Psychopathy has traditionally been linked with heightened criminality, but the relationship of psychopathy with increased risk for dangerousness is contested. To address this debated issue, we conducted an umbrella review (PROSPERO CRD42020214761) of all available meta-analyses of psychopathy and indices of ‘dangerousness’ (e.g., violent or sexual recidivism, self-reported aggression). We searched PsycINFO, Web of Science, PubMed, and Scopus from inception to August 19, 2022, to identify systematic reviews and meta-analyses on psychopathy and dangerousness. Our review included 33 studies, with the quantitative synthesis including 17 effect sizes extracted from 10 studies (N = ∼77,000 participants). Overall, we observed a pooled correlation coefficient r = 0.284 [95% CI = 0.233, 0.336] for the association of psychopathy with dangerousness (equivalent to Cohen's d = 0.592). Despite considerable heterogeneity (I² = 89.9%), leave-one-out analyses had minimal impact. The effect was robust to examination of potential moderators such as study quality and sample ages. However, the relationship was stronger when psychopathy was assessed using self-report compared to clinical rating scales. The association should be interpreted as meaningful in both the short-term and the long-term and suggests that psychopathy is one of the strongest predictors of dangerousness in the realm of psychopathology.

精神病态传统上与犯罪率增加有关，但精神病态与危险风险增加之间的关系存在争议。为了解决这个有争议的问题，我们对所有可用的精神病和“危险”指数（例如，暴力或性累犯、自我报告的攻击行为）的荟萃分析进行了综合审查 (PROSPERO CRD42020214761)。我们搜索了从开始到 2022 年 8 月 19 日的 PsycINFO、Web of Science、PubMed 和 Scopus，以确定关于精神病和危险性的系统评论和荟萃分析。我们的审查包括 33 项研究，定量综合包括从 10 项研究（N  = ~77,000 名参与者）中提取的 17 种效应量。总体而言，我们观察到合并相关系数r  = 0.284 [ 95% CI = 0.233, 0.336] 对于精神病与危险性的关联（相当于 Cohen 的d  = 0.592）。尽管存在相当大的异质性 ( I ²  = 89.9%)，留一法分析的影响很小。这种效果对于检查潜在的调节因素（例如研究质量和样本年龄）是稳健的。然而，与临床评定量表相比，使用自我报告评估精神病时，这种关系更强。这种关联应该被解释为在短期和长期都有意义，并表明精神病态是精神病理学领域最强烈的危险预测因子之一。