According to an ecological perspective in psychology and in line with social cognitive theory, smoking behaviour is determined by different social contexts (for example, peers, family and school) providing adolescents with important role models. This paper investigates the effects of personal characteristics as well as family, peer and school context variables on youth smoking behaviour. We hypothesize that school smoking policy variables predict adolescents' smoking in addition to other context variables. Data were obtained from a self-report survey administered to 3364 students in 40 secondary schools in Bavaria, Germany. For both younger (10-15 years) and older (16-21 years) students, strong associations were found between smoking behaviour and (i) smoking best friends and friends in general, (ii) other substance use and (iii) school performance. In the younger age group, the non-existence of smoking bans for students was associated with an increased risk of being smoker. For the older students, a positive association was found between the presence of smoking teachers on school grounds and smoking behaviour. Results are discussed considering methodological aspects and public health concerns.

Smokers with smoking-related diseases who are hospitalized in rehabilitation centers should be offered smoking cessation. This is the first study evaluating whether telephone booster sessions after intensive inpatient treatment are an effective strategy. The present study was conducted in 13 rehabilitation centers for somatic disorders as a prospective multicenter study with a randomized treatment-control group design. We compared abstinence rates after hospital discharge from treatment that included a group smoking cessation program with (treatment group) and without telephone booster sessions (control group). Data from 290 smokers were analyzed. After 6 and 12 months the treatment group achieved abstinence rates twice as high as those of the control group. Men profited more from telephone booster sessions than did women. Results indicated that telephone booster sessions were highly effective (even) after an inherently intensive group program during a hospital stay. Further research should focus on the special needs of women receiving telephone counseling.

<B>Ziel:</B> Mit dem Epidemiologischen Suchtsurvey 2003 werden die Beobachtungen des Tabakkonsums bei Erwachsenen in Deutschland fortgesetzt. </P><P> <B>Methode:</B> Die Stichprobe umfasst 8.061 Personen im Alter von 18 bis 59 Jahren. Die Datenerhebung erfolgte schriftlich, die Antwortrate betrug 55 %. </P><P> <B>Ergebnisse:</B> 37 % der befragten Männer und 31 % der Frauen sind aktuelle Raucher. Seit 1995 ist bei jungen Frauen wieder ein Anstieg der Konsumprävalenz festzustellen. Am Arbeitsplatz waren 27 % der erwerbstätigen und auszubildenden Nicht- und Exraucher täglich Passivrauchen ausgesetzt. </P><P> <B>Schlussfolgerungen:</B> Die im internationalen Vergleich hohe Konsumprävalenz in Deutschland erfordert verstärkte Präventionsmaßnahmen, insbesondere für Frauen. Die Arbeitsstättenverordnung zum Schutz der Nichtraucher am Arbeitsplatz bedarf weiterer Maßnahmen zur Durchsetzung.

Background/Aims: This study applied a multilevel approach to examine the associations between school smoking policy and student smoking. It was tested whether individual characteristics are mediators of school policy effects. Methods: On the basis of cross-sectional data from 3,364 students and school principals from 40 schools in Germany, two multilevel nonlinear regression models were computed for current smoking. Results: In the first model, controlling for individual factors not influenced by school, smoking bans for students and evidence-based prevention activities were negatively associated with smoking prevalence. The second model included student characteristics potentially influenced by school (e.g. school engagement, peer smoking). As school variables remained significant, these characteristics obviously do not mediate school context effects. Conclusion: School context variables influence students’ smoking independent of individual factors. The data suggest that individual variables rather mediate family influences.

No abstract