Baseline characteristics
A total of 1,695 patients with ACS were included in this study. A total of 429 (25.3%), 434 (25.6%) patients with NSTEMI, and 832 (49.1%) patients had STEMI, NSTEMI, and UA, respectively. The median patient age was 62 (54,71) years. There were 1,208 (71.3%) male patients and 959 (56.6%) male patients with DM. Patients with ACS were divided into three groups according to non-HDL cholesterol tertiles. Patients in the T3 group were younger and had higher LDL-C, TC, TG, and BMI than patients in the T1 and T2 groups (all P < 0.05). Detailed baseline characteristics are shown in Table 1.
Table 1
Baseline characteristics according to tertiles of the non-HDL-C
| Total (N = 1,695) | T1 (N = 577) | T2 (N = 553) | T3 (N = 565) | P |
Glycemia(mmol/L) | 7.71(5.83,11.45) | 7.24(5.62,11.20) | 7.76(5.86,11.39) | 8.24(6.02,11.66) | 0.006 |
Age(years) | 62 (54,71) | 64 (56,72) | 63 (54,71) | 58 (52,69) | < 0.001 |
BMI(kg/m2) | 23.88 (21.80,26.42) | 23.51(21.45,26.11) | 23.88 (21.93,26.12) | 24.04 (22.04,26.89) | 0.017 |
Creatinine(µmol/L) | 78 (66,94) | 81 (69,99) | 78 (65,92) | 74 (63,92) | < 0.001 |
hs-CRP(mg/L) | 3.77(1.21,14.30) | 3.46(1.10,14.22) | 4.10(1.49,17.07) | 3.77(1.20,11.99) | 0.086 |
SBP (mmHg) | 123 (111,136) | 122 (111,135) | 122 (111,135) | 125 (112,138) | 0.254 |
DBP (mmHg) | 75 (67,84) | 75 (68,83) | 75 (67,84) | 76 (68,86) | 0.107 |
LVEF (%) | 56 (46,61) | 56 (47,61) | 55 (44,60) | 56 (47,60) | 0.174 |
eGFR (mL/min/1.73 m2) | 87.73(70.43,105.03) | 84.30(68.13,99.87) | 88.35(70.97,104.67) | 92.06(73.07,108.71) | < 0.001 |
Hb (g/L) | 138 (125,151) | 136 (122,147) | 138 (125,150) | 142 (130,154) | < 0.001 |
HbA1c (%) | 6.50 (5.80,8.00) | 6.50 (5.80,7.60) | 6.50 (5.70,8.30) | 6.70 (5.80,8.40) | 0.018 |
HDL-C (mmol/L) | 1.09 (0.93,1.27) | 1.03 (0.87,1.20) | 1.09 (0.93,1.26) | 1.16 (1.01,1.31) | < 0.001 |
LDL-C (mmol/L) | 2.96 (2.35,3.60) | 2.13 (1.81,2.38) | 3.00 (2.76,3.25) | 3.89 (3.60,4.29) | < 0.001 |
TC (mmol/L) | 4.85 (3.97,5.76) | 3.65 (3.20,4.01) | 4.86 (4.56,5.18) | 6.19 (5.74,6.83) | < 0.001 |
TG (mmol/L) | 1.88 (1.26,2.89) | 1.39 (1.02,2.16) | 1.83 (1.30,2.83) | 2.47 (1.71,3.95) | < 0.001 |
Uric acid(µmol/l) | 353 (294,431) | 349 (295,425) | 350 (291,428) | 364 (299,441) | 0.129 |
Male | 1,208 (71.3) | 445 (77.1) | 382 (69.1) | 381 (67.4) | < 0.001 |
Smoking | | | | | < 0.001 |
Current | 714 (42.1) | 235 (40.7) | 233 (42.1) | 246 (43.5) | |
Former | 210 (12.4) | 108 (18.7) | 51 (9.2) | 51 (9.0) | |
Never | 771 (45.5) | 234 (40.6) | 269 (48.6) | 268 (47.4) | |
Previous MI | 559 (33) | 249 (43.2) | 160 (28.9) | 150 (26.5) | < 0.001 |
Previous CABG | 14 (0.8) | 3 (0.5) | 6 (1.1) | 5 (0.9) | 0.567 |
Previous PCI | 190 (11.2) | 115 (19.9) | 45 (8.1) | 30 (5.3) | < 0.001 |
Hypertension grade | | | | | 0.468 |
1 | 97 (5.7) | 31 (5.4) | 33 (6.0) | 33 (5.8) | |
2 | 294 (17.3) | 109 (18.9) | 91 (16.5) | 94 (16.6) | |
3 | 643 (37.9) | 226 (39.2) | 218 (39.4) | 199 (35.2) | |
Dyslipidaemia | 754 (44.5) | 126 (21.8) | 226 (40.9) | 402 (71.2) | < 0.001 |
Previous stroke | 216 (12.7) | 101 (17.5) | 59 (10.7) | 56 (9.9) | < 0.001 |
Family history | 8 (0.5) | 0 (0.0) | 2 (0.4) | 6 (1.1) | 0.029 |
Insulin | 433 (25.5) | 146 (25.3) | 136 (24.6) | 151 (26.7) | 0.706 |
Oral hypoglycaemic drugs | 763 (45) | 264 (45.8) | 238 (43.0) | 261 (46.2) | 0.517 |
DM | 959 (56.6) | 321 (55.6) | 305 (55.2) | 333 (58.9) | 0.378 |
Diagnosis on admission | | | | | < 0.001 |
STEMI | 429 (25.3) | 100 (17.3) | 157 (28.4) | 172 (30.4) | |
NSTEMI | 434 (25.6) | 126 (21.8) | 152 (27.5) | 156 (27.6) | |
UA | 832 (49.1) | 351 (60.8) | 244 (44.1) | 237 (41.9) | |
Gensini score | 50 (32,78) | 45 (25,72) | 52 (32,80) | 52 (35,80) | < 0.001 |
Data are presented as means ± SDs, medians (interquartile ranges), or n (%).
Clinical outcomes
Table 2 shows the relationships between all patients with ACS and high Gensini scores. In Model 1, single-factor logistic regression analysis showed that non-HDL-C was correlated with high Gensini scores (OR = 1.09; 95%CI = 1.02–1.18; P = 0.016). The T3 group had a 1.37 times greater risk of high Gensini scores compared to the T1 group (OR = 1.37; 95%CI = 1.08–1.73; P = 0.008). In Model 2, adjusted for potential risk factors, multifactorial logistic regression analysis showed that non-HDL-C was an independent risk factor for high Gensini scores (OR = 1.20; 95%CI = 1.10–1.31; P < 0.001). The T2 group had a 1.61 times higher risk (OR = 1.61; 95%CI = 1.25–2.07; P < 0.001) and the T3 group had a 1.8 times higher risk (OR = 1.80; 95%CI = 1.37–2.36; P < 0.001) of developing high Gensini scores compared to the T1 group.
The results of the restricted cubic spline analysis are shown in Fig. 2, which demonstrates a dose-response relationship between non-HDL-C and high Gensini scores in patients with ACS (nonlinear P = 0.513).
Table 2
Associations between non-HDL-C and Gensini score
| Events/N | Model 1 | | | | Model 2 | | |
| | OR | 95CI% | P | | OR | 95CI% | P |
non-HDL-C | 834/1,695 | 1.09 | 1.02–1.18 | 0.016 | | 1.20 | 1.10–1.31 | < 0.001 |
T1 | 251/577 | Reference | | | | Reference | | |
T2 | 293/553 | 1.46 | 1.16–1.85 | 0.001 | | 1.61 | 1.25–2.07 | < 0.001 |
T3 | 290/565 | 1.37 | 1.08–1.73 | 0.008 | | 1.80 | 1.37–2.36 | < 0.001 |
P for trend | | | | 0.008 | | | | < 0.001 |
Model 1: Unadjusted.
Model 2: Adjusted for age, BMI, SBP, DBP, TG, eGFR, creatinine, glycemia, uric acid, haemoglobin, hs-CRP, HbA1c, LVEF, sex, smoking, Previous MI, Previous CABG, Previous PCI, Hypertension, Previous stroke, family history, and insulin.
Subgroup analyses
Further categorisation was performed according to sex, age groups, and DM status (diabetic or non-diabetic) to explore the correlation between non-HDL-C and high Gensini scores in the subgroups.
Table 3 shows the results of the analysis of patients with ACS with different DM statuses. The results showed no significant interaction between sex and non-HDL-C levels (interaction P-value = 0.110). After adjusting for potential risk factors, multifactorial logistic regression analysis (OR = 1.15; 95%CI = 1.02–1.29; P = 0.024) showed that non-HDL-C was an independent risk factor for high Gensini scores in patients with DM in the T2 group (OR = 1.73; 95%CI = 1.22–2.44; P = 0.002) and T3 group (OR = 1.65; 95%CI = 1.14–2.38; P = 0.007), with risks 1.73 and 1.65 times higher than that in the T1 group, respectively. In non-DM patients, after adjusting for potential risk factors, non-HDL-C remained an independent risk factor for high Gensini scores (OR = 1.29; 95%CI = 1.11–1.49; P < 0.001), with the T3 group exhibiting a 1.98 times higher risk (OR = 1.98; 95%CI = 1.30–3.04; P = 0.002) compared to the T1 group.
Table 3
Associations between non-HDL-C and Gensini scores according to different diabetes statuses
| Events/N | Mode 1 | P | | | Model 2 | | | |
| | OR | 95CI% | P | | OR | 95CI% | P | P for interaction |
Glucose metabolism state | | | | | | | | | 0.110 |
DM | 461/959 | 1.03 | 0.94–1.13 | 0.528 | | 1.15 | 1.02–1.29 | 0.024 | |
T1 | 138/321 | Reference | | | | Reference | | | |
T2 | 165/305 | 1.56 | 1.14–2.14 | 0.006 | | 1.73 | 1.22–2.44 | 0.002 | |
T3 | 158/333 | 1.20 | 0.88–1.63 | 0.252 | | 1.65 | 1.14–2.38 | 0.007 | |
Non-DM | 373/736 | 1.20 | 1.07–1.35 | 0.002 | | 1.29 | 1.11–1.49 | < 0.001 | |
T1 | 113/256 | Reference | | | | Reference | | | |
T2 | 128/248 | 1.35 | 0.95–1.92 | 0.093 | | 1.43 | 0.97–2.10 | 0.069 | |
T3 | 132/232 | 1.67 | 1.17–2.39 | 0.005 | | 1.98 | 1.30–3.04 | 0.002 | |
Model 1: Unadjusted.
Model 2: Adjusted for age, BMI, SBP, DBP, TG, eGFR, creatinine, glycemia, uric acid, haemoglobin, hs-CRP, HbA1c, LVEF, sex, smoking, Previous MI, Previous CABG, Previous PCI, Hypertension, Previous stroke, family history, and insulin.
Table 4 presents the results of the analysis of patients with ACS across different age groups. The results showed no significant interaction between age and non-HDL-C level (interaction P-value = 0.258). In elderly patients, after adjusting for potential risk factors, multifactorial logistic regression analysis (OR = 1.18; 95%CI = 1.04–1.33; P = 0.009) showed that non-HDL-C was an independent risk factor for high Gensini scores, and the risk of developing high Gensini scores in the T3 group (OR = 1.45; 95%CI = 1.00-2.10; P = 0.048) was 1.45 times higher than that in the T1 group. In the non-elderly age group, after adjusting for potential risk factors, multifactorial logistic regression analysis showed that non-HDL-C was an independent risk factor for high Gensini scores (OR = 1.21; 95%CI = 1.06–1.40; P = 0.007). The T2 group (OR = 2.23; 95%CI = 1.47–3.39; P < 0.001) and T3 group (OR = 2.56; 95%CI = 1.66–3.95; P < 0.001) exhibited 2.23 and 2.56 times higher risks than the T1 group, respectively, for developing high Gensini scores.
Table 4
Associations between non-HDL-C and Gensini scores according to different ages
| Events/N | Model 1 | | | | Model 2 | | | |
| | OR | 95CI% | P | | OR | 95CI% | P | P for interaction |
Age-years | | | | | | | | | 0.258 |
Age ≥ 60 | 477 /937 | 1.07 | 0.97–1.18 | 0.192 | | 1.18 | 1.04–1.33 | 0.009 | |
T1 | 180/369 | Reference | | | | Reference | | | |
T2 | 166/312 | 1.19 | 0.88–1.61 | 0.250 | | 1.36 | 0.98–1.89 | 0.069 | |
T3 | 131/256 | 1.10 | 0.80–1.51 | 0.557 | | 1.45 | 1.00-2.10 | 0.048 | |
Age < 60 | 357/758 | 1.15 | 1.03–1.29 | 0.012 | | 1.21 | 1.06–1.40 | 0.007 | |
T1 | 71/208 | Reference | | | | Reference | | | |
T2 | 127/241 | 2.15 | 1.47–3.15 | < 0.001 | | 2.23 | 1.47–3.39 | < 0.001 | |
T3 | 159/309 | 2.05 | 1.42–2.94 | < 0.001 | | 2.56 | 1.66–3.95 | < 0.001 | |
Model 1: Unadjusted.
Model 2: Adjusted for BMI, SBP, DBP, TG, eGFR, creatinine, glycemia, uric acid, Hb, hs-CRP, HbA1c, LVEF, sex, smoking, Previous MI, Previous CABG, Previous PCI, Hypertension, Previous stroke, family history, and insulin.
Table 5 shows the results of the analysis of patients with ACS stratified by sex. No significant interaction between sex and non-HDL-C levels was observed (interaction P = 0.491). In male patients, after adjusting for potential risk factors, multifactorial logistic regression analysis showed that non-HDL-C was an independent risk factor for high Gensini scores (OR = 1.15; 95%CI = 1.04–1.28; P = 0.010). The T2 group (OR = 1.57; 95%CI = 1.16–2.11; P = 0.003) and T3 group (OR = 1.60; 95%CI = 1.16–2.22; P = 0.004) exhibited 1.57 and 1.60 times higher, respectively, for developing high Gensini scores compared to the T1 group. In female patients, after adjusting for potential risk factors, multifactorial logistic regression analysis showed that non-HDL-C was an independent risk factor for high Gensini scores (OR = 1.31; 95%CI = 1.10–1.57; P = 0.002). The T2 group (OR = 1.90; 95%CI = 1.14–3.17; P = 0.014) and T3 group (OR = 2.54; 95%CI = 1.48–4.35; P < 0.001) exhibited risks 1.90 and 2.54 times higher, respectively, for developing high Gensini scores compared to the T1 group.
Table 5
Associations between non-HDL-C and Gensini scores according to different sexes
| Events/N | Model 1 | | | | Model 2 | | | |
| | OR | 95CI% | P | | OR | 95CI% | P | P for interaction |
Sex | | | | | | | | | 0.491 |
Male | 601 /1208 | 1.07 | 0.98–1.17 | 0.116 | | 1.15 | 1.04–1.28 | 0.010 | |
T1 | 199 /445 | Reference | | | | Reference | | | |
T2 | 207 /382 | 1.46 | 1.11–1.92 | 0.007 | | 1.57 | 1.16–2.11 | 0.003 | |
T3 | 195 /381 | 1.30 | 0.99–1.71 | 0.064 | | 1.60 | 1.16–2.22 | 0.004 | |
Female | 233 /487 | 1.16 | 1.01–1.33 | 0.034 | | 1.31 | 1.10–1.57 | 0.002 | |
T1 | 52 /132 | Reference | | | | Reference | | | |
T2 | 86 /171 | 1.56 | 0.98–2.47 | 0.059 | | 1.90 | 1.14–3.17 | 0.014 | |
T3 | 95 /184 | 1.64 | 1.04–2.58 | 0.032 | | 2.54 | 1.48–4.35 | < 0.001 | |
Model 1: Unadjusted.
Model 2: Adjusted for age, BMI, SBP, DBP, TG, eGFR, creatinine, glycemia, uric acid, haemoglobin, hs-CRP, HbA1c, LVEF, smoking, Previous MI, Previous CABG, Previous PCI, Hypertension, Previous stroke, family history, and insulin.
Non-HDL-C predicts the incremental effect of high Gensini scores in ACS patients
In the analysis of patients with ACS, an ROC curve was constructed (Fig. 3), and the ability of the baseline risk model, non-HDL-C model (baseline risk model + non-HDL-C), and TC model (baseline risk model + TC) to predict high Gensini scores in patients with ACS was further assessed. Table 6 shows the C-statistics, NRI, and IDI for the different models. The C-statistic showed that the non-HDL-C model had an incremental predictive effect over the baseline risk model (0.6584 [95%CI; 0.6326–0.6841] vs 0.6483 [95%CI; 0.6224–0.6743], P = 0.043). NRI (0.0382 [95%CI = 0.0065–0.0699]; P = 0.018) and IDI (0.0091 [95%CI = 0.0047–0.0136]; P < 0.001] were computed, both showing a significant incremental predictive effect of incorporating non-HDL-C into the baseline risk model.
Table 7 shows the results of the internal validation of the model using the bootstrap method with 1000 resamplings with an AUC value of 0.6234; the AUC value of the non-HDL-C model constructed for the study was 0.6584, which is a similar result and is considered to indicate a certain level of robustness of the model.
Table 6
Incremental predictive value and predictive power of various models with NRI, IDI, and C-statistics
Model | C-statistic (95%Cl) | P | NRI (95%Cl) | P | IDI (95%Cl) | P |
Baseline risk model | 0.6483(0.6224–0.6743) | Ref | Ref | | Ref | |
+TC | 0.6566(0.6308–0.6824) | 0.065 | 0.0300(-0.0005-0.0604) | 0.054 | 0.0073(0.0033–0.0113) | < 0.001 |
+Non-HDL-C | 0.6584(0.6326–0.6841) | 0.043 | 0.0382(0.0065–0.0699) | 0.018 | 0.0091(0.0047–0.0136) | < 0.001 |
Baseline risk model includes age, BMI, SBP, DBP, TG, eGFR, creatinine, glycemia, uric acid, Hb, hs-CRP, HbA1c, LVEF, sex, smoking, Previous MI, Previous CABG, Previous PCI, hypertension, Previous stroke, family history, and insulin.
Table 7
Results of bootstrap-based internal validation of models
AUC | Sensitivity | Specificity |
0.6182 | 0.6234 | 0.4513 |