Metformin in Kidney Disease: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Metformin in Kidney Disease (NCT02252081), a randomized phase 2 study. The primary outcome is Change is Leptin to Adiponectin Ratio (LAR).

2. Clinical Question

3. Trial Design

This was a Phase 2, randomized, parallel-group trial comparing metformin with placebo. The registry lists an enrollment of 125 and a completed status. The recorded start date is 2014-10-01, and the primary completion date is 2019-12-31. The sponsor was the VA Office of Research and Development.

The registry records quadruple masking. That label indicates a high level of masking in the study design, but the record does not specify which roles were masked or describe how masking was implemented. The allocation method is listed as randomized; details of the sequence generation, concealment, stratification, or any blocks are not reported in the registry record.

The treatment period was 16 weeks. Metformin dosing depended on eGFR: the recorded regimen was 500 to 1500 mg orally per day for eGFR > 45 ml/min, and 500 to 1000 mg orally per day for eGFR <= 45 ml/min. The comparator was placebo pill(s) taken orally per day for the same period. The record does not state how many participants were assigned to each arm or whether all enrolled participants began treatment.

The stated primary outcome was change in LAR at 16 weeks after treatment began. Secondary outcomes listed in the record were change in flow-mediated dilation (FMD) and aortic pulse-wave velocity (aPWV). The record reports four outcome measures overall, but does not identify all four in the available outcome details. It does not report a protocol, the rationale for the enrollment target, or a detailed analysis plan.

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:

Analysis population for the primary outcome, as defined in the registry: analyzed as log transformed using ANCOVA of change, ng/µg (log transformed value)

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 is Leptin to Adiponectin Ratio (LAR), measured over 16 weeks after start of treatment.

The primary endpoint was change in the leptin-to-adiponectin ratio (LAR), measured in ng/micrograms, at 16 weeks after treatment began. The registry describes this change as a biomarker of insulin resistance in CKD. LAR combines information about two circulating hormones; the trial used the ratio as a biomarker rather than as a direct clinical event such as hospitalization or death.

The registry lists the endpoint parameter as a median with inter-quartile range. It also states that the primary analysis used log-transformed values and ANCOVA of change. These descriptions refer to different aspects of measurement and analysis: the median and inter-quartile range characterize the outcome distribution, while the stated model analyzes change on a transformed scale. The record does not provide the outcome values, the estimated treatment contrast, its uncertainty, or a p-value. It also does not explain how the reported median and inter-quartile range relate to the log-transformed ANCOVA analysis.

As a general statistical consideration, a biomarker can help characterize a biological process, but a change in a biomarker is not by itself evidence of a change in symptoms, disease progression, or clinical events. Interpretation depends on the magnitude and precision of the estimated difference, the analysis population, and the relationship between the biomarker and outcomes important to patients. Those primary-result details are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does randomization contribute?

As a general statistical principle, random assignment aims to make treatment groups comparable, on average, with respect to both measured and unmeasured characteristics at the point of assignment. This supports interpretation of between-group differences as effects of assignment, provided the analysis preserves the randomized comparison and other design features are handled appropriately. Randomization does not guarantee that the realized groups are identical, and the registry record does not report the allocation sequence or concealment procedures.

The trial is listed as parallel-group: participants were assigned to one of the treatment groups rather than crossing between treatments as part of the recorded model. The registry records quadruple masking, but does not specify exactly who was masked, how masking was maintained, or whether masking was assessed. Those details cannot be inferred from the masking label.

What does the endpoint description imply?

The primary outcome is change in LAR, with the record specifying an analysis on a log-transformed scale using ANCOVA. In general, a log transformation changes the scale on which values are modeled and can be useful when outcomes are skewed or when variation grows with the size of the measurement. Model estimates on a transformed scale do not automatically have the same interpretation as raw-scale differences. The record does not explain how any model estimate would be translated back to the original scale, and it gives no primary estimate to interpret.

ANCOVA is a general modeling approach that can estimate a group contrast while accounting for other variables included in the model. The registry names ANCOVA of change but does not identify covariates, model specifications, diagnostics, or whether the model included a baseline measurement. Those elements of the analysis plan are not reported in the registry record. No assumptions about them should be made.

The endpoint is described using a median and inter-quartile range. In general, the median represents the midpoint of an ordered distribution, while the inter-quartile range describes the span of its middle half. These summaries can be useful when a distribution is skewed or contains extreme values. The registry's mention of a median and inter-quartile range does not itself reveal the treatment contrast, and no values for either group are reported in the available results.

Why does uncertainty matter?

In general, an estimate describes the size and direction of a contrast, while a confidence interval communicates the range of values reasonably compatible with the data under the model and its assumptions. A narrow interval indicates greater precision than a wide interval, but precision is not the same as clinical importance. Neither a confidence interval nor another measure of uncertainty for the primary comparison is reported in the registry record.

A p-value, when reported, summarizes how compatible the observed data are with a specified null model under the analysis assumptions. It is not the probability that the null hypothesis is true, nor does it describe the size of an effect. The registry record does not report a primary p-value, so there is no basis for describing statistical significance of the primary endpoint.

How should planning calculations be understood?

The registry describes sample-size planning illustrations based on the observed effect, with a two-sided significance level of 0.05, equal allocation, and a normal approximation. It describes illustrations for detecting stated fractions of the observed effect. Such calculations are planning scenarios, not observed trial results and not evidence that the study had a particular probability of detecting an effect after the trial was completed. The numerical planning outputs are not reported in the registry record.

In general, planning calculations depend on assumptions about the target effect, variability, allocation, and analysis method. If those assumptions differ from the realized data or intended analysis, the calculations may not describe the trial's actual precision. The record does not report the original design justification or explain whether the planning illustrations correspond to the trial's initial design.

The record states that time-to-event planning calculations use a total event count, while other calculations use participants per arm. This distinction is general to the planning description and does not mean that the primary LAR endpoint was a time-to-event outcome. The primary endpoint is a measured change at 16 weeks.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary endpoint, secondary outcomes, and four outcome measures overall. It also indicates that the primary outcome has several categories or time points and was not reduced to one comparison. This creates an interpretive limitation: without a clearly identified primary contrast, it is difficult to tell which estimate was intended to carry the main confirmatory interpretation. The record does not report a multiplicity adjustment, a hierarchy of hypotheses, or rules for interpreting multiple endpoints and time points.

As a general statistical principle, testing multiple outcomes or time points can increase the chance of at least one apparently unusual result if each is interpreted without accounting for the set of tests. Methods such as prespecified outcome hierarchies or multiplicity adjustments can address that issue, but the registry record does not say whether such methods were planned or used. No claim about the trial's actual multiplicity handling can be made from the available information.

The registry describes the primary endpoint analysis as ANCOVA of change using log-transformed values, but it does not report the analysis population, exclusions, missing-data methods, or the number of participants contributing outcome data. The enrolled total is 125; the record does not state how many participants completed the endpoint assessment or were included in the primary analysis. Reasons for missing measurements, if any, are not reported in the registry record.

In general, missing outcomes can affect an estimate when the likelihood of missingness is related to participant characteristics, treatment, or the unobserved outcome. The direction and extent of any such effect depend on the pattern of missingness and the analysis assumptions. Without participant flow and missing-data details, the robustness of the primary analysis cannot be assessed.

11. 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 3% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: pneumonia & pneumonia related complications (6.7%), acute coronary syndrome (3.3%), pericarditis & chest pain (3.3%).

The registry lists 30 participants at risk for serious adverse events in each group. It records 2 participants affected in the metformin group and 2 in the placebo group. For other adverse events, it records 16 affected in the metformin group and 6 in the placebo group, with 30 at risk in each group. The registry also lists 18 deaths in the metformin group and 8 in the placebo group.

These entries should be treated as descriptive registry counts, not as a causal comparison. The available record does not explain the period covered by each count, whether deaths were considered treatment-related, or how the counts relate to participant follow-up. It also does not report an adjusted safety analysis or a comparative uncertainty measure. The number at risk is recorded, but the listed information is not enough to establish a complete account of exposure or follow-up.

The serious-event terms listed are pneumonia and pneumonia-related complications at 6.7%, acute coronary syndrome at 3.3%, and pericarditis and chest pain at 3.3%. The registry does not specify the treatment groups associated with these terms in the available information. The percentages should not be interpreted as arm-specific rates or compared with one another without that context. Event counts alone do not establish whether an event was caused by treatment.

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

The record documents a Phase 2 randomized trial comparing metformin with placebo over 16 weeks, with change in LAR as the stated primary endpoint. It gives the treatment regimens, the endpoint definition, and the intended log-transformed ANCOVA approach. These details describe the question and design; they do not supply the primary result.

A missing single primary comparison is consequential. Without an estimate, a measure of uncertainty, and a clear account of the analysis population, readers cannot tell whether the observed data favored one group, how large any contrast was, or how much uncertainty surrounded it. The endpoint description does not answer those questions. Nor can a result for a biomarker alone be treated as a result for clinical events or patient experience.

The safety entries add descriptive information about serious and other adverse events and deaths. However, the record does not give the context needed to attribute these events to the intervention or to judge comparative safety precisely. The serious-event terms and percentages are not linked to treatment arms in the available details. The counts should therefore remain descriptive rather than be used to infer causation.

In general, a trial record can clarify design and planned outcomes while still leaving important evidence questions unresolved when results or analysis details are absent. Here, the record supports describing the randomized comparison, its treatment duration, and the biomarker endpoint. It does not support a claim that metformin improves or worsens the endpoint, cardiovascular measures, kidney disease, or clinical outcomes. Those conclusions would require results and analyses not reported in the registry record.

13. Why This Trial Matters Statistically

This trial is statistically notable for pairing a randomized, masked treatment comparison with a biomarker endpoint in CKD. The primary outcome was change in LAR, which the registry describes as a biomarker of insulin resistance. FMD and aPWV are also listed as secondary outcomes, linking the recorded outcome set to vascular measurements. These features make clear why the estimand and analysis details matter: a measured biomarker change and a clinical outcome are not interchangeable forms of evidence.

The registry specifies log-transformed ANCOVA of change while also identifying the endpoint summaries as median and inter-quartile range. Both descriptions can be relevant, but they answer different statistical questions. To interpret a result well, readers need to know the scale of the estimate, the model's covariates, the relationship between the modeled contrast and the raw-scale outcome, and the uncertainty around that estimate. Those additional analysis details are not reported in the registry record.

The record also illustrates the importance of a clearly identifiable primary comparison when an outcome involves several categories or time points. Without one reported contrast, it is not possible to determine which result is central or how multiple comparisons were handled. The record gives no multiplicity strategy, missing-data approach, or analysis population, so statistical robustness cannot be assessed from the available information.

Finally, the safety counts demonstrate why denominators, follow-up context, event definitions, and attribution matter. Counts can be described accurately without treating them as evidence of causation. In this trial, the registry provides some group-level counts but not the context or comparative analysis needed for strong safety conclusions.

14. 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 most important limitation of the available record is that it does not provide a single primary-outcome comparison or its numerical result. The record explicitly indicates that the primary outcome has several categories or time points and was not reduced to one comparison. The endpoint's planned measure and analysis are described, but estimates, uncertainty, and a p-value are not reported. As a result, the primary treatment contrast cannot be interpreted quantitatively.

The record lists four outcome measures overall but only identifies the primary endpoint and two secondary outcomes in the available details. Results for those outcomes are not reported in the registry record. The analysis population, participant flow, group allocation counts, and endpoint completion counts are also not reported. Although enrollment is listed as 125, this does not establish how many participants contributed to each outcome analysis.

Important model details are absent. The record names ANCOVA of change on log-transformed values but does not specify covariates, model diagnostics, handling of missing outcomes, or the interpretation scale for an estimate. It does not report a multiplicity strategy despite noting multiple categories or time points for the primary outcome. The original sample-size rationale and any interim analyses are not reported in the registry record.

Safety information is limited to descriptive counts, listed serious-event terms and percentages, and death counts. The record does not provide event timing, causal attribution, exposure duration by participant, or a comparative safety analysis. It also does not associate the listed serious-event terms with a treatment arm in the available details. These limitations restrict conclusions about both efficacy and safety to descriptions of what the registry records.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Metformin in Kidney Disease — NCT02252081.
• Registry record last updated 2021-11-04; status last verified 2021-10.
• Results first posted 2021-11-04.
• 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.

18. Record Summary

NCT02252081, titled “Metformin in Kidney Disease,” is listed as a completed Phase 2 randomized, quadruple-masked, parallel-group trial sponsored by the VA Office of Research and Development. The registry records enrollment of 125, a start date of 2014-10-01, and a primary completion date of 2019-12-31.

The trial compared oral metformin with oral placebo over 16 weeks. Its primary endpoint was change in the leptin-to-adiponectin ratio, described as a biomarker of insulin resistance in CKD and analyzed using log-transformed ANCOVA of change. FMD and aPWV are listed as secondary outcomes.

The registry record does not report a single primary comparison or numerical primary result. It also does not report detailed analysis populations, missing-data methods, or a multiplicity approach. Adverse-event counts and selected serious-event terms are recorded, but the available information does not establish causation or provide a complete comparative safety analysis.