In our study, 59.9% of the children were reported ETS exposure at home. In the United States the prevalence of children living in homes with a smoker has been estimated to be 43%, with state specific estimates of exposure at home ranging from 12% to 34% [5
]. Thirty-five percent of children in the United States live in homes in which residents or visitors smoke on a regular basis [17
]. Similarly, about 43% of Australian children [18
], 33% of Canadian children [19
], 41% of British children [20
], 59.2% of the students in Mexico [21
] and 89.0% of Turkish children [22
] are exposed to environmental tobacco smoke.
In the present study, 89 (42.7%) of the students pointed out both parents smoking at home, while 80 (38.4%) reported only their father, 26 (12.7%) reported 3 or more household members and 13 (6.2%) reported only their mother. In a study conducted by Celik et al. [23
], it was found that 12.6% of the mothers, 49.2% of fathers, and overall in 42.2% of the families, one or more persons smoked at home.
We found that mothers with higher education levels were more likely to smoke at home, although not statistically significant. In another study conducted in Kahramanmaras, Celik et al. [23
] showed that rates of homes with at least one smoking household member and rates of smoking mothers were positively correlated with the educational levels of mothers. Moreover, in a study representing whole the Turkish population, it was reported that smoking more than 10 cigarettes was most common among women with at least a high school education (45%) [24
]. The fact that studies in Western countries reveal a negative correlation between women's education levels and smoking status, unlike studies from Turkey, may be because of the conservative socio-cultural environment of the country [25
Socioeconomic factors also are known to be related to cotinine levels. Parental education and family income both may be indicators of the prevalence of smoking in the community in which the child lives and plays [27
]. The effect of maternal smoking on child salivary cotinine level has been reported to be greater than the effect of paternal smoking in England and Wales, especially with high levels of cigarette consumption [29
]. Housing characteristics have also been previously described as being associated with cotinine levels, with smaller homes predicting higher levels among smoke-exposed children [28
]. In our study, the proportion of self-reported exposure did not differ significantly between sexes, grades or parents' educational status, as well as between urinary cotinine levels and number of cigarettes exposed to and the daily duration of exposure. In contrast to our study, it has been reported by 3 different studies that the number of cigarettes that parents smoke is a major determinant of salivary cotinine concentration in children [27
]. This contradiction may be attributable to the choice of method in measuring cotinine. Consistent with our results, Smyth et al. [31
] have reported that salivary cotinine was more closely related to family smoking behavior than urinary cotinine concentrations.
Most estimates of the exposure of infants and young children to tobacco smoke are based on adult reports. Self-reports have the advantage of low cost and ease of administration, but raise questions of reliability and validity. Brunekreef et al. [32
] emphasize that underreporting of ETS exposure by parents of study children varies, and may depend on the instrument used, population studied, age, and symptom status, underlining the need for questionnaire validation in specific study settings. In addition to urinary cotinine tests, we used a questionnaire to assess subjects' ETS exposure levels. Besides those with 33 (9.5%) of the students, all of the others cotinine test results were consistent with the self-reports. In interpreting our results, the age group of our subjects and the fact that they are private school students must be considered.
In our study, mean urinary cotinine levels of two intervention groups measured nine months after the initial urinary tests were similar, and in both groups, urinary cotinine levels of the majority of the students were below 10 ng/ml (Figure ). We suggest that a brief intervention incorporating feedback of children's urinary cotinine levels via a letter would display a similar positive effect on parents' attitude on restricting smoking at home as feedback plus intensive counselling. In another study with a similar design to ours, Wakefield et al. [33
] reported no significant change in smoking habits of parents between intervention (written and verbal feedback about child's urinary cotinine level) and control (usual advice about smoking) groups. Their study differed from ours in that children in their intervention group were predominantly from low-income families, and 69.5% of mothers were smoking at home. Three other studies concerning children's exposure to ETS at home have revealed that counselling and advice, without feedback on children's urinary cotinine levels, did not change the children's exposure to ETS [27
]. In this context, Hovell et al.'s study [15
] is an exception, in which authors reported that an intervention involving more intensive counselling was associated with significant decreases in children's exposure to ETS. The results of this study were obtained with a group of low-income families, in which families were paid as an incentive to participate.