High Dose Vitamin B1 to Reduce Abusive Alcohol Use: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for High Dose Vitamin B1 to Reduce Abusive Alcohol Use (NCT00680121), a randomized phase 4 study. The primary outcome is Change in Average Daily Alcohol Consumption.

2. Clinical Question

3. Trial Design

The registry identifies this as a completed, randomized, quadruple-masked, parallel-group Phase 4 study sponsored by the University of Kansas. It lists enrollment as 120. The interventions were benfotiamine and placebo, with the control group receiving placebo and the benfotiamine arm listed as receiving 600 mg. The record does not define which groups of people were masked, or provide details about the randomization sequence or allocation concealment.

The study began in 2008-07, and primary completion is recorded as 2011-08. The brief summary describes examining drinking patterns over 6 months of continued supplement use. The primary outcome is change from baseline to 6 months in average daily alcohol consumption, expressed as standard drinks per day. Posted arm-level outcome values are based on n=33 for control and n=37 for benfotiamine. These outcome counts should not be treated as a complete account of participant flow: reasons for the counts, missing outcome data, and analysis-population definitions are not reported in the registry record.

The registry lists three secondary measures at 6 months: the Alcoholism Severity Scale, Barrett Impulsivity Scale: Total Impulsiveness, and Symptom Checklist-90 Global Severity Index. It reports no analyses in the analyses field. The protocol, detailed statistical analysis plan, original sample-size justification, and interim-analysis information are not reported in the registry record. Randomization and masking describe features of the planned comparison, but they do not by themselves resolve questions about missing data, outcome measurement, or how analyses were selected and conducted.

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 Change in Average Daily Alcohol Consumption, measured over Change from Baseline to 6 Months.

The primary endpoint is the change in average daily alcohol consumption from baseline to 6 months. It is measured as alcoholic drinks per day, described in the registry as standard drinks of alcohol per day. A change score compares a participant’s follow-up measurement with their baseline measurement; a group mean summarizes those changes across the participants included in that group’s posted result.

This outcome directly represents the drinking pattern that the study set out to examine. Comparing mean changes between the benfotiamine and placebo groups focuses the question on whether the average change differed between assigned groups over the stated follow-up period. It does not show how every participant changed, how changes varied over time, or whether any observed change was sustained beyond the recorded timeframe.

The registry reports a mean and standard deviation for each group. The standard deviation describes variability among the observations contributing to a group mean; it is not the uncertainty around the group mean itself. The comparative estimate is a mean difference, with the posted direction corresponding to benfotiamine minus control. The endpoint record does not report a separate clinical threshold for interpreting a difference, or details about how drinking was measured, verified, or handled when measurements were missing. Those details are not reported in the registry record.

7. Results Posted in the Registry

• Benfotiamine: mean -3.00 versus -3.40 on control; difference 0.40 (95% CI -1.52 to 2.32), standardized effect d = 0.10; z-test p = 0.684.

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.

Alcoholism Severity Scale

• 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:
  • Benfotiamine: mean 10.70 versus 14.00 on control; difference -3.30 (95% CI -5.67 to -0.93), standardized effect d = -0.65; z-test p = 0.006.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Benfotiamine than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Barrett Impulsivity Scale: Total Impulsiveness

• 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:
  • Benfotiamine: mean 65.40 versus 65.00 on control; difference 0.40 (95% CI -4.53 to 5.33), standardized effect d = 0.04; z-test p = 0.874.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Benfotiamine and the control. The interval gives the range of effects that remain compatible with the data.

Symptom Checklist-90 (SCL-90): Global Severity Index

• 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:
  • Benfotiamine: mean 1.04 versus 1.02 on control; difference 0.02 (95% CI -0.34 to 0.38), standardized effect d = 0.03; z-test p = 0.914.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Benfotiamine and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes add context about severity, impulsivity, and psychological symptoms at 6 months, but they do not replace the primary alcohol-consumption endpoint. Across these measures, the posted pattern is not uniform: the Alcoholism Severity Scale comparison has an unadjusted interval excluding zero, whereas the impulsivity and SCL-90 comparisons have intervals that include zero. This suggests that the secondary results do not present a consistent pattern across all measured domains.

The Alcoholism Severity Scale result is notable as a secondary finding, but it should be interpreted alongside the primary result rather than as a resolution of the primary comparison. It measures a different construct at a follow-up time, while the primary endpoint is change in average daily alcohol consumption from baseline. A difference on a severity score does not establish the same contrast in drinking behavior, and a result on one secondary measure cannot by itself establish a broad pattern across outcomes.

Unadjusted secondary tests carry less weight when several outcomes are examined because the chance of obtaining at least one apparently unusual result can increase as more comparisons are considered. The registry lists three secondary outcomes, but it does not report a multiplicity adjustment, a hierarchy among them, or a prespecified strategy for interpreting them. It also does not list reported analyses. Without that information, the isolated secondary finding should be treated as exploratory rather than definitive.

The registry does not explain how the secondary scales were selected, whether they were measured at other times, or how missing data were handled for them. Their results can help frame questions for further evaluation, but the posted record does not support treating them as independent confirmation of a primary effect. Their interpretation is strongest when considered with the primary estimate, uncertainty, and the incomplete reporting of analysis methods.

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

• Alcoholism Severity Scale
• Barrett Impulsivity Scale: Total Impulsiveness
• Symptom Checklist-90 (SCL-90): Global Severity Index

9. Statistical Interpretation

10. Statistical Methods Explained

Continuous outcomes are summarized by a mean and standard deviation in each group. The mean describes the average observed value or change among the participants contributing data. The standard deviation describes how spread out those observations are around the mean. These are different summaries: a group can have a modest mean difference while individual values vary widely. The standard deviation is not a confidence interval and should not be read as a direct measure of how precisely the mean difference is estimated.

What does the mean difference represent?

For the primary outcome, the comparison is the benfotiamine group mean change minus the control group mean change. The posted estimate is 0.4 alcoholic drinks per day. A positive value means the benfotiamine group’s average change was numerically higher than the control group’s average change, according to that subtraction. Because both group means are negative, “higher” does not mean that the benfotiamine group had a larger decrease. Interpretation depends on the endpoint’s direction and on whether the contrast is large enough to matter; no threshold for clinical importance is reported in the registry record.

A group contrast is not a comparison of each participant with a matched participant in the other arm. Nor does it describe the proportion of participants whose drinking increased or decreased. Individual-level data, distribution plots, and responder definitions are not reported in the registry record. The result is therefore an average contrast, not a description of every person’s outcome.

How should the uncertainty be read?

A confidence interval expresses uncertainty about an estimated quantity under a statistical model and its assumptions. A narrower interval indicates greater precision than a wider one, while an interval that includes zero leaves open the possibility of no group difference as well as differences in either direction. The primary interval extends from -1.525 to 2.325 alcoholic drinks per day. The confidence level and the exact method used to construct the posted interval are not reported in the registry record, so those details should not be inferred.

The unadjusted p-value is 0.684. A p-value is calculated under a null model and assumptions about the data; it is not the probability that a treatment has no effect, nor a measure of the size or clinical importance of an effect. The word “unadjusted” matters: the posted comparison does not account for multiple outcomes or any other adjustment. The record does not state whether a different sponsor analysis was conducted outside the registry’s reported analyses.

Why does variation matter?

Alcohol consumption and change in consumption can vary substantially among participants. When observations are more dispersed, the group means may be estimated less precisely unless there is enough information to offset that variability. In general, over-dispersion means that observed variation is greater than a model expects. If that feature is present but ignored, uncertainty can be understated. The posted record gives standard deviations, but it does not provide diagnostic checks or a model assessment that would establish whether over-dispersion was considered.

For a continuous outcome, a mean-and-standard-deviation summary is most informative when the distribution and analysis are appropriate for those summaries. The registry does not report the distribution of the drinking measure, transformations, robust methods, or a model specification. It is therefore not possible to determine from the record whether the simple summary-based comparison captures all relevant features of the data.

What can and cannot be inferred from the posted analysis?

The comparative values are simple, unadjusted recalculations from the posted group summaries. No reported analyses are listed. The discrepancy flag is false, so there is no indicated conflict between these recalculations and a sponsor analysis in the record. This does not supply the missing details of a formal analysis plan. The record does not report whether the primary comparison was prespecified in the same form, whether baseline values were included in a model, how missing observations were handled, or which participants were included in each analysis.

Random allocation supports a planned comparison of assigned groups, but the effect estimate still depends on what data were observed and how they were analyzed. Masking may reduce some forms of bias, yet the record does not specify the masking procedures or their success. These design labels do not show whether participants adhered to the intervention or whether outcome assessment was complete.

Finally, the posted confidence intervals and p-values are not guarantees about future studies. They quantify uncertainty within a statistical approach, not the likelihood that a result will reproduce. A reader should consider the magnitude and direction of the estimate, the interval, the variability, the completeness of reporting, and the broader pattern of outcomes rather than treating a single significance test as a full interpretation.

11. Confidence Intervals and Precision

The primary mean difference is 0.4 alcoholic drinks per day, and its confidence interval runs from -1.525 to 2.325. The interval includes zero, so the posted result is compatible with no average difference as well as with a contrast in either direction. Its span indicates that the data do not pin down the primary contrast to a narrow range. The interval is more informative than the point estimate alone because it makes the uncertainty visible.

The endpoint means are based on n=33 in control and n=37 in benfotiamine. The standard deviations are 3.6 and 4.6, respectively. These show that participant-level changes varied within each arm; they do not themselves define the precision of the contrast. Precision also depends on the amount of information available and on the analysis assumptions. The record does not report the confidence level, interval-construction method, or a fuller analysis specification, so the interval should be interpreted as posted rather than attributed to an undocumented method.

The interval does not establish that every value within its bounds is equally plausible, nor does it show the probability that the true difference lies inside it. It also does not determine whether a difference would be clinically meaningful: the registry record does not report a threshold for that judgment. Because the interval includes values favoring either group, the estimate is not precise enough to settle the direction of the average contrast on its own. The p-value of 0.684 is consistent with that uncertainty, but it is not a substitute for considering the estimate and interval together.

Precision should also be distinguished from validity. A narrow interval would not correct bias from missing data, measurement problems, or selective analysis. The registry record does not report enough detail to assess those issues fully. The posted interval characterizes uncertainty in the simple comparison, not every source of uncertainty in the study.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm, not total participants or total events. The stated planning 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 a stated fraction of the observed effect. These are planning illustrations based on the observed effect, not the trial’s original design assumptions; the original sample-size justification is not reported in the registry record.

The 80% power column describes the number of participants per arm that a planning calculation estimates would be needed to have an 80% chance of detecting the chosen effect under the stated assumptions, if that effect were real and the assumptions held. The 90% column uses a higher chance of detection and therefore gives a larger planned sample for the same effect. These figures are not a guarantee that a study of that size will produce a statistically significant result.

The rows for planning for 100%, 75%, and 50% of the observed effect illustrate how the planned sample changes when the target effect is reduced. Smaller effects are harder to distinguish from random variation and therefore need more participants. The estimates for the primary endpoint are especially large relative to the posted observed contrast; they should be treated as illustrations, not as a recommendation for a future study. This is not observed power.

The planning calculation uses the observed effect, which is uncertain and may be an unstable basis for forecasting. It also uses assumptions that may not match a future study’s outcome distribution, allocation, or analysis. The registry record does not report the original planned effect, target power, or design assumptions, so the illustration cannot be used to reconstruct the study’s original rationale.

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 = 0.400; pooled standard deviation σ = 4.159.
2. n = 2 × (1.96 + 0.842)² × 4.159² / 0.400² = 1,697.4, rounded up to 1,698 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 <- 0.4                      # observed difference in means
sigma <- 4.15946829181186                     # 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(680121)
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 = 1,698 (80% power), 2,273 (90% power)
 75% of observed effect: n per arm = 3,018 (80% power), 4,040 (90% power)
 50% of observed effect: n per arm = 6,790 (80% power), 9,090 (90% power)
power.t.test: 1698.4 
Simulated power at n = 1,698 per arm: 0.802 (20000 replicates)

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

14. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary endpoint and three secondary outcomes. It does not report a multiplicity adjustment, a testing hierarchy, or a plan specifying how the outcomes were to be interpreted together. When several statistical comparisons are made, the possibility of at least one apparently notable result by chance can increase. This general concern is especially relevant to the secondary findings because the posted values are unadjusted. The registry record does not say whether any adjustment or outcome-selection procedure was conducted outside the analyses it reports.

The analyses field lists no reported analyses. As a result, the registry record does not describe the exact statistical models, analysis populations, covariate adjustments, or whether the primary comparison followed a prespecified plan. The comparison estimates are simple unadjusted recalculations from group summary statistics. The discrepancy flag is false, but that does not provide missing details about an analysis plan.

Missing-data information is also limited. The enrollment is listed as 120, while the primary arm-level results use n=33 for control and n=37 for benfotiamine. Those values alone do not explain participant flow or establish why some enrolled participants are not represented in the posted outcome summaries. The record does not report reasons for missing outcomes, timing of withdrawal, or methods such as imputation or sensitivity analyses. It also does not state whether the outcome summaries include all randomized participants or a subset.

Further details that are not reported in the registry record include an intention-to-treat definition, adherence information, handling of protocol deviations, baseline adjustment, and the treatment of outliers. These omissions do not prove that such procedures were absent from the conduct of the study; they mean that the registry record does not allow them to be evaluated. Consequently, the posted results are best read as available summary comparisons rather than as a fully documented analysis of all randomized participants.

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.

The registry lists 60 participants at risk in each group for serious adverse events and for other adverse events. It records zero affected participants for serious adverse events and zero affected participants for other adverse events in both the control and benfotiamine groups. These are the posted safety counts for the registry’s stated categories.

Zero recorded affected participants does not establish that an intervention has no safety risks. The counts describe what was entered for these categories, not the likelihood of every possible adverse outcome. The registry record does not provide event definitions, event-specific details, observation time for these safety summaries, or information on how events were collected. It also does not report death counts: those entries are not reported in the registry record. Safety interpretation should therefore remain limited to the recorded figures and their stated categories.

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

The primary result describes a small estimated difference between the posted group mean changes, with uncertainty spanning values on either side of zero. It does not establish that the groups had exactly the same average change, and it does not establish a clear difference between them. The interval and p-value should be read as a description of the posted comparison, not as a verdict about whether benfotiamine works or does not work.

The result concerns average change in daily alcohol consumption from baseline to 6 months. It does not describe every participant’s response, drinking patterns beyond that period, or outcomes for people outside the population described in the registry. The record does not report a threshold for a clinically important difference, so the practical importance of possible values within the interval cannot be determined from this result alone.

The secondary outcomes provide mixed evidence across different measures, and their unadjusted nature limits the weight they can carry. The Alcoholism Severity Scale result is distinct from the primary drinking measure; the other posted secondary comparisons do not show the same statistical pattern. Taken together, these findings do not establish a consistent effect across all recorded outcomes.

Important details needed for a broader conclusion are not reported in the registry record, including the analysis plan, missing-data handling, participant flow, and measurement procedures. The study’s randomization and masking are relevant design features, but they do not remove uncertainty arising from limited reporting. The results are most appropriately understood as estimates from the posted summaries, with the precision and limitations described, rather than as a complete account of all possible benefits, risks, or effects.

17. Why This Trial Matters Statistically

This trial illustrates why a group mean difference should be read together with its uncertainty. The primary point estimate is close to zero, but the confidence interval is wide enough to include contrasts in either direction. Focusing only on the point estimate or only on the p-value would leave out important information about the range of values compatible with the posted comparison.

It also demonstrates how variability matters. The primary outcome’s standard deviations are larger than the estimated difference between group means, so the average contrast does not capture the spread of individual changes. In clinical research, a mean can be useful for comparing groups while still concealing variation in responses. Distributional summaries and individual-level analyses would add information, but they are not reported in the registry record.

The secondary outcomes offer a further lesson about multiplicity. An isolated unadjusted result among several outcomes can appear compelling unless its place in the outcome hierarchy and testing plan is known. Here, the registry does not report such a plan, making cautious interpretation especially important. The secondary measures also address different constructs, so their results should not be treated as interchangeable evidence for the primary question.

Finally, registry results are easier to assess when analysis populations, missing-data procedures, and statistical methods are fully documented. Their absence does not demonstrate that the study was improperly conducted, but it limits what readers can verify. This record provides useful design and outcome information while leaving important aspects of the analysis and participant flow unresolved.

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 offers limited information about how alcohol consumption was measured, including whether it relied on self-report, how standard drinks were defined for participants, and how measurements were verified. These details matter because measurement error can affect both group averages and the uncertainty around their comparison. The record also does not describe adherence to benfotiamine or placebo, or whether adherence was assessed.

Participant flow is not fully described. Enrollment is listed as 120, while the posted primary results use n=33 in control and n=37 in benfotiamine. The record does not explain the relationship between those counts, identify reasons for missing outcome data, or specify the analysis population. No missing-data method or sensitivity analysis is reported.

The trial record does not provide the original sample-size rationale or the assumptions behind the study’s design. The available planning illustrations are based on the observed effect and should not be mistaken for that rationale. The analyses field is empty, and the exact model, confidence-interval method, baseline adjustment, and handling of outliers are not reported. The record also does not describe how masking was implemented or assessed.

The primary follow-up is change through 6 months. Longer-term drinking outcomes are not reported in the posted primary endpoint. The secondary measures are recorded at 6 months, but their analysis plan and multiplicity handling are not reported. These limitations concern what can be assessed from the registry record; they do not establish that the procedures were absent from the study itself.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: High Dose Vitamin B1 to Reduce Abusive Alcohol Use — NCT00680121.
• Registry record last updated 2014-09-25; status last verified 2014-09.
• Results first posted 2014-09-25.
• 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.

22. Record Summary

NCT00680121 is a completed, randomized, quadruple-masked, parallel-group Phase 4 trial sponsored by the University of Kansas. It compared benfotiamine, listed as 600 mg, with placebo for the primary outcome of change in average daily alcohol consumption from baseline to 6 months. Enrollment is listed as 120; the posted primary outcome summaries use n=33 for control and n=37 for benfotiamine.

The primary mean difference was 0.4 alcoholic drinks per day, with a confidence interval from -1.525 to 2.325 and an unadjusted p-value of 0.684. The interval includes zero and leaves uncertainty about the direction and size of the average contrast. Three secondary outcomes are listed, with no reported analyses in the registry record. Safety entries record zero affected participants for serious and other adverse events in each group; death counts are not reported in the registry record.