Training Effects on Skeletal Muscle Fatty Acid Metabolism: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Training Effects on Skeletal Muscle Fatty Acid Metabolism (NCT00786487), a randomized na study. The primary outcome is Insulin Sensitivity as Measured by Hyperinsulinemic Euglycemic Clamp at a Single Time Point (6 Hrs) After Intralipid or Glycerol Infusion.

2. Clinical Question

The trial’s stated scientific interest is how skeletal muscle processes fat and how this may relate to insulin resistance. Its summary describes the rationale that insulin resistance precedes and predicts type 2 diabetes, and that pharmaceutical-grade fat infusion can induce insulin resistance. These statements explain the investigators’ motivation; they are not results of the posted comparisons.

The reported endpoint is the M value from a hyperinsulinemic euglycemic clamp, measured at a single time point after infusion began. The registry presents separate comparisons of lipid with glycerol within the trained and untrained groups. This structure addresses whether the infusion groups had different endpoint summaries within each training category. It does not, by itself, answer whether training changes the effect of lipid relative to glycerol; a direct test of that interaction is not reported in the registry record.

The record labels the study randomized, but it does not report enough detail to determine what aspects of study participation were randomized, how allocation was generated, or whether training status was assigned. The research question should therefore be read as a comparison of the four recorded groups, not as proof that every difference between those groups was caused by the infusion or by training.

3. Trial Design

The registry describes a randomized, single-masked, parallel study sponsored by the University of Minnesota. Its status is completed. The record gives a start of 2009-01 and primary completion of 2014-01, and lists enrollment as 33. It lists the phase as NA. The identity of the masked party, the allocation procedure, and the timing or conduct of masking are not reported in the registry record.

The four recorded groups combine infusion type and training category: lipid infusion in trained participants, glycerol infusion in trained participants, lipid infusion in untrained participants, and glycerol infusion in untrained participants. The registry reports 10.0 participants in the lipid-trained group and 7.0 in the glycerol-trained group. It reports 9.0 in the lipid-untrained group and 6.0 in the glycerol-untrained group for the primary endpoint. These are group-specific counts; the record does not explain any differences between endpoint and safety counts.

The intervention descriptions identify a 20% lipid infusion and a glycerol infusion. The summary characterizes glycerol as a control infusion without the fat component. Further details about infusion dose, duration, administration, adherence, or co-interventions are not reported in the registry record. The record also does not explain whether participants’ trained or untrained status was determined before enrollment, how it was measured, or whether it changed during the study.

The primary endpoint was measured at 6 hours after starting the lipid or glycerol infusion. The registry describes a single-time-point measurement, rather than a reported change from baseline or a series of measurements. It gives group means and standard deviations and reports a two-sided t-test for each lipid-versus-glycerol comparison within training category. It does not report the allocation concealment method, a protocol for handling departures from assigned treatment, or the analysis population for these comparisons.

The record identifies the study condition as healthy volunteers and gives no secondary outcomes in the posted results. The intended research rationale discusses muscle energy function and fat metabolites, but the posted primary outcome is insulin sensitivity. Whether those other concepts were measured or analyzed is not reported in the registry record.

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: we had anticipated enrollment of 36 subjects but ended up finishing the study with 33 subjects

Overall Study

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Insulin Sensitivity as Measured by Hyperinsulinemic Euglycemic Clamp at a Single Time Point (6 Hrs) After Intralipid or Glycerol Infusion, measured over at 6 hours after starting lipid/glycerol infusion.

The primary endpoint is insulin sensitivity measured using the M value from a hyperinsulinemic euglycemic clamp. The registry defines the M value as glucose infusion rate per kilogram of fat-free mass per minute. It was assessed at 6 hours after starting the lipid or glycerol infusion, at a single time point.

In general, a euglycemic clamp uses controlled insulin and glucose administration to assess how much glucose must be infused to maintain the target glucose level under the test conditions. The M value expresses that glucose infusion rate relative to fat-free mass. In general, greater glucose infusion requirements under a clamp are interpreted as greater insulin-mediated glucose uptake, which is why the M value is used as a measure of insulin sensitivity. This is a description of the measurement concept, not a claim about clinical benefit.

A clamp is a physiological measurement rather than a direct measure of later disease or everyday functioning. The registry’s endpoint concerns insulin sensitivity at the stated post-infusion time point; it does not establish how long any measured difference might persist. A single time-point endpoint also cannot describe the trajectory of the response or distinguish an immediate effect from a lasting one.

The posted group summaries are means with standard deviations. A mean describes the group’s average M value, while a standard deviation describes the spread of individual measurements around that average. A standard deviation is not a confidence interval and does not, on its own, quantify the precision of a comparison. The registry does not provide confidence intervals for the reported comparisons.

The endpoint’s physiological focus fits the stated interest in insulin resistance and fat infusion. Its interpretation remains bounded by the specific measurement, timing, and healthy-volunteer population recorded. It should not be treated as a direct measurement of diabetes risk, muscle fat metabolites, or clinical outcomes, none of which is reported as the posted primary endpoint.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the reported t-test assess?

In general, a two-sided t-test compares group means while allowing for a difference in either direction. Its null hypothesis is that the population means being compared are equal. A two-sided test does not require the observed difference to follow the direction expected by the investigators. The registry reports this method separately for lipid versus glycerol among trained participants and for lipid versus glycerol among untrained participants.

A p value is calculated under a statistical model that assumes the null hypothesis and the test’s other assumptions. In general, it describes how incompatible the observed data, or data more extreme under the test statistic, are with that model. It is not the probability that the null hypothesis is true, and it does not measure the probability that the result arose “by chance” in an everyday sense. The registry reports p <0.05 for each comparison, rather than an exact p value.

Why does the test direction matter?

A two-sided test considers departures on both sides of equality. That is relevant when differences in either direction count as evidence against equal means. The record does not report a directional hypothesis or a prespecified one-sided test. The p-value category indicates that the registry’s reported test met its stated threshold, but the record does not give the test statistic or enough detail to reproduce the sponsor’s analysis.

What do means and standard deviations tell the reader?

For a continuous endpoint, the mean is the arithmetic average of the recorded measurements in a group. The standard deviation describes how dispersed those measurements are around the mean. In general, a larger standard deviation indicates more spread among individual observations, but it does not indicate the precision of the mean by itself. Precision depends on both variability and the amount of information available, as well as the analysis assumptions.

The standard deviation should not be read as a range containing every participant, nor as a confidence interval for the group mean or comparison. Two group means and standard deviations do not reveal the raw observations, their distribution, possible outliers, or whether the variability differs across groups. Those features can matter for evaluating a t-test, especially with small groups. The registry does not report diagnostic checks or a sensitivity analysis.

What would a confidence interval add?

In general, a confidence interval gives a range of effect estimates compatible with the data under a specified statistical procedure. It communicates the direction and scale of an estimated contrast as well as its uncertainty. A narrow interval indicates greater precision than a wide interval, subject to the method’s assumptions. The registry record does not report confidence intervals or numerical estimates of the between-group contrasts, so the precision of the comparisons cannot be judged from an interval.

A p-value threshold and a confidence interval answer related but different questions. A thresholded p value indicates whether the test crossed a specified decision boundary; it does not communicate the range of effect sizes that remain compatible with the data. The reported p <0.05 values therefore do not replace an estimate and interval, nor do they show whether any difference is large enough to matter outside the statistical test.

Why analyze the training categories separately?

The registry reports two within-category comparisons: lipid against glycerol among trained participants, and lipid against glycerol among untrained participants. This approach allows each infusion contrast to be examined within a recorded training category. It does not directly test whether the infusion contrast itself changes according to training category.

In general, a formal interaction test asks whether treatment contrasts vary across categories of another factor. The presence of two separate p values, including whether each crosses a threshold, does not answer that question. It is also not valid to infer an interaction merely because one estimate appears different from another or because significance results differ. The registry record does not report an interaction estimate or test.

The record identifies the study as randomized but does not say whether training status was randomized or how allocation related to the four groups. Randomization can support causal comparisons for the factor actually assigned, when implementation and analysis are appropriate. Without details about what was randomized, it is not possible to assume that differences between trained and untrained participants have the same causal interpretation as the within-category infusion comparisons.

What assumptions and design details remain uncertain?

In general, a t-test relies on assumptions about independent observations and the sampling distribution of the comparison. Depending on the version of the test, assumptions about equal variances may also apply. The registry names a t-test but does not identify a variance assumption, report assumption checks, or describe adjustments for covariates. The analysis cannot be reconstructed fully from the record.

Over-dispersion is a concern mainly for outcomes whose variability exceeds what a chosen model expects; for example, counts can vary more than a simple count model allows. The primary endpoint is continuous, and the registry reports means and standard deviations. A general lesson is that model assumptions should match the outcome and its observed variability. The record does not report whether a particular variance model was evaluated, so no claim about over-dispersion in these endpoint measurements is warranted.

Sample-size calculations are planning illustrations rather than observed power. In general, a planning calculation depends on an assumed effect, variability, significance level, allocation, and target power. The registry record states assumptions for illustrations based on the observed effect, including a two-sided significance level of 0.05, equal allocation, and a normal approximation. It does not report the trial’s original sample-size justification or establish that these illustrative assumptions were used to design the study. A calculation based on an observed effect should not be interpreted as the trial’s observed power.

Multiplicity is relevant because the registry reports more than one comparison for the primary endpoint. In general, performing several tests can increase the chance of at least one false-positive result if each test is judged at the same threshold. Whether and how to adjust depends on the prespecified question and analysis plan. The registry record does not report a multiplicity strategy or establish whether these comparisons were prespecified as a family. The reported p values should therefore be read as the reported individual tests, not as evidence that any familywise adjustment was made.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Lipid Trained", "Glycerol Trained", "Lipid Untrained", "Glycerol Untrained")
n   <- c(10, 7, 9, 6)
m   <- c(8, 12, 6, 11)
sp  <- c(3, 2, 2, 4)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Lipid Trained  10  8.000  3.000  5.854  10.146
CARM  Glycerol Trained  7  12.000  2.000  10.150  13.850
CARM  Lipid Untrained  9  6.000  2.000  4.463  7.537
CARM  Glycerol Untrained  6  11.000  4.000  6.802  15.198

11. Analyses Reported by the Sponsor

The registry reports the sponsor’s analyses as two-sided t-tests, with p <0.05 for the lipid-versus-glycerol comparison among trained participants and p <0.05 for the corresponding comparison among untrained participants. It does not report numerical contrast estimates, confidence intervals, or test statistics. The posted group means and standard deviations describe the endpoint in each group but do not supply those missing comparison details.

A recalculation of a continuous two-group contrast from summary statistics would be an unadjusted calculation and could depend on assumptions not fully documented in the registry. In this record, the comparison-level values needed for a recalculated estimate are missing or not estimable. No recalculated contrast or p value is therefore available to compare with the sponsor’s analysis.

The registry does not flag a discrepancy between sponsor and recalculated results. Because no comparison-level recalculation is available, the sponsor’s reported tests remain the only posted inferential results. Their interpretation is limited by the absence of effect estimates, confidence intervals, and details about the test implementation. The reported p-value categories should not be treated as substitutes for those quantities.

12. Multiplicity, Missing Data and Analysis Populations

The registry reports two statistical tests for the primary endpoint, comparing lipid with glycerol separately among trained and untrained participants. It does not report whether these tests were prespecified as co-primary comparisons, whether one was primary and the other supportive, or whether a multiplicity adjustment was planned or applied. The reported p values should therefore be understood as the individual results recorded for those tests. Whether they control an overall error rate across the comparisons is not reported in the registry record.

In general, when several hypotheses are tested, the chance of at least one false-positive result can rise if each test is interpreted independently at the same threshold. An adjustment can address that issue, but its appropriateness depends on the study questions and analysis plan. The record does not identify a family of hypotheses or a decision rule, so it is not possible to determine whether adjustment was needed or performed. It also does not report an interaction test comparing the infusion contrasts across training categories.

The registry gives group counts for the primary endpoint and separate counts at risk for adverse-event reporting. It does not explain how missing endpoint measurements were handled, whether any participants were excluded from the efficacy analysis, or whether the analyses followed an intention-to-treat or per-protocol principle. The number and reasons for missing endpoint observations are not reported in the registry record. No imputation procedure or sensitivity analysis is described.

The record does not report the analysis population definition, the timing of any exclusions, or how participants who did not complete the infusion or assessment were treated. It also does not provide a participant flow account sufficient to reconcile all group counts. These omissions limit assessment of whether the reported group summaries represent everyone assigned to a group or only those with an observed endpoint.

General statistical interpretation distinguishes a prespecified analysis from an exploratory one because decisions made after seeing data can affect the meaning of nominal p values. The registry does not report the protocol or statistical analysis plan details needed to establish prespecification. That absence does not show that decisions were made after seeing results; it means the registry record does not resolve the issue.

13. 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.

Most frequent non-serious adverse event terms

The adverse-event record lists zero serious adverse events for each of the four groups. It provides no serious adverse-event terms. These entries describe what is recorded in the registry; they do not establish that no unrecorded event occurred. Death counts are not reported in the record, so a missing death value should not be interpreted as a count of zero.

For other adverse events, the registry lists 1 affected participant in the lipid-trained group and 1 in the glycerol-trained group, with 0 in each untrained group. The corresponding numbers at risk are listed as 10 for lipid-trained, 7 for glycerol-trained, 9 for lipid-untrained, and 6 for glycerol-untrained in the other-event entries. The record gives no event descriptions or assessment of relatedness, severity, or timing.

The serious-event entries list 10, 7, 9, and 7 participants at risk across the same group order, while the primary endpoint counts are 10.0, 7.0, 9.0, and 6.0. The registry does not explain the difference in the final group’s at-risk count between these entries. It should not be resolved by assuming a correction or a particular participant flow.

In general, safety counts are descriptive and are affected by the period of observation, event definitions, and completeness of reporting. Those details are not reported in the registry record. The small group counts and absence of event descriptions limit comparisons of safety between infusions; the entries do not support broader conclusions about risk.

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

The posted results show the group means and standard deviations for insulin sensitivity measured by the M value at 6 hours after infusion began. In both training categories, the recorded glycerol group mean is higher than the corresponding lipid group mean. The registry reports p <0.05 from a two-sided t-test for each within-category comparison. These are descriptions of the recorded endpoint and reported analyses.

They do not establish the size or precision of either contrast because the registry does not report a numerical contrast estimate or confidence interval. The p-value categories indicate that the reported tests crossed the stated threshold, but they do not quantify clinical importance, establish the probability that a hypothesis is true, or explain the reason for the observed group pattern.

The results do not show whether the infusion contrast differs between trained and untrained participants. That requires a direct comparison of contrasts, such as a reported interaction analysis, which is not included in the registry record. The record also does not establish whether training status was randomized, so the summaries cannot be used to infer a causal effect of training.

The endpoint is a physiological measure assessed at a single time point. It does not directly report long-term insulin sensitivity, diabetes development, muscle energy function, or muscle fat metabolites. Although the study summary discusses those topics as motivation, the registry posts no secondary outcomes. The endpoint therefore cannot be used to claim effects on outcomes that were not reported.

The registered population is healthy volunteers. The results should not be assumed to describe people with insulin resistance or diabetes, or any other population not characterized in the record. Further participant characteristics and eligibility details are not reported in the registry record.

In general, statistical evidence, precision, clinical relevance, and causal interpretation are separate considerations. Here, the registry provides thresholded test results and descriptive group summaries, but limited information on precision, analysis choices, and participant flow. A careful reading keeps the conclusion close to the measured endpoint, the recorded groups, and the reported tests.

15. Why This Trial Matters Statistically

This trial is statistically interesting because it records a continuous physiological endpoint in a design that combines infusion type with training category. The group summaries make it possible to see the observed M values within each recorded group, while the separate two-sided t-tests address lipid-versus-glycerol comparisons within each training category. Keeping those questions distinct is important: two within-category tests do not automatically answer whether the infusion contrast varies by training status.

The record also illustrates the difference between a test result and an effect estimate. A thresholded p value conveys evidence relative to a statistical test’s null model, while an estimate with a confidence interval would communicate the contrast and its precision. The registry reports p <0.05 for both comparisons but does not report numerical contrasts or intervals. Readers therefore have less information about the range of effects compatible with the data than would be available from complete comparison estimates.

Small groups make transparent reporting especially useful. Group means and standard deviations are helpful summaries, but they cannot reveal individual observations or test assumptions. Reporting the analysis population, missing-data handling, allocation process, and confidence intervals would help readers evaluate robustness and uncertainty. Those details are not reported in the registry record.

The study also highlights the importance of distinguishing a measured biomarker from broader research motivation. The M value provides a defined physiological endpoint at a specified time after infusion. It does not by itself answer questions about disease outcomes or other mechanisms named in the rationale. Statistically, a conclusion should track the endpoint and contrast actually analyzed rather than extend to related but unreported outcomes.

These are general lessons in reading a small clinical experiment: identify the comparison, separate descriptive patterns from inferential tests, examine precision and assumptions, and avoid treating subgroup results as an interaction test. The registry’s limited reporting makes those distinctions particularly important for this study.

16. Limitations

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

The registry does not report the randomization sequence, allocation concealment, or which party was masked. Although the trial is labeled randomized and single-masked, the record does not identify the masked party or explain how randomization was implemented. The record also does not say whether training status was assigned, measured before the infusion, or defined using a stated criterion.

The primary endpoint is reported at a single time point, 6 hours after infusion began. No baseline M value or repeated measurement is included in the posted outcome description. The results therefore do not describe within-participant change over time or the persistence of the measured response. Details about infusion dose beyond the 20% lipid description, infusion duration, adherence, and other treatment procedures are not reported in the registry record.

For the statistical comparisons, the record names a two-sided t-test but does not provide the test statistic, exact p value, contrast estimate, confidence interval, variance assumption, covariate adjustment, or diagnostic checks. It also does not report an interaction test. The posted summaries cannot establish whether the analysis assumptions were assessed or whether alternative analyses would yield similar conclusions.

The registry does not explain missing-data handling, exclusions, or the analysis population. The group counts in the adverse-event entries do not fully match the primary endpoint counts, and the reason is not reported. There is no participant flow account in the posted record sufficient to explain this difference. The record does not provide details of outcome assessment completeness.

The two reported tests raise a multiplicity question, but no adjustment or prespecified testing hierarchy is described. The registry also does not report the original sample-size justification. The stated planning assumptions concern illustrations based on an observed effect, not evidence of the trial’s design rationale or observed power.

Finally, the registry identifies healthy volunteers but gives no further participant characteristics. That limits assessment of applicability beyond the recorded population. The study summary discusses additional physiological interests, but no secondary outcomes are posted, so those topics cannot be evaluated from the reported results.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Training Effects on Skeletal Muscle Fatty Acid Metabolism — NCT00786487.
• Registry record last updated 2015-07-23; status last verified 2015-07.
• Results first posted 2015-07-23.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

20. Record Summary

NCT00786487, titled “Training Effects on Skeletal Muscle Fatty Acid Metabolism,” is a completed randomized, single-masked, parallel study sponsored by the University of Minnesota. The registry lists healthy volunteers, enrollment of 33, a start of 2009-01, and primary completion of 2014-01. Phase is listed as NA.

The study compares 20% lipid infusion with glycerol infusion in trained and untrained groups. Its primary endpoint is the M value from a hyperinsulinemic euglycemic clamp at 6 hours after infusion began. The registry reports group means and standard deviations and two-sided t-tests for the infusion comparisons within each training category, with p <0.05 for both.

Numerical contrast estimates, confidence intervals, an interaction test, multiplicity procedures, missing-data methods, and details of the analysis population are not reported in the registry record. No secondary outcomes are posted. The results describe the recorded endpoint and comparisons but do not resolve the broader mechanistic questions in the study rationale.