- Email: [email protected]
Cardiovascular disease in asthma and COPD: A population-based retrospective cross-sectional study
Respiratory Medicine (2012) 106, 249e256
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/rmed
Cardiovascular disease in asthma and COPD: A population-based retrospective cross-sectional study Mario Cazzola a,b,*, Luigino Calzetta a,b, Germano Bettoncelli c, Claudio Cricelli c, Francesco Romeo a, Maria Gabriella Matera d, Paola Rogliani a a
Department of Internal Medicine, University of Rome Tor Vergata, Rome, Italy Department of Pulmonary Rehabilitation, IRCCS, San Raffaele Pisana Hospital, Rome, Italy c Health Search Institute, Italian College of General Practitioners, Florence, Italy d Department of Experimental Medicine, Second University, Naples, Italy b
Received 7 March 2011; accepted 27 July 2011
KEYWORDS Asthma; COPD; Cardiovascular disease; Population-based retrospective crosssectional study
Summary We conducted a large population-based retrospective cross-sectional study for determining the extent of clinically recognized chronic obstructive pulmonary disease (COPD) and asthma, and the prevalence of associated cardiovascular diseases (CVDs), using information obtained from the Health Search Database (HSD) owned by the Italian College of General Practitioners (SIMG). Our study provides further evidence that patients with the diagnosis of COPD are at increased association with the diagnosis of most CVDs. It also documents that age clusters between 35 and 54 years are those at highest association of simultaneous presence of the diagnosis of CVD and that of COPD, with a progressive significant reduction in older age clusters. Moreover, it shows that the diagnosis of asthma is modestly associated with the diagnosis of different CV morbidities. ª 2011 Elsevier Ltd. All rights reserved.
* Corresponding author. Cattedra di Malattie Respiratorie, Dipartimento di Medicina Interna, Universita ` di Roma Tor Vergata, Via Montpellier 1, IT 00133 Roma, Italy. Tel.: þ39 348 6412311; fax: þ39 06 72596621. E-mail address: [email protected] (M. Cazzola). 0954-6111/$ - see front matter ª 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.rmed.2011.07.021
250
Introduction Chronic obstructive pulmonary disease (COPD) and asthma are conditions associated with many comorbidities at the time of diagnosis.1e5 In particular, there is solid evidence that patients with COPD are at increased risk of cardiovascular disease (CVD),3,6e8 whereas the link between asthma and CVD seems to be less strong,5,9e11 and some researchers deny the existence of this link.12,13 In the current study, we determined the extent of clinically recognized COPD and asthma, and the prevalence of associated CVDs using information obtained from the Health Search Database (HSD) owned by the Societa ` Italiana Medici Generici (SIMG), the Italian College of General Practitioners, which stores information on about 1.5% of the total Italian population served by general practitioners (GPs).
Patients and methods Collection of information We conducted a population-based retrospective crosssectional study using information obtained from the HSD. Primary-care data were obtained from the HSD of SIMG. Established in 1998 with the primary aim to carry out observational studies on the incidence and prevalence of chronic diseases and with the input of GPs who accept to participate on a voluntary basis, the HSD contains information provided by more than 650 GPs from all Italian regions. It included a total of 909,638 individuals (429,962 men and 479,676 women) aged >14 years and registered in the HSD at the end of December 2009.14 After intensive training, all participating GPs had to use specially designed software to record data during their normal daily clinical practice (Millewin). The software system codes all the diagnostic records by using the Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM).15 Prescription records are also coded according to the anatomical therapeutic chemical (ATC) classification system. Data are recorded in the HSD with the consent of the patient, lawfully stored, managed according to privacy rules and can be used for scientific studies without any further authorization from an ethics committee. It is well known that before using any data resource, particularly one that is based on computer records, it is necessary to determine the quality and completeness of the available information. Therefore, data are subject to a range of quality checks. Any variations within agreed ranges are investigated and submitted to each participating GP. Physicians who fail to meet standard quality criteria are not considered for epidemiological studies.16 We must highlight that comparative analyses have demonstrated the validity of the information gathered in the HSD.17
Ascertainment of COPD, asthma and comorbidities Cases of COPD were identified on the basis of the ICD-9 codes 491, 492, and 496, and those of asthma on the basis of the ICD-9 code 493. The number of patients classified on
M. Cazzola et al. the basis of the ICD-9 code 493.2 was really small and for this reason we have not considered it separately also for avoiding a statistical bias. Unfortunately we have not information related to temporality in HSD. Always using the ICD-9 codes, we regarded as cardiovascular comorbidities cases of hypertensive disease (401e405), ischemic heart disease (410e414), diseases of pulmonary circulation (415e417), other forms of heart disease (420e429), cardiac arrhythmia (427), heart failure (428), cerebrovascular diseases (430e438).
Analysis The study was carried out by a cross-sectional and observational field method. It permitted to describe the situation at the time of data collection, at the end of December 2009, and it allowed determining the prevalence of investigated outcomes. The studied population was divided by several clusters: age (35e44, 45e54, 55e64, 65e74, 75e84 and older than 85 years old), gender (male and female), and comorbidities. We evaluated the odds ratio (OR) that was chosen instead of relative risk (RR) because asthma and COPD have a prevalence that is less than 10%.18,19 Positive and negative OR values mean a co-relative positive and negative association between evaluated variables. Data were adjusted for sex and age by Mantel-Haenszel method, that permits a precision-based estimates of the OR. Furthermore, if the prevalence of comorbidities were more than 10%, and the OR was more than 2.5 or less than 0.5, correction of the OR was carried out.18,19 In order to calculate statistical and epidemiological values, computer software GraphPad Prism (CA, USA), SPSS (Chicago, IL, USA) and OpenEpi (Dean AG, Sullivan KM, Soe MM. Open Source Epidemiologic Statistics for Public Health) were used.
Results We excluded the age cluster 15e34 years because in this population the risk of COPD was not relevant.14 We identified a population of 690,489 individuals aged >34 years and registered in the HSD at the end of December 2009. There were 39,741 (5.76% of the entire population) individuals suffering from asthma and 25,281 (3.66% of the entire population) patients with a diagnosis of COPD. The prevalence of asthma was greater in women than in men with a men/women ratio of 0.71, whereas the prevalence of COPD was greater in men than in women with a men/ women ratio of 1.64, and increased with age in both men and women.
Association between asthma, COPD and CV morbidities Table 1 shows the demographic characteristics of the study population by age distribution, asthma, COPD and CV morbidities. All individuals that were not identified on the basis of the ICD-9 code 493 were considered “non-asthmatics”, whereas all subjects that were not identified on the basis of the ICD-9 codes 491, 492, and 496 were considered “non-COPD patients”.
A population-based retrospective cross-sectional study
251
Table 1 Demographic characteristics of study population by age distribution, the diagnosis of asthma, COPD and CV morbidities. A: number of subjects, B: prevalence (%). A
General population (total)
General population (non-asthmatic and nonCOPD patients)
Asthma
COPD
35e44 45e54 55e64 65e74 75e84 >85 Total
167,191 156,279 135,376 113,465 84,196 33,982 690,489
156,284 144,866 123,471 99,742 71,589 29,515 625,467
10,123 9434 7709 6568 4506 1401 39,741
784 1979 4196 7155 8101 3066 25,281
B
General population (total)
General population (non-asthmatic and non-COPD patients)
Asthma
COPD
General population (total)
General population (non-asthmatic and non-COPD patients)
Asthma
COPD
0.06 0.59 1.28 2.51 4.37 4.28
0.13 1.67 3.55 5.37 6.01 5.25
35e44 45e54 55e64 65e74 75e84 >85
Angina and coronary disease 0.12 0.11 0.65 0.61 2.40 2.24 4.63 4.24 7.24 6.54 8.17 7.46
0.13 0.91 2.97 5.91 9.92 11.92
0.89 2.27 6.22 8.93 11.95 13.27
Acute or old myocardial infarction 0.11 0.11 0.56 0.54 1.69 1.66 2.88 2.73 3.96 3.70 3.92 3.76
35e44 45e54 55e64 65e74 75e84 >85
Cardiac arrhythmia 2.12 3.32 5.11 9.35 15.64 17.38
2.07 3.18 4.85 8.79 14.68 16.32
2.88 5.12 7.64 12.85 20.06 24.13
2.17 4.70 7.96 13.92 21.64 24.49
Heart failure 0.04 0.03 0.14 0.12 0.51 0.41 1.49 1.17 4.38 3.58 8.66 7.41
0.08 0.29 1.02 2.57 7.15 15.92
0.38 0.91 2.62 5.02 9.91 17.32
35e44 45e54 55e64 65e74 75e84 >85
Hypertension disease 6.78 6.58 19.96 19.37 39.54 38.59 54.87 53.79 62.73 61.92 58.63 57.56
9.52 26.08 48.88 63.11 69.09 68.81
12.24 33.80 50.45 62.35 66.30 64.22
Cerebrovascular disease 0.09 0.09 0.33 0.31 1.07 1.03 3.04 2.92 6.96 6.91 10.11 10.29
0.13 0.49 1.60 3.90 9.21 13.49
0.26 0.71 1.43 3.80 6.17 6.88
35e44 45e54 55e64 65e74 75e84 >85
Other heart disease 0.06 0.12 0.18 0.34 0.50 0.62
0.14 0.24 0.29 0.58 0.84 0.50
0.38 0.35 0.38 0.80 0.96 0.85
0.06 0.13 0.25 0.53 1.11 1.28
0.13 0.20 0.36 0.55 1.10 0.91
0.05 0.10 0.17 0.29 0.42 0.60
Table 2 shows the data of CV morbidities (acute or old myocardial infarction, angina and coronary disease, cardiac arrhythmia, heart failure, cerebrovascular disease, other heart disease, or pulmonary embolism) with their prevalences in the general population and in patients with the diagnosis of asthma or COPD. We also show the data of hypertensive disease with its prevalence in the general population and in patients with the diagnosis of asthma or COPD. The diagnosis of COPD was significantly associated with an increased likelihood of diagnosis of CVD and hypertensive disease (Fig. 1B). On the contrary, although CVD and hypertensive disease were more prevalent in patients with diagnosis of asthma than in the general population (Table
Pulmonary embolism 0.04 0.04 0.07 0.07 0.13 0.12 0.28 0.24 0.70 0.63 0.82 0.78
2), apparently asthma had a weak association with these diseases (Fig. 1A). Intriguingly, the adjusted OR for acute or old myocardial infarction was 0.98 (95% CI 0.90e1.07) in patients with the diagnosis of asthma and 2.91 (95% CI 2.74e3.09) in those with the diagnosis of COPD.
Impact of gender on CVD in asthma and COPD CV morbidities in the general population affected slightly more women than men, with an adjusted men/women ratio of 0.98. The ration was balanced in COPD (1.00), but it decreased in asthma (0.95) indicating a higher prevalence
252
M. Cazzola et al.
Table 2 Demographic characteristics of study population by prevalence of CV morbidities in the general population, and in patients with the diagnosis of asthma or COPD (%).
Angina and coronary disease Acute or old myocardial infarction Cardiac arrhythmia Heart failure Hypertension disease Cerebrovascular disease Other heart disease Pulmonary embolism
General population (total)
General population (non-asthmatic and non-COPD patients)
Asthma
Total
Total
Total
Men
Women
Men
Women
COPD
Men
Women
Total
Men
Women
2.69
3.46
2.03
2.39
3.07
1.80
3.35
4.10
2.89
9.20
10.42
7.48
1.63
2.56
0.83
1.52
2.38
0.77
1.47
2.47
0.85
4.81
6.30
2.69
6.56 1.35 33.46 2.15 0.22 0.22
6.29 1.32 31.78 1.91 0.23 0.20
6.80 1.37 34.92 2.37 0.22 0.24
6.06 1.06 32.13 2.04 0.19 0.20
5.72 0.98 30.33 1.95 0.19 0.18
6.36 1.13 33.68 2.12 0.20 0.22
8.68 2.08 38.79 2.62 0.36 0.35
8.33 2.03 36.02 2.25 0.37 0.32
8.90 2.12 40.49 2.85 0.35 0.37
15.60 7.21 58.08 4.19 0.74 0.70
15.53 7.13 55.79 0.76 0.76 0.61
15.71 7.33 61.33 9.03 0.71 0.82
of CVD in asthmatic women. There was always a prevalence of hypertensive disease in women, with a greater impact in COPD (men/women ratio: 0.91 in the general population, 0.91 in COPD, and 0.89 in asthma, respectively). Among patients with the diagnosis of asthma, there was not difference in gender for association with all CVDs, excluded for angina and coronary disease, for which the OR was higher in women compared to men (Fig. 1A). Among patients with the diagnosis of COPD, there were not differences in gender for association with all CVDs, excluded for acute or old myocardial infarction and angina and coronary disease, for which the OR was higher in women compared to men (Fig. 1B).
Impact of age on CVDs in patients with the diagnosis of asthma or COPD Both in patients with the diagnosis of asthma and in those with the diagnosis of COPD, an increase in age resulted in a progressive increase in the prevalence of the diagnosis of CVD and hypertensive disease. However, while the different age clusters did not affect differently the CV morbidities in asthma, clusters between 35 and 54 years were those at highest correlation with the simultaneous presence of the diagnoses of CV morbidities and COPD, with a progressive significant reduction in older age clusters. Intriguingly, the age cluster 35e44 years caused the highest ORs for angina and coronary disease, heart failure, and cerebrovascular disease, and that 45e54 years caused the highest correlation with acute or old myocardial infarction. Age did not influence differently the association with cardiac arrhythmia, although there was a decrease of adjusted OR in the age cluster 35e44. Fig. 2 illustrates the ORs of diagnoses of CVD and hypertensive both in patients with the diagnosis of asthma and in those with the diagnosis of COPD according to the different age clusters.
Impact of smoking habits on CVD in patients with the diagnosis of asthma or COPD Table 3 shows the demographic characteristics of the study population by smoking status (never smokers, current-
smokers and former smokers) and by the diagnosis of asthma and COPD. As expected, there were differences of smoking habit between general population and COPD, on the contrary the prevalence of smoking habits in asthmatics was analogous to that in general population. In particular, smoking prevalence was higher in COPD compared to asthmatics (COPD/asthma ratio: 1.45, current-smokers; 1.63, former smokers), whereas never smokers were the majority in asthmatics (COPD/asthma ratio: 0.57, never smokers). In patients with the diagnosis of asthma, the smoking habit did not modify the prevalence of CVD, compared to the general population. On the contrary, in patients with the diagnosis of COPD the overall prevalence of CVD was more than double in current and former smokers and it increased by more than five-fold in never smokers, compared to the general population (Table 4). The analysis of the putative interference of smoking habit on the association with CVD showed that the smoking habit represented neither a bias nor a confounding factor. In fact, the diagnosis of COPD increased by 2.6 folds and significantly (P < 0.05) the overall correlation with CVD in never-smokers compared to current and former smokers, versus the general population. On the contrary, the diagnosis of asthma did not modify the OR of CVD by clustering for smoking habit (Table 5).
Discussion This large observational study is the first, to the best of our knowledge, that has explored the difference in prevalence of the diagnosis of CVD between patients with and without diagnoses of COPD or asthma performing stratified analyses by age or gender. It documents that the diagnosis of asthma is modestly correlated with the diagnosis of different CV morbidities and provides further evidence that patients with the diagnosis of COPD are highly associated with the diagnoses of CVDs. Moreover, it suggests that the diagnosis of COPD is strongly associated with that of CVD, regardless the smoking habit. We believe that the additional information to the possible, although weak, association between the diagnoses of asthma and CVD is important because, as already
A population-based retrospective cross-sectional study
Figure 1 Adjusted odds ratios and 95% confidence limits of CV morbidities in patients with the diagnosis of asthma (A) and COPD (B). "Total” adjusted for age and sex, “Men” and “Women” adjusted for age. p < 0.05 vs non-asthmatic (A) and vs non-COPD (B) population.
mentioned, literature on this topic is conflicting.5,8e13 Our data confirm a higher prevalence of the diagnosis of CVD in women with that of asthma, show that, at least in men, the diagnosis of asthma is associated with diagnosis of angina and coronary disease but not with that of acute or old myocardial infarction, and do not confirm the reported association between the diagnoses of asthma and cerebrovascular disease.13 We must admit that we are unable to explain why there was an association of the diagnosis of asthma with that of angina and coronary disease but not with that of myocardial infarction in men, although the prevalence of the diagnosis of angina and coronary disease and that of myocardial infarction in asthmatic men were similar to those in men in the non-asthmatic population. It is likely that the low number of events could have generated information that larger numbers could deny, although again we must emphasize the breadth of our sample. In any
253 case, we cannot exclude that the observed increased ORs for the diagnoses of cardiac arrhythmia and heart failure, as well as other heart diseases, may suggest that the consequences of those diseases on the lung could have been confounded with asthma. However, we must stress that we have only considered comorbidities that were present in the same patient simultaneously with the diagnosis of asthma. Another finding that is difficult to explain is the impact of gender in the association between the diagnoses of asthma or COPD and CVD. Several reports suggest a role for female sex hormones in promoting the asthmatic phenotype.20 On the other hand, female smokers appear to be more susceptible than male smokers to developing COPD. Women have a more vigorous inflammatory reaction and generate more oxidative stress in the airways than men21 likely because estrogens could cause an imbalance in the formation and removal of toxic intermediates of cigarette smoke components.22 All this would suggest a stronger association between the diagnoses of asthma or COPD and CVD in women, but often our data did not confirm it. Soriano and colleagues1 suggested that the occurrence of CVD events in their study was generally lower in asthma patients compared to COPD patients likely because of the younger age distribution. Also in our population the diagnosis of asthma was prevalent in younger people and this could indicate that the age is the main association factor of CVD in asthma or COPD. However, although an increase in age resulted in a progressive increase in the prevalence of the diagnoses of CVD and hypertensive disease both in the diagnoses of asthma and COPD, the various age clusters did not affect differently the association of the diagnosis of CV morbidities with that of asthma, whereas cluster 3544 years was that at highest correlation with the simultaneous presence of CV morbidities and COPD, with a progressive significant reduction in older age clusters. Also Feary and colleagues found that the association between diagnoses of COPD and CVD was strongest in the younger age groups, persisted across all ages but reduced as age increased.23 It is also difficult to explain the higher association of the diagnosis of CVDs with the diagnosis of COPD among younger than older patients that we have found in our study. It has been suggested that it may be mediated by the use of inhaled bronchodilators24 because these agents were shown to impose a larger relative risk in people with unrecognized CVD than in people with known disorders.25 We do not agree with this hypothesis because our data show that the adjusted ORs of cardiac arrhythmia, a common side effect of bronchodilators, were almost constant in all age clusters. We believe that our results are interesting because they come from a large, validated, population-based database that illustrates what can be observed in real life and can be considered representative of the general population. This is the major strength of the present study. Moreover, use of routine medical records should have eliminated differential misclassification due to recall or interviewer bias. The main limitations of our study are common to retrospective database analyses. These limitations might be associated mainly with bias towards the selection of more highly motivated, or better organized, practices. Consequently, our findings might likely be an overestimate of the
254
M. Cazzola et al.
Figure 2 Adjusted odds ratios and 95% confidence limits of CV morbidities in patients with the diagnosis of asthma or COPD in different strata of age in years. p < 0.05 vs comorbidities in non-asthmatic and vs non-COPD population. Values adjusted for sex.
Table 3 Demographic characteristics of study population by smoking habits, and the diagnosis of asthma or COPD (%).
Never smokers Current-smokers Former smokers
General population
Asthma
COPD
53.62 25.51 20.88
54.98 25.19 19.83
31.23 36.46 32.32
general picture on the one hand. Nonetheless, on the other hand, it is well known that patients vary in their use of health care and it is accepted that a significant proportion of patients who are ill may not consult their doctor and this can lead to underestimation of the problem. Several reports have highlighted that diagnosed prevalence and morbidity data underestimate the total burden of COPD because the disease is usually not diagnosed until it is clinically apparent, and there is considerable variation in
reported prevalence.26e31 This is likely due to a lack of awareness about COPD by patients. It is only 25e50% of COPD patients are known to their doctor.27 Patients may well experience symptoms, but do not always present these symptoms to their family physician.27 Poor recognition of the disease may lead patients to believe that their respiratory symptoms are nothing more than a smoker’s cough, rather than an early sign of disease pathology, causing them to delay contact with a healthcare professional.26 In any case, the majority of patients with a decreased FEV1 do not complain of any bronchial symptoms.32 Nonetheless, the underestimation of the actual burden of COPD may also be due to the underdiagnosis of the disease by healthcare professionals, particularly in the early stages of the disease,33 mainly because in family practice the diagnostic process is primarily based on symptoms and signs presented by the patient and, consequently, early-stage COPD can easily be veiled by signs of asthma.34 Also a lack of awareness among physicians of
Table 4 Influence of smoking status on CV morbidities in the diagnosis of asthma or COPD related to the general population (values are expressed as rates). Asthma vs General population
Ischemic cardiopathy Myocardial infarction Cardiac arrhythmia Heart Failure Hypertension Cerebrovascular disease Other heart disease Pulmonary embolism
COPD vs General population
Neversmokers
Current smokers
Former smokers
Neversmokers
Current smokers
Former smokers
1.21 0.88 1.29 1.51 1.13 1.19 1.56 1.53
1.26 0.91 1.34 1.57 1.17 1.23 1.61 1.59
1.31 0.95 1.39 1.63 1.22 1.28 1.68 1.65
5.87 5.05 4.08 9.20 2.98 5.41 5.67 5.33
2.39 2.06 1.66 3.75 1.21 2.20 2.31 2.17
2.21 1.90 1.54 3.46 1.12 2.04 2.13 2.00
A population-based retrospective cross-sectional study Table 5 Adjusted odds ratios and 95% confidence limits of overall CV morbidities in never-smoker population with the diagnosis of asthma or COPD related to current and former smokers. Data are adjusted for CV morbidities. Asthma Never-smokers/Current smokers vs general population Never-smokers/Former smokers vs general population
COPD
0.96 2.45 (0.80e1.13) (1.58e3.33)* 0.93 2.66 (0.76e1.10) (1.80e3.51)*
255 asthma. It is clear, therefore, that there is a strong need to look beyond the lungs in treating patients with asthma or COPD. In any case, we must highlight that since the GPs usually state diagnosis of COPD, asthma or CVD by using ICD codes and a diagnosis is not true disease identification, the relationships here presented have been described as relationship between diagnoses, and not between diseases.
Conflict of interest None declared.
*p < 0.05 vs general population.
Acknowledgments recommended treatment practices can justify the underdiagnosis of COPD.33,34 In any case, underestimation of the actual burden of COPD is an important issue because it may cause a bias when trying to study the real relationships between diseases, a bias that register studies cannot control and correct for. We must highlight that GPs are faced with an array of alternative codes for the same or similar conditions and, in effect, practicing physicians have little faith in the accuracy of the ICD-9 diagnoses entered for prescription purposes and this can question the real value of diagnoses in the HSD. Therefore, there may have been misclassification when our GPs have used specific diagnostic categories and this likely because there are no standardized methods of applying diagnostic labels in general practice. In particular, a gender bias in the diagnosis of COPD and asthma is to be expected when the epidemiology of a disease makes a dramatic shift in terms of maleefemale prevalence. Physicians are at risk for assuming COPD in men and asthma in women.14 We did not use the lung function data for our analyses because only one third of patients with COPD and less than 20% of those with asthma had spirometry records. This finding fits perfectly with the results of the National Committee for Quality Assurance (NCQA) study that suggested that approximately 32% of a broad range of patients with a new COPD diagnosis had undergone spirometry within the previous 2 years to 6 months following diagnosis.35 In addition, spirometric testing appeared to decrease with increasing age. A study that investigated the degree of use of spirometry to establish the diagnosis of COPD in Italy documented that many GPs do not use spirometry to establish the diagnosis of COPD because they think that spirometry is not necessary for the diagnosis of COPD or there are logistical limitations to the access of the patients to lung function laboratories.36 Regardless of anything else, we must mention that in the analysis of Feary and colleagues that used data from primary care to investigate the prevalence of major comorbidities in subjects with COPD, no clinically important difference was seen in the results after excluding patients with COPD but without spirometry records from the analysis.37 In conclusion, our study clearly indicates that the diagnosis of CVD is strongly associated with the diagnosis of COPD and, albeit in a less strong manner, even with that of
This study is based on data from the Health Search database owned by the Italian College of General Practitioners. However, the interpretation and conclusions contained in this study are those of the authors alone. The original idea for the study were formulated by Prof. Cazzola, with subsequent advice from Drs. Bettoncelli and Cricelli, and statistical input from Dr. Calzetta, who also performed the statistical analyses. Prof. Cazzola drafted the manuscript, with extensive contributions and revision from all authors. Access to the Health Search database was offered free by the Italian College of General Practitioners.
References 1. Soriano JB, Visick GT, Muellerova H, Payvandi N, Hansell AL. Patterns of comorbidities in newly diagnosed COPD and asthma in primary care. Chest 2005;128:2099e107. 2. Fabbri LM, Luppi F, Beghe ´ B, Rabe KF. Complex chronic comorbidities of COPD. Eur Respir J 2008;31:204e12. 3. Cazzola M, Bettoncelli G, Sessa E, Cricelli C, Biscione G. Prevalence of comorbidities in patients with chronic obstructive pulmonary disease. Respiration 2010;80:112e9. 4. Gershon AS, Wang C, Guan J, To T. Burden of comorbidity in individuals with asthma. Thorax 2010;65:612e8. 5. Cazzola M, Calzetta L, Bettoncelli G, Cricelli C, Novelli L, Rogliani P. Asthma and comorbid medical illness. Eur Respir J 2011;38:42e9. 6. Mannino DM, Thorn D, Swensen A, Holguin F. Prevalence outcomes of diabetes, hypertension, and cardiovascular disease in COPD. Eur Respir J 2008;32:962e9. 7. Rutten FH, Cramer MJ, Lammers JW, Grobbee DE, Hoes AW. Heart failure and chronic obstructive pulmonary disease: an ignored combination? Eur J Heart Fail 2006;8:706e11. 8. Soriano JB, Rigo F, Guerrero D, Yan ˜ez A, Forteza JF, Frontera G, et al. High prevalence of undiagnosed airflow limitation in patients with cardiovascular disease. Chest 2010; 137:333e40. 9. Appleton S, Ruffin S, Wilson D, Taylor A, Adams R. Asthma is associated with cardiovascular disease in a representative population sample. Obes Res Clin Pract 2008;2:91e9. 10. Iribarren C, Tolstykh IV, Eisner MD. Are patients with asthma at increased risk of coronary heart disease? Int J Epidemiol 2004; 33:743e8. 11. Onufrak SJ, Abramson JL, Austin HD, Holguin F, McClellan WM, Vaccarino LV. Relation of adult-onset asthma to coronary heart disease and stroke. Am J Cardiol 2008;101:1247e52. 12. Enright PL, Ward BJ, Tracy RP, Lasser EC. Asthma and its association with cardiovascular disease in the elderly. The
256
13.
14.
15.
16.
17.
18.
19. 20.
21.
22.
23.
24.
M. Cazzola et al. Cardiovascular Health Study Research Group. J Asthma 1996; 33:45e53. Schanen JG, Iribarren C, Shahar E, Punjabi NM, Rich SS, Sorlie PD, et al. Asthma and incident cardiovascular disease: the Atherosclerosis risk in communities study. Thorax 2005;60: 633e8. Cazzola M, Puxeddu E, Bettoncelli G, Novelli L, Segreti A, Cricelli C, et al. The prevalence of asthma and COPD in Italy: a practice-based study. Respir Med 2011;105: 386e91. World Health Organization. International classification of diseases. Manual of the international statistical classification of diseases: injuries and causes of deaths (9th revision). Geneva: World Health Organization; 1977. Lawrenson R, Williams T, Farmer R. Clinical information for research: the use of general practice databases. J Public Health Med 1999;21:299e304. Filippi A, Bignamini AA, Sessa E, Samani F, Mazzaglia G. Secondary prevention of stroke in Italy: a cross-sectional survey in family practice. Stroke 2003;34:1010e4. Zhang J, Yu KF. What’s the relative risk? A method of correcting the odds ratio in cohort studies of common outcomes. JAMA 1998;280:1690e1. Schmidt OC, Kohlmann T. When to use the odds ratio or the relative risk? Int J Public Health 2008;53:165e7. Card JW, Zeldin DC. Hormonal influences on lung function and response to environmental agents: lessons from animal models of respiratory disease. Proc Am Thorac Soc 2009;6: 588e95. Ben-Zaken Cohen S, Pare ´ PD, Man SFP, Sin DD. The growing burden of chronic obstructive pulmonary disease and lung cancer in women. Examining sex differences in cigarette smoke metabolism. Am J Respir Crit Care Med 2007;176: 113e20. Sin DD, Cohen SB, Day A, Coxson H, Pare PD. Understanding the biological differences in susceptibility to chronic obstructive pulmonary disease between men and women. Proc Am Thorac Soc 2007;4:671e4. Johnston AK, Mannino DM, Hagan GW, Davis KJ, Kiri VA. Relationship between lung function impairment and incidence or recurrence of cardiovascular events in a middle-aged cohort. Thorax 2008;63:599e605. Macie C, Wooldrage K, Manfreda J, Anthonisen N. Cardiovascular morbidity and the use of inhaled bronchodilators. Int J Chron Obstruct Pulmon Dis 2008;3:163e9.
25. Camp PG, Goring SM. Gender and the diagnosis, management, and surveillance of chronic obstructive pulmonary disease. Proc Am Thorac Soc 2007;4:686e91. 26. Britton M. The burden of COPD in the U.K.: results from the Confronting COPD survey. Respir Med 2003;97(Suppl C):S71e9. 27. van Schayck CP, Chavannes NH. Detection of asthma and chronic obstructive pulmonary disease in primary care. Eur Respir J 2003;39(Suppl:1):6se22s. 28. Mohangoo AD, van der Linden MW, Schellevis FG, Raat H. Prevalence estimates of asthma or COPD from a health interview survey and from general practitioner registration: what’s the difference? Eur J Public Health 2006;16:101e5. 29. Lindberg A, Bjerg A, Ro ¨nmark E, Larsson LG, Lundba ¨ck B. Prevalence and underdiagnosis of COPD by disease severity and the attributable fraction of smoking report from the obstructive lung disease in Northern Sweden studies. Respir Med 2006; 100:264e72. 30. Vandevoorde J, Verbanck S, Gijssels L, Schuermans D, Devroey D, De Backer J, Kartounian J, Vincken W. Early detection of COPD: a case finding study in general practice. Respir Med 2007;101:525e30. 31. Tinkelman DG, Price D, Nordyke RJ, Halbert RJ. COPD screening efforts in primary care: what is the yield? Prim Care Respir J 2007;16:41e8. 32. White P. Prevalence of COPD in primary care: no room for complacency. Fam Pract 2009;26:1e2. 33. Nacul L, Soljak M, Samarasundera E, Hopkinson NS, Lacerda E, Indulkar T, et al. COPD in England: a comparison of expected, model-based prevalence and observed prevalence from general practice data. J Public Health 2011;33:108e16. 34. Albers M, Schermer T, Molema J, Kloek C, Akkermans R, Heijdra Y, et al. Do family physicians’ records fit guideline diagnosed COPD? Fam Pract 2009;26:81e7. 35. Han MK, Kim MG, Mardon R, Renner P, Sullivan S, Diette GB, et al. Spirometry utilization for COPD: how do we measure up? Chest 2007;132:403e9. 36. Caramori G, Bettoncelli G, Tosatto R, Arpinelli F, Visona ` G, Invernizzi G, et al. Underuse of spirometry by general practitioners for the diagnosis of COPD in Italy. Monaldi Arch Chest Dis 2005;63:6e12. 37. Feary JR, Rodrigues LC, Smith CJ, Hubbard RB, Gibson JE. Prevalence of major comorbidities in subjects with COPD and incidence of myocardial infarction and stroke: a comprehensive analysis using data from primary care. Thorax 2010;65: 956e62.
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/rmed
Cardiovascular disease in asthma and COPD: A population-based retrospective cross-sectional study Mario Cazzola a,b,*, Luigino Calzetta a,b, Germano Bettoncelli c, Claudio Cricelli c, Francesco Romeo a, Maria Gabriella Matera d, Paola Rogliani a a
Department of Internal Medicine, University of Rome Tor Vergata, Rome, Italy Department of Pulmonary Rehabilitation, IRCCS, San Raffaele Pisana Hospital, Rome, Italy c Health Search Institute, Italian College of General Practitioners, Florence, Italy d Department of Experimental Medicine, Second University, Naples, Italy b
Received 7 March 2011; accepted 27 July 2011
KEYWORDS Asthma; COPD; Cardiovascular disease; Population-based retrospective crosssectional study
Summary We conducted a large population-based retrospective cross-sectional study for determining the extent of clinically recognized chronic obstructive pulmonary disease (COPD) and asthma, and the prevalence of associated cardiovascular diseases (CVDs), using information obtained from the Health Search Database (HSD) owned by the Italian College of General Practitioners (SIMG). Our study provides further evidence that patients with the diagnosis of COPD are at increased association with the diagnosis of most CVDs. It also documents that age clusters between 35 and 54 years are those at highest association of simultaneous presence of the diagnosis of CVD and that of COPD, with a progressive significant reduction in older age clusters. Moreover, it shows that the diagnosis of asthma is modestly associated with the diagnosis of different CV morbidities. ª 2011 Elsevier Ltd. All rights reserved.
* Corresponding author. Cattedra di Malattie Respiratorie, Dipartimento di Medicina Interna, Universita ` di Roma Tor Vergata, Via Montpellier 1, IT 00133 Roma, Italy. Tel.: þ39 348 6412311; fax: þ39 06 72596621. E-mail address: [email protected] (M. Cazzola). 0954-6111/$ - see front matter ª 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.rmed.2011.07.021
250
Introduction Chronic obstructive pulmonary disease (COPD) and asthma are conditions associated with many comorbidities at the time of diagnosis.1e5 In particular, there is solid evidence that patients with COPD are at increased risk of cardiovascular disease (CVD),3,6e8 whereas the link between asthma and CVD seems to be less strong,5,9e11 and some researchers deny the existence of this link.12,13 In the current study, we determined the extent of clinically recognized COPD and asthma, and the prevalence of associated CVDs using information obtained from the Health Search Database (HSD) owned by the Societa ` Italiana Medici Generici (SIMG), the Italian College of General Practitioners, which stores information on about 1.5% of the total Italian population served by general practitioners (GPs).
Patients and methods Collection of information We conducted a population-based retrospective crosssectional study using information obtained from the HSD. Primary-care data were obtained from the HSD of SIMG. Established in 1998 with the primary aim to carry out observational studies on the incidence and prevalence of chronic diseases and with the input of GPs who accept to participate on a voluntary basis, the HSD contains information provided by more than 650 GPs from all Italian regions. It included a total of 909,638 individuals (429,962 men and 479,676 women) aged >14 years and registered in the HSD at the end of December 2009.14 After intensive training, all participating GPs had to use specially designed software to record data during their normal daily clinical practice (Millewin). The software system codes all the diagnostic records by using the Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM).15 Prescription records are also coded according to the anatomical therapeutic chemical (ATC) classification system. Data are recorded in the HSD with the consent of the patient, lawfully stored, managed according to privacy rules and can be used for scientific studies without any further authorization from an ethics committee. It is well known that before using any data resource, particularly one that is based on computer records, it is necessary to determine the quality and completeness of the available information. Therefore, data are subject to a range of quality checks. Any variations within agreed ranges are investigated and submitted to each participating GP. Physicians who fail to meet standard quality criteria are not considered for epidemiological studies.16 We must highlight that comparative analyses have demonstrated the validity of the information gathered in the HSD.17
Ascertainment of COPD, asthma and comorbidities Cases of COPD were identified on the basis of the ICD-9 codes 491, 492, and 496, and those of asthma on the basis of the ICD-9 code 493. The number of patients classified on
M. Cazzola et al. the basis of the ICD-9 code 493.2 was really small and for this reason we have not considered it separately also for avoiding a statistical bias. Unfortunately we have not information related to temporality in HSD. Always using the ICD-9 codes, we regarded as cardiovascular comorbidities cases of hypertensive disease (401e405), ischemic heart disease (410e414), diseases of pulmonary circulation (415e417), other forms of heart disease (420e429), cardiac arrhythmia (427), heart failure (428), cerebrovascular diseases (430e438).
Analysis The study was carried out by a cross-sectional and observational field method. It permitted to describe the situation at the time of data collection, at the end of December 2009, and it allowed determining the prevalence of investigated outcomes. The studied population was divided by several clusters: age (35e44, 45e54, 55e64, 65e74, 75e84 and older than 85 years old), gender (male and female), and comorbidities. We evaluated the odds ratio (OR) that was chosen instead of relative risk (RR) because asthma and COPD have a prevalence that is less than 10%.18,19 Positive and negative OR values mean a co-relative positive and negative association between evaluated variables. Data were adjusted for sex and age by Mantel-Haenszel method, that permits a precision-based estimates of the OR. Furthermore, if the prevalence of comorbidities were more than 10%, and the OR was more than 2.5 or less than 0.5, correction of the OR was carried out.18,19 In order to calculate statistical and epidemiological values, computer software GraphPad Prism (CA, USA), SPSS (Chicago, IL, USA) and OpenEpi (Dean AG, Sullivan KM, Soe MM. Open Source Epidemiologic Statistics for Public Health) were used.
Results We excluded the age cluster 15e34 years because in this population the risk of COPD was not relevant.14 We identified a population of 690,489 individuals aged >34 years and registered in the HSD at the end of December 2009. There were 39,741 (5.76% of the entire population) individuals suffering from asthma and 25,281 (3.66% of the entire population) patients with a diagnosis of COPD. The prevalence of asthma was greater in women than in men with a men/women ratio of 0.71, whereas the prevalence of COPD was greater in men than in women with a men/ women ratio of 1.64, and increased with age in both men and women.
Association between asthma, COPD and CV morbidities Table 1 shows the demographic characteristics of the study population by age distribution, asthma, COPD and CV morbidities. All individuals that were not identified on the basis of the ICD-9 code 493 were considered “non-asthmatics”, whereas all subjects that were not identified on the basis of the ICD-9 codes 491, 492, and 496 were considered “non-COPD patients”.
A population-based retrospective cross-sectional study
251
Table 1 Demographic characteristics of study population by age distribution, the diagnosis of asthma, COPD and CV morbidities. A: number of subjects, B: prevalence (%). A
General population (total)
General population (non-asthmatic and nonCOPD patients)
Asthma
COPD
35e44 45e54 55e64 65e74 75e84 >85 Total
167,191 156,279 135,376 113,465 84,196 33,982 690,489
156,284 144,866 123,471 99,742 71,589 29,515 625,467
10,123 9434 7709 6568 4506 1401 39,741
784 1979 4196 7155 8101 3066 25,281
B
General population (total)
General population (non-asthmatic and non-COPD patients)
Asthma
COPD
General population (total)
General population (non-asthmatic and non-COPD patients)
Asthma
COPD
0.06 0.59 1.28 2.51 4.37 4.28
0.13 1.67 3.55 5.37 6.01 5.25
35e44 45e54 55e64 65e74 75e84 >85
Angina and coronary disease 0.12 0.11 0.65 0.61 2.40 2.24 4.63 4.24 7.24 6.54 8.17 7.46
0.13 0.91 2.97 5.91 9.92 11.92
0.89 2.27 6.22 8.93 11.95 13.27
Acute or old myocardial infarction 0.11 0.11 0.56 0.54 1.69 1.66 2.88 2.73 3.96 3.70 3.92 3.76
35e44 45e54 55e64 65e74 75e84 >85
Cardiac arrhythmia 2.12 3.32 5.11 9.35 15.64 17.38
2.07 3.18 4.85 8.79 14.68 16.32
2.88 5.12 7.64 12.85 20.06 24.13
2.17 4.70 7.96 13.92 21.64 24.49
Heart failure 0.04 0.03 0.14 0.12 0.51 0.41 1.49 1.17 4.38 3.58 8.66 7.41
0.08 0.29 1.02 2.57 7.15 15.92
0.38 0.91 2.62 5.02 9.91 17.32
35e44 45e54 55e64 65e74 75e84 >85
Hypertension disease 6.78 6.58 19.96 19.37 39.54 38.59 54.87 53.79 62.73 61.92 58.63 57.56
9.52 26.08 48.88 63.11 69.09 68.81
12.24 33.80 50.45 62.35 66.30 64.22
Cerebrovascular disease 0.09 0.09 0.33 0.31 1.07 1.03 3.04 2.92 6.96 6.91 10.11 10.29
0.13 0.49 1.60 3.90 9.21 13.49
0.26 0.71 1.43 3.80 6.17 6.88
35e44 45e54 55e64 65e74 75e84 >85
Other heart disease 0.06 0.12 0.18 0.34 0.50 0.62
0.14 0.24 0.29 0.58 0.84 0.50
0.38 0.35 0.38 0.80 0.96 0.85
0.06 0.13 0.25 0.53 1.11 1.28
0.13 0.20 0.36 0.55 1.10 0.91
0.05 0.10 0.17 0.29 0.42 0.60
Table 2 shows the data of CV morbidities (acute or old myocardial infarction, angina and coronary disease, cardiac arrhythmia, heart failure, cerebrovascular disease, other heart disease, or pulmonary embolism) with their prevalences in the general population and in patients with the diagnosis of asthma or COPD. We also show the data of hypertensive disease with its prevalence in the general population and in patients with the diagnosis of asthma or COPD. The diagnosis of COPD was significantly associated with an increased likelihood of diagnosis of CVD and hypertensive disease (Fig. 1B). On the contrary, although CVD and hypertensive disease were more prevalent in patients with diagnosis of asthma than in the general population (Table
Pulmonary embolism 0.04 0.04 0.07 0.07 0.13 0.12 0.28 0.24 0.70 0.63 0.82 0.78
2), apparently asthma had a weak association with these diseases (Fig. 1A). Intriguingly, the adjusted OR for acute or old myocardial infarction was 0.98 (95% CI 0.90e1.07) in patients with the diagnosis of asthma and 2.91 (95% CI 2.74e3.09) in those with the diagnosis of COPD.
Impact of gender on CVD in asthma and COPD CV morbidities in the general population affected slightly more women than men, with an adjusted men/women ratio of 0.98. The ration was balanced in COPD (1.00), but it decreased in asthma (0.95) indicating a higher prevalence
252
M. Cazzola et al.
Table 2 Demographic characteristics of study population by prevalence of CV morbidities in the general population, and in patients with the diagnosis of asthma or COPD (%).
Angina and coronary disease Acute or old myocardial infarction Cardiac arrhythmia Heart failure Hypertension disease Cerebrovascular disease Other heart disease Pulmonary embolism
General population (total)
General population (non-asthmatic and non-COPD patients)
Asthma
Total
Total
Total
Men
Women
Men
Women
COPD
Men
Women
Total
Men
Women
2.69
3.46
2.03
2.39
3.07
1.80
3.35
4.10
2.89
9.20
10.42
7.48
1.63
2.56
0.83
1.52
2.38
0.77
1.47
2.47
0.85
4.81
6.30
2.69
6.56 1.35 33.46 2.15 0.22 0.22
6.29 1.32 31.78 1.91 0.23 0.20
6.80 1.37 34.92 2.37 0.22 0.24
6.06 1.06 32.13 2.04 0.19 0.20
5.72 0.98 30.33 1.95 0.19 0.18
6.36 1.13 33.68 2.12 0.20 0.22
8.68 2.08 38.79 2.62 0.36 0.35
8.33 2.03 36.02 2.25 0.37 0.32
8.90 2.12 40.49 2.85 0.35 0.37
15.60 7.21 58.08 4.19 0.74 0.70
15.53 7.13 55.79 0.76 0.76 0.61
15.71 7.33 61.33 9.03 0.71 0.82
of CVD in asthmatic women. There was always a prevalence of hypertensive disease in women, with a greater impact in COPD (men/women ratio: 0.91 in the general population, 0.91 in COPD, and 0.89 in asthma, respectively). Among patients with the diagnosis of asthma, there was not difference in gender for association with all CVDs, excluded for angina and coronary disease, for which the OR was higher in women compared to men (Fig. 1A). Among patients with the diagnosis of COPD, there were not differences in gender for association with all CVDs, excluded for acute or old myocardial infarction and angina and coronary disease, for which the OR was higher in women compared to men (Fig. 1B).
Impact of age on CVDs in patients with the diagnosis of asthma or COPD Both in patients with the diagnosis of asthma and in those with the diagnosis of COPD, an increase in age resulted in a progressive increase in the prevalence of the diagnosis of CVD and hypertensive disease. However, while the different age clusters did not affect differently the CV morbidities in asthma, clusters between 35 and 54 years were those at highest correlation with the simultaneous presence of the diagnoses of CV morbidities and COPD, with a progressive significant reduction in older age clusters. Intriguingly, the age cluster 35e44 years caused the highest ORs for angina and coronary disease, heart failure, and cerebrovascular disease, and that 45e54 years caused the highest correlation with acute or old myocardial infarction. Age did not influence differently the association with cardiac arrhythmia, although there was a decrease of adjusted OR in the age cluster 35e44. Fig. 2 illustrates the ORs of diagnoses of CVD and hypertensive both in patients with the diagnosis of asthma and in those with the diagnosis of COPD according to the different age clusters.
Impact of smoking habits on CVD in patients with the diagnosis of asthma or COPD Table 3 shows the demographic characteristics of the study population by smoking status (never smokers, current-
smokers and former smokers) and by the diagnosis of asthma and COPD. As expected, there were differences of smoking habit between general population and COPD, on the contrary the prevalence of smoking habits in asthmatics was analogous to that in general population. In particular, smoking prevalence was higher in COPD compared to asthmatics (COPD/asthma ratio: 1.45, current-smokers; 1.63, former smokers), whereas never smokers were the majority in asthmatics (COPD/asthma ratio: 0.57, never smokers). In patients with the diagnosis of asthma, the smoking habit did not modify the prevalence of CVD, compared to the general population. On the contrary, in patients with the diagnosis of COPD the overall prevalence of CVD was more than double in current and former smokers and it increased by more than five-fold in never smokers, compared to the general population (Table 4). The analysis of the putative interference of smoking habit on the association with CVD showed that the smoking habit represented neither a bias nor a confounding factor. In fact, the diagnosis of COPD increased by 2.6 folds and significantly (P < 0.05) the overall correlation with CVD in never-smokers compared to current and former smokers, versus the general population. On the contrary, the diagnosis of asthma did not modify the OR of CVD by clustering for smoking habit (Table 5).
Discussion This large observational study is the first, to the best of our knowledge, that has explored the difference in prevalence of the diagnosis of CVD between patients with and without diagnoses of COPD or asthma performing stratified analyses by age or gender. It documents that the diagnosis of asthma is modestly correlated with the diagnosis of different CV morbidities and provides further evidence that patients with the diagnosis of COPD are highly associated with the diagnoses of CVDs. Moreover, it suggests that the diagnosis of COPD is strongly associated with that of CVD, regardless the smoking habit. We believe that the additional information to the possible, although weak, association between the diagnoses of asthma and CVD is important because, as already
A population-based retrospective cross-sectional study
Figure 1 Adjusted odds ratios and 95% confidence limits of CV morbidities in patients with the diagnosis of asthma (A) and COPD (B). "Total” adjusted for age and sex, “Men” and “Women” adjusted for age. p < 0.05 vs non-asthmatic (A) and vs non-COPD (B) population.
mentioned, literature on this topic is conflicting.5,8e13 Our data confirm a higher prevalence of the diagnosis of CVD in women with that of asthma, show that, at least in men, the diagnosis of asthma is associated with diagnosis of angina and coronary disease but not with that of acute or old myocardial infarction, and do not confirm the reported association between the diagnoses of asthma and cerebrovascular disease.13 We must admit that we are unable to explain why there was an association of the diagnosis of asthma with that of angina and coronary disease but not with that of myocardial infarction in men, although the prevalence of the diagnosis of angina and coronary disease and that of myocardial infarction in asthmatic men were similar to those in men in the non-asthmatic population. It is likely that the low number of events could have generated information that larger numbers could deny, although again we must emphasize the breadth of our sample. In any
253 case, we cannot exclude that the observed increased ORs for the diagnoses of cardiac arrhythmia and heart failure, as well as other heart diseases, may suggest that the consequences of those diseases on the lung could have been confounded with asthma. However, we must stress that we have only considered comorbidities that were present in the same patient simultaneously with the diagnosis of asthma. Another finding that is difficult to explain is the impact of gender in the association between the diagnoses of asthma or COPD and CVD. Several reports suggest a role for female sex hormones in promoting the asthmatic phenotype.20 On the other hand, female smokers appear to be more susceptible than male smokers to developing COPD. Women have a more vigorous inflammatory reaction and generate more oxidative stress in the airways than men21 likely because estrogens could cause an imbalance in the formation and removal of toxic intermediates of cigarette smoke components.22 All this would suggest a stronger association between the diagnoses of asthma or COPD and CVD in women, but often our data did not confirm it. Soriano and colleagues1 suggested that the occurrence of CVD events in their study was generally lower in asthma patients compared to COPD patients likely because of the younger age distribution. Also in our population the diagnosis of asthma was prevalent in younger people and this could indicate that the age is the main association factor of CVD in asthma or COPD. However, although an increase in age resulted in a progressive increase in the prevalence of the diagnoses of CVD and hypertensive disease both in the diagnoses of asthma and COPD, the various age clusters did not affect differently the association of the diagnosis of CV morbidities with that of asthma, whereas cluster 3544 years was that at highest correlation with the simultaneous presence of CV morbidities and COPD, with a progressive significant reduction in older age clusters. Also Feary and colleagues found that the association between diagnoses of COPD and CVD was strongest in the younger age groups, persisted across all ages but reduced as age increased.23 It is also difficult to explain the higher association of the diagnosis of CVDs with the diagnosis of COPD among younger than older patients that we have found in our study. It has been suggested that it may be mediated by the use of inhaled bronchodilators24 because these agents were shown to impose a larger relative risk in people with unrecognized CVD than in people with known disorders.25 We do not agree with this hypothesis because our data show that the adjusted ORs of cardiac arrhythmia, a common side effect of bronchodilators, were almost constant in all age clusters. We believe that our results are interesting because they come from a large, validated, population-based database that illustrates what can be observed in real life and can be considered representative of the general population. This is the major strength of the present study. Moreover, use of routine medical records should have eliminated differential misclassification due to recall or interviewer bias. The main limitations of our study are common to retrospective database analyses. These limitations might be associated mainly with bias towards the selection of more highly motivated, or better organized, practices. Consequently, our findings might likely be an overestimate of the
254
M. Cazzola et al.
Figure 2 Adjusted odds ratios and 95% confidence limits of CV morbidities in patients with the diagnosis of asthma or COPD in different strata of age in years. p < 0.05 vs comorbidities in non-asthmatic and vs non-COPD population. Values adjusted for sex.
Table 3 Demographic characteristics of study population by smoking habits, and the diagnosis of asthma or COPD (%).
Never smokers Current-smokers Former smokers
General population
Asthma
COPD
53.62 25.51 20.88
54.98 25.19 19.83
31.23 36.46 32.32
general picture on the one hand. Nonetheless, on the other hand, it is well known that patients vary in their use of health care and it is accepted that a significant proportion of patients who are ill may not consult their doctor and this can lead to underestimation of the problem. Several reports have highlighted that diagnosed prevalence and morbidity data underestimate the total burden of COPD because the disease is usually not diagnosed until it is clinically apparent, and there is considerable variation in
reported prevalence.26e31 This is likely due to a lack of awareness about COPD by patients. It is only 25e50% of COPD patients are known to their doctor.27 Patients may well experience symptoms, but do not always present these symptoms to their family physician.27 Poor recognition of the disease may lead patients to believe that their respiratory symptoms are nothing more than a smoker’s cough, rather than an early sign of disease pathology, causing them to delay contact with a healthcare professional.26 In any case, the majority of patients with a decreased FEV1 do not complain of any bronchial symptoms.32 Nonetheless, the underestimation of the actual burden of COPD may also be due to the underdiagnosis of the disease by healthcare professionals, particularly in the early stages of the disease,33 mainly because in family practice the diagnostic process is primarily based on symptoms and signs presented by the patient and, consequently, early-stage COPD can easily be veiled by signs of asthma.34 Also a lack of awareness among physicians of
Table 4 Influence of smoking status on CV morbidities in the diagnosis of asthma or COPD related to the general population (values are expressed as rates). Asthma vs General population
Ischemic cardiopathy Myocardial infarction Cardiac arrhythmia Heart Failure Hypertension Cerebrovascular disease Other heart disease Pulmonary embolism
COPD vs General population
Neversmokers
Current smokers
Former smokers
Neversmokers
Current smokers
Former smokers
1.21 0.88 1.29 1.51 1.13 1.19 1.56 1.53
1.26 0.91 1.34 1.57 1.17 1.23 1.61 1.59
1.31 0.95 1.39 1.63 1.22 1.28 1.68 1.65
5.87 5.05 4.08 9.20 2.98 5.41 5.67 5.33
2.39 2.06 1.66 3.75 1.21 2.20 2.31 2.17
2.21 1.90 1.54 3.46 1.12 2.04 2.13 2.00
A population-based retrospective cross-sectional study Table 5 Adjusted odds ratios and 95% confidence limits of overall CV morbidities in never-smoker population with the diagnosis of asthma or COPD related to current and former smokers. Data are adjusted for CV morbidities. Asthma Never-smokers/Current smokers vs general population Never-smokers/Former smokers vs general population
COPD
0.96 2.45 (0.80e1.13) (1.58e3.33)* 0.93 2.66 (0.76e1.10) (1.80e3.51)*
255 asthma. It is clear, therefore, that there is a strong need to look beyond the lungs in treating patients with asthma or COPD. In any case, we must highlight that since the GPs usually state diagnosis of COPD, asthma or CVD by using ICD codes and a diagnosis is not true disease identification, the relationships here presented have been described as relationship between diagnoses, and not between diseases.
Conflict of interest None declared.
*p < 0.05 vs general population.
Acknowledgments recommended treatment practices can justify the underdiagnosis of COPD.33,34 In any case, underestimation of the actual burden of COPD is an important issue because it may cause a bias when trying to study the real relationships between diseases, a bias that register studies cannot control and correct for. We must highlight that GPs are faced with an array of alternative codes for the same or similar conditions and, in effect, practicing physicians have little faith in the accuracy of the ICD-9 diagnoses entered for prescription purposes and this can question the real value of diagnoses in the HSD. Therefore, there may have been misclassification when our GPs have used specific diagnostic categories and this likely because there are no standardized methods of applying diagnostic labels in general practice. In particular, a gender bias in the diagnosis of COPD and asthma is to be expected when the epidemiology of a disease makes a dramatic shift in terms of maleefemale prevalence. Physicians are at risk for assuming COPD in men and asthma in women.14 We did not use the lung function data for our analyses because only one third of patients with COPD and less than 20% of those with asthma had spirometry records. This finding fits perfectly with the results of the National Committee for Quality Assurance (NCQA) study that suggested that approximately 32% of a broad range of patients with a new COPD diagnosis had undergone spirometry within the previous 2 years to 6 months following diagnosis.35 In addition, spirometric testing appeared to decrease with increasing age. A study that investigated the degree of use of spirometry to establish the diagnosis of COPD in Italy documented that many GPs do not use spirometry to establish the diagnosis of COPD because they think that spirometry is not necessary for the diagnosis of COPD or there are logistical limitations to the access of the patients to lung function laboratories.36 Regardless of anything else, we must mention that in the analysis of Feary and colleagues that used data from primary care to investigate the prevalence of major comorbidities in subjects with COPD, no clinically important difference was seen in the results after excluding patients with COPD but without spirometry records from the analysis.37 In conclusion, our study clearly indicates that the diagnosis of CVD is strongly associated with the diagnosis of COPD and, albeit in a less strong manner, even with that of
This study is based on data from the Health Search database owned by the Italian College of General Practitioners. However, the interpretation and conclusions contained in this study are those of the authors alone. The original idea for the study were formulated by Prof. Cazzola, with subsequent advice from Drs. Bettoncelli and Cricelli, and statistical input from Dr. Calzetta, who also performed the statistical analyses. Prof. Cazzola drafted the manuscript, with extensive contributions and revision from all authors. Access to the Health Search database was offered free by the Italian College of General Practitioners.
References 1. Soriano JB, Visick GT, Muellerova H, Payvandi N, Hansell AL. Patterns of comorbidities in newly diagnosed COPD and asthma in primary care. Chest 2005;128:2099e107. 2. Fabbri LM, Luppi F, Beghe ´ B, Rabe KF. Complex chronic comorbidities of COPD. Eur Respir J 2008;31:204e12. 3. Cazzola M, Bettoncelli G, Sessa E, Cricelli C, Biscione G. Prevalence of comorbidities in patients with chronic obstructive pulmonary disease. Respiration 2010;80:112e9. 4. Gershon AS, Wang C, Guan J, To T. Burden of comorbidity in individuals with asthma. Thorax 2010;65:612e8. 5. Cazzola M, Calzetta L, Bettoncelli G, Cricelli C, Novelli L, Rogliani P. Asthma and comorbid medical illness. Eur Respir J 2011;38:42e9. 6. Mannino DM, Thorn D, Swensen A, Holguin F. Prevalence outcomes of diabetes, hypertension, and cardiovascular disease in COPD. Eur Respir J 2008;32:962e9. 7. Rutten FH, Cramer MJ, Lammers JW, Grobbee DE, Hoes AW. Heart failure and chronic obstructive pulmonary disease: an ignored combination? Eur J Heart Fail 2006;8:706e11. 8. Soriano JB, Rigo F, Guerrero D, Yan ˜ez A, Forteza JF, Frontera G, et al. High prevalence of undiagnosed airflow limitation in patients with cardiovascular disease. Chest 2010; 137:333e40. 9. Appleton S, Ruffin S, Wilson D, Taylor A, Adams R. Asthma is associated with cardiovascular disease in a representative population sample. Obes Res Clin Pract 2008;2:91e9. 10. Iribarren C, Tolstykh IV, Eisner MD. Are patients with asthma at increased risk of coronary heart disease? Int J Epidemiol 2004; 33:743e8. 11. Onufrak SJ, Abramson JL, Austin HD, Holguin F, McClellan WM, Vaccarino LV. Relation of adult-onset asthma to coronary heart disease and stroke. Am J Cardiol 2008;101:1247e52. 12. Enright PL, Ward BJ, Tracy RP, Lasser EC. Asthma and its association with cardiovascular disease in the elderly. The
256
13.
14.
15.
16.
17.
18.
19. 20.
21.
22.
23.
24.
M. Cazzola et al. Cardiovascular Health Study Research Group. J Asthma 1996; 33:45e53. Schanen JG, Iribarren C, Shahar E, Punjabi NM, Rich SS, Sorlie PD, et al. Asthma and incident cardiovascular disease: the Atherosclerosis risk in communities study. Thorax 2005;60: 633e8. Cazzola M, Puxeddu E, Bettoncelli G, Novelli L, Segreti A, Cricelli C, et al. The prevalence of asthma and COPD in Italy: a practice-based study. Respir Med 2011;105: 386e91. World Health Organization. International classification of diseases. Manual of the international statistical classification of diseases: injuries and causes of deaths (9th revision). Geneva: World Health Organization; 1977. Lawrenson R, Williams T, Farmer R. Clinical information for research: the use of general practice databases. J Public Health Med 1999;21:299e304. Filippi A, Bignamini AA, Sessa E, Samani F, Mazzaglia G. Secondary prevention of stroke in Italy: a cross-sectional survey in family practice. Stroke 2003;34:1010e4. Zhang J, Yu KF. What’s the relative risk? A method of correcting the odds ratio in cohort studies of common outcomes. JAMA 1998;280:1690e1. Schmidt OC, Kohlmann T. When to use the odds ratio or the relative risk? Int J Public Health 2008;53:165e7. Card JW, Zeldin DC. Hormonal influences on lung function and response to environmental agents: lessons from animal models of respiratory disease. Proc Am Thorac Soc 2009;6: 588e95. Ben-Zaken Cohen S, Pare ´ PD, Man SFP, Sin DD. The growing burden of chronic obstructive pulmonary disease and lung cancer in women. Examining sex differences in cigarette smoke metabolism. Am J Respir Crit Care Med 2007;176: 113e20. Sin DD, Cohen SB, Day A, Coxson H, Pare PD. Understanding the biological differences in susceptibility to chronic obstructive pulmonary disease between men and women. Proc Am Thorac Soc 2007;4:671e4. Johnston AK, Mannino DM, Hagan GW, Davis KJ, Kiri VA. Relationship between lung function impairment and incidence or recurrence of cardiovascular events in a middle-aged cohort. Thorax 2008;63:599e605. Macie C, Wooldrage K, Manfreda J, Anthonisen N. Cardiovascular morbidity and the use of inhaled bronchodilators. Int J Chron Obstruct Pulmon Dis 2008;3:163e9.
25. Camp PG, Goring SM. Gender and the diagnosis, management, and surveillance of chronic obstructive pulmonary disease. Proc Am Thorac Soc 2007;4:686e91. 26. Britton M. The burden of COPD in the U.K.: results from the Confronting COPD survey. Respir Med 2003;97(Suppl C):S71e9. 27. van Schayck CP, Chavannes NH. Detection of asthma and chronic obstructive pulmonary disease in primary care. Eur Respir J 2003;39(Suppl:1):6se22s. 28. Mohangoo AD, van der Linden MW, Schellevis FG, Raat H. Prevalence estimates of asthma or COPD from a health interview survey and from general practitioner registration: what’s the difference? Eur J Public Health 2006;16:101e5. 29. Lindberg A, Bjerg A, Ro ¨nmark E, Larsson LG, Lundba ¨ck B. Prevalence and underdiagnosis of COPD by disease severity and the attributable fraction of smoking report from the obstructive lung disease in Northern Sweden studies. Respir Med 2006; 100:264e72. 30. Vandevoorde J, Verbanck S, Gijssels L, Schuermans D, Devroey D, De Backer J, Kartounian J, Vincken W. Early detection of COPD: a case finding study in general practice. Respir Med 2007;101:525e30. 31. Tinkelman DG, Price D, Nordyke RJ, Halbert RJ. COPD screening efforts in primary care: what is the yield? Prim Care Respir J 2007;16:41e8. 32. White P. Prevalence of COPD in primary care: no room for complacency. Fam Pract 2009;26:1e2. 33. Nacul L, Soljak M, Samarasundera E, Hopkinson NS, Lacerda E, Indulkar T, et al. COPD in England: a comparison of expected, model-based prevalence and observed prevalence from general practice data. J Public Health 2011;33:108e16. 34. Albers M, Schermer T, Molema J, Kloek C, Akkermans R, Heijdra Y, et al. Do family physicians’ records fit guideline diagnosed COPD? Fam Pract 2009;26:81e7. 35. Han MK, Kim MG, Mardon R, Renner P, Sullivan S, Diette GB, et al. Spirometry utilization for COPD: how do we measure up? Chest 2007;132:403e9. 36. Caramori G, Bettoncelli G, Tosatto R, Arpinelli F, Visona ` G, Invernizzi G, et al. Underuse of spirometry by general practitioners for the diagnosis of COPD in Italy. Monaldi Arch Chest Dis 2005;63:6e12. 37. Feary JR, Rodrigues LC, Smith CJ, Hubbard RB, Gibson JE. Prevalence of major comorbidities in subjects with COPD and incidence of myocardial infarction and stroke: a comprehensive analysis using data from primary care. Thorax 2010;65: 956e62.