Abstract：
Bipolar disorder is associated with increased rates of many physical disorders, but the effects of medication are unclear. We systematically investigated the associations between sustained use of first line maintenance agents, lithium versus lamotrigine and valproate, and the risk of physical disorders using a nation-wide population-based target trial emulation covering the entire 5.9 million inhabitants in Denmark. We identified two cohorts. Cohort 1: patients with a diagnosis of bipolar disorder prior to first purchase (N = 12.607). Cohort 2: all 156.678 adult patients who had their first ever purchase (since 1995) of either lithium, lamotrigine or valproate between 1997 and 2021 regardless of diagnosis. Main analyses investigated the effect of sustained exposure defined as exposure for all consecutive 6-months periods during a 10-year follow-up. Outcomes included a diagnosis of incident stroke, arteriosclerosis, angina pectoris, myocardial infarction, diabetes mellitus, myxedema, osteoporosis, dementia, Parkinson\'s disease, chronic kidney disease and cancer (including subtypes). In both Cohorts 1 and 2, there were no systematic statistically significant differences in associations between sustained use of lithium versus lamotrigine and valproate, respectively, and any physical disorder, including subtypes of disorders, except myxedema, for which exposure to lithium increased the absolute risk of myxedema with 7-10 % compared with lamotrigine or valproate. In conclusion, these analyses emulating a target trial of \"real world\" observational register-based data show that lithium does not increase the risk of developing any kind of physical disorders, except myxedema, which may be a result of detection bias.
摘要：
双相情感障碍与许多身体疾病的发病率增加有关，但药物的效果尚不清楚。我们系统地调查了持续使用一线维护剂之间的关联，锂与拉莫三嗪和丙戊酸盐，以及使用覆盖丹麦全部590万居民的全国性基于人群的目标试验模拟来评估身体疾病的风险。我们确定了两个队列。队列1：首次购买前诊断为双相情感障碍的患者（N=12.607）。队列2：所有156.678名成年患者首次购买（自1995年以来）锂，1997年至2021年之间的拉莫三嗪或丙戊酸钠，无论诊断如何。主要分析调查了持续暴露的影响，定义为10年随访期间所有连续6个月的暴露。结果包括意外中风的诊断，动脉硬化,心绞痛,心肌梗塞,糖尿病,粘液水肿，骨质疏松,痴呆症,帕金森病,慢性肾脏疾病和癌症（包括亚型）。在队列1和2中，持续使用锂与拉莫三嗪和丙戊酸盐之间的关联没有系统的统计学显着差异。分别,和任何身体疾病，包括疾病的亚型，除了粘液水肿,与拉莫三嗪或丙戊酸盐相比，锂暴露会使粘液水肿的绝对风险增加7-10％。总之,这些模拟“现实世界”观察记录数据的目标试验的分析表明，锂不会增加发展为任何类型的身体疾病的风险，除了粘液水肿,这可能是检测偏差的结果。
