Background: Chronic Obstructive Pulmonary disease (COPD) is an under-diagnosed disease with a prevalence of approximately 10%, highly dependent on age and smoking habits. Comorbidities are common in COPD and of these, cardiovascular diseases (CVD) are the most common. COPD is the fourth leading cause of death globally, and CVD probably contribute to the high mortality. Within CVD, Ischemic Heart Disease (IHD) is the most common. It is highly clinically relevant to identify signs of ischemic heart disease, other cardiac conditions, and risk factors for CVD in COPD. Electrocardiogram (ECG) is a simple but still major diagnostic tool in clinical cardiology, including disturbances in the electric conduction system and ischemia. Due to the under-diagnosis of COPD, there is limited knowledge regarding the prevalence and prognostic impact of ECG abnormalities in COPD. Arterial stiffness is a risk factor for CVD, which has raised an increased interest, however not evaluated in population based studies of COPD.
Aim: The overall aim was to describe cardiovascular aspects on COPD, with a specific focus on arterial stiffness, prevalence and prognostic impact of ischemic ECG abnormalities and prolonged QT interval, by comparing subjects with and without obstructive lung function impairment in a population-based cohort.
Methods: The thesis is based on the Obstructive Lung Disease in Northern Sweden (OLIN) COPD study; a population-based longitudinal cohort study. During the years 2002-2004, all participants in clinical examinations from previously recruited large population-based cohorts were invited to re-examination including spirometry and a structured interview. All subjects with obstructive lung function impairment (n=993) were identified, together with 993 age and sex-matched referents without airway obstruction. The study population (n=1986) has been invited to annual examinations since 2005 including spirometry and structured interview. Papers I-III are based on data from 2005 when electrocardiogram (ECG) was recorded in addition to the basic program. All ECGs were Minnesota coded and QT-time was measured. Paper IV is based data from 2010 when non-invasive measurements of arterial stiffness, assessed as pulse wave velocity (PWV), was added to the program. Spirometric data were classified as normal lung function (NLF), restrictive spirometric pattern (RSP) and airway obstruction (COPD). The following spirometric criteria for COPD were used: post-bronchodilator FEV1/VC<0.70 (papers I-IV, in paper III labelled GOLD-COPD) and lower limit of normal, LLN (LLN-COPD) (paper III). Spirometric classification of COPD severity was based on FEV1 % predicted as a continuous variable or according to the Global Initiative for Obstructive Lung Disease (GOLD), divided into GOLD 1-4.
Results: The prevalence of ischemic heart disease (IHD), both self-reported and assessed as probable and possible ischemic ECG abnormalities (I-ECG) according to the Whitehall criteria, was similar among subjects with NLF and COPD. The prevalence of both self-reported and probable (I-ECG) according to Whitehall increased by GOLD grade. Among those with COPD, self-reported IHD was associated with disease severity, assessed as FEV1 % predicted also after adjustment for age and sex (paper I).
In both COPD and NLF, those with I-ECG had a higher cumulative mortality over 5 years than those without I-ECG (29.6 vs. 10.6%, p<0.001 and 17.1 vs. 6.3 %, p=0.001). When analysed in a multivariate model, the Mortality Risk Ratio (MRR, 95%CI) was increased for subjects with COPD and I-ECG (2.4, 1.5-3.9), and non-significantly so for NLF with I-ECG (1.65, 0.94-2.90), when compared to NLF without I-ECG. When analyzed separately among subjects with COPD, the increased risk for death associated with I-ECG persisted independent of age, sex, BMI-class, smoking habits and disease severity assessed as FEV1 % predicted (1.89, 1.20-2.99). The proportion without reported IHD was high among those with I-ECG; 72.4% in NLF and 67.3% in COPD. The pattern was similar also among them; I-ECG was associated with an increased risk for death in COPD and non-significantly so in NLF (paper II).
Mean corrected QT-time (QTc) and prevalence of QTc prolongation was higher in RSP than NLF but similar in NLF and GOLD-COPD. The prevalence of borderline as well as prolonged QTc increased by GOLD grade (test for trend p=0.012 for both groups). Of those with GOLD-COPD, 52% fulfilled the LLN-criterion (LLN-COPD). When comparing LLN-COPD and NLF, the pattern was similar as when comparing NLF and GOLD-COPD. The cumulative mortality over 5 years was higher among subjects with borderline and prolonged QTc than those with normal QTc in subjects with GOLD-COPD and LLN-COPD but not in NLF and RSP (paper III).
Arterial stiffness, assessed as PWV, was higher in GOLD 3-4 compared to non-COPD (10.52 vs. 9.13 m/s, p=0.042). Reported CVD and age >60 were both associated with significantly higher PWV in COPD as well as in non-COPD. In a multivariate model, GOLD 3-4 remained associated with higher PWV when compared with non-COPD, also when adjusted for sex, age group, smoking habits, blood pressure, reported CVD and pulse rate (paper IV).
Conclusion: In this population-based study, the prevalence of ischemic ECG abnormalities was similar among subjects with normal lung function and COPD, but increased by disease severity among subjects with COPD. Ischemic ECG abnormalities were associated with an increased mortality among subjects with COPD, independent of common confounders and disease severity, also among those without known heart disease. Whilst the prevalence of QTc prolongation was similar in NLF, COPD and LLN-COPD, it was associated with an increased mortality only in the COPD-groups. ECG is a simple non-invasive method and seems to identify findings of prognostic importance among subjects with COPD. Central arterial stiffness, a known risk factor for cardiovascular disease, was increased among subjects with severe and very severe COPD when compared to subjects without COPD independent of common confounders.