The multitude of gambling activities has given rise to heterogeneous ways of analyzing these behaviors and may partially underlie the lack of replication in gambling research. The current study used complementary analyses to investigate the structure, typology and etiology of gambling behaviors in a discovery sample of 2,116 twins (54.86% female; M_age = 24.90) and a replication sample of 619 siblings (30.37% female; M_age = 28.00). Our approach was twofold. First, we used confirmatory factor analyses to investigate the structure across the frequency of eight gambling activities. Second, we used factor mixture models to identify gambling frequency subtypes. We assessed associations with gambling frequency as well as conducted genetically informed analyses to estimate the role of genetic and environmental influences. Across samples, a two-factor model fit the data best, with a Common Gambling factor influencing all activities and a separate factor for Skill Gambling. Our study identified four gambling frequency subtypes, which resembled the typology from the Pathways Model. We found distinct demographic, psychiatric, behavioral and genetic risk profiles for the different gambling factors and subtypes with robust associations observed for male sex, risk-taking, sensation seeking, alcohol dependence and problem gambling. Controlling for shared genetic and environmental influences (via co-twin control modeling), we found that sensation seeking directly increased Common Gambling frequency. In sum, we illustrated the utility of multi-dimensional statistical techniques for disentangling the structure and typology from complex multivariate gambling data.

大量的赌博活动导致了分析这些行为的不同方法，这可能是赌博研究缺乏可复制性的部分原因。目前的研究使用补充分析来调查 2,116 名双胞胎（54.86% 女性；M_{年龄= 24.90）的发现样本和 619 名兄弟姐妹（30.37% 女性；} M_年龄 ）的复制样本中赌博行为的结构、类型和病因 = 28.00)。我们的方法是双重的。首先，我们使用验证性因素分析来调查八种赌博活动频率的结构。其次，我们使用因子混合模型来识别赌博频率亚型。我们评估了与赌博频率的关联，并进行了遗传信息分析，以估计遗传和环境影响的作用。在样本中，双因素模型最适合数据，其中一个共同赌博因素影响所有活动，一个单独因素影响技能赌博。我们的研究确定了四种赌博频率亚型，类似于通路模型中的类型。我们发现不同赌博因素和亚型的人口统计学、精神病学、行为和遗传风险概况与男性、冒险、感觉寻求，酒精依赖和问题赌博。控制共享的遗传和环境影响（通过双胞胎控制模型），我们发现寻求感觉直接增加了常见赌博频率。总之，我们说明了多维统计技术在从复杂的多元赌博数据中解开结构和类型的效用。