Second hand smoking is well known for its deteriorating effects in man. It increases significantly recurrent colds, cough and phlegm, throat problems, and the presence of respiratory symptoms by roughly a factor of 2 [1, 2]. It even has been linked to increase the risk for lung cancer by the factor of 1.3, particularly when the exposure occurs before the age of 25 [3]. Cigarette smoking is still quite common in many parts of the world. In parallel, exposure to second-hand smoke continues to be widespread despite declines in smoking in developed countries and despite evidence of serious health effects in infants and children. In addition to whatever effects direct and postnatal second-hand tobacco smoke exposure have on the occurrence of asthma, bronchitis, cardio-vascular diseases and impaired levels and growth of lung function in adolescents, there is an underlying alteration in the prenatal and early postnatal development of the structural and mechanical characteristics of the lung that contribute substantially to these deficits. These developmental effects are important contributors to the future risks for impaired pulmonary function in adolescence [4, 5].

While the detrimental effect of parental smoking exposure on the respiratory health of children is well known since years [6], information on the long-lasting consequence on human health, particularly regarding the development of diseases like chronic obstructive pulmonary disease (COPD) is sparse [7, 8]. Consequently, the purpose of this study was to assess the risk of environmental smoke exposure at home during childhood a) for COPD severity, b) for disease instability defined as higher exacerbation rate in later adulthood, and c) whether it might have an add-on effect to yearlong active smoke exposure.

This study has been approved by the ethics committee of the Saxonian Physician Association, Dresden/Germany (Code No.: EK-AMG-89/2003). Patients who refuse to participate or who were unable to understand the written informed consent form were excluded.

A structured 11 pages questionnaire was used to question prospectively COPD patients for their family and social history, previous occupation and possible occupational exposure status, parental and personal smoking habits. Diseases during childhood and adolescence were recorded as well as the year of COPD diagnosis and frequency of exacerbations, defined as those that required hospitalization and treatment with antibiotics and/or systemic corticosteroids [9]. At the time of the interview, following parameter were gathered from the patient: anthropometric data, dyspnoea using MRC-scale, lung function, blood gas analysis, and a current medication. To exclude concurrent diseases as defined below, patients were examined using EKG, echocardiography, X-ray of the chest.

Inclusion criteria were: COPD according to the GOLD (Global obstructive Lung Disease [10]) definition of all stages, age ≥ 30 years, negative reversibility testing (FEV1 < 12% and < 200 ml, 30 minutes after inhalation with salbutamol 400 μg). Patients were regarded non-eligible, when the following applied: severe concomitant co-morbidities in an acute deterioration or in chronic disease stage such as heart failure (NYHA ≥ 3), myocardial infarction ≤ 6 months, pneumonia and other acute infectious disease, asthma as defined by GINA (Global Initiative for Asthma [11]), interstitial and other lung disease except COPD, chronic kidney disease requiring dialysis, sleep apnoea syndrome, any malignant diseases, occupational disease of the lung (e.g. silicosis, asbestosis). For this bicentre trial COPD patients were recruited a) when admitted to the in-patient or out-patient unit of our centre (Robert-Koch-Hospital), an academic teaching hospital specialized in the management of acute and chronic pulmonary diseases including lung cancer, and b) in a private practice of a pulmonologist.

In contrast to never-exacerbaters (NE), patients with ≥ 1 exacerbation (1E) reported that they had to be hospitalized about 3-times (mean) within the last 24 years because of severe deterioration of their disease. At the time of interview 58% of all patients were exsmokers (= non-smoking ≥ 1 year), 22% smoked actively and 19.2% claimed to be live long non-smoker.

The fathers smoked longer and more cigarettes than the mothers (Table 2). Paternal smoking habits had no influence on FEV₁ of our COPD patients. In contrast, patients who's mothers smoked, exhibit significantly (P = 0.037) lower FEV₁ values than those raised without maternal smoke exposure (Figure 1). Hospitalization rate due to COPD exacerbations remained unaffected through smoking parents: Difference due to smoking vs. non-smoking fathers P = 0.999, and smoking vs. non-smoking mothers P = 0.547. No significant differences between HE and 1E were found (data not shown).

Parameter	Father	Mother
Smoker/non-smoker (%)	75.0/25.0	15.4/84.6
Cumulative cigarette exposure
(% positive answers)
< 10 py	5.7	32.5
10-20 py	23.7	55.0
> 20 py	70.6	12.5

py = pack/years (one pack year [py] = smoking 1 cigarette pack/day, everyday within 1 year).

Ca.10% of adult smokers demonstrate an increased decline of lung function with age [15, 16]. The effects of passive smoke exposure vary with genetic factors, gender, race and exposure to other pollutants [7, 8]. Exposure to environmental tobacco smoke during childhood and subsequent active smoking in later adulthood both further aggravate symptoms and have a negative effect on lung function, among others [17, 18].

This paper focuses on the effects of second-hand smoke in addition of direct exposure on the respiratory health of the lung function and frequency of exacerbations in COPD patients. Once birth cohort studies on this subject are unrealistic due to necessary long observation time, we choose an interview based approach in which COPD patients comment on their active cigarette smoking habits and second hand exposure during their childhood at home. Interestingly, maternal smoking resulted in further lung function decline in our patients, and thus has an additional effect to active smoking in later life. Although fathers smoked much more than theirs wives, paternal smoking was no contributor to the future risk for impaired pulmonary function. Based on earlier observations [19], we may speculate: Because mostly mothers but not their fathers raise children, the exposure to maternal cigarette smoke are much more intense than from paternal second hand smoke. Our patients could not know whether their mothers also smoked during pregnancy from obvious reasons. Assuming that this was the case in most instances, intra-uterine and inhalant second hand exposure may have contributed to lung function reduction as observed in our cohort [20]. Jurado et al [21] and others reported that maternal smoking increased the risk of wheezing with colds, cough with phlegm and asthma in the offspring up to 3-fold. The detrimental effect depended on the intensity of smoke exposure and if it occurs only prenatal, postnatal or both. Smoking fathers did not exert this effect or their influence was significantly smaller which was confirmed by our data [22–24].

Our results are very well in concurrence with published data indicating that direct and postnatal secondhand tobacco smoke exposure promotes the incidence of asthma and bronchitis, and impairs the growth of the lung [20, 25]. Tobacco smoke alters very early in life prenatal and early postnatal development of the structural and mechanical characteristics of the lung that contribute substantially to these deficits. These developmental effects are therefore important contributors to the future risks for impaired pulmonary function and COPD [26, 27].

Lower FEV₁ in the cohort with smoking mothers did not correlate with increased rate of severe exacerbations requiring hospitalization. This amazes considering, that patients with FEV₁ < 50% predicted has an 2.6-fold higher risk to become again hospitalized due to an COPD exacerbation [28]. Inversely reduces a FEV₁ increase of 10% the exacerbation rate by 28% [29]. FEV₁ values are by and large poor predictors for future exacerbations rates, because many patients with low FEV₁ may be stable and patients with lower disease stages may frequently exacerbate [30, 31]. Good therapeutic management as demonstrated in studies using tiotropium bromide and inhaled corticosteroids or smoking cessation may help reduce exacerbations [9, 32–34]. All these factors could have diminished in our study a possible influence of environmental tobacco smoke exposure in childhood.

Our study has various disadvantages. Although the data were collected prospectively, we relayed predominantly on individual mental capacity for incidents dating back decades in the past. Objective methods to verify patient's strength to remember this kind of information is, to our knowledge, not on hand. We omitted therefore intelligence tests and related assessments. The study design did not allow the differentiation between in utero and post-natal smoke exposure. We only can assume that smoking mothers smoke also during pregnancy and not just after the delivery of their child. According to previous reports, about one third of smoking mothers stopped during pregnancy [35]. But after successful smoking cession during pregnancy approximately 10 - 60% of all women relapse to smoking during the next 6 months. 34% of all children are exposed to maternal smoking [35, 36]. Considering these data, one can reckon that over 70% of our patients reporting second hand smoke exposure during childhood were also exposed in utero.

We hereby demonstrate the reliability of data collection from COPD patients even if the event date back many decades, and we confirmed for the first time with this approach that maternal smoke exposure effects COPD severity, which even adds up to active smoking.

The study was supported by Boehringer Ingelheim Pharma KG, Germany.