Analysis of factors influencing falls in elderly patients with multimorbidity in the chronic phase of cerebral infarction and construction of a nomogram prediction model

Liu Dandan; Li Yuanyuan; Sun Aihua; Yu Xiujuan; Zhao Yanqiu; Chen Jin

doi:10.3969/j.issn.1672-5921.2026.02.003

Abstract：Objective To analyze factors influencing falls in elderly patients with multimorbidity in the chronic phase of cerebral infarction and construct a nomogram model for predicting the risk of falls.Methods Consecutive elderly patients(aged≥60 years)with multimorbidity(co-occurrence of≥2 mental conditions)in chronic phase of cerebral infarction(latest onset to admission time≥1 month)who were admitted to the Geriatrics Department of the Second People's Hospital of Hefei from August 2022 to October 2024 were retrospectively included in this study.General and clinical data were collected from patients enrolled,including gender,age,body mass index,number(≥3 or 2)and types of comorbidities(hypertension,diabetes,coronary heart disease,chronic heart failure,sarcopenia,osteoporosis,and other chronic diseases such as hypothyroidism and hyperplasia of prostate,et al),nutritional status(nutritional risk screening 2002[NRS-2002]scale score),and fall risk score(Tinetti performance-oriented mobility assessment[Tinetti POMA]scale),limb function(modified Rankin scale[mRS]),cognitive status(Montreal cognitive assessment[MoCA]scale score),lower limb hemiplegia(Brunnstrom stage),and incidence of falls during follow-up period(6months after admission).On the morning following admission,fasting venous blood samples were collected to measure the levels of glycated hemoglobin,fasting blood glucose,albumin,prealbumin,triglyceride,total cholesterol,low-density lipoprotein,high-density lipoprotein,creatinine,and uric acid.According to whether falls occurred during 6 months follow-up period,the patients were divided into the fall group and the non-fall group.General data,clinical data and laboratory indexes of the two groups were compared.Factors with statistically differences found by univariate analysis were included in multivariate Logistic regression to identify the influencing factors of falls in elderly patients with multimorbidity in the chronic phase of cerebral infarction,and a nomogram prediction model was constructed.The receiver operating characteristic(ROC)curve,calibration curve,and clinical decision curve were used to assess the predictive efficacy of the nomogram model.Results A total of 147 elderly patients with coexistence of multiple diseases in chronic phase of cerebral infarction were included,including 59 males and 88 females,age ranged from 61 to 93 years old,with an average age of(77±7)years old.The time from last onset to admission was 17-122months,with an average of(63±17)months.The number of comorbidities was 2-6types,with a median value of 3(2,4)kinds.The top three comorbidity types were hypertension(94.56%[139/147]),diabetes(57.14%[84/147]),sarcopenia(42.86%[63/147]),and osteoporosis(42.86%[63/147]).During the follow-up period,54(36.73%)patients experienced falls and were assigned to the fall group.The remaining 93 patients were assigned to the non-fall group.(1)Compared with the non-fall group,the fall group had significantly older age(82[76,85]years vs.75[69,82]years),higher NRS-2002 scale score(4[3,4]points vs.3[2,3]points),higher incidence rate of≥3 types of comorbidities(94.44%[51/54]vs.44.09%[41/93]),higher prevalence of osteoporosis(70.37%[38/54]vs.26.88%[25/93])and sarcopenic(81.48%[44/54]vs.20.43%[19/93]);and,significantly lower Tinetti POMA scale score([16.59±1.09]points vs.[17.41±1.02]points)and albumin level(37.65[35.60,39.73]g/L vs.39.90[37.55,41.95]g/L;all P＜0.01).(2)Multivariate Logistic regression analysis showed that osteoporosis(OR,3.059,95%CI 1.063-8.797),sarcopenia(OR,23.698,95%CI 5.157-108.912),having≥3 types of comorbidities(OR,16.839,95%CI 2.227-127.348),and high NRS-2002 scale score(OR,5.346,95%CI 2.035-14.047)were risk factors for falls in elderly patients with multimorbidity in the chronic phase of cerebral infarction(all P＜0.05),while high Tinetti POMA scale score was a protective factor(OR,0.434,95%CI 0.217-0.868,P=0.018).(3)A nomogram for falls in elderly patients with multimorbidity in the chronic phase of cerebral infarction was plotted based on multivariate Logistic regression analysis results.ROC curve showed that the area under the curve of the nomogram model for predicting falls in elderly patients with coexistence of multiple diseases in chronic phase of cerebral infarction was 0.894(95%CI 0.840-0.948).The sensitivity and specificity were 77.80%and 90.30%,respectively.Hosmer-Lemeshow goodness of fit test showed that the nomogram model fitted well(χ2=9.612,P=0.293).The decision curve showed that better clinical benefits could be achieved at the threshold between 0.01 and 0.99.Conclusions Hypertension and diabetes are common comorbidities in elderly patients with multimorbidity in the chronic phase of cerebral infarction.Such patients have a relatively higher incidence of falls.Moreover,the incidence of falls is related to factors such as osteoporosis,sarcopenia,high number of comorbidities,high nutritional risk,and low Tinetti POMA score.The nomogram model constructed based on above factors demonstrates good predictive efficacy and can serve as a quantitative tool for prediction of falls and generation of prevention strategies in clinical practice.