Skip to main content

Table 3 Multivariable logistic regression analysis in the training cohorts

From: Development, validation, and visualization of a web-based nomogram to predict 5-year mortality risk in older adults with hypertension

Variables Full Modela Final Modelb
  OR (95% CI) P value OR (95% CI) P value
Factors Selected
 Age, years 1.101 [1.079, 1.124] < 0.001 1.105 [1.085, 1.126] < 0.001
Sex
 Male Ref.   Ref.  
 Female 0.452 [0.346, 0.587] < 0.001 0.504 [0.397, 0.639] < 0.001
Diabetes
 Absence Ref.   Ref.  
 Presence 1.409 [0.982, 2.007] 0.060 1.415 [1.112, 1.798] 0.005
Cardiovascular disease
 Absence Ref.   Ref.  
 Presence 1.821 [1.443, 2.296] < 0.001 1.814 [1.449, 2.268] < 0.001
Body mass index, kg/m2
 < 25 Ref.   Ref.  
 25-30 0.728 [0.562, 0.942] 0.016 0.714 [0.554, 0.919] 0.009
 ≥30 0.737 [0.549, 0.990] 0.043 0.733 [0.550, 0.976] 0.034
Smoking
 Never Ref.   Ref.  
 Former 1.315 [1.043, 1.659] 0.021 1.313 [1.043, 1.653] 0.021
 Current 2.459 [1.653, 3.635] < 0.001 2.687 [1.830, 3.921] < 0.001
Lipid-lowering drugs
 Absence Ref.   Ref.  
 Presence 0.676 [0.520, 0.873] 0.003 0.672 [0.522, 0.861] 0.002
 Systolic blood pressure, mmHg 1.005 [1.000, 1.010] 0.061 1.005 [1.000, 1.010] 0.030
 Hemoglobin, g/dL 0.898 [0.823, 0.981] 0.017 0.904 [0.833, 0.981] 0.016
 Albumin, g/dL 0.926 [0.892, 0.962] < 0.001 0.924 [0.890, 0.958] < 0.001
 Blood urea nitrogen, mg/dL 1.083 [1.037, 1.131] < 0.001 1.087 [1.050, 1.125] < 0.001
Factors Not Selected
Ethnicity
 Non-Hispanic White Ref.   Ref.  
 Non-Hispanic Black 0.875 [0.628, 1.213] 0.427 NA NA
 Other Hispanic 0.556 [0.240, 1.157] 0.139 NA NA
 Other races 0.878 [0.627, 1.220] 0.444 NA NA
Education
 less than high school Ref.   Ref.  
 high school or above 0.942 [0.743, 1.196] 0.624 NA NA
The income to poverty ratios
 ≤1.3 Ref.   Ref.  
 > 1.3 1.156 [0.903, 1.477] 0.248 NA NA
 Diastolic blood pressure, mmHg 1.003 [0.994, 1.011] 0.528 NA NA
Antihypertensive drugs
 Absence Ref.   Ref.  
 Presence 0.970 [0.765, 1.230] 0.798 NA NA
Hypoglycemic agents
 Absence Ref.   Ref.  
 Presence 1.141 [0.763, 1.714] 0.522 NA NA
Antiplatelet drugs
 Absence Ref.   Ref.  
 Presence 0.841 [0.537, 1.294] 0.440 NA NA
Marital status
 Married Ref.   Ref.  
 Widowed or divorced 1.230 [0.963, 1.572] 0.097 NA NA
 Single 1.041 [0.586, 1.776] 0.886 NA NA
 Total cholesterol /High density lipoprotein cholesterol 1.005 [0.925, 1.091] 0.901 NA NA
 White blood cells, 1000 cells/uL 1.020 [0.984, 1.054] 0.213 NA NA
 HbA1C, % 0.952 [0.826, 1.091] 0.490 NA NA
 Platelet, 1000 cells/uL 1.001 [0.999, 1.002] 0.460 NA NA
 eGFR, mg/min/1.73 m2 0.999 [0.992, 1.006] 0.788 NA NA
 Model performance Full Model   Final Model  
 AIC 2305.8   2286.0  
 C-index (95%CI) 0.762(0.739-0.784)   0.759(0.736-0.782)  
 Hosmer-Lemeshow testc χ2 = 6.065 (P value =0.640)   χ2 = 10.976 (P value =0.203)  
  1. aThe full model included 24 predictors: age, sex, ethnicity, education, marital status, the income to poverty ratios, smoking, cardiovascular disease, diabetes, antihypertensive, hypoglycemic, lipid-lowering, and antiplatelet drugs, BMI, SBP, DBP, TC/HDLC, HbA1C, white blood cells, hemoglobin, platelet, albumin, BUN, and eGFR
  2. bThe final model included 11 predictors: age, sex, cardiovascular disease, diabetes, BMI, smoking, lipid-lowering drugs, SBP, hemoglobin, albumin, and BUN
  3. cThe Hosmer-Lemeshow test was applied to test the goodness of fit and a P value > 0.05 was considered good fit HbA1C hemoglobin, eGFR estimated glomerular filtration rate, CI confidence interval, AIC Akaike information criterion
Back to article page

BMC Geriatrics

ISSN: 1471-2318

Contact us