Slomkowski et al. (2005) have used the DeFries-Fulker regression model to test whether sibling effects on smoking may reflect social or genetic processes. They included not only the siblings phenotype but also the social connectedness between the siblings and showed that shared environmental sibling effects on smoking were significant.

The authors used the measure ‘number of days smoked over the past 30 days’. Although the authors do not report the distribution of this variable, it is probably not normally distributed. It is expected that a large group will score 0 days (non-smokers) and a group will score 30 days (daily smokers). Probably a small group will report a number between 1 and 29 (chippers, non-daily smokers). The authors examined the individual differences on this variable and transformed it to approximately a normal distribution. This measure is especially useful in the critical period of taking up smoking. After smoking is initiated other measures, such as number of cigarettes per day, number of quit attempts, years smoked and nicotine dependence, become more important. It is interesting whether the results of Slomkowski et al. can be replicated for these types of measure. In a sample of Dutch twins and their family members data on smoking and social contact were collected (Boomsma et al. 2002). For the maximum number of cigarettes smoked per day, the environmental effects are stronger in twins who have contact on a daily/weekly basis than twins who contact each other less than weekly (Table 1).

Table 1.  Estimation of genetic and shared environmental influences on maximum number of cigarettes smoked per day in total group and separately for twin pairs who have daily/weekly contact and twin pairs who contact each other less than weekly. Genetic and shared environmental influences are estimated using the DeFries-Fulker regression model (DeFries & Fulker 1985). Regression coefficient b3 = h2 and regression coefficient b1 = c2. Sample: twins participating in the survey of 2000 and/or the survey of 2002 of the Netherlands Twin Registry, mean age: 32 years.
 b 3 (h 2 )b 1 (c 2 )
Total group0.54 (= 0.000)0.15 (= 0.032)
Twin pairs: daily/ weekly contact0.33 (= 0.018)0.38 (= 0.001)
Twin pairs: less than weekly/no contact0.61 (= 0.000)0.06 (= 0.510)

Both the results of Slomkowski et al. and the results of our Dutch data are in agreement with previous studies. Rose et al. (1990) concluded that social contact contributed to sibling resemblance for alcohol consumption. Slomkowski et al. state that the critical addition of their report is the demonstration that sibling social connectedness is a significant moderator of the shared environmental effect on smoking adolescence using a genetically informative design. However, Rose et al. (1990) have already introduced a twin-model including social contact. They considered and answered the question of what the direction of the effect is: does similarity in alcohol consumption lead to increased contact or does more contact lead to similarity in alcohol consumption? They concluded that more social contact contributed to higher sibling similarity in alcohol consumption (Rose et al. 1990). Several important smoking measures are dichotomous, such as current smoking (yes/no) or ever smoked (yes/no). Carey (1992) has introduced a method of analysis that permit reciprocal twin interactions for a dichotomous trait. This model provided a better fit to their data (on antisocial behaviour) and yielded lower estimates of heritability compared to the traditional twin-model. Those threshold models including sibling interaction could also be applied to dichotomous smoking data. Using different methods and models can shed light on the nature, timing and specificity of sibling effects on smoking behaviour.


  1. Top of page
  2. References
  • Boomsma, D. I., Vink, J. M., Beijsterveldt, T. C. , De Geus E. J. , Beem, A. L., Mulder, E. J. et al. (2002) Netherlands Twin Register: a focus on longitudinal research. Twin Research, 5, 401406.
  • Carey, G. (1992) Twin imitation for antisocial behavior: implications for genetic and family environment research. Journal of Abnormal Psychology, 101, 1825.
  • DeFries, J. C. & Fulker, D. W. (1985) Multiple regression analysis of twin data. Behavior Genetics, 15, 467473.
  • Rose, R. J., Kaprio, J., Williams, C. J., Viken, R. & Obremski, K. (1990) Social contact and sibling similarity: facts, issues, and red herrings. Behavior Genetics, 20, 763777.
  • Slomowski, C., Rende, R., Novak, S., Lloyd-Richardson, E. & Niaura, R. (2005) Sibling effects on smoking in adolescence: evidence for social influence from a genetically informative design. Addiction, 100, 430438.