Assessment of an School-based Intervention in Eating Habits and Physical Activity in Schoolchildren: the AVall Study: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Assessment of an School-based Intervention in Eating Habits and Physical Activity in Schoolchildren: the AVall Study (NCT01156805), a randomized na study. The primary outcome is Anthropometric Variables.

2. Clinical Question

In statistical terms, the trial asks whether the pattern of BMI over the study period differs between children assigned to the school-based educational intervention and those assigned to no intervention. That framing makes the comparison about change over time, not simply a single follow-up measurement. The registry identifies the analysis as a superiority analysis using mixed models and reports p = 0.05. It does not provide a numerical estimate of the difference in BMI progression, a confidence interval, or enough detail to establish exactly how the outcome was represented in the model. The clinical question is therefore identifiable, but the magnitude and precision of the estimated contrast cannot be assessed from the reported result alone.

3. Trial Design

The AVall Study is listed as a randomized, parallel-group trial with no masking. Its status is completed, and the registry records enrollment of 509 participants. The intervention was an educational program focused on food habits and physical activity among schoolchildren. The comparator is described as receiving no intervention. The study is sponsored by Hospital de Granollers.

The record identifies the design as cluster-randomized. In a cluster-randomized trial, groups such as schools, rather than individual participants, are assigned to study conditions. Participants within a school may share experiences and environments, so their outcomes may be more alike than outcomes from participants assigned independently. This dependence affects how uncertainty should be estimated. The identity of the randomization unit, the number of clusters, and the allocation procedure are not reported in the registry record.

The record lists the study start as 2006-09 and primary completion as 2008-09. The primary outcome is described as BMI progression in the intervention and control groups, with a timeframe of every two years. The intervention description says training was received during the study period, but the duration and intensity of the educational activities are not reported.

There is some ambiguity in the arm descriptions. The registry describes one group as receiving educational training and another as receiving no intervention, but arm names do not consistently distinguish these conditions. The intended comparison is discernible from the descriptions, although the labels should be read cautiously. Participant flow, numbers analyzed for the primary outcome, exclusions after randomization, and the handling of departures from assigned groups are not reported in the registry record. These omissions limit assessment of implementation and analysis populations, even though the broad randomized, parallel, cluster-level structure is stated.

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: Qualitative variables are described in absolute and relative frequencies (percentages). Continuous variables are described as mean and SD.

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 Anthropometric Variables, measured over every two years.

The primary outcome is titled “Anthropometric Variables,” and its description specifies BMI progression in the intervention and control groups. BMI is expressed in kg/m^2 and is treated as a continuous measure. The registry says continuous variables are described using means and standard deviations, and gives the measurement timeframe as every two years. It does not state the exact measurement visits or give the observed mean BMI values.

BMI is used as an anthropometric indicator that can be measured repeatedly and compared across groups. A progression outcome can capture whether the measure changes differently over time, rather than relying only on a value at one follow-up. That general rationale does not establish whether BMI is an adequate measure for every individual child or whether any observed change has a particular clinical meaning. The registry does not report the model’s exact outcome specification, such as whether the analysis tested a group-by-time pattern or another contrast.

The primary-outcome description also says that changes in eating habits and physical activity were secondary outcomes. Elsewhere, the listed secondary outcomes are height, father’s height, and mother’s height. The record therefore gives differing descriptions of secondary outcomes. The primary outcome remains identifiable as BMI progression, but the exact set and definition of secondary endpoints are not fully clear from the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The statistical method named for the reported comparison is mixed models analysis. The registry identifies the endpoint as a continuous measure and says that continuous variables are described as mean and standard deviation. It does not report the fixed effects, random effects, covariance structure, estimation procedure, or exact hypothesis tested. The explanation below describes the general statistical ideas, not unreported details of the sponsor’s implementation.

Why use a mixed model?

General explanation: mixed models can represent measurements that are related because they come from the same participant or because participants belong to the same group. They combine effects describing average patterns across the study with terms that account for variation among groups or individuals. In a study with repeated BMI measurements, observations from one child may be related over time. In a cluster-randomized trial, children attending the same school may also have related outcomes. A mixed model can account for such dependencies when its structure matches the design and data.

The registry calls the trial cluster-randomized, but does not identify the cluster unit or describe how clustering was represented in the model. It also does not say whether repeated measurements were modeled directly. The method name alone is not enough to determine whether the analysis fully reflected both the school-level assignment and the repeated outcome measurements.

Why does clustering matter?

General explanation: when observations within a cluster resemble one another, treating every participant as statistically independent can make the data appear to contain more independent information than they do. That can lead to uncertainty estimates that are too small and tests that are too favorable. Appropriate cluster-aware methods adjust the analysis to account for within-cluster similarity. The relevant correction depends on the cluster structure and the analysis model. The number of clusters and other details needed to assess that structure are not reported in the registry record.

The record explicitly states that an individual-level recalculation would be misleading for this cluster-randomized design. For that reason, the sponsor’s own analysis is the relevant analysis reported for interpretation. No unadjusted re-analysis should be treated as an alternative estimate of the trial result.

What does a p-value convey?

General explanation: a p-value measures how unusual data at least as incompatible with a null hypothesis as the observed data would be, assuming that null hypothesis and the analysis assumptions. It does not state the probability that the null hypothesis is true. It also does not show the magnitude of a difference, the probability that an intervention is beneficial, or whether an effect matters in practice.

The registry reports p = 0.05. It gives no confidence interval or numerical effect estimate alongside that value. A confidence interval would describe a range of effect estimates compatible with the data and statistical model at a stated confidence level. Without such an interval, readers cannot assess the precision of the estimated contrast from the registry result. The exact test definition and decision threshold are not reported.

Why report means and standard deviations?

General explanation: a mean describes the average value in a group, while a standard deviation describes how spread out individual values are around that average. These summaries can help characterize a continuous measure such as BMI. They do not by themselves describe the uncertainty in a between-group contrast, and they do not account for clustering or repeated measurements. The registry states that means and standard deviations were used to describe continuous variables, but does not report the relevant group summaries for the primary outcome.

What is not available about the model?

The registry does not report the model formula, the time points included, the contrast used to test superiority, or the treatment of baseline measurements. It also does not report how missing observations were handled, whether assumptions were checked, or whether the analysis used an intention-to-treat population. These details are important for assessing model fit and the relationship between the analysis and the randomized design.

The sample-size assumptions described for planning illustrations are not the trial’s own design justification. Such illustrations are planning tools based on an effect assumption and specified testing conditions; they are not observed power and do not replace an effect estimate with its uncertainty. The trial’s own sample-size justification is not reported in the registry record. No planning value should be interpreted as a result observed among participants.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Educational Intervention, Control", "Non Intervention")
n   <- c(272, 237)
m   <- c(16.4, 17)
sp  <- c(2.8, 2.4)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Educational Intervention, Control  272  16.400  2.800  16.066  16.734
CARM  Non Intervention  237  17.000  2.400  16.693  17.307

11. Analyses Reported by the Sponsor

The registry reports the sponsor’s analysis as a mixed models analysis, with p = 0.05 for a superiority comparison. It does not report a numerical effect estimate, confidence interval, or enough model detail to reconstruct that analysis. As a result, the sponsor’s reported test is the only reported inferential result for the primary comparison.

The record identifies the trial as cluster-randomized and cautions that a simple individual-level recalculation would be misleading. That warning matters because outcomes within a cluster may be correlated, and ignoring that dependence can distort standard errors and hypothesis tests. Accordingly, the sponsor’s own analysis takes precedence; an unadjusted recalculation should not be used to contradict or replace it.

This does not mean that every modeling choice can be independently verified. The registry does not describe the model specification, cluster unit, or analysis population, so the assumptions behind the reported test cannot be fully assessed. Nor does the reported p-value provide the missing effect size or its precision. The sponsor’s result can be reported as a test with p = 0.05, while the magnitude and direction of the BMI contrast remain unreported in the registry record.

12. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary outcome concerning BMI progression and also refers to secondary outcomes. In the primary-outcome description, changes in eating habits and physical activity are called secondary outcomes. A separate list names height, father’s height, and mother’s height. The record therefore does not give one fully consistent account of the secondary endpoints.

Multiplicity arises when several outcomes, time points, subgroups, or statistical tests are examined. General explanation: as the number of tests grows, the chance of obtaining at least one apparently unusual result under a collection of null hypotheses can increase. Methods such as prespecifying a hierarchy or adjusting significance thresholds can address some multiplicity concerns. The registry does not report whether the trial used a multiplicity strategy, how secondary outcomes were prioritized, or whether the primary comparison was protected from other analyses. Interim analyses and any associated adjustment are also not reported in the registry record.

Missing-data handling is not reported. The record does not say how many participants contributed primary-outcome measurements, whether observations were missing at particular visits, or what assumptions and methods were used to address missing outcomes. Those details matter in repeated-measurement analyses because the analysis may rely on assumptions about why values are unavailable. The named mixed models method does not, by itself, establish how missing data were handled.

The analysis population is also not specified. The registry does not state whether the primary analysis included all randomized participants according to assigned group, only participants with outcome data, or another defined population. It does not report exclusions after randomization, protocol deviations, or how participants were treated if they did not receive the assigned condition. These omissions prevent assessment of whether the reported comparison preserves the randomized assignment in the analysis.

These are limits of the available record, not evidence that multiplicity, missingness, or analysis-population issues were necessarily mishandled. The relevant methods and participant-flow details are simply not reported in the registry record.

13. 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 events for 237 participants at risk in the educational-intervention group and 272 participants at risk in the non-intervention group. It records 0 serious adverse events in each group. The corresponding figures for other adverse events are 0 in each group, with 237 and 272 participants at risk, respectively. Deaths are recorded as 0 in both groups.

These counts describe what is recorded in the registry; they do not establish that no adverse experiences occurred outside the categories or reporting process represented in the record. The registry does not provide serious adverse-event terms, and the list of leading serious-event terms is empty. It does not describe how events were collected, the observation period for safety, or whether reporting thresholds differed by event type.

The at-risk counts are not necessarily equivalent to the number of participants contributing to the primary BMI analysis. The record does not give safety follow-up details or explain whether all enrolled participants were followed for the same duration. Therefore, the figures should be read as the registry’s reported counts, not as a complete characterization of every possible safety outcome or of the trial’s overall safety monitoring.

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

The registry describes a randomized, parallel-group, cluster-randomized study of a school-based educational program intended to improve food habits and physical activity. The primary question concerns BMI progression in schoolchildren assigned to the educational intervention compared with a group receiving no intervention. The reported statistical result is a mixed models superiority analysis with p = 0.05.

That result does not show the size or direction of the estimated BMI contrast. The registry gives no group means for the primary outcome, no numerical estimate of the difference in progression, and no confidence interval. A p-value alone cannot establish whether a difference was large or small, or how precisely it was estimated. It also cannot establish that the intervention caused a particular change, especially when the analysis details and implementation information are not available.

General explanation: statistical evidence and practical importance are distinct. A test result concerns compatibility with a statistical model and null hypothesis; practical interpretation depends on the magnitude and uncertainty of the estimated effect, the outcome’s meaning, and the study context. Those elements cannot be fully evaluated from the reported primary result because its effect estimate and interval are not reported.

The cluster-randomized design further means that the analysis must account for the way participants were grouped. The registry names mixed models, but does not provide enough information to assess the model specification. A simple individual-level recalculation would be misleading, so the sponsor’s own analysis takes precedence. The absence of a recalculation is not evidence for or against the reported result; it reflects the need to respect the trial’s cluster structure.

The record also leaves questions about the secondary outcomes, missing data, analysis population, and multiplicity unresolved. These omissions do not demonstrate that the study was conducted improperly. They limit what readers can verify and what conclusions can be drawn from the registry result alone. The appropriate reading is narrow: the registry reports a test of BMI progression with p = 0.05, but does not provide the effect size, direction, or precision needed for a fuller assessment.

15. Why This Trial Matters Statistically

The AVall Study illustrates why the design of a trial must shape its statistical analysis. In a cluster-randomized study, the assignment occurs at the group level, and participants from the same cluster may share influences that make their outcomes correlated. A method that treats all participants as independent can misrepresent the uncertainty. The registry’s explicit warning against a simple individual-level recalculation is therefore an important part of interpreting its reported result.

The study also makes a useful distinction between identifying a statistical test and understanding its result. The record names mixed models analysis and reports p = 0.05, but does not provide an effect estimate or confidence interval. Without these, readers cannot judge the magnitude, direction, or precision of the BMI comparison. A p-value is not a substitute for reporting the estimated contrast and its uncertainty.

Repeated measurement is relevant as well. The primary outcome is described as BMI progression, with a measurement timeframe of every two years. A longitudinal question concerns how values change across time, so clarity about measurement occasions, model structure, and missing observations is essential. The registry does not provide those details. This shows how a concise result can identify the broad analytic approach while leaving important questions about implementation unanswered.

Finally, the record presents differing descriptions of secondary outcomes and ambiguous arm labels. Such inconsistencies do not determine the validity of the primary analysis, but they make it harder to understand the complete outcome and intervention framework. Clear reporting of cluster structure, endpoint definitions, model specification, effect estimates, and participant flow would make the statistical evidence easier to evaluate and reproduce.

16. 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 contains limited information for evaluating the primary analysis. It reports the method as mixed models analysis and gives p = 0.05, but does not provide an effect estimate, confidence interval, group-level BMI summaries, or the exact contrast tested. As a result, the direction, magnitude, and precision of the primary comparison cannot be assessed from the reported result.

The cluster-randomized design is stated, but the record does not identify the randomization unit, number of clusters, or allocation procedure. Although mixed models are named, the model specification is not reported. The available information is therefore insufficient to determine how clustering and repeated BMI measurements were represented in the analysis. A simple individual-level recalculation is not appropriate for this design, so the sponsor’s analysis takes precedence.

The descriptions of trial arms are not fully consistent. The intervention is described as educational training intended to improve food habits and physical activity, while another arm description says no intervention was made; the arm titles themselves are not consistently informative. The exact intervention content, delivery, and adherence are not reported. These gaps make it difficult to assess how the planned contrast was implemented in practice.

There are also differences in the descriptions of secondary outcomes. The primary-outcome text refers to eating habits and physical activity, while the listed secondary outcomes are height, father’s height, and mother’s height. The record does not report a multiplicity plan. Missing-data methods, participant flow, analysis-population definitions, and exclusions after randomization are not reported in the registry record.

These limitations concern what can be learned from the registry entry. They should not be treated as proof that the trial lacked additional analyses or procedures; those details are simply not reported in the registry record. The available information supports a cautious description of the design and reported test, rather than a detailed assessment of effect size or analytic robustness.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Assessment of an School-based Intervention in Eating Habits and Physical Activity in Schoolchildren: the AVall Study — NCT01156805.
• Registry record last updated 2020-08-17; status last verified 2017-08.
• Results first posted 2019-12-13.
• 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.

20. Record Summary

The AVall Study is a completed, randomized, parallel-group, cluster-randomized trial sponsored by Hospital de Granollers. The registry records enrollment of 509 participants and describes a school-based educational intervention focused on food habits and physical activity, compared with no intervention. The primary outcome is BMI progression, expressed in kg/m^2, with a measurement timeframe of every two years.

The reported primary analysis is a mixed models superiority analysis with p = 0.05. The registry does not report an effect estimate or confidence interval, and the model details are not available. Because a simple individual-level recalculation would be misleading for the cluster-randomized design, the sponsor’s own analysis takes precedence. The registry records 0 serious adverse events, 0 other adverse events, and 0 deaths in each group.

Important details—including the cluster structure, analysis population, missing-data methods, multiplicity approach, and numerical BMI results—are not reported in the registry record. The result can therefore be described as a reported test, but its estimated magnitude and precision cannot be determined from the record.