The Impact of Real-time Feedback on Physical Activity Patterns in Flemish Employees: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for The Impact of Real-time Feedback on Physical Activity Patterns in Flemish Employees (NCT01432327), a randomized na study. The primary outcome is Physical Activity Level.

2. Clinical Question

The study description also frames the broader question as whether electronic self-monitoring can increase daily physical activity, how those values compare with the 10.000 step program, and whether real-time feedback differs when accompanied by personal coaching. The posted primary-outcome results address group comparisons in Physical Activity Level. They do not report a separate statistical comparison of the 10.000 step program as a target or threshold.

3. Trial Design

The registry describes a randomized, parallel-group intervention study with no masking. It lists the study as completed, with enrollment of 227, and identifies KU Leuven as the sponsor. The registry lists the start as 2010-08 and primary completion as 2015-02. The study phase is recorded as NA.

The four named groups differed in the feedback participants received. The Control Group received no feedback during the intervention period; participants wore the SenseWear Armband for 4 weeks, with results discussed after the intervention period. The Step Group received daily step-count feedback by pedometer. The Display Group received real-time feedback on energy expenditure, minutes of physical activity, and step count through the SenseWear Display. The Coaching Group received real-time feedback on energy expenditure, step count, and activity minutes, along with weekly meetings with a Personal Coach to discuss progress.

The primary outcome was Physical Activity Level, measured at the one-year timeframe and expressed in metabolic equivalents. Its description defines PAL as total energy expenditure divided by basal metabolic rate. The record describes intention-to-treat analysis, including participants who did not complete the study. Details of the randomization procedure, allocation concealment, intervention duration for each feedback group, and the registry's original design assumptions are not reported in the registry record.

The registry lists four secondary outcomes: Daily Energy Expenditure in Physical Activity, Percent of Participants Losing Fat Percentage, Stages of Motivational Readiness for Physical Activity, and Step Count. Although those outcomes are listed, their analyzed results are not reported in the registry record. The record lists 5 outcome measures and 0 analyses.

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: Students and seniors were excluded from the study. Participants with an average Physical Activity Level above 1.71 times their resting metabolic rate were excluded from the trial

Baseline

Posttest 3 Months

Reasons for not completing this period:

Posttest 6 Months

Reasons for not completing this period:

Posttest 1 Year

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intention to treat analysis was used and analyses data from all participants, including those who did not complete the study

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 Physical Activity Level, measured over One year.

The primary endpoint is Physical Activity Level (PAL), reported as a mean in metabolic equivalents. The registry defines PAL as the 24-hour energy requirement, measured as total energy expenditure, divided by basal metabolic rate. Expressing energy expenditure relative to basal metabolic rate accounts for differences in body size and composition, according to the outcome description.

The record describes a desirable PAL as including regular physical activity at work or during spare time, at an intensity and duration associated with reducing risk of overweight and chronic diseases associated with obesity. It identifies PAL values of 1.75 and higher as corresponding to that description. This is contextual information in the registry, not a reported trial finding or a separate outcome threshold analysis.

The endpoint is a continuous measure, so the registry reports group means and standard deviations. The primary outcome timeframe is one year. The posted arm summaries are 1.45 for the Control Group, 1.4 for the Step Group, 1.46 for the Display Group, and 1.46 for the Coaching Group. Their standard deviations are 0.15, 0.16, 0.17, and 0.18, respectively. These summaries describe the recorded outcome; on their own, they do not establish that feedback caused a change.

7. Results Posted in the Registry

• Step Group: mean 1.40 versus 1.45 on control; difference -0.05 (95% CI -0.11 to 0.01), standardized effect d = -0.32; z-test p = 0.089.
• Display Group: mean 1.46 versus 1.45 on control; difference 0.01 (95% CI -0.05 to 0.07), standardized effect d = 0.06; z-test p = 0.741.
• Coaching Group: mean 1.46 versus 1.45 on control; difference 0.01 (95% CI -0.05 to 0.07), standardized effect d = 0.06; z-test p = 0.750.

8. Statistical Interpretation

9. Statistical Methods Explained

What does a mean difference represent?

The primary outcome is continuous, and its arm summaries are reported as means with standard deviations. A mean is the arithmetic average of the measured values in a group. A standard deviation describes how spread out individual measurements are around that average; it is not the uncertainty in the mean itself. In this study, the reported mean differences compare the average PAL in each feedback group with the average in the Control Group. A negative estimate indicates a lower group mean than control, while a positive estimate indicates a higher group mean.

The reported estimates are -0.05 for Step versus Control, 0.01 for Display versus Control, and 0.01 for Coaching versus Control. These estimates are unadjusted comparisons. They describe differences between group averages and do not, by themselves, account for any other factors or establish a cause-and-effect explanation.

What does a confidence interval convey?

General: A confidence interval expresses uncertainty around an estimate. A narrower interval indicates greater precision, while a wider interval leaves more uncertainty about the size and direction of the underlying difference. When an interval includes zero for a mean difference, the data are compatible with no difference as well as with some positive or negative differences within the interval.

All three intervals in the registry include zero. The exact confidence level and the statistical procedure used to form these intervals are not reported in the registry record. Their interpretation should therefore stay with the information the record gives: each estimated difference is uncertain, and none of the intervals excludes no mean difference.

How should the p-values be understood?

General: A p-value describes how unusual data at least as incompatible with a specified null model would be, if that model were true and the assumptions of the analysis held. It is not the probability that the null hypothesis is true, nor does it measure the size or practical importance of an effect. A p-value also does not establish that a result is reproducible.

The registry reports unadjusted p-values of 0.089 for Step, 0.741 for Display, and 0.75 for Coaching, each compared with Control. The inferential method behind these values is not reported in the registry record. The values should be read alongside the mean differences and confidence intervals, rather than as standalone measures of evidence or clinical importance.

Why can variability and analysis choices matter?

General: The standard deviation describes variability among observed measurements. More variability can make group means harder to distinguish precisely, particularly when the number of observations is limited. A standard deviation is not the same as a standard error or a confidence interval. The registry gives group standard deviations and comparison intervals, but does not explain the precise procedure used to calculate those intervals.

The record states that intention-to-treat analysis was used and that participants who did not complete the study were included. That describes the intended analysis population, but it does not specify how missing one-year measurements were handled. Options such as imputation, modeling, or analysis of observed values cannot be inferred from the record. The missing-data method, the observed endpoint counts by arm, and any sensitivity analyses are not reported in the registry record.

General: In a multi-arm study, comparing several groups against one control creates multiple statistical comparisons. If each is assessed separately, the chance of at least one apparently notable result across the set can differ from the chance attached to an individual comparison. Methods that account for multiplicity can be used when appropriate, but the registry record does not report a multiplicity adjustment, a testing hierarchy, or whether one was planned. The three p-values should therefore be understood as unadjusted comparisons.

What cannot be inferred about the design?

The registry identifies the study as randomized, parallel, and unmasked. It does not report how the random sequence was generated, whether allocation was concealed, or how adherence was assessed. The original sample-size justification and statistical analysis plan are not reported in the registry record. The analysis method used by the sponsor, including any covariate adjustment or model specification, is also not described. As a result, the comparison statistics can be explained as posted, but the record does not support reconstructing all of the choices behind them.

10. Confidence Intervals and Precision

Precision concerns how tightly the data locate the estimated group difference. The Step Group estimate was -0.05, with a confidence interval that extends from below zero to above zero. The Display Group estimate was 0.01, and its interval also crosses zero. The Coaching Group estimate was 0.01, with an interval that crosses zero as well. In each comparison, the interval therefore includes no mean difference and also permits differences in either direction.

The unadjusted p-values are 0.089 for Step, 0.741 for Display, and 0.75 for Coaching. These values are consistent with the fact that none of the intervals excludes zero, but they do not show how likely any particular effect is or how important a difference might be. The registry record does not report the confidence level, so it should not be inferred from convention.

The standard deviations are 0.15 in Control, 0.16 in Step, 0.17 in Display, and 0.18 in Coaching. They describe the spread of PAL measurements within each group, rather than precision of the between-group estimates. The interval widths reflect uncertainty in the comparisons, but details of the interval calculation are not reported in the registry record.

These intervals are compatible with a range of effects, including no difference. The posted estimates are near zero for Display and Coaching, while Step has a negative estimate; uncertainty remains for all comparisons. The registry does not identify a minimum clinically important difference, so the intervals cannot be classified against such a threshold using the record alone.

11. Planning a Similar Trial: Sample Size

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

The planning figures are participants per arm, not total events. They are calculated as illustrations from the observed effect, not as the trial's original design calculation. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; the effect to detect is the stated fraction of the observed effect. The trial's own sample-size assumptions are not reported in the registry record.

For each comparison, the planning rows use 100%, 75%, and 50% of the observed effect. The columns for 80% and 90% power indicate the number of participants per arm that the planning calculation estimates would be needed to detect the selected effect with those target power levels under the listed assumptions. Power is a design property for a specified effect and analysis setup; it is not the probability that the observed result is true.

Smaller target effects require more participants per arm to detect, all else equal. This pattern is visible when the planning effect is reduced from the full observed effect to a smaller fraction. It is particularly important to read the numbers in light of the observed differences: estimates close to zero can produce very large planning requirements when the calculation is based on detecting that small observed effect. Such calculations are sensitive to the observed estimate and should not be treated as a recommendation for a future study without a separate design rationale.

These figures are planning illustrations, not observed power. They do not show the probability that this completed study detected its true effect, and they do not replace the registry's unreported original sample-size justification. No time-to-event endpoint is being used for these calculations.

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 = -0.050; pooled standard deviation σ = 0.155.
2. n = 2 × (1.96 + 0.842)² × 0.155² / -0.050² = 151.1, rounded up to 152 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 <- -0.05                      # observed difference in means
sigma <- 0.155126465129217                     # 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(1432327)
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 = 152 (80% power), 203 (90% power)
 75% of observed effect: n per arm = 269 (80% power), 360 (90% power)
 50% of observed effect: n per arm = 605 (80% power), 810 (90% power)
power.t.test: 152.1 
Simulated power at n = 152 per arm: 0.803 (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 primary outcome description says that an intention-to-treat analysis was used, including participants who did not complete the study. This indicates the stated analysis population, but the record does not describe how missing one-year PAL measurements were handled. It does not report whether missing values were imputed, modeled, or excluded from a particular calculation, nor does it give the number of observed primary-outcome measurements by arm. The details of missingness and any sensitivity analyses are not reported in the registry record.

The registry lists four secondary outcomes: Daily Energy Expenditure in Physical Activity, Percent of Participants Losing Fat Percentage, Stages of Motivational Readiness for Physical Activity, and Step Count. The record lists 0 analyses, and results for those secondary outcomes are not reported in the registry record. The number of outcome measures listed should not be taken to mean that results for each measure are available.

The posted primary comparisons assess Step, Display, and Coaching against Control, with unadjusted p-values. The registry record does not report whether the comparisons were selected in advance as a family, whether one comparison was designated as primary among the group contrasts, or whether any adjustment for multiple testing was planned or used. No corrected p-values or multiplicity procedure are reported. Consequently, the p-values should be interpreted as unadjusted results, and no family-wise claim can be inferred from them.

Although the study is identified as randomized, the randomization procedure, allocation concealment, and statistical model are not reported in the registry record. It also does not report the sponsor's full analysis plan or whether the posted group summaries reflect any adjustments. Those omissions limit assessment of the analysis choices, but they do not alter the basic descriptive reading of the posted means, standard deviations, estimates, intervals, and unadjusted p-values.

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 adverse-event counts separately from the primary activity outcome. For serious adverse events, the Control and Step groups each have 0 participants listed as at risk; the Display Group has 3 and the Coaching Group has 5 listed as at risk. The number affected by serious adverse events is recorded as 0 in all four groups. The serious-event risk counts should be read as recorded; zero listed as at risk is not the same as a broad assurance that no serious events occurred.

For other adverse events, the registry lists 56 participants at risk in Control and 57 in each feedback group. The number affected is recorded as 0 for each group. Death counts are null for all groups, which means deaths are not reported in the registry record; a null entry should not be interpreted as a reported zero.

These registry entries describe what was recorded in the adverse-event fields. The record does not provide event terms, collection procedures, follow-up completeness, or further safety analysis. They should not be read as a comprehensive evaluation of safety or used to infer more than the counts and reporting status shown.

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

The posted results describe one-year Physical Activity Level in a randomized comparison of no feedback with step-count feedback, real-time activity feedback, and real-time feedback accompanied by personal coaching. The arm means were similar in magnitude: 1.45 for Control, 1.4 for Step, and 1.46 for both Display and Coaching. The estimated differences from Control were -0.05, 0.01, and 0.01, respectively. Each confidence interval includes zero.

These estimates do not establish that feedback caused an increase or decrease in activity. They are group summaries and unadjusted comparisons, with uncertainty around each estimate. The p-values do not measure clinical importance, and the registry does not report a threshold for a meaningful difference in PAL. The descriptive reference to PAL values of 1.75 and higher is part of the outcome definition; the record does not report a separate test of whether participants reached that level.

The result also does not establish equivalence among the groups. A conclusion of equivalence would require an appropriate design and a prespecified margin, neither of which is reported in the registry record. Similarly, the close mean values do not prove that the feedback approaches have identical effects. The intervals allow for a range of possible differences, and the record does not report the confidence level or complete analysis details.

The study's broader aim mentions electronic self-monitoring, comparison with a 10.000 step program, and real-time feedback with or without coaching. The posted primary outcome summaries provide group comparisons for PAL, but the record does not report the secondary outcome analyses that might address related activity measures. The interpretation should therefore remain focused on the primary endpoint and the reported comparisons, rather than extending to unreported outcomes or longer-term effects.

16. Why This Trial Matters Statistically

This study is statistically informative because it compares different feedback formats while using a continuous measure of activity as its primary endpoint. The parallel-group design makes the group means and contrasts straightforward to describe: each feedback group is compared with a no-feedback control, and uncertainty is visible through the reported intervals. The results also illustrate why an estimate should be considered alongside its precision rather than judged only by its direction.

The comparisons are close to zero for Display and Coaching, while Step has a negative estimate. All three confidence intervals include zero, and the unadjusted p-values range from 0.089 to 0.75. These features make it important to distinguish a small observed difference from evidence that the true difference is exactly zero. The record supports neither a claim of clear separation nor a claim of equivalence.

The study also highlights the statistical cost of targeting small effects in planning. The planning illustrations based on the observed differences call for more participants per arm when the target effect is reduced. Those calculations are not the study's own design justification, and their dependence on observed estimates is a reason to treat them as illustrations rather than definitive future-study requirements.

Finally, the registry record makes transparent several limits on interpretation: the secondary outcomes are listed but their analyses are not reported; the handling of missing values and the multiple-comparison strategy are not reported; and the analysis details beyond intention-to-treat are absent. These are important distinctions for researchers reading registry results, because clear descriptive estimates can coexist with incomplete reporting of the methods needed to fully evaluate their uncertainty.

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 registry record identifies a randomized study but gives limited detail about its implementation. The randomization sequence, allocation concealment, recruitment setting, adherence, and deviations from assigned interventions are not reported in the registry record. Masking is listed as none, so expectations or awareness of group assignment could potentially affect behavior or measurement; the record does not describe any steps taken to address this.

The intervention descriptions are not equally detailed. The control description specifies wearing the SenseWear Armband for 4 weeks, but the duration of the feedback interventions is not reported in the registry record. The primary outcome is set at one year, and the record does not explain the timing of outcome measurement relative to the feedback period or what activity monitoring was used to derive the one-year PAL values.

The record does not report the analysis procedure, confidence level, missing-data handling, outcome counts by arm, or a multiplicity strategy. It also lists secondary outcomes without reporting their analyses. The original sample-size justification and any prespecified clinically important difference are not reported. These omissions constrain evaluation of the precision, robustness, and practical meaning of the estimates.

The registry enrollment is 227, while arm-level outcome summaries report their own sample-size fields. The record does not explain the relationship between enrollment, assignment, follow-up, and the values contributing to each summary. The result should therefore be read from the reported endpoint summaries and not as a complete account of participant flow. Applicability beyond the employee population named in the study title is also not established by this record.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: The Impact of Real-time Feedback on Physical Activity Patterns in Flemish Employees — NCT01432327.
• Registry record last updated 2015-03-13; status last verified 2015-02.
• Results first posted 2012-04-18.
• 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 record describes a completed, randomized, parallel-group study with no masking, sponsored by KU Leuven. It lists enrollment of 227 and a one-year primary outcome of Physical Activity Level, expressed as a metabolic-equivalent value. Participants were assigned to Control, Step, Display, or Coaching groups, with the latter groups receiving different forms of activity feedback.

The reported primary means were 1.45 for Control, 1.4 for Step, and 1.46 for Display and Coaching. The unadjusted mean differences from Control were -0.05, 0.01, and 0.01, with confidence intervals that include zero. The registry states that intention-to-treat analysis included participants who did not complete the study, but does not report missing-data handling or detailed analysis methods.

Four secondary outcomes are listed, but their analyses are not reported in the registry record. The original sample-size justification and the confidence level are also not reported.