Alzheimer's Disease Acitretin Medication: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Alzheimer's Disease Acitretin Medication (NCT01078168), a randomized phase 2 study. The primary outcome is Difference in Cerebrospinal Fluid (CSF) Soluble Alpha-clevaed Amyloid Precursor Protein (APPsα) Concentration at Visit 3 Compared to Baseline.

2. Clinical Question

3. Trial Design

This was a phase 2, randomized, double-masked, parallel-group study comparing oral acitretin with placebo in people with mild to moderate Alzheimer disease. The registry describes the study as multicenter and prospective. Acitretin was given at 30 mg per day from day 1 to day 28; the placebo arm also followed a day 1 to day 28 schedule. The registry record lists an enrollment of 22 and a completed status.

The primary outcome concerned CSF soluble alpha-secretase-cleaved amyloid precursor protein, APPsα. Samples were collected at baseline and at visit 3, described as 4 weeks. The outcome was expressed as a fold change from baseline to visit 3, with group means and standard deviations. Western blotting was used, with normalization based on hSA levels in individual samples.

The primary outcome population was limited to patients with successful punctures at both visit 1 and visit 3. The registry notes that one acitretin-treated patient did not have a visit 3 puncture, and reports an outcome group size of 10 for acitretin. The posted outcome also gives 10 participants for placebo. How participants without a successful pair of samples were handled beyond this description is not reported in the registry record.

The registry identifies one primary outcome measure and one posted analysis. It records a two-sided t-test for the sponsor’s comparison of final values. Details of the original statistical plan, including randomization procedures, allocation concealment, sample-size justification, interim analyses, and any prespecified analysis population, are not reported in the registry record. The registry lists K. Lieb as sponsor and gives a completed status.

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: neuropsychiatric testing, general exclusion of e.g. depression or severe diseases affecting the trial

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All patients were included with successful punctation at V1 and V3. For one Acitretin-treated patient not punctation at V3 was obtained (resulting n of 10).

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 Difference in Cerebrospinal Fluid (CSF) Soluble Alpha-clevaed Amyloid Precursor Protein (APPsα) Concentration at Visit 3 Compared to Baseline, measured over baseline and 4 weeks (visit 3).

The primary endpoint is the difference between groups in the mean CSF APPsα fold change at visit 3 compared with baseline. APPsα is the soluble alpha-secretase-cleaved form of amyloid precursor protein. The trial description states that the study aimed to assess alpha-secretase activity through changes in this CSF measure. It also states that samples were assessed by Western blotting and normalized using hSA levels from the individual samples.

A fold-change measure expresses a follow-up value relative to a baseline value, rather than reporting only the follow-up concentration. The registry identifies the measurement timeframe as baseline and 4 weeks, corresponding to visit 3. The reported endpoint is continuous, and the registry gives a group mean and standard deviation for each arm.

Using a CSF biomarker as the primary endpoint addresses the study’s stated biological question: whether the marker changed under the assigned treatment. It is not, by itself, a direct measure of everyday functioning, symptom burden, or disease progression. The registry record does not report a clinical outcome as a primary measure, and it does not establish how a change in this biomarker relates to a clinical benefit. That distinction matters when interpreting the result: the statistical comparison concerns the specified laboratory measure over the stated interval.

7. Results Posted in the Registry

• Acitretin: mean 1.25 versus 0.87 on control; difference 0.38 (95% CI 0.05 to 0.70), standardized effect d = 1.02; z-test p = 0.022.

8. Statistical Interpretation

9. Statistical Methods Explained

What kind of outcome was analyzed?

The primary outcome is continuous: a fold-change measure of CSF APPsα from baseline to visit 3. For a continuous outcome, a group can be described by its mean and a measure of spread. The registry reports the mean and standard deviation for each arm. The mean describes the group’s average on the measurement scale; the standard deviation describes how spread out individual observations are around that mean. A standard deviation is not the uncertainty in the estimated group mean, and it does not show the full distribution of the observations.

The endpoint is based on samples taken at two visits, while the posted comparison is described as a difference in final values. The registry record does not say whether the sponsor’s analysis adjusted for baseline values, or specify how the fold-change values were derived for each participant beyond naming the baseline and visit 3 comparison. These details can affect interpretation of an analysis of repeated measurements, so they should not be assumed.

What does the mean difference represent?

The reported mean difference is the estimated difference between the acitretin and placebo group means on the endpoint scale. The re-calculated estimate is 0.377. A mean difference retains the units of the outcome, in this case the fold-change scale; it is not a percentage difference. The registry also gives a pooled standard deviation of 0.369 and a standardized effect size, d, of 1.022 for the unadjusted re-calculation. A standardized effect expresses a mean contrast relative to a measure of variability, but its numerical size does not determine clinical importance.

For a simple two-group comparison, a t-test evaluates whether the data provide evidence of a difference in means under the test’s assumptions. The sponsor’s posted method is a two-sided t-test. A two-sided test allows departures in either direction, rather than testing only for an increase or only for a decrease. The registry does not report whether the sponsor used a pooled-variance or unequal-variance version of the t-test, nor does it report diagnostic checks for the assumptions.

Why do the analysis details matter?

In general, a t-test relies on an appropriate independent-group comparison and on a suitable model for the outcome data. With a small outcome population, the estimate and test can be sensitive to individual observations and to distributional features. If the distribution is strongly skewed or contains influential values, a mean-based method may not summarize the data well. The registry record does not report the individual observations, distributional diagnostics, or a robustness analysis, so those features cannot be assessed from the posted summary statistics.

The unadjusted re-calculation uses the posted group summaries to estimate a contrast and its uncertainty without incorporating covariates or an adjustment for other outcomes. It is distinct from the sponsor’s reported analysis, which is described as a t-test of final values. The re-calculated estimate is not a substitute for a prespecified sponsor analysis. Since the registry does not flag a discrepancy, the two results can be presented as separate accounts of the same endpoint rather than resolving their different interval and p-value values by assumption.

In general, adjustment for baseline values can improve precision when baseline measurements are related to follow-up values, but whether such adjustment was planned or used is not reported in the registry record. Likewise, the record does not describe transformations, outlier handling, or any alternative analysis. No such procedures should be inferred from the summary values alone.

How should variability and uncertainty be distinguished?

The group standard deviations, 0.453 for acitretin and 0.259 for placebo, describe observed variability in the outcome values within each group. The reported standard errors for the group means are 0.143 and 0.082, respectively. In general, a standard error describes the sampling uncertainty of an estimate under a statistical model; it tends to be smaller than the underlying individual-level variability because it concerns an estimated mean. The registry’s summary values do not give individual-level data, so the spread and shape of the distributions cannot be reconstructed from them.

A confidence interval for a mean difference combines an estimate with its uncertainty under the calculation method. The re-calculated 95% interval is 0.053 to 0.700. The sponsor reports a different 95% interval, from 0.030 to 0.724. An interval that is wider reflects less precision under that particular calculation, though interval width alone does not capture every limitation of a study. Confidence intervals do not correct for missing outcome measurements, selective reporting, or choices made after looking at data.

What is the role of the p-value?

The re-calculated unadjusted p-value is 0.022; the sponsor reports 0.035. In general, a p-value is calculated relative to a null hypothesis and a statistical model. Smaller p-values indicate that data like those observed would be less compatible with that null model, if its assumptions hold. A p-value does not state the probability that the result is due to chance, and it does not convey the probability that a treatment effect exists. It also does not measure the size, importance, or reproducibility of the observed contrast.

The registry does not report an adjustment for testing multiple outcomes or analyses. The primary endpoint is identified, and one analysis is posted, but that information does not establish whether a multiplicity plan was prespecified. The record also does not report interim analyses or a statistical analysis plan. The stated p-values should therefore be understood as the values from their respective reported or re-calculated analyses, not as a complete account of all possible analyses.

10. Confidence Intervals and Precision

The re-calculated mean difference is 0.377, with a 95% confidence interval from 0.053 to 0.700. The sponsor’s result gives the same point estimate and a 95% confidence interval from 0.030 to 0.724. Both intervals are above zero, but they differ in their endpoints and width. The sponsor’s reported interval is wider. This indicates that the estimated contrast has uncertainty; it should not be read as a precise single value.

In general, a confidence interval describes the range of values compatible with the observed data under the method used and its assumptions. It is not a guarantee that the true effect lies within the interval, and it does not mean there is a 95% probability that the particular interval contains the true value. It also does not incorporate uncertainty from every aspect of trial conduct, including missing measurements or analytical decisions that are not documented.

There are 10 participants reported in each primary-outcome group, and the group standard deviations are 0.453 for acitretin and 0.259 for placebo. With limited outcome information, the estimated difference may be sensitive to the observed values and to modeling choices. The interval quantifies sampling uncertainty according to its calculation, not uncertainty about whether APPsα is a clinically meaningful endpoint. The registry does not provide additional precision measures for the biological or clinical significance of this change.

The re-calculated interval and the sponsor’s interval are both relevant descriptions of uncertainty, but they come from different reported approaches. The registry does not explain the exact reason for the difference. The sponsor’s analysis is the trial’s reported analysis; the re-calculation is an unadjusted analysis from the posted summary statistics. Neither interval can be used to infer results beyond the measured outcome, population, and timeframe.

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 estimates are participants per arm, not total event counts. They are illustrations based on the observed mean difference, with two-sided significance level 0.05, equal allocation between arms, and a normal approximation. The effect to detect is set as a stated fraction of the observed effect. These are planning assumptions for the illustration; the original trial’s own sample-size assumptions are not reported in the registry record.

For planning around 100% of the observed effect, the estimates are 16 participants per arm for 80% power and 21 participants per arm for 90% power. For 75% of the observed effect, the corresponding estimates are 27 and 36 participants per arm. For 50% of the observed effect, they are 61 and 81 participants per arm. The 80% column means the design would have an 80% chance of detecting the specified effect under the illustration’s assumptions if that effect were true. The 90% column expresses the same idea using a higher target chance of detection.

The pattern reflects a general principle: smaller effects are harder to distinguish from random variation, so planning for smaller effects requires more participants. Increasing the desired chance of detecting an effect also requires a larger planned sample under otherwise similar assumptions. These figures are not observed power and do not describe what the completed trial achieved. They are prospective planning illustrations, not evidence about the reliability or clinical importance of the observed result.

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 = 0.377; pooled standard deviation σ = 0.369.
2. n = 2 × (1.96 + 0.842)² × 0.369² / 0.377² = 15.0, rounded up to 16 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 <- 0.3769                      # observed difference in means
sigma <- 0.368925791996168                     # 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(1078168)
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 = 16 (80% power), 21 (90% power)
 75% of observed effect: n per arm = 27 (80% power), 36 (90% power)
 50% of observed effect: n per arm = 61 (80% power), 81 (90% power)
power.t.test: 16.1 
Simulated power at n = 16 per arm: 0.799 (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. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary comparison as a two-sided t-test of final values. The reported mean difference is 0.377, with a 95% confidence interval from 0.030 to 0.724 and p = 0.035. This is the trial’s own posted analysis and should be distinguished from an unadjusted re-calculation based on group summary statistics.

The re-calculated analysis gives the same mean difference, 0.377, but a 95% confidence interval from 0.053 to 0.700 and an unadjusted p-value of 0.022. The difference in interval limits and p-values shows that the uncertainty calculations are not identical. The registry record does not describe enough analytical detail to account for the difference, so the two sets of inferential values should not be treated as interchangeable.

The sponsor’s analysis is the appropriate reference for what the trial reported. The re-calculation is useful as a separate, transparent summary-statistics-based comparison, but it does not replace the sponsor’s analysis or establish that the trial used the re-calculation’s assumptions. The registry does not give a fuller statistical analysis plan or the exact choices underlying the sponsor’s test.

14. Multiplicity, Missing Data and Analysis Populations

The registry identifies one primary outcome measure and posts one analysis. It does not report secondary outcomes. Although the record names a primary endpoint, it does not report whether the analysis was selected in advance through a published protocol or statistical analysis plan, or whether any testing hierarchy was specified. Adjustment for multiplicity, if any, is not reported in the registry record.

Multiplicity matters when several outcomes, time points, subgroups, or analytic approaches are tested. In general, testing many hypotheses can increase the chance of at least one apparently unusual result under a setting where no effects exist. Methods for controlling that issue depend on the intended family of tests and the prespecified plan. The registry does not provide enough information to determine whether any such plan applied to this study. The posted p-values are therefore described as reported for their respective analyses and should not be assumed to be multiplicity-adjusted.

The primary outcome population was restricted to patients with successful punctures at both visit 1 and visit 3. The registry notes that one acitretin-treated patient had no visit 3 puncture, resulting in an acitretin outcome group size of 10. The placebo outcome group size is also reported as 10. The registry does not describe any imputation for missing samples, sensitivity analysis, or other missing-data method. It also does not state whether the analysis followed an intention-to-treat principle or another prespecified population definition.

Restricting analysis to participants with paired measurements can affect interpretation if the availability of those measurements is related to outcomes or treatment assignment. In general, missingness can reduce precision and may introduce bias if the participants with observed data differ systematically from those without observed data. The registry record does not provide enough information to assess the reason for missingness or its likely impact.

The record does not report interim analyses, stopping rules, protocol deviations, or a detailed account of exclusions. These omissions do not establish that such procedures were absent; they mean that they cannot be evaluated from the registry entry. Interpretation should remain tied to the described outcome population and the analysis details that are actually reported.

15. 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.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: persistent diarrhea (9.1%).

For serious adverse events, the registry reports 1 affected participant among 11 at risk in the acitretin group and 0 affected among 10 at risk in the placebo group. The listed serious adverse event term is persistent diarrhea, reported as 9.1%. The registry does not provide enough detail in these figures to determine the full event-level context, timing, or relationship assessment.

The same record lists other adverse events in 8 of 11 participants at risk in the acitretin group and 5 of 10 in the placebo group. These are separate figures from the serious adverse event counts. They describe the recorded counts and denominators; they do not show whether events were mild, related to treatment, or resolved. Death counts are not reported in the registry record.

In general, adverse-event counts from a small study are descriptive and can be imprecise. A zero count in an arm means no affected participants were recorded in that group; it does not establish that the underlying risk is zero. The registry states that safety and tolerability were study aims, but the posted counts alone do not provide a complete safety profile or support conclusions beyond the observed events and recorded denominators.

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

The primary result describes a difference in mean CSF APPsα fold change between the acitretin and placebo groups at visit 3 relative to baseline. The registry reports a mean of 1.251 for acitretin and 0.874 for placebo, with a re-calculated mean difference of 0.377. The sponsor’s reported estimate is also 0.377. The confidence intervals and p-values indicate statistical uncertainty around that estimated contrast, and the sponsor and re-calculated intervals are not identical.

This is a result about a laboratory biomarker measured in CSF, not a direct result about symptoms, daily function, or the course of Alzheimer disease. The trial description frames APPsα as a measure relevant to alpha-secretase activity. A difference in the marker does not by itself establish that a biological process changed in a clinically important way, and the registry does not report evidence linking this result to a clinical outcome.

The study was randomized and double-masked, but its primary outcome population comprised 10 participants in each group with successful punctures at both specified visits. One acitretin-treated participant lacked a visit 3 puncture. The registry does not report further missing-data methods or enough analysis detail to resolve all questions about the handling of the outcome population. The estimates should therefore be interpreted in the context of the participants and measurements included in the posted outcome.

The study’s stated interval was baseline to 4 weeks, and the treatment schedule ran from day 1 to day 28. The result does not establish what would happen over longer follow-up, in other populations, or with other dosing schedules. Nor does a p-value establish the probability that a treatment effect exists or that the observed difference is important. The most bounded reading is that the posted analysis estimates a group contrast in the specified biomarker over the reported interval, with uncertainty described by the reported confidence intervals.

17. Why This Trial Matters Statistically

This trial is statistically instructive because it pairs a biological rationale with a continuous laboratory endpoint. The primary question concerns a mean difference in CSF APPsα fold change, making the distinction between the measured biomarker and a clinical outcome important. Statistical evidence about the biomarker does not, by itself, answer a clinical question about symptoms or function.

The registry also makes clear that the analyzed population depends on successful sampling at two visits. The acitretin outcome group excludes a participant without a visit 3 puncture, while the posted primary outcome includes 10 participants in each arm. This illustrates how sample availability can shape an analysis population and why the handling of missing measurements matters. The registry does not report imputation or sensitivity analyses that would clarify the effect of missing samples.

There is also a useful distinction between the sponsor’s reported test and a summary-statistics re-calculation. Both give the same point estimate, while their confidence intervals and p-values differ. That makes it important to report the analysis method alongside the estimate, rather than treating a p-value as a result independent of how it was calculated. The record does not contain enough detail to explain every difference.

Finally, the sample-size illustrations demonstrate how planning depends on the effect chosen for detection and on the desired probability of detecting it. They are not evidence about observed power or a substitute for the trial’s original design rationale. Together, these features show why effect estimates, precision, endpoint meaning, missingness, and analysis choices should be considered alongside a nominal test result.

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 registry record describes a limited primary-outcome population: 10 participants in each group with successful punctures at visit 1 and visit 3. It notes that one acitretin-treated patient did not have a visit 3 puncture. It does not explain why the sample was unavailable, whether missingness was related to outcome, or whether any analysis addressed the missing measurement. This limits assessment of how representative the analyzed participants are of the randomized groups.

The study used a biomarker endpoint over a short stated timeframe, baseline to 4 weeks. The registry does not report a primary clinical outcome, longer-term follow-up result, or evidence connecting the observed biomarker contrast to a clinical effect. The primary endpoint therefore answers a focused laboratory question rather than a broader question about disease course.

Analytical reporting is also limited. The sponsor’s analysis is described as a two-sided t-test of final values, but the registry does not specify the variance version, baseline adjustment, diagnostic checks, or handling of outliers. The re-calculated confidence interval and p-value differ from the sponsor’s values, and the record does not explain the difference. No statistical analysis plan, multiplicity strategy, interim analysis information, or sample-size justification is reported in the registry record.

The safety results are counts from the recorded at-risk groups and do not provide a complete account of event timing, severity beyond the serious-event classification, or relationship to treatment. Death counts are not reported. The small outcome groups also mean that descriptive rates and estimates can be sensitive to individual observations. These limitations do not negate the reported data, but they constrain how far the statistical and safety findings can be generalized.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Alzheimer's Disease Acitretin Medication — NCT01078168.
• Registry record last updated 2018-02-05; status last verified 2017-07.
• Results first posted 2018-02-05.
• 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

NCT01078168 is a completed phase 2 randomized, double-masked, parallel-group study titled Alzheimer’s Disease Acitretin Medication. The registry describes it as a multicenter prospective placebo-controlled study sponsored by K. Lieb, with an enrollment of 22. The study compared oral acitretin at 30 mg per day from day 1 to day 28 with oral placebo.

The primary outcome was CSF APPsα fold change at visit 3 compared with baseline, assessed by Western blotting and normalized using hSA levels. The posted outcome includes 10 participants in each group. The sponsor reports a two-sided t-test of final values; an unadjusted re-calculation gives the same mean difference but different confidence limits and p-value.

The registry lists a start of 2010-03, primary completion of 2013-01, and completed status. Further details about the original sample-size rationale, missing-data methods, multiplicity plan, and statistical analysis plan are not reported in the registry record.