Girls on the Move Intervention to Increase Physical Activity Among Middle School Girls: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Girls on the Move Intervention to Increase Physical Activity Among Middle School Girls (NCT01503333), a randomized na study. The primary outcome is Minutes of Moderate to Vigorous Physical Activity (MVPA) Post-intervention.

2. Clinical Question

The trial also named cardiovascular fitness, body mass index z-score, and percent body fat as secondary outcomes measured after the 17-week intervention. Those outcomes address related but distinct questions: whether the intervention groups differed in aerobic performance and in two body-composition measures at that assessment. The registry summary also states hypotheses concerning activity at a later follow-up, but the primary outcome analyzed in the posted results is the post-intervention activity measure. Results for a later follow-up are not reported in the registry record.

The primary question is about a group-level difference in an activity measure, not about whether every participant changed in the same way. The posted arm means describe average activity for the two groups. A comparison estimate and its confidence interval describe the estimated separation between those averages and the uncertainty around it. They do not, by themselves, establish why any observed group pattern occurred or describe each girl’s individual experience.

3. Trial Design

The registry describes a randomized, parallel-group trial of the Girls on the Move intervention among middle school girls. The listed enrollment is 1543, and the trial status is completed. The record identifies Michigan State University as sponsor. The listed phase is not applicable. The registry record gives a start of 2011-09 and primary completion of 2016-05.

Girls assigned to the physical activity intervention received a school nurse counseling component, tailored feedback from a computer program, and an after-school physical activity club. The summary describes the overall intervention as 17 weeks. The control group completed data collection activities and received usual school offerings. Masking is listed as none; the record does not describe blinding procedures beyond that designation.

The trial used a parallel model, so the two groups were compared as separate intervention and control arms. The posted primary outcome summaries include 680 participants in control and 706 in the intervention arm. These are the group counts attached to the reported primary outcome, and the record does not establish that these counts represent every randomized participant or an intention-to-treat analysis population.

Physical activity was assessed after the intervention using ActiGraph GT3X+ accelerometers worn at the right hip for 7 consecutive days. The registry describes data collection across weekdays and weekend days, processing into 15-second epochs, and use of established intensity cut-points. It also reports an imputation approach based on available hour-block data and standardization of wear time. Those details help define how the measured endpoint was constructed, but the registry does not provide a full analysis protocol or enough information to reconstruct all decisions in the reported mixed-model analysis.

The registry lists 11 outcome measures and 11 analyses. Three secondary post-intervention outcomes are posted in the record. A detailed sample-size justification, the allocation procedure, the intended analysis population, and a prespecified statistical analysis plan are not reported in the registry record.

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: The majority (1386 \[post-intervention\] of 1519 girls \[baseline\], 91.24%) provided at least 8 hours of data on 3 weekdays and 1 weekend day.

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 Minutes of Moderate to Vigorous Physical Activity (MVPA) Post-intervention, measured over Minutes of moderate to vigorous physical activity per hour at post-intervention (after 17-week intervention).

The primary endpoint is minutes of moderate to vigorous physical activity (MVPA) per hour after the 17-week intervention. It is a continuous measure, summarized as a mean with standard deviation in the posted arm results. The registry describes activity measurement using an ActiGraph GT3X+ accelerometer worn on an elastic belt at the right hip for 7 consecutive days, covering 5 weekdays and 2 weekend days. The collected data were re-integrated into 15-second epochs and processed using established intensity cut-points.

The endpoint is intended to quantify time spent at moderate-to-vigorous activity intensity from the accelerometer record. Standardizing the wear-time windows is relevant because the amount of time observed can affect the opportunity to record activity. The registry states that wear time was standardized to 14 hours per weekday and 10 hours per weekend day, and that an imputation approach used available data in hour blocks across the 7 days. The record says that 1386 girls at post-intervention, out of 1519 girls at baseline, provided at least 8 hours of data on 3 weekdays and 1 weekend day.

As a continuous outcome, the mean gives a group-level average, while the standard deviation describes variation among observations. The endpoint can therefore support a comparison of average measured activity between randomized groups. It does not directly show how many girls reached a particular activity level, and the registry record does not specify a threshold-based primary outcome. The registry’s brief summary expresses its immediate post-intervention hypothesis in minutes per week, whereas the posted primary outcome is labeled in mean minutes per hour; these are the units used in their respective registry descriptions and should not be treated as interchangeable.

7. Results Posted in the Registry

• Physical Activity Intervention: mean 3.27 versus 3.27 on control; difference 0.00 (95% CI -0.16 to 0.16), standardized effect d = 0.00; z-test p = 1.000.

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.

Cardiovascular Fitness (Aerobic Performance)

• 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:
  • Physical Activity Intervention: mean 37.51 versus 37.32 on control; difference 0.19 (95% CI -0.36 to 0.74), standardized effect d = 0.04; z-test p = 0.499.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Physical Activity Intervention and the control. The interval gives the range of effects that remain compatible with the data.

Body Mass Index (BMI) Z-score

• 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 errors.
• Result:
  • Physical Activity Intervention: mean 0.99 versus 1.01 on control; difference -0.02 (95% CI -0.05 to 0.01), standardized effect d = -0.08; z-test p = 0.157.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Physical Activity Intervention and the control. The interval gives the range of effects that remain compatible with the data.

Percent Body Fat

• 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 errors.
• Result:
  • Physical Activity Intervention: mean 30.30 versus 30.67 on control; difference -0.37 (95% CI -0.76 to 0.02), standardized effect d = -0.10; z-test p = 0.062.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Physical Activity Intervention and the control. The interval gives the range of effects that remain compatible with the data.

Across the three posted secondary outcomes, the estimates do not all point in the same direction or have the same degree of uncertainty. Cardiovascular fitness is slightly higher in the intervention arm in the posted means, while the two body-composition measures are lower in the intervention arm. Those directions are descriptive; the secondary outcomes are distinct measures and should not be combined into a single overall effect.

The secondary analyses offer context alongside the primary activity result, but they do not replace it. A result on aerobic performance or body composition does not directly answer the primary question about accelerometer-measured MVPA. Likewise, the primary activity comparison cannot be used to infer the secondary outcomes. Each outcome has its own measurement, summary, and uncertainty.

The registry lists 11 outcome measures and 11 analyses, while three secondary post-intervention outcomes are posted. When multiple outcomes are examined, the chance of at least one apparently noteworthy result can increase if each comparison is considered independently. General statistical note: multiplicity adjustments or a prespecified hierarchy can help control the interpretation of multiple tests, but the registry record does not report whether such procedures were used for these secondary results. Their unadjusted results should therefore be treated as exploratory or contextual unless a fuller analysis plan establishes otherwise.

The registry’s secondary mixed-model results and the simple unadjusted comparisons are also methodologically distinct. The reported model estimates should be interpreted as the sponsor’s analyses, and the unadjusted calculations as direct comparisons from posted summaries. The sample-size illustrations are planning exercises derived from observed effects, not evidence about the reliability of the findings or the trial’s original design. Taken together, the secondary outcomes describe a varied pattern of group estimates, with uncertainty and multiplicity considerations that counsel against a broad conclusion based on any one isolated result.

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

• Cardiovascular Fitness (Aerobic Performance)
• Body Mass Index (BMI) Z-score
• Percent Body Fat

9. Statistical Interpretation

10. Statistical Methods Explained

The registry reports continuous outcomes as means, with standard deviations for the primary outcome and cardiovascular fitness and standard errors for the body-composition outcomes. A mean is an average across the observations included in a group summary. A standard deviation describes variation among observations within a group. A standard error describes the uncertainty of an estimated mean, rather than the spread of individual observations. These are different summaries and should not be interpreted as interchangeable measures of variability.

What does the unadjusted comparison represent?

The unadjusted mean difference compares the posted group means directly, without incorporating model-based adjustment. For the primary outcome, the posted means are 3.27 in each group, so the recalculated unadjusted difference is 0.0. This is a simple comparison of the available summary statistics. It does not reproduce every possible analysis of participant-level data, and it should not be described as the registry’s mixed-model result.

For continuous outcomes, a mean difference is expressed in the outcome’s measurement units. A positive estimate indicates a higher mean in the intervention group than in control under the stated group ordering; a negative estimate indicates a lower mean. The sign alone does not indicate importance, and clinical interpretation depends on the outcome and context. The registry does not report a threshold for a meaningful difference in the primary activity measure.

What does a confidence interval add?

A confidence interval communicates uncertainty around an estimate. In general, a 95% confidence interval is produced by a procedure that, under its assumptions, would cover the target value in 95% of repeated applications. It is not a probability statement that the particular interval contains the true value, nor does it list all effects that are possible. A wider interval reflects less precision than a narrower interval, all else equal. The interval should be read alongside the estimate and the analysis method that produced it.

For a difference between groups, an interval that includes zero is compatible with no difference under that analysis, as well as with values on either side of zero within the interval. This does not prove the groups are identical. Demonstrating equivalence requires a prespecified margin and an analysis designed to assess whether the interval falls within that margin. Such a margin for the primary outcome is not reported in the registry record.

Why are there two kinds of estimates?

The record includes a recalculated unadjusted comparison based on posted group summaries and a sponsor-reported mixed models analysis. A mixed model is a model-based approach that can account for structured or repeated observations when the analysis specifies such features. The exact model specification, covariates, repeated-measure structure, and handling of missing observations are not reported in the registry record. It is therefore not appropriate to infer those details from the method label alone.

The two estimates may differ because they are not necessarily calculated from the same information or under the same assumptions. The unadjusted calculation uses posted group means and sample sizes. The sponsor’s estimate is reported as a parameter estimate from mixed models. The record does not provide enough detail to reconcile the difference or to establish whether the model estimate targets precisely the same quantity as the unadjusted difference. The reported analysis and the transparent summary-statistic calculation should be kept distinct.

Why does dispersion matter?

When observations vary substantially around a group mean, the mean difference is estimated less precisely than it would be with less variation, given the same sample size and other conditions. The standard deviation is relevant to this uncertainty for individual-level measurements. For measures reported with standard errors, the standard error describes the precision of the group mean as posted. The registry’s primary outcome reports a standard deviation of 1.49 in each group.

General statistical note: over-dispersion occurs when variability is greater than a model assumes. If present and not handled appropriately, it can make model-based uncertainty estimates too small. The registry’s primary endpoint is continuous and the record does not report an over-dispersion assessment; this concept should not be taken as a finding about this trial.

How should planning illustrations be read?

The sample-size illustrations associated with secondary outcomes are planning calculations based on the observed effects, not observed power and not the trial’s own design justification. They assume a two-sided significance level of 0.05, equal allocation, and a normal approximation, with the detectable effect set to a stated fraction of the observed effect. The listed results are participants per arm for 80% power and participants per arm for 90% power. Such illustrations depend strongly on the assumed effect: planning for a smaller fraction of an observed effect requires more participants per arm than planning for the full observed effect. These values are not evidence that the completed trial did or did not have adequate power.

General statistical note: power is a design property calculated under specified assumptions about the effect, variability, analysis, and error rate. It is not a measure of the probability that a finding is true. Retrospective power calculated from an observed estimate is not a useful substitute for examining the estimate and its confidence interval. The planning illustrations should therefore be treated as hypothetical design aids, not as a judgment about the trial’s result.

What information is missing for a full audit?

The registry record does not report a detailed statistical analysis plan, a primary-outcome sample-size justification, an interim-analysis plan, an equivalence margin for the primary endpoint, or full model specifications. It also does not give a clear account of the analysis population or all participant-level missing-data decisions. These omissions limit how fully an independent reader can reproduce or audit the reported models. They do not invalidate the posted summaries, but they constrain how much can be inferred beyond the statistics and methods explicitly described.

11. Confidence Intervals and Precision

The primary unadjusted difference in means is 0.0 mean minutes per hour, with a 95% confidence interval from -0.157 to 0.157. The point estimate is the center of the reported interval, while the interval communicates uncertainty around that estimate under the unadjusted calculation. Because the interval includes zero and values on either side, the results do not identify a clear direction of difference for this comparison. The interval is relatively narrow in relation to the displayed primary group means, but it does not establish that the groups are equivalent.

The sponsor-reported mixed-model estimate is -0.08, with a 95% confidence interval from -0.21 to 0.05. This interval also includes zero. It is a model-based result and differs from the unadjusted estimate; its meaning depends on model details that are not reported in the registry record. The two intervals should not be combined or treated as if they were produced by the same method.

Precision is not the same as importance. A narrow interval can still include values that might matter in context, while a point estimate near zero does not by itself establish that meaningful differences have been ruled out. The registry does not report a primary-outcome threshold for an important difference or a prespecified equivalence margin. Consequently, the posted intervals describe statistical uncertainty but do not settle the practical importance of all values they contain.

Confidence intervals also depend on the assumptions and data used to calculate them. The unadjusted interval is based on the posted group summaries; the sponsor interval comes from the reported mixed models analysis. Unreported details about the model and missing observations limit a full assessment of the latter interval’s construction. The confidence intervals are best read as method-specific estimates of precision, rather than as interchangeable summaries.

12. Analyses Reported by the Sponsor

The registry reports a sponsor analysis using mixed models, with a parameter estimate of -0.08, a 95% confidence interval from -0.21 to 0.05, and p = “.207”. The separate recalculated unadjusted comparison from the posted primary arm summaries is 0.0, with a 95% confidence interval from -0.157 to 0.157 and p = 1.0. The two estimates are not identical.

The unadjusted calculation is a direct comparison of the reported arm means and sample sizes. The sponsor result is identified as a mixed-model parameter estimate. The registry record does not provide sufficient model detail to determine precisely why the estimates differ, whether the model includes other measurements or covariates, or whether the model targets exactly the same contrast as the simple comparison. The findings should be described by their respective methods rather than forced into a single number.

Neither the sponsor-reported confidence interval nor the unadjusted confidence interval excludes zero. The sponsor-reported p value is “.207”; the unadjusted p value is 1.0. These results do not constitute a demonstration of equivalence, because the registry record does not report a primary-outcome equivalence margin. The registry’s notation of the analysis type does not provide that missing margin or replace the need to understand the analysis plan.

The record does not flag a discrepancy. The sponsor’s analysis is therefore reported as its own model-based result, while the recalculation is a transparent, simple summary-statistic comparison. The reported mixed-model analysis is the appropriate description of the sponsor’s stated method; the recalculation is useful as a separate check on the posted group summaries.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists 11 outcome measures and 11 analyses, including the primary endpoint and a set of secondary measures. The record does not report a plan for controlling multiplicity, such as a testing hierarchy, an adjustment method, or a rule for deciding which outcomes would support confirmatory conclusions. Without such information, multiple unadjusted tests should not be interpreted as though each were the sole prespecified question. General statistical note: examining many outcomes can increase the chance of at least one result appearing notable by chance, even if each test is considered separately.

The registry record does not report a detailed statistical analysis plan, a primary-outcome sample-size justification, or an interim-analysis plan. It also does not specify the intended analysis population in terms such as intention-to-treat or per-protocol. The posted primary outcome counts are 680 in control and 706 in the intervention arm; the record does not establish how those counts relate to every randomized participant or to the full enrolled population.

Missing outcome data require particular attention because the primary endpoint depended on accelerometer wear. The registry states that 1386 girls at post-intervention, out of 1519 girls at baseline, provided at least 8 hours of data on 3 weekdays and 1 weekend day. It also says an imputation approach used all available data in hour blocks across all 7 days. The precise imputation method, assumptions about missingness, and handling of incomplete records in the mixed-model analysis are not reported in the registry record.

General statistical note: if the likelihood of missing data is related to activity level or differs between groups, estimates can be affected unless the analysis assumptions and methods adequately address that pattern. This is a general concern, not evidence that such bias occurred in this trial. The registry does not provide enough information to assess the plausibility of missing-data assumptions or to compare results under alternative assumptions.

The record also does not report a primary-outcome equivalence margin. Although the mixed-model result is labeled as an equivalence analysis in the registry, the margin needed to interpret equivalence is not stated for the primary endpoint. An interval that includes zero is not by itself evidence of equivalence. These omissions limit confirmatory interpretation, particularly across multiple outcomes and analysis approaches.

14. Safety Results

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

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 777 participants at risk in the control group and zero among 766 participants at risk in the physical activity intervention group. It also reports zero other adverse events among 777 at risk in control and zero among 766 at risk in the intervention group. No serious adverse-event terms are listed. These figures describe the posted adverse-event counts and the group-specific numbers recorded as at risk.

The adverse-event denominators are not the same as the primary outcome group counts. They refer to the safety figures recorded in the registry and should not be substituted for the participants included in the primary outcome summaries. The record does not explain how the at-risk populations were defined, how adverse events were collected, or whether reporting covered all events or only events meeting specified criteria.

Deaths are not reported in the registry record. Zero recorded events in the posted categories should be read as a description of those registry counts, not proof that no event of any kind occurred. The registry figures also do not, by themselves, establish a comparative safety profile or support medical advice. Interpretation depends on event definitions, ascertainment, follow-up, and completeness of reporting, details that are not fully described in the registry record.

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

The primary posted outcome describes average accelerometer-measured MVPA per hour after the 17-week intervention. The two posted arm means are both 3.27. The unadjusted difference is 0.0, and its confidence interval includes values below and above zero. The registry’s mixed-model analysis reports a different estimate, -0.08, with a confidence interval that also includes zero. Those results describe the data and analyses recorded; they do not justify a broad claim about every participant or every form of physical activity.

The result does not show that the groups were exactly identical. A failure to identify a clear difference is not the same as proving no difference. Nor does the interval establish equivalence: a margin defining how much difference would count as practically negligible is not reported for the primary endpoint. The registry also does not state a threshold for a meaningful difference in the activity measure.

The result does not establish why the group summaries took the values they did. Randomization is the stated design feature intended to support comparison between groups, but the registry record does not report all allocation details or the full analysis population. The posted result also depends on accelerometer data processing, wear-time standardization, and imputation, with important details about the missing-data method not reported.

The secondary outcomes address cardiovascular fitness and body composition, not the same activity construct as the primary endpoint. Their estimates provide additional outcome-specific information, but multiple unadjusted secondary comparisons carry more risk of chance findings, and the registry does not report a multiplicity plan. A stated hypothesis about activity at a later follow-up is present in the summary, but later follow-up results are not reported in the registry record.

In practical terms, the registry supports a cautious account of the post-intervention estimates and their uncertainty. It does not support medical advice, a claim of treatment efficacy or lack of efficacy, a definitive equivalence conclusion, or a generalization to populations beyond those described in the trial record. Those judgments would require context and information beyond the posted results.

16. Why This Trial Matters Statistically

This trial illustrates why an outcome’s measurement scale matters. The registry summary states an immediate post-intervention hypothesis in minutes per week, while the posted primary outcome is expressed as mean minutes per hour. The endpoint description and the hypothesis therefore need to be kept distinct; values in one unit should not be casually restated in the other. Clear alignment between the stated question, outcome definition, and reported unit is essential for interpretation.

It also illustrates the difference between a direct calculation from arm summaries and a model-based estimate. The posted arm means produce an unadjusted difference of 0.0, whereas the sponsor-reported mixed-model estimate is -0.08. These are not contradictory versions of the same calculation: they are results reported using different approaches, and the model details needed for a complete reconciliation are absent. Separating methods prevents a simple comparison from being mistaken for the sponsor’s full analysis.

The trial also shows why confidence intervals matter alongside p values. Intervals communicate the range of estimates compatible with an analysis, while a p value alone does not express the size or precision of a difference. In this record, both the recalculated and reported primary intervals include zero, but that does not prove equality or equivalence. Equivalence requires a prespecified margin and an appropriate analysis.

Finally, accelerometer-based endpoints require attention to wear time, processing, and missing observations. The registry describes several of these steps and reports an imputation approach, but not the complete statistical plan. This makes the trial a useful example of how measurement details, model transparency, multiplicity, and missing-data reporting shape the strength of conclusions that can be drawn from a clinical-trial registry.

17. Limitations

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

Several limitations specific to this record constrain interpretation. The primary endpoint is reported in mean minutes per hour, while the registry’s immediate hypothesis is expressed in minutes per week. The registry does not explain the relationship between those descriptions. The posted primary arm means are identical at the displayed precision, while the mixed-model analysis reports a parameter estimate of -0.08; full model specifications are not reported, so the difference between these analyses cannot be fully audited.

The accelerometer outcome depended on usable wear data. Although the registry states the threshold and gives the counts meeting it, it does not describe the complete missing-data assumptions or provide sensitivity analyses. The record says that hour-block imputation was used, but does not give enough information to reproduce the procedure or determine how it affected the estimate. The posted outcome group counts also do not establish the analysis population relative to all enrolled or randomized participants.

The registry does not report a primary-outcome sample-size justification, a detailed statistical analysis plan, an interim-analysis plan, or an equivalence margin for the primary outcome. It also does not report whether multiplicity was controlled across the listed outcomes. These gaps limit assessment of prespecification, the role of secondary tests, and whether the analysis was aligned with the stated hypotheses.

The record’s summary includes a later follow-up hypothesis, but results for that follow-up are not reported in the registry record. Safety reporting lists zero events in the specified categories but does not fully explain ascertainment or definitions, and deaths are not reported. These are limits of the registry information; they should not be interpreted as evidence that unreported procedures or events did or did not occur.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Girls on the Move Intervention to Increase Physical Activity Among Middle School Girls — NCT01503333.
• Registry record last updated 2018-11-21; status last verified 2018-11.
• Results first posted 2018-11-21.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

21. Record Summary

The registry describes a completed, randomized, parallel-group trial of a 17-week school-based physical activity intervention for middle school girls. The intervention combined school nurse counseling, tailored computer feedback, and an after-school physical activity club; controls received usual school offerings and completed data collection.

The primary endpoint was post-intervention mean minutes of MVPA per hour, measured using accelerometers. Both posted group means were 3.27. The recalculated unadjusted difference was 0.0, with a 95% confidence interval from -0.157 to 0.157. The sponsor-reported mixed-model estimate was -0.08, with a 95% confidence interval from -0.21 to 0.05. The registry does not provide enough model detail to fully reconcile the two approaches.

Three secondary post-intervention outcomes and adverse-event counts are also posted. The record does not report a complete statistical analysis plan, a primary-outcome equivalence margin, a multiplicity procedure, or later follow-up results. The estimates are best read with these reporting limits in mind.