This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record.
1. Trial at a Glance
The LIFE Study was a randomized phase 3 parallel-group trial comparing a physical activity behavioral intervention with a successful aging behavioral intervention in adults studied for sedentary lifestyle, risk of disability, and aging. The primary outcome was major mobility disability, defined around the ability to complete a 400-meter walk.
| Feature | LIFE |
|---|---|
| Trial name | LIFE Study — Lifestyle Interventions and Independence for Elders |
| Phase | Phase 3 |
| Conditions | Sedentary Lifestyle; Risk of Disability; Aging |
| Design | Randomized, parallel-group, single-masked |
| Primary purpose | Prevention |
| Enrollment | 1635 |
| Interventions | Physical Activity (behavioral); Successful Aging (behavioral) |
| Primary endpoint | Major Mobility Disability, Defined as Incapacity to Walk 400 Meters |
| Primary analysis | Cox proportional-hazards model |
| Effect measure | Hazard ratio |
| Hypothesis type | Superiority |
| ClinicalTrials.gov | NCT01072500 |
2. Clinical Question
The central question was whether the physical activity intervention differed from the successful aging intervention with respect to the time to major mobility disability in a population characterized by sedentary lifestyle, risk of disability, and aging.
Population
Participants enrolled in the LIFE Study for whom the trial addressed sedentary lifestyle, risk of disability, and aging.
Intervention
Physical Activity, classified in the registry as a behavioral intervention.
Comparator
Successful Aging, also classified in the registry as a behavioral intervention.
Primary question
Does physical activity change the time-to-event outcome of major mobility disability compared with successful aging?
3. Trial Design
Physical Activity
- Behavioral intervention
- Compared directly with the successful aging intervention
- Primary statistical comparison used a Cox proportional-hazards model
Successful Aging
- Behavioral intervention
- Comparator for the primary superiority analysis
- Primary statistical comparison used a Cox proportional-hazards model
Trial timing
Study start
The registry records February 2010 as the study start.
Primary completion
The registry records December 2013 as the primary completion date.
4. Endpoints
Primary Endpoint: Major Mobility Disability
The registered primary outcome was Major Mobility Disability, Defined as Incapacity to Walk 400 Meters. The registered time frame was median 2.7 years / average 2.6 years.
Participants were asked to walk 400 m at their usual pace, without overexerting, on a 20-meter course for 10 laps. Each lap was 40 meters. Participants could stop for up to 1 minute for fatigue or related symptoms.
When major mobility disability could not be objectively measured because the participant could not come to the clinic and a suitable walking course was unavailable at the participant's home, institution, or hospital, an alternative adjudication was used. This was based on objective inability to walk 4 meters in less than 10 seconds, or self-, proxy-, or medical-record-reported inability to walk across a room.
The registry notes that participants meeting these alternative criteria would not be able to complete the 400-meter walk within 15 minutes.
Secondary Endpoint
The posted secondary analysis evaluated Persistent Mobility Disability (Assessed Every 6 Months) over the same registered time frame of median 2.7 years / average 2.6 years.
| Endpoint | Time frame | Statistical type | Posted method |
|---|---|---|---|
| Major Mobility Disability, Defined as Incapacity to Walk 400 Meters | Median 2.7 years / Average 2.6 years | Time-to-event | Cox proportional-hazards model |
| Persistent Mobility Disability (Assessed Every 6 Months) | Median 2.7 years / Average 2.6 years | Time-to-event | Cox proportional-hazards model |
5. Statistical Methodology
Cox proportional-hazards model
The registry reports a Cox proportional-hazards model for the primary analysis. This is a time-to-event regression model designed for outcomes where both the occurrence of an event and the timing of that event contribute information.
The hazard ratio is obtained from the relative hazard represented by the fitted regression coefficient. In this trial, the reported effect measure was the hazard ratio comparing Physical Activity with Successful Aging.
Stratified analysis
The primary analysis notes that the intervention comparison used a likelihood ratio test from a Cox regression model, stratified by field center and sex. Stratification allows the baseline hazard to differ across the specified strata while estimating the treatment comparison within the Cox framework.
Superiority hypothesis
The registry classifies the hypothesis type as superiority. The statistical question is therefore whether the time-to-event experience differs between the randomized interventions, rather than whether one intervention is merely sufficiently close to the other under a prespecified non-inferiority margin.
Two-sided confidence interval
The primary hazard ratio was accompanied by a 95% two-sided confidence interval. A confidence interval provides information about the statistical precision of the estimated hazard ratio and should be interpreted alongside the point estimate rather than replaced by the P-value.
Likelihood ratio testing
The primary analysis notes that the intervention comparison used a likelihood ratio test from the Cox regression model. Conceptually, a likelihood ratio test compares the statistical fit of models that differ in the treatment effect being tested. In a randomized trial, the resulting inference is anchored to the randomized treatment comparison, while the Cox model supplies the time-to-event framework.
6. Primary Result: Major Mobility Disability
The registry posts a formal statistical analysis for the primary endpoint. The comparison was Physical Activity versus Successful Aging using a Cox proportional-hazards model stratified by field center and sex.
Hazard ratio for major mobility disability
95% CI: 0.69–0.98 · P = 0.03
Comparison: Physical Activity vs Successful Aging
| Primary endpoint | Physical Activity vs Successful Aging | 95% CI | P-value | Analysis |
|---|---|---|---|---|
| Major Mobility Disability, Defined as Incapacity to Walk 400 Meters | HR 0.82 | 0.69–0.98 | 0.03 | Cox proportional-hazards model; stratified by field center and sex |
The hazard ratio of 0.82 means that, under the fitted Cox model, the estimated instantaneous rate of major mobility disability was lower in the Physical Activity group than in the Successful Aging group during the analyzed follow-up.
The estimate is a relative time-to-event measure. It does not mean that 82% of participants avoided disability, that 18% of participants necessarily benefited, or that every participant experienced the same proportional change in risk.
The 95% confidence interval of 0.69–0.98 describes uncertainty around the estimated hazard ratio. It is relatively close to the null value of 1, so the point estimate should be interpreted together with its interval rather than in isolation.
The P-value of 0.03 addresses the statistical evidence against the null hypothesis under the specified analysis. It does not measure the magnitude or clinical importance of the treatment effect.
Because the analysis is based on a Cox proportional-hazards model, interpretation of a single hazard ratio also depends on the model's proportional-hazards framework. The reported analysis additionally used stratification by field center and sex, which accounts for those factors in the baseline hazard structure.
7. Secondary Result: Persistent Mobility Disability
The posted secondary analysis evaluated persistent mobility disability assessed every 6 months. It used the same general time-to-event framework, with Physical Activity compared with Successful Aging using a Cox proportional-hazards model.
Hazard ratio for persistent mobility disability
95% CI: 0.57–0.91 · P = 0.006
Comparison: Physical Activity vs Successful Aging
| Secondary endpoint | Effect estimate | 95% CI | P-value | Analysis |
|---|---|---|---|---|
| Persistent Mobility Disability (Assessed Every 6 Months) | HR 0.72 | 0.57–0.91 | 0.006 | Cox proportional-hazards model |
The hazard ratio of 0.72 indicates a lower estimated instantaneous rate of persistent mobility disability for Physical Activity relative to Successful Aging under the reported Cox model.
As with the primary endpoint, the hazard ratio is not a probability and does not directly state the proportion of participants who will remain free of persistent mobility disability.
The 95% confidence interval of 0.57–0.91 expresses uncertainty around the estimated relative hazard. The interval is entirely below 1, while the P-value of 0.006 summarizes the statistical evidence under the reported hypothesis-testing framework.
The secondary endpoint should nevertheless be distinguished from the primary endpoint. A secondary result can provide additional evidence about the intervention's effect on mobility, but it does not change the definition or statistical role of the prespecified primary outcome.
8. Statistical Methods Explained
Why was a Cox proportional-hazards model used?
Both posted analyses are time-to-event analyses. The outcome is not simply whether mobility disability occurred; the timing of the event also contributes information. The Cox model is designed for this setting and produces a hazard ratio comparing the event rate between groups over follow-up while allowing censoring to contribute information up to the point at which participants are no longer observed.
What does a hazard ratio of 0.82 mean?
A hazard ratio of 0.82 indicates that the estimated instantaneous event rate under the fitted model was lower for Physical Activity than for Successful Aging. It is not a relative risk, an absolute risk reduction, or a statement that exactly the same proportional reduction applies to every participant at every time point.
Why is the confidence interval important?
The primary estimate of 0.82 is only one estimate of the underlying treatment effect. The 95% confidence interval of 0.69–0.98 communicates the statistical uncertainty around that estimate. A reader should therefore consider both the location and width of the interval, not only whether the P-value crosses a conventional threshold.
Why was the analysis stratified by field center and sex?
The registry states that the primary likelihood ratio test came from a Cox regression model stratified by field center and sex. Stratification allows the baseline hazard to vary across these strata rather than requiring a common baseline hazard for all participants. The treatment comparison is then estimated within the Cox framework while accounting for those stratification factors.
What does the P-value of 0.03 tell us?
The P-value quantifies the statistical evidence against the null hypothesis under the specified model and testing procedure. It does not tell us that there is a 3% probability that the null hypothesis is true, and it does not quantify the size of the treatment effect. The hazard ratio and its confidence interval provide the effect-size information.
Why is this a time-to-event endpoint rather than a simple binary endpoint?
The registry classifies the primary endpoint as binary but the posted statistical analysis as time-to-event. The distinction matters because the analysis uses when major mobility disability occurs, not merely whether it ever occurred. This permits Cox regression to incorporate the timing of events and censoring during the follow-up period.
9. Understanding the Primary Endpoint More Closely
The primary endpoint has an unusually concrete functional definition: inability to complete a 400-meter walk within 15 minutes under specified conditions. That definition connects the statistical event directly to a measurable functional capability.
Objective assessment
The registered endpoint used a 400-meter walk test, with specific requirements for completion time, sitting, assistance, pacing, and permitted cane use.
Alternative adjudication
When the standard walk could not be objectively performed because of access or course availability, specified alternative functional criteria could determine the outcome.
Time-to-event structure
The statistical analysis uses the timing of major mobility disability rather than treating all observed events as if they occurred at the same time.
Clinical meaning
The endpoint represents a functional milestone rather than a laboratory measurement, making endpoint definition and assessment conditions especially important to interpretation.
10. Serious Adverse Events
The registry reports serious adverse events by randomized arm as affected participants among those at risk.
| Intervention | Affected / At risk |
|---|---|
| Physical Activity | 404 / 818 |
| Successful Aging | 373 / 817 |
The reported serious-adverse-event counts should be interpreted as a safety comparison between the randomized intervention groups. The registry gives the affected and at-risk counts, but the clinical meaning of an adverse-event comparison depends on the definition, timing, severity, duration, and attribution of the individual events.
The efficacy hazard ratios and serious-adverse-event counts answer different statistical questions. A time-to-event efficacy estimate summarizes the relative occurrence of the mobility endpoint, whereas the safety counts describe participants affected by serious adverse events.
11. What the Hazard Ratio Does — and Does Not — Mean
The primary hazard ratio of 0.82 summarizes the relative difference in the instantaneous rate of major mobility disability estimated by the Cox model. It is a model-based effect measure for the observed follow-up period.
A hazard ratio cannot by itself tell a reader how many additional participants avoided major mobility disability. Absolute event probabilities require additional information about the event-time distribution.
The hazard ratio describes a treatment-group comparison. It does not imply that each individual participant has an identical proportional change in risk or that an individual's outcome can be predicted from the hazard ratio alone.
The confidence interval of 0.69–0.98 should be read alongside the point estimate of 0.82. The interval communicates the statistical precision of the estimated treatment comparison.
12. Stratification and the Role of Field Center and Sex
The primary analysis used a Cox regression model stratified by field center and sex. This is an important statistical detail because the model did not simply treat all participants as belonging to one homogeneous risk stratum.
The reported likelihood ratio test was based on a Cox regression model stratified by field center and sex.
In a stratified Cox model, each stratum can have its own baseline hazard function while the treatment effect is estimated across the strata. This can be useful when the baseline event process is expected to differ systematically across important design factors.
Importantly, stratification does not mean that the analysis estimates a separate treatment effect for every field center and sex category. Rather, the factors define the stratification structure of the Cox model used for the primary comparison.
13. Censoring and Time-to-Event Interpretation
Time-to-event analysis is useful when participants are followed for different lengths of time or when some participants have not experienced the event by the end of observation. Such participants can remain informative through their observed follow-up rather than being treated simply as event-free for the entire study.
For the LIFE primary endpoint, the registry reports a median follow-up time frame of 2.7 years and an average of 2.6 years. The Cox analysis uses the event timing contained within that follow-up structure.
14. Primary and Secondary Analyses Together
| Feature | Primary endpoint | Secondary endpoint |
|---|---|---|
| Outcome | Major Mobility Disability, Defined as Incapacity to Walk 400 Meters | Persistent Mobility Disability (Assessed Every 6 Months) |
| Time frame | Median 2.7 years / Average 2.6 years | Median 2.7 years / Average 2.6 years |
| Effect measure | Hazard ratio | Hazard ratio |
| Estimate | 0.82 | 0.72 |
| 95% CI | 0.69–0.98 | 0.57–0.91 |
| P-value | 0.03 | 0.006 |
| Hypothesis type | Superiority | Superiority |
The two estimates address related but distinct mobility outcomes. The primary analysis focuses on the registered major mobility disability definition, while the secondary analysis evaluates persistent mobility disability assessed every 6 months. Their shared Cox framework makes the hazard ratios comparable in form, but the endpoint definitions remain different.
15. Statistical Interpretation of Confidence Intervals and P-values
The primary estimate is HR 0.82, with a 95% confidence interval of 0.69–0.98 and a P-value of 0.03. These three quantities play different roles.
Hazard ratio
The hazard ratio is the estimated relative treatment effect from the Cox model.
Confidence interval
The confidence interval describes uncertainty and precision around the estimated hazard ratio.
P-value
The P-value describes the evidence against the null hypothesis under the specified statistical test.
Clinical meaning
None of these quantities alone describes the complete clinical importance of the intervention.
The same distinction applies to the secondary analysis. Its estimate of 0.72, confidence interval of 0.57–0.91, and P-value of 0.006 should be read as a coordinated statistical description rather than as interchangeable measures.
16. What the Registry Reports About Statistical Analyses
The registry reports 2 statistical analyses and 2 posted outcome measures. One analysis corresponds to the primary endpoint and one to the secondary endpoint. The primary analysis includes an estimate and confidence interval.
| Analysis characteristic | Registry information |
|---|---|
| Results posted | Yes |
| Outcome measures posted | 2 |
| Statistical analyses posted | 2 |
| Primary endpoint analyses | 1 |
| Primary analyses with estimate + CI | 1 |
| Effect measure | Hazard ratio |
| Hypothesis type | Superiority |
This structure provides a useful distinction between the endpoint definition and the statistical analysis used to evaluate it. The primary endpoint is expressed as a functional mobility event, while the statistical analysis treats that event as a time-to-event outcome.
17. Limitations and Interpretation Issues
- Endpoint definition: major mobility disability is operationalized through a specific walking test and specified alternative criteria. Interpretation therefore depends on consistent application of those criteria.
- Time-to-event modeling: the hazard ratio is a model-based summary and should be interpreted within the Cox proportional-hazards framework.
- Stratification: the primary analysis was stratified by field center and sex. The resulting hazard ratio is therefore not simply an unadjusted comparison of event counts.
- Confidence-interval precision: the primary confidence interval of 0.69–0.98 spans a range of plausible relative effects and should be considered alongside the point estimate.
- Secondary endpoint: persistent mobility disability is a distinct endpoint from the registered primary outcome and should not be treated as interchangeable with major mobility disability.
- Safety interpretation: serious adverse-event counts are reported separately from efficacy outcomes and do not by themselves establish an overall benefit-risk balance.
- Incomplete descriptive detail: the registry information summarized here does not provide additional baseline or subgroup estimates beyond the posted analyses described above.
18. Why This Trial Matters Statistically
The LIFE Study is a useful teaching example because it connects a clinically concrete functional endpoint to a formal time-to-event analysis. The primary outcome is easy to conceptualize as an event—major mobility disability—but the analysis goes beyond a simple event/no-event comparison by incorporating when the event occurs.
| Statistical concept | How it appears in LIFE |
|---|---|
| Randomization | Participants were randomized to Physical Activity or Successful Aging. |
| Parallel-group design | The two behavioral interventions were compared in parallel. |
| Time-to-event endpoint | Major mobility disability was analyzed according to time to occurrence. |
| Cox proportional-hazards model | Used for the posted primary and secondary statistical analyses. |
| Hazard ratio | Used as the effect measure for both posted analyses. |
| Stratified analysis | The primary Cox model was stratified by field center and sex. |
| Confidence interval | The primary HR was reported with a 95% two-sided CI. |
| P-value | Reported for both the primary and secondary analyses. |
| Superiority testing | The registry identifies the hypothesis type as superiority. |
The trial therefore illustrates an important principle in clinical biostatistics: the endpoint definition, follow-up structure, analysis model, effect measure, confidence interval, and hypothesis test all contribute different pieces of the statistical interpretation.
19. Related Tutorials
Learn more about the methods used in this trial:
20. Related Calculators
21. Sources
- ClinicalTrials.gov: NCT01072500 — LIFE Study.
- PubMed: PMID 23685766.
- PubMed: PMID 23589390.
- PubMed: PMID 23716501.
- PubMed: PMID 24222666.
- PubMed: PMID 24049442.
Continue through Clinical Biostats
Explore statistical tutorials, clinical-trial methods, and calculators that connect trial endpoints to the underlying biostatistical concepts.
22. Record Summary
The LIFE Study provides a clear example of a randomized clinical-trial analysis in which a functional outcome is evaluated as a time-to-event endpoint. The primary comparison of Physical Activity versus Successful Aging produced a hazard ratio of 0.82 for major mobility disability, with a 95% confidence interval of 0.69–0.98 and a P-value of 0.03. The secondary analysis of persistent mobility disability produced a hazard ratio of 0.72, with a 95% confidence interval of 0.57–0.91 and a P-value of 0.006.
The statistical interpretation depends on understanding the endpoint definition, the time-to-event framework, the stratified Cox model, and the distinction between a hazard ratio, confidence interval, and P-value. Together, these elements describe not only the direction of the randomized comparison but also the uncertainty and modeling assumptions surrounding the estimated treatment effect.