Abstract: Objective To systematically evaluate the influencing factors of fractures (FR) in school-age children and adolescents, and to assess the association based on a dose-response Meta-analysis. Methods Observational studies on the FR in school-age children and adolescents published from Jan.2000 to Jan.2022 were retrived from the databases of CBM, CNKI, Wanfang Data, and VIP.The articles were screened strictly to extract valid data.Stata 16.0 was applied to combine the odds ratio (OR) of effect sizes, Q-test and I² test were used to evaluate the heterogeneity of outcome indicators in the literature, and Begg's test was used to detect publication bias.Finally, the dose-response relationship between FR and influencing factors was performed. Results A total of 14 articles (2 cohort studies, 3 case-control studies and 9 cross-sectional studies) involving 30 769 cases and 45 857 controls were included.Meta-analysis showed that boys (OR=2.91, 95%CI: 2.76 - 3.06), moderate-to-vigorous physical activities (OR=1.55, 95%CI: 1.33 - 1.77), FR history (OR=3.56, 95%CI: 2.18 - 4.94), July/August (OR=4.19, 95% CI: 3.30 - 5.08) and rural area (OR=2.17, 95%CI: 1.02 - 3.32) were risk factors for the developing of FR, while high secondary school (OR=0.39, 95%CI: 0.11 - 0.67) was a protective factor for FR.The dose-response analysis showed a non-linear correlation between age and FR risks (χ²=13.26, P<0.01).The correlations of body mass index (χ²=6.71, P=0.082), physical activity time (χ²=3.18, P=0.107) and household income (χ²=1.47, P=0.690) with FR risks were linear. Conclusion Gender, physical activity time, fracture history, month, and urban/rural area may lead to the differences in the incidence of FR in school-age children and adolescents, and the effect of age on FR showed a non-linear relationship of "increasing and then decreasing", which needs to be confirmed by further clinical studies and epidemiological surveys.

Key words: school-age children and adolescents, fracture, Meta-analysis, dose-response analysis

摘要： 目的 用Meta分析的方法系统评价学龄儿童和青少年骨折(FR)发生的影响因素,并对FR发生与影响因素间的剂量-反应关系进行定量评价。方法 由两名研究人员独立检索中国生物医学文献数据库(CBM)、中国知网(CNKI)、万方数据(Wanfang Data)和维普中文科技期刊全文数据库(VIP)2000年1月-2022年1月发表的与学龄儿童、青少年骨折发生相关的观察性研究,严格筛选文献,提取有效数据,应用Stata 16.0对效应量的比值比(OR)进行合并;采用Q检验和I²检验评价文献结局指标的异质性,并利用Begg's检验对发表偏倚情况进行检测,最后对学龄儿童骨折及其危险因素之间的剂量-反应关系进行定量分析。结果 共纳入14篇文献,包括2篇队列研究,3篇病例对照研究和9篇横断面研究,骨折组30 769例和对照组45 857例。Meta分析结果显示,男生(OR=2.91,95% CI:2.76~3.06)、中高强度体育活动(OR=1.55,95% CI:1.33~1.77)、骨折史(OR=3.56,95% CI:2.18~4.94)、7/8月份(OR=4.19,95% CI:3.30~5.08)、农村地区(OR=2.17,95% CI:1.02~3.32)为FR发生的危险因素,高中学段(OR=0.39,95% CI:0.11~0.67)为其保护因素。进一步的剂量-反应分析结果显示,年龄与FR发病风险之间呈非线性关系(χ²=13.26,P<0.01),体质量指数(χ²=6.71,P=0.082)、体育活动时间(χ²=3.18,P=0.107)和家庭收入水平(χ²=1.47,P=0.690)与FR发病风险之间呈线性关系。结论 性别、体育活动时间、骨折史、月份、城乡分布等因素会导致学龄儿童和青少年FR发生率的差异,而年龄对FR的影响呈现一种“先上升再下降”的非线性关系,上述结论尚需进一步临床研究及流行病学调查加以确证。

关键词: 学龄儿童, 青少年, 骨折, Meta分析, 剂量-反应分析