INTRODUCTION
Previous studies have shown that there is a range of FEV1 trajectories across the lifespan. Yet, how trajectories defined by the combination of both FEV1 and FVC relate to airflow obstruction (AO; FEV1/FVC<0.7 and FEV1< 80% ref.) or Preserved Ratio Impaired Spirometry (PRISm; FEV1/FVC ? 0.7 and FEV1 < 80% ref.) is unknown.
METHODS
We applied Bayesian non-parametric mixture modeling to data from the Framingham Offspring Cohort (FOC). Trajectory subgroups were identified by simultaneously modeling both pre-bronchodilator FEV1 and FVC (in liters) as a function of age and height. We modeled all males >20 yr. (N=1,333). The prevalence of AO or PRISm and age of onset was compared across trajectories.
RESULTS
This analysis identified seven FEV1 and FVC defined trajectories. Their description (Table1) was based on both measurements, which were in accordance with each other.
|
Description (based on FEV1 and FVC) |
N individuals |
Airflow limitation (n, % traj, % of all AO) |
Mean age of onset for AO |
PRISm (n, % traj) | Mean age of onset for PRISm |
| Supranormal |
44 (3.3%) |
- | - | - | - |
| Above Average |
359 (26.9%) |
1 (0.28%, 0.61%) |
51 |
- | - |
| Above average, accelerated decline | 13 (1%) | 10 (77%, 6.1%) |
55 |
2 (15.4%) |
52 |
| Average |
549 (41.2%) |
32 (5.8%, 19.6%) |
57 |
6 (1.09%) |
49 |
| Below Average | 271 (20.3%) |
79 (29.2%, 48.5%) |
51 |
124 (45.8%) |
37 |
| Persistently Low | 55 (4.1%) |
22 (40%, 13.5%) |
54 |
46 (83.6%) |
35 |
| Below average, accelerated decline | 14 (1.1%) | 8 (57.1%, 4.9%) | 46 | 9 (64.3%) | 32 |
CONCLUSIONS
Bayesian non-parametric trajectory modeling identified seven combined FEV1 and FVC life-course trajectories differentially associated to AO or PRISm.