Epidemiological studies of asthma and allergic diseases in teenagers : methodological aspects and tobacco use

Parental reports are often used in studies of asthma and allergic diseases in children. A change in respondent from parent to index subject usually occurs during adolescence. Little is known about the effects this change in method might have on the outcomes of a longitudinal study. Smoking is a major cause of respiratory symptoms among adults and environmental tobacco smoke (ETS) is a risk factor for asthma among children. Less is known about these associations among teenagers. In order to improve prevention of smoking, it is important to identify populations at risk of becoming smokers. The aim of this thesis were to 1) evaluate the methodological change from parental to self-completion of a questionnaire about asthma and allergic diseases, and 2) to study determinants for, and respiratory health effects of ETS and personal smoking in teenagers. In 1996, a longitudinal study of asthma and allergic diseases among schoolchildren started within the Obstructive Lung Disease in Northern Sweden (OLIN) studies. All children in first and second grades (aged 7-8 years) in three municipalities, Luleå, Kiruna and Piteå (n=3,525) were invited and 97% participated by parental completion of a questionnaire. The cohort has been followed with annual questionnaires until age 16-17 years and with high participation rates (>91%). From age 12-13 years, the teenagers were the respondents and questions about their tobacco use were included. In addition to the questionnaire completed by the teenagers at age 13-14 years, a questionnaire was also distributed to a random sample of 10% of the parents and 294 participated (84%). The parents and the teenagers reported a similar prevalence of asthma, respiratory symptoms, rhinitis, eczema and environmental factors. Two statistically significant differences were found: the teenagers reported a higher prevalence of wheezing during or after exercise (14% vs 8%, p<0.05), and having a dog in the home in the last 12 months (42% vs 29%, p<0.001). Answer agreement between parents and teenagers on questions about asthma was almost perfect with kappa values of 0.8-0.9. Corresponding kappa values for questions about respiratory symptoms and rhinitis were 0.3-0.6 and for eczema 0.5-0.6. Agreement about environmental factors varied from 0.2-0.9. Kappa values for parental smoking were 0.8-0.9. The risk factor pattern for allergic diseases was similar regardless of respondent, ie parent or teenager. The prevalence of smoking increased from 3% at 12-13 years to 6% at 14-15 years. Smoking was more common among girls, while the use of snus was more common among boys. Significant risk factors related to smoking among teenagers were smoking family members, female sex and living in an apartment. Having physician-diagnosed asthma did not prevent the teenagers from becoming smokers. Factors related to using snus were a smoking mother and male sex. Daily smokers aged 16-17 years (9%) reported a significantly higher prevalence of wheezing and physician-diagnosed asthma compared to non-smokers. There was a significant dose-response association with higher prevalence of wheeze among those who smoked ≥11 cigarettes per day compared to those who smoked ≤10 per day. In multivariate analyses, maternal environmental tobacco smoke exposure was a significant risk factor for ever wheeze and physician-diagnosed asthma at age 16-17 years, while daily smoking was a risk factor for current wheeze. In conclusion, the methodological change of questionnaire respondent from parent to index subject did not substantially alter the findings of this longitudinal study. There were significant sex differences in the tobacco use: smoking was more common among girls and snus was more common among boys. The most important factor related to tobacco use was presence of family members who smoke. Both maternal ETS exposure and personal smoking was associated with asthma and wheeze in adolescence. ETS was associated with lifetime symptoms but daily smoking was more strongly associated with current symptoms