Cleaner Air for Better Cardiac Biomarkers: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Cleaner Air for Better Cardiac Biomarkers (NCT05379400), a randomized na study. The primary outcome is Percent Change in Circulating Concentration of Tumor Necrosis Factor Alpha (TNFα).

2. Clinical Question

3. Trial Design

The registry describes a randomized, double-blind, parallel-group, sham-controlled trial of portable air cleaners. Participants were assigned to an active device or a sham device. The active device used a true HEPA filter; the sham used the same model with the filter removed. The registry says the devices had identical appearance and sound, a feature intended to help maintain blinding. Whether blinding was successful is not reported in the registry record.

The study is listed as completed, with a start date of 2022-08-11 and a primary completion date of 2023-08-24. Its phase is listed as NA. The registry records enrollment of 48. The study summary describes a planned cohort of 74 adults with hypertension, and the posted primary outcome results list 22 participants in each group. Reasons for the difference between the described cohort and the recorded enrollment, or for the outcome group counts, are not reported in the registry record.

The intervention period described in the study summary was four weeks, with biomarker measurements before and after use of the air cleaner. The primary outcome timeframe is listed as Baseline, Week 4. The registry also describes home blood-pressure tracking as exploratory. Detailed information on allocation procedures, adherence, and the trial's original sample-size justification is 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.

Before assignment: 48 participants were enrolled, n=2 did not start the study.

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 Percent Change in Circulating Concentration of Tumor Necrosis Factor Alpha (TNFα), measured over Baseline, Week 4.

The primary endpoint is the percent change in circulating tumor necrosis factor alpha (TNFα) from baseline to Week 4. TNFα is an inflammatory biomarker, and the study description identifies it as an established marker selected for an a priori specified analysis. Concentrations were to be measured using the Olink Explore 384 Cardiometabolic panel.

In general, measuring a biomarker can provide information about a biological response to an intervention, but a biomarker result is not itself a measure of a clinical event or a patient's overall health. The endpoint therefore addresses a focused biological question: whether the groups differed in their change in TNFα over the study period. It does not, by itself, establish that any biomarker change led to a change in blood pressure or cardiovascular outcomes.

The posted outcome is summarized as a mean, with standard deviations describing variability within each group. The registry does not report further details about the exact calculation of each participant's percent change or how observations with missing measurements were handled.

7. Results Posted in the Registry

• Portable Air Cleaner (PAC) Group: mean -1.03 versus 2.00 on control; difference -3.03 (95% CI -6.84 to 0.78), standardized effect d = -0.47; z-test p = 0.120.

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.

Change in Indoor Particulate Matter < 2.5 um in Diameter (PM2.5) Concentration

• 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:
  • Portable Air Cleaner (PAC) Group: mean 2.70 versus 11.75 on control; difference -9.05 (95% CI -17.86 to -0.25), standardized effect d = -0.61; z-test p = 0.044.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Portable Air Cleaner (PAC) Group than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

There is one posted secondary outcome: change in indoor PM2.5 concentration. Its estimate points toward a lower change in the active-device group than in the sham group, while the primary TNFα comparison remains uncertain. Taken together, the outcomes present a pattern in which the environmental measure is more clearly separated between groups than the primary biomarker measure.

The two outcomes address different steps in the study's rationale. Indoor particulate matter is an exposure-related measure, while TNFα is a circulating biomarker. A difference in the environmental measure does not demonstrate that exposure change produced a biological response, and the posted results do not establish mediation. The study summary mentions home blood-pressure tracking for exploratory analysis, but blood-pressure results are not reported in the registry record.

The secondary result carries less weight for a confirmatory conclusion because it is an additional outcome and its statistical interpretation is unadjusted. When multiple outcomes are examined, the chance of seeing a seemingly notable result by chance can increase. In general, planned multiplicity control helps address that concern; whether such a plan existed or was applied is not reported in the registry record. The secondary finding can be treated as informative for follow-up research, but it should not be treated as conclusive confirmation of the primary biomarker question.

The registry lists two outcome measures overall and posts one secondary outcome result. No secondary analyses are listed. The record does not report whether the secondary outcome was designated exploratory in the formal analysis plan, how missing measurements were handled, or whether any additional outcomes were collected but not posted. Those gaps limit conclusions about the full set of analyses considered.

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

• Change in Indoor Particulate Matter < 2.5 um in Diameter (PM2.5) Concentration

9. Statistical Interpretation

10. Statistical Methods Explained

The registry posts group means, standard deviations, and a mean difference for the primary continuous outcome. These summaries describe average change and the variability of change within each randomized group. They do not show the distribution of individual participant results or explain how the analysis handled missing measurements.

What does the mean difference represent?

The reported mean difference is the active-group mean minus the sham-group mean. Its value of -3.03 indicates that the observed average percent change was lower in the active group by that amount on the endpoint's scale. The sign depends on the order of subtraction; it does not by itself establish whether the difference is clinically important or reliable.

In general, a mean difference is easiest to interpret when the outcome scale has a clear substantive meaning. Here the scale is percent change in circulating TNFα, but the registry does not report a threshold for a clinically meaningful difference. A difference in biomarker change should not be translated into a change in cardiovascular risk or another clinical outcome without supporting evidence.

How should standard deviations be read?

The standard deviation describes how spread out participant-level values are around a group mean. The posted standard deviations are 8.35 in the active group and 3.69 in the sham group. They indicate that participant changes varied, particularly in the active group. A standard deviation is not a confidence interval for the group mean, and it does not show the full shape of the data. Participant-level observations and distribution plots are not reported in the registry record.

In general, greater variability makes a group mean difference less precisely estimated when sample sizes are otherwise similar. It can also make results sensitive to unusual observations. The registry does not report diagnostics, transformations, or methods for handling outliers, so their role in this result cannot be assessed.

What does the confidence interval add?

A confidence interval describes uncertainty around an estimated group difference under the statistical method used. The primary interval runs from -6.845 to a value above zero. Because it crosses zero, the results are compatible with no average difference under the interval's assumptions, while also allowing a range of nonzero differences. The interval does not assign a probability to the true effect being inside its limits.

In general, a confidence interval is affected by the estimated variability, sample size, and analysis method. It gives more information about the range of values compatible with the data than a p-value alone. The registry does not report enough detail to verify the interval's calculation independently, and the posted result should be read as the registry's reported analysis.

What does the p-value say?

The primary unadjusted p-value is 0.12. In general, a p-value describes how incompatible the observed data are with a specified null model, assuming that model and the test's assumptions hold. It is not the probability that the null hypothesis is true, nor does it measure the size or importance of an effect. The registry does not report a multiplicity-adjusted p-value or a full statistical analysis plan.

The comparison is unadjusted in the posted result. The record does not state whether the primary analysis included covariates, how repeated measurements were modeled, or how participants were handled if one of their measurements was missing. It would be inappropriate to infer a paired test, a particular regression model, or a specific analysis population from the outcome summary alone.

Why can unequal variability matter?

In general, over-dispersion means that observed variability is greater than a model expects. It matters when a model's standard errors rely on a variance assumption that does not fit the data, because uncertainty can then be understated or overstated. The primary endpoint is continuous, and the registry reports different group standard deviations, but it does not report a formal over-dispersion assessment or a model that would allow that issue to be evaluated. A difference between group standard deviations alone does not establish over-dispersion.

The reported comparison also includes a standardized effect estimate, but the registry does not explain its role in the primary analysis. Standardized effects express a difference relative to a variability scale and can help compare outcomes measured in different units. They are not a substitute for the original-unit estimate or its confidence interval, and their interpretation depends on how the standardizing variability was defined.

All statistical explanations above are general. The registry record does not report the full protocol or statistical analysis plan, so details beyond the posted summaries remain unknown.

11. Confidence Intervals and Precision

The primary estimate is a mean difference of -3.03, but its confidence interval extends from -6.845 to a value above zero. That interval crosses the no-difference value. It therefore does not isolate a single direction of effect: the data are compatible with a negative difference, with little or no difference, and with a positive difference within the interval's limits.

The interval is important because the point estimate alone can look more definite than the evidence warrants. The estimate identifies the center of the reported range, while the interval communicates uncertainty about the group difference. Its upper end is close to zero on the endpoint scale, but the registry's posted interval does not exclude zero. The unadjusted p-value of 0.12 is consistent with that uncertainty.

Precision is limited by the modest number of participants listed for each group and the variability in participant changes, especially in the active group. This does not show that the observed estimate is wrong; it means the registry results leave a broad range of possible group differences. A larger study could produce a narrower interval if its design, variability, and analysis supported that result, but the outcome cannot be predicted from these data.

In general, a confidence interval is not a guarantee that the true effect lies within its bounds, and it is not a range of likely individual responses. It characterizes uncertainty in an estimated group-level quantity under the analysis assumptions. The registry does not report a prespecified clinically meaningful threshold for TNFα change, so statistical precision cannot be translated into a judgment about clinical importance.

12. 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 illustrations based on the observed effect, not the trial's original design justification. That original design justification and its assumptions are not reported in the registry record.

The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The calculation considers detecting the stated fraction of the observed effect. For the primary TNFα outcome, planning for 100% of the observed effect gives 72 participants per arm for 80% power and 96 participants per arm for 90% power. Planning for 75% of the observed effect gives 127 participants per arm for 80% power and 170 participants per arm for 90% power. Planning for 50% of the observed effect gives 285 participants per arm for 80% power and 382 participants per arm for 90% power.

The secondary PM2.5 illustration gives 43 and 57 participants per arm for 80% and 90% power, respectively, when planning for 100% of its observed effect. For 75% of that effect, the corresponding figures are 76 and 102 participants per arm; for 50%, they are 171 and 228 participants per arm.

In plain terms, 80% power describes a design's probability of detecting an effect of the specified size under the planning assumptions if that effect is real; 90% power uses a higher detection probability. Smaller effects require more participants to reach either target because they are harder to distinguish from random variation. These numbers are planning illustrations, not observed power for the completed trial.

13. 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 = -3.030; pooled standard deviation σ = 6.455.
2. n = 2 × (1.96 + 0.842)² × 6.455² / -3.030² = 71.2, rounded up to 72 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 <- -3.03                      # observed difference in means
sigma <- 6.45517621757919                     # 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(5379400)
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 = 72 (80% power), 96 (90% power)
 75% of observed effect: n per arm = 127 (80% power), 170 (90% power)
 50% of observed effect: n per arm = 285 (80% power), 382 (90% power)
power.t.test: 72.2 
Simulated power at n = 72 per arm: 0.796 (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.

14. Multiplicity, Missing Data and Analysis Populations

The registry identifies an a priori specified primary TNFα analysis and lists an additional secondary outcome for indoor PM2.5. It does not report a multiplicity strategy, such as a procedure for controlling the chance of false-positive findings across outcomes or analyses. The secondary result is therefore best interpreted cautiously alongside the primary result rather than as a separate confirmatory conclusion.

In general, multiplicity becomes relevant when researchers test several outcomes, time points, subgroups, or alternative models. If many tests are considered, some may appear notable by chance even when no corresponding effect exists. A prespecified hierarchy or adjustment can clarify how evidence across tests should be interpreted. Whether the investigators specified such an approach is not reported in the registry record.

Missing-data methods are also not reported in the registry record. The primary outcome results list 22 participants in each group, while the overall enrollment is recorded as 48. The record does not explain the relationship between these counts, whether all randomized participants had both measurements, or whether any participants were excluded from the endpoint analysis. It does not state whether missing observations were imputed, analyzed using a model, or handled by another method.

The registry does not identify an intention-to-treat, per-protocol, or other analysis population for the posted primary results. It also does not report the number of participants who completed the intervention, adherence to device use, or reasons for any missing outcome data. These details matter because exclusions or missing measurements can affect how representative an analysis is of the randomized comparison.

Without the analysis plan and missing-data details, readers cannot determine whether every planned outcome and analysis is represented in the registry results or whether the posted estimates follow the intended primary analysis. This is a limit on interpretation, not evidence that the analysis was mishandled.

15. 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 posted safety counts list zero serious adverse events, zero other adverse events, and zero deaths in each group. Each group lists 22 participants at risk for serious events and 22 at risk for other adverse events.

These figures describe the events recorded in the registry for the groups and observation period. A count of zero means no events of those types were posted; it does not prove that a device has no risks or that an event could not occur in a larger or longer study. The registry does not provide further detail about event definitions, monitoring procedures, or the completeness of safety reporting. The figures should therefore be read as a record of reported events in this trial, not as a general safety assessment.

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

The primary result estimates a lower mean percent change in TNFα in the active air-cleaner group than in the sham group. However, the confidence interval crosses zero and the unadjusted p-value is 0.12. The posted data therefore leave uncertainty about the size and direction of the group difference. They do not establish a statistically clear difference, and they do not establish equivalence or absence of an effect.

The secondary indoor PM2.5 result points toward a group difference in the expected direction for an air-filtration intervention. That environmental finding does not prove that a change in indoor particulate matter caused a change in TNFα. The registry does not post blood-pressure results, and the primary biomarker outcome is not a clinical cardiovascular endpoint. Accordingly, these results do not establish an effect on blood pressure, cardiovascular disease, or other clinical outcomes.

In general, a randomized comparison can help estimate the effect of assignment to an intervention when its conduct and analysis support that interpretation. The registry identifies random allocation and a sham comparator, but it does not report detailed allocation procedures, adherence, analysis populations, or missing-data handling. Those unreported details constrain how fully the result can be evaluated.

The study summary describes a pilot project and a four-week period of home air-cleaner use. A short-term biomarker result should not be extrapolated to longer-term exposure, sustained biomarker changes, or health outcomes. The observed estimates can help frame questions for further research, but they do not settle the biological or clinical questions on their own.

17. Why This Trial Matters Statistically

This trial is statistically informative because it places an environmental exposure measure and an inflammatory biomarker in the same randomized comparison. That structure makes it possible to examine whether assignment to active filtration is associated with changes in both a measure of indoor particulate matter and a biological marker. The posted results also illustrate that evidence can differ across related endpoints: the primary biomarker estimate is uncertain even though the secondary environmental estimate is more distinct.

The trial also highlights why effect estimates should be read with their uncertainty. A point estimate gives a concise summary, but the primary confidence interval crosses zero and the group measurements show variability. Reporting the interval alongside the estimate makes clear that the result does not pinpoint a precise average difference.

As a pilot, the study can contribute information for planning later research, including the observed scale and variability of outcomes. The planning illustrations based on these observed effects show how the required participants per arm change when the target effect is smaller or the desired power is higher. These illustrations are distinct from the trial's original design rationale, which is not reported in the registry record.

Finally, the record underscores the value of complete reporting. Clear information on analysis populations, missing data, multiplicity, and adherence would help readers evaluate how robustly the posted outcomes represent the randomized comparison. Such details are particularly useful when interpreting a small study with multiple biological and environmental questions.

18. 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 posted primary outcome groups each include 22 participants, and the primary confidence interval spans values on both sides of zero. That combination limits precision and leaves uncertainty about the size and direction of the average difference. The registry records enrollment of 48, while the study summary describes a planned cohort of 74 adults; reasons for this difference are not reported in the registry record.

The intervention period was four weeks, so the results address a short-term measurement window. The primary endpoint is a biomarker rather than a clinical outcome. Although the registry summary mentions home blood-pressure tracking as exploratory, results for that measure are not reported in the registry record. No conclusions about blood pressure or cardiovascular events can be drawn from the posted biomarker result.

The record does not report adherence to device use, the number of participants completing the intervention, missing-data methods, or the analysis population. It also does not report detailed procedures for allocation, the success of blinding, or the original sample-size rationale. These omissions prevent a full assessment of potential bias and of how the posted outcome group counts relate to all enrolled participants.

The secondary outcome is an additional analysis, and the registry does not report a multiplicity strategy. Its result should not be interpreted as independent confirmation of the primary finding. The registry also does not provide participant-level data or further distributional details, limiting assessment of unusual observations and the shape of outcome variability.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Cleaner Air for Better Cardiac Biomarkers — NCT05379400.
• Registry record last updated 2024-12-06; status last verified 2024-10.
• Results first posted 2024-12-06.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

22. Record Summary

NCT05379400 is a completed, randomized, double-blind, parallel, sham-controlled trial sponsored by NYU Langone Health. It compared a portable air cleaner with a sham device in a study described as involving adults with hypertension and four weeks of home use.

The primary outcome was percent change in circulating TNFα from baseline to Week 4. The posted mean difference favored a lower change in the active group, but its confidence interval crossed zero and the unadjusted p-value was 0.12. One secondary outcome, indoor PM2.5 concentration, was also posted. The registry does not report the full statistical plan, missing-data methods, or the reasons for the difference between the described planned cohort and recorded enrollment.