Deutsch Arielle R, Slutske Wendy S, Lynskey Michael T, Bucholz Kathleen K, Madden Pamela A F, Heath Andrew C, Martin Nicholas G
University of Missouri-Columbia.
Washington University School of Medicine.
Dev Psychopathol. 2017 Aug;29(3):845-861. doi: 10.1017/S0954579416000523. Epub 2016 Jul 15.
The current study examined a stage-based alcohol use trajectory model to test for potential causal effects of earlier drinking milestones on later drinking milestones in a combined sample of two cohorts of Australian monozygotic and same-sex dizygotic twins (N = 7,398, age M = 30.46, SD = 2.61, 61% male, 56% monozygotic twins). Ages of drinking, drunkenness, regular drinking, tolerance, first nontolerance alcohol use disorder symptom, and alcohol use disorder symptom onsets were assessed retrospectively. Ages of milestone attainment (i.e., age-of-onset) and time between milestones (i.e., time-to-event) were examined via frailty models within a multilevel discordant twin design. For age-of-onset models, earlier ages of onset of antecedent drinking milestones increased hazards for earlier ages of onset for more proximal subsequent drinking milestones. For the time-to-event models, however, earlier ages of onset for the "starting" milestone decreased risk for a shorter time period between the starting and the "ending" milestone. Earlier age of onset of intermediate milestones between starting and ending drinking milestones had the opposite effect, increasing risk for a shorter time period between the starting and ending milestones. These results are consistent with a causal effect of an earlier age of drinking milestone onset on temporally proximal subsequent drinking milestones.
本研究检验了一个基于阶段的酒精使用轨迹模型，以测试在澳大利亚单卵双胞胎和同性双卵双胞胎的两个队列的合并样本中（N = 7398，年龄M = 30.46，标准差 = 2.61，61%为男性，56%为单卵双胞胎），早期饮酒里程碑对后期饮酒里程碑的潜在因果效应。通过回顾性评估饮酒、醉酒、经常饮酒、耐受性、首个非耐受性酒精使用障碍症状以及酒精使用障碍症状发作的年龄。在多水平不一致双胞胎设计中，通过脆弱模型检查里程碑达成年龄（即发病年龄）和里程碑之间的时间（即事件发生时间）。对于发病年龄模型，先前饮酒里程碑的较早发病年龄增加了更接近的后续饮酒里程碑较早发病的风险。然而，对于事件发生时间模型，“开始”里程碑的较早发病年龄降低了从开始到“结束”里程碑之间较短时间段的风险。在开始和结束饮酒里程碑之间的中间里程碑的较早发病年龄具有相反的效果，增加了从开始到结束里程碑之间较短时间段的风险。这些结果与较早的饮酒里程碑发病年龄对时间上接近的后续饮酒里程碑的因果效应一致。
