Implementation and Evaluation of a Before School Physical Activity Program in Revere, MA: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Implementation and Evaluation of a Before School Physical Activity Program in Revere, MA (NCT03805295), a randomized na study. The primary outcome is Change in Body Mass Index (kg/m^2).

2. Clinical Question

The primary question is about a difference in group mean changes, not simply whether either group's average changed from baseline. The posted means are 0.14 for Spring 2018 Participation and 0.01 for Fall 2018 Participation. The reported mean difference, calculated as Fall relative to Spring, is -0.13. The comparison also includes a confidence interval and an unadjusted p-value. These summarize uncertainty around the posted group comparison, but the record does not establish that the result reflects a causal effect of program participation. The registry describes the study as evaluating implementation and outcomes, and records randomized allocation, no masking, and a sequential model.

3. Trial Design

The registry record describes a completed randomized study evaluating implementation of BOKS in three schools serving K-8 students in Revere, Massachusetts. The sponsor is Massachusetts General Hospital. The registry lists enrollment of 188, classifies the allocation as randomized, masking as none, and the model as sequential. The study is listed as not having an applicable phase.

The record describes two participation periods. Students in Spring 2018 Participation were to take part in BOKS in Winter-Spring 2018 and serve as the control group in Fall 2018. Students in Fall 2018 Participation were to take part in BOKS in Fall 2018. Each arm description gives a target of 40 students per school, up to 120 total. These are descriptions of the planned participation groups; they should not be confused with the outcome-specific counts in the posted results.

The primary outcome is change in body mass index, measured at baseline and follow-up for each session over 12 weeks. The registry also names change in executive function and change in PedsQL quality of life as secondary outcomes. It lists three outcome measures and two posted secondary outcomes. The detailed assessment procedures for the secondary measures are not reported in the registry record.

The record gives the study start as 2018-02-16, primary completion as 2018-12-30, and status as completed. It describes the project as a cohort study while also recording randomized allocation. The registry does not report further details on sequence generation, concealment of assignment, how the sequential model was implemented, whether participants contributed in more than one period, or how within-student measurements were handled. Those design details should not be inferred from the labels alone.

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

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 Body Mass Index (kg/m^2), measured over Baseline, Follow-up for each session (12 weeks).

The primary endpoint is change in body mass index, expressed in kg/m^2. BMI is calculated from height and weight; the registry states that the reported change is calculated from those measurements. The stated timeframe is baseline and follow-up for each session, with follow-up at 12 weeks. The reported outcome type is a continuous measure summarized by its mean and standard deviation.

A change score represents the difference between a participant's measurement at follow-up and the baseline measurement, although the precise calculation convention, including the direction of subtraction, is not reported in the registry record. The group means therefore summarize the posted changes; they are not the groups' final BMI levels. The comparison of interest is the difference between those average changes.

Using a continuous endpoint retains the size and direction of measurements rather than reducing them to categories. General statistical context: a mean change can describe the average shift in a group, while its variability indicates how much individual changes differ around that average. A group mean alone does not show the distribution of individual responses. The record does not report the clinical interpretation assigned to a particular BMI change, nor does it provide a threshold for a meaningful change. The endpoint can describe the posted anthropometric outcome, but it does not by itself summarize the program's other stated aims concerning executive function, quality of life, physical activity, or fitness.

7. Results Posted in the Registry

• Fall 2018 Participation: mean 0.01 versus 0.14 on control; difference -0.13 (95% CI -0.41 to 0.15), standardized effect d = -0.20; z-test p = 0.370.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous: BMI change in kg/m^2. The registry reports a mean and standard deviation for each participation group, followed by a mean difference, confidence interval, and unadjusted p-value. These summaries answer related but distinct questions. Group means describe the average observed change; the mean difference describes how far apart the two averages are; the confidence interval describes uncertainty around that estimated difference; and the p-value summarizes compatibility with a null comparison under a statistical model.

What does the mean difference represent?

The comparison is Fall 2018 Participation relative to Spring 2018 Participation. Its estimate is -0.13. The negative sign indicates that the posted mean change in the Fall group was lower than the posted mean change in the Spring group. This is a comparison of average changes, not a comparison of final BMI values. The group means are 0.01 and 0.14, respectively. The registry does not report the exact convention used to calculate each participant's change score, so interpretation should remain tied to the stated outcome rather than assume a particular subtraction order beyond the direction used for the reported group comparison.

How should the confidence interval be understood?

General statistical explanation: a confidence interval gives a range of values that are compatible with the observed estimate and the assumptions of the interval procedure. Its width reflects precision: a wider interval represents less precise estimation than a narrower one. Here, the reported interval runs from a negative value to a positive value and therefore includes 0. That makes the estimate uncertain about the direction of the group difference. The exact confidence level is not reported in the registry record, so it is not appropriate to label this as a 95% interval.

What does the p-value say?

The unadjusted p-value is 0.37. General statistical explanation: a p-value measures how unusual the observed data, or data more extreme, would be under a specified null model. It is not the chance that the null model is true, and it does not measure the probability that the observed difference is due to chance. The registry does not report the trial's original analysis plan, including its prespecified hypothesis test or decision threshold. The p-value should consequently be reported as an unadjusted result rather than treated as a complete description of the analysis.

Why does variability matter?

The standard deviations are 0.63 for Spring 2018 Participation and 0.7 for Fall 2018 Participation. General statistical explanation: standard deviation describes dispersion of individual observations around a group mean. When individual outcomes vary substantially, the mean may be estimated less precisely unless there are enough observations to offset that variability. A mean difference should therefore be read alongside both group sizes and dispersion, not in isolation. The comparison reports outcome counts of 76 and 31, and a pooled standard deviation of 0.651. The registry does not explain why these counts differ from one another or from total enrollment.

What assumptions affect the comparison?

General statistical explanation: a simple two-group comparison relies on assumptions about the observations and the way variability is represented. Independence is particularly relevant when data may include repeated measurements from the same students or students grouped within schools. The registry records a sequential model and identifies Spring participants as a control group in Fall, but it does not report how repeated observations or school-level clustering were treated in the comparison. It also does not report covariate adjustment, an analysis population definition, or a missing-data method. The posted estimate and unadjusted p-value should not be mistaken for evidence that these issues were addressed in a particular way.

For the planning illustration, the stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect set at the listed fraction of the observed effect. These assumptions belong to the planning illustration, not necessarily to the original trial design. The original design's statistical assumptions and sample-size justification are not reported in the registry record.

10. Confidence Intervals and Precision

The mean difference in BMI change is -0.13. The reported confidence interval spans negative and positive values, including 0. This means the data do not pin down the direction of the difference precisely: the interval is compatible with a lower average change in Fall relative to Spring and also with a higher average change. The confidence level is not reported in the registry record, so the interval should not be assigned a confidence percentage that is absent from the record.

General statistical explanation: precision concerns the range of effect sizes that remain reasonably compatible with the data under the interval method. It is distinct from the point estimate. An estimate can be close to 0 but still have meaningful uncertainty, and a confidence interval that includes 0 does not show that the groups are equivalent. Establishing equivalence would require an appropriate prespecified margin and analysis, which are not reported in the registry record.

The outcome-specific counts are 76 for Spring 2018 Participation and 31 for Fall 2018 Participation. The reported standard deviations are 0.63 and 0.7. The smaller Fall outcome count and the observed variability are relevant context for the uncertainty in the comparison, but the registry does not report enough detail to identify the contribution of each factor or to reconstruct the interval method. The unadjusted p-value of 0.37 offers a separate summary and does not replace the interval's information about possible effect sizes.

Precision is also limited by gaps in the reported analysis details. The record does not state how missing measurements were handled, whether the comparison accounted for school grouping or repeated participation, or which population was included in the analysis. These omissions do not change the arithmetic of the posted estimate, but they limit the ability to assess how robustly the interval represents uncertainty under the study's design.

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. The illustration assumes a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. It uses the observed effect as its reference and considers planning for 100%, 75%, and 50% of that observed effect. These assumptions describe the planning illustration; the original trial's sample-size justification and design assumptions are not reported in the registry record.

For 100% of the observed effect, the illustration gives 394 participants per arm for 80% power and 527 participants per arm for 90% power. For 75% of the observed effect, it gives 700 participants per arm for 80% power and 937 participants per arm for 90% power. For 50% of the observed effect, it gives 1,574 participants per arm for 80% power and 2,107 participants per arm for 90% power.

In plain language, the 80% power column describes a planning target intended to give a specified chance of detecting the assumed effect under the listed statistical assumptions if that effect is present. The 90% power column uses a higher target chance of detection and therefore calls for more participants under the same planning framework. Smaller effects are harder to distinguish from random variation, so the illustration requires more participants per arm as the assumed effect becomes smaller. These figures are planning illustrations, not observed power.

The figures do not describe how many students were enrolled or analyzed in this study. Registry enrollment is 188, while the posted primary outcome groups contain 76 and 31 participants. The planning numbers should not be interpreted as a report that these larger numbers took part in the trial, nor as evidence by themselves about the study's success or failure.

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 = -0.130; pooled standard deviation σ = 0.651.
2. n = 2 × (1.96 + 0.842)² × 0.651² / -0.130² = 393.4, rounded up to 394 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 <- -0.13                      # observed difference in means
sigma <- 0.650768776140958                     # 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(3805295)
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 = 394 (80% power), 527 (90% power)
 75% of observed effect: n per arm = 700 (80% power), 937 (90% power)
 50% of observed effect: n per arm = 1,574 (80% power), 2,107 (90% power)
power.t.test: 394.3 
Simulated power at n = 394 per arm: 0.804 (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. Multiplicity, Missing Data and Analysis Populations

The registry names BMI change as the primary outcome and lists change in executive function and change in PedsQL quality of life as secondary outcomes. It reports three outcome measures and two posted secondary outcomes. The posted primary outcome comparison has an unadjusted p-value of 0.37. The registry record does not report a multiplicity adjustment, a hierarchy of hypotheses, or a plan for interpreting results across the primary and secondary outcomes. It is therefore not possible to determine whether or how the study controlled the increased chance of false-positive findings that can arise when several outcomes or comparisons are examined.

The record does not report results for the two named secondary outcomes in the available outcome information. Their absence from the posted results should not be read as evidence that they were not measured, nor should any unreported result be inferred. The registry lists no analyses in its analysis listing, while the outcome information includes a group comparison for BMI change. The analysis methods, including any adjusted models or additional planned analyses, are not reported in the registry record.

Missing-data handling is not reported. The record does not state how many participants had missing baseline or follow-up measurements, whether incomplete cases were excluded, or whether any method was used to account for missing values. The analysis population is also not described: there is no reported definition of intention-to-treat, per-protocol, or another population. The outcome counts of 76 and 31 therefore cannot by themselves establish which participants were eligible for the analysis or why those counts differ from total enrollment of 188.

The registry also does not report whether the statistical analysis accounted for the three-school setting, the sequential model, or potential repeated observations among students participating across periods. These are design and analysis considerations, not proof that the posted results are invalid. Without the analysis plan and fuller reporting, however, the available result cannot establish how these features were handled or how they might affect uncertainty.

14. Safety Results

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

Deaths

Other (non-serious) adverse events

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

The serious adverse event figures list 86 participants at risk in Spring 2018 Participation and 42 at risk in Fall 2018 Participation. The recorded serious adverse event count is 0 in each group, and deaths are also recorded as 0 in each group. The other adverse event figures list the same at-risk counts, with 0 recorded in each group.

These figures mean that the registry records no serious adverse events, no other adverse events, and no deaths in the listed at-risk groups. They do not establish that no event of any kind occurred outside the period or definitions used for registry reporting. The record does not provide event definitions, monitoring procedures, or additional detail about how events were collected. It also does not explain why the safety at-risk counts differ from the primary outcome counts of 76 and 31.

Adverse event counts are descriptive safety information. They are not a comparison of risk-adjusted rates, and no statistical safety comparison is reported. The available counts should be interpreted within the reporting scope of the registry rather than as a complete account of every possible event.

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

The posted primary result describes a mean BMI change of 0.14 for Spring 2018 Participation and 0.01 for Fall 2018 Participation. The mean difference, Fall relative to Spring, is -0.13. The reported confidence interval includes 0, and the unadjusted p-value is 0.37. Taken together, these summaries show a modest numerical difference in the posted group means with substantial uncertainty about the direction and size of the underlying difference.

This result does not establish that the groups had identical outcomes. An interval that includes 0 is not a demonstration of equivalence, and the registry does not report an equivalence margin or a design intended to establish equivalence. Nor does the result establish that BOKS caused a change in BMI. The interpretation is constrained by the study's seasonal participation groups, the stated control role for the Spring group in Fall, and missing details about repeated participation, school-level clustering, and the analysis population.

The record describes a broader evaluation that includes implementation, executive function, and quality of life, in addition to BMI. The primary BMI comparison cannot answer whether implementation was successful or whether other outcomes changed. Although the registry names executive function and PedsQL quality of life as secondary outcomes, their results are not reported in the available outcome information. No conclusion about them follows from the BMI figures.

General statistical context: a trial result is one estimate accompanied by uncertainty, not a universal prediction for every student or school. The posted means summarize groups, and they do not show the range of individual changes. The record also does not report a threshold for interpreting the size of BMI change. Accordingly, the numerical estimate should not be translated into a claim about clinical importance or individual benefit.

The appropriate reading is narrow: for the participants included in the posted primary outcome comparison, the recorded mean changes were 0.14 and 0.01, with an estimated group difference of -0.13 and uncertainty spanning values on both sides of 0. The registry does not provide enough methodological detail to extend that finding to broader causal or outcome claims.

16. Why This Trial Matters Statistically

This trial illustrates why a reported mean difference should not be read alone. The primary outcome is continuous, and the registry provides both group means and a comparison estimate. The estimate of -0.13 conveys the direction and scale of the group contrast, while the confidence interval communicates that the estimate is uncertain and includes 0. The unadjusted p-value of 0.37 is another summary, but it cannot replace the interval or the design details needed to interpret the result.

The study also makes clear that trial labels and analysis details are not interchangeable. The registry records randomized allocation and a sequential model, describes Spring participants as a control group in Fall, and calls the project a cohort study. These descriptions do not, by themselves, reveal whether participants were measured repeatedly, how school-level grouping was handled, or how allocation translated into the analysis. Those details matter when assessing the independence assumptions and precision of a comparison.

Sample-size illustrations are also easy to misread. They are based on a specified effect and assumptions, and the resulting participants-per-arm figures change with the effect size being planned for. They are not observed trial results and do not reveal the original sample-size justification. Distinguishing planning quantities from realized enrollment helps prevent a calculated target from being mistaken for evidence about the trial's actual power.

Finally, outcome reporting affects what can be concluded. BMI change is the posted primary comparison, while executive function and PedsQL quality of life are named secondary outcomes without results in the available outcome information. Missing-data rules, analysis populations, and multiplicity handling are not reported. These gaps are statistically important because they shape how broadly and confidently a reader can interpret a result, even when the numerical estimate itself is clear.

17. Limitations

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

The primary outcome comparison is based on 76 participants in Spring 2018 Participation and 31 in Fall 2018 Participation, whereas registry enrollment is 188. Safety reporting lists 86 and 42 participants at risk. The registry does not explain these different counts, so participant flow, outcome availability, and the relationship between the analysis and enrolled population remain unclear.

The groups are tied to different participation periods, and the record says that Spring participants would serve as the control group in Fall. It does not explain whether the same students contributed in more than one period or how any repeated observations were treated. With three schools involved, school-level grouping is another potential design feature, but the registry does not report whether the comparison accounted for it. These unknowns limit assessment of the assumptions behind the reported comparison.

The record does not report the exact method used to construct the confidence interval, its confidence level, the original analysis plan, the sample-size justification, or the original analysis threshold. It also does not state how missing values were handled, define the analysis population, or describe adjustment for multiple outcomes. The posted p-value is explicitly unadjusted. These omissions make it difficult to judge the completeness and prespecification of the analysis.

The available outcome information does not report results for the named secondary outcomes, and the registry's analysis listing contains no analyses. The absence of those details limits conclusions about executive function, quality of life, and the study's broader implementation aims. The primary result also has no reported threshold for a clinically meaningful BMI change. It should therefore be treated as a registry-reported group comparison with uncertainty, not as a complete account of all trial outcomes or a definitive causal evaluation.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Implementation and Evaluation of a Before School Physical Activity Program in Revere, MA — NCT03805295.
• Registry record last updated 2025-12-23; status last verified 2025-12.
• Results first posted 2025-12-23.
• 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

NCT03805295 is a completed randomized evaluation of the BOKS before-school physical activity program in three K-8 schools in Revere, Massachusetts. The sponsor is Massachusetts General Hospital. The registry reports enrollment of 188, no masking, a sequential model, and study dates of 2018-02-16 for start and 2018-12-30 for primary completion.

The primary outcome is change in BMI in kg/m^2, measured at baseline and follow-up for each 12-week session. The posted means are 0.14 for Spring 2018 Participation and 0.01 for Fall 2018 Participation, with a mean difference of -0.13, a confidence interval spanning negative and positive values, and an unadjusted p-value of 0.37. Outcome counts are 76 and 31.

Executive function and PedsQL quality of life are named secondary outcomes, but their results are not reported in the available outcome information. Serious adverse events, other adverse events, and deaths are each recorded as 0 in both listed safety groups. Analysis details, missing-data handling, and the relationship among enrollment, outcome, and safety counts are not reported in the registry record.