- Split View
-
Views
-
Cite
Cite
Linda Ringlever, Marieke Hiemstra, Rutger C. M. E. Engels, Onno C. P. van Schayck, Roy Otten, Engaging parents of children with and without asthma in smoking-specific parenting: results from a 3-year Randomized Controlled Trial evaluation, Health Education Research, Volume 31, Issue 6, December 2016, Pages 760–770, https://doi.org/10.1093/her/cyw050
- Share Icon Share
The present study evaluated long-term effects of a home-based smoking prevention program targeting smoking-specific parenting in families with children with and without asthma. A total of 1398 non-smoking children (mean age 10.1) participated, of which 197 (14.1%) were diagnosed with asthma. Families were blinded to group assignment. The intervention group (n = 684) received booklets with assignments that actively encouraged parents to engage in smoking-specific parenting strategies. Control families (n = 714) received booklets containing basic information about youth smoking. Latent growth curve modeling was used to calculate intercepts and slopes to examine whether there was change in the different parenting aspects over the study period. Regression analyses were used to examine whether a possible change was different for intervention and control condition families with and without a child with asthma. For those smoking-specific parenting aspects that changed over time, families in the intervention and control condition increased similarly. Families with a child with asthma did not engage in parenting at higher levels due to the intervention program than parents of non-asthmatic children. This prevention program did not affect smoking-specific parenting in the Netherlands. Future prevention research could focus on other risk factors for smoking initiation among adolescents with asthma.
Trial registration: Netherlands Trial Register NTR1465.
Introduction
Adolescents with asthma are as likely or even more likely to engage in smoking, when compared with their peers without asthma [1]. Specific health consequences associated with asthma and smoking include less responsiveness to asthma medication, and the potential to develop chronic obstructive pulmonary disease [2]. Risk factors for smoking onset among young children have been studied widely and involve demographic, peer, family and intrapersonal influences [3–5]. Research specifically focusing on risk factors for youth with asthma showed that adolescents with asthma share similar risk factors for smoking [6], among which family smoking and parental approval of smoking. Prevention programs could target these parental factors as the involvement of parents in smoking prevention programs for children has been suggested as an effective strategy [7, 8]. The present study examined the effectiveness of a smoking prevention program targeting smoking-specific parenting strategies used by parents of non-smoking children with and without asthma.
Smoking-specific parenting
Smoking-specific parenting concerns all parenting strategies and interactions influencing the development of children’s cognitions and norms against smoking [9]. In the last decade, the concept of smoking-specific parenting has received more attention in the literature [10–15] as specific parenting is most observable [16] and might be easier to modify than the more general parenting style or goals. Examples of smoking-specific parenting include communicating about smoking and setting house rules. Associations with adolescent smoking have been frequently found [10, 11]; however, hardly any studies concentrated on smoking-specific parenting in families with a child with asthma. Otten et al. [15] examined five different aspects of smoking-specific parenting and found that adolescents with asthma reported their parents to talk more often about non-smoking [15]. Two other studies found no difference between families with and without a child with asthma for parental approval of smoking [6, 17].
It is unknown whether it is effective to encourage parents of children with asthma to engage in smoking-specific parenting. It is conceivable that parents who have children with asthma are more willing to complete a family-based smoking prevention program when offered in order to prevent their offspring smoking. Therefore, the aim of the present study is to evaluate whether the Randomized Controlled Trial (RCT) ‘Smoke-free Kids’ activated parents’ antismoking socialization efforts. The Dutch Smoke-free Kids is based on the original version found to be effective among smoking parents in the United States [12]. Although we did not find long-term effects after 24 and 36 months on smoking onset [18], short-term results on hypothesized mechanisms were found in the Netherlands as well [19]. Therefore, we aim to test what the effects of the program are on the development of smoking-specific parenting practices over the course of 3 years. Because of the troubling finding that children with asthma often engage in smoking irrespective of the increased associated health risks, it is important to evaluate this prevention program by comparing the effectiveness on parenting between families with and without a child with asthma.
Method
Procedure
Inclusion criteria for the Smoke-free Kids trial in the Netherlands were (i) only one child per household providing data, aged between 9 and 11 years old, (ii) the participating parent had to be the mother or the female guardian and (iii) both mother and child had to be able to read and speak Dutch. The main recruitment strategy was the distribution of recruitment letters via primary schools. For the remaining, health professionals were asked to place information posters and flyers in their waiting rooms and announcements were placed in local newspapers and health related websites. After baseline assessment participants were randomized. To avoid contamination between the intervention and the control conditions, randomization took place at school level. An independent statistician performed the randomization and stratified for school and asthma. For more details, see our RCT-registration [Dutch Trial Register NTR1465] and Hiemstra et al. [20]. The study was approved by the ethical committee of the Faculty of Social Sciences at Radboud University Nijmegen.
Intervention
Participants
Power calculations indicated that a minimum of 856 families was necessary to evaluate the Smoke-free Kids trial [20]. To examine whether the effects of Smoke-free Kids were different for children with asthma when compared with children without asthma, a required minimum of 200 children with asthma or asthma symptoms was estimated. Eventually, the study was over-enrolled with 1478 mothers and children providing baseline data. At baseline 80 children (5.4%) reported to have ever smoked a puff of a cigarette and this percentage was higher among children with asthma [21]. As smoking initiation was the primary outcome of Smoke-free Kids to prevent, these children were excluded from all analyses evaluating the effectiveness of Smoke-free Kids. Therefore, 1398 mothers and their children participated in this study. The intervention condition comprised of more girls (52.7%) than the control condition [47.3%, χ2(1,1398) = 8.61, P = 0.003] and included more children with asthma (16.1%) than the control condition [12.2%, χ2(1,1394) = 4.30, P = 0.04]. Children were aged between 8 and 12 years old at baseline measurement (M = 10.11, SD = 0.79).
Measures
Covariates were child’s age and gender, parental education and parental smoking. Parental education was measured at baseline on a 9-point scale ranging from 1 = primary school to 9 = university. Mother reported about the educational level of father and mother. Answers were recoded into lower- (i.e. up to preparatory secondary school for vocational education), middle- (i.e. intermediate vocational education and pre-university education) and higher-educated (i.e. higher professional or university education) fathers and mothers. For parental smoking, children reported at baseline whether their father or mother currently smoked, never smoked or had smoked in the past. A combined variable was created with three categories (both parents non-daily smokers, one parent daily smoker, both parents daily smokers) [15].
Asthma Items from the parent version of the International Study of Asthma and Allergies in Childhood (ISAAC) questionnaire were used to screen for asthma [22, 23]. Mothers indicated whether their child ever had asthma and whether asthma was ever diagnosed by a physician. Only if both answers were positive, these children were categorized as having asthma. Four mothers indicated their child to have asthma, but did not indicate a diagnosis made by a physician. Therefore, these four families were not included in the analyses.
Smoking-specific parenting strategies
Frequency of communication Frequency of smoking-specific communication between mother and child was measured with seven items (e.g. during the last 12 months, how often did your mother talk to you about how to resist peer pressure to use tobacco use?) [11,14]. Response options were 1= never, 2 = sometimes or 3 = a lot. Higher mean scores on this scale represent a higher frequency of communication between mother and child. αs were 0.79, 0.77, 0.82, 0.81 and 0.79, respectively.
Quality of communication Quality of smoking-specific communication between mother and child was measured with six items [11]. An example item is ‘My mother and I are interested in each other’s opinion about smoking’. Children could respond whether these statements were 1 = not true, 2 = sometimes true or 3 = true. Higher mean scores on this scale represent a higher quality of communication between mother and child. Cronbach’s α were 0.64, 0.65, 0.68, 0.66 and 0.71 for the subsequent time points.
The following smoking-specific parenting concepts are all measured by separate items:
Agreement assessed whether children had a non-smoking agreement with their parents (1 = no, 2 = yes).
Availability assessed how often cigarettes were available for children in their homes (1 = all the time, 2 = sometimes, 3 = never).
Influence inquired whether the child thinks his/her mother can have influence on their smoking behavior (1 = totally disagree, 2 = disagree, 3 = agree, 4 = totally agree).
House rule inside concerns the rule that only adults are allowed to smoke at home, not children (1 = not true, 2 = sometimes true, 3 = true).
House rule outside assessed whether smoking is only allowed outside (1 = not true, 2 = sometimes true, 3 = true).
Conflict measured the child’s perception that his/her mother would be angry as a reaction to smoking behavior by the child (1 = not true, 2 = maybe, 3 = true).
Disappointment measured the child’s perception that his/her mother would be disappointed as a reaction to smoking behavior by the child (1 = not true, 2 = maybe, 3 = true).
Withdrawal measured the child’s perception that his/her mother would have no difficulty with his/her smoking when she found out her child would be smoking (1 = not true, 2 = maybe, 3 = true).
Statistical analysis
First, descriptive and attrition findings are provided. Second, development of the different anti-smoking parenting concepts was modeled by estimating separate latent growth curves. Latent growth curve modeling permits to capture not only the initial levels of individuals at the beginning of a developmental period, but also individual changes over a developmental period [24]. In the present study, intercepts represented initial levels of a specific parenting construct while the slopes represented the rates of change in the same constructs. Mplus provides both the means and the variances of the intercept and the slope. If the slope growth factor mean is significant, this means that it is significantly different from zero, which indicates development over time on average. If the slope growth factor variance is significant, this indicates that not all individuals grow at the same rate, but that there is significant variability in their growth rates. For the purpose of this study, we aimed to establish whether there was change over time, and whether this change could be predicted by condition or asthma. Those parenting concepts that showed significant growth over time were included in the next step. In that step, we tested whether the different growth curves could be predicted by regressing the intercepts and the slopes one by one on the condition (i.e. intervention, control) and on asthma status, while controlling for the aforementioned set of covariates. In a final step, we examined whether the effects were different for adolescents with and without asthma by including an interaction term. As we tested many interactions, we took a more conservative approach and used a significance level of P < 0.01. The Satorra–Bentler Scaled χ2 difference test was used to test model comparisons [25]. To determine model fit we used the comparative fit index (CFI, critical value ≥0.90) [26], the Tucker Lewis Index (TLI, critical value ≥0.90) [27] and the root mean squared estimate of approximation (RMSEA, critical value ≤0.08) [28].
Results
Descriptive analyses
The attrition rate at final follow-up was 11.4% (Fig. 1). Dropouts did not differ from those still participating after 36 months on age, gender or asthma status. Dropouts were more often children from lower socio-economic status (SES) families [dropout increased from 8.0% among higher-educated mothers to 16.7% among lower educated mothers, χ2(1,1470) = 14.39, P < 0.001, and from 9.2 to 13.6% for paternal education, χ2(1,1437) = 5.65, P = 0.059]. Less children dropped out who had two non-smoking parents (9.7%) than children with one (14.3%) or two (18.1%) smoking parents [χ2(1,1478) = 12.26, P = 0.002]. Finally, more children from the intervention than from the control condition dropped out [15.4% versus 8.1%, χ2(1,1478) = 18.80, P < 0.001]. More detailed analyses showed that this difference appeared directly after offering the intervention material to the families (Fig. 1).
More boys (17.6%) than girls (11.0%) ever had asthma [χ2(1,1394) = 12.25, P < 0.001] and children with asthma were slightly older than their peers without asthma [mean age 10.2 versus 10.1 years, t(1390) = −2.236, P = 0.04]. There were no differences in SES or parental smoking.
Table I provides the means (SD) for all smoking-specific parenting measures at baseline. Children with asthma only differed in their reports on parenting with regard to a higher reported frequency of communication than their peers without asthma.
. | No asthma (n = 1197) . | Asthma (n = 197) . | Total (n = 1394) . | Significant difference no asthma and asthma . |
---|---|---|---|---|
. | Mean (SD) . | Mean (SD) . | Mean (SD) . | P values . |
Agreement | 1.21 (0.41) | 1.22 (0.42) | 1.21 (0.41) | 0.66 |
Availability | 2.53 (0.75) | 2.62 (0.69) | 2.55 (0.74) | 0.09 |
Influence | 3.32 (0.74) | 3.24 (0.80) | 3.31 (0.75) | 0.22 |
House rule inside | 2.08 (0.90) | 2.01 (0.94) | 2.07 (0.91) | 0.34 |
House rule outside | 2.43 (0.83) | 2.43 (0.83) | 2.43 (0.83) | 0.96 |
Conflict | 2.41 (0.78) | 2.43 (0.77) | 2.41 (0.78) | 0.69 |
Disappointment | 2.77 (0.53) | 2.81 (0.50) | 2.78 (0.53) | 0.43 |
Withdrawal | 2.83 (0.48) | 2.80 (0.53) | 2.82 (0.49) | 0.39 |
Frequency communication | 1.75 (0.45) | 1.83 (0.47) | 1.76 (0.45) | 0.02 |
Quality communication | 2.67 (0.35) | 2.67 (0.33) | 2.67 (0.34) | 0.91 |
. | No asthma (n = 1197) . | Asthma (n = 197) . | Total (n = 1394) . | Significant difference no asthma and asthma . |
---|---|---|---|---|
. | Mean (SD) . | Mean (SD) . | Mean (SD) . | P values . |
Agreement | 1.21 (0.41) | 1.22 (0.42) | 1.21 (0.41) | 0.66 |
Availability | 2.53 (0.75) | 2.62 (0.69) | 2.55 (0.74) | 0.09 |
Influence | 3.32 (0.74) | 3.24 (0.80) | 3.31 (0.75) | 0.22 |
House rule inside | 2.08 (0.90) | 2.01 (0.94) | 2.07 (0.91) | 0.34 |
House rule outside | 2.43 (0.83) | 2.43 (0.83) | 2.43 (0.83) | 0.96 |
Conflict | 2.41 (0.78) | 2.43 (0.77) | 2.41 (0.78) | 0.69 |
Disappointment | 2.77 (0.53) | 2.81 (0.50) | 2.78 (0.53) | 0.43 |
Withdrawal | 2.83 (0.48) | 2.80 (0.53) | 2.82 (0.49) | 0.39 |
Frequency communication | 1.75 (0.45) | 1.83 (0.47) | 1.76 (0.45) | 0.02 |
Quality communication | 2.67 (0.35) | 2.67 (0.33) | 2.67 (0.34) | 0.91 |
. | No asthma (n = 1197) . | Asthma (n = 197) . | Total (n = 1394) . | Significant difference no asthma and asthma . |
---|---|---|---|---|
. | Mean (SD) . | Mean (SD) . | Mean (SD) . | P values . |
Agreement | 1.21 (0.41) | 1.22 (0.42) | 1.21 (0.41) | 0.66 |
Availability | 2.53 (0.75) | 2.62 (0.69) | 2.55 (0.74) | 0.09 |
Influence | 3.32 (0.74) | 3.24 (0.80) | 3.31 (0.75) | 0.22 |
House rule inside | 2.08 (0.90) | 2.01 (0.94) | 2.07 (0.91) | 0.34 |
House rule outside | 2.43 (0.83) | 2.43 (0.83) | 2.43 (0.83) | 0.96 |
Conflict | 2.41 (0.78) | 2.43 (0.77) | 2.41 (0.78) | 0.69 |
Disappointment | 2.77 (0.53) | 2.81 (0.50) | 2.78 (0.53) | 0.43 |
Withdrawal | 2.83 (0.48) | 2.80 (0.53) | 2.82 (0.49) | 0.39 |
Frequency communication | 1.75 (0.45) | 1.83 (0.47) | 1.76 (0.45) | 0.02 |
Quality communication | 2.67 (0.35) | 2.67 (0.33) | 2.67 (0.34) | 0.91 |
. | No asthma (n = 1197) . | Asthma (n = 197) . | Total (n = 1394) . | Significant difference no asthma and asthma . |
---|---|---|---|---|
. | Mean (SD) . | Mean (SD) . | Mean (SD) . | P values . |
Agreement | 1.21 (0.41) | 1.22 (0.42) | 1.21 (0.41) | 0.66 |
Availability | 2.53 (0.75) | 2.62 (0.69) | 2.55 (0.74) | 0.09 |
Influence | 3.32 (0.74) | 3.24 (0.80) | 3.31 (0.75) | 0.22 |
House rule inside | 2.08 (0.90) | 2.01 (0.94) | 2.07 (0.91) | 0.34 |
House rule outside | 2.43 (0.83) | 2.43 (0.83) | 2.43 (0.83) | 0.96 |
Conflict | 2.41 (0.78) | 2.43 (0.77) | 2.41 (0.78) | 0.69 |
Disappointment | 2.77 (0.53) | 2.81 (0.50) | 2.78 (0.53) | 0.43 |
Withdrawal | 2.83 (0.48) | 2.80 (0.53) | 2.82 (0.49) | 0.39 |
Frequency communication | 1.75 (0.45) | 1.83 (0.47) | 1.76 (0.45) | 0.02 |
Quality communication | 2.67 (0.35) | 2.67 (0.33) | 2.67 (0.34) | 0.91 |
Latent growth curves
Latent growth curves were estimated for each of the anti-smoking socialization constructs using Mplus 6.0 [24]. Model fit indices (i.e. CFI, TLI, RMSEA) for all latent growth models were good (for CFI and TLI > 0.90; for RMSEA <0.07). Expected influence (Meanslope = 0.25, P = 0.00), house rules outside (Meanslope = 0.65, P = 0.00), and disappointment (Meanslope = 0.41, P = 0.01) did show an increase over time. For house rules inside, frequency and quality of communication descriptive statistics as well as the Chi-Square Difference Test [25] indicated a non-linear trend. Specifically, for house rules inside (‘adults are allowed to smoking inside, not children’: 1 = not true, 2 = sometimes true, 3 = true), an initial increase is followed by decrease (Meanslope = 0.12; Meanquadratic trend = −0.21). Regarding frequency of communication, statistics showed that communication about smoking occurred more frequently during the intervention than before and after the intervention (Meanslope = 0.78; Meanquadratic trend = −0.57). Quality of communication was rated higher during the intervention than before and after the intervention (Meanslope = 0.86; Meanquadratic trend = −0.61). No significant overall change over time (slope) was found for agreement (Meanslope = 0.13, P = 0.14), availability (Meanslope = −0.22, P = 0.12), conflict (Meanslope = −0.02, P = 0.76) and withdrawal (Meanslope = −0.11, P = 0.58), so these parenting concepts were not included in the regression analyses in the next step.
Tables II and III show the results of six separate regressions in which we tested the effects of asthma, the intervention versus control condition, and the interaction (asthma × condition) on the intercepts and slopes for the aforementioned significant anti-smoking socialization concepts, while controlling for the covariates. (In order to save space we did not show the results of the covariates. Findings can be obtained from the first author.) In some of the regressions there was an effect of asthma or the condition on the intercept or the slope, although effects of condition on the intercept could only be subscribed to coincidence. The slope of disappointment was significant, showing an overall tendency for children to expect their mothers to react more disappointed if they would find out smoking. Having asthma was predictive of a less steeper increase in expected disappointment over time (Bslope = −0.18). Regarding frequency and quality of communication, we found support for a quadratic trend, indicating an increase followed by a decrease. Being in the control condition was associated with lower rates of frequency of communication (Bslope = −0.10). In families with a child with asthma the decrease in frequency of communication was less steep (Bquadratic = −0.10). Finally, being in the control condition was associated with a less steep increase in quality of communication over time (Bslope = −0.23) and with a stronger decrease after the intervention (Bquadratic = 0.17). However, what stands out are the overall non-significant interactions of asthma on the relationship between condition and changes in smoking-specific parenting, indicating that the effects of the intervention on smoking-specific parenting aspects were similar for adolescents with and without asthma. (We also took a less conservative approach by running the regression models with those parenting concepts that had not shown a significant change over time as outcome, but with significant variance. However, then also no significant effects could be identified for condition, asthma, or the interplay between these two.)
Aspects of anti-smoking socialization . | Independent variables . | Intercept . | Slope . | |||||||
---|---|---|---|---|---|---|---|---|---|---|
B . | b . | SE . | P . | B . | b . | SE . | P . | |||
Influence | Step 1 | Condition | −0.04 | −0.04 | 0.05 | 0.43 | 0.01 | 0.00 | 0.02 | 0.89 |
Asthma | −0.06 | −0.09 | 0.07 | 0.24 | 0.11 | 0.05 | 0.03 | 0.10 | ||
Step 2 | Interaction | −0.30 | 0.22 | 0.16 | 0.16 | −0.36 | −0.08 | 0.06 | 0.19 | |
Houserule outside | Step 1 | Condition | −0.04 | −0.08 | 0.06 | 0.21 | 0.10 | 0.00 | 0.02 | 0.86 |
Asthma | 0.05 | 0.14 | 0.09 | 0.10 | −0.07 | −0.04 | 0.02 | 0.07 | ||
Step 2 | Interaction | −0.11 | −0.14 | 0.18 | 0.43 | −0.05 | −0.01 | 0.05 | 0.84 | |
Disappointment | Step 1 | Condition | 0.04 | 0.06 | 0.07 | 0.41 | −0.03 | −0.01 | 0.02 | 0.74 |
Asthma | 0.09 | 0.18 | 0.10 | 0.08 | −0.18 | −0.08 | 0.03 | 0.02 | ||
Step 2 | Interaction | −0.04 | −0.03 | 0.19 | 0.87 | −0.20 | −0.04 | 0.06 | 0.53 |
Aspects of anti-smoking socialization . | Independent variables . | Intercept . | Slope . | |||||||
---|---|---|---|---|---|---|---|---|---|---|
B . | b . | SE . | P . | B . | b . | SE . | P . | |||
Influence | Step 1 | Condition | −0.04 | −0.04 | 0.05 | 0.43 | 0.01 | 0.00 | 0.02 | 0.89 |
Asthma | −0.06 | −0.09 | 0.07 | 0.24 | 0.11 | 0.05 | 0.03 | 0.10 | ||
Step 2 | Interaction | −0.30 | 0.22 | 0.16 | 0.16 | −0.36 | −0.08 | 0.06 | 0.19 | |
Houserule outside | Step 1 | Condition | −0.04 | −0.08 | 0.06 | 0.21 | 0.10 | 0.00 | 0.02 | 0.86 |
Asthma | 0.05 | 0.14 | 0.09 | 0.10 | −0.07 | −0.04 | 0.02 | 0.07 | ||
Step 2 | Interaction | −0.11 | −0.14 | 0.18 | 0.43 | −0.05 | −0.01 | 0.05 | 0.84 | |
Disappointment | Step 1 | Condition | 0.04 | 0.06 | 0.07 | 0.41 | −0.03 | −0.01 | 0.02 | 0.74 |
Asthma | 0.09 | 0.18 | 0.10 | 0.08 | −0.18 | −0.08 | 0.03 | 0.02 | ||
Step 2 | Interaction | −0.04 | −0.03 | 0.19 | 0.87 | −0.20 | −0.04 | 0.06 | 0.53 |
Note. Standardized and unstandardized estimates, standard errors and P values of the program effects are shown. In all analyses, we controlled for age, sex, education level father, education level mother and parental smoking. Condition: 1 = intervention condition and 2 = control condition, 1 = no asthma, 2 = asthma.
Aspects of anti-smoking socialization . | Independent variables . | Intercept . | Slope . | |||||||
---|---|---|---|---|---|---|---|---|---|---|
B . | b . | SE . | P . | B . | b . | SE . | P . | |||
Influence | Step 1 | Condition | −0.04 | −0.04 | 0.05 | 0.43 | 0.01 | 0.00 | 0.02 | 0.89 |
Asthma | −0.06 | −0.09 | 0.07 | 0.24 | 0.11 | 0.05 | 0.03 | 0.10 | ||
Step 2 | Interaction | −0.30 | 0.22 | 0.16 | 0.16 | −0.36 | −0.08 | 0.06 | 0.19 | |
Houserule outside | Step 1 | Condition | −0.04 | −0.08 | 0.06 | 0.21 | 0.10 | 0.00 | 0.02 | 0.86 |
Asthma | 0.05 | 0.14 | 0.09 | 0.10 | −0.07 | −0.04 | 0.02 | 0.07 | ||
Step 2 | Interaction | −0.11 | −0.14 | 0.18 | 0.43 | −0.05 | −0.01 | 0.05 | 0.84 | |
Disappointment | Step 1 | Condition | 0.04 | 0.06 | 0.07 | 0.41 | −0.03 | −0.01 | 0.02 | 0.74 |
Asthma | 0.09 | 0.18 | 0.10 | 0.08 | −0.18 | −0.08 | 0.03 | 0.02 | ||
Step 2 | Interaction | −0.04 | −0.03 | 0.19 | 0.87 | −0.20 | −0.04 | 0.06 | 0.53 |
Aspects of anti-smoking socialization . | Independent variables . | Intercept . | Slope . | |||||||
---|---|---|---|---|---|---|---|---|---|---|
B . | b . | SE . | P . | B . | b . | SE . | P . | |||
Influence | Step 1 | Condition | −0.04 | −0.04 | 0.05 | 0.43 | 0.01 | 0.00 | 0.02 | 0.89 |
Asthma | −0.06 | −0.09 | 0.07 | 0.24 | 0.11 | 0.05 | 0.03 | 0.10 | ||
Step 2 | Interaction | −0.30 | 0.22 | 0.16 | 0.16 | −0.36 | −0.08 | 0.06 | 0.19 | |
Houserule outside | Step 1 | Condition | −0.04 | −0.08 | 0.06 | 0.21 | 0.10 | 0.00 | 0.02 | 0.86 |
Asthma | 0.05 | 0.14 | 0.09 | 0.10 | −0.07 | −0.04 | 0.02 | 0.07 | ||
Step 2 | Interaction | −0.11 | −0.14 | 0.18 | 0.43 | −0.05 | −0.01 | 0.05 | 0.84 | |
Disappointment | Step 1 | Condition | 0.04 | 0.06 | 0.07 | 0.41 | −0.03 | −0.01 | 0.02 | 0.74 |
Asthma | 0.09 | 0.18 | 0.10 | 0.08 | −0.18 | −0.08 | 0.03 | 0.02 | ||
Step 2 | Interaction | −0.04 | −0.03 | 0.19 | 0.87 | −0.20 | −0.04 | 0.06 | 0.53 |
Note. Standardized and unstandardized estimates, standard errors and P values of the program effects are shown. In all analyses, we controlled for age, sex, education level father, education level mother and parental smoking. Condition: 1 = intervention condition and 2 = control condition, 1 = no asthma, 2 = asthma.
Smoking-specific parenting . | Independent variables . | Intercept . | Slope . | Quadratic trend . | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
B . | B . | SE . | P . | B . | b . | SE . | P . | B . | b . | SE . | P . | |||
House rule Inside | Step 1 | Condition | 0.08 | 0.01 | 0.07 | 0.88 | −0.01 | 0.02 | 0.09 | 0.85 | 0.01 | 0.00 | 0.02 | 0.85 |
Asthma | −0.08 | −0.15 | 0.09 | 0.11 | 0.02 | 0.04 | 0.013 | 0.77 | −0.01 | 0.00 | 0.03 | 0.99 | ||
Step 2 | Interaction | –a | – | – | – | – | – | – | – | – | – | – | – | |
Frequency communication | Step 1 | Condition | −0.10 | −0.05 | 0.02 | 0.03 | −0.08 | −0.03 | 0.02 | 0.14 | 0.05 | 0.01 | 0.01 | 0.30 |
Asthma | 0.06 | 0.05 | 0.04 | 0.18 | 0.08 | 0.05 | 0.03 | 0.12 | −0.10 | −0.02 | 0.01 | 0.03 | ||
Step 2 | Interaction | −0.34 | −0.12 | 0.07 | 0.09 | 0.17 | 0.05 | 0.07 | 0.46 | −0.06 | −0.01 | 0.01 | 0.77 | |
Quality communication | Step 1 | Condition | 0.05 | 0.02 | 0.02 | 0.25 | −0.23 | −0.04 | 0.02 | 0.03 | 0.17 | 0.01 | 0.00 | 0.03 |
Asthma | .04 | .02 | .03 | 0.37 | 0.01 | 0.00 | 0.02 | 0.95 | 0.04 | 0.00 | 0.01 | 0.57 | ||
Step 2 | Interaction | 0.15 | 0.04 | 0.05 | 0.45 | −0.37 | −0.04 | 0.05 | 0.42 | 0.16 | 0.01 | 0.01 | 0.64 |
Smoking-specific parenting . | Independent variables . | Intercept . | Slope . | Quadratic trend . | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
B . | B . | SE . | P . | B . | b . | SE . | P . | B . | b . | SE . | P . | |||
House rule Inside | Step 1 | Condition | 0.08 | 0.01 | 0.07 | 0.88 | −0.01 | 0.02 | 0.09 | 0.85 | 0.01 | 0.00 | 0.02 | 0.85 |
Asthma | −0.08 | −0.15 | 0.09 | 0.11 | 0.02 | 0.04 | 0.013 | 0.77 | −0.01 | 0.00 | 0.03 | 0.99 | ||
Step 2 | Interaction | –a | – | – | – | – | – | – | – | – | – | – | – | |
Frequency communication | Step 1 | Condition | −0.10 | −0.05 | 0.02 | 0.03 | −0.08 | −0.03 | 0.02 | 0.14 | 0.05 | 0.01 | 0.01 | 0.30 |
Asthma | 0.06 | 0.05 | 0.04 | 0.18 | 0.08 | 0.05 | 0.03 | 0.12 | −0.10 | −0.02 | 0.01 | 0.03 | ||
Step 2 | Interaction | −0.34 | −0.12 | 0.07 | 0.09 | 0.17 | 0.05 | 0.07 | 0.46 | −0.06 | −0.01 | 0.01 | 0.77 | |
Quality communication | Step 1 | Condition | 0.05 | 0.02 | 0.02 | 0.25 | −0.23 | −0.04 | 0.02 | 0.03 | 0.17 | 0.01 | 0.00 | 0.03 |
Asthma | .04 | .02 | .03 | 0.37 | 0.01 | 0.00 | 0.02 | 0.95 | 0.04 | 0.00 | 0.01 | 0.57 | ||
Step 2 | Interaction | 0.15 | 0.04 | 0.05 | 0.45 | −0.37 | −0.04 | 0.05 | 0.42 | 0.16 | 0.01 | 0.01 | 0.64 |
Note. Standardized and unstandardized estimates, standard errors and P-values of the program effects are shown. In all analyses, we controlled for age, sex, education level father, education level mother and parental smoking. Condition: 1 = intervention condition and 2 = control condition, 1 = no asthma, 2 = asthma.
aA model with an interaction term did not convert.
Smoking-specific parenting . | Independent variables . | Intercept . | Slope . | Quadratic trend . | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
B . | B . | SE . | P . | B . | b . | SE . | P . | B . | b . | SE . | P . | |||
House rule Inside | Step 1 | Condition | 0.08 | 0.01 | 0.07 | 0.88 | −0.01 | 0.02 | 0.09 | 0.85 | 0.01 | 0.00 | 0.02 | 0.85 |
Asthma | −0.08 | −0.15 | 0.09 | 0.11 | 0.02 | 0.04 | 0.013 | 0.77 | −0.01 | 0.00 | 0.03 | 0.99 | ||
Step 2 | Interaction | –a | – | – | – | – | – | – | – | – | – | – | – | |
Frequency communication | Step 1 | Condition | −0.10 | −0.05 | 0.02 | 0.03 | −0.08 | −0.03 | 0.02 | 0.14 | 0.05 | 0.01 | 0.01 | 0.30 |
Asthma | 0.06 | 0.05 | 0.04 | 0.18 | 0.08 | 0.05 | 0.03 | 0.12 | −0.10 | −0.02 | 0.01 | 0.03 | ||
Step 2 | Interaction | −0.34 | −0.12 | 0.07 | 0.09 | 0.17 | 0.05 | 0.07 | 0.46 | −0.06 | −0.01 | 0.01 | 0.77 | |
Quality communication | Step 1 | Condition | 0.05 | 0.02 | 0.02 | 0.25 | −0.23 | −0.04 | 0.02 | 0.03 | 0.17 | 0.01 | 0.00 | 0.03 |
Asthma | .04 | .02 | .03 | 0.37 | 0.01 | 0.00 | 0.02 | 0.95 | 0.04 | 0.00 | 0.01 | 0.57 | ||
Step 2 | Interaction | 0.15 | 0.04 | 0.05 | 0.45 | −0.37 | −0.04 | 0.05 | 0.42 | 0.16 | 0.01 | 0.01 | 0.64 |
Smoking-specific parenting . | Independent variables . | Intercept . | Slope . | Quadratic trend . | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
B . | B . | SE . | P . | B . | b . | SE . | P . | B . | b . | SE . | P . | |||
House rule Inside | Step 1 | Condition | 0.08 | 0.01 | 0.07 | 0.88 | −0.01 | 0.02 | 0.09 | 0.85 | 0.01 | 0.00 | 0.02 | 0.85 |
Asthma | −0.08 | −0.15 | 0.09 | 0.11 | 0.02 | 0.04 | 0.013 | 0.77 | −0.01 | 0.00 | 0.03 | 0.99 | ||
Step 2 | Interaction | –a | – | – | – | – | – | – | – | – | – | – | – | |
Frequency communication | Step 1 | Condition | −0.10 | −0.05 | 0.02 | 0.03 | −0.08 | −0.03 | 0.02 | 0.14 | 0.05 | 0.01 | 0.01 | 0.30 |
Asthma | 0.06 | 0.05 | 0.04 | 0.18 | 0.08 | 0.05 | 0.03 | 0.12 | −0.10 | −0.02 | 0.01 | 0.03 | ||
Step 2 | Interaction | −0.34 | −0.12 | 0.07 | 0.09 | 0.17 | 0.05 | 0.07 | 0.46 | −0.06 | −0.01 | 0.01 | 0.77 | |
Quality communication | Step 1 | Condition | 0.05 | 0.02 | 0.02 | 0.25 | −0.23 | −0.04 | 0.02 | 0.03 | 0.17 | 0.01 | 0.00 | 0.03 |
Asthma | .04 | .02 | .03 | 0.37 | 0.01 | 0.00 | 0.02 | 0.95 | 0.04 | 0.00 | 0.01 | 0.57 | ||
Step 2 | Interaction | 0.15 | 0.04 | 0.05 | 0.45 | −0.37 | −0.04 | 0.05 | 0.42 | 0.16 | 0.01 | 0.01 | 0.64 |
Note. Standardized and unstandardized estimates, standard errors and P-values of the program effects are shown. In all analyses, we controlled for age, sex, education level father, education level mother and parental smoking. Condition: 1 = intervention condition and 2 = control condition, 1 = no asthma, 2 = asthma.
aA model with an interaction term did not convert.
Discussion
The present study evaluated the long-term effects of a home-based smoking prevention program on smoking-specific parenting among families with children with and without asthma. Hardly any attention has been paid to smoking prevention efforts for this subgroup of adolescents [29] despite additional negative consequences of smoking [2].
In general, no differential intervention effects on smoking-specific parenting were found in this 3-year RCT for families with and without a child with asthma. What we did find for the total sample (i.e. both families with and without asthma) is an effect of the program on quality of smoking communication. Being in the control condition was associated with less increase of quality of this communication over time and a stronger decrease after the intervention. The original study of Smoke-free kids also found some significant effects on parenting [9] among a sample including solely smoking parents. Children in their intervention condition were less likely to have initiated smoking [9]. In the Netherlands, the steeper increase in quality of communication currently found for children in the intervention condition did not translate into less smoking behavior 3 years post-baseline [18]. When the Smoke-free Kids program was replicated in the United States among a non-smoking parent sample [30], the authors showed that high engagement in the program did enhance recall of parental efforts in anti-smoking socialization by the children after 3 years [30]. However, they also did not replicate their effects on smoking initiation. Jackson and Dickinson postulate the young age of the children at final follow-up as one reason for the differential intervention effects among smoking [9] and non-smoking parents [30]. Children were 7–8 years old at baseline in their smoking-parent sample [9] and it is found that children of smoking parents are more likely to early initiate smoking relative to children with non-smoking parents [31]. In our study, children’s mean age at baseline was 10.1 and at final follow-up 13.1 years old. A longer follow-up period including the age of 14 and 15, in which children in the Netherlands show a strong uptake in ever smoking when compared with 12- and 13-year olds [32]. An intervention evaluation including children aged 10–14 also faced with low baseline and growth rates of substance use [33]. They conclude that this may have suppressed detection of interventions effects [33]. Another study found that adolescents of parents who smoked reported similar levels of anti-smoking practices measured at seventh grade. When these adolescents became older they reported lower levels of punishment for smoking and less smoking rules at home than children of non-smoking parents [34]. Finally, a school-based intervention did not show effects during the elementary school years; however, an effect appeared during secondary school, in favor for children in the intervention condition [35]. A study design including substantial years of follow-up assessments is preferable. Also, motivating parents to uphold their learned smoke-free parenting efforts may be an effective strategy [34], for instance by sending additional booster modules.
We reported some short-term effects of the intervention on parenting [19], while current analyses do not support long-term effects. As shortly discussed by Wiehe et al. [36], short-term effects may be a by-product of the intervention period. We found higher frequencies of a non-smoking agreement in the intervention condition after 6 months [19], while this non-smoking agreement was one of the assignments in the intervention booklets. When asked at the 6 months follow-up answers can be more positive due to higher retention of the learned material than at 3-year follow-up. This argument questions the timing of the intervention, which could arguably be closer to the likelihood the prevented behavior will first occur.
Interventions that aim at encouraging smoking-specific parenting strategies may benefit from emphasis on the health dangers [37]. Parents who view smoking as less dangerous to health are less likely to take action by means of smoking-specific parenting than parents who view smoking as more dangerous [37]. Parents often tend to underestimate their offspring’s engagement in smoking [38]. Due to respiratory problems, parents of a child with asthma might underestimate child smoking to a larger extend and might not see necessity for smoking rules. Parents of smoke-exposed children with asthma might believe that exposure to smoke has limited or no negative effect on their child’s asthma [39]. Given the voluntary nature of subscribing to this smoking prevention program, the already positive reports of children at baseline concerning smoking-specific involvement of their parents, and the non-significant increases in development of this parenting, it seems that those parents most motivated to prevent their children from smoking actually enrolled in this study. It is warranted to get underrepresented families involved in health promotion programs [40].
Specific parenting practices take place in the context of a general parenting style, which may alter the effectiveness of specific parenting practices on child’s development [41]. The present study exclusively focused on smoking-specific parenting, while, for instance, a general parenting style of low control or strictness has also been linked to adolescent smoking initiation [15, 42, 43]. Adolescents with asthma are found to perceive their parents as stricter and more involved than adolescents without asthma [15]. Although it may be more feasible to tailor a program on smoking-specific parenting practices, programs targeting only at smoking-specific practices and not general parenting may not be sufficient [42].
Some limitations should be acknowledged. First, the sample consisted of relatively high-educated parents. SES is associated with different styles of smoking-specific parenting. Higher-educated parents are, for instance, less permissive about smoking at home and place higher value on their child staying smoke-free [37]. Although we included parental education as a covariate, including a sample consisting of mainly higher-educated mothers might have underestimated the strengths of relationships examined. Second, we only included child self-reports. Child reports of parenting may differ from parental reports [38]. On the other hand, efforts by parents to keep offspring smoke-free may only be effective when children perceive these efforts, providing support for the use of children’s reports. Third, as stated elsewhere [13, 37], we were limited to abbreviated measures of some parenting strategies (i.e. one item measures or few answer options) due to the large survey and young age of the participants. This might have underestimated the strength of detecting parenting effects, although on the other hand, Jackson and Dickinson [12] did find results among similar aged children with simplified smoking-specific parenting measures. Fourth, only maternal and not paternal parenting was evaluated. Research indicates similar associations for maternal and paternal reports of smoking-specific parenting with adolescent smoking [11]. However, evidence for differences between father’s and mother’s communications about substance use have been found as well, depending on the substance and gender of the child [44]. Specifically, it was found that for sons, paternal communication was protective against marijuana, and maternal communication was protective against smoking. For daughters, neither paternal nor maternal communications were protective against substance use [44]. Based on these findings, it is essential to include fathers’ contributions to smoking-specific parenting and to examine the relations of parenting for sons and daughters separately. Fifth, we did not differentiate between families who engaged in the whole program (i.e. all five booklets) or had only gone through a few of them. On the other hand, program compliance in this sample seemed high in both conditions.
Conclusion
This prevention program did not support a differential effect on anti-smoking strategies held by parents with and without a child with asthma. Substantial smoking rates are consistently found for adolescents with asthma [1]. Parents of a child with asthma might underestimate this risk. Therefore, active engagement in anti-smoking socialization by parents of adolescents with asthma should still be encouraged. As there are hardly any effective prevention programs available for adolescents with asthma, future prevention research could focus also on other risk factors for smoking initiation, such as peers or interpersonal factors (e.g. peer smoking, refusal self-efficacy skills) [3–5].
Funding
This work was supported by a grant from ZonMw, The Netherlands Organization for Health Research and Development [project no. 50-50110-96-511]; and the Dutch Lung Foundation [project no. 3.4.07.041].
Conflict of interest statement
None declared.
References
STIVORO – voor een rookvrije toekomst: Smoking, the hard facts: Youth 2011. [In Dutch: Roken, de harde feiten: Jeugd 2011]. Den Haag: STIVORO – voor een rookvrije toekomst.