Huerta Michael, Chodick Gabriel, Balicer Ran D, Davidovitch Nadav, Grotto Itamar
Israel Defence Force Medical Corps, Army Health Branch, Military Post 02149, Israel.
Prev Med. 2005 Aug;41(2):646-50. doi: 10.1016/j.ypmed.2005.01.011.
Many studies use questionnaires to determine smoking status and age of smoking onset. This study aimed to determine the reliability of self-reported smoking history and age of smoking initiation.
The proportion of inconsistent answers and correlation coefficients of reported age of initial smoking were measured by an answer-reanswer analysis of questionnaires in an ongoing, two-step, population-based survey of health behavior. Interviews were conducted on the day of recruitment to and the day of discharge from mandatory military service in Israel among a sample of 25,437 young men and women recruited between 1986 and 2000.
Of 7276 participants reporting current or past smoking upon recruitment, 559 (7.7%) reported never having smoked upon discharge, thus demonstrating prima facie inconsistency. Variables significantly associated with reliable reporting in a multivariate logistic regression model were female gender (P = 0.04) and more than 4 years of military service (P < 0.01). 6010 subjects who reported a positive smoking history at both recruitment and discharge were available for analysis of reliability of reported age at smoking onset. Intraclass correlation coefficients for recruitment/discharge consistency in reported age at first cigarette were 0.73 (95% CI: 0.71-0.74) and 0.76 (95% CI: 0.74-0.78) for men and women, respectively. Eastern origin, lower subject education level, and lower paternal education level were also associated with lower reliability.
Our results showed a relatively high level of answer-reanswer reliability, with some variance attributable to personal characteristics. These results suggest that self-reported age at onset of tobacco use is practical and reliable in normative, young adult populations. However, time elapsed between questionnaires and demographic and lifestyle characteristics may affect reliability rates, and thus should be carefully regarded in future studies.
许多研究使用问卷来确定吸烟状况和开始吸烟的年龄。本研究旨在确定自我报告的吸烟史和开始吸烟年龄的可靠性。
在一项正在进行的、分两步的、基于人群的健康行为调查中，通过对问卷的回答-再回答分析来测量不一致答案的比例和报告的首次吸烟年龄的相关系数。在以色列，对1986年至2000年间招募的25437名年轻男女样本进行了征兵日和义务兵役退伍日的访谈。
在7276名招募时报告当前或过去吸烟的参与者中，559人（7.7%）在退伍时报告从未吸烟，因此初步显示出不一致。在多变量逻辑回归模型中，与可靠报告显著相关的变量是女性（P = 0.04）和超过4年的兵役（P < 0.01）。6010名在招募和退伍时都报告有阳性吸烟史的受试者可用于分析报告的开始吸烟年龄的可靠性。首次吸烟年龄报告的招募/退伍一致性的组内相关系数，男性为0.73（95%CI：0.71 - 0.74），女性为0.76（95%CI：0.74 - 0.78）。东方血统、较低的受试者教育水平和较低的父亲教育水平也与较低的可靠性相关。
我们的结果显示出相对较高的回答-再回答可靠性水平，部分差异可归因于个人特征。这些结果表明，在规范的年轻成年人群中，自我报告的开始使用烟草的年龄是实用且可靠的。然而，问卷之间的时间间隔以及人口统计学和生活方式特征可能会影响可靠性率，因此在未来研究中应予以仔细考虑。
