Systemic inflammation has been suggested to increase the risk of future acute exacerbations of COPD (AE).

Subjects with COPD were recruited in a stable phase in three Swedish regions. Baseline data included anthropometry, spirometry, venous blood sampling, COPD Assessment Test (CAT), and self-reported data on medication, comorbidities and smoking. Medical records were used to identify AEs during 1 year prior to baseline and during the following 3 years. The outcome was annualised AEs during follow-up (aAE).

Of 572 subjects (59 % female; mean ± SD age 69 ± 8 years, predicted FEV₁ 57 ± 17 %), 167 (29 %) had ?1 AE prior to baseline. After categorising aAEs into three groups (0, >0 - <1, and ?1 yearly AEs), a multivariate ordinal logistic regression model was fitted, including previously described predictors of AEs. After adjustment for medication, prior AE yes/no [odds ratio 3.94; 95 % confidence interval 2.65 ? 5.84], FEV₁ [0.97; 0.96 ? 0.98], CAT score [1.04; 1.02 ? 1.07], underweight versus normal [2.40; 1.01 ? 5.69], and heart failure (HF) [2.32; 1.06 ? 5.08] were associated with aAEs, whereas asthma [1.15; 0.78 ? 1.70] and current smoking [1.22; 0.82 ? 1.81] were not. Blood and plasma inflammatory biomarkers, including eosinophils [1.97; 0.68 ? 5.72], neutrophil-to-lymphocyte ratio [0.99; 0.87 ? 1.14], systemic inflammatory index (platelets x neutrophils / lymphocytes) [1.00; 1.00 ? 1.00], C-reactive protein [0.99; 0.96 ? 1.01], and fibrinogen [1.13; 0.86 ? 1.49] were added to the model, one by one, but none was associated with aAEs or changed the estimates of the basic model.

In conclusion, stable-phase biomarkers of systemic inflammation did not associate with the risk of future AEs.