Characteristics of participants at baseline
Totally, 2568 normoglycemic participants (810 males and 1758 females) were recruited in 2011 (Fig. 1), with an average age of 62.68 ± 6.97 years. After 7 years of follow-up, 52 subjects had developed newly diagnosed diabetes, and 423 individuals had developed prediabetes. Among participants with prediabetes, 278 had IGT, 105 had IFG, and 40 had IGT + IFG. The incidences of prediabetes and diabetes were 115.3 per 1000 person-years and 14.2 per 1000 person-years, respectively. As can be seen in Table 1 and Table 2, the average age of subjects who developed diabetes (64.06 ± 5.23 years old) or prediabetes (64.06 ± 6.94 years old) was older than that of participants who did not develop diabetes (62.65 ± 7.04 years old and 62.36 ± 6.98 years old). As expected, the levels of systolic pressure, WHtR, weight, BMI, WC, WHR, TG, FBG, PBG and HbA1c among newly diagnosed diabetes patients were more likely to be higher than the levels among those who did not develop diabetes (p = 0.019, p = 0.001, p = 0.002, p = 0.001, p = 0.001, p = 0.004, p = 0.002, p < 0.001, p = 0.014 and p < 0.001). While the result of HDL-c was the opposite (p < 0.001). Moreover, the family history of diabetes was significantly different among different groups (p = 0.046). The rate of current employment was much lower while the rate of regular exercise and diabetes family history were much higher in the group of people who had developed prediabetes than in the other groups (Table 2). The levels of WHtR (p < 0.001), weight (p < 0.001), BMI (p < 0.001), WC (p < 0.001), HC (p = 0.016), WHR (p < 0.001), systolic pressure (p < 0.001), diastolic pressure (p = 0.001), TG ( < = 0.001), LDL-c (p = 0.008), FBG (p < 0.001), PBG (p < 0.001) and HbA1c (p < 0.001) among those with prediabetes were more likely to be higher than those among people who did not develop prediabetes. The result of HDL-c was also the opposite (p = 0.002).
Table 1
Baseline characteristics of study population grouped by diabetes and nondiabetes
| Total (n = 2568) | Diabetes (n = 52) | Nondiabets (n = 2516) | p-value |
Age(years old) | 62.68 ± 6.97 | 64.06 ± 5.23 | 62.65 ± 7.04 | p = 0.149 |
Gender | male | 810(31.54%) | 20(38.46%) | 790(31.40%) | p = 0.293 |
female | 1758(68.46%) | 32(61.54%) | 1726(68.60%) |
Marital status(n, %) | married | 2433(94.74%) | 48(92.31%) | 2385(94.79%) | p = 0.350 |
not married /widowed | 135(5.26%) | 4(7.69%) | 131(5.21%) |
Current employment(n, %) | yes | 591(23.01%) | 6(11.54%) | 585(23.25%) | p = 0.065 |
no | 1977(76.99%) | 46(88.46%) | 1931(76.75%) |
Education(n, %) | illiteracy | 17(0.66%) | 1(1.92%) | 16(0.64%) | p = 0.209 |
primary schlool | 72(2.80%) | 1(1.92%) | 72(2.86%) |
junior high school | 737(28.70%) | 20(38.46%) | 716(28.46%) |
senior high school | 1237(48.17%) | 19(36.54%) | 1218(48.41%) |
college | 505(19.67%) | 11(21.16%) | 494(19.63%) |
Current smoker(n, %) | yes | 471(18.34%) | 12(23.08%) | 459(18.24%) | p = 0.469 |
no | 2097(81.66%) | 40(76.92%) | 2057(81.76%) |
Current alcohol consumption(n, %) | yes | 803(31.27%) | 15(28.85%) | 788(31.32%) | p = 0.764 |
no | 1765(68.73%) | 37(71.15%) | 1728(68.68%) |
Current tea consumption(n, %) | yes | 1647(64.14%) | 38(73.08%) | 1609(63.95%) | p = 0.191 |
no | 921(35.86%) | 14(26.92%) | 907(36.05%) |
Regular exercise(n, %) | yes | 282(10.98%) | 5(9.62%) | 277(11.01%) | p = 0.829 |
no | 2286(89.02%) | 47(90.38%) | 2239(88.99%) |
sleeptime(h) | 7.5(7.0, 8.0) | 8.0(7.0, 8.38) | 7.5(7.0, 8.0) | p = 0.868 |
Diabetes family history | yes | 588(22.90%) | 18(34.62%) | 570(22.66%) | p = 0.046* |
no | 1980(77.10%) | 34(65.38%) | 1946(77.34%) |
SBP(mmHg) | 126.67(117.00, 137.00) | 130.17(122.50, 142.50) | 126.33(117.00, 137.67) | p = 0.019* |
DBP(mmHg) | 74.67(68.33, 80.75) | 78.00(70.08, 82.17) | 74.67(68.33, 80.67) | p = 0.075 |
Pulse(bp) | 76.33(70.00, 83.67) | 76.83(70.75, 84.00) | 76.33(70.00, 83.67) | p = 0.549 |
WHtR | 0.50(0.47, 0.53) | 0.52(0.49, 0.56) | 0.50(0.47, 0.53) | p = 0.001* |
Height(cm) | 161,00(156.00, 167.00) | 160.50(157.00, 167.30) | 161.00(156.00, 167.00) | p = 0.667 |
Weight(kg) | 64.00(58.00, 72.08) | 70.00(60.75, 77.75) | 64.00(58.00, 72.00) | p = 0.002* |
BMI | 24.77(22.78, 27.08) | 26.31(24.20, 28.55) | 24.76(22.74, 27.04) | p = 0.001* |
WC(cm) | 81.00(75.00, 86.00) | 85.00(79.25, 90.88) | 80.55(75.00, 86.00) | p = 0.001* |
HC(cm) | 92.00(89.00, 97.00) | 92.00(90.00, 100.00) | 92.00(89.00, 96.50) | p = 0.056 |
WHR | 0.87(0.83, 0.91)) | 0.89 ± 0.06 | 0.87(0.83, 0.91) | p = 0.004* |
GGT | 17.75(13.50, 25.00) | 22.80(18.20, 37.65) | 17.60(13.50, 24.80) | p < 0.001* |
Cr | 63.70(56.40, 74.00) | 62.00(56.20, 74.65) | 63.70(56.40, 73.98) | p = 0.968 |
TC(mmol/l) | 5.08(4.54, 5.70) | 5.21(4.79, 5.69) | 5.08(4.53, 5.70) | p = 0.447 |
TG(mmol/l) | 1.12(0.81, 1.58) | 1.41(1.04, 1.93) | 1.11(0.81, 1.57) | p = 0.002* |
HDL-c(mmol/l) | 1.46(1.25, 1.73) | 1.31(1.13, 1.55) | 1.47(1.25, 1.73) | p < 0.001* |
LDL-c(mmol/l) | 3.09(2.62, 3.58) | 3.31(3.00, 3.58) | 3.08(2.63, 3.58) | p = 0.063 |
FBG(mmol/l) | 5.22(4.96, 5.52) | 5.49(5.29, 5.74) | 5.21(4.96, 5.51) | p < 0.001* |
PBG(mmol/l) | 6.06(5.31, 6.76) | 6.33(5.89, 6.95) | 6.05(5.30, 6.76) | p = 0.014* |
HbA1C(%) | 5.70(5.40, 5.90) | 5.90(5.70, 6.10) | 5.70(5.40, 5.90) | p < 0.001* |
Data were mean ± SD or median (IQR) for skewed variables or numbers (proportions) for categorical variables |
* indicate significance of P value (P < 0.05) |
SBP: systolic blood pressure; DBP: diastolic blood pressure; GGT: gamma-glutamyl transferase; TC: total cholesterol; TG: triglycerides; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol; FBG: fasting blood glucose; PBG: postprandial blood glucose; HbA1c: hemoglobin A1c. |
Table 2
Baseline characteristics of study population grouped by prediabetes and normal
| Total (n = 2516) | Prediabetes (n = 423) | Normal (n = 2093) | p-value |
Age(years old) | 62.65 ± 7.00 | 64.06 ± 6.94 | 62.36 ± 6.98 | p < 0.001* |
Gender | male | 790(31.40%) | 146(34.52%) | 644(30.77%) | p = 0.135 |
female | 1726(68.60%) | 277(65.48%) | 1449(69.23%) |
Marital status(n, %) | married | 2385(94.79%) | 404(95.51%) | 1981(94.65%) | p = 0.478 |
not married /widowed | 131(5.21%) | 19(4.49%) | 112(5.35%) |
Current employment(n, %) | yes | 577(22.93%) | 66(15.60%) | 511(24.41%) | p < 0.001* |
no | 1939(77.07%) | 357(84.40%) | 1582(75.59%) |
Education(n, %) | illiteracy | 16(0.64%) | 3(0.71%) | 13(0.62%) | p = 0.002* |
primary schlool | 72(2.86%) | 11(2.60%) | 61(2.91%) |
junior high school | 716(28.46%) | 155(36.64%) | 561(26.81%) |
senior high school | 1218(48.41%) | 181(42.79%) | 1037(49.55%) |
college | 494(19.63%) | 73(17.26%) | 421(20.11%) |
Current smoker(n, %) | yes | 459(18.24%) | 84(19.86%) | 375(17.92%) | p = 0.369 |
no | 2057(81.76%) | 339(80.14%) | 1718(82.08%) |
Current alcohol consumption(n, %) | yes | 788(31.32%) | 148(35.00%) | 640(30.58%) | p = 0.075 |
no | 1728(68.68%) | 275(65.00%) | 1453(69.42%) |
Current tea consumption(n, %) | yes | 1609(63.95%) | 263(62.17%) | 1346(64.31%) | p = 0.406 |
no | 907(36.05%) | 160(37.83%) | 747(35.69%) |
Regular exercise(n, %) | yes | 277(11.01%) | 65(15.37%) | 212(10.13%) | p = 0.002* |
no | 2239(88.99%) | 358(84.63%) | 1881(89.87%) |
sleeptime(h) | 7.50(7.00, 8.00) | 7.50(7.00, 8.00) | 7.50(7.00, 8.00) | p = 0.345 |
Diabetes family history | yes | 570(22.66%) | 116(27.42%) | 454(21.69%) | p = 0.011* |
no | 1946(77.34%) | 307(72.58%) | 1639(78.31%) |
SBP(mmHg) | 126.33(117.00, 136.67) | 130.00(120.00, 140.00) | 125.67(116.33, 136.00) | p < 0.001* |
DBP(mmHg) | 74.67(68.33, 80.67) | 76.00(70.00, 81.33) | 74.33(68.00, 80.67) | p = 0.001* |
Pulse(bp) | 76.33(70.00, 83.67) | 76.67(69.00, 85.00) | 76.33(70.00, 83.67) | p = 0.573 |
WHtR | 0.50(0.47, 0.53) | 0.511(0.48, 0.54) | 0.50(0.46, 0.53) | p < 0.001* |
Height(cm) | 161.00(156.00, 167.00) | 160.50(155.80, 166.50) | 161.00(156.00, 167.00) | p = 0.598 |
Weight(kg) | 64.00(58.00, 72.00) | 66.00(60.00, 74.00) | 64.00(57.00, 71.50) | p < 0.001* |
BMI | 24.76(22.74, 27.04) | 25.83(23.57, 27.82) | 24.61(22.60, 26.82) | p < 0.001* |
WC(cm) | 80.55(75.00, 86.00) | 83.00(78.00, 89.00) | 80.00(74.00, 86.00) | p < 0.001* |
HC(cm) | 92.00(89.00, 96.50) | 93.00(89.00, 97.00) | 92.00(88.50, 96.00) | p = 0.016* |
WHR | 0.87(0.83, 0.91) | 0.89(0.85, 0.93) | 0.87(0.83, 0.91) | p < 0.001* |
GGT | 17.60(13.50, 24.80) | 19.40(14.40, 27.00) | 17.30(13.40, 24.25) | p < 0.001* |
Cr | 63.70(56.40, 73.98) | 65.00(56.40, 75.30) | 63.50(56.40, 73.70) | p = 0.135 |
TC(mmol/l) | 5.08(4.53, 5.70) | 5.15(4.64, 5.78) | 5.06(4.51, 5.69) | p = 0.058 |
TG(mmol/l) | 1.11(0.81, 1.57) | 1.21(0.86, 1.81) | 1.09(0.80, 1.52) | p < 0.001* |
HDL-c(mmol/l) | 1.47(1.25, 1.73) | 1.41(1.22, 1.68) | 1.48(1.26, 1.74) | p = 0.002* |
LDL-c(mmol/l) | 3.08(2.62, 3.58) | 3.16(2.69, 3.69) | 3.06(2.60, 3.54) | p = 0.008* |
FBG(mmol/l) | 5.21(4.96, 5.51) | 5.42(5.11, 5.74) | 5.18(4.94, 5.45) | p < 0.001* |
PBG(mmol/l) | 6.06(5.30, 6.76) | 6.44(5.57, 6.99) | 5.98(5.24, 6.70) | p < 0.001* |
HbA1C(%) | 5.70(5.40, 5.90 | 5.80(5.60, 6.00) | 5.70(5.40, 5.90) | p < 0.001* |
Data were mean ± SD or median (IQR) for skewed variables or numbers (proportions) for categorical variables |
* indicate significance of P value (P < 0.05) |
SBP: systolic blood pressure; DBP: diastolic blood pressure; GGT: gamma-glutamyl transferase; TC: total cholesterol; TG: triglycerides; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol; FBG: fasting blood glucose; PBG: postprandial blood glucose; HbA1c: hemoglobin A1c. |
Association of anthropometric values with incident diabetes risk
To further analyze the function of these anthropometrics, we used stratified analyses for different sexes. To better discuss the relationship, we divided the WHtR, BMI, WC and WHR into tertiles. The results are shown in Table 3. Although all of these anthropometric values are meaningful in the general population, the results change after stratification by gender. The results showed that these 4 indices were significantly different between diabetic and nondiabetic participants in the female group, but the difference disappeared in the male group. After adjusting for age in the general population, a difference also existed in Tertile 3, while no significant difference was observed in the multivariate-adjusted model. A difference in BMI existed in the age-adjusted model in both female and male (female: Tertile 3, OR: 2.88, 95% CI: 1.13–7.33, p = 0.026; male: Tertile 2, OR: 5.04, 95% CI: 1.09–23.25, p = 0.038), while this difference disappeared in the multivariate-adjusted model. No differences were found in other anthropometric values.
Table 3
Age-adjusted and multivariable association of different anthropometric indices with the risk of diabetes
| Diabetes %(n/total) | Nondiabetes %(n/total) | p-value | Age-adjusted | Multivariate-adjusted |
OR(95% CI) | p-value | OR(95% CI) | p-value |
Tota(n = 2568) | WHtR | Tertile 1 | 19.23%(10/52) | 33.87%(852/2516) | p = 0.019* | reference | | reference | |
Tertile 2 | 30.77%(16/52) | 33.30%(838/2516) | 1.59(0.72, 3.53) | p = 0.256 | 1.24(0.55, 2.82) | p = 0.6 |
Tertile 3 | 50%(26/52) | 32.83%(826/2516) | 2.53(1.20, 5.34) | p = 0.014* | 1.56(0.71, 3.40) | p = 0.27 |
BMI | Tertile 1 | 19.23%(10/52) | 33.74%(849/2516) | p = 0.01* | reference | | reference | |
Tertile 2 | 28.85%(15/52) | 33.30%(838/2516) | 1.50(0.67, 3.35) | p = 0.327 | 1.17(0.51, 2.67) | p = 0.72 |
Tertile 3 | 51.92%(27/52) | 32.96%(829/2516) | 2.71(1.30, 5.64) | p = 0.008* | 1.50(0.69, 3.28) | p = 0.31 |
WC | Tertile 1 | 17.31%(9/52) | 35.10%(883/2516) | p = 0.003* | reference | | reference | |
Tertile 2 | 30.77%(16/52) | 33.46%(842/2516) | 1.80(0.79, 4.12) | p = 0.162 | 1.44(0.62, 3.37) | p = 0.39 |
Tertile 3 | 51.92%(27/52) | 31.44%(791/2516) | 3.17(1.47, 6.86) | p = 0.003* | 1.96(0.86, 4.49) | p = 0.11 |
WHR | Tertile 1 | 17.31%(9/52) | 33.74%(849/2516) | p = 0.01* | reference | | reference | |
Tertile 2 | 32.69%(17/52) | 34.18%(860/2516) | 1.81(0.80, 4.09) | p = 0.156 | 1.23(0.53, 2.85) | p = 0.627 |
Tertile 3 | 50%(26/52) | 32.08%(807/2516) | 2.86(1.32, 6.23) | p = 0.008* | 1.53(0.67, 3.48) | p = 0.316 |
Female(n = 1758) | WHtR | Tertile 1 | 21.88%(7/32) | 33.66%(581/1726) | p = 0.023* | reference | | reference | |
Tertile 2 | 21.88%(7/32) | 33.49%(578/1726) | 0.96(0.34, 2.77) | p = 0.946 | 0.73(0.25, 2.15) | p = 0.569 |
Tertile 3 | 56.24%(18/32) | 32.85%(567/1726) | 2.26(0.91, 5.60) | p = 0.079 | 1.40(0.54, 3.68) | p = 0.49 |
BMI | Tertile 1 | 18.75%(6/32) | 33.55%(579/1726) | p = 0.021* | reference | | reference | |
Tertile 2 | 25%(8/32) | 33.49%(578/1726) | 1.32(0.46, 3.83) | p = 0.609 | 0.99(0.34, 2.94) | p = 0.989 |
Tertile 3 | 56.25%(18/32) | 32.96%(569/1726) | 2.88(1.13, 7.33) | p = 0.026* | 1.73(0.64, 4.64) | p = 0.28 |
WC | Tertile 1 | 21.88%(7/32) | 37.43%(646/1726) | p = 0.043* | reference | | reference | |
Tertile 2 | 28.12%(9/32) | 32.50%(561/1726) | 1.43(0.53, 3.86) | p = 0.484 | 1.12(0.41, 3.10) | p = 0.826 |
Tertile 3 | 50%(16/32) | 30.07%(519/1726) | 2.48(1.00, 6.18) | p = 0.051 | 1.67(0.64, 4.36) | p = 0.3 |
WHR | Tertile 1 | 25%(8/32) | 33.72%(582/1726) | p = 0.001* | reference | | reference | |
Tertile 2 | 12.5%(4/32) | 33.55%(579/1726) | 0.47(0.14, 1.58) | p = 0.223 | 0.31(0.09, 1.06) | p = 0.063 |
Tertile 3 | 62.5%(20/32) | 32.73%(565/1726) | 2.24(0.96, 5.24) | p = 0.063 | 1.27(0.51, 3.13) | p = 0.606 |
Male(n = 810) | WHtR | Tertile 1 | 25%(5/20) | 33.42%(264/790) | p = 0.515 | reference | | reference | |
Tertile 2 | 30%(6/20) | 33.80%(267/790) | 1.18(0.36, 3.92) | p = 0.787 | 0.98(0.27, 3.55) | p = 0.979 |
Tertile 3 | 45%(9/20) | 32.78%(259/790) | 1.82(0.60, 5.52) | p = 0.287 | 1.18(0.35, 4.00) | p = 0.789 |
BMI | Tertile 1 | 10%(2/20) | 33.67%(266/790) | p = 0.061 | reference | | reference | |
Tertile 2 | 50%(10/20) | 32.91%(260/790) | 5.04(1.09, 23.25) | p = 0.038* | 6.25(1.03, 38.10) | p = 0.047* |
Tertile 3 | 40%(8/20) | 33.42%(264/790) | 3.94(0.83, 18.80) | p = 0.085 | 3.12(0.49, 19.82) | p = 0.228 |
WC | Tertile 1 | 35%(7/20) | 35.44%(280/790) | p = 0.611 | reference | | reference | |
Tertile 2 | 25%(5/20) | 34.18%(270/790) | 0.73(0.23, 2.33) | p = 0.592 | 0.58(0.17, 2.02) | p = 0.396 |
Tertile 3 | 40%(8/20) | 30.38%(240/790) | 1.30(0.47, 3.66) | p = 0.614 | 0.83(0.27, 2.58) | p = 0.746 |
WHR | Tertile 1 | 30%(6/20) | 33.16%(262/790) | p = 0.814 | reference | | reference | |
Tertile 2 | 40%(8/20) | 32.66%(258/790) | 1.36(0.46, 3.96) | p = 0.579 | 1.12(0.35, 3.58) | p = 0.85 |
Tertile 3 | 30%(6/20) | 34.18%(270/790) | 0.95(0.30, 2.99) | p = 0.928 | 0.57(0.17, 1.97) | p = 0.375 |
Data were mean ± SD or median (IQR) for skewed variables or numbers (proportions) for categorical variables |
* indicate significance of P value (P < 0.05) |
Multivariable Logistic models to adjust for the following confounding factors: age, SBP, diabetes family history, GGT, TG, HDL-c and HbA1C. |
Association of different anthropometric values with incident prediabetes risk
In order to better discuss the relationship between different anthropometric values and blood glucose status, we further divided people into prediabetic and normal blood glucose groups. The results are shown in Table 4. The age-adjusted model in the general population showed that the WHtR has the highest OR (Tertile 2 OR: 1.85, 95% CI 1.39–2.45, p < 0.001; Tertile 3 OR: 2.25, 95% CI: 1.68–2.94, p < 0.001) compared with other values, which means that people with a higher WHtR have a higher risk of prediabetes. Moreover, as the WHtR increased, the OR increased. This phenomenon was also true in the multivariate-adjusted model (Tertile 2 OR: 1.72, 95% CI 1.28–2.30, p < 0.001; Tertile 3 OR: 1.90, 95% CI: 1.41–2.56, p < 0.001). The association between the WHtR and prediabetes was the strongest among all four indices. To summarize, these four indices in the prediabetes group tended to follow the following pattern: WHtR < WC < WHR < BMI.
Table 4
Age-adjusted and multivariable association of different anthropometric indices with the risk of prediabetes
| Prediabetes %(n/total) | Normal %(n/total) | p-value | Age-adjusted | Multivariate-adjusted |
OR(95% CI) | p-value | OR(95% CI) | p-value |
Tota(n = 2516) | WHtR | Tertile 1 | 20.80%(88/423) | 36.12%(756/2093) | p < 0.001* | reference | | reference | |
Tertile 2 | 35.70%(151/423) | 32.54%(681/2093) | 1.85(1.39, 2.45) | p < 0.001* | 1.72(1.28, 2.30) | p < 0.001* |
Tertile 3 | 43.50%(184/423) | 31.34%(656/2093) | 2.25(1.68, 2.94) | p < 0.001* | 1.90(1.41, 2.56) | p < 0.001* |
BMI | Tertile 1 | 23.40%(99/423) | 35.64%(746/2093) | p < 0.001* | reference | | reference | |
Tertile 2 | 32.39%(137/423) | 33.16%(694/2093) | 1.46(1.11, 1.93) | p = 0.008* | 1.33(1.00, 1.77) | p = 0.053 |
Tertile 3 | 44.21%(187/423) | 31.20%(653/2093) | 2.11(1.62, 2.76) | p < 0.001* | 1.65(1.23, 2.21) | p < 0.001* |
WC | Tertile 1 | 24.11%(102/423) | 37.31%(781/2093) | p < 0.001* | reference | | reference | |
Tertile 2 | 35.70%(151/423) | 33.02%(691/2093) | 1.59(1.21, 2.09) | p = 0.001* | 1.53(1.15, 2.03) | p = 0.003* |
Tertile 3 | 40.19%(170/423) | 29.67%(621/2093) | 1.92(1.47, 2.52) | p < 0.001* | 1.69(1.25, 2.29) | p = 0.001* |
WHR | Tertile 1 | 22.93%(97/423) | 35.16%(736/2093) | p < 0.001* | reference | | reference | |
Tertile 2 | 35.93%(152/423) | 34.59%(724/2093) | 1.52(1.15, 2.00) | p = 0.003* | 1.35(1.01, 1.79) | p = 0.042* |
Tertile 3 | 41.13%(174/423) | 30.24%(633/2093) | 1.89(1.44, 2.50) | p < 0.001* | 1.53(1.13, 2.06) | p = 0.006* |
Female(n = 1726) | WHtR | Tertile 1 | 19.86%(55/277) | 36.03%(522/1449) | p < 0.001* | reference | | reference | |
Tertile 2 | 35.38%(98/277) | 32.78%(475/1449) | 1.90(1.34, 2.71) | p < 0.001* | 1.74(1.21, 2.51) | p = 0.003* |
Tertile 3 | 44.76%(124/277) | 31.19%(452/1449) | 2.34(1.65, 3.33) | p < 0.001* | 2.02(1.38, 2.95) | p < 0.001* |
BMI | Tertile 1 | 20.94%(58/277) | 35.96%(521/1449) | p < 0.001* | reference | | reference | |
Tertile 2 | 35.38%(98/277) | 32.78%(475/1449) | 1.84(1.30, 2.61) | p = 0.001* | 1.65(1.15, 2.37) | p = 0.006* |
Tertile 3 | 43.68%(121/277) | 31.26%(453/1449) | 2.31(1.64, 3.24) | p < 0.001* | 1.81(1.25, 2.61) | p = 0.002* |
WC | Tertile 1 | 25.63%(71/277) | 39.68%(575/1449) | p < 0.001* | reference | | reference | |
Tertile 2 | 35.02%(97/277) | 32.02%(464/1449) | 1.65(1.19, 2.30) | p = 0.003* | 1.53(1.09, 2.15) | p = 0.015* |
Tertile 3 | 39.35%(109/277) | 28.30%(410/1449) | 1.95(1.40, 2.72) | p < 0.001* | 1.69(1.18, 2.42) | p = 0.004* |
WHR | Tertile 1 | 23.83%(66/277) | 35.61%(516/1449) | p = 0.001* | reference | | reference | |
Tertile 2 | 37.55%(104/277) | 32.51%(471/1449) | 1.65(1.18, 2.30) | p = 0.004* | 1.46(1.03, 2.07) | p = 0.035* |
Tertile 3 | 38.63%(107/277) | 31.88%(462/1449) | 1.62(1.15, 2.27) | p = 0.005* | 1.26(0.88, 1.81) | p = 0.205 |
Male(n = 790) | WHtR | Tertile 1 | 23.97%(35/146) | 35.56%(229/644) | p = 0.021* | reference | | reference | |
Tertile 2 | 35.62%(52/146) | 32.76%(211/644) | 1.63(1.02, 2.60) | p = 0.042* | 1.55(0.95, 2.52) | p = 0.08 |
Tertile 3 | 40.41%(59/146) | 31.68%(204/644) | 1.90(1.20, 3.02) | p = 0.006* | 1.57(0.96, 2.59) | p = 0.074 |
BMI | Tertile 1 | 26.71%(39/146) | 34.94%(225/644) | p = 0.004* | reference | | reference | |
Tertile 2 | 28.08%(41/146) | 34.32%(221/644) | 1.11(0.69, 1.79) | p = 0.679 | 1.00(0.62, 1.64) | p = 0.991 |
Tertile 3 | 45.21%(66/146) | 30.74%(198/644) | 2.12(1.29, 3.14) | p = 0.002* | 1.55(0.96, 2.51) | p = 0.075 |
WC | Tertile 1 | 28.08%(41/146) | 37.10%(239/644) | p = 0.071 | reference | | reference | |
Tertile 2 | 34.93%(51/146) | 34.01%(219/644) | 1.40(0.89, 2.20) | p = 0.146 | 1.34(0.83, 2.15) | p = 0.227 |
Tertile 3 | 36.99%(54/146) | 28.89%(186/644) | 1.75(1.12, 2.75) | p = 0.015* | 1.43(0.88, 2.34) | p = 0.153 |
WHR | Tertile 1 | 19.17%(28/146) | 36.34%(234/644) | p < 0.001* | reference | | reference | |
Tertile 2 | 32.88%(48/146) | 32.61%(210/644) | 1.92(1.16, 3.17) | p = 0.011* | 1.89(1.13, 3.16) | p = 0.016* |
Tertile 3 | 47.95%(70/146) | 31.05%(200/644) | 3.03(1.88, 4.91) | p < 0.001* | 2.62(1.58, 4.33) | p < 0.001* |
Data were mean ± SD or median (IQR) for skewed variables or numbers (proportions) for categorical variables |
* indicate significance of P value (P < 0.05) |
Multivariable Logistic models to adjust for the following confounding factors: age, employment, education, exercise, diabetes family history, SBP, DBP, GGT, TG, HDL-c and HbA1C. |
Results are similar in females. In the age-adjusted model, compared to subjects in the lowest WHtR tertile, the risk of prediabetes was 1.90 times higher for subjects in the second tertile (95% CI 1.34–2.71, p < 0.001) and 2.34 times higher for subjects in the highest tertile (95% CI 1.65–3.33, p < 0.001). After controlling potential confounders, the relationship between anthropometric indices and prediabetes decreased but remained significant, except for WHR. In the multivariate-adjusted model, the OR of WHtR was the highest. The results are different in the group of males. In the age-adjusted model, all four anthropometric indices were significantly different between the prediabetic and normal blood glucose groups. In the multivariate-adjusted model, only WHR has obvious difference. Compared to subjects in the lowest WHR tertile, the risk of prediabetes was 1.89 times higher for subjects in the second tertile (95% CI 1.13–3.16, p = 0.016) and 2.62 times higher for subjects in the highest tertile (95% CI 1.58–4.33, p < 0.001).
Receiver operating characteristic (ROC) curve analysis
ROC curve analyses was performed to compare the predictive values of the anthropometric measures. The respective areas under the ROC curves (AUC) was calculated for the prediction of prediabetes. The results are presented in Table 5 and Fig. 2. In the general population, the AUC of WHtR, BMI, WC and WHR were 0.598, 0.593, 0.589 and 0.588, respectively, and the WHtR was the strongest predictor of prediabetes (p < 0.001). In females, the AUC of WHtR was the largest (AUC = 0.600) compared with other anthropometric values, including BMI (AUC = 0.592), WC (AUC = 0.593) and WHR (AUC = 0.577). It was also larger than it was in the general population. Although the AUC of WHtR was the largest both in the general population and in the group of females, this was not true in the group of males. The results showed that WHR (AUC = 0.626) had a higher predictive value than other indices in males. The AUC of WHtR, BMI and WC were 0.590, 0.592 and 0.575, respectively.
Table 5
AUC and corresponding 95% CI of different anthropometric indices in men and women
Variables | Total | Female | Male |
AUC (95% CI) | p-value | AUC (95% CI) | p-value | AUC (95% CI) | p-value |
WHtR | 0.598(0.569, 0.626) | p < 0.001* | 0.600(0.564, 0.635) | p < 0.001* | 0.590(0.540, 0.640) | p = 0.001* |
BMI | 0.593(0.564, 0.622) | p < 0.001* | 0.592(0.556, 0.627) | p < 0.001* | 0.592(0.541, 0.643) | p = 0.001* |
WC | 0.589(0.560, 0.617) | p < 0.001* | 0.593(0.558, 0.628) | p < 0.001* | 0.575(0.525, 0.626) | p = 0.004* |
WHR | 0.588(0.559, 0.618) | p < 0.001* | 0.577(0.541, 0.612) | p < 0.001* | 0.626(0.575, 0.676) | p < 0.001* |
* indicate significance of P value (P < 0.05) |
Optimal cut-off points of anthropometric indices in the prediction of prediabetes
Table 6 shows the best cut-off points and corresponding sensitivity and specificity of each index in the prediction of prediabetes. For females, the optimal cut-off point of WHtR was 0.48, with the largest Youden index and highest sensitivity, BMI was 23.5 kg/m2, WC was 75.6 cm, WHR was 0.84; for males, the best cut-off point of WHtR was 0.50, BMI was 25.5 kg/m2, WC was 88.9 cm, WHR was 0.93.
Table 6
Cutoff point for different anthropometric indices in predicting prediabetes
Variables | Female | Male | |
Cutoff point | Sensitivity | Specifity | Youden index | Cutoff point | Sensitivity | Specifity | Youden index | |
WHtR | 0.4795 | 76.2 | 40.5 | 0.167 | 0.4956 | 68.5 | 45.5 | 0.14 | |
BMI(kg/m2) | 23.5273 | 75.1 | 41.1 | 0.162 | 25.4657 | 65.1 | 53.9 | 0.19 | |
WC(cm) | 75.75 | 74.4 | 39.9 | 0.143 | 88.9 | 45.9 | 67.2 | 0.131 | |
WHR | 0.8399 | 70.8 | 43.1 | 0.139 | 0.9259 | 52.7 | 65.5 | 0.182 | |