Clinical, physiologic, and radiographic factors contributing to development of hypoxemia in moderate to severe COPD: a cohort study

Abstract

Background: Hypoxemia is a major complication of COPD and is a strong predictor of mortality. We previously identified independent risk factors for the presence of resting hypoxemia in the COPDGene cohort. However, little is known about characteristics that predict onset of resting hypoxemia in patients who are normoxic at baseline. We hypothesized that a combination of clinical, physiologic, and radiographic characteristics would predict development of resting hypoxemia after 5-years of follow-up in participants with moderate to severe COPD METHODS: We analyzed 678 participants with moderate-to-severe COPD recruited into the COPDGene cohort who completed baseline and 5-year follow-up visits and who were normoxic by pulse oximetry at baseline. Development of resting hypoxemia was defined as an oxygen saturation ≤88% on ambient air at rest during follow-up. Demographic and clinical characteristics, lung function, and radiographic indices were analyzed with logistic regression models to identify predictors of the development of hypoxemia.

Results: Forty-six participants (7%) developed resting hypoxemia at follow-up. Enrollment at Denver (OR 8.30, 95%CI 3.05-22.6), lower baseline oxygen saturation (OR 0.70, 95%CI 0.58-0.85), self-reported heart failure (OR 6.92, 95%CI 1.56-30.6), pulmonary artery (PA) enlargement on computed tomography (OR 2.81, 95%CI 1.17-6.74), and prior severe COPD exacerbation (OR 3.31, 95%CI 1.38-7.90) were independently associated with development of resting hypoxemia. Participants who developed hypoxemia had greater decline in 6-min walk distance and greater 5-year decline in quality of life compared to those who remained normoxic at follow-up.

Conclusions: Development of clinically significant hypoxemia over a 5-year span is associated with comorbid heart failure, PA enlargement and severe COPD exacerbation. Further studies are needed to determine if treatments targeting these factors can prevent new onset hypoxemia.

Trial registration: COPDGene is registered at ClinicalTrials.gov: NCT00608764 (Registration Date: January 28, 2008).

Fig. 1

Patient Flow Diagram. Abbreviations: PRISm = Preserved Ratio, Impaired Spirometry

Fig. 2

Distribution of change in oxygen saturation over a 5-year period among COPD cases without hypoxemia at baseline (n = 678)

Fig. 3

Development of hypoxemia is associated with worsened QOL and reduced exercise tolerance. Compared to normoxic participants at Phase 2, participants who developed hypoxemia had a greater rate of change in a SGRQ total score, b SGRQ activity score, c SGRQ impact score, but no difference of change in d SGRQ symptom score or breathlessness measured by e MMRC. Likewise, these participants have a greater decrease in f 6MWD change compared to participants who remained normoxic. Error bars represent standard error. SGRQ = St George’s Respiratory Questionnaire, 6MWD = 6-min walk distance. MMRC = modified medical research council score. *P < 0.05, **P < 0.01, ***P < 0.001