Fibroblast Growth Factors 19 and 21 in Gestational Diabetes Mellitus: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Fibroblast Growth Factors 19 and 21 in Gestational Diabetes Mellitus (NCT01858597), a non-randomized na study. The primary outcome is The Concentration of Serum FGF19.

2. Clinical Question

The registry summary presents FGF19 and FGF21 as endocrine regulators involved in glucose, lipid and energy metabolism, and describes a hypothesis that their signaling may be relevant to insulin resistance in gestational diabetes mellitus. That is the biological rationale, not a reported finding from this study. The stated primary endpoint focuses on serum FGF19; it should not be treated as a direct measurement of insulin resistance or as evidence that a signaling pathway caused any observed difference. Although FGF21 appears in the study title and rationale, its specific outcome measure is not described in the listed primary outcome. The record also lists expression of FGF19 in term placenta as a secondary outcome. How either outcome relates to a prespecified statistical comparison is not reported in the registry record.

3. Trial Design

The registry identifies NCT01858597 as a completed study sponsored by First Affiliated Hospital, Sun Yat-Sen University. The recorded start is 2013-03 and primary completion is 2015-03. Enrollment is listed as 90. The phase is recorded as NA. These administrative details do not establish how participants were recruited, how many contributed data to each outcome, or whether the enrollment figure represents the analysis population.

The design fields describe the study as non-randomized, single-masked and crossover. The group descriptions, however, identify women with gestational diabetes mellitus and healthy pregnant women. The registry record does not explain who was masked, what information was masked, or how masking was maintained. It also does not describe a sequence of interventions or explain how the crossover label applies to these groups. Those design details should therefore be treated as unresolved rather than inferred from the labels.

The intervention is listed as biochemical detection, a diagnostic test, rather than as a treatment assignment. The two named groups are not randomized arms, and the registry explicitly cautions that they are not a like-for-like control comparison. Differences between groups can therefore reflect other characteristics as well as gestational diabetes mellitus. The record does not state whether confounders were measured or adjusted for.

For the primary outcome, the registry specifies a mean serum FGF19 concentration in pg/mL, measured in the fasting state during 24-28 gestational weeks, with full range as the dispersion. It lists expression of FGF19 in term placenta as a secondary outcome. The record does not describe the timing or analysis population for that secondary outcome, the laboratory procedures, or the statistical model. It also does not report an analysis plan or numerical outcome results. Completion status alone does not fill in these missing details.

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

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 The Concentration of Serum FGF19, measured over in fasting state during 24-28 gestational weeks.

The primary endpoint is the concentration of serum FGF19, expressed in pg/mL, in a fasting state during 24-28 gestational weeks. The registry describes the endpoint as a mean and specifies full range as its dispersion. A mean gives the arithmetic average of the measured values; a range describes the lowest and highest observed values. The range can help convey how widely observations extend, but it does not by itself show the shape of the distribution or the precision of the mean.

FGF19 is the focus of the primary measure because the study rationale concerns endocrine signaling and its possible relationship with insulin resistance in gestational diabetes mellitus. A circulating concentration is a biomarker measurement, not a direct clinical outcome and not, by itself, a measure of insulin resistance. The registry record does not report the assay, laboratory quality procedures, or a threshold defining an abnormal concentration. It also does not provide the observed mean or range for either group, a group contrast, or an uncertainty interval.

The secondary outcome is expression of FGF19 in term placenta. Its numerical result and detailed measurement definition are not reported in the registry record. The listed outcomes therefore describe what was intended to be measured, but do not establish what the study observed or how precisely any group difference was estimated.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies a mean as the primary outcome summary and full range as its dispersion. These are descriptive statistics. The mean condenses a set of measurements into an average, while the range records the extremes. The range can be strongly affected by extreme observations and does not show whether most values cluster near the mean or are spread across the interval. General statistical interpretation is stronger when a center and a measure of uncertainty are both available. The registry record does not report a confidence interval or another uncertainty estimate for the primary outcome.

A confidence interval is, in general, a range of estimates constructed by a statistical procedure intended to have a stated coverage rate over repeated use of that procedure. It conveys uncertainty about an estimated quantity; it is not a range containing a guaranteed proportion of individual measurements, and it does not establish that a result is clinically important. No confidence interval is reported for serum FGF19 in the registry record. A test result or a confidence interval cannot be reconstructed from the endpoint description alone.

The group comparison is non-randomized. In general, randomization helps prevent systematic differences in baseline characteristics from driving the contrast between groups. Without random assignment, measured or unmeasured characteristics may differ between women with gestational diabetes mellitus and healthy pregnant controls. Statistical adjustment can sometimes address measured differences, but it cannot automatically remove bias from unmeasured factors or poorly measured covariates. The registry record does not report covariates, an adjusted model, or an adjustment rationale. No adjusted or causal interpretation should be attributed to the study on the basis of the design labels alone.

The design also carries a crossover label. In general, a crossover study has participants receive different conditions in an ordered sequence, so within-person comparisons may be possible; interpretation then depends on details such as sequence, timing and possible carryover. The registry record does not describe such sequences or conditions, and it does not explain how that label applies to the disease and healthy-control groups. It is therefore not possible to determine whether a within-participant comparison was performed or whether crossover-specific methods were relevant.

The record specifies single masking but does not say who was masked or what was concealed. Masking can, in general, reduce the risk that knowledge of group assignment influences measurement or assessment. Its relevance depends on what was masked and how the outcome was collected. A serum concentration is a laboratory measure, but the record does not specify whether laboratory personnel were masked, how specimens were coded, or whether the measurement procedure could be influenced by group knowledge. The masking label alone does not resolve these questions.

The registry's planning assumptions describe an illustration based on an observed effect, a two-sided significance level of 0.05, equal allocation between two arms and a normal approximation. The stated effect to detect is a fraction of the observed effect. In general, a planning calculation translates assumptions about an effect, variability, error rates and design into a target sample size. It is not a report of what the study actually found. A planning illustration based on an observed effect also depends on that effect estimate, which can be unstable, particularly when it comes from a small or imprecise study. No observed effect value or numerical planning result is reported in the registry record. The illustration must not be read as observed power.

For time-to-event calculations, the record says the planning tables express a total number of events; other planning calculations express participants per arm. This is a distinction in the form of a design calculation, not evidence that this biomarker study had a time-to-event endpoint. The listed primary endpoint is a serum concentration, not a time-to-event measure. The record does not report a time-to-event analysis for the primary outcome.

Over-dispersion is a general concern when observed variability exceeds what a statistical model assumes. If a model understates variability, its uncertainty estimates can be too narrow and its tests can appear more decisive than warranted. The registry record does not identify a probability model for the primary endpoint or report a dispersion assessment, so over-dispersion cannot be evaluated from the recorded methods. The general point is that a model's assumptions should be matched to the observed data, especially when measurements are skewed or include extreme values. The posted description of a mean and full range is not enough to determine the distribution or select an appropriate inferential method.

Finally, the record lists no analysis entries and does not report a statistical analysis plan, test, model, confidence interval, missing-data procedure or analysis population. These omissions prevent a reader from assessing how the primary endpoint was compared, whether adjustments were prespecified, or how incomplete measurements were handled. The registry's outcome definitions can be described, but the inferential method and its assumptions cannot be filled in by general statistical practice.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists the concentration of serum FGF19 as the primary outcome and expression of FGF19 in term placenta as a secondary outcome. It reports two outcome measures, but does not describe a multiplicity procedure. When a study examines multiple outcomes or multiple comparisons, the chance of at least one apparently notable result can increase. General approaches include adjusting the testing threshold or specifying a hierarchy in advance, but the registry record does not state whether either approach was used. No test results are listed, so it is not possible to assess how multiplicity affected the interpretation.

Missing-data handling is not reported in the registry record. The record does not state how many participants had serum measurements available, whether samples were missing, why any data might be absent, or whether an imputation or complete-case approach was used. A complete-case analysis uses only records with the measurements needed for the analysis; depending on why data are missing, this can yield a biased estimate. Imputation methods also rely on assumptions that should be described. Without information on missingness and the analysis procedure, the effect of incomplete data cannot be evaluated.

The analysis populations are not defined. Enrollment is listed as 90, and the adverse-event entries separately list 30 participants at risk in the gestational diabetes mellitus group and 60 in the control group. Those entries do not establish how many contributed to the primary endpoint or secondary outcome, and they should not be treated as outcome analysis denominators. The registry does not report whether the analysis included all enrolled participants, only those with evaluable measurements, or another defined population.

No analysis entries are listed, and no statistical plan is reported. It is therefore not possible to determine whether the primary and secondary outcomes were analyzed as planned, whether subgroup analyses were conducted, or how any exclusions were handled. These are reporting gaps, not evidence that a particular method was or was not used.

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 0% reporting threshold.

The registry's adverse-event entries list the gestational diabetes mellitus group with 30 at risk and the control group with 60 at risk. For each group, serious events are recorded as 0, other events as 0, and deaths as 0. These figures describe what the registry record reports for its safety categories; they do not provide a detailed account of how events were defined, collected or monitored.

Zero recorded events does not establish that no adverse events occurred outside the scope or reporting procedures represented in the registry. The record does not describe event ascertainment, the observation period for safety, or whether all participants were assessed in the same way. The at-risk counts are safety denominators as recorded, not necessarily the number with a completed serum FGF19 measurement or the analysis population for an outcome.

The intervention is listed as biochemical detection, not as a therapeutic treatment assignment. The safety entries therefore should not be read as evidence about the safety of a treatment. They also do not establish comparative safety between the groups, which were not randomized. The figures are limited registry counts and do not alter the fact that numerical results for the primary biomarker outcome are not reported.

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

The study record describes an investigation of serum FGF19 in the context of gestational diabetes mellitus, with a group of healthy pregnant controls. It identifies a fasting measurement during 24-28 gestational weeks as the primary endpoint and names placental FGF19 expression as a secondary outcome. These details establish the intended focus of the study, not the direction or size of its findings.

The registry does not report numerical values for either outcome, a between-group estimate, a confidence interval or a statistical test. The data shown in the registry record therefore do not support a conclusion about whether serum FGF19 concentrations differed, how large a difference might have been, or how precisely it could be estimated. Missing results should not be interpreted as a null finding. They also cannot be replaced by the biological rationale, enrollment description or completion status.

Even if a group difference were available, the non-randomized design would limit causal interpretation. In general, a difference between women with gestational diabetes mellitus and healthy pregnant women could reflect the condition, other differences between the groups, or a combination. The registry does not report the covariates or analysis methods needed to evaluate those possibilities. The crossover label is not explained in relation to the group structure, so it cannot be used to infer a within-participant contrast.

FGF19 is a biomarker in this study, not a clinical endpoint that directly establishes a change in health. A measured association would not, by itself, demonstrate that FGF19 signaling causes insulin resistance or that changing FGF19 would change pregnancy outcomes. The record's description of metabolic pathways is background and hypothesis, not an outcome result. The listed safety counts likewise do not establish a treatment effect or answer the biomarker question.

The appropriate reading is therefore narrow: the record documents a completed, non-randomized study with a specified primary measure and a secondary placental measure, but it does not report the outcome estimates or inferential analyses needed to judge the observed relationship. Further claims about magnitude, precision, causal mechanism or clinical significance are not supported by the registry record.

13. Why This Trial Matters Statistically

This study raises a useful statistical distinction between a biological hypothesis and an estimable result. The rationale links FGF19 and FGF21 signaling with metabolic regulation, while the primary endpoint is a measured serum concentration. A statistical analysis would need to connect that biomarker measurement to a clearly defined group contrast and communicate uncertainty around the estimate. The registry identifies the endpoint but does not report the numerical estimate or its uncertainty, leaving the central comparison unanswered in the posted record.

The non-randomized comparison also illustrates why design labels matter. Comparing women with gestational diabetes mellitus with healthy pregnant women can describe an association, but group membership is not randomly assigned. General causal reasoning therefore requires attention to other characteristics that may differ between groups. The registry does not report an adjustment strategy, so readers cannot assess whether the analysis addressed potential confounding.

Another statistical issue is interpretability of the design. The record lists both a crossover model and disease-based groups, without explaining how a crossover structure was implemented. Such labels have methodological implications: a crossover design may call for within-person comparisons and attention to sequence, whereas a comparison of distinct disease and control groups usually poses a different contrast. The missing explanation prevents a clear assessment of which design assumptions apply.

Finally, the registry distinguishes outcome definitions from outcome reporting. A mean and full range are named, but no actual values or analysis are listed. The safety counts are reported, but they do not substitute for primary outcome estimates. This record is a reminder that a defined endpoint, completed status and planning assumptions cannot replace transparent reporting of observed data, analysis populations and uncertainty.

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 main limitation for interpreting this particular record is that numerical results for the primary and secondary outcomes are not reported. Without group summaries, a contrast estimate or an uncertainty measure, the size, direction and precision of any association cannot be assessed. The record also lists no analysis entries, so the statistical procedure used for the outcomes is not described.

The comparison is non-randomized, and the registry explicitly cautions that the groups are not a like-for-like control comparison. The record does not report participant characteristics, potential confounders or adjusted analyses. This leaves open whether group differences in a measured outcome could be associated with factors beyond gestational diabetes mellitus.

The design information is incomplete. Single masking is recorded without identifying who was masked, and crossover is recorded without a description of sequences or how it relates to the disease and control groups. The intervention is labeled biochemical detection, but laboratory assay details and measurement quality procedures are not reported. The primary outcome specifies a mean and full range, but the record does not report distributional information or an inferential method.

Analysis populations and missing-data procedures are also not reported. Enrollment and safety at-risk counts do not establish which participants contributed to each outcome. The record does not describe exclusions, incomplete measurements, multiplicity handling, subgroup analyses or a statistical analysis plan. These limitations concern what can be evaluated from the registry record; they do not demonstrate that an unreported procedure was omitted from the conduct of the study.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Fibroblast Growth Factors 19 and 21 in Gestational Diabetes Mellitus — NCT01858597.
• Registry record last updated 2023-05-30; status last verified 2022-07.
• Results first posted 2023-05-30.
• 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

NCT01858597 is recorded as a completed, non-randomized study sponsored by First Affiliated Hospital, Sun Yat-Sen University. Its listed groups are women with gestational diabetes mellitus and healthy pregnant women. The registry identifies serum FGF19 concentration in the fasting state during 24-28 gestational weeks as the primary outcome, reported as a mean with full range, and expression of FGF19 in term placenta as a secondary outcome.

Enrollment is listed as 90, and the recorded study dates are 2013-03 to 2015-03 for start and primary completion. The design is labeled single-masked and crossover, but those features are not explained in relation to the group descriptions. Numerical outcome results, analysis methods, missing-data procedures and analysis populations are not reported in the registry record. The adverse-event entries record zero serious events, zero other events and zero deaths in both groups.