Dietary Carbohydrate and GERD in Veterans: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Dietary Carbohydrate and GERD in Veterans (NCT02384551), a randomized na study. The primary outcome is Gastroesophageal pH.

2. Clinical Question

3. Trial Design

The registry identifies NCT02384551 as a completed, randomized, single-masked, parallel-group trial titled “Dietary Carbohydrate and GERD in Veterans.” Its official title is “Dietary Carbohydrate Effects on GERD in Obese Veterans: Nutritional or Hormonal?” The recorded enrollment is 150. The study began on 2016-01-25, and the primary completion date is 2022-07-01. The phase is listed as NA, which is consistent with the registry not assigning a phase category.

The design has four diet arms, defined by high or low total carbohydrate and high or low total simple carbohydrate. HTHS is the reference arm in the posted comparisons. Random assignment can, in general, help balance participant characteristics across groups, but the registry record does not describe the allocation procedure or whether allocation was concealed. Masking is recorded as single; who was masked is not reported in the registry record. The parallel model means participants were assigned to a group rather than each participant receiving every diet in sequence.

The registry lists gastroesophageal pH as the primary outcome, measured over 24 hours. GERDQ and GSAS symptom scores are listed as secondary outcomes at 9 weeks. Although the record reports enrollment of 150, the primary outcome summaries list 22 participants in HTHS, 26 in HTLS, 22 in LTHS, and 25 in LTLS. The reason those outcome counts are below the planned enrollment is not reported in the registry record. The record does not provide further details on recruitment, follow-up, adherence, or participant flow.

The sponsor is the VA Office of Research and Development. The registry marks the study completed, but does not describe the trial’s original sample-size assumptions, its prespecified statistical analysis plan, or any interim analyses. Those omissions matter when interpreting whether the posted comparisons were planned, how missing observations were handled, and how the four-arm design informed the analysis.

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

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 Gastroesophageal pH, measured over 24 hours.

The primary endpoint is the mean percentage of a 24-hour monitoring period during which gastroesophageal pH was below 4. It is recorded as ambulatory pH monitoring for percent time gastroesophageal pH below 4 in 24 hours. The registry gives the unit as percent time pH below 4 in 24 hours, the summary measure as a mean, and the dispersion measure as a standard deviation.

In general, ambulatory pH monitoring captures acid exposure over an extended period rather than at one moment. Summarizing the fraction of monitored time below a pH threshold gives a continuous measure that can be compared across diet groups. A group mean describes the average observed value, while its standard deviation describes how dispersed individual measurements were within that group. These are different quantities: a spread of observations is not a confidence interval for the group mean.

The primary outcome is relevant to the trial’s stated interest in dietary carbohydrate and GERD because it is an objective physiologic measure of gastroesophageal acidity. It is not itself a direct measure of symptoms, quality of life, or medication use. The registry separately lists GERDQ and GSAS scores as secondary symptom outcomes. How the investigators defined a clinically meaningful change in pH exposure, and whether the endpoint was selected before enrollment, are not reported in the registry record.

7. Results Posted in the Registry

• HTLS: mean 6.60 versus 7.30 on control; difference -0.70 (95% CI -6.70 to 5.30), standardized effect d = -0.07; z-test p = 0.819.
• LTHS: mean 3.70 versus 7.30 on control; difference -3.60 (95% CI -8.95 to 1.75), standardized effect d = -0.40; z-test p = 0.187.
• LTLS: mean 4.90 versus 7.30 on control; difference -2.40 (95% CI -7.78 to 2.98), standardized effect d = -0.27; z-test p = 0.382.

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.

Gastroesophageal Reflux Disease Questionnaire (GERDQ) Score

• 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:
  • HTLS: mean 3.20 versus 5.10 on control; difference -1.90 (95% CI -3.89 to 0.09), standardized effect d = -0.55; z-test p = 0.061.
  • LTHS: mean 4.50 versus 5.10 on control; difference -0.60 (95% CI -3.06 to 1.86), standardized effect d = -0.14; z-test p = 0.633.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between HTLS and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between LTHS and the control. The interval gives the range of effects that remain compatible with the data.

Gastroesophageal Scale Assessment Symptom (GSAS) Score

• 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:
  • HTLS: mean 4.10 versus 5.60 on control; difference -1.50 (95% CI -2.76 to -0.24), standardized effect d = -0.68; z-test p = 0.020.
  • LTHS: mean 5.60 versus 5.60 on control; difference 0.00 (95% CI -1.39 to 1.39), standardized effect d = 0.00; z-test p = 1.000.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for HTLS than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between LTHS and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists two symptom measures at 9 weeks: the Gastroesophageal Reflux Disease Questionnaire (GERDQ) score and the Gastroesophageal Scale Assessment Symptom (GSAS) score. Both are continuous outcomes summarized by group means and standard deviations. The posted pattern is not identical across outcomes and comparisons, which is a reminder that symptom measures and ambulatory pH are related but distinct ways of describing the study question.

The symptom summaries add context to the physiologic primary endpoint, but they do not directly validate or negate it. A pH measure captures monitored acidity over 24 hours, whereas questionnaire scores capture reported symptoms using their respective scales. Associations between these measures, their clinical interpretation, and the meaning of score changes are not reported in the registry record. The secondary results therefore should be considered complementary descriptive evidence rather than interchangeable measures of one outcome.

These secondary comparisons carry less weight than a prespecified primary analysis when the registry does not report a testing hierarchy or multiplicity adjustment. Each additional outcome and contrast creates another opportunity for a result that appears notable by chance. Here, the record lists no reported analyses, and it does not say which secondary comparisons were planned or how the investigators intended to control interpretation across outcomes. The unadjusted estimates are useful for describing the posted data but cannot establish a confirmatory secondary finding on their own.

The secondary outcomes also have limits as posted: the registry does not report baseline scores, participant-level changes, or details about missing observations. Comparisons of final group means are not the same as analyses of change from baseline or repeated measurements. No such method should be assumed. Overall, these measures broaden the outcome picture but do not resolve the uncertainty in the primary pH contrasts or provide an independent basis for a definitive conclusion.

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

• Gastroesophageal Reflux Disease Questionnaire (GERDQ) Score
• Gastroesophageal Scale Assessment Symptom (GSAS) Score

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous, and the registry reports a mean and standard deviation for each arm. A mean difference is an intuitive contrast: it subtracts the reference-arm mean from the other arm’s mean. Negative estimates correspond to lower observed mean time with pH below 4 than in HTHS. The difference is expressed in the same units as the endpoint, percent time during the 24-hour monitoring period.

The posted comparisons are unadjusted recalculations from the group-level summary statistics. “Unadjusted” means the estimates do not account for other participant characteristics or covariates in the information reported. These are not identified as the investigators’ own adjusted model results. The registry lists zero analyses, so the investigators’ statistical model, if any, is not reported in the registry record. Because discrepancy is false, no disagreement between the recalculated contrasts and a sponsor analysis is flagged; the registry nevertheless does not supply a sponsor-authored analysis for comparison.

A confidence interval represents a range of values that are reasonably compatible with the observed data under the statistical method used. It conveys uncertainty about an estimated group difference, rather than the spread of individual measurements. A wider interval indicates less precision. If an interval includes zero, the data do not clearly distinguish a positive difference from a negative difference at the interval’s stated confidence level. That does not prove there is no difference. The confidence intervals for the posted primary contrasts all include zero.

A p-value describes, under a specified null hypothesis and statistical model, how unusual results at least as extreme as the observed result would be if that null hypothesis were true. It is not the probability that the null hypothesis is true, nor a measure of the clinical importance of an effect. The primary comparison p-values are explicitly unadjusted. Because the record does not report the investigators’ test or a multiplicity plan, they should not be interpreted as a complete family-wise assessment of all possible comparisons in a four-arm trial.

The standard deviation describes variation among participants’ measurements within an arm. When an outcome has substantial variation, the mean can be estimated less precisely for a given number of observations. This is one reason that an observed difference can be numerically nonzero while its confidence interval remains broad. The primary endpoint has standard deviations of 12.1 in HTHS, 8.4 in HTLS, 4.2 in LTHS, and 4.7 in LTLS. The registry does not describe the distributional shape of the measurements or report analyses designed to address skewness or unusual values.

In general, over-dispersion means that observations vary more than a statistical model assumes. If a model understates that variation, its standard errors can be too small and its confidence intervals too narrow. Conversely, accounting for greater variation can widen uncertainty intervals. This is a general statistical point, not a finding that over-dispersion was diagnosed in this trial. The registry reports standard deviations but does not report a model assessment or a specific correction for over-dispersion.

Randomization is intended, in general, to make groups comparable on average at baseline, including on characteristics that are not measured. It does not guarantee that the groups will be balanced in a modest sample, and it does not resolve missing outcomes or nonadherence. Baseline characteristics, adherence measures, and the method used to preserve the randomized comparison are not reported in the registry record. It is therefore not possible to assess these features from the posted summary alone.

The primary contrasts compare each diet arm with HTHS. This choice gives a clear reference for the reported mean differences, but it does not answer every question about the four-arm structure. In a factorial design, investigators might separately estimate effects associated with total carbohydrate, effects associated with simple carbohydrate, and whether those factors interact. The record calls the model parallel and lists four combinations, but does not report a factorial analysis. It would be inappropriate to infer that such effects were tested or to reconstruct them from the listed contrasts.

The secondary endpoints use continuous scores summarized by means and standard deviations. Their reported contrasts are also unadjusted recalculations. The registry record does not report a multiplicity adjustment, a hierarchy of secondary hypotheses, or a prespecified strategy for interpreting multiple endpoints. General statistical practice treats numerous unadjusted comparisons cautiously because, as the number of tests increases, the chance of at least one apparently unusual result can increase even if there are no underlying effects. This is a reason to consider the full testing context, not evidence that any particular result is false.

Finally, a statistical result depends on which participants contribute observations and how missing values are handled. The posted outcome counts are not identical across arms, and are below the recorded enrollment. The registry does not explain why, identify an analysis population such as intention-to-treat, or describe imputation or complete-case procedures. Those choices can affect both the estimate and its uncertainty. The reported summaries are informative descriptions of the posted observations, but they cannot substitute for details of the analysis plan and participant flow.

11. Confidence Intervals and Precision

The primary estimates are not highly precise: their confidence intervals span values on both sides of zero. The HTLS comparison has an estimated mean difference of -0.7, with a 95% confidence interval from -6.699 to 5.299. The LTHS comparison has an estimate of -3.6, with an interval from -8.952 to 1.752. The LTLS comparison has an estimate of -2.4, with an interval from -7.781 to 2.981. Each interval therefore allows for a lower mean in the comparison arm, little or no difference, and a higher mean.

In general, a confidence interval gives a range of effect sizes compatible with the data and the method, not a guarantee that the true effect lies in that range. A broad interval signals that the estimate is uncertain. It should not be reduced to a binary conclusion based on whether it crosses zero: the range of plausible values and the relevance of those values are important. The registry does not report a clinically meaningful threshold for the primary endpoint, so the practical implications of the interval’s endpoints cannot be classified from this record.

Precision depends in part on the amount of information in the comparison, including the number of observed outcomes and their variability. The primary arm counts are 22, 26, 22, and 25, while the standard deviations vary from 4.2 to 12.1. The outcome counts are much smaller than the recorded enrollment of 150; why is not reported in the registry record. These factors help contextualize the width of the intervals, but do not identify the cause of any missing precision.

Confidence intervals for these recalculated contrasts are based on the posted summary information. They do not incorporate an adjustment for multiple comparisons, and no such adjustment is reported in the registry record. The intervals quantify uncertainty under the recalculation’s assumptions; they do not cover uncertainty from unreported design choices, missing-data handling, or possible differences between the posted summaries and a fuller analysis.

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 for each two-arm contrast. They are not total event counts: the endpoint is continuous, and the listed figures refer to participants per arm. The planning assumptions described in the registry are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect set to the stated fraction of the observed effect. The trial’s original design assumptions and sample-size justification are not reported in the registry record.

The 80% power column gives the illustrative number of participants per arm that would yield an 80% chance of detecting the specified effect under those assumptions, if that effect and model were correct. The 90% power column gives the corresponding illustration for a 90% chance. These are planning calculations, not observed power, and they are not a retrospective measure of whether the completed trial was adequately designed.

The rows use the full observed effect and smaller fractions of that effect. Smaller effects are harder to distinguish from random variation, so they require more participants per arm under the same assumptions. These figures are tied to the particular observed contrast and should not be treated as universal requirements for a future study. They also do not account for issues that are not part of the stated assumptions, such as missing outcome data, a different analysis method, or adjustment for multiple comparisons.

Because the estimates come from the trial’s observed group differences, planning illustrations can be unstable when those estimates are uncertain. A small observed effect can produce very large projected sample sizes; a larger observed effect can yield smaller illustrations. The registry record does not report the original design target or its rationale, so the illustrations should be kept separate from claims about what the investigators planned before the study began.

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 = -0.700; pooled standard deviation σ = 10.256.
2. n = 2 × (1.96 + 0.842)² × 10.256² / -0.700² = 3,369.8, rounded up to 3,370 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 <- -0.7                      # observed difference in means
sigma <- 10.2560798511441                     # 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(2384551)
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 = 3,370 (80% power), 4,512 (90% power)
 75% of observed effect: n per arm = 5,991 (80% power), 8,020 (90% power)
 50% of observed effect: n per arm = 13,480 (80% power), 18,045 (90% power)
power.t.test: 3370.8 
Simulated power at n = 3,370 per arm: 0.803 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Multiplicity, Missing Data and Analysis Populations

A four-arm study can generate several possible comparisons. The posted primary results compare HTLS, LTHS, and LTLS separately with HTHS. Secondary results also include comparisons of certain arms for GERDQ and GSAS, but not every possible contrast is posted. The registry record does not report a prespecified multiplicity strategy, such as a hierarchy of hypotheses or an adjustment to account for multiple testing. The primary p-values are unadjusted, and no analysis is listed in the record.

In general, testing multiple hypotheses without adjustment can increase the probability of at least one apparently unusual result when all null hypotheses are true. This does not mean an individual result is necessarily spurious; it means the testing context is relevant to how much weight to place on isolated p-values. Without a reported analysis plan, it is not possible to determine which results were intended as confirmatory and which were exploratory. The absence of a reported plan should not be mistaken for evidence that no plan existed outside the registry.

The registry lists 150 enrolled participants, while the primary-outcome summaries list arm counts of 22, 26, 22, and 25. Secondary outcome counts follow those same arm-specific counts. Why the posted outcome counts are lower than the enrollment is not reported in the registry record. Participant flow, exclusions, withdrawals, and outcome-specific missingness are not described.

The record also does not identify the analysis population. It does not say whether all randomized participants were analyzed in their assigned groups, whether only participants with observed outcomes were included, or whether missing values were imputed. In general, these choices can affect both the estimated group difference and its uncertainty, especially if missingness relates to outcomes or group assignment. No claim about such a pattern can be made from the available record.

Baseline adjustment, adherence-based analyses, sensitivity analyses, and interim analyses are also not reported in the registry record. Their absence from the record means that the posted arm-level summaries cannot show whether those approaches were performed elsewhere. The conclusions supported by these data should therefore remain limited to the observed summaries and their unadjusted uncertainty intervals.

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.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: pill stuck in throat (2.5%).

The registry reports serious adverse event figures separately by arm. The number at risk was 35 for HTHS, 40 for HTLS, 36 for LTHS, and 39 for LTLS. One participant in HTLS was recorded as affected by a serious adverse event; zero were recorded as affected in each of the other arms. The registry lists zero deaths in every arm.

For other adverse events, the number at risk matches those serious-event group counts, and zero participants were recorded as affected in each arm. The registry’s listed serious event term was “pill stuck in throat,” with 2.5% shown as the term’s frequency. The record does not provide further detail on event timing, severity criteria, relatedness to the diet, or how the term frequency was calculated.

These figures describe events recorded in the registry; they do not establish that an intervention caused or prevented an event. The number at risk in the safety summaries differs from the outcome counts in the efficacy summaries, and the reason is not reported in the registry record. With these data alone, differences between arms should be interpreted cautiously, especially because the event counts are sparse. The registry record does not provide a comparative safety analysis.

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

The posted data describe the mean percentage of monitored time with gastroesophageal pH below 4 in each diet arm and the unadjusted differences from HTHS. The three estimated differences are negative, but all three primary confidence intervals include zero and extend across both negative and positive values. Thus, the registry summaries do not precisely determine whether any of the comparison-arm means are lower or higher than HTHS in the population.

This does not establish equivalence or show that there is no difference. Equivalence requires a prespecified margin and an analysis designed to assess whether the uncertainty is narrow enough to exclude differences considered important. No equivalence margin or clinical threshold is reported in the registry record. Likewise, an unadjusted p-value is not a measure of the size or importance of a difference, and a value above a conventional cutoff does not prove that the groups are the same.

The primary measure is an objective physiologic endpoint, but it is not a complete measure of GERD experience. The registry separately reports symptom questionnaire outcomes, and it mentions symptoms and medication use in the brief summary. Medication use is not among the posted outcomes described in the record. The posted pH and symptom results therefore should not be taken as a comprehensive account of all patient-relevant outcomes or of the trial’s full analyses.

The data also cannot establish why any group differences arose. The registry does not report enough information about adherence, participant characteristics, missing observations, or analysis populations to evaluate those explanations. The study’s randomized design is important context, but the summaries alone do not replace details about conduct and analysis.

The appropriate interpretation is limited: the registry reports group means and uncertain unadjusted contrasts for the posted outcomes. It does not support a treatment recommendation, a claim of benefit or lack of benefit, or a conclusion about clinical importance. Those judgments would require more complete information about the analysis plan, outcome interpretation, participant flow, and the broader evidence base.

17. Why This Trial Matters Statistically

This trial is statistically interesting because it compares four combinations of carbohydrate amount and type rather than a single dietary intervention with a single control. That structure raises questions about how to distinguish the contribution of each dietary dimension and whether the dimensions interact. The registry reports pairwise contrasts against HTHS, but does not report a factorial analysis or an overall test across the arms. The distinction between the design’s structure and the posted comparisons is central to reading the results correctly.

The example also illustrates how descriptive differences, uncertainty intervals, and p-values answer different questions. The group means show the observed data pattern; the confidence intervals indicate that the estimated primary contrasts remain imprecise; and the unadjusted p-values do not resolve the uncertainty or establish practical importance. Treating any one of these quantities as a complete verdict would discard important information.

The difference between recorded enrollment and the outcome counts is another statistical issue, not just a reporting detail. Without participant flow and missing-data methods, readers cannot determine why fewer observations appear in the outcome summaries or whether the analysis preserves the intended randomized comparison. Similarly, the absence of a reported multiplicity plan limits how strongly secondary findings can be interpreted.

Finally, the planning illustrations show why sample-size calculations depend strongly on the assumed effect size. They are based on observed contrasts and stated generic planning assumptions, not the trial’s documented original design. Keeping design planning separate from observed results prevents retrospective calculations from being mistaken for evidence that a study was powered to detect a particular effect.

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.

Several trial-specific limitations follow from the registry record. First, the primary outcome summaries include 22 participants in HTHS, 26 in HTLS, 22 in LTHS, and 25 in LTLS, compared with recorded enrollment of 150. The reason for the difference, including whether it reflects attrition, eligibility for the outcome, or another process, is not reported in the registry record. Participant flow and outcome-specific missingness are unavailable.

Second, the registry does not report baseline characteristics, baseline values for the outcomes, adherence, or medication use results. These omissions limit assessment of group comparability, interpretation of final scores, and understanding of whether dietary exposure differed as intended. The record’s masking designation is single, but it does not identify who was masked.

Third, no statistical analyses are listed. The registry does not report the original analysis plan, covariate adjustment, an overall four-group test, factorial effects, handling of missing data, multiplicity control, or interim analyses. The posted pairwise recalculations therefore cannot be treated as a substitute for a complete prespecified analysis. The primary outcome intervals are broad and include zero, leaving substantial uncertainty about the direction and size of the contrasts.

Finally, there are multiple outcome domains and potential contrasts in a four-arm study. The posted secondary comparisons are not exhaustive of every possible arm comparison, and the record does not explain the selection or priority of those contrasts. The registry also gives no threshold for a clinically meaningful pH or symptom-score difference. These limitations constrain both statistical and clinical interpretation without implying that unreported procedures were or were not performed outside the registry.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Dietary Carbohydrate and GERD in Veterans — NCT02384551.
• Registry record last updated 2024-09-19; status last verified 2024-08.
• Results first posted 2024-09-19.
• 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

NCT02384551 is a completed randomized, single-masked, parallel-group study of four dietary carbohydrate patterns in veterans with obesity and GERD. Recorded enrollment was 150. The primary endpoint was mean percent time with gastroesophageal pH below 4 during 24-hour ambulatory monitoring. GERDQ and GSAS scores at 9 weeks were listed as secondary outcomes.

The primary arm means ranged from 3.7 to 7.3. Each posted comparison with HTHS had a confidence interval that included zero. The registry lists no analyses, and does not report the original statistical plan, participant flow, or missing-data methods. The results describe the posted observations but leave uncertainty about the size and direction of the underlying group differences.