Detrended fluctuation analysis of peak expiratory flow and exacerbation frequency in COPD

We investigated whether exacerbation frequency in chronic obstructive pulmonary disease (COPD) was related to an exponent α which quantifies self-similarity in daily peak expiratory flow (PEF) and is calculated using detrended fluctuation analysis (DFA). We examined data from COPD patients who recorded an increase in respiratory symptoms and post-bronchodilator PEF on daily diary cards. We also investigated PEF data from a double-blind, placebo-controlled trial of the anti-cholinergic agent, tiotropium. In the observational study there were 308 patients with COPD (195 males; mean ± sd age 68.3 ± 8.4 yrs, forced expiratory volume in 1 s (FEV(1)) 1.12 ± 0.46 L, FEV(1) % predicted 44.5 ± 16.4%). The mean ± sd α over the first year was 0.944 ± 0.19 and it was positively related to the frequency of exacerbations per year (p=0.009). In the clinical trial, α was lower in COPD patients randomised to tiotropium, mean ± sd 0.87 ± 0.21 (n=48) than on placebo, mean ± sd 0.95 ± 0.19 (n=52; p=0.035). Power analysis showed that fewer patients would be required for clinical studies with α as the outcome measure than exacerbation frequency. DFA shows that daily PEF in COPD has long-term correlations which are related to exacerbation frequency. Monitoring of PEF and use of α may result in smaller COPD patient sample sizes in trials.