This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Nattokinase Atherothrombotic Prevention Study (NCT02080520), a randomized na study. The primary outcome is Progression of Subclinical Atherosclerosis.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT02080520 |
| Phase | NA |
| Condition | Prevention of Subclinical Atherosclerosis Progression, Prevention of Cognitive Decline |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Quadruple |
| Enrollment | 265 |
| Arms in the primary outcome | 2 |
| Lead sponsor | University of Southern California |
| Results posted | Yes |
2. Clinical Question
Population
The study concerned prevention in healthy, non-demented women and men. The registry description frames the research around people without established dementia and the progression of subclinical atherosclerosis. Further eligibility details are not reported in the registry record.
Intervention
Participants assigned to the nattokinase arm received oral nattokinase at 2,000 fibrinolytic units daily. Nattokinase is classified in the registry as a dietary supplement.
Comparator
The comparison group received a matched placebo daily. The registry identifies the study as randomized, quadruple masked, and parallel group.
Primary question
The primary endpoint assessed the rate of change in distal common carotid artery far-wall intima-media thickness, measured in mm per year. The registry describes this as a co-primary trial endpoint and identifies cognitive change as an additional outcome. The central question was whether assignment to nattokinase, compared with placebo, was associated with a different rate of subclinical atherosclerosis progression. The posted result reports an arm-specific mean for each group and a mixed-model comparison with p = 0.31. This is a comparison of measured outcomes, not by itself a demonstration that either assignment changes cardiovascular or cognitive health.
3. Trial Design
The Nattokinase Atherothrombotic Prevention Study is a randomized, parallel-group trial comparing nattokinase with a matched placebo. The registry lists quadruple masking, meaning that the design is recorded as masked across four roles or groups; which specific roles were masked is not reported in the registry record. The study is marked completed. Its listed start date is 2014-04, and its primary completion date is 2019-07-31.
The intervention was oral nattokinase at 2,000 fibrinolytic units daily, while the comparator was a matched placebo taken daily. The registry lists enrollment of 265 participants: 132 in the nattokinase arm and 133 in the placebo arm for the primary outcome. Those are the posted outcome group sizes and should not be confused with the proposed enrollment described in the brief summary, which refers to randomizing 240 people. The registry does not explain the difference between the proposed and recorded enrollment.
The primary endpoint is the rate of change in distal common carotid artery far-wall intima-media thickness, expressed in mm per year. Its measurement schedule is described as baseline x 2 and then every 6 months, up to 3 years. The outcome description calls this a co-primary endpoint. The registry also lists change in global cognition as a secondary outcome and reports four outcome measures overall. Details of the other outcome measures are not reported in the registry record.
The trial's phase is listed as NA. A detailed protocol, recruitment criteria, allocation procedure, and planned statistical analysis beyond the reported mixed-model method are 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: 1189 individuals were screened by telephone, 857 did not meet inclusion criteria. 332 individuals were screened in research clinic, 67 were excluded (51 did not meet inclusion criteria; 16 declined to participate). 265 individuals were randomized.
Overall Study
| Stage | Nattokinase | Placebo |
|---|---|---|
| STARTED | 132 | 133 |
| COMPLETED | 118 | 116 |
| NOT COMPLETED | 14 | 17 |
Reasons for not completing this period:
| Reason | Nattokinase | Placebo |
|---|---|---|
| Self-reported symptoms | 5 | 6 |
| Unable to contact | 2 | 1 |
| Initiated aspirin | 2 | 0 |
| Atrial fibrillation | 1 | 1 |
| Cancer | 1 | 2 |
| Not interested in study | 1 | 3 |
| Too busy | 1 | 1 |
| Work-related injury | 1 | 0 |
| Personal reasons | 0 | 3 |
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.
| Characteristic | Unit | Nattokinase | Placebo | Total |
|---|---|---|---|---|
| Participants analysed | 132 | 133 | 265 | |
| Age, Customized | years | 65.4 (60.8 to 72.3) | 65.2 (60.6 to 71.7) | 65.3 (60.7 to 71.9) |
| Sex: Female, Male: Female | Participants | 83 | 81 | 164 |
| Sex: Female, Male: Male | Participants | 49 | 52 | 101 |
| Race/Ethnicity, Customized: White, non-Hispanic | Participants | 94 | 93 | 187 |
| Race/Ethnicity, Customized: Black, non-Hispanic | Participants | 11 | 4 | 15 |
| Race/Ethnicity, Customized: Asian, non-Hispanic | Participants | 10 | 15 | 25 |
| Race/Ethnicity, Customized: Hispanic | Participants | 16 | 19 | 35 |
| Race/Ethnicity, Customized: Other | Participants | 1 | 2 | 3 |
| Region of Enrollment: United States | participants | 132 | 133 | 265 |
6. Primary Endpoint
The registry lists the primary outcome as Progression of Subclinical Atherosclerosis, measured over Baseline x 2 and then every 6 months, up to 3 years.
| Element | Registry entry |
|---|---|
| Result type | Mean |
| Measure of spread | 95% Confidence Interval |
| Unit | mm per year |
| Endpoint classification | Continuous Endpoint |
The primary endpoint is the rate of change in the far wall of the distal common carotid artery, measured as intima-media thickness in mm per year. Intima-media thickness is an ultrasound measure of the arterial wall. In general, tracking change over time can characterize progression of a subclinical vascular measure before a clinical event is the endpoint. The registry describes computer image processed B-mode ultrasonograms and specifies repeated assessment from baseline, then every 6 months, up to 3 years.
The endpoint is reported as a mean for each arm with a 95% confidence interval. The nattokinase arm's mean rate is 0.013 mm per year, with a 95% confidence interval from 0.01 to 0.015. The placebo arm's mean rate is 0.011 mm per year, with a 95% confidence interval from 0.009 to 0.013. These are arm-specific summaries; they are not a confidence interval for the between-arm contrast. The sponsor's mixed-model comparison is reported separately.
The registry calls the measure a co-primary endpoint, while the listed secondary outcome is change in global cognition. Why this vascular measure was selected, and how it relates to any other co-primary endpoint, are not described in detail in the registry record. In general, a surrogate or subclinical measure can provide information about change in a biological or imaging outcome, but its results do not by themselves establish effects on clinical events or cognition.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The posted arm-specific mean rates of carotid intima-media thickness change are 0.013 mm per year for nattokinase and 0.011 mm per year for placebo. The respective 95% confidence intervals are 0.01 to 0.015 and 0.009 to 0.013. The reported sponsor analysis is a mixed-model analysis, with p = 0.31 for the superiority comparison.
The p-value describes how compatible the observed comparison is with the null hypothesis under the model and its assumptions. A p-value of 0.31 is not conventionally small evidence against that null hypothesis. It does not establish that the arms are identical, nor does it indicate the probability that the null hypothesis is true. The registry does not report a model-based between-arm estimate or its confidence interval, so the magnitude and precision of the adjusted contrast cannot be read directly from the posted p-value.
The arm-level confidence intervals describe uncertainty around each arm's mean rate, not the uncertainty around the difference between arms. Their endpoints should therefore not be used as a substitute for a confidence interval on the treatment comparison. The registry reports no recalculated two-arm contrast for this continuous outcome because the needed comparison values are missing or not estimable from the posted summary statistics.
The reported means are close in the displayed units, but the clinical meaning of any difference is not specified in the registry record. Nor does the record state a threshold for a clinically important change. The result addresses an imaging-based measure of subclinical atherosclerosis, not a direct count of cardiovascular events. The record also lists change in global cognition as a secondary outcome, but its numerical result is not included in the registry record for this interpretation.
The registry does not report the analysis population, how missing or unusable ultrasound measurements were handled, or whether the mixed model included covariate adjustment. Those details affect how precisely the reported comparison can be understood. No conclusion about effects on dementia, cardiovascular disease, or other clinical outcomes follows from the primary endpoint p-value alone.
In general, statistical evidence and clinical importance are related but distinct. A result can be imprecise even when arm means differ, and a small p-value would not by itself establish a meaningful clinical effect. In this record, the displayed summaries and reported comparison should be read as estimates of the specified imaging outcome in the analyzed data, with important modeling and missing-data details not reported.
9. Statistical Methods Explained
The outcome is continuous: the registry defines it as a mean rate of change in intima-media thickness, with units of mm per year. The arm summaries are means accompanied by 95% confidence intervals. In general, a confidence interval expresses the uncertainty associated with an estimate under the method used to construct it. It is not a range that contains a fixed proportion of individual participants' measurements, and it does not directly give the interval for the difference between arms unless that contrast is specifically calculated.
What does the mixed-model result represent?
The sponsor reports a mixed-model analysis for the comparison of nattokinase with placebo and gives p = 0.31. In general, mixed models are designed for data with repeated measurements or other forms of related observations. They can account for the fact that multiple measurements from the same participant are not independent. This can be useful when an outcome is assessed repeatedly over time, as the registry describes for carotid ultrasound measurements.
The registry record does not report the model formula, covariance structure, fixed or random effects, adjustment variables, or the precise contrast tested. It also does not give a model-based effect estimate or confidence interval for the between-arm contrast. Those omissions limit the ability to reconstruct or fully assess the analysis from the posted summaries. The p-value should be attributed to the sponsor's reported mixed-model comparison rather than treated as a test calculated from the two arm means alone.
Why is a direct arm comparison different from comparing the arm intervals?
The two listed confidence intervals are attached to separate arm means. A treatment comparison instead concerns a contrast between the arms and has its own uncertainty, which depends on the variability and structure of the underlying data and the analysis method. In general, the overlap or non-overlap of separate confidence intervals is not a replacement for a formal comparison. Here, the sponsor's mixed-model p-value is the posted inferential result for the primary comparison.
The registry's recalculation status says that a two-arm comparison for this continuous outcome could not be estimated from the available summary statistics. Thus, there is no independent unadjusted contrast or recalculated confidence interval to set alongside the sponsor's model result. The absence of such a recalculation does not mean that the sponsor's analysis is unavailable; its p-value is reported. The record marks no discrepancy between a recalculation and the sponsor's analysis.
What do the confidence intervals say about precision?
The nattokinase mean is reported as 0.013 mm per year, with interval 0.01 to 0.015. The placebo mean is 0.011 mm per year, with interval 0.009 to 0.013. These intervals communicate uncertainty around the arm means. They do not indicate the distribution of individual responses or show what would happen for a particular participant. They also cannot alone establish whether a difference is clinically important.
Precision is also distinct from statistical significance. A narrow interval can indicate a more precise estimate, while a broad interval indicates greater uncertainty, but interpretation requires knowing what parameter the interval estimates. Because no confidence interval for the model-based contrast is posted, precision of the treatment comparison cannot be directly assessed from the record's primary analysis result.
Why can repeated measurements and over-dispersion matter?
In general, repeated measurements on the same participant tend to be correlated. A method that treats every measurement as independent may underestimate uncertainty if it ignores that dependence. Mixed models can represent such correlation, depending on the model specification. The registry's use of a mixed-model analysis is consistent with the repeated measurement schedule, but the details needed to evaluate its specific assumptions are not reported in the registry record.
Over-dispersion is a general issue when observed variation is greater than a model assumes. If not accounted for, standard errors and confidence intervals can be misleading. For a continuous repeated-measures outcome, the relevant variance and correlation assumptions should be specified and checked. The registry record does not state whether over-dispersion was considered or how model assumptions were assessed, so no claim about those diagnostics can be made.
How should planning illustrations be read?
Any sample-size values elsewhere in the statistical material are planning illustrations based on the observed effect, not observed power and not the trial's original design justification. The stated illustration assumptions are a two-sided significance level of 0.05, equal allocation, a normal approximation, and detection of a stated fraction of the observed effect. For non-time-to-event outcomes, those illustrations are expressed as participants per arm for 80% power or another stated power target. Such calculations are hypothetical planning aids: they do not change the trial's observed enrollment or show the probability that the study detected a true effect.
The original sample-size justification, power assumptions, and any prespecified interim analyses are not reported in the registry record. The proposed enrollment in the brief summary is 240, whereas recorded enrollment is 265. The record does not explain this difference, and the posted information does not establish why the final enrollment took that value.
10. Analyses Reported by the Sponsor
| Method | Estimate | 95% CI | P-value |
|---|---|---|---|
| Mixed Models Analysis | not reported | 0.310 |
The sponsor's own reported comparison is a mixed-model analysis with p = 0.31. The registry posts arm-specific mean rates and confidence intervals, but it does not provide the model-based contrast estimate or its confidence interval. A simple unadjusted re-calculation of the two-arm continuous-outcome comparison is not estimable from the posted summary statistics, so no recalculated effect or p-value can be compared with the sponsor result.
The registry marks no discrepancy between a recalculation and the sponsor analysis. In this case, there is no estimable recalculation that could disagree with the reported result; the sponsor's mixed-model finding is therefore the available inferential comparison. The arm means are useful descriptive summaries, but they do not reproduce the repeated-measures model analysis.
In general, a model-based result can depend on the participant-level measurements, their timing, the model structure, and how incomplete observations are handled. The registry does not report those details or the mixed-model contrast estimate. The reported p-value should be interpreted as the sponsor's analysis of the primary comparison, not as a calculation from the posted arm means and confidence intervals. Its magnitude does not establish equivalence between the assignments or rule out effects outside the observed dataset's precision.
11. Multiplicity, Missing Data and Analysis Populations
The registry lists four outcome measures and three analyses. It describes the primary endpoint as a co-primary endpoint and lists change in global cognition as a secondary outcome. However, the full outcome hierarchy, the complete set of planned comparisons, and any multiplicity adjustment are not reported in the registry record. It is therefore not possible to determine from the record whether the reported p-value was adjusted for testing multiple endpoints or analyses, or how such testing was intended to be interpreted.
In general, testing several hypotheses can increase the chance of obtaining at least one apparently notable result by chance. Prespecified multiplicity procedures can control that issue, but no such procedure should be inferred without documentation. The posted primary comparison is p = 0.31; no claim about a multiplicity-adjusted value is available.
Missing-data methods are also not reported in the registry record. The registry gives group sizes for the primary outcome, but does not state whether all randomized participants contributed measurements at every scheduled time, how incomplete or unusable scans were handled, or whether any imputation or sensitivity analyses were used. A mixed model can use repeated observations under specified assumptions, but the presence of a mixed model alone does not establish how missingness was addressed or whether its assumptions were met.
The analysis population is not reported in the registry record. The available information does not say whether the primary analysis followed intention-to-treat principles, used a per-protocol set, or used another defined group. Nor are exclusions after randomization described. These details matter because the population included in an analysis may differ from the people assigned at random, and that can affect interpretation of the comparison.
Accordingly, the sponsor's p-value should be read with the reported analysis method and the limits of the available reporting in mind. The record does not support a conclusion about multiplicity control, missing-data robustness, or the exact analysis population.
12. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Nattokinase | 13 | 132 | 9.8% |
| Placebo | 15 | 133 | 11.3% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Nattokinase | 0 of 132 |
| Placebo | 1 of 133 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Nattokinase | 58 of 132 |
| Placebo | 57 of 133 |
Most frequent serious adverse event terms
| Event term | Organ system | Nattokinase | Placebo |
|---|---|---|---|
| Prostate cancer | Neoplasms benign, malignant and unspecified (incl cysts and polyps) | 4 of 132 | 1 of 133 |
| Knee arthroplasty | Surgical and medical procedures | 1 of 132 | 2 of 133 |
| Pancreatic cancer | Neoplasms benign, malignant and unspecified (incl cysts and polyps) | 0 of 132 | 2 of 133 |
| Non-fatal myocardial infarction | Cardiac disorders | 0 of 132 | 2 of 133 |
| Coronary artery bypass graft surgery | Surgical and medical procedures | 1 of 132 | 1 of 133 |
| Cerebrovascular accident | Nervous system disorders | 1 of 132 | 1 of 133 |
| Pulmonary embolism | Respiratory, thoracic and mediastinal disorders | 0 of 132 | 2 of 133 |
| Chest pain | Cardiac disorders | 1 of 132 | 0 of 133 |
| Transient ischemic attack | Nervous system disorders | 0 of 132 | 1 of 133 |
| Deep vein thrombosis | Vascular disorders | 0 of 132 | 1 of 133 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Nattokinase | Placebo |
|---|---|---|---|
| Basal cell carcinoma | Neoplasms benign, malignant and unspecified (incl cysts and polyps) | 4 of 132 | 4 of 133 |
| Dyspepsia | Gastrointestinal disorders | 1 of 132 | 3 of 133 |
| Fall | Injury, poisoning and procedural complications | 1 of 132 | 3 of 133 |
| Hypoaesthesia | Nervous system disorders | 2 of 132 | 2 of 133 |
| Cataract operation | Surgical and medical procedures | 3 of 132 | 1 of 133 |
| Knee arthroplasty | Surgical and medical procedures | 2 of 132 | 2 of 133 |
| Atrial fibrillation | Cardiac disorders | 1 of 132 | 2 of 133 |
| Vertigo | Ear and labyrinth disorders | 1 of 132 | 2 of 133 |
| Urinary tract infection | Infections and infestations | 0 of 132 | 3 of 133 |
| Contusion | Injury, poisoning and procedural complications | 2 of 132 | 1 of 133 |
Serious events with the highest arm-level rate: prostate cancer (3.0%), pancreatic cancer (1.5%), non-fatal myocardial infarction (1.5%), pulmonary embolism (1.5%), knee arthroplasty (1.5%).
The serious adverse event summaries list 13 affected participants among 132 at risk in the nattokinase arm and 15 among 133 at risk in the placebo arm. The registry separately lists other adverse events affecting 58 of 132 participants in the nattokinase arm and 57 of 133 in the placebo arm. Deaths are recorded as 0 in the nattokinase arm and 1 in the placebo arm.
The serious-event terms listed most frequently include prostate cancer at 3.0%, pancreatic cancer at 1.5%, non-fatal myocardial infarction at 1.5%, pulmonary embolism at 1.5%, knee arthroplasty at 1.5%, and chest pain at 0.8%. These are registry-listed terms and percentages; the record does not provide further detail about event timing, relatedness, severity grading, or how terms were selected for display.
These figures are descriptive counts, not proof that an assignment caused or prevented an event. The registry does not report a formal between-arm safety analysis or an account of adjudication procedures. In general, rare events and differences in follow-up can make safety comparisons uncertain, and an event count alone does not establish causation. The posted summaries should be read as reported events among the listed participants, not as a complete assessment of safety or a basis for individual medical decisions.
13. What the Result Does, and Does Not, Mean
The result addresses a specific measurement: the mean annual rate of change in distal common carotid artery far-wall intima-media thickness. The nattokinase arm's rate is reported as 0.013 mm per year and the placebo arm's as 0.011 mm per year. The sponsor's mixed-model comparison has p = 0.31. These findings describe the observed endpoint and the reported statistical comparison in this trial.
They do not show that the two assignments are identical. A non-small p-value is not an equivalence test, and the registry does not report an equivalence margin or a model-based confidence interval for the difference. The result also does not establish that any possible difference is clinically important or unimportant; a clinical-importance threshold is not reported.
The endpoint is an ultrasound measure of subclinical atherosclerosis progression. It is not itself a count of cardiovascular events, a measure of dementia incidence, or a direct assessment of how participants felt or functioned. The registry lists global cognition as a secondary outcome, but its numerical result is not available in the registry record for this analysis. Therefore, the primary endpoint result cannot answer the cognition question by itself.
General statistical interpretation distinguishes an observed estimate from a causal conclusion. Randomization is a design feature intended to support a fair comparison, while the analysis, missing observations, adherence, and outcome measurement also affect what can be inferred. The record does not provide enough analysis detail to assess all of those components. The result should therefore be understood as the reported comparison for the specified endpoint, rather than a broad claim about all cardiovascular or cognitive outcomes.
Nor does a result from this trial necessarily generalize to people outside the target population described in the registry. Eligibility details and the characteristics of enrolled participants are not reported in the registry record. The study's recorded enrollment and endpoint summaries do not, on their own, establish how the findings apply to other populations or settings.
14. Why This Trial Matters Statistically
This trial illustrates why an outcome's definition and scale matter. Its primary measure is a rate of change in carotid intima-media thickness, stated in mm per year, rather than a clinical event count. That choice permits the study to quantify change in a subclinical imaging measure over repeated assessments, but interpretation must stay anchored to that measure.
It also shows the difference between descriptive arm summaries and a modeled comparison. Separate means and confidence intervals describe each arm. The mixed-model p-value comes from an analysis of the comparison, and the two forms of reporting are not interchangeable. Because the registry does not post a model-based contrast estimate or interval, readers cannot judge the comparison's estimated magnitude and precision from the sponsor result alone.
The record is also a useful reminder that statistical reporting needs to include more than a p-value. Model specifications, analysis populations, missing-data handling, multiplicity procedures, and clinically meaningful thresholds help readers understand what an estimate does and does not establish. Those details are not reported in the registry record for this trial.
Finally, the listed cognitive outcome and the endpoint description's co-primary wording make outcome hierarchy important. Without a complete prespecified analysis plan or results for every outcome, it is not possible to reconstruct the full inferential framework. The posted result remains informative as a report of the specified vascular measure, but its scope is narrower than the study's broader prevention motivation.
15. 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 primary comparison is reported as a mixed-model p-value without a model-based between-arm estimate or confidence interval. The arm-specific means and confidence intervals do not fill that gap. The posted summaries also do not support a recalculated two-arm continuous-outcome contrast.
The registry does not report model specification, covariate adjustment, covariance structure, assumption checks, missing-data handling, or the analysis population. It also does not provide a full account of the outcome hierarchy, multiplicity adjustment, interim analyses, or sample-size justification. These omissions limit independent assessment of the inferential result.
Although the endpoint is called a co-primary trial endpoint in its description, the registry lists global cognition as a secondary outcome. The complete primary endpoint set and numerical results for all outcome measures are not reported in the information. The cognitive result therefore cannot be interpreted alongside the vascular result.
The brief summary describes a proposed enrollment of 240, while recorded enrollment is 265. The registry does not explain that difference. Participant eligibility details, baseline characteristics, treatment adherence, and follow-up completeness are also not reported in the registry record. These gaps limit judgments about representativeness and the role of incomplete follow-up.
The endpoint is a subclinical imaging measure, and the record does not identify a clinically important threshold for its rate of change. The reported result should not be extended to clinical events or cognitive outcomes without those outcomes' own results. Safety summaries are descriptive and lack reported causal attribution or a formal comparative analysis.
16. Related Tutorials
17. Related Statistical Calculators
18. Sources
- ClinicalTrials.gov: Nattokinase Atherothrombotic Prevention Study — NCT02080520.
- Registry record last updated 2024-03-19; status last verified 2024-03.
- Results first posted 2024-03-19.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
19. Record Summary
The Nattokinase Atherothrombotic Prevention Study is recorded as a completed, randomized, quadruple-masked, parallel-group trial sponsored by the University of Southern California. Its recorded enrollment is 265. Participants were assigned to oral nattokinase at 2,000 fibrinolytic units daily or matched placebo daily.
The primary endpoint is the mean rate of change in distal common carotid artery far-wall intima-media thickness, measured in mm per year. The arm means are 0.013 for nattokinase and 0.011 for placebo; the reported mixed-model comparison has p = 0.31. A two-arm recalculation is not estimable from the posted summary statistics.
The registry also lists global cognition as a secondary outcome and provides descriptive adverse-event counts. Detailed model specifications, missing-data methods, analysis population, multiplicity handling, and the numerical cognitive result are not reported in the registry record.