基于Kupperman指数的石化女工围绝经期综合征的影响因素分析和风险列线图的建立

Kupperman index-based analysis of factors influencing perimenopausal symptoms and establishment of risk nomogram in female petrochemical workers

[背景]女性的整个生命周期中面临较多的生殖健康问题,而石化行业有害因素暴露更可能对职业女性的健康和生育能力的影响产生叠加效应。[目的]分析石油化工行业女性职工围绝经期综合征(PMS)的影响因素,并建立女工出现PMS风险的列线图模型,为女工提供方便快捷的健康监测评估方法。[方法]选择某石油化工企业所有年龄为45~55岁符合条件的围绝经期女工共计2653人作为研究对象,进行现场问卷调查,收集女工的基本情况、职业情况、工作心理状况、生殖及生育健康信息。利用Kupperman指数量表评估女工PMS(Kupperman评分≥7分)的发生情况,采用疲劳量表评估躯体疲劳和脑力疲劳程度。采用多因素logistic回归建立列线图预测模型,绘制列线图,使用C-index验证模型的区分度,再采用Bootstrap方法从研究对象内部验证风险模型,评价模型的校准度。[结果]2653名女工中,发生PMS者共1306例,占49.2%;症状以疲乏(79.95%)、烦躁易怒(71.32%)、失眠(66.79%)为多。不同工龄、体重指数(BMI)和工作体位的女工PMS发生率不同(P<0.05);有饮酒习惯,母系存在过早或过晚绝经,患高血压,缺乏体育锻炼,搬运重物,近6个月因身体原因请假,粉尘、化学物、噪声[>80 d B(A)]、电磁或职业有害因素联合暴露,以及未配戴防护口罩和手套或防护耳塞的女工PMS发生率均较高(P<0.05);睡眠时间≤6 h的女工PMS发生率高于>6 h者(P<0.05),躯体和脑力疲劳的女工PMS发生率高于无疲劳者(P<0.05)。logistic回归分析结果显示母系存在过早或过晚绝经(OR=1.572,95%CI:1.320~1.872)、高血压(OR=1.579,95%CI:1.127~2.213)、饮酒(OR=1.286,95%CI:1.080~1.532)、缺乏体育锻炼(OR=1.598,95%CI:1.330~1.920)、睡眠时间≤6 h(OR=1.853,95%CI:1.518~2.263)、近6个月因身体原因请假(OR=1.614,95%CI:1.226~2.123)、躯体疲劳(OR=2.384,95%CI:1.887~3.012)、脑力疲劳(OR=5.649,95%CI:4.382~7.283)、职业有害因素联合暴露(OR=1.329,95%CI:1.108~1.593)、工作体位为长时间坐姿(OR=2.014,95%CI:1.271~3.190)、搬运重物(OR=1.505,95%CI:1.178~1.923)的女工发生PMS的风险较高(P<0.05)。列线图模型ROC曲线的C-index为0.748(95%CI:0.729~0.766),Bootstrap方法内部验证结果显示校准图形中标准曲线和预测曲线贴合良好,绝对误差为0.008,说明本列线图模型校准度良好。[结论]石化女工出现PMS是多因素长时间暴露的结果。本研究建立的列线图模型有良好预测能力,作为监测评估石油行业女性生殖健康的工具有一定应用价值。

[Background]Women face more reproductive health problems in their whole life cycle.Occupational exposure to harmful factors in the petrochemical industry may have a synergistic effect on women’s existing health problems.[Objective]To analyze the influencing factors of perimenopausal syndrome(PMS)in female workers in petrochemical industry,and establish a nomogram model of the risk of PMS in female workers,so as to provide a easy and quick health monitoring and evaluation method for female workers.[Methods]A total of 2653 perimenopausal female workers aged 45-55 years old were selected from a petrochemical enterprise.A questionnaire survey was conducted to collect information on demographic characteristics,occupational characteristics,psychological status,and reproductive health information.The prevalence of PMS of female workers was evaluated by the Kupperman Index Scale,the physical fatigue and mental fatigue were evaluated by the Fatigue Scale.A linear graph prediction model was established by multiple logistic regression.A nomogram was presented and C-index was used to verify the differentiation of the model.Then Bootstrap method was used for internal validation.[Results]Among the 2653 female worker,a total of 1306 cases(49.2%)presented PMS with a Kupperman score≥7.The main symptoms were fatigue(79.95%),irritability(71.32%),and insomnia(66.79%).Significant differences in PMS prevalence were found among female workers of different age,body mass index,and working posture groups(P<0.05).The participants with alcohol drinking,maternal premature or late menopause,hypertension,lack of physical exercise,heavy lifting,sick leave in the last 6 months,combined occupational exposures to dust,chemicals,noise[>80 dB(A)],or electromagnetic field,and not wearing protective masks,gloves or protective earplugs reported higher prevalence rates of PMS(P<0.05).The prevalence rate of PMS in female workers with sleep duration≤6 h was higher than that with>6 h(P<0.05),and higher in female workers with physical and mental fatigue than in those without(P<0.05).The results of logistic regression analysis showed that those with maternal premature or late menopause(OR=1.572,95%CI:1.320-1.872),hypertension(OR=1.579,95%CI:1.127-2.213),alcohol drinking(OR=1.286,95%CI:1.080-1.532),no physical exercise(OR=1.598,95%CI:1.330-1.920),sleep duration≤6 h(OR=1.853,95%CI:1.518-2.263),sick leave in recent 6 months(OR=1.614,95%CI:1.226-2.123),physical fatigue(OR=2.384,95%CI:1.887-3.012),mental fatigue(OR=5.649,95%CI:4.382-7.283),combined exposure to occupational harmful factors(OR=1.329,95%CI:1.108-1.593),long-time sitting(OR=2.014,95%CI:1.271-3.190),and heavy lifting(OR=1.505,95%CI:1.178-1.923)showed a higher risk of reporting PMS(P<0.05).The C-index from the ROC curve of the nomogram model was 0.748(95%CI:0.729-0.766).The results of Bootstrap validation showed that the standard curve and the predicted curve almost overlapped,and the absolute error was 0.008,indicating that the model fitness was good.[Conclusion]PMS in female petrochemical workers may occur due to long-term exposures to multiple factors.The established nomogram model has good predictive ability and could be applied to monitor and evaluate female reproductive health in petroleum industry.