Main Clinical Characteristics Of The Study Population
In the total population of 166 patients, 62% were male, mean age was 58.9 ± 10.8 years, BMI 24.8 ± 2.6 kg/m2, and HbA1c 7.8 ± 1.0%. TyG index was between 7.80 and 10.96, with a mean value of 9.64 ± 0.63.
Clinical and biochemical characteristics according to TyG tertiles
Table 1
Clinical and biological characteristics according to the TyG tertiles
| Lowest tertile (7.80–9.37) (n = 56) | Mid tertile (9.38–9.99) (n = 55) | Highest tertile (10.00- 10.96) (n = 55) | p value |
Age (years) | 58.5 ± 10.1 | 60.6 ± 11.4 | 57.5 ± 11.0 | 0.306 |
Gender (M/F) | 36/20 | 34/21 | 33/22 | 0.897 |
BMI (kg/m2) | 23.6 ± 2.8 | 25.0 ± 2.5 | 25.8 ± 2.1 | < 0.001 |
Waist circumference (cm) | 86.0 ± 11.7 | 91.2 ± 10.9 | 94.0 ± 9.9 | 0.001 |
Male | 90.2 ± 11.9 | 94.9 ± 9.9 | 96.4 ± 9.1 | 0.041 |
Female | 78.4 ± 6.1 | 85.2 ± 10.2 | 90.4 ± 10.2 | < 0.001 |
Systolic blood pressure (mmHg) | 138.9 ± 21.3 | 139.2 ± 22.9 | 143.7 ± 23.1 | 0.456 |
Diastolic blood pressure (mmHg) | 85.9 ± 13.4 | 85.4 ± 14.1 | 85.6 ± 14.0 | 0.975 |
Duration of diabetes (years) | 4.7 ± 2.8 | 5.8 ± 4.2 | 5.9 ± 4.1 | 0.006 |
Practicing physical activity | 24 (42.9%) | 12 (21.8%) | 7 (12.7%) | 0.001 |
Smoking | 15 (26.8%) | 22 (40%) | 24 (43.6%) | 0.152 |
Metabolic syndrome | 23 (41.1%) | 42 (76.4%) | 44 (80.0%) | < 0.001 |
Total cholesterol (mg/dL) | 207.1 ± 46.3 | 224.1 ± 55.2 | 232.8 ± 52.7 | 0.03 |
Triglycerides (mg/dL) | 146.9 ± 42.2 | 209.9 ± 44.2 | 330.8 ± 70.4 | < 0.001 |
HDL-cholesterol (mg/dL) | 50.0 ± 12.4 | 41.9 ± 8.2 | 40.1 ± 7.8 | < 0.001 |
LDL-cholesterol (mg/dL) | 126.1 ± 36.7 | 130.4 ± 44.0 | 146.0 ± 51.5 | 0.042 |
Non-HDL cholesterol (mg/dl) | 157.1 ± 43.3 | 182.2 ± 54.6 | 192.7 ± 51.5 | < 0.001 |
Fasting plasma glucose (mg/dL) | 111.9 ± 27.7 | 159.7 ± 29.3 | 189.0 ± 43.2 | < 0.001 |
HbA1c (%) | 7.3 ± 0.6 | 8.0 ± 0.9 | 8.3 ± 1.0 | < 0.001 |
TyG index | 8.95 ± 0.37 | 9.67 ± 0.19 | 10.33 ± 0.23 | < 0.001 |
Insulin (mU/L) | 20.3 ± 11.6 | 27.7 ± 13.0 | 28.1 ± 12.5 | < 0.001 |
HOMA-IR index | 5.43 ± 3.36 | 10.50 ± 5.57 | 13.04 ± 6.84 | 0.001 |
eGFR (mL/min/1.73 m2) | 110.7 ± 162.0 | 91.2 ± 70.5 | 86.2 ± 31.7 | 0.46 |
Coronary artery stenosis (%) | 16 (28.6) | 16 (29.1) | 28 (50.9) | 0.021 |
Drug therapy | | | | |
Metformin | 55 (98.2%) | 53 (96.4%) | 55 (100%) | 0.359 |
Sulphonylurea | 50 (89.3%) | 51 (92.7%) | 48 (87.3%) | 0.634 |
Thiazolidinedione | 1 (1.8%) | 0 (0%) | 0 (0%) | 0.372 |
DPP-4 inhibitors | 1 (1.8%) | 1 (1.8%) | 3 (5.5%) | 0.432 |
α-glucosidase inhibitor | 4 (7.1%) | 12 (21.8%) | 14 (25.5%) | 0.029 |
Insulin | 2 (3.6%) | 0 (0%) | 1 (1.8%) | 0.369 |
Antiplatelet | 30 (53.6%) | 32 (58.2%) | 32 (58.2%) | 0.852 |
Statin | 33 (58.9%) | 32 (58.2%) | 31 (56.4%) | 0.961 |
We separated the population in tertiles of TyG index. Among the patients in the highest tertiles, BMI, waist circumference, total cholesterol, triglycerides, LDL-cholesterol, non-HDL cholesterol, plasma glucose and HbA1c levels were higher, HDL-cholesterol levels were lower, and lesser patients practiced physical activity. There was no significant difference for the current treatments across TyG tertiles.
Association of TyG index with the metabolic syndrome and HOMA-IR index
A metabolic syndrome affected 109 patients (65.7%). The percentage of patients with the metabolic syndrome was higher in the highest TyG tertiles (Table 1). TyG index was significant in detecting a metabolic syndrome with an area under the ROC curve of 0.745 (95% CI: 0.660–0.830, p < 0.001) (Fig. 1). The cut-off point of 9.145 for TyG index offered a sensitivity of 93.6% and specificity of 52.6%, positive and predictive values of 77.3% and 79.4%, respectively. TyG index correlated significantly with HOMA-IR index (r = 0,569, p < 0.001) (Fig. 2).
Results of coronary computed tomography angiography and association with TyG index
Among the 166 patients, 60 had CS ≥ 50%, including 32 with CS ≥ 70%. Twenty-one patients had 2- or 3- vessel disease (with stenosis ≥ 50%).
Compared to the patients without significant CS, the patients with CS ≥ 50% were older, had a longer diabetes duration, higher waist circumference, blood pressure, HbA1c and TyG index without significant difference for HOMA-IR, and more of them had a metabolic syndrome, were smokers, and were on statin or antiplatelet treatments (Table 2).
Table 2
Comparison between the patients with and without coronary artery stenosis (CS)
| No significant CS (n = 106) | Significant CS (n = 60) | p value |
Age (years) | 57.2 ± 10.7 | 61.9 ± 10.4 | 0.007 |
Gender (M/F) | 63/43 | 40/20 | 0.356 |
BMI (kg/m2) | 24.5 ± 2.7 | 25.3 ± 2.4 | 0.061 |
Waist circumference (cm) | 88.8 ± 11.3 | 93.1 ± 11.0 | 0.020 |
Systolic blood pressure (mmHg) | 130.2 ± 18.5 | 158.9 ± 16.2 | 0.001 |
Diastolic blood pressure (mmHg) | 80.3 ± 11.1 | 95.2 ± 12.8 | 0.001 |
Duration of diabetes (years) | 3.9 ± 3.0 | 9.3 ± 3.3 | 0.001 |
Practicing physical activity | 30 (28.3%) | 13 (21.7%) | 0.349 |
Smoking | 27 (25.5%) | 34 (56.7%) | < 0.001 |
Metabolic syndrome | 60 (56.6%) | 49 (81.7%) | < 0.001 |
Total cholesterol (mg/dl) | 222.7 ± 53.3 | 218.6 ± 50.9 | 0.624 |
Triglycerides (mg/dl) | 218.3 ± 94.3 | 247.2 ± 89.5 | 0.055 |
HDL-cholesterol (mg/dl) | 45.2 ± 11.3 | 42.0 ± 8.9 | 0.061 |
LDL-cholesterol (mg/dl) | 130.5 ± 44.9 | 141.1 ± 44.9 | 0.146 |
Non-HDL cholesterol | 177.5 ± 52.7 | 176.6 ± 50.9 | 0.910 |
Fasting plasma glucose (mg/dl) | 146.9 ± 43.5 | 165.3 ± 50.3 | 0.015 |
HbA1c (%) | 7.57 ± 0.78 | 8.45 ± 1.08 | 0.001 |
TyG index | 9.56 ± 0.61 | 9.78 ± 0.63 | 0.028 |
HOMA-IR index | 9.12 ± 6.17 | 10.54 ± 6.41 | 0.163 |
Drug therapy | | | |
Metformin | 104 (98.1%) | 59 (98.3%) | 0.919 |
Sulphonylurea | 91 (85.8%) | 58 (96.7%) | 0.027 |
Thiazolidinedione | 1 (1%) | 0 (0%) | 0.450 |
DPP4 inhibitors | 0 (0%) | 5 (8%) | 0.003 |
Alpha glucosidase inhibitor | 5 (4.7%) | 25 (41.7%) | 0.001 |
Insulin | 0 (0.0%) | 3 (5.0%) | 0.02 |
Antiplatelet | 46 (43.4%) | 48 (80%) | 0.001 |
Statin | 47 (44.3%) | 49 (81.7%) | 0.001 |
The prevalence of CS ≥ 50% was higher in the patients of the highest TyG tertiles (50.9% vs 28.6 and 29.1% in the lowest and mid tertiles, p = 0.021) (Table 1). In multivariate logistic regression analyses, TyG index was associated (p < 0.05) with CS independently of duration of diabetes, systolic blood pressure, waist circumference, LDL-cholesterol and HbA1c (Table 3).
Table 3
Multivariate regression analysis for significant coronary artery stenosis
| Odds ratio (95% CI) | p value |
Waist circumference > 90 cm male, > 80 cm female* | 0.637 (0.242–1.675) | 0.361 |
HbA1c > 7% | 1.725 (0.455–6.550) | 0.423 |
Duration of diabetes > 10 years | 10.463 (2.963–36.946) | 0.0001 |
TyG index > 10 | 2.831 (1.050–7.633) | 0.040 |
LDL-cholesterol > 100 mg/dL | 2.957 (1.042–8.394) | 0.042 |
Systolic blood pressure > 140 mmHg | 23.596 (7.526–73.978) | 0.0001 |
* Thresholds for abdominal obesity in the Asian population according to the IDF (20) |
TyG index was significant in predicting the presence of CS with an area under the ROC curve of 0.678 (95% CI: 0.582–0.775, p = 0.002) (Fig. 3). The cut-point of 9.63 offered 75% sensitivity, 44% specificity, 34% positive predictive value and 56% negative predictive value. When adding TyG index to age and HbA1c in the prediction model, the AROC was higher (0.780) than considering age or HbA1c alone (0.622 and 0.748, respectively) (p < 0.001 for both comparisons).
A higher number of diseased vessels (p = 0.04) and the degree of CS (p < 0.005) were associated with a higher level of TyG index (Tables 4 and 5).
Table 4
TyG index according to the number of coronary vessels with stenosis ≥ 50%
Number of vessels with stenosis | Number of patients | TyG index | 95% CI | p value |
0 | 106 | 9.56 ± 0.62 | 9.44–9.68 | 0.04 |
1 | 39 | 9.72 ± 0.71 | 9.49–9.95 |
2 or 3 | 21 | 9.92 ± 0.44 | 9.72–10.12 |
Table 5
TyG index according to the degree of coronary artery stenosis
Degree of stenosis | Number of patients | TyG index | 95% CI | p value |
< 50% | 106 | 9.56 ± 0.61 | 9.45–9.68 | < 0.005 |
50–69% | 28 | 9.58 ± 0.71 | 9.30–9.85 |
≥ 70% | 32 | 9.97 ± 0.49 | 9.79–10.15 |
Subgroup analyses showed that TyG index was associated with CS in the patients aged ≥ 60 years, with HbA1c ≥ 7%, on statin or antiplatelet therapy (Fig. 4).