WOMEN'S DRINKING: SECONDARY ANALYSES OF 20-YEAR DATA

这项申请提出了新的分析独特的20年全国调查数据的妇女饮酒和 它的前因后果。全国妇女健康和生活体验研究(NSHLEW) 在1981年、1991年、1996年和2001年采访了具有代表性的美国女性样本。2001年的调查 数据使评估对女性饮酒行为和饮酒的长期影响的新模型成为可能 轨迹，并提供关于受访者使用处方精神活性药物的详细信息， 关于老年女性健康的新信息，以及复制和延长这些的终生饮酒史 是在20年前获得的。 拟议的分析将(1)测试长期影响妇女危险的多波模型 饮酒；(2)确定女性饮酒轨迹的模式、前因和后果；(3)评估 (A)追溯饮酒史的可靠性和实用性(例如，确定饮酒轨迹)，(B)背景 就业对女性饮酒的影响，(C)终生饮酒模式与老年女性饮酒模式的关系 健康，以及(D)妇女同时使用酒精和处方药的前因后果； (4)比较从典型数量频率问题和分级频率问题中获得的饮酒自我报告 问题。 将回顾数据和三波调查数据相结合的结构方程模型将用于 检验假设对女性饮酒的影响：(1)儿童早期经历，(2)性侵害 儿童期和成年期；(3)成人亲密关系的特点。饮酒轨迹(按 潜在增长混合模型和先验标准)将分析与先行因素的关联(如 儿童)、随之而来的变化(如婚姻状况)和结果(如健康问题)。轨迹 从终生饮酒史得出的数据，如果被证明是可靠的，将进行类似的评估。其他分析将测试 关于(1)就业对女性饮酒背景影响的历史变化，(2)女性的健康状况的假设 有不同终生饮酒史的老年女性，以及(3)联合饮酒和 开出的精神活性药物。一项方法论实验将显示女性是否报告更高的水平 在回答分级频率的问题时的饮酒量比回答关于典型饮酒的问题时的饮酒量 频次和数量。这里提出的二次分析的结果将提高对女性 终生饮酒及其前因后果，有助于设计对性别更敏感的 以及有效的治疗、预防和酒精政策方法。

This application proposes new analyses of unique 20-year national survey data on women's drinking and its antecedents and outcomes. The National Study of Health and Life Experiences of Women (NSHLEW) interviewed representative national samples of U.S. women in 1981, 1991, 1996, and 2001. The 2001 survey data make it possible to evaluate new models of long-term influences on women's drinking behavior and drinking trajectories, and also provide detailed information about respondents' use of prescribedpsychoactive medications, new information about older women's health, and lifetime drinking histories replicating and extending those obtained 20 years earlier. The proposed analyses will (1) test multi-wave models of long-term influences on women's hazardous drinking; (2) identify patterns, antecedents, and consequences of women's drinking trajectories; (3) evaluate (a) the reliability and utility of retrospective drinking histories (e.g., to identify drinking trajectories), (b) contextual influences of employment on women's drinking, (c) associations of lifetime drinking patterns with older women's health, and (d) antecedents and consequencesof women's combined use of alcohol and prescribed medications; and (4) compare drinking self-reports obtained from typical quantity-frequency questions vs. graduated-frequency questions. Structural equation models combining retrospective data and three waves of survey data will be used to test hypothesized effects on women's drinking from (1) early childhood experiences, (2) sexual victimization in childhood and adulthood, and (3) characteristics of adult intimate relationships. Drinking trajectories (classified by latent growth mixture modeling and by a priori criteria) will be analyzedfor associations with antecedents (as in childhood), accompanying changes (as in marital status), and outcomes (such as health problems). Trajectories derived from lifetime drinking histories, if shown to be reliable, will be evaluated similarly. Other analyses will test hypotheses about (1) historical change in contextual effects of employment on women's drinking, (2) the health of older women with different lifetime drinking histories, and (3) consequencesof combined use of alcohol with prescribed psychoactive medications. A methodological experiment will show whether women report higher levels of alcohol consumption in response to graduated-frequency questions than to questions about typical drinking frequency and quantity. Results of the secondary analyses proposed here will improve knowledge about women's lifetime drinking and its antecedents and consequences, in ways that can aid the design of more gender-sensitive and effective approaches to treatment, prevention, and alcohol policy.