Antioxidant Supplementation in Pregnant Women: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Antioxidant Supplementation in Pregnant Women (NCT01232205), a randomized phase 2, phase 3 study. The primary outcome is Preeclampsia.

2. Clinical Question

The study summary says the purpose was to investigate the benefit of several antioxidant micronutrients in women with low antioxidant status and changes in cell-free mRNA. Cell-free mRNA is listed as a secondary outcome. The primary endpoint is preeclampsia, so the posted comparison addresses an event count rather than a continuous measurement of antioxidant status or mRNA. That distinction matters: the main result describes how many participants in each reported arm had the defined condition, not the degree of change in a biomarker.

3. Trial Design

The registry identifies a randomized, double-masked, parallel-group trial with phase designations of PHASE2 and PHASE3. Participants were assigned to a micronutrient antioxidant group or a control group. The registry lists an enrollment of 168. The posted primary-outcome arm counts are 52.0 and 58.0, respectively. The record does not explain the relationship between the enrollment figure and the denominators shown for this outcome, including whether the difference reflects follow-up, outcome availability, exclusions, or another reason.

The intervention was described as milk enriched with vitamins and minerals, including Cu, Zn, Mn, Fe, carotene, vitamins B6, B12, C, and E, selenium, and calcium. The control arm is named, but its description is blank. The record therefore does not establish the control product’s composition or whether it was matched to the intervention in appearance, taste, or other features. Double masking is recorded, but who was masked and how masking was maintained are not reported in the registry record.

The listed start date is 2001-06, the primary completion date is 2009-12, and the study status is COMPLETED. The primary outcome timeframe is 9 months. The registry does not give further detail about when the outcome was assessed in relation to pregnancy, the timing of individual follow-up visits, or how outcome information was collected. Those details should not be inferred from the timeframe alone.

The record lists 3 outcome measures and 1 posted secondary outcome, cell-free mRNA. It lists zero analyses. This indicates that the registry record contains outcome information but does not provide a set of analysis records. It does not establish which analyses were planned or performed outside the registry. The sponsor is Showa University.

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: There were 110 (65.4 percent) eligible subjects from 168 pregnant women, because of ferric reducing ability of plasma (FRAP) level measured

Overall 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 Preeclampsia, measured over 9 months.

The primary endpoint is preeclampsia, recorded as the number of participants with the event. The registry gives a timeframe of 9 months and defines preeclampsia as gestational hypertension, with systolic pressure at least 140 mmHg or diastolic blood pressure at least 90 mmHg on at least 2 occasions after 20 weeks’ gestation, together with proteinuria above 0.3 g/day. It also defines severe preeclampsia using specified severe blood-pressure or proteinuria criteria. These criteria set the condition’s operational definition for the record; the posted results do not separately report severe preeclampsia counts.

In general, a binary endpoint records whether each participant experienced an event, rather than measuring how much a continuous quantity changed. It can therefore be summarized as event counts and compared as a difference in event risk between groups. Here the arm results are counts of 1.0 and 8.0, with reported arm denominators of 52.0 and 58.0. The registry’s comparative estimate is a risk difference, which expresses the contrast between the event risks in the two groups.

A defined clinical endpoint makes the comparison interpretable in terms of the condition named in the trial question. However, the registry does not report adjudication procedures, assessor training, or details of how blood pressure and proteinuria were measured. Those implementation details could help readers judge consistency of outcome ascertainment, but they are not reported in the registry record.

7. Results Posted in the Registry

• Micronutrient Antioxidant: 1.9% (1/52) versus 13.8% (8/58) on control; risk difference -11.9 percentage points (95% CI -21.5 to -2.2), risk ratio 0.14 (0.02–1.08); unadjusted two-proportion z-test p = 0.023.

8. Statistical Interpretation

9. Statistical Methods Explained

What does the risk difference describe?

In general, a risk difference subtracts the event risk in one group from the event risk in the comparison group, according to the stated direction of the contrast. A value of zero represents no difference in event risk. The reported value for the micronutrient antioxidant arm is -0.119. In this record’s direction, the negative sign corresponds to a lower observed event risk in the intervention arm. The value is an absolute contrast, not a relative percentage reduction. No conversion to a percentage or other effect scale is needed to interpret its sign.

The event counts underlying the comparison are 1.0 in the intervention arm and 8.0 in control, with reported denominators of 52.0 and 58.0. The registry also gives the arm risks as 0.019 and 0.138. This kind of estimate is directly interpretable as a difference between risks, but it does not by itself explain why the observed risks differ. In a randomized trial, random assignment supports a comparative design, but the registry record alone does not provide enough information to assess every aspect of conduct, follow-up, or analysis.

How should a confidence interval be read?

In general, a confidence interval describes a range of values compatible with the data and the method used to calculate the interval. It is not a guarantee that the true effect lies within the particular limits, nor does it show the range of outcomes for individual participants. For the risk difference, the posted interval runs from -0.215 to -0.022. Both limits are below zero. The interval’s width conveys uncertainty about the magnitude of the difference; it does not remove uncertainty about study conduct, missing outcomes, or whether the analysis matches a prespecified plan.

Intervals for relative measures use a different null value. A risk ratio or odds ratio of 1 represents no relative difference. The posted risk-ratio interval extends from 0.018 to 1.077, and the odds-ratio interval extends from 0.015 to 1.016. Because each interval extends across 1, those interval summaries include the relative null value. The different interval patterns are not a contradiction in arithmetic: risk difference and relative measures use different scales and interval methods. The registry does not report enough methodological detail to identify the precise interval procedure.

What does the p-value add?

In general, a p-value measures how incompatible the observed data are with a specified null hypothesis under the assumptions of the test. It is not the probability that the null hypothesis is true, and it does not measure the probability that the result occurred by chance in a broad, assumption-free sense. The registry lists an unadjusted p-value of 0.023. The word unadjusted indicates that the posted value is not presented as adjusted for other comparisons or covariates. The registry does not report the test used, the null hypothesis in full, or whether an adjusted analysis was performed elsewhere.

A p-value does not describe the size or clinical importance of a difference. The risk-difference estimate and its interval provide information about the scale and uncertainty of the observed contrast. Because the relative-effect intervals and risk-difference interval do not have the same relationship to their respective null values, readers should not reduce the interpretation to a single significance threshold. Event counts, denominators, the effect estimate, interval, and analysis context are all relevant.

Why do sparse events and over-dispersion matter?

In general, sparse binary outcomes mean that relatively few participants experience the event. With few events, a change in one or a small number of outcomes can substantially affect the estimated group risks and comparative effect. Estimates can consequently be imprecise, and some statistical procedures may behave differently from their large-sample approximations. The posted event counts are small, so cautious interpretation is appropriate. This is a general statistical consideration, not evidence that a particular procedure failed in this trial.

Over-dispersion is a general concern when observed variation exceeds what a model assumes. It is often discussed for count or clustered data, where unmodeled extra variation can make standard errors too small if a model assumes less variability than the data contain. The posted primary endpoint is a participant-level binary event count, and the registry does not report a model or any over-dispersion assessment. There is therefore no basis in the registry record to say that over-dispersion was present or that a correction was used.

What can be inferred from randomization and masking?

In general, random allocation is intended to reduce systematic differences between groups at assignment, and masking is intended to limit influences on care, outcome assessment, or reporting. The record describes this study as randomized and double-masked. Those labels do not by themselves establish how the sequence was generated, whether allocation was concealed, who was masked, or whether masking was maintained. Those details are not reported in the registry record.

The record does not provide an analysis plan, adjustment details, or a description of analysis populations. It lists zero analyses. Thus the posted effect estimates can be described, but the registry does not support a detailed reconstruction of the statistical methods used or a judgment that all planned analyses have been reported.

10. Confidence Intervals and Precision

The primary risk-difference estimate is -0.119, and its 95% confidence interval extends from -0.215 to -0.022. The interval excludes zero, the no-difference value on the risk-difference scale. Its limits nevertheless span a range of possible magnitudes, so the estimate should not be read as an exact measure of the contrast. The interval captures sampling uncertainty under its calculation method; it does not account automatically for possible bias, outcome misclassification, missing data, or selective reporting.

The relative measures give a different precision picture. The registry reports a risk ratio of 0.139 with an interval from 0.018 to 1.077 and an odds ratio of 0.122 with an interval from 0.015 to 1.016. Both intervals include 1, the null value for those measures. Risk differences and relative measures use different scales, and their intervals can differ in their relation to the null, particularly with sparse event counts. The interval methods are not reported, so the reason for the exact pattern cannot be established from the record.

In general, confidence intervals are useful because they display a range of effects compatible with the data rather than presenting only a point estimate. They do not show the probability that a particular value is true, and they do not guarantee reproducibility in another study. Here, the few posted events and the unreported explanation for the outcome denominators limit how fully precision can be judged. The registry also does not report whether the intervals were prespecified or whether they account for any features of the analysis beyond the stated comparison.

The unadjusted p-value is 0.023, but a p-value does not replace the interval as a description of effect-size uncertainty. Readers should consider the event counts, denominators, risk difference, relative estimates, and their intervals together. The posted information supports describing the direction of the observed risk-difference estimate, while leaving meaningful uncertainty about its magnitude and the robustness of the result.

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 sample-size figures are participants per arm, not total event counts. They are planning illustrations calculated from the observed effect, rather than the trial’s own design targets. The listed 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 original trial’s own sample-size assumptions and justification are not reported in the registry record.

The 80% power column gives the illustrated number of participants per arm for a design with an 80% chance of detecting the specified effect under the stated assumptions, if that effect and those assumptions were correct. The 90% power column gives the corresponding planning count for a 90% chance. These are prospective planning quantities. They are not observed power and do not describe the probability that the completed trial’s result is true.

The rows use the full observed effect and smaller fractions of that effect. For planning at 100% of the observed effect, the figures are 80 participants per arm for 80% power and 106 participants per arm for 90% power. For planning at 75% of the observed effect, they are 167 and 223 participants per arm. For planning at 50% of the observed effect, they are 430 and 574 participants per arm. Smaller effects need more participants to detect under the same planning assumptions, which is why the illustrated requirements rise as the target effect becomes smaller.

These calculations use an effect estimated from the trial’s posted results. Such an estimate is uncertain, and a planning illustration based on it should not be treated as a reliable forecast for a future study. The figures do not reveal the sample-size calculation used when this trial was designed, and they do not establish what the trial was powered to detect.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.138; treatment proportion p₁ = 0.019; average p̄ = 0.079.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.3805 = 0.7458.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.3712 = 0.3124.
4. n = (0.7458 + 0.3124)² / (-0.1187)² = 79.5, rounded up to 80 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
p0 <- 0.137931034482759                          # control proportion
p1 <- 0.0192307692307692                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(1232205)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 80 (80% power), 106 (90% power)
 75% of observed effect: n per arm = 167 (80% power), 223 (90% power)
 50% of observed effect: n per arm = 430 (80% power), 574 (90% power)
power.prop.test: 79.5 
Simulated power at n = 80 per arm: 0.837 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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 3 outcome measures and 1 posted secondary outcome, cell-free mRNA. The primary outcome is preeclampsia. It lists zero analyses, so the record does not contain a detailed set of statistical analysis results or a description of how outcomes were prioritized. The number of listed measures does not show how many comparisons were planned or performed.

Multiplicity arises in general when several outcomes, time points, subgroup comparisons, or statistical tests are examined. Repeated testing can increase the chance of at least one result appearing notable under a null setting. Researchers may address this with a prespecified hierarchy or adjustment, but the registry does not report whether such an approach was planned or used. The unadjusted p-value of 0.023 should therefore not be assumed to account for other outcomes or analyses.

The primary outcome arm denominators are 52.0 and 58.0, while registered enrollment is 168. The record does not explain the difference. It does not report participant flow, missing primary outcome data, reasons for missingness, exclusions, or the populations included in the analysis. It is therefore not possible to tell whether all randomized participants were analyzed, whether the reported denominators reflect an available-case analysis, or whether another analysis population was used.

In general, missing outcome data can affect interpretation if the chance of an outcome being observed is related to the outcome itself or to group assignment. The registry provides no missing-data handling method, sensitivity analysis, or imputation details. This absence does not establish that data were missing or mishandled; it means that these issues cannot be assessed using the registry record.

The record also does not report protocol deviations, adherence, allocation concealment, masking procedures, or adjustments for baseline characteristics. These are not grounds to assume a problem, but they are relevant elements of trial interpretation that remain undocumented in the posted information. The registry’s zero analysis entries should not be read as proof that no analyses were conducted; it indicates only that no analyses are listed in the record.

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

The posted results show 1.0 preeclampsia event in the micronutrient antioxidant arm and 8.0 in the control arm. The risk-difference estimate is -0.119, with a 95% confidence interval from -0.215 to -0.022 and an unadjusted p-value of 0.023. This is the result recorded for the primary endpoint, under the comparison direction used in the registry. It describes the observed difference between the reported groups; it does not alone establish why that difference arose.

The result does not show that the supplement works or does not work. The registry record lacks information needed to assess fully how participants were followed, how outcomes were measured, why the endpoint denominators differ from enrollment, and what analysis populations or statistical procedures were used. Randomization and double masking are recorded, but the registry does not report their implementation details. The relative-effect intervals also extend across 1, while the risk-difference interval lies below zero, underscoring that interpretation depends on the measure and its uncertainty.

The study summary concerns antioxidant micronutrients and changes in cell-free mRNA, but the posted primary result is preeclampsia. Cell-free mRNA is listed as a secondary outcome, and no secondary analysis results appear in the registry record. The primary endpoint result therefore should not be treated as evidence about the secondary biomarker outcome or about all possible effects of the intervention.

General statistical interpretation requires separating a numerical association from a causal conclusion. Random assignment is intended to support causal comparisons, but this registry entry does not provide the detail needed to evaluate implementation, missingness, adherence, or selective reporting. Nor does it show whether the observed estimate would be reproduced in another population or under another delivery protocol. The narrow, defensible statement is that the posted comparison reports fewer primary endpoint events in the intervention arm and an estimated negative risk difference with the stated interval. Broader conclusions go beyond what the record reports.

15. Why This Trial Matters Statistically

This trial is statistically informative as an example of why a point estimate should be read with its underlying event counts and uncertainty. The primary comparison is based on 1.0 and 8.0 reported events, and its risk-difference interval lies below zero. The relative-effect intervals, however, extend across their null value. Looking at more than one effect scale helps show that a result’s apparent precision can depend on how the comparison is expressed.

The registry also illustrates the importance of distinguishing enrollment from the denominators used in an outcome analysis. Enrollment is listed as 168, while the primary endpoint arm counts use denominators of 52.0 and 58.0. Without participant-flow and missing-data information, readers cannot determine why these figures differ or how the analyzed groups were defined. That uncertainty is a reporting limitation, not proof of a particular cause.

Another important statistical distinction is between a trial’s original design and retrospective planning illustrations based on an observed effect. The sample-size figures use the observed result and stated planning assumptions. They can illustrate how smaller target effects require larger designs, but they do not reveal the trial’s original design rationale or its planned detectable effect.

Finally, a completed randomized, double-masked trial record can still leave important analytical questions unanswered. The registry does not include analysis entries, details on multiplicity, missing outcomes, or a statistical plan. These gaps demonstrate why event counts, effect estimates, intervals, p-values, design labels, and reporting completeness each contribute different information. No one item is a substitute for the others.

16. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

Several trial-specific details needed to evaluate the posted comparison are not reported in the registry record. The control arm has no descriptive text, so its contents and how it was delivered cannot be assessed. The quantities of the micronutrients in the enriched milk are also not listed. Although the study is labeled double-masked, the record does not identify who was masked or explain the masking procedure.

The registry provides an enrollment of 168, but the primary-outcome denominators are 52.0 and 58.0. It does not provide a participant-flow account explaining this difference. There is no reported information about outcome availability, losses to follow-up, exclusions, adherence, or the analysis population. The record also does not describe how preeclampsia was assessed in practice, whether outcome assessors were blinded, or how the stated definition was applied.

The statistical methods are not described in enough detail to reconstruct the analysis. The registry gives a risk difference, relative estimates, intervals, and an unadjusted p-value, but does not state the tests or interval procedures used, whether covariates were included in any analysis, or whether sensitivity analyses were conducted. It lists zero analyses. The record does not state whether multiplicity was addressed or whether an analysis plan was prespecified.

Adverse-event fields list zero for participants at risk and affected in both the serious and other adverse-event categories. The record does not explain the scope or completeness of adverse-event collection. Deaths are not reported as numerical results in the record. These fields should be read as registry entries, not as a complete account of safety monitoring.

These limitations restrict the conclusions that can be drawn from this registry entry. They do not establish that a particular procedure was absent or that a specific bias occurred; they identify information that is not reported in the registry record.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Antioxidant Supplementation in Pregnant Women — NCT01232205.
• Registry record last updated 2010-12-02; status last verified 2009-12.
• Results first posted 2010-12-02.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

20. Record Summary

The registry describes NCT01232205, Antioxidant Supplementation in Pregnant Women, as a completed randomized, double-masked, parallel-group study with PHASE2 and PHASE3 designations. Enrollment is listed as 168. The intervention is milk enriched with vitamins and minerals, and the comparator is named Control. The primary endpoint is preeclampsia over a stated timeframe of 9 months.

The posted primary outcome counts are 1.0 in the micronutrient antioxidant arm, with a denominator of 52.0, and 8.0 in control, with a denominator of 58.0. The risk difference is -0.119, with a 95% confidence interval from -0.215 to -0.022 and an unadjusted p-value of 0.023. The relative-effect intervals extend across 1.

Cell-free mRNA is the listed secondary outcome. The record does not report a statistical analysis plan, participant flow, the reason for the difference between enrollment and outcome denominators, or details of the control product and masking procedures. No analysis entries are listed.