| 摘要 |
背景 老年抑郁症的临床表现不典型，在临床中识别率较低，尤其是重度发作可能会增加自杀的风险，因此需要构建能够快速识别老年抑郁重度发作的预测模型来提高早期识别率。方法 选取2021年1月至2023年6月期间在上海市宝山区精神卫生中心就诊的老年抑郁症患者为调查对象，进行量表评估和血液学检测，采用Logistic 回归分析探讨老年抑郁症重度发作的影响因素，用Risk score法构建老年抑郁症重度发作简易预测模型，并检验其预测效果。结果 共有402例老年抑郁症纳入研究，378例被试完成基线调查和血液检测，其中重度发作有93例（24.60%）。多因素Logistic 回归分析结果显示HAMA总分[OR=1.218,95%CI(1.161~1.278)]、白介素-6[OR=1.009,95%CI(1.002~1.017)]、LDL [OR=1.701,95%CI(1.213~2.385)]均是老年抑郁症重度发作的危险因素（P＜0.05或P＜0.001）。基于Logistic回归建立的Risk score预测模型为：Risk Score = IL-6+21.89 HAMA+59 LDL，该模型ROC曲线下面积（AUC）为0.922 (p < 0.001, 95%CI = 0.892~0.951)。约登指数最大时为 0.735, 截断值为705.165分，灵敏度为90.3%,特异度为83.2%。结论 老年抑郁症患者重度发作比率较高，HAMA总分、白介素-6和低密度脂蛋白是其危险因素，基Logistic回归建立的Risk score预测模型预测老年抑郁症重度发作的灵敏度为90.3%,特异度为83.2%。
|
| Abstract |
Background The clinical manifestations of depression in the elderly are atypical, and the recognition rate in the clinic is low, especially severe episodes may increase the risk of suicide, so it is necessary to construct a prediction model that can quickly identify severe episodes of depression in the elderly to improve the early recognition rate.Methods The elderly patients with depression who were treated in Shanghai Baoshan District Mental Health Center from January 2021 to June 2023 were selected as the survey subjects, and the scale evaluation and hematology tests were carried out, and the influencing factors of severe onset of geriatric depression were explored by logistic regression analysis, and a simple prediction model of severe onset of geriatric depression was constructed by risk score method, and its prediction effect was tested.Results A total of 402 cases of geriatric depression were included in the study, and 378 subjects completed the baseline survey and blood tests, of which 93 (24.60%) had severe episodes. The results of multivariate logistic regression analysis showed that HAMA total score [OR=1.218, 95%CI (1.161~1.278)], interleukin-6 [OR=1.009, 95%CI (1.002~1.017)], and LDL [OR=1.701, 95%CI (1.213~2.385)] were risk factors for severe episodes of depression in the elderly (P<0.05 or P<0.001). The risk score prediction model based on logistic regression was as follows: Risk score = IL-6+21.89 HAMA+59 LDL, and the area under the ROC curve (AUC) of the model was 0.922 (p < 0.001, 95%CI = 0.892~0.951). The maximum Youden index was 0.735, the cut-off value was 705.165 points, the sensitivity was 90.3%, and the specificity was 83.2%.Conclusion The risk score prediction model based on logistic regression has a sensitivity of 90.3% and a specificity of 83.2% for predicting severe episodes of depressive disorder in elderly patients with high rates of severe attacks, with HAMA total score, interleukin-6 and low-density lipoprotein as risk factors.
|
| KeyWord |
senile depression, severe attacks, risk factors, prediction, prediction model
|