Table 2.
Association of cardiovascular disease, coronary heart disease, and cerebrovascular disease events competing with non-event death across quartiles of triglyceride glucose.
| Incident Event | Model | TyG index | ||||
|---|---|---|---|---|---|---|
| Quartile 1 (<8.32) n = 1523 | Quartile 2 (8.32–8.61) n = 1521 | Quartile 3 (8.61–8.89) n = 1515 | Quartile 4 (≥8.90) n = 1519 | P for trend of linearity | ||
| CVD | CVD | 121 (7.94) | 128 (8.42) | 177 (11.68) | 279 (18.37) | <0.001 |
| Non-CVD death | 41 (2.69) | 47 (3.09) | 54 (3.56) | 41 (2.70) | NA | |
| Model 1 | Reference | 1.05 (0.82–1.35) | 1.49 (1.18–1.88) | 2.38 (1.92–2.96) | <0.001 | |
| Model 2 | Reference | 0.99 (0.77–1.28) | 1.33 (1.05–1.68) | 1.72 (1.37–2.16) | <0.001 | |
| Model 3 | Reference | 1.00 (0.80–1.25) | 1.17 (0.94–1.45) | 1.61 (1.31–1.99) | <0.001 | |
| CHD | CHD | 71 (4.66) | 86 (5.65) | 118 (7.79) | 225 (14.81) | <0.001 |
| Non-CHD death | 56 (3.68) | 54 (3.55) | 69 (4.55) | 52 (3.42) | NA | |
| Model 1 | Reference | 1.19 (0.87–1.64) | 1.68 (1.25–2.26) | 3.28 (2.51–4.30) | <0.001 | |
| Model 2 | Reference | 1.14 (0.83–1.57) | 1.50 (1.11–2.03) | 2.32 (1.75–3.08) | <0.001 | |
| Model 3 | Reference | 1.22 (0.93–1.60) | 1.26 (0.96–1.66) | 2.05 (1.58–2.64) | <0.001 | |
| Cerebrovascular disease | Cerebrovascular disease | 52 (3.41) | 45 (2.96) | 63 (4.16) | 74 (4.87) | 0.006 |
| Non-cerebrovascular disease death | 46 (3.02) | 50 (3.29) | 63 (4.16) | 50 (3.29) | NA | |
| Model 1 | Reference | 0.87 (0.58–1.29) | 1.24 (0.86–1.80) | 1.46 (1.02–2.09) | 0.008 | |
| Model 2 | Reference | 0.81 (0.54–1.21) | 1.12 (0.77–1.63) | 1.12 (0.78–1.62) | 0.308 | |
| Model 3 | Reference | 0.67 (0.45–0.99) | 1.12 (0.79–1.59) | 1.20 (0.84–1.72) | 0.183 | |
Values are n (%) or subhazard ratio (95% confidence interval) using the first (lowest) quartile as the reference. Model 1 was adjusted by age and sex; model 2 was model 1 plus living alone, current smoking, alcohol consumption, exercise, body mass index, resting heart rate, systolic blood pressure, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, and diabetic status; model 3 was model 2 with time-varying repeated measures of TyG. P values for linear trend across TyG quartiles were evaluated by a median value within each quartile as a continuous variable. TyG = triglyceride glucose; CVD = cardiovascular disease; CHD = coronary heart disease; NA = not applicable.