Effect of Metformin on Vascular and Mitochondrial Function in Type 1 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effect of Metformin on Vascular and Mitochondrial Function in Type 1 Diabetes (NCT01813929), a randomized phase 4 study. The primary outcome is Insulin Sensitivity by Hyperinsulinemic Euglycemic Clamp.

2. Clinical Question

The study’s stated rationale is that insulin resistance may contribute to cardiovascular disease risk in type 1 diabetes. Its purpose was to measure metformin’s effects on insulin sensitivity, vascular function and compliance, and mitochondrial function. The primary endpoint was insulin sensitivity; vascular, mitochondrial, inflammatory, heart-rate, and continuous-glucose measures were among the secondary outcomes listed in the record.

The registry reports mean insulin-sensitivity values of 3.27 mg/kg/min for metformin and 3.46 mg/kg/min for placebo. The simple unadjusted mean difference is -0.19 mg/kg/min, with a 95% confidence interval from -1.452 to 1.072. Those estimates describe the posted summaries; they do not establish that the treatment periods differed, nor do they establish equivalence. The record does not report a trial-conducted analysis accounting for the crossover structure.

3. Trial Design

This was a completed phase 4 study sponsored by the University of Colorado, Denver. The registry describes it as randomized, double-masked, and crossover, with metformin and placebo as the two drug interventions. Enrollment was 23. The study began in 2011-06, and primary completion was 2017-03-24. The registry status is completed.

In a crossover design, participants receive more than one intervention in separate periods, so comparisons can potentially use within-person information. That is a general feature of crossover studies, not a description of the analysis actually used. The treatment order, sequence allocation details, washout interval, adherence, and handling of period or carryover effects are not reported in the registry record. The descriptions of the metformin and placebo arms contain no further treatment details.

The primary measure was assessed at the end of each 6 week intervention period. The listed secondary measures also use that timeframe. The record lists 30 outcome measures and 22 posted secondary outcomes, but reports no analyses under its trial-analysis listing. The registry reports outcome summaries by intervention group; it does not describe how participants contributed to each summary or how the arm-level counts relate to the 23 enrolled participants. For the primary outcome, those counts are 16 for metformin and 17 for placebo.

The registry record does not report the original sample-size justification, the statistical analysis plan, any prespecified interim analyses, or whether a period-specific or paired model was planned. The unadjusted comparisons discussed in the results are simple calculations from the posted summary statistics, rather than a reported crossover analysis. That distinction matters because the simple calculation does not reflect the design’s possible within-participant structure.

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: 3 participants screen failed before starting the study. 3 more participants withdrew after randomization, but before starting the study.

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.

Population: Baseline information was collected for all participants for which outcome measure data was able to be collected. One subject only completed the placebo arm and is not included in paired t-tests. Initial random order, double-blinded design was revised to placebo-metformin, single-blinded (participants) design due to concern for carryover effects from metformin.

6. Primary Endpoint

The registry lists the primary outcome as Insulin Sensitivity by Hyperinsulinemic Euglycemic Clamp, measured over End of each 6 week intervention period.

The primary endpoint is insulin sensitivity measured by a hyperinsulinemic euglycemic clamp. During the clamp, insulin is infused while glucose is adjusted to maintain a target blood sugar level. The reported quantity is the glucose infusion rate during the final 30 minutes, normalized to total body weight and expressed in mg/kg/min. The registry describes an insulin infusion of 40 mU/m2 surface area, blood sugar checks every 5 minutes, and a target glucose level of 90 mg/dL for 2 hours.

As a general physiological interpretation of this measure, a higher glucose infusion rate under the stated clamp conditions corresponds to greater insulin sensitivity: more glucose is needed to maintain the target level when insulin is being infused. The registry also notes that there is no strictly defined normal range for this measure. It is therefore most appropriate to interpret the reported values as measurements compared between intervention periods, rather than to classify them against a normal threshold.

The endpoint was selected to measure the study’s stated primary question about metformin and insulin sensitivity in type 1 diabetes. It is a direct, experimentally controlled measure of glucose handling during the clamp, but the posted endpoint summaries alone do not establish whether any observed difference is attributable to the intervention. The registry does not report a trial-specific analysis accounting for the crossover design.

7. Results Posted in the Registry

• Metformin: mean 3.27 versus 3.46 on control; difference -0.19 (95% CI -1.45 to 1.07), standardized effect d = -0.10; z-test p = 0.768.

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.

Arterial Stiffness by PWV

• 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:
  • Metformin: mean 8.64 versus 8.56 on control; difference 0.08 (95% CI -1.95 to 2.11), standardized effect d = 0.03; z-test p = 0.939.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Metformin and the control. The interval gives the range of effects that remain compatible with the data.

Arterial Stiffness by AI@75

• 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:
  • Metformin: mean 15.89 versus 14.96 on control; difference 0.93 (95% CI -6.57 to 8.43), standardized effect d = 0.08; z-test p = 0.808.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Metformin and the control. The interval gives the range of effects that remain compatible with the data.

Inflammatory Marker: hsCRP

• 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:
  • Metformin: mean 2.13 versus 2.95 on control; difference -0.82 (95% CI -2.54 to 0.90), standardized effect d = -0.32; z-test p = 0.351.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Metformin and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists secondary outcomes spanning arterial stiffness, mitochondrial measures, inflammation, heart-rate variability, and continuous glucose monitoring. The posted summaries include numerical comparisons for three secondary measures: arterial stiffness by PWV, arterial stiffness by AI@75, and hsCRP. For each of these simple unadjusted comparisons, the confidence interval includes zero. Together, those results do not establish a clear, consistent difference across the reported secondary measures.

The secondary outcomes are related to the study’s broader question about vascular, mitochondrial, and metabolic measures, while the primary outcome directly assesses insulin sensitivity. Their pattern does not change the interpretation of the primary comparison: the primary interval itself is wide and includes zero, and the registry does not report a crossover-specific analysis. Secondary outcomes can give context about whether a pattern appears across related measures, but they do not substitute for the primary analysis.

The registry lists 22 posted secondary outcomes, while numerical summaries are not reported for many of the named measures in the registry record. These include oxygen consumption, electron transport chain protein expression, heart-rate variability, and continuous glucose monitor measures. The lack of numerical results for these outcomes means their direction, size, and precision cannot be assessed from this record.

As a general statistical issue, evaluating many outcomes increases the opportunity for chance findings when each is considered separately. Unadjusted secondary comparisons do not account for that multiplicity, and no multiplicity strategy is reported in the registry record. For that reason, an isolated secondary result would carry less weight than a prespecified primary outcome supported by an appropriate analysis. Here, the reported secondary estimates are also simple calculations from group summaries, not analyses reported by the trial investigators.

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

• Arterial Stiffness by PWV
• Arterial Stiffness by AI@75
• Mitochondrial Measures: Oxygen Consumption
• Mitochondrial Measures: Protein Expression Levels of Electron Transport Chain Complexes
• Inflammatory Marker: hsCRP
• Heart Rate Variability
• Continuous Glucose Monitor Measures of Mean Glucose
• Continuous Glucose Monitor Measures of Hypoglycemia

9. Statistical Interpretation

10. Statistical Methods Explained

The posted primary summaries are means and standard deviations for a continuous outcome. A mean is the arithmetic average of the reported measurements. A standard deviation describes how dispersed measurements are around that average; it is not a confidence interval and does not by itself show how precisely the treatment contrast is estimated. The reported values are 3.27 mg/kg/min with a standard deviation of 1.51 for metformin, and 3.46 mg/kg/min with a standard deviation of 2.15 for placebo.

What does the mean difference represent?

The mean difference is expressed as metformin minus placebo. The simple unadjusted estimate is -0.19 mg/kg/min, so the posted metformin mean is lower than the placebo mean. This is a descriptive contrast between the reported group summaries. The direction of the estimate is not proof of a treatment effect, and the result should be interpreted together with its uncertainty and with the design limitations.

What is a confidence interval?

In general, a confidence interval is a range of values compatible with the data and the assumptions of the method used to calculate it. The 95% interval for this simple estimate extends from -1.452 to 1.072 mg/kg/min. Because that range includes zero, it does not distinguish the estimate from no difference under the calculation’s assumptions. It also extends to values on both sides of zero, so the posted summaries do not narrowly constrain the direction or size of the contrast. A confidence interval is not a guarantee that the true effect lies within its endpoints, and its interpretation depends on the analysis model being appropriate.

What does the p-value tell us?

The unadjusted p-value is 0.768. In general, a p-value describes how unusual data at least as discrepant as those observed would be under a specified null model and its assumptions. It is not the probability that the null hypothesis is true, nor does it measure clinical importance. Here, it is tied to the simple unadjusted comparison, not a reported crossover analysis. The registry record contains no trial-reported analysis to use in its place.

Why does the crossover design matter?

As a general statistical point, crossover data can contain within-participant comparisons. Measurements from the same person may be related, and accounting for that relationship can improve estimation when the design and assumptions support it. A simple comparison of intervention-group summaries does not use that pairing. It also does not account for possible period effects or carryover from one intervention period to the next. The treatment sequence, washout information, and trial analysis model are not reported in the registry record, so it is not possible to assess those features from the posted summaries.

How should variability be read?

The standard deviations describe the spread of the outcome values within each intervention group. Greater variability generally makes a group mean less precise for a given amount of information. But these group standard deviations do not capture the correlation between repeated measurements on the same participants, which is important in crossover designs. The simple re-calculation uses the posted summaries and is not a paired analysis. Individual data and a reported model would be needed to characterize within-person variability and its contribution to the treatment contrast.

General statistical interpretation also depends on how observations were selected and handled. The registry lists enrollment of 23, while the primary group summaries list 16 metformin and 17 placebo observations. The record does not explain the relation between those counts, whether all participants completed both periods, how missing data were handled, or which participants were included in the outcome analysis. These omissions prevent a complete assessment of the analysis population and the representativeness of the posted outcome summaries.

The registry states that there are no reported analyses. Thus the p-value, confidence interval, and mean-difference calculation are simple unadjusted re-calculations from the posted summary statistics, not results of a trial-reported statistical model. The registry does not report whether any adjusted, paired, or sensitivity analyses were performed. If a trial-reported analysis were available, it would be needed to understand the investigators’ treatment-effect estimate and its assumptions.

Finally, a statistical result is not a clinical threshold. The registry states that no strictly defined normal range applies to the clamp measure. Neither the mean difference nor its p-value alone indicates the importance of a change for patients. General interpretation would consider the endpoint’s meaning, the size and precision of the estimate, the quality of the design and analysis, and the completeness of outcome reporting. The record does not state a threshold for a clinically important difference.

11. Confidence Intervals and Precision

The primary unadjusted mean difference is -0.19 mg/kg/min, with a 95% confidence interval from -1.452 to 1.072. The interval is wide relative to the point estimate and crosses zero. This means the simple calculation does not estimate the treatment contrast precisely enough to rule out values on either side of zero. The p-value of 0.768 is consistent with the interval’s inclusion of zero, but it should not be interpreted as evidence of equivalence or as a measure of the probability that there is no effect.

Confidence intervals describe uncertainty under the assumptions of the calculation. The interval comes from the simple unadjusted comparison of posted group summaries. It does not reflect a model that accounts for within-person pairing in the crossover design. If the treatment periods are correlated within participants, a crossover-specific analysis could produce a different estimate of uncertainty. Such an analysis is not reported in the registry record.

The primary arm summaries list 16 observations for metformin and 17 for placebo, while enrollment is 23. The registry does not explain how the arm-level counts relate to enrolled participants or whether all participants contributed measurements in both periods. Without this information, the effective amount of paired information and the influence of missing data cannot be assessed. The confidence interval should therefore be read as precision for the simple re-calculation, not necessarily as the precision of the trial’s intended treatment comparison.

In general, greater precision comes from more informative data and an appropriate analysis; a narrower interval alone would not correct bias or design-model problems. The reported interval is informative about the limits of the posted summary-based comparison, but not a complete characterization of uncertainty under the crossover design.

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 events. They are illustrations based on the observed effect, rather than the trial’s original design assumptions; those original assumptions are not reported in the registry record. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. Each planning scenario targets the stated fraction of the observed effect.

For the primary outcome, planning for 100% of the observed effect gives 1,518 participants per arm for 80% power and 2,032 participants per arm for 90% power. Planning for 75% of the observed effect gives 2,698 participants per arm for 80% power and 3,611 participants per arm for 90% power. Planning for 50% of the observed effect gives 6,069 participants per arm for 80% power and 8,125 participants per arm for 90% power.

In general, power is the chance that a study will detect a specified effect under the planning assumptions if that effect is present. The 80% and 90% columns indicate the per-arm sample sizes that would be needed under those assumptions for the corresponding power targets. A 90% target requires more participants than an 80% target in these illustrations. Smaller effects need more participants to detect, which is why the per-arm figures rise as the target effect is reduced.

These are planning illustrations, not observed power. They use an effect estimated from the posted results and do not reproduce the original crossover design or its possible within-participant structure. They should not be read as a retrospective claim about the chance that this trial detected an effect.

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 = -0.190; pooled standard deviation σ = 1.868.
2. n = 2 × (1.96 + 0.842)² × 1.868² / -0.190² = 1,517.2, rounded up to 1,518 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 <- -0.19                      # observed difference in means
sigma <- 1.86790809333899                     # 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(1813929)
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 = 1,518 (80% power), 2,032 (90% power)
 75% of observed effect: n per arm = 2,698 (80% power), 3,611 (90% power)
 50% of observed effect: n per arm = 6,069 (80% power), 8,125 (90% power)
power.t.test: 1518.2 
Simulated power at n = 1,518 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 record lists 30 outcome measures and 22 posted secondary outcomes. It does not report a multiplicity adjustment, a hierarchy among secondary outcomes, or a plan for deciding which comparisons would be treated as confirmatory. As a general statistical matter, testing many outcomes can increase the chance that at least one result appears noteworthy by chance. Unadjusted secondary comparisons should therefore be treated as exploratory unless the trial’s analysis plan establishes otherwise. That plan is not reported in the registry record.

The record also reports no trial analyses. It does not describe the statistical analysis plan, the planned model for the crossover design, the analysis population, or any interim analyses. The simple unadjusted comparisons are not a substitute for those details. In particular, the registry does not report whether the primary comparison was intended to use within-participant pairing or how period and carryover effects would be considered.

Missing-data handling is not reported. The registry lists 23 participants enrolled, but the primary outcome summaries list 16 observations in the metformin group and 17 in the placebo group. The record does not explain whether those counts refer to participants, intervention-period observations, or another analysis unit; whether participants completed both periods; or why the counts differ from enrollment. It also does not state whether missing outcome values were imputed, excluded, or handled by another method.

The registry does not identify an intention-to-treat, per-protocol, or other analysis population. Nor does it report participant flow through the intervention periods, treatment adherence, or the number of participants with complete paired measurements. These omissions limit assessment of selection and missingness and prevent a full evaluation of how the reported summaries represent the randomized participants. They also make it difficult to judge whether the simple unadjusted calculation corresponds to the intended estimand.

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 5% reporting threshold.

The registry reports zero serious adverse events in the metformin group, among 16 participants at risk, and zero in the placebo group, among 17 participants at risk. It also reports zero deaths in both groups. The listed counts for other adverse events recorded as related are zero in each group, with 16 and 17 participants at risk, respectively.

These are counts in the registry record, not proof that no adverse events of any kind occurred. The record does not provide event-level descriptions, severity details for non-serious events, or a fuller account of how events were collected. The absence of reported serious events in these groups should be interpreted in the context of the reported group sizes and the limited safety detail, not as evidence that an intervention has no risks. No medical advice follows from these counts.

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

The posted primary results show similar group means, with a simple unadjusted mean difference of -0.19 mg/kg/min. The 95% confidence interval, from -1.452 to 1.072 mg/kg/min, includes zero and spans values in both directions. The result is therefore inconclusive about the direction and size of the treatment contrast under that simple calculation. It does not demonstrate equivalence, and it does not establish that the interventions had the same effect.

The comparison also needs to be understood in relation to the crossover design. A simple analysis of intervention-group summaries does not account for possible within-participant pairing, period effects, or carryover. Because the registry reports no trial analyses, the posted comparison cannot be treated as a definitive estimate from the study’s intended model. A crossover-specific analysis could be more appropriate, but its results and assumptions are not reported.

The endpoint measures glucose infusion during a controlled clamp and is described as a measure of insulin sensitivity. The registry notes that no strictly defined normal range applies. The record also does not report a threshold for a clinically important difference, so the numerical contrast cannot be translated into a judgment about clinical importance from these data alone.

The secondary outcomes provide limited additional context: the reported simple comparisons have intervals that include zero, while numerical results for many listed outcomes are not reported. The registry does not report a multiplicity plan, missing-data approach, or analysis population. Those gaps, together with the small enrolled sample and the unclear relation between enrollment and arm-level counts, restrict the strength of conclusions that can be drawn.

Overall, the registry record describes the outcome summaries and their uncertainty but does not support a firm conclusion about the treatment effect. It neither establishes benefit nor establishes absence of an effect. Further interpretation would require the trial’s own analysis and fuller information about participant-level crossover data and outcome reporting.

17. Why This Trial Matters Statistically

This trial illustrates why the design and the analysis need to be read together. A crossover study can make within-participant comparisons, but the posted primary result is a simple unadjusted comparison of intervention-group summaries. Without a reported model that accounts for repeated measurements, the calculated interval and p-value may not represent the most relevant analysis for the design.

It also demonstrates the difference between an estimate and its precision. The mean difference is close to zero, but the confidence interval includes values on both sides of zero. A point estimate alone can obscure that uncertainty. Conversely, a non-small p-value does not establish that the interventions are equivalent or that a meaningful effect is absent.

The registry record further shows why participant counts and analysis populations matter. Enrollment is 23, while the primary group summaries list 16 and 17 observations. The record does not explain how those counts relate to participants or paired periods. That missing context affects interpretation of precision and possible missing-data bias.

Finally, the outcome list is broader than the results with numerical summaries, and no multiplicity strategy is reported. This makes it important to distinguish the primary question from exploratory secondary comparisons and to avoid overinterpreting isolated findings. The record is a useful example of how results can be reported without enough analysis detail to resolve these statistical questions.

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 directly from what the registry record does and does not report. The sample is small, with 23 enrolled participants, and the primary group summaries list 16 observations for metformin and 17 for placebo. The relation between these counts and the crossover periods is not explained, so the amount of complete paired information cannot be determined.

The registry describes randomization, double masking, and crossover allocation, but it does not report the treatment sequence, washout period, or details needed to assess period and carryover effects. The simple unadjusted comparison does not account for within-participant pairing. No trial-reported analysis is available in the registry record to clarify the intended model or the results from one.

The dose and further treatment details are not reported in the posted results. Neither the original sample-size justification nor the statistical analysis plan is reported. The record also does not report how missing observations were handled, which analysis population was used, or whether all participants completed both intervention periods. The original rationale and eligibility information are not a substitute for these analysis details.

Reporting is incomplete across the outcome set: numerical summaries are available for the primary measure and three secondary measures, while many listed secondary outcomes do not have numerical results in the registry record. A multiplicity strategy is not reported. The adverse-event record gives group counts but limited detail. These limitations do not demonstrate that the trial was conducted improperly; they identify what cannot be assessed from the registry results alone.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Effect of Metformin on Vascular and Mitochondrial Function in Type 1 Diabetes — NCT01813929.
• Registry record last updated 2022-01-21; status last verified 2021-12.
• Results first posted 2022-01-21.
• 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

This completed phase 4 randomized, double-masked crossover study enrolled 23 participants with type 1 diabetes and compared metformin with placebo. The primary endpoint was insulin sensitivity measured by glucose infusion rate during a hyperinsulinemic euglycemic clamp at the end of each 6 week intervention period.

The posted primary means were 3.27 mg/kg/min for metformin and 3.46 mg/kg/min for placebo. A simple unadjusted comparison estimated a mean difference of -0.19 mg/kg/min, with a 95% confidence interval from -1.452 to 1.072 and an unadjusted p-value of 0.768. The registry reports no trial analyses, so this is not a crossover-specific result.

Numerical summaries are posted for three secondary outcomes, and the registry reports zero serious adverse events and zero deaths in both groups. The trial’s original design assumptions, statistical plan, analysis population, missing-data handling, and several crossover details are not reported in the registry record.