Chronic obstructive pulmonary disease (COPD) is a slowly progressive disease characterised by irreversible airflow limitation, which is caused by a chronic inflammatory immune response to inhaled gases and particles.1 2 In chronic bronchitis, belonging to the clinical spectrum of COPD, the inflammatory process is located in central airways and it is associated with increased mucus production, reduced mucociliary clearance and disrupted airspace epithelial barrier. In emphysema, another clinical manifestation of COPD, the distal airspaces are enlarged because of destruction of airway walls. In the third subtype, obstruction of small airways (<2 mm diameter) only extensive damage will be recorded in lung function tests because small airways only account for 10–15% of total airway resistance.1 Tobacco smoking is the single commonest risk factor of COPD but many other factors are known.1 In smokers of more than 40 pack-years, the prevalence of COPD has been reported to be over 50%.3 It is a shared risk factor for COPD and lung cancer, but familial aggregation of the two conditions suggests that tobacco smoking alone does not explain the co-morbidity.4 5

α-Antitrypsin deficiency is an established genetic risk factor of emphysema, and it is a classic example on a common risk allele, a paradigm used for the present genomics of common diseases.6 7 Homozygous carriers of the susceptibility allele have plasma levels of α-antitrypsin of only 10–15% of the levels in non-carriers and they are at a high risk of emphysema if they smoke.8^–10 Many other candidate susceptibility genes for COPD have been tested individually or following linkage studies but no consensus have been reached.11 12 Many family and twin studies have been carried out but most of them are very old with a few exceptions.12^–14 For example, a small family study on three index cases of emphysema, published in 2002, cited 13 papers on the similar theme with publication years ranging from 1921 to 1986.15 Studies in which the diagnosis in family members have not been medically verified are known to have a notoriously poor accuracy.16 However, to date, familial studies on COPD have been hampered by the inability to obtain the necessary information about a sample sufficiently large to permit reliable conclusions. The present study set out to define risks for COPD between singleton siblings and twins, with reassuring patient numbers. It covers all hospitalised patients for chronic lower respiratory diseases (used here synonymously with COPD) from Sweden during years 1986 and 2004, a total of 19 637 patients. Data on family relationships and birth dates allow us to calculate risks for COPD in siblings, twins and spouses. In view of the ongoing genetic studies, reliable family data are needed for the dissection of the assumed heritable and environment components and their interactions. The Swedish family-disease datasets have been recently used in the study of many familial diseases.17^–19

METHODS

The COPD research database, used for this study, was constructed by linking several national Swedish registers. Statistics Sweden, the Swedish government-owned statistics bureau, provided the Multigeneration Register where people (second generation) born in Sweden in 1932 and later are linked to their parents (first generation), registered shortly after birth. Families could be defined by linking all the children to their parents. Sibships can only be defined for the second generation, which was the present study population. Linkages were carried out to national census data in order to obtain individual occupational status. The final link was made by adding individual data from the Swedish Hospital Discharge Register that records complete data on all discharges with dates of hospitalisation and diagnoses since 1986 and from the Causes of Death Register. All linkages were performed by the use of an individual national identification number that is assigned to each person in Sweden for their lifetime. This number was replaced by a serial number for each person in order to provide anonymity. The serial number was used to check that each individual was only entered once, for his or her first appearance with a COPD diagnosis. Over seven million individuals with their siblings were included in this database.

Outcome variable

Details of COPD patients were retrieved from hospital discharges or cause of death notifications, reported according to the 9th (1987–1996) and 10th (1997–2004) version of the International Classification of Diseases (ICD). The subtypes of COPD were defined as obstructive chronic bronchitis (codes in ICD-9 and ICD-10: 491.2 and J44), emphysema (492, J43). The following codes of COPD were not shown in the present study: 2060 cases of other chronic bronchitis (491.0, 491.1, 491.W, 491,X, J41, J42), 2171 cases of unspecified bronchitis (490, J40), 281 cases of bronchiectasis (494, J47) and 759 cases of others (495, 496). Data on α₁-antitrypsin deficiency were not available. A total of 19 637 patients were identified, 17 752 (90%) from the Hospital Discharge Register and 1885 (10%) from the Causes of Death Register.

Individual variables controlled for in the analysis

• Gender: men and women

• Sibling: those who had common parents

• Age at diagnosis: categorised as follows: 0–4, 5–9, 10–19, and ⩾20 years; the groups were merged as necessary

• Occupational status for the men and women was divided into six groups according to occupation: (1) farmers, (2) unskilled/skilled workers, (3) white-collar workers, (4) professionals, (5) self-employed and (6) all others

• Region: divided into three groups: (1) large cities—Stockholm, Gothenburg and Malmo, (2) southern Sweden and (3) northern Sweden, allowing adjustment for regional differences in hospitalisation.

• Spouse: defined for the population older than 25 years through common children.

Statistical analysis

Person-years were calculated from start of follow-up on 1 January 1987 until hospitalisation for COPD, death, emigration or closing date, on 31 December 2004. Age-specific incidence rates were calculated for the whole follow-up period, divided into six 3-year periods. Standardised incidence ratios (SIRs) were calculated as the ratio of observed (O) to expected (E) number of cases. The expected number of cases was calculated for age (5-year groups), sex, period (3-year groups), region and occupational status-specific standard incidence rates. Sibling risks were calculated for singleton men and women (n = 554, of whom 518 came from families of two affected siblings and 36 from families of three affected siblings) or separately for twins (25 affected twin pairs, lacking zygosity data), with siblings affected with COPD, compared with men and women whose siblings were not affected by these conditions, using the cohort methods as described.20 Confidence intervals (95% CI) were calculated assuming a Poisson distribution, and they were adjusted for dependence between the sibling pairs.20 Age-adjusted incidence rates were calculated based on the European standard population. Age strata were tested for heterogeneity by the χ² test.

RESULTS

Obstructive chronic bronchitis was the commonest type of COPD causing hospitalisation in Sweden, with 6431 male and 6762 female patients, 554 with a sibling history of disease (table 1). Emphysema was much rarer and it showed a male excess of cases.

A total of 554 singleton sibling pairs were diagnosed with COPD, giving an overall familial SIR of 3.02, somewhat higher for men than for women (table 2). Heterogeneity tests showed p values of <0.01 men and women separately and for all. When one sibling was diagnosed with any COPD, the risk for a specific subtype in the other sibling was highest, 5.89, for emphysema compared to obstructive chronic bronchitis, 4.54 (table 3). When both siblings were diagnosed with a specific subtype, emphysema showed by far the highest risk of 14.22; however, only 12 individuals were affected (table 4). The SIR for concordant obstructive chronic bronchitis in siblings was 4.45 (448 affected individuals), somewhat lower than that for the discordant obstructive chronic bronchitis-emphysema pairs (SIRs 5.56 and 5.10). Only 25 twin pairs were diagnosed with COPD (data not shown), giving a SIR of 11.87 for obstructive chronic bronchitis (n = 24, 95% CIs 5.10 to 16.94). Only one twin pair was diagnosed with emphysema.

Two kinds of analyses were carried out in order to test for the extent of environmental sharing on the observed risks. Firstly, SIRs were analysed for siblings according to their age difference. Overall, the age difference (<5 years vs >5 years) had little effect (p values for heterogeneity of 0.94 for obstructive chronic bronchitis and 0.95 for emphysema). As a second test, risks were calculated for spouses to be hospitalised for COPD (table 5). The SIR for spouses was 1.55 to 1.57 when the wife/the husband was hospitalised for the same reason.

DISCUSSION

Tobacco smoking is the most important environmental risk factor for COPD. Based on spouse correlation for lung cancer with a SIR of about 1.5, smoking habit aggregates in families.21 22 Although twin correlation for smoking habit suggests heritability of some 50%,23 the risk for lung cancer between siblings is no higher than 2.2 in Sweden.24 Thus the present sibling risk of 4.54 for obstructive chronic bronchitis and 5.89 for emphysema (14.22 for siblings with concordant emphysema) appears far higher than could be explained by familial smoking habits alone. These SIRs are clearly higher than the sibling risk for asthma (3.2) found in the present database25 or in several of the published studies.26

The analysis of sibling risks for COPD by their age difference, assuming that a small age difference was a proxy for sharing of habits, showed no indication of environmental sharing. For lung cancer, the reported SIRs have been 2.19 and 2.04 for small and large age difference, respectively, thus suggesting some sharing.27 The spouse SIRs were about 1.6, in line with spouse risks for lung cancer. It is interesting that spouse correlation suggested environmental sharing, as opposed to the analysis of siblings by age difference, which could be explained, for example, by higher correlation of smoking by spouses than by siblings. The present sibling risks concerned singleton births and separate analysis was carried out for twins. The risk for obstructive chronic bronchitis in twins was 11.87 but because only 24 pairs of unknown zygosity were available, no detailed analysis was possible. Anyway, the higher risk for twins compared to singleton siblings is not inconsistent with the contribution of heritable effects in COPD. Obviously α-antitrypsin deficiency is one and perhaps a dominant cause for the high sibling risk for emphysema. To what extent other genes contribute to this effects is beyond the present genetic epidemiological data.

The present study is the only nationwide family study on COPD; the studied population was seven million. The design of using hospitalised COPD patients affords advantages and disadvantages. The disadvantages include the reduction in the number of cases and the possibility that some selective factors operate in the process of hospitalisation that would favour siblings being hospitalised. However, the analysis of spouses showed no indications of such selective factors. Because the follow-up was 18 years, only risks between siblings rather than between parents and offspring could be determined. Affordability or accessibility of health care is unlikely to be a selective factor as there is a universal access to health care in Sweden. If siblings in general had a tendency to seek or receive hospitalisation, we would assume that siblings close in age would show the highest risk, which was contrary to the data. The advantages include complete nationwide coverage in a country of high medical standards and a medical diagnosis of both siblings, often by a specialist during extended examinations in the clinic. Importantly, the diagnoses are those with which the patients are discharged and thus are highly accurate. Any misclassification would decrease familial risks.

The diagnostics of COPD in primary care may be a problem. Patient characteristics, such as cough and smoking, are independently related to mild COPD, and it may not be possible for the primary care physician to exclude COPD.28 Therefore, knowledge about the familiar risk, in addition to patient symptoms and characteristics, may help to identify early mild COPD.

In conclusion, the far higher familial risks between singleton siblings and twins compared to spouses provide strong genetic epidemiological evidence for the overall heritable effects for COPD, which are stronger for emphysema than for obstructive chronic bronchitis. α-Antitrypsin deficiency and its interaction with environmental factors may be an overwhelming genetic determinant of these findings, particularly in emphysema. Discovering other genetic factors and gene-environment interactions will be a major challenge, which calls for large well characterised patient series.