Population selection
Of the 211,833 participants, 5,746 were excluded for lack of baseline TG values, while 4,789 were excluded because of extreme TG or FPG values, leaving 201,298 subjects for final data analysis.
Baseline parameters of study population
A total of 201,298 subjects (54.3% male and 45.7% female) were involved in this study. Average age and BMI of the population were 42.08 ± 12.67 years and 23.19 ± 3.32 kg/m2, respectively. After an average follow-up of 3.12 years (SD, 0.94), 3,389 participants were reported to have diabetes. The average TyG index was 8.35 ± 0.57, and baseline TyG index in diabetic patients was obviously higher than subjects without diabetes (8.90 ± 0.52 vs. 8.34 ± 0.57; p<0.001). Table 1 displayed baseline parameters of the population by TyG index quartiles (<7.93, 7.93–8.31, 8.31–8.73, ≥8.73). Except for HDL-C, which was not statistically different among the TyG quartiles, participants with higher TyG index generally had higher age, BMI, LDL-C, TC, ALT, AST, Scr, SBP, DBP, family history of diabetes, higher rates of smokers and drinkers.
Univariate analysis
The univariate analyses of potential risk factors were presented in Table 2. The results revealed that these factors, including age, BMI, LDL-C, TC, TG, FPG, TyG index, ALT, AST, Scr, SBP, DBP, drinking, smoking and family history of diabetes, were positively related to future risk of diabetes, whereas HDL-C was not correlated with future risk. Besides, compared with males, females showed a lower risk of diabetes.
Table 1. Baseline parameters of population (N = 201,298)
TyG index
|
Q1 (<7.93)
|
Q2 (≥7.93 to <8.31)
|
Q3 (≥8.31 to <8.73)
|
Q4 (≥8.73)
|
P-value
|
Participants
|
49,413
|
50,766
|
49,852
|
51,267
|
|
Age (years, mean ± SD)
|
37.32 ± 9.93
|
40.48 ± 12.00
|
43.76 ± 13.26
|
46.61 ± 13.20
|
<0.001
|
Sex, n (%).
|
|
|
|
|
<0.001
|
Male
|
16,016 (32.41)
|
24,391 (48.05)
|
30,683 (61.55)
|
38,146 (74.41)
|
|
Female
|
33,397 (67.59)
|
26,375 (51.95)
|
19,169 (38.45)
|
13,121 (25.59)
|
|
BMI (kg/m2, mean ± SD)
|
21.26 ± 2.59
|
22.40 ± 2.95
|
23.71 ± 3.13
|
25.31 ± 3.11
|
<0.001
|
SBP (mmHg, mean ± SD)
|
112.37 ± 14.02
|
116.50 ± 15.29
|
120.82 ± 16.07
|
125.62 ± 16.64
|
<0.001
|
DBP (mmHg, mean ± SD)
|
69.98 ± 9.53
|
72.45 ± 10.09
|
75.10 ± 10.51
|
78.47 ± 10.92
|
<0.001
|
FPG (mg/dL, mean ± SD)
|
83.49 ± 9.29
|
86.88 ± 9.40
|
89.61 ± 9.75
|
93.58 ± 10.48
|
<0.001
|
TC (mg/dL, mean ± SD)
|
165.41 ± 28.64
|
175.71 ± 30.75
|
185.95 ± 32.89
|
198.45 ± 34.94
|
<0.001
|
TG (mg/dL, mean ± SD)
|
51.00 ± 11.26
|
78.87 ± 12.03
|
113.51 ± 18.14
|
199.82 ± 58.97
|
<0.001
|
LDL-C (mg/dL, mean ± SD)
|
79.28 ± 22.93
|
85.06 ± 27.52
|
90.81 ± 31.54
|
95.18 ± 34.56
|
<0.001
|
HDL-C (mg/dL, mean ± SD)
|
53.17 ± 11.91
|
53.09 ± 11.82
|
53.08 ± 11.90
|
53.07 ± 11.79
|
0.769
|
ALT (IU/L, median (Q1-Q3)
|
13.90 (10.90-18.60)
|
16.00 (12.00-22.80)
|
19.05 (14.00-28.00)
|
25.40 (18.00-38.30)
|
<0.001
|
AST (IU/L, mean ± SD)
|
21.74 ± 6.05
|
22.25 ± 6.41
|
22.95 ± 6.75
|
24.25 ± 8.22
|
<0.001
|
Scr (mg/dL, mean ± SD)
|
0.73 ± 0.15
|
0.77 ± 0.16
|
0.81 ± 0.16
|
0.84 ± 0.16
|
<0.001
|
Smoker
|
|
|
|
|
<0.001
|
Now
|
1011 (2.05)
|
2096 (4.13)
|
3140 (6.30)
|
4824 (9.41)
|
|
Once
|
304 (0.62)
|
536 (1.06)
|
732 (1.47)
|
847 (1.65)
|
|
Never
|
10,591 (21.43)
|
11,372 (22.40)
|
10,978 (22.02)
|
10,556 (20.59)
|
|
Not recorded
|
37,507 (75.90)
|
36,762 (72.41)
|
35,002 (70.21)
|
35,040 (68.35)
|
|
Drinker
|
|
|
|
|
<0.001
|
Now
|
108 (0.22)
|
225 (0.44)
|
329 (0.66)
|
581 (1.13)
|
|
Once
|
1180 (2.39)
|
1831 (3.61)
|
2357 (4.73)
|
3128 (6.10)
|
|
Never
|
10,618 (21.48)
|
11,948 (23.54)
|
12,164 (24.40)
|
12,518 (24.42)
|
|
Not recorded
|
37,507 (75.91)
|
36,762 (72.41)
|
35,002 (70.21)
|
35,040 (68.35)
|
|
Family history of diabetes, n (%)
|
|
|
|
|
<0.001
|
No
|
48,526 (98.20)
|
49,708 (97.92)
|
48,804 (97.90)
|
50,158 (97.84)
|
|
Yes
|
887 (1.80)
|
1058 (2.08)
|
1048 (2.10)
|
1109 (2.16)
|
|
Values are presented as mean ± SD, median (Q1–Q3) or n (%).
BMI Body-mass index, SBP Systolic blood pressure, DBP Diastolic blood pressure, FPG Fasting plasma glucose, TC Total cholesterol, TG Triglyceride, LDL-CLow-density lipoprotein cholesterol, HDL-C High-density lipoprotein cholesterol, ALT Alanine aminotransferase, AST Aspartate transaminase, Scr Serum creatinine
Table 2. The results of univariate analysis
|
Statistics
|
HR (95% CI)
|
P value
|
Age (y)
|
42.08 ± 12.67
|
1.07 (1.06-1.07)
|
<0.0001
|
Gender
|
|
|
|
Male
|
109,236 (54.27%)
|
Ref
|
|
Female
|
92,062 (45.73%)
|
0.51 (0.47-0.55)
|
<0.0001
|
BMI (kg/m2)
|
23.19 ± 3.32
|
1.24 (1.23-1.25)
|
<0.0001
|
SBP (mmHg)
|
118.88 ± 16.31
|
1.04 (1.04-1.04)
|
<0.0001
|
DBP (mmHg)
|
74.03 ± 10.76
|
1.05 (1.04-1.05)
|
<0.0001
|
LDL-C (mg/dL)
|
87.64 ± 30.11
|
1.01 (1.00-1.02)
|
<0.0001
|
HDL-C (mg/dL)
|
53.10 ± 11.85
|
1.00 (1.00-1.01)
|
0.4881
|
TC (mg/dL)
|
181.51 ± 34.19
|
1.01 (1.01-1.01)
|
<0.0001
|
TG (mg/dL)
|
111.41 ± 64.72
|
1.01 (1.01-1.01)
|
<0.0001
|
FPG (mg/dL)
|
88.43 ± 10.43
|
1.15 (1.14-1.15)
|
<0.0001
|
TyG index
|
8.35 ± 0.57
|
5.78 (5.44-6.14)
|
<0.0001
|
Scr (mg/dL)
|
0.79 ± 0.17
|
2.87 (2.34, 3.52)
|
<0.0001
|
ALT (U/L)
|
23.52 ± 19.95
|
1.01 (1.01-1.01)
|
<0.0001
|
AST (U/L)
|
22.81±6.98
|
1.01 (1.01-1.02)
|
<0.0001
|
Smoker
|
|
|
|
Now
|
11,071 (5.50%)
|
Ref
|
|
Once
|
2,419 (1.20%)
|
0.75 (0.56-1.00)
|
0.0506
|
Never
|
43,497 (21.61%)
|
0.46 (0.40-0.53)
|
<0.0001
|
Not recorded
|
144,311 (71.69%)
|
0.63 (0.56-0.71)
|
<0.0001
|
Drinker
|
|
|
|
Now
|
1,243 (0.62%)
|
Ref
|
|
Once
|
8,496 (4.22%)
|
0.48 (0.33-0.70)
|
0.0001
|
Never
|
47,248 (23.47%)
|
0.50 (0.35-0.70)
|
<0.0001
|
Not recorded
|
144,311 (71.69%)
|
0.54 (0.39-0.76)
|
0.0003
|
Family history of diabetes
|
|
|
|
No
|
197,196 (97.96%)
|
Ref
|
|
Yes
|
4,102 (2.04%)
|
1.74 (1.47-2.06)
|
<0.0001
|
Data were expressed as mean ± SD or n (%).
BMI Body-mass index, SBP Systolic blood pressure, DBP Diastolic blood pressure, TC Total cholesterol, TG Triglyceride, LDL-C Low-density lipoprotein cholesterol, HDL-C High-density lipoprotein cholesterol, FPG Fasting plasma glucose, TyG index triglyceride-glucose index, Scr Serum creatinine, ALT Alanine aminotransferase, AST Aspartate transaminase
As shown in Figure 1, Kaplan-Meier curve revealed that the cumulative risk of incident diabetes was markedly different among the TyG index quartiles (log-rank test, P<0.001) and increased gradually with increase of TyG index, resulting in maximum risk of diabetes in the highest quartile.
Relationship between TyG index and future diabetes risk
As shown in Table 3, the effect of TyG index on risk of diabetes was assessed by cox proportional hazard model. In crude model, TyG index demonstrated a strongly positive association with future diabetes risk (HR, 5.78; 95% CI, 5.44–6.14). In model Ⅰ (adjusted for sex, age and BMI), the positive correlation became relatively weaker (HR, 3.31; 95% CI, 3.09–3.55). In model Ⅱ (further adjusted for TC, LDL-C, AST, ALT, Scr, DBP, SBP, drinking, smoking and family history of diabetes), the correlation did not change significantly compared with the minor adjustment model (HR, 3.34; 95% CI, 3.11–3.60).
Table 3. Relationship between TyG index and risk of diabetes
Outcomes
|
Crude model
|
|
Model Ⅰ
|
|
Model Ⅱ
|
HR (95% CI)
|
P
|
|
HR (95% CI)
|
P
|
HR (95% CI)
|
P
|
TyG index
|
5.78 (5.44-6.14)
|
<0.0001
|
|
3.31 (3.09-3.55)
|
<0.0001
|
|
3.34 (3.11-3.60)
|
<0.0001
|
TyG (quartile)
|
|
|
|
|
|
|
|
|
Q 1
|
Ref
|
|
|
Ref
|
|
|
Ref
|
|
Q 2
|
2.88 (2.34-3.53)
|
<0.0001
|
|
1.84 (1.50-2.27)
|
<0.0001
|
|
1.83 (1.49-2.26)
|
0.0293
|
Q 3
|
7.45 (6.17-9.00)
|
<0.0001
|
|
3.29 (2.70-3.99)
|
<0.0001
|
|
3.29 (2.70-4.01)
|
0.0004
|
Q 4
|
19.94 (16.64-23.88)
|
<0.0001
|
|
6.36 (5.26-7.70)
|
<0.0001
|
|
6.26 (5.15-7.60)
|
<0.0001
|
P for trend
|
<0.0001
|
|
|
<0.0001
|
|
|
<0.0001
|
|
Crude model: adjusted for none.
Model Ⅰ: adjusted for age, sex and BMI
Model Ⅱ: adjusted for age, sex, BMI, LDL-C, TC, Scr, AST, ALT, SBP, DBP, drinking, smoking and family history of diabetes
To ensure the robustness of the results, TyG index was processed a categorical variable (quartiles) for sensitivity analysis. There was a graded and positive correlation of TyG index with future risk of diabetes. Compared with the lowest quartile (Q1), increasing TyG index (Q2, Q3, and Q4) was related to increased HR estimates of incident diabetes [HR (95% CI), 1.83 (1.49–2.26); 3.29 ( 2.70–4.01) and 6.26 (5.15–7.60), respectively].
Nonlinear relationship
As shown in Figure 2, after adjusting for sex, age, BMI, LDL-C, TC, ALT, AST, Scr, SBP, DBP, smoking, drinking, and family history of diabetes, a significant nonlinear relationship was found between TyG index and risk of future diabetes (P<0.001), and the slope of the curve showed an upward tendency with the increase of TyG index.
Subgroup analysis
To further investigate the impact of other risk factors on the correlation of TyG index with future diabetes risk, subgroup analyses were carried on according to the following stratification variables: sex, age, BMI, DBP, SBP, smoking, drinking and family history of diabetes. The results of subgroup analyses and interactions were summarized in Table 4. The additive interactions between TyG index and diabetes risk were observed in sex, age, BMI, and SBP (P-value for interaction < 0.05). Stronger correlations were found in participants with age < 40 years, BMI ≥ 18.5 kg/m2 and < 24 kg/m2, or SBP < 140 mmHg, or in females. However, significant interactions were not found in DBP, smoking, drinking, or family history of diabetes.
Table 4. Effect of magnitude of TyG index on diabetes risk stratified by subgroups
Characteristics
|
No. of participants
|
HR (95%CI)
|
P-value
|
P for interaction
|
Age (year)
|
|
|
|
<0.0001
|
<40
|
106,447
|
4.53 (3.76-5.45)
|
<0.0001
|
|
>=40, <60
|
71,176
|
3.54 (3.19-3.93)
|
<0.0001
|
|
>=60
|
23,675
|
2.67 (2.37-3.00)
|
<0.0001
|
|
Sex
|
|
|
|
0.0150
|
Male
|
109,236
|
3.16 (2.90-3.45)
|
<0.0001
|
|
Female
|
92,062
|
3.84 (3.37-4.37)
|
<0.0001
|
|
BMI(kg/m2)
|
|
|
|
0.0075
|
<18.5
|
11,593
|
3.64 (1.53-8.64)
|
0.0034
|
|
>=18.5, <24
|
112,241
|
4.13 (3.62-4.71)
|
<0.0001
|
|
>=24, <28
|
60,886
|
3.22 (2.90-3.58)
|
<0.0001
|
|
>=28, <32
|
14,388
|
3.11 (2.64, 3.68)
|
<0.0001
|
|
>=32
|
2,190
|
3.36 (2.33, 4.85)
|
<0.0001
|
|
SBP(mmHg)
|
|
|
|
<0.0001
|
<140
|
181,383
|
3.48 (3.19-3.79)
|
<0.0001
|
|
>=140
|
19,915
|
2.89 (2.53-3.29)
|
<0.0001
|
|
DBP(mmHg)
|
|
|
|
0.9984
|
<90
|
185,636
|
3.29 (3.04-3.56)
|
<0.0001
|
|
>=90
|
15,661
|
3.43 (2.90-4.05)
|
<0.0001
|
|
Smoker
|
|
|
|
0.6979
|
Now
|
11,071
|
3.03 (2.36-3.90)
|
<0.0001
|
|
Once
|
2,419
|
4.27 (2.30-7.91)
|
<0.0001
|
|
Never
|
43,497
|
3.40 (2.84-4.08)
|
<0.0001
|
|
Not recorded
|
144,311
|
3.39 (3.13-3.69)
|
<0.0001
|
|
Drinker
|
|
|
|
0.2174
|
Now
|
1,243
|
5.31 (2.34-12.05)
|
<0.0001
|
|
Once
|
8,496
|
3.65 (2.52-5.28)
|
<0.0001
|
|
Never
|
47,248
|
3.33 (2.84-3.90)
|
<0.0001
|
|
Not recorded
|
144,311
|
3.39 (3.13-3.69)
|
<0.0001
|
|
Family history of diabetes
|
|
|
|
0.1175
|
No
|
197,196
|
3.39 (3.15-3.65)
|
<0.0001
|
|
Yes
|
4,102
|
3.07 (2.10-4.50)
|
<0.0001
|
|
Note 1: the model was adjusted for sex, age, BMI, LDL-C, TC, Scr, ALT, AST, SBP, DBP, drinking, smoking and family history of diabetes.
Note 2: the model was adjusted for all above variables except the corresponding stratification variable.