Aims To assess the overall relative risk of diabetes in individuals with prediabetes based on updated diagnostic criteria, as compared with individuals with normoglycaemia; and to identify the study characteristics associated with the heterogeneity between studies. Design Meta‐analysis, meta‐regression. Data sources PubMed, CINAHL, British Nursing Index. Search time frame: December 1998–December 2018. Review Methods The pooled relative risk of developing diabetes among individuals with prediabetes compared with those with normoglycaemia was calculated under a random effects model. Studies reported the natural progression from prediabetes to diabetes were included in this review. Sources of study heterogeneity were examined by a meta‐regression. Results Fifty‐nine eligible studies were systematically identified. The pooled relative risk for diabetes among individuals with prediabetes as compared with normoglycaemia was 5.88 (95% CI: 5.02‐6.89). The annualized incidence rate (per 1,000 person‐year) for diabetes among individuals with prediabetes defined by different criteria varied from 2.20–212.15, with high heterogeneity between studies (I2 = 96.64%, Q test: p < .001). In the multivariable meta‐regression analysis, Asian population significantly increased RR compare to Caucasians. Besides, people with ‘elevated glycated haemoglobin A1c or impaired fasting glucose’ had highest relative risk compare to people with other types of prediabetes. Conclusion Individuals with prediabetes had higher risk of developing diabetes than those with normoglycaemia. Races and diagnostic criteria of prediabetes were associated with the magnitude of the estimated risk. Impact Prediabetes is a precursor of diabetes. To screen people with prediabetes as early as possible, practitioners could consider haemoglobin A1c test as an alternative to fasting plasma glucose test. Nurses should educate people especially Asians with prediabetes for the prevention of progression to diabetes. 摘要 目的 根据更新的诊断标准,与正常血糖的个体相比,评估糖尿病前期个体患糖尿病的总体相对风险;并确定与研究间异质性相关的研究特征。 设计 荟萃分析、荟萃回归。 数据来源 PubMed、CINAHL、英国护理学索引。搜索时间范围:1998年12月至2018年12月。 审查方法 在随机效应模型下,计算了与正常血糖个体相比,糖尿病前期个体患糖尿病的合并相对风险度。研究报告指出,从糖尿病前期发展为糖尿病的自然发展被纳入本审查中。通过荟萃回归检查了研究的异质性来源。 结果 系统化地确定了59项符合条件的研究。与正常血糖个体相比,糖尿病前期个体患糖尿病的合并相对风险度为5.88(95%置信区间:5.02‐6.89)。按不同标准定义的糖尿病前期个体的糖尿病年发病率(每1000人‐年)在2.20‐212.15不等,研究间异质性较高(I2=96.64%,Q检验:P<.001)。在多变量荟萃回归分析中,与白种人相比,亚洲人的相对风险度显著增加。此外,‘糖化血红蛋白A1c升高或空腹血糖受损’的人群与其他类型的糖尿病前期人群相比,相对风险度最高。 结论 糖尿病前期个体患糖尿病的风险血糖正常个体。糖尿病前期的个体类别和诊断标准与估计的风险量级有关。 影响 糖尿病前期是糖尿病的前兆。为了尽早筛查糖尿病前期人群,医生可以考虑用血红蛋白A1c测试来代替空腹血糖检测。护士应提醒人们,尤其是糖尿病前期的亚洲人,以防止发展成糖尿病。