Elsevier

Drug and Alcohol Dependence

Volume 84, Issue 1, 1 September 2006, Pages 40-47
Drug and Alcohol Dependence

Correspondence between proxy and self-reports on smoking in a full family study

https://doi.org/10.1016/j.drugalcdep.2005.11.026Get rights and content

Abstract

Objectives

The present study investigated the reliability of proxy reports obtained from family members with self-reports on adolescent and parental lifetime and current smoking status.

Methods

Data were assessed from 416 families, consisting of both biological parents and two adolescent siblings aged 13–17 years. These families were assessed at baseline and 1 year later. Sensitivity, specificity, positive predictive value and negative predictive value were calculated to test whether proxy reports corresponded with self-reports.

Results

Mothers scored higher than fathers on most measures on lifetime and current smoking status of both children. The sensitivity was low for parental reports, but moderate to high for children's reports. Specificity and positive predictive value were high in all proxy reports. The negative predictive value was moderate (parents as proxy reporters) to low (children as proxy reporters) on lifetime smoking, but high on current smoking.

Conclusions

Adolescents, aged 13–17 years, can be used as a reliable source to assess the smoking status of their mothers and fathers. Parents, however, appeared to accurately identify the smoking status of their adolescent children less reliably.

Introduction

Cigarette smoking is one of the main preventable health hazards and causes of high mortality rates. To gain insight into prevalence numbers of cigarette smoking among adults and adolescents in countries all over the world, research projects principally use self-reports. Although the validity of self-reports has often been questioned, it has been argued that self-reports on cigarette smoking appear to have high levels of sensitivity and specificity as compared with results of biochemical verification (Patrick et al., 1994). If confidentiality and anonymity are guaranteed, the smoking status of adults and adolescents can be accurately assessed through self-reports (Dolcini et al., 1996, Hunter et al., 1980).

In a study by Gilpin et al. (1994), in an adult population, it was argued that proxy reports should be considered in the overall estimate of smoking prevalence. Their study shows that one adult could provide data on smoking status for all household members. Especially children/siblings, spouse/partner and parent/guardian were more accurate as proxy reporters than other relatives or unrelated persons, implying that immediate relatives have a closer relationship with the individual and provide more accurate information on smoking status of the other family member (Gilpin et al., 1994). However, they used older children, from 18 years upwards, as proxy reporters.

The aim of this study was to investigate the reliability of adolescents’ and parents’ proxy reports (i.e., reporting about the other person) on parental and adolescent cigarette smoking. It was important to investigate the reliability of adolescent children's proxy reports on parental cigarette smoking, because many studies concluded that children imitated or modeled their parents’ smoking behavior through observation, starting from the assumption that children knew whether or not their parents smoked. However, few studies investigated whether children are able to accurately identify their parents’ smoking status. One of the first studies indicated that elementary schoolchildren, aged 10–13 years, were able to provide reliable reports on the smoking status of the parent with whom they are living (Barnett et al., 1997). The reliability of adolescent reports on parental smoking has not been investigated, as far as we know. We assumed that older adolescents were probably less reliable than pre-teens and younger adolescents. In the adolescence period, friends and peers become more important for adolescents although parents remain influential, obviously the time spent with parents decreases, as the adolescents grow older (e.g., Chassin et al., 1984, Larson et al., 1996, Urberg et al., 1991). Therefore, older adolescents will be more engaged with their friends and spend time outside home with friends and peers, in contrast to younger children who will be more frequently in the presence of their parents. Consequently, as younger adolescents are at home more than their older adolescent siblings they will probably be more able to accurately identify and report the frequency and intensity of parents’ smoking. In this study we investigated both older and younger siblings.

The present study is as far as we know one of the first to examine whether parents are able to provide reliable reports on their children's smoking status. Adolescent smoking is often covert behavior, because adolescents experiment or regularly smoke, in secrecy, outside their homes (Engels, 1998), in contrast to parents’ smoking, which is usually overt behavior displayed at home. It is relevant to investigate whether parents know their children's smoking status for two reasons. Firstly, if parents are accurate in reporting their offspring's cigarette smoking, it would be sufficient to only use parents’ reports in population surveys to estimate adolescent smoking prevalence. This would be more convenient and would reduce costs (Gilpin et al., 1994, Hatch et al., 1991). Secondly, school-based educational interventions to prevent and discourage children from smoking and mass media have often targeted the children themselves.

Recently, prevention campaigns have also focused on the parents by means of parent education programs (Lantz et al., 2000). Therefore, it is important to estimate the extent to which parents are aware of the smoking status of their children. It appears that parents who exert control and monitor their children's behavior are less likely to have children who will initiate smoking (Engels et al., in press). Perhaps when parents are not able to provide reliable reports on their offspring's smoking, this might be seen as a sign of ‘inappropriate parenting’ as they do not monitor and control their children's behavior, and therefore the children will end up initiating and continuing smoking (Cohen et al., 1994, Duncan et al., 1998, Simons-Morton et al., 1999). In addition, when parents are not aware of their children's attempts at smoking or that a child is a current smoker, they will probably not react or undertake appropriate actions to discourage or stop them from smoking (Engels and Willemsen, 2004, Harakeh et al., 2005). Parents will almost always undertake action when they perceive that their offspring initiated smoking; for example, it will prompt them to talk more frequently about smoking-related issues or make a non-smoking agreement with their children.

This paper presents data from a full family study in which each family consisted of both biological parents and two adolescent siblings. Proxy reports in this study were obtained from mothers, fathers, and older and younger siblings. Further, data were collected again a year later to assess the stability of the original results; the two sets of data are referred to below as data for the first and second “waves”. It was important to study whether the reliability of a specific proxy report was stable or whether it changed over time, and whether the accuracy of proxy reports increased or decreased over a period of time (12 months).

Thus, the purposes of the study were to test (1) whether mothers and fathers were able to provide accurate reports on the older and younger children's smoking status; (2) whether older and younger children were able to provide valid reports on their mother and father's smoking status; (3) whether these proxy reports was stable over a 1-year period.

Section snippets

Design and sample

A total of 428 Dutch families were selected. This “Family and Health” project has been described previously (Harakeh et al., 2005). Each family consisted of a mother, father and two adolescent children in the age range 13–17. The addresses of the families were selected from registers of 22 municipalities in the Netherlands. The families received a letter in which was asked to return the included consent response form, if they were willing to participate. A total of 885 families consented to

Prevalence of smoking

Table 1 shows the percentages of lifetime smoking and current smoking for the four family members at baseline and 12 months later. Self-reports of children showed that a higher percentage of older children had tried smoking or were current smokers at T1 and T2. Parental self-reports indicated that at both waves an equal percentage of mothers and fathers had ever tried smoking. More fathers than mothers were current smokers. Furthermore, as Table 1 shows, in general, parents underestimated the

Discussion

Our findings indicated that proxy reports of mothers and fathers were fairly inaccurate indicators of adolescents’ lifetime and current smoking status. Parents often thought that their child never tried smoking or was not currently smoking while in fact many of the children had tried smoking or were current smokers. The study of Gilpin et al. (1994) indicated that close family members such as parents were able to accurately report the smoking status of their adult children, however, this study

Acknowledgement

This study was funded by a grant from the Netherlands Organization for Scientific Research (NWO).

References (30)

  • L. Chassin et al.

    Predicting the onset of cigarette smoking in adolescents: a longitudinal study

    J. Appl. Soc. Psychol.

    (1984)
  • D.A. Cohen et al.

    Parent participation in an adolescent drug abuse prevention program

    J. Drug Educ.

    (1995)
  • D.A. Cohen et al.

    Parenting behaviors and the onset of smoking and alcohol use: a longitudinal study

    Pediatrics

    (1994)
  • H. De Vries et al.

    Parents’ and friends’ smoking status as predictors of smoking onset: findings from six European Countries

    Health Educ. Res.

    (2003)
  • M.M. Dolcini et al.

    Factors influencing agreement between self-reports and biological measures of smoking among adolescents

    J. Res. Adolesc.

    (1996)
  • Cited by (0)

    View full text