Youth experiences of secondhand smoke exposure in New Zealand: evidence from 5 national surveys (2000 to 2008)

People are regularly exposed to secondhand smoke (SHS). Worldwide in 2004 it was estimated that 40% of children aged 0-to-14-years, and 33% of non-smoking adult males and 35% of non-smoking adult females were regularly exposed to SHS indoors, resulting in an estimated 603,000 deaths and 10·9 million disability-adjusted life-years (DALYs).1In New Zealand (NZ), SHS has been estimated to kill around 300 people per year,2-4 and cause a substantial burden of morbidity, particularly for children.5 The harmful effects of SHS in young people have been extensively researched, showing that exposure increases the risk of respiratory illnesses, ear problems, asthma, lung function,6-9 and, more recently, poorer mental health.10 In NZ, a greater risk of exposure has been found among low income individuals and for Māori.11, 12Given the association between SHS exposure and poorer health, it is important to monitor SHS exposure and examine trends over time. Decreases in SHS exposure have been seen both overseas13 and in NZ,12but SHS is still thought of as one of the "most common indoor pollutants worldwide" (p.144).1Despite reductions in overall exposure, there are still some areas where SHS may impact young people, in particular smoking in the home and smoking in vehicles. These are the most significant sites of SHS exposure for most children in the USA,14 and in NZ, 10 to 14% of young people 15-to-19-years were exposed to SHS in their home in 2009, significantly higher than older age groups. A similar number have also been exposed to SHS in a vehicle in the past week.15Efforts to reduce the harm caused by SHS have included both legislative as well as voluntary policies. With regard to legislation, in NZ the Smoke-free Environments Act was introduced in 1990 with subsequent amendments made in 2003 banning smoking in public places including work places, restaurants and bars.These legislative smokefree policies have been successful at reducing SHS exposure and improving indoor air quality.16 Furthermore, social marketing campaigns have fostered voluntary adoption of smokefree homes and cars policies by the public. Data show that these campaigns have been successful in changing behaviour in both NZ,17,29,30 as well as internationally.9 The main benefit is reduced SHS exposure for those living in the home.6,14 However, additional benefits for young people exist including reduced experimentation with smoking,23 and lower likelihood of smoking uptake.17 These findings appear to be stronger for strict smoking bans compared with partial bans, consistent with a recent review of the effect of home smoking restrictions on youth behaviour.17Research in NZ suggests that there is support for smokefree homes.18,19 There is also support for banning smoking in cars,20,21 but this has not been given a high priority by policy makers.22,23 Recently the NZ Māori Affairs Select Committee published their recommendations to the Government on achieving the goal of NZ being smokefree by 2025, and emphasising the extension of the Smoke-free Environments Act to include banning smoking in vehicles, particularly those carrying children.24The aims of this study were to describe trends in young people's exposure to SHS, and to their exposure to household rules around smoking in NZ over the period 2000 to 2008.Methods Sample selection—The Youth Lifestyle Study (YLS) and Youth In-depth Survey (YIS) used methods and key measures from the international Global Youth Tobacco Survey (GYTS), developed by the World Health Organization and the Centers for Disease Control and Prevention to monitor tobacco use among youth across countries.25 The study data came from the 2000, 2002 and 2004 YLS surveys, and the 2006 and 2008 YIS surveys of Year 10 high school students from randomly selected secondary schools in NZ, undertaken by the Health Sponsorship Council. A two-stage cluster sample design, with random selection of participating classes was used. The sampling strategy was intended to result in a representative sample of Year 10 school students with survey weights at the individual student level used for this purpose. More details are contained elsewhere.26-28 The survey used a self-report questionnaire for students administered during school class time. The response rate for schools was 80.5% and from students was 84.9%, giving an overall response rate of 68.3% for 2008.27The response rate for schools was 78.0% and from students of 83.7, giving an overall response rate of 65.3% in 2006.27 The overall response rate in 2004 was 74.6%29 and in 2002 was 58.2%.30 Table 1 describes the sample of students taking part each year. Ethical approval for this study was given by the Ministry of Health's multi-regional ethics committee in 2008. Measures—The YLS and YIS surveys assessed self-reported smoking attitudes, behaviours, and knowledge and information on youth culture and lifestyle. Student age, sex and ethnicity data were also collected. Exposure to SHS around young people in their home was examined through the question: During the past 7 days, on how many days have people around you smoked in your home? From 2004 participants were also asked who the people were that had been smoking around them in their home. The response options in the 2004 survey included: mother; father; best friend; brothers and sisters; family friends; other relatives or caregivers. The two more recent surveys included grandparents and referred to older brothers and sisters. Smoking around young people in vehicles was examined through the following two questions in 2006 and 2008: During the past 7 days, did anyone smoke in your presence while you were travelling in cars or vans?Participants were also asked: During the past 7 days, which of the following people have smoked around you while you were travelling in cars or vans? The list of options was the same for smoking exposure. To assess what rules were in place around smoking at home, young people were asked: At your home is smoking allowed anywhere inside, only in set areas, or nowhere inside your home? A similar question was asked for smoking outside on the property. Analysis—Descriptive statistics are provided by year of survey for all variables, including both sample characteristics and key measures. Regression models were used to investigate changes in key measures over time, adjusting for demographic variables (sex; age groups of: up to and including 13, 14, 15, and 16 and older; and ethnicity using prioritised ethnicity with the order of priority being from highest to lowest: Māori, Pacific Peoples, Asian, Other, and NZ European) as predictors of interest in themselves and to reflect changes in sample characteristics over time when looking at trends. Ordinal logistic regression was used to model days exposed to SHS at home and elsewhere and household smoking rules with outcomes dichotomised when proportionality did not hold. Exposure to SHS in vehicles, specific people contributing to SHS, and other specified locations where SHS exposure has occurred were modelled using binary logistic regression. Interactions between survey year and other predictors (including household rules) were investigated and retained where statistically significant. All analyses included a random effect for school to model cluster effects. All analyses were performed using Stata v11.1 software.31 All significance tests were two-sided, with p<0.05 considered statistically significant. Adjusted odds ratios (aORs) and 95% confidence interval (CI) are presented. Table 1. Demographic characteristics of total sample, 2000-2008 (percentages) Variables Total sample 2000 n=1610 2002 n=2756 2004 n=3400 2006 n=3200 2008 n=3066 Sex Males Females 47.8 52.2 49.2 50.8 50.7 49.3 49.3 50.7 50.8 49.2 Age (years) <14 14 15 16+ 0.27 56.4 39.4 3.9 1.4 74.6 22.7 1.3 1.1 61.7 36.2 1.0 0.6 64.5 34.0 0.81 0.79 61.1 37.0 1.2 Ethnicity NZE* Māori Pacific Asian Other 58.5 17.2 10.8 4.8 8.6 56.2 20.6 10.5 6.9 5.8 48.5 24.9 11.6 7.6 7.4 57.1 20.8 8.6 4.7 8.8 52.2 25.0 11.8 9.0 2.0 * NZE represents New Zealand European. Results Sample characteristics—The total sample consisted of approximately equal number of males and females, with most participants being 14 years of age and of NZ European ethnicity (see Table 1 above). Exposure to secondhand smoke in the home—Table 2 shows there has been a downward linear trend (adjusted linear trend p<0.001) in SHS exposure for young people in their home from 49% in 2000 to 35% in 2008. In terms of the number of days exposed, with the exception of 5-6 days, the percentage decreases seem proportional at almost all levels which appears to support the hypothesis that exposure is not shifting to a lower category as much as it is ceasing altogether (Figure 1). Table 2. Exposure to SHS in the home, vehicles, and places other than their home in past week, and household smoking rules (%) Variables Year of survey P value 2000 2002 2004 2006 2008 SHS exposure In home In vehicles 48.6 46.5 45.7 42.1 32.9 29.0 34.9 30.9 p<0.001 p<0.001 Is smoking allowed inside your home? Anywhere inside In set inside areas Nowhere inside 17.1 12.3 70.6 9.3 10.7 80.0 7.0 10.4 82.6 p<0.001 Is smoking allowed outside your home? Anywhere outside In set outside areas Nowhere outside 71.4 10.7 17.9 60.5 11.5 28.1 55.8 12.6 31.6 p<0.001 Smokefree home and property Total smoking ban 17.2 27.5 30.9 p<0.001 Note: empty cells indicate years when that question was not asked. Females were more likely (aOR 1.1, CI: 1.1 to 1.2, p=0.001) to be exposed to SHS in their home than males. Overall, ethnicity was a significant predictor over all years (overall p<0.001) with Māori (aOR 3.2, CI: 2.9 to 3.5, p<0.001) and Pacific (aOR 2.0, CI: 1.7 to 2.3, p<0.001) young people more likely to be exposed to SHS in their home compared with NZ European (Table 3). Those identifying as Asian were less likely (aOR 0.51, CI: 0.43 to 0.61, p<0.001) to be exposed to SHS in their home than NZ European. Participants reported who had smoked around them in their home in 2008. Mothers were reported as smoking around them most often at 46%, and one-third (35%) had fathers who smoked around them in their homes. One-quarter (25%) of older brothers and 19% of older sisters were also reported to have smoked around young people in their homes. In terms of friends, 16% of best friends and 21% of other close friends were reported as smoking in the presence of the participants in their home. Figure 1. Percentage of young people exposed to SHS in their home 2000-2008 for numbers of days per week (excluding exposures on no days) Exposure to secondhand smoke while travelling in vehicles—There was a downward linear trend in young people's exposure to SHS in vehicles from 42% in 2004 to 31% in 2008 (adjusted linear trend p<0.001) (Table 2). Females were more likely to be exposed to SHS in vehicles than males (aOR 1.3, CI: 1.1 to 1.4, p<0.001). Overall, ethnicity was a significant predictor over all years (overall p<0.001) with Māori (aOR 3.1 CI: 2.7 to 3.5, p<0.001) and Pacific (aOR 2.3 CI: 1.9 to 2.7, p<0.001) young people more likely to be exposed to SHS in vehicles compared with NZ European. Those identifying as Asian were less likely (aOR 0.55, CI: 0.43 to0 .70, p<0.001) to be exposed to SHS in vehicles than NZ European (Table 3). Overall, age was a significant predictor over all years (overall p<0.001) with higher SHS exposure in vehicles for those 16 years and older compared with students under the age of 14 years (aOR 2.5, CI: 1.4 to 4.5, p=0.003). As with smoking in homes, parents were the most significant contributors of SHS in vehicles; 41% of mothers and 33% of fathers smoked around young people in vehicles in the past week in 2008. One-fifth (19%) of older brothers and 18% of older sisters were also reported to have smoked around young people in vehicles. A significant source of SHS exposure for young people in vehicles was people other than family members; one-third (32%) reported other people, such as visitors, smoked around them in vehicles and one-quarter (26%) of respondents had a family friend smoke in their presence while travelling in vehicles. In terms of friends, 16% of best friends and 22% of other close friends were reported as smoking in the presence of the participants in vehicles in 2008. Household rules around smoking—There is evidence that both unrestricted indoor and outdoor smoking has declined between 2004 and 2008 (adjusted linear trends p<0.001 for both). Combining these, there is evidence of an increase in homes and properties being entirely smokefree from 17% in 2004 to 31% in 2008 (adjusted linear trend p<0.001). Females were less likely to live at a smokefree