, and the registry does not say that a count model or over-dispersion adjustment was used. Over-dispersion should therefore not be presented as a feature of the reported primary analysis; it is a general concept that may matter in other statistical settings.

The registry supplies no sample-size justification or detailed statistical analysis plan. A planning illustration based on an observed effect, if presented among the statistical displays, is a hypothetical design calculation and not observed power. Such an illustration cannot establish the study’s original design rationale or show the probability that the study would detect an effect. The registry record does not report that rationale.

10. Analyses Reported by the Sponsor

The sponsor’s own analysis is reported as a mixed models comparison of mean final values, with an estimated mean difference of 5.3, a 95% confidence interval from 0.04 to 10.8, and p = 0.048. The registry also posts separate mean changes from baseline for each arm: -8.6 pounds in online behavioral weight loss and -3.3 pounds in usual care. These are distinct summaries, and the sponsor’s model-based comparison should not be reconstructed by subtracting or otherwise combining the arm summaries.

The record does not flag a discrepancy between the reported analysis and a recalculation. The sponsor’s reported analysis is therefore the available inferential result, while the posted arm-level values provide descriptive context. The analysis is labeled superiority, but the estimate and interval still need to be interpreted with their uncertainty and with the limitations of the reported methods. The registry record does not include enough model detail to independently reproduce the sponsor’s result or assess every modeling assumption.

The reported p-value is near the conventional 0.05 reference point, and the lower confidence-limit is close to zero. The estimate should not be reduced to a binary significant-or-not label. The interval communicates that the size of the mean difference is uncertain, while the arm-level summaries indicate the direction of average changes within each group.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists seven outcome measures and four analyses. It names three secondary outcomes: study completion rate between arms, intervention-arm weeks with one or more logins, and intervention-arm weeks with daily data reporting on at least five of seven days. The primary endpoint is mean weight change from baseline to 16-week follow-up. The other listed outcome measures and the relationship between all outcomes and analyses are not described in the registry record from the registry record.

General statistical context: When several outcomes or comparisons are tested, the chance of at least one small p-value can increase. A prespecified approach to multiplicity can address that issue, for example by adjusting the threshold or controlling an error rate across tests. The registry record does not report whether any multiplicity adjustment was planned or applied, whether the primary endpoint was designated before analysis, or how the analyses relate to the seven outcome measures. The p-value for the primary comparison should therefore be read as reported, without assuming it was adjusted or that it was the only test conducted.

The enrollment is 89, while the primary endpoint counts are 40 for the online-program arm and 20 for usual care. The registry does not explain this difference or report a participant-flow account that identifies randomization, endpoint availability, exclusions, or analysis inclusion. It is not possible to determine from these counts alone whether the primary analysis was intention-to-treat, per-protocol, or based on another analysis population. The population used for the mixed models analysis is not reported in the registry record.

Missing-data handling is also not reported in the registry record. General statistical context: Missing outcomes can affect an estimate when the chance of being observed is related to participants’ characteristics or outcomes. Different methods rely on different assumptions, and a mixed-model label does not by itself establish how missing data were handled or whether those assumptions are appropriate. The record does not report the amount or pattern of missing primary outcome data, imputation methods, or sensitivity analyses. These omissions limit assessment of how robust the reported comparison is to incomplete follow-up.

12. Safety Results

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

Deaths

Other (non-serious) adverse events

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

The registry reports zero serious adverse events among 40 participants at risk in the online behavioral weight-loss group and among 20 at risk in usual care. It also reports zero other adverse events in those two groups. For the run-in phase, 89 participants were at risk, with zero serious and zero other adverse events. No deaths are reported in any of the three groups. The record lists no top serious adverse-event terms.

These figures describe events recorded in the registry for the stated groups and at-risk counts. They do not show that no events of any kind occurred outside the categories captured in the report, or establish that an intervention is free of risk. Zero recorded events in a study does not rule out events that are uncommon or that were not captured or reported. The registry record does not give event definitions, the adverse-event collection period, severity criteria beyond the serious/other categories, or details of monitoring.

The run-in count should be kept separate from the two endpoint arm counts. The record does not explain how the run-in group relates to the randomized arms. These safety figures do not provide a comparison adjusted for exposure time, and the registry does not report exposure-time estimates. They are best read as the posted counts, with the limits of the available safety-reporting detail.

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

The posted primary result is an estimate of average weight change between two randomized study arms over the baseline-to-16-week measurement period. Mean change was -8.6 pounds in the online behavioral weight-loss arm and -3.3 pounds in usual care. The sponsor’s mixed models analysis reported a mean difference in final values of 5.3, with a 95% confidence interval from 0.04 to 10.8 and p = 0.048. Taken together, the posted estimates point toward greater average weight reduction in the online-program arm during the study period.

The confidence interval shows that the exact size of the between-arm estimate is uncertain. The lower bound is close to zero and the upper bound is substantially larger, so the estimate should not be described as precise. The p-value is not a measure of the probability that the program is effective, nor does it determine whether the size of the difference matters to participants. The registry does not report a threshold for a clinically meaningful change in weight.

The result does not establish that each participant lost weight, that all participants engaged with the program, or that any weight change continued beyond 16 weeks. The endpoint is a group mean, and the registry record does not provide individual outcome distributions or longer-term follow-up results. It also does not answer the separate feasibility questions about completion and program use, despite listing those as secondary outcomes.

The result should not be generalized automatically to all postpartum women. The registry describes a population with cardiovascular risk factors but does not provide detailed eligibility criteria or enough information about recruitment and participant characteristics to judge how closely the enrolled group represents other populations. In addition, enrollment was 89, while endpoint arm counts were 40 and 20, and the reasons for those counts are not reported. That uncertainty about the analyzed population matters when assessing how broadly to interpret the estimate.

In short, this is a randomized comparison with a reported model-based result favoring greater average reduction in the online-program arm, accompanied by considerable uncertainty about the estimate’s size and incomplete reporting of analysis details. It is not proof of a lasting effect, a statement about individual outcomes, or a basis for medical advice.

14. Why This Trial Matters Statistically

This study brings a continuous endpoint into a randomized comparison of a postpartum behavioral program and usual care. The mean weight-change estimates convey the direction and average magnitude of observed change in each arm, while the mixed models estimate and its confidence interval address the between-arm comparison. Keeping those summaries distinct is an important statistical practice: within-arm change and between-arm difference answer related but different questions.

The study also illustrates why estimates should be read with uncertainty rather than reduced to a p-value. The reported p-value is 0.048, but the 95% confidence interval extends from 0.04 to 10.8. The interval communicates a range of plausible effect sizes under the model and reminds readers that the evidence does not pin down the size precisely. The point estimate alone would conceal that uncertainty, while the p-value alone would not describe the estimated effect.

Feasibility is part of the stated aim, and the registry lists completion and online engagement outcomes. These outcomes could answer questions different from weight change, such as whether participants completed the study or used the program. The primary weight endpoint does not substitute for those measures. The record lists three secondary outcomes but does not provide their results in the registry record.

The trial also highlights the importance of transparent reporting. Enrollment, endpoint arm counts, and the run-in safety count are not interchangeable, and the record does not explain their relationship. More detail about participant flow, analysis populations, missing outcomes, model specification, and multiplicity would help readers assess the strength and scope of the primary result. These are not merely technical details: they determine which participants and comparisons the estimate represents.

15. 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 registry record does not report the detailed eligibility criteria, baseline characteristics, or the method used to define cardiovascular risk. This limits assessment of how the enrolled participants compare with other postpartum populations. It also does not report the number assigned to each arm as a planned design target or explain why the posted primary endpoint counts are 40 and 20 when enrollment is 89.

The primary comparison is described as a mixed models analysis, but the model specification is not reported in the registry record. Covariates, repeated-measure structure, assumptions, and handling of missing observations are not available. The record also does not state which analysis population was used, whether all randomized participants were included, or whether sensitivity analyses were conducted. These gaps make independent reproduction and robustness assessment impossible from the reported information.

The registry lists seven outcome measures and four analyses, but does not report the full outcome set or an approach to multiplicity. It is unclear from the record how the listed analyses relate to the outcomes or whether the primary comparison was selected before looking at results. The sample-size justification and detailed statistical analysis plan are not reported in the registry record.

The posted endpoint covers baseline to 16-week follow-up. Longer-term outcomes are not reported in the registry record, so the primary result cannot establish whether weight changes continued. The separate feasibility and recruitment aims are also not resolved by the primary endpoint statistics. The record describes planned email and mailer recruitment strategies but does not provide their comparative results in the registry record relevant to this analysis.

Safety reporting lists zero serious and zero other adverse events in the named groups, but definitions, collection details, and exposure-time information are not reported. The absence of recorded events in this dataset should not be interpreted as proof that no risks exist.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Postpartum Weight Loss for Women at Elevated Cardiovascular Risk — NCT04914819.
• Registry record last updated 2023-07-06; status last verified 2023-06.
• Results first posted 2023-07-06.
• 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.

19. Record Summary

The EmPOWER-Mom Study was a completed randomized, parallel-group trial sponsored by the University of Pennsylvania. It compared a 16-week online behavioral weight-loss program with usual care among postpartum women described as having cardiovascular risk factors. Enrollment was 89, and the registry reports no masking.

The primary endpoint was mean weight change in pounds from baseline to 16-week follow-up. Posted arm means were -8.6 pounds for online behavioral weight loss and -3.3 pounds for usual care. The sponsor reported a mixed models mean difference in final values of 5.3, with a 95% confidence interval from 0.04 to 10.8 and p = 0.048.

The registry also lists feasibility and engagement outcomes and recruitment-strategy aims. The detailed statistical plan, missing-data methods, analysis population, and reasons for the difference between enrollment and endpoint counts are not reported in the registry record. The primary result suggests a between-arm difference in average change, but its size and broader interpretation remain uncertain.