Baseline Characteristics of participants by GGT quintiles
In this study, a total of 15,464 people enrolled in the study. Male 7034 (45.49%), female 8430 (54.51%), the average age of the population was 43.71 years, the average GGT was 20.31 IU/L, and the average follow-up time was 6.05y. The study population was divided into 5 groups according to the GGT quintiles. As can be seen from Table 1, age, BMI, WC, ALT, AST, body weight, HDL-c, TG, HbA1c, FPG, DBP and SBP were positively correlated with GGT, and the P-value of the trend test was significant. The proportion of fatty liver in GGTQ4-5 was much higher than GGTQ1-3, and this difference was statistically significant. There was also a positive correlation between alcohol intaking and GGT. For the factor of smoking, the proportion of people with high GGT was higher than that of people with low GGT in the current and past smokers. Among non-smokers, there were no differences among the different GGT groups. During the follow-up period, the proportion of patients diagnosed with diabetes was higher in GGTQ5 than in GGTQ1-4. The cumulative incidence of diabetes stratified by GGT quintiles was shown in Fig. 1. It can be seen that the risk of diabetes mellitus in GGTQ5, GGTQ4 was much higher than GGTQ1, GGTQ2, GGTQ3. We classified GGT into GGTQ4-5 ( > = 16 IU/L) and GGTQ1-3 (< 16 IU/), with GGT16 IU/L as the boundary.
Table 1
Baseline Characteristics of participants by γ-glutamyltranspeptidase quintiles (N = 15464)
Characteristics | Q1 | Q2 | Q3 | Q4 | Q5 | P value |
< 11.00 IU/L | 11.00 to < 12.00 IU/L | 12.00 to < 16.00 IU/L | 16.00 to < 24.00 IU/L | 24.00t o < 399.00 IU/L |
No. of participants | 2885 | 2328 | 3674 | 3304 | 3273 | |
Age (years) | 42.18 ± 8.49 | 42.60 ± 8.65 | 43.56 ± 9.18 | 44.36 ± 9.00 | 45.35 ± 8.66 | < 0.001 |
BMI (kg/m^2) | 20.67 ± 2.40 | 21.07 ± 2.66 | 21.66 ± 2.88 | 22.83 ± 3.12 | 23.93 ± 3.21 | < 0.001 |
WC (cm) | 70.54 ± 7.27 | 72.79 ± 7.53 | 75.37 ± 8.08 | 79.37 ± 8.35 | 82.63 ± 8.53 | < 0.001 |
ALT (IU/L) | 12.86 ± 5.23 | 14.48 ± 5.38 | 16.94 ± 6.61 | 21.67 ± 9.97 | 31.91 ± 23.59 | < 0.001 |
AST (IU/L) | 14.96 ± 4.23 | 16.10 ± 4.57 | 17.32 ± 5.16 | 19.12 ± 6.66 | 23.56 ± 14.32 | < 0.001 |
Body Weight (kg) | 53.36 ± 8.15 | 55.57 ± 9.29 | 59.18 ± 10.26 | 64.64 ± 11.17 | 68.25 ± 11.51 | < 0.001 |
HDL (mmol/L) | 1.59 ± 0.37 | 1.56 ± 0.39 | 1.49 ± 0.41 | 1.37 ± 0.40 | 1.33 ± 0.39 | < 0.001 |
TC (mmol/L) | 4.90 ± 0.83 | 4.96 ± 0.82 | 5.06 ± 0.83 | 5.21 ± 0.87 | 5.43 ± 0.86 | < 0.001 |
TG (mmol/L) | 0.61 ± 0.31 | 0.67 ± 0.41 | 0.80 ± 0.48 | 1.04 ± 0.67 | 1.34 ± 0.88 | < 0.001 |
HbA1c (mmol/mol) | 32.33 ± 3.49 | 32.80 ± 3.36 | 33.07 ± 3.39 | 33.32 ± 3.51 | 33.46 ± 3.70 | < 0.001 |
FPG (mmol/L) | 4.95 ± 0.38 | 5.02 ± 0.39 | 5.14 ± 0.39 | 5.27 ± 0.38 | 5.37 ± 0.38 | < 0.001 |
DBP (mmHg) | 66.29 ± 8.90 | 67.97 ± 9.32 | 70.96 ± 9.93 | 73.90 ± 9.98 | 77.18 ± 10.36 | < 0.001 |
SBP (mmHg) | 107.11 ± 12.95 | 109.68 ± 13.44 | 113.76 ± 14.31 | 117.66 ± 14.08 | 122.08 ± 14.84 | < 0.001 |
Male | 2467 (85.51%) | 1642 (70.53%) | 1751 (47.66%) | 743 (22.49%) | 431 (13.17%) | < 0.001 |
Fatty liver | 86 (2.98%) | 130 (5.58%) | 436 (11.87%) | 828 (25.06%) | 1261 (38.53%) | < 0.001 |
Exercise | 486 (16.85%) | 382 (16.41%) | 699 (19.03%) | 620 (18.77%) | 522 (15.95%) | < 0.001 |
Alcohol consumption | | | | | | < 0.001 |
None | 2699 (93.55%) | 2062 (88.57%) | 2999 (81.63%) | 2292 (69.37%) | 1753 (53.56%) | |
Light | 130 (4.51%) | 183 (7.86%) | 385 (10.48%) | 503 (15.22%) | 557 (17.02%) | |
Moderate | 52 (1.80%) | 73 (3.14%) | 238 (6.48%) | 381 (11.53%) | 616 (18.82%) | |
Heavy | 4 (0.14%) | 10 (0.43%) | 52 (1.42%) | 128 (3.87%) | 347 (10.60%) | |
Smoking status | | | | | | < 0.001 |
Never | 2340 (81.11%) | 1707 (73.32%) | 2246 (61.13%) | 1475 (44.64%) | 1263 (38.59%) | |
Past | 277 (9.60%) | 278 (11.94%) | 632 (17.20%) | 840 (25.42%) | 925 (28.26%) | |
Current | 268 (9.29%) | 343 (14.73%) | 796 (21.67%) | 989 (29.93%) | 1085 (33.15%) | |
Year of follow up (years) | 7.17(0.45–12.90) | 5.15(0.50-12.96) | 5.03(0.52–12.94) | 5.06(0.53–12.90) | 5.50(0.50-12.96) | < 0.001 |
Diabetes was diagnosed during follow-up (%) | 20 (0.7%) | 26 (1.1%) | 47 (1.3%) | 105 (3.2%) | 175 (5.3%) | < 0.001 |
The results of the relationship between GGT and diabetes mellitus
The COX regression analysis model was used to estimate the correlation between GGT and diabetes mellitus. As can be seen from Table 2, in the unadjusted model, there was a positive correlation between GGT and the incidence of diabetes. In Model I, we adjusted gender and age, and we found that there was still a positive correlation between GGT and the incidence of diabetes. In Model II, after adjusting sex, age, fatty liver, BMI, ALT, AST, WC, body weight, exercise, TC, TG, HbA1c, alcohol consumption, smoking status, FPG, DBP and SBP, the risk of diabetes increased by 6% for every SD raised by GGT (per SD increase, HR: 1.06, 95%CI (1.02–1.16), P = 0.042). For sensitivity analysis, we converted GGT to a categorical variable (quintile). In the adjusted II model, compared with GGTQ1, the HR for diabetes in the GGTQ5 group was 1.49 (95%:1.13–2.66, P = 0.047) and GGTQ4 was 1.46 (95%CI:1.07–2.56, P = 0.044). For further analysis, with GGT16 IU/L as the limit, GGTQ4-5 was classified into one group, and GGTQ1-3 was classified into another group. It was noted that compared to GGTQ1-3 in a fully adjusted model, the HR for Q4-5 diabetes progression was 1. 40 (95%, 1.05–1.86, P = 0.021).
Table 2
Effect modification of GGT on incident of diabetes
Exposure | Non-adjusted | Adjust I | Adjust II |
| HR (95%CI) | P value | HR (95%CI) | P value | HR (95%CI) | P value |
GGT (IU/L) | 1.01 (1.01, 1.01) | < 0.001 | 1.01 (1.01, 1.01) | < 0.001 | 1.00 (1.00, 1.01) | 0.042 |
GGT (IU/L) per SD | 1.23 (1.18, 1.27) | < 0.001 | 1.17 (1.12, 1.23) | < 0.001 | 1.06 (1.02, 1.16) | 0.042 |
Categories | | | | | | |
GGT Q1 | 1.0 | | 1.0 | | 1.0 | |
GGT Q2 | 2.02 (1.13, 3.61) | 0.018 | 1.92 (1.07, 3.46) | 0.029 | 1.34 (0.74, 2.41) | 0.334 |
GGT Q3 | 2.49 (1.47, 4.20) | < 0.001 | 2.22 (1.30, 3.79) | < 0.001 | 0.98 (0.56, 1.71) | 0.949 |
GGT Q4 | 5.95 (3.68, 9.60) | < 0.001 | 5.00 (3.00, 8.34) | < 0.001 | 1.46 (1.07, 2.56) | 0.044 |
GGT Q5 | 9.27(5.83,14.72) | < 0.001 | 7.51 (4.54, 12.41) | < 0.001 | 1.49 (1.13, 2.66) | 0.047 |
Categories | | | | | | |
GGT Q1-Q3 | 1.0 | | 1.0 | | 1.0 | |
GGT Q4-Q5 | 4.27 (3.38, 5.40) | < 0.001 | 3.50 (2.69, 4.56) | < 0.001 | 1.40 (1.05, 1.86) | 0.021 |
Adjust I model adjust for sex, age. |
Adjust II model adjust for sex, age, fatty liver, BMI, ALT, AST, WC, body weight, exercise, TC, TG, HbA1c, alcohol consumption, smoking status, FPG, DBP, SBP. CI, confidence interval; SD, standard deviation |
The results of subgroup analyses
After stratification, we analyzed based on the main covariates known to affect diabetes, including sex, age, BMI, smoking status, alcohol consumption, WC, exercise, fatty liver. As shown in Fig. 2, GGTQ4-5 had an increased risk of developing diabetes in all subgroups compared to GGTQ1-3. For further sensitivity analysis, we stratified by age and gender. It was subdivided into 5 groups based on age and 2 groups based on gender. As can show in Table 3, in the adjusted II model, the HR of GGTQ4-5 was 1.31 (95%CI:1.01–1.71) or 1.13 (95%CI:1.02–1.38) compared to GGTQ1-3 in the 40-50y male or female. In other age groups, the risk of diabetes was not statistically significant. Because the confidence interval for β values spanned 1.
Table 3
Effect modification of GGT (Q4-5 vs. Q1–3) on incident of diabetes, stratified by age and gender.
Age, y | Sex | Unadjusted | Adjusted |
Q1-Q3 (%) | Q4-Q5 (%) | HR (95%CI) | P value | HR (95%CI) | P value |
20to < 30 | | | | | | | |
| Male | 0/251(0%) | 0/17(0%) | 1.00 (1.00, 1.00) | NS | 1.00 (1.00, 1.00) | NS |
| Female | 1/69(1.45%) | 0/74(0%) | 0.00 (0.00, Inf) | NS | 0.00 (0.00, Inf) | NS |
30to < 40 | | | | | | | |
| Male | 15/2039 (0.74%) | 2/262 (0.76%) | 1.10 (0.67, 1.80) | NS | 0.90 (0.50, 1.62) | NS |
| Female | 6/1154 (0.52%) | 47/1720 (2.73%) | 1.76 (1.33, 2.34) | < 0.001 | 1.05 (0.76, 1.45) | NS |
40to < 50 | | | | | | | |
| Male | 19/2324 (0.82%) | 15/409 (3.67%) | 1.77 (1.41, 2.22) | < 0.001 | 1.31 (1.01, 1.71) | 0.046 |
| Female | 15/993 (1.51%) | 114/2060 (5.53%) | 1.59 (1.33, 1.90) | < 0.001 | 1.13 (1.02, 1.38) | 0.037 |
50to < 60 | | | | | | | |
| Male | 13/1078 (1.21%) | 15/413 (3.63%) | 1.57 (1.23, 2.02) | < 0.001 | 1.17 (0.86, 1.61) | NS |
| Female | 13/643 (2.02%) | 65/1241(5.24%) | 1.41 (1.16, 1.72) | < 0.001 | 1.16 (0.93, 1.45) | NS |
>= 60 | | | | | | | |
| Male | 3/163 (1.84%) | 5/73 (6.85%) | 1.49 (0.93, 2.41) | 0.099 | 1.67 (0.80, 3.49) | NS |
| Female | 8/168 (4.76%) | 17/308 (5.52%) | 1.08 (0.82, 1.43) | 0.588 | 1.11 (0.77, 1.60) | NS |
Adjust for fatty liver, BMI, ALT, AST, WC, body weight, exercise, TC, TG, HbA1c, alcohol consumption, smoking status, FPG, DBP, SBP. |
The results of propensity score matching
We matched 686 non-diabetic patients and 343 patients diagnosed with diabetes. Their baseline characteristics are presented in Table 4. From the results, we know that diabetes and non-diabetes groups do not have sex, age, fatty liver, BMI, ALT, AST, WC, body weight, exercise, TC, TG, HbA1c, alcohol consumption, smoking status, FPG, DBP, SBP significance. Diabetes patients have higher GGT than non-diabetics. (30.55 ± 26.47 IU/L VS 28.20 ± 22.62 IU/L). The results before matching are presented in Table S1, the diabetic group and the non-diabetic group. All factors had significance.
Table 4
Baseline characteristics of participants stratified by outcome after PSM
| No Diabetes (n = 686) | Diabetes(n = 343) | P value |
Age (years) | 47.64 ± 9.11 | 47.10 ± 8.62 | NS |
BMI (kg/m^2) | 24.73 ± 3.37 | 24.83 ± 3.77 | NS |
WC (cm) | 84.28 ± 9.06 | 84.52 ± 10.08 | NS |
ALT (IU/L) | 27.41 ± 17.52 | 30.66 ± 19.46 | NS |
AST (IU/L) | 21.72 ± 10.35 | 22.18 ± 9.76 | NS |
Body Weight (kg) | 68.52 ± 12.24 | 69.16 ± 13.22 | NS |
HDL-c (mmol/L) | 1.27 ± 0.38 | 1.20 ± 0.34 | 0.038 |
TC (mmol/L) | 5.44 ± 0.89 | 5.43 ± 0.91 | NS |
TG (mmol/L) | 1.36 ± 0.85 | 1.50 ± 1.00 | NS |
HbA1c (mmol/mol) | 36.75 ± 3.32 | 36.49 ± 3.79 | NS |
FPG (mmol/L) | 5.58 ± 0.34 | 5.59 ± 0.36 | NS |
DBP (mmHg) | 77.07 ± 10.60 | 77.21 ± 10.12 | NS |
SBP (mmHg) | 122.12 ± 15.67 | 121.99 ± 15.56 | NS |
Male | 484 (70.6) | 262 (76.4) | NS |
Fatty liver | 381 (55.5) | 194 (56.6) | NS |
Exercise | 109 (15.9) | 49 (14.3) | NS |
Alcohol consumption | | | NS |
None | 473 (69) | 243 (70.8) | |
Light | 80 (11.7) | 37 (10.8) | |
Moderate | 87 (12.7) | 33 (9.6) | |
Heavy | 46 (6.7) | 30 (8.7) | |
Smoking status | | | NS |
Never | 283 (41.3) | 140 (40.8) | |
Past | 162 (23.6) | 69 (20.1) | |
Current | 241 (35.1) | 134 (39.1) | |
GGT (IU/L) | 28.20 ± 22.62 | 30.55 ± 26.47 | 0.031 |
NS, no significance |