Development and Validation of a New Nomogram Model for Predicting Acute Ischemic Stroke in Elderly Patients with Non-Valvular Atrial Fibrillation: A Single-Center Cross-Sectional Study

Figures & data

Figure 1 Flow chart of the study selection.

Abbreviation: NVAF, non-valvular atrial fibrillation.

Table 1 Comparison of Baseline Data in the Non-Stroke Group and Stroke Group

Variables	Non-Stroke Group (n=158)	Stroke Group (n=143)	P
General clinical data
Age (years)	75.49±8.83	78.74±8.77	0.002*
Male [n (%)]	77(48.7)	69(48.3)	0.933
Smoking [n (%)]	37(23.4)	46(32.2)	0.090
Drinking [n (%)]	31(19.6)	27(18.9)	0.871
Hypertension [n (%)]	100(63.3)	107(74.8)	0.031*
Diabetes [n (%)]	35(22.2)	42(29.4)	0.152
Coronary heart disease [n (%)]	36(22.8)	36(25.2)	0.627
Heart failure [n (%)]	41(25.9)	44(30.8)	0.354
Stroke [n (%)]	31(19.6)	82(57.3)	<0.001*
Hyperlipemia [n (%)]	8(5.1)	19(13.3)	0.013*
Systolic pressure (mmHg)	130.40±19.24	143.97±23.87	<0.001*
Diastolic pressure (mmHg)	79.37±14.54	83.04±13.83	0.442
Type of atrial fibrillation
Paroxysmal	82(51.9)	58(40.6)	0.049*
Persistent	76(48.1)	85(59.4)	<0.001*
CHA2DS2-VASc score	3.37±1.46	5.39±1.13
Laboratory data
TG (mmol/L)	1.06(0.81,1.45)	1.06(0.83,1.43)	0.987
TC (mmol/L)	3.59(3.05,4.44)	3.81(3.16,4.73)	0.049*
HDL-C (mmol/L)	1.21±0.36	1.10±0.26	0.002*
LDL-C (mmol/L)	2.16±0.79	2.32±0.82	0.098
Lp(a) (mg/L)	139.04(76.10,212.44)	251.71(154.40,426.63)	<0.001*
HCY (µmol/L)	14.92(12.46,19.95)	13.91(11.15,17.71)	0.059
FBG (mmol/L)	5.80±1.82	6.03±2.15	0.307
UA (µmol/L)	322.83±95.12	331.96±83.12	0.378
SCR (µmol/L)	82.10±23.45	83.63±18.69	0.528
Cystatin C (mg/L)	1.19(1.05,1.40)	1.23(1.09,1.40)	0.245
GFR (mL/min)	73.52±19.86	68.13±15.10	0.008*
D-dimer (mg/L)	0.39(0.22,0.75)	0.76(0.36,1.62)	<0.001*
NC (×109/L)	3.52(2.55,4.51)	4.15(3.43,5.60)	<0.001*
MC (×109/L)	0.50±0.21	0.54±0.20	0.109
LC (×109/L)	1.56±0.70	1.37±0.57	0.010*
PLT (×109/L)	183.28±69.49	183.3±52.25	0.998
NLR	2.33(1.69,3.39)	3.21(2.28,5.28)	<0.001*
HCT (%)	40.02±4.30	39.68±4.83	0.521
RDW (%)	12.92±0.73	13.68±1.27	<0.001*
Echocardiography data
LAD (mm)	43.18±4.94	44.59±5.20	0.016*
LVEDD (mm)	49.38±4.18	49.29±4.14	0.858
LVEF (%)	62.81±6.51	61.86±6.34	0.201

Note: *P<0.05.

Abbreviations: TG, triglyceride; TC, total cholesterol; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; Lp(a), lipoprotein(a); HCY, homocysteine; FBG, fasting blood glucose; SCr, serum creatinine; GFR, glomerular filtration rate; NC, neutrophil count; LC, lymphocyte count; PLT, platelet count; HCT, hematocrit; RDW, red cell distribution width; NLR, neutrophil-to-lymphocyte ratio; LAD, left atrial diameter; LVEDD, left ventricular end-diastolic dimension; LVEF, left ventricular ejection fraction.

Figure 2 (A and B) Predictor plots screened by least absolute shrinkage and selection operator (LASSO) regression analysis.

Table 2 Multivariate Logistic Regression Analysis of Influencing Factors for AIS in Elderly Patients with NVAF

Variables	β	SE	OR (95% CI)	P
Systolic pressure	0.027	0.008	1.027	<0.001*
Stroke	1.448	0.335	4.256	<0.001*
Hyperlipemia	1.472	0.582	4.357	0.011*
HDL-C	−1.800	0.573	0.165	0.002*
Lp(a)	0.005	0.001	1.005	<0.001*
RDW	0.547	0.195	1.729	0.005*
NLR	0.227	0.088	1.255	0.010*
D- dimer	0.143	0.097	1.154	0.140

Note: *P<0.05.

Abbreviations: HDL-C, high-density lipoprotein cholesterol; Lp(a), lipoprotein(a); RDW, red cell distribution width; NLR, neutrophil-to-lymphocyte ratio.

Figure 3 Nomogram for predicting the acute ischemic stroke (AIS) in elderly patients with non-valvular atrial fibrillation (NVAF).

Abbreviations: NLR, neutrophil-to-lymphocyte ratio; RDW, red cell distribution width; Lp(a), lipoprotein(a); systolic pressure, history of stroke, hyperlipidemia; HDL-C, high-density lipoprotein cholesterol.

Figure 4 (A) Receiver operating characteristic (ROC) curve of nomogram model in predicting the efficacy of acute ischemic stroke (AIS) in elderly patients with non-valvular atrial fibrillation (NVAF) (training sample). (B) ROC curve of nomogram model in predicting the efficacy of AIS in elderly patients with NVAF (validation sample).

Figure 5 (A) Calibration curve of the training sample. (B) Calibration curve of validation sample. (C) Clinical decision curve analysis (DCA) of training sample for nomogram models. (D) Clinical decision curve analysis (DCA) of validation sample for nomogram models.

Table 3 Comparison of ROC Curve Between Nomogram Predictive Model and CHA₂DS₂-VASc Score

Predictive Model	AUC	95% CI	Sensitivity (%)	Specificity (%)	P
Nomogram predictive model	0.8811	0.8430–0.9193	0.7552	0.8797	<0.001*
CHA2DS2-VASc score	0.8571	0.8177–0.8965	0.7832	0.7841	<0.001*

Note: *P<0.05.

Abbreviations: AUC, area under the curve; CI, confidence interval.

Figure 6 (A) Comparison of ROC curve between nomogram predictive model and CHA2DS2-VASc score. (B) Comparison of clinical decision curve analysis (DCA) between nomogram predictive model and CHA2DS2-VASc score. Model 1: nomogram predictive model. Model 2: CHA2DS2-VASc score.