Effectiveness of Intensive Lipid Modification Medication in Preventing the Progression of Peripheral Arterial Disease (The ELIMIT Study): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effectiveness of Intensive Lipid Modification Medication in Preventing the Progression of Peripheral Arterial Disease (The ELIMIT Study) (NCT00687076), a randomized phase 4 study. The primary outcome is Effect of Intensive Lipid Modification Medication Therapy on Progression of Atherosclerosis and Restenosis of Femoral Arteries Measured Using High Resolution Magnetic Resonance Imaging (MRI) to Examine the Femoral Artery for Progression of Atherosclerosis.

2. Clinical Question

The question concerns an imaging measure of arterial structure after an endovascular procedure. The registry’s primary outcome description identifies change in superficial femoral artery (SFA) wall volume as the primary endpoint. It also says that changes in SFA lumen volume and total vessel volume were analyzed. These measures focus on changes in the treated arterial region rather than symptoms or events. A secondary outcome listed in the record is change in total cholesterol from baseline to Month 12.

The comparison is between two medication strategies given with standard medical care. It is not a comparison of medication with no treatment. The record describes a randomized, triple-masked, parallel-group study and lists peripheral arterial disease as its condition. It does not report a separate clinical question about symptom monitoring as a measured endpoint; the primary question recorded for the trial is about MRI-measured arterial change.

3. Trial Design

The ELIMIT trial was a randomized, triple-masked, parallel-group Phase 4 study sponsored by Baylor College of Medicine. The registry lists 102 participants as enrolled and the study status as completed. The recorded start is 2004-04, and primary completion is 2010-12. The record does not describe which study roles were masked, so the exact meaning of triple masking for participants, clinicians, and assessors is not reported in the registry record.

Participants were assigned to one of two medication strategies. The intensive strategy combined a statin with ezetimibe and Niaspan, alongside standard medical care. The comparison strategy used standard lipid-modifying medications with standard medical care and placebo ezetimibe and placebo Niaspan. Thus, the randomized contrast is between an intensive combination and a standard medication approach, not between active treatment and no care.

The primary outcome was change in SFA wall volume measured by MRI over 24 months. The description says that baseline and 24-month measurements were planned and that changes in SFA lumen and total vessel volumes were also analyzed. The posted arm summaries list 45 participants for Triple Therapy and 46 for Mono Therapy. The primary outcome description separately reports 87 participants completed baseline MRI and that multilevel models used all available imaging data from 91 participants. These counts describe different aspects of enrollment and imaging; the registry does not provide a complete participant-flow account that reconciles them.

The record also lists change in total cholesterol from baseline to Month 12 as a secondary outcome. It does not report the original design assumptions, the planned recruitment target rationale, or the detailed allocation procedure beyond stating that allocation was randomized.

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: Multilevel models were used to describe changes over time in the MRI outcome variables and to compare the drug therapy groups. The advantage of multilevel models is the capability to use data with missing or irregularly timed observations, due to death or loss to follow-up, on the outcome variable.

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.

Population: Baseline characteristics were compared using analysis of variance, chi-square tests. Multilevel statistical models were used to analyze the primary outcome. Complete baseline characteristics were availabe for 95 patients. Between randomization and baseline visit, 1 patient withdrew from study, and 6 patients declined blood collection.

6. Primary Endpoint

The registry lists the primary outcome as Effect of Intensive Lipid Modification Medication Therapy on Progression of Atherosclerosis and Restenosis of Femoral Arteries Measured Using High Resolution Magnetic Resonance Imaging (MRI) to Examine the Femoral Artery for Progression of Atherosclerosis, measured over Measured at baseline and 24 Months.

The primary endpoint was change in superficial femoral artery (SFA) wall volume over 24 months, determined by MRI. The registry gives the unit as mm^3 and describes measurements at baseline and 24 months. It also states that the 24-month changes in SFA lumen volume and total vessel volume were analyzed, but identifies SFA wall volume change as the primary outcome.

The SFA is an artery in the leg, making its imaging measurements relevant to the trial’s focus on peripheral arterial disease after endovascular intervention. In general, MRI can provide repeated images of vessel structure, allowing researchers to quantify changes in an anatomical measure rather than relying only on symptoms. That is the general rationale for using an imaging endpoint; the registry does not report validation details or a threshold for a clinically important change.

The posted results summarize the primary continuous measure with arm means and standard errors. The registry’s arm values are 58.1 for Triple Therapy and 60.5 for Mono Therapy, and the reported mean difference is 2.4. The displayed summary does not specify in its arm labels whether these values are adjusted or give a fuller description of how the change score was encoded. The outcome description does identify the modeled outcome as change in SFA wall volume, so interpretation should stay tied to that stated measure rather than assuming a direction of benefit or a clinical consequence.

7. Results Posted in the Registry

• Mono Therapy: mean 60.50 versus 58.10 on control; difference 2.40 (95% CI -12.30 to 17.10), standardized effect d = 0.07; z-test p = 0.749.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous: it represents a measured quantity of SFA wall volume rather than a category such as event or no event. The registry reports means and standard errors for the two groups, along with a mean difference and confidence interval. The general purpose of these summaries is to express the estimated group-level contrast and the uncertainty around that estimate.

What does the mean difference describe?

The mean difference compares the posted arm summaries. The registry reports 58.1 for Triple Therapy and 60.5 for Mono Therapy, and expresses the contrast as Mono Therapy relative to Triple Therapy: 2.4. A positive value under that contrast convention means the Mono Therapy summary is numerically higher. This is a description of the posted measure, not a judgment about clinical benefit. The registry’s primary outcome description frames the endpoint as change in SFA wall volume, but does not provide a separate interpretation threshold for the magnitude of change.

Why use a multilevel model?

The registry says multilevel models were used to describe changes over time in MRI outcomes and compare the medication groups. In general, multilevel models can represent repeated observations from the same participant and account for the fact that measurements within a participant are related. The registry specifically notes that the approach can use outcome data with missing or irregularly timed observations, including observations affected by death or loss to follow-up. This capability does not mean that missing data have no consequences: the validity of an analysis still depends on assumptions about the observation process and model. The registry does not report the detailed model specification or the assumptions used to address missingness.

The primary analysis used all available imaging data from 91 participants. The description says this included 20 participants who completed baseline imaging only and 4 who completed at least 2 imaging visits other than baseline. It also reports that 87 participants completed baseline MRI. The record does not give the number of observations at each time point or a complete account of all missing scans, so the pattern of available data cannot be reconstructed from the record summary alone.

What is the role of the standard error?

A standard error describes the estimated uncertainty in a group mean, under the statistical model used. It is not the same as the spread of individual participant measurements. The posted arm summaries give standard errors of 5.5 and 5.1. The confidence interval for the mean difference is the more direct summary for judging uncertainty in the between-group contrast, because it incorporates uncertainty in that contrast rather than presenting the arm means separately.

What does the p-value tell the reader?

The reported unadjusted p-value is 0.749. In general, a p-value is calculated under a specified null hypothesis and model; it measures how compatible the observed data are with that null model in the particular test. It is not a measure of the probability that one medication strategy is effective, nor does it show that the groups are identical. The registry labels the result unadjusted. It does not report whether the primary analysis was adjusted for other variables or how any multiple-outcome testing was handled.

Why can over-dispersion matter?

General statistical note: over-dispersion arises when observed variability is greater than a model assumes. If a model understates variability, standard errors can be too small and confidence intervals too narrow. This concern is especially relevant to some count-data models; the primary endpoint is continuous and the registry describes multilevel models. The record does not report an over-dispersion assessment, so no specific claim about it can be made for this analysis.

The registry does not report the detailed model formula, covariance structure, covariates, estimation method, handling of values outside expected ranges, or sensitivity analyses. Nor does it report the original statistical analysis plan. These omissions limit independent assessment of how modeling choices affected the estimate. The methods that can be described are therefore the stated use of multilevel models with available MRI data and the posted arm summaries and contrast.

10. Confidence Intervals and Precision

The estimated mean difference is 2.4, with a 95% confidence interval from -12.3 to 17.1. A confidence interval gives a range of estimates compatible with the observed data under the analysis model and its assumptions. It is not a range containing 95% of individual participant values, and it does not assign a 95% probability to this particular interval after it has been calculated.

The interval includes zero, which corresponds to no difference in the direction defined for the contrast. Its endpoints also extend to values below and above zero. Thus, the posted estimate does not pin down whether the group means differ in either direction, nor does it tightly determine the size of a possible difference. The point estimate of 2.4 should be read alongside that uncertainty rather than as a definitive measure of the underlying contrast.

The arm means are reported with standard errors of 5.5 for Triple Therapy and 5.1 for Mono Therapy. Standard errors describe uncertainty in estimated means, not the variation among individual participants. The confidence interval for the difference is more directly useful for assessing the uncertainty in the comparison itself.

The unadjusted p-value is 0.749. It is consistent with the confidence interval’s indication that the data do not provide a precise estimate of a nonzero group difference under the stated analysis. Neither the p-value nor the interval establishes equivalence. The registry does not report a prespecified equivalence margin, a minimal clinically important difference, or a precision target for this endpoint. Consequently, the interval’s clinical implications cannot be determined from the registry 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 participants and not total events. The exception described for time-to-event tables does not apply to this continuous outcome. The registry’s planning illustration uses a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It considers effects equal to 100%, 75%, and 50% of the observed effect. These are planning scenarios based on the observed estimate, not the original design assumptions for the trial.

For the scenario using 100% of the observed effect, the illustration gives 3,483 participants per arm for 80% power and 4,663 participants per arm for 90% power. For 75% of the observed effect, it gives 6,192 participants per arm for 80% power and 8,289 participants per arm for 90% power. For 50% of the observed effect, it gives 13,931 participants per arm for 80% power and 18,650 participants per arm for 90% power.

In plain language, 80% power describes a design that would have an 80% chance of detecting the assumed effect under the planning model if that effect were the true one; 90% power means a 90% chance under those same conditions. These are hypothetical design probabilities, not probabilities that the observed result is correct. The sample-size figures are not observed power.

Smaller assumed effects require more participants per arm to achieve the same target power, because a smaller group difference is harder to distinguish from random variation. The trial’s own original sample-size justification and design assumptions are not reported in the registry record. The posted planning illustration should therefore be understood as a separate calculation based on the observed result, not as evidence about how the trial was designed.

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 = 2.400; pooled standard deviation σ = 35.748.
2. n = 2 × (1.96 + 0.842)² × 35.748² / 2.400² = 3,482.7, rounded up to 3,483 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 <- 2.4                      # observed difference in means
sigma <- 35.7481358043066                     # 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(687076)
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 = 3,483 (80% power), 4,663 (90% power)
 75% of observed effect: n per arm = 6,192 (80% power), 8,289 (90% power)
 50% of observed effect: n per arm = 13,931 (80% power), 18,650 (90% power)
power.t.test: 3483.7 
Simulated power at n = 3,483 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.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists two outcome measures and one posted secondary outcome: change in total cholesterol from baseline to Month 12. The primary outcome description also says that SFA lumen volume and SFA total vessel volume were analyzed alongside SFA wall volume. The record does not report a multiplicity adjustment, a hierarchy for testing outcomes, or a prespecified strategy for interpreting multiple comparisons. It is therefore not possible to determine from the registry record how the risk of false-positive findings across outcomes was handled. The primary result is labeled with an unadjusted p-value.

For missing data, the primary outcome description reports 102 randomized participants, 87 participants completing baseline MRI, and multilevel models using all available imaging data from 91 participants. It also describes 20 participants with baseline imaging only and 4 with at least two imaging visits other than baseline. Between randomization and baseline, the record states that 1 participant withdrew, 8 opted out of baseline imaging, and 6 declined baseline blood collection. These are reported as separate details; the registry does not establish whether the categories overlap.

The registry explains that multilevel models can use missing or irregularly timed outcome observations. In general, that can make more use of incomplete repeated-measures data than an analysis limited to complete cases. It does not eliminate the need to consider why measurements are missing or whether the model’s assumptions are appropriate. The record does not report missing-data sensitivity analyses, imputation methods, or a detailed account of observations by visit.

The record does not give a full analysis-population definition or state whether every randomized participant was included in all analyses. It also does not report protocol deviations or exclusions beyond the imaging and baseline details. The original statistical plan, interim analyses, and analysis-population rules are not reported in the registry record.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: myocardial infarction (6.4%), major stroke (4.3%), coronary revascularization (4.3%).

The adverse-event record lists 47 participants at risk in Triple Therapy and 48 in Mono Therapy for serious adverse events. Serious events affected 7 and 4 participants, respectively. For other adverse events, the listed at-risk counts are 47 and 48, and the affected counts are 10 and 8. Deaths are listed as 4 in each group. These are descriptive counts; the record does not provide an analysis comparing safety outcomes between groups or enough detail to infer the cause of any event or death.

The registry lists myocardial infarction at 6.4%, major stroke at 4.3%, and coronary revascularization at 4.3% among the leading serious event terms. The record excerpt does not state the denominator or event-counting convention for those percentages, so they should be read as listed rather than recalculated or treated as directly comparable group-specific rates.

At-risk counts differ from the posted primary-outcome arm counts, and the record does not explain the difference. The adverse-event summaries do not establish that each event term represents a unique participant, or that the listed events were caused by a study medication. They describe what the registry reports, not a complete safety assessment.

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

The primary result is an estimate of the difference between two medication strategies on MRI-measured change in SFA wall volume over 24 months. The reported mean difference is 2.4, and its 95% confidence interval ranges from -12.3 to 17.1. The interval includes zero and allows values in either direction, so the posted data do not establish a precise difference between the group summaries.

This is not the same as showing that the strategies are equivalent. A conclusion of equivalence requires a prespecified range of differences considered acceptably small and an analysis designed to assess that range. Neither is reported in the registry record. The result also does not prove that no difference exists; the uncertainty interval includes a range of possibilities.

The endpoint is an imaging measure, not a direct measure of leg symptoms, mobility, heart attack, stroke, or survival. The registry lists total cholesterol change as a secondary outcome, but the registry entries described for this analysis do not include its result. The primary estimate therefore should not be extended to outcomes that are not reported.

The analysis used multilevel models and all available MRI data from 91 participants, according to the registry description. The enrolled total was 102, and the record separately states that 87 completed baseline MRI. The differing counts and incomplete imaging mean the primary model’s available data are not simply the same as the total randomized group. The registry does not provide detailed sensitivity analyses to show how the estimate might change under alternative assumptions about missing measurements.

Overall, the posted result is best understood as an uncertain estimate for a specific imaging endpoint and comparison. It does not support a broad conclusion about all consequences of either medication strategy, and it is not medical advice.

16. Why This Trial Matters Statistically

This trial illustrates why an effect estimate should be read together with its uncertainty. The reported mean difference is 2.4, but the 95% confidence interval extends from -12.3 to 17.1. A point estimate can look simple, while the interval shows that the data do not determine the direction or magnitude of the contrast closely. The unadjusted p-value of 0.749 does not replace that assessment.

It also demonstrates the importance of knowing which participants and observations contribute to a longitudinal analysis. The registry reports 102 randomized participants, 87 with baseline MRI completion, and a multilevel model using all available imaging from 91 participants. Those counts are not interchangeable. Repeated imaging can be incomplete or irregularly timed, and multilevel models are one way to use available observations; interpretation still depends on modeling and missing-data assumptions.

The contrast is between two active care strategies. Both arms received standard medical care, while the medication combinations differed. This distinction matters when describing the trial question: the result is not a comparison against no treatment.

Finally, the record shows the limits of registry-level interpretation. It describes the primary endpoint and broad analysis approach, but does not report the original sample-size justification, detailed model specification, or multiplicity plan. A careful statistical reading distinguishes those unreported design details from the separate planning illustration based on the observed effect. Together, these features make the trial useful for understanding how endpoint choice, incomplete longitudinal measurements, and uncertainty shape interpretation of a randomized comparison.

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 does not report the original sample-size justification or the assumptions used to plan recruitment. The sample-size illustration is based on the observed effect and cannot substitute for the trial’s original design rationale. The detailed allocation procedure, statistical analysis plan, interim analyses, and multiplicity strategy are also not reported.

The primary outcome description gives useful but incomplete information about imaging availability. It reports 102 randomized participants, 87 completing baseline MRI, and 91 contributing available imaging data to the multilevel models. The registry does not provide a full flow diagram, the number of MRI observations at each time point, or an explanation reconciling these counts. It also does not report sensitivity analyses for missing data or enough detail to assess the modeling assumptions independently.

The posted result does not specify a clinically important threshold for SFA wall volume change, and the record does not give an equivalence margin. Consequently, the confidence interval cannot be translated into a determination of clinical importance or equivalence from the registry record alone. Details of the multilevel model, including covariate adjustment and covariance structure, are not reported.

The safety summaries include counts and selected serious-event terms, but the registry excerpt does not give event-level timing, causal attribution, or a group-comparison analysis. The denominators for the listed serious-event percentages are not stated in the registry record for interpretation. These limitations mean that the counts should be treated as descriptive registry summaries rather than a complete comparative safety analysis.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Effectiveness of Intensive Lipid Modification Medication in Preventing the Progression of Peripheral Arterial Disease (The ELIMIT Study) — NCT00687076.
• Registry record last updated 2020-02-06; status last verified 2020-01.
• Results first posted 2014-08-25.
• 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 ELIMIT study was a randomized, triple-masked, parallel-group Phase 4 trial sponsored by Baylor College of Medicine and listed as completed. It enrolled 102 participants with peripheral arterial disease who had undergone endovascular intervention. The comparison was intensive lipid modification with a statin, ezetimibe, and Niaspan versus standard lipid-modifying medications with placebo ezetimibe and placebo Niaspan, with standard medical care in both arms.

The primary endpoint was change in SFA wall volume measured by MRI over 24 months. The reported mean difference was 2.4, with a 95% confidence interval from -12.3 to 17.1 and an unadjusted p-value of 0.749. The primary analysis used multilevel models and all available imaging data from 91 participants. The registry also lists change in total cholesterol from baseline to Month 12 as a secondary outcome.

The confidence interval indicates substantial uncertainty about the direction and size of the imaging outcome difference. The registry does not report the original sample-size rationale, detailed model specification, or multiplicity plan.