Prevalence PheDASOdds ratio
Loading evidence…
Disease feature
PheCode 401.1
PheCode 401.1
Clinical phenotype
Primary presentation
Severe vs None
Incidence PheDASHazard ratio
Loading evidence…
Adjustment ladder
How the estimate changes from M1 to M4
M4 is emphasized as the primary presentation model. M3 excludes BMI.
Prevalence
Odds ratios
Loading model comparison…
Incidence
Hazard ratios
Loading model comparison…
Attenuation between M3 and M4 may be scientifically informative because BMI can be conceptualized differently across questions. The atlas presents both estimates without assigning a causal role.
Dose pattern
M4 associations across OSA severity
Each row compares an OSA severity group with AHI <5 in the referred cohort.
Prevalence
OR by severity
| Contrast | OR | Interval | FDR |
|---|
Incidence
HR by severity
| Contrast | HR | Interval | FDR |
|---|
Existing disease evidence
L1-regularized logistic models summarize disease presence before index. Odds ratios are shrunk and intervals are non-classical.
Disclosure notePer-feature case counts are not present in the current combined output. Public data downloads remain disabled pending approved disclosure checks.
New-onset disease evidence
Ridge cause-specific Cox models follow at-risk participants after index. Disease-specific at-risk and event counts, stability, and proportional-hazards diagnostics appear beside each result.
Model noteIncidence M3 also includes baseline encounter rate; model labels are analysis-specific.
Association, not causation
The reference is a symptomatic referral group. These estimates do not measure population OSA prevalence and should not be described as causal, protective, or independent risk effects.
Return to this feature in the scan →