Agreement between proband and parental self-report of smoking behavior and nicotine dependence

Although investigators have used family history methods to investigate familial clustering of disorders such as depressive disorder, alcoholism, coronary heart disease, and cancer, research of this type is relatively new to the field of smoking. We examined agreement between proband report of parental smoking behavior and parent's self-reported smoking behavior in 126 proband-parent pairs. Probands were either never, current, or exsmokers; parents were either current or exsmokers. Agreement between proband and parent was better for smoking behaviors when the parent was a current smoker. We found good proband-parent agreement for some smoking behaviors when the parent was a current smoker (e.g., age started smoking [mean (SD) difference between proband and parental report, 1.36 years (9.07 years)], and cigarettes per day and brand smoked [kappa=.49 and.56, respectively]) but poor agreement for items that may represent more complex or less observable indicators of nicotine dependence, regardless of parental smoking status. Reliability was excellent for probands of either current- or exsmoker parents for smoking status (kappa=.92). As a result of probands' difficulty in reliably answering several items that comprise the Fagerström Tolerance Questionnaire (FTQ) (e.g., number of minutes to parent's first cigarette of the day, whether parent smoked more in the morning than during the rest of the day) and the DSM-IV (Diagnostic and Statistical Manual of Mental Disorders, 4th edition) diagnosis (e.g., sleep difficulties and difficulty concentrating during nicotine withdrawal), reliability for an FTQ score >/=7 (kappa=.067) and the DSM-IV nicotine dependence diagnosis (kappa=.28) was poor. Our findings should aid investigators in defining the limits of proband reports of parental smoking behavior and identify opportunities for the development of better approaches for the assessment of familial nicotine dependence.