-
The sex-based characteristics according to the glycemic status are summarized in Table 1. Compared with NGT participants, prediabetic and diabetic participants were older, more obese, and more likely to have higher blood pressure and lower eGFR. All glycemic and lipid indices (FPG, 2h-PG, HbA1c, FINS, TG, and LDL-C) increased as glucose tolerance worsened (all P < 0.001). The SUA levels were significantly higher in men than in women in all groups. Moreover, NGT individuals had the lowest SUA levels, regardless of the sex. In comparison, prediabetic men and diabetic women had the highest SUA levels and prevalence of hyperuricemia.
Characteristics Men Women NGT Prediabetes Diabetes NGT Prediabetes Diabetes Number of cases (n/N) 11,243/30,941 12,374/30,941 7,324/30,941 27,025/62,361 23,387/62,361 11,949/62,361 Age (years) 59.5 ± 9.0 61.1 ± 8.9 62.3 ± 8.8 57.2 ± 8.1 59.7 ± 8.4 62.2 ± 8.2 < 50 16.8% 12.2% 9.0% 20.0% 12.9% 7.1% 50–60 35.2% 32.5% 30.5% 44.9% 39.6% 34.3% 60–70 35.1% 38.7% 40.6% 27.9% 35.3% 40.9% > 70 12.9% 16.6% 19.8% 7.3% 12.2% 17.6% BMI (kg/m2) 24.1 ± 3.2 24.8 ± 3.3 25.5 ± 3.4 23.8 ± 3.4 24.8 ± 3.6 25.6 ± 3.8 Underweight 3.7% 3.1% 3.1% 4.1% 2.9% 2.9% Normal weight 47.2% 37.8% 29.9% 52.1% 40.5% 32.0% Overweight 38.2% 43.9% 46.7% 34.1% 40.1% 42.1% Obese 10.9% 15.2% 20.3% 9.7% 16.5% 23.0% WC (cm) 85.2 ± 9.6 87.6 ± 9.3 90.0 ± 9.3 81.2 ± 9.2 84.6 ± 9.5 87.3 ± 9.8 SBP (mmHg) 130.6 ± 18.0 136.0 ± 18.1 138.4 ± 18.9 125.7 ± 18.1 132.5 ± 18.6 137.1 ± 19.5 DBP (mmHg) 79.1 ± 11.2 81.5 ± 11.2 81.4 ± 11.4 75.6 ± 0.6 78.3 ± 10.8 77.9 ± 11.0 Scr (mg/dL) 81.6 ± 14.5 82.3 ± 15.3 83.8 ± 17.3 66.5 ± 9.2 67.3 ± 9.9 68.9 ± 13.5 TG (mg/dL) 1.5 ± 1.1 1.8 ± 1.4 2.0 ± 1.8 1.5 ± 0.9 1.8 ± 1.2 2.1 ± 1.5 LDL-C (mg/dL) 2.9 ± 0.8 3.0 ± 0.8 3.0 ± 0.8 3.1 ± 0.8 3.2 ± 0.9 3.2 ± 0.9 FPG (mg/dL) 5.1 ± 0.3 5.8 ± 0.5 7.5 ± 2.4 5.1 ± 0.3 5.7 ± 0.5 7.4 ± 2.4 2h-PG (mg/dL) 5.8 ± 1.2 7.7 ± 1.8 13.1 ± 4.8 6.0 ± 1.0 8.0 ± 1.6 13.1 ± 4.6 HbA1c (%)* 5.4 (0.4) 5.5 (0.5) 6.3 (1.4) 5.4 (0.5) 5.6 (0.5) 6.4 (1.2) FINS (pmol/L)* 5.1 (3.8) 6.0 (4.3) 6.7 (5.4) 6.0 (3.7) 7.4 (4.7) 8.4 (6.1) SUA (μmol/L) 378.8 ± 88.8 392.1 ± 91.9 379.0 ± 95.5 295.8 ± 73.4 314.4 ± 79.1 327.1 ± 88.5 Hyperuricemia (%) 30.4% 36.5% 31.8% 18.2% 26.4% 33.1% Current smoker (%) 59.3% 56.3% 55.2% 1.8% 1.8% 2.1% Current drinker (%) 38.4% 44.3% 42.2% 3.2% 3.5% 2.7% Physically active (%) 10.7% 11.4% 6.5% 6.9% 8.1% 4.9% Urban residence 56.0% 49.1% 56.8% 62.4% 54.7% 60.4% eGFR (mL/min/1.73 m2) 90.6 ± 12.7 89.0 ± 13.2 87.2 ± 14.2 90.6 ± 11.5 88.1 ± 12.1 85.0 ± 13.8 Note. Values are expressed as mean ± SD or n (%), unless otherwise indicated.
* Values are expressed as median (range).
Abbreviations: NGT, normal glucose tolerance; FPG, fasting plasma glucose; 2h-PG, 2-hour postload glucose; HbA1c, glycated hemoglobin; FINS, fasting insulin; BMI, body mass index; WC, waist circumference; SBP, systolic blood pressure; DBP, diastolic blood pressure; Scr, serum creatinine; TG, triglyceride; LDL-C, low-density lipoprotein cholesterol; SUA, serum uric acid; eGFR, estimated glomerular filtration rate.Table 1. Baseline characteristics according to glycemic status by sex
In the current study, the total prevalence of hyperuricemia were 31.2% for men and 24.1% for women; the prevalence according to demographic factors is presented in Figure 1. The prevalence increased with increase in BMI in both sexes; moreover, it was higher in individuals living in urban areas than in those living in rural areas. Nevertheless, a positive association was observed between age and prevalence of hyperuricemia in women, but a negative association was observed in men.
As mentioned earlier, the prevalence of hyperuricemia differed between men and women according to their glucose tolerance status. Therefore, the prevalence of hyperuricemia was calculated further according to glucose values (Figure 2). The prevalence initially increased with the increase in glucose levels and decreased after reaching the peak. Men with FPG levels of 6–7 mmol/L and 2h-PG of 10–12 mmol/L and women with FPG levels of 8–9 mmol/L and 2h-PG of 14–16 mmol/L had the highest prevalence of hyperuricemia. The prevalence of hyperuricemia was also the highest in men and women with an HbA1c of 6%–7%.
To identify the potential nonlinearity, the associations between glycemic indices and SUA levels were explored separately for men and women with unadjusted GAM (Figure 3A – 3C). Similar to the shapes in Figure 1, the SUA levels showed an inverted U-shaped relationship with major glycemic indices. SUA increased with an increase in FPG, 2h-PG, and HbA1c before the inflection points and then decreased with a further increase in these glycemic indices. The inflection points differed between sexes, with a threshold of 6.5 mmol/L for FPG, 11.0 mmol/L for 2h-PG, and 6.1% for HbA1c in men and 8.0 mmol/L for FPG, 14.0 mmol/L for 2h-PG, and 6.5% for HbA1c in women. The trends of associations between glycemic indices (FPG, 2h-PG, and HbA1c) and uric acid levels in all four models are visualized using GAM in Supplementary Figures S1–S2 (available in www.besjournal.com).
Figure S1. The nonlinear association between glycemic indices and uric acid in men by unadjusted and adjusted GAM model. The y-axis represents the spline function. Shade scope indicate 95% confidence bounds
Figure S2. The nonlinear association between glycemic indices and uric acid in women by unadjusted and adjusted GAM model. The y-axis represents the spline function. Shade scope indicate 95% confidence bounds
Stratified analysis was performed to verify the results of the pre-specified subgroups. As shown in Table 2, positive correlations were observed between glycemic indices and SUA levels in participants whose values did not exceed the abovementioned cut-off points. Meanwhile, a negative correlation was observed in participants whose glucose levels were greater than the cut-off points in crude models and most adjusted models. In fully adjusted models, negative associations were observed between FPG, HbA1c and SUA levels before the inflection points after accounting for serum insulin levels (−0.90 and −2.66, respectively) in men. All regression coefficients were significant in all four models
Glycemic index Model 1 Model 2 Model 3 Model 4 β SE P β SE P β SE P β SE P Men FPG (mmol/L) ≤ 6.5 6.86 1.09 < 0.0001 7.49 1.09 < 0.0001 2.16 1.04 < 0.0001 −0.90 1.08 < 0.0001 > 6.5 −7.99 0.55 < 0.0001 −8.15 0.55 < 0.0001 −9.73 0.52 < 0.0001 −9.99 0.55 < 0.0001 2h-PG (mmol/L) ≤ 11 5.06 0.31 < 0.0001 5.64 0.31 < 0.0001 3.07 0.30 < 0.0001 2.94 0.31 < 0.0001 > 11 −4.27 0.32 < 0.0001 −4.29 0.32 < 0.0001 −5.06 0.31 < 0.0001 −5.29 0.32 < 0.0001 HbA1c (%) ≤ 6.1 10.45 1.61 < 0.0001 13.22 1.62 < 0.0001 0.91 1.54 < 0.0001 −2.66 1.59 < 0.0001 > 6.1 −15.60 0.93 < 0.0001 −15.73 0.93 < 0.0001 −17.20 0.87 < 0.0001 −18.83 0.94 < 0.0001 Women FPG (mmol/L) ≤ 8 12.76 0.46 < 0.0001 10.23 0.46 < 0.0001 5.36 0.44 < 0.0001 0.60 0.45 < 0.0001 > 8 −6.10 0.67 < 0.0001 −5.89 0.67 < 0.0001 −7.63 0.63 < 0.0001 −7.76 0.66 < 0.0001 2h-PG (mmol/L) ≤ 14 7.30 0.15 < 0.0001 6.48 0.15 < 0.0001 4.61 0.14 < 0.0001 3.77 0.15 < 0.0001 > 14 −4.34 0.41 < 0.0001 −4.29 0.41 < 0.0001 −5.36 0.39 < 0.0001 −5.55 0.41 < 0.0001 HbA1c (%) ≤ 6.5 36.19 0.82 < 0.0001 30.93 0.84 < 0.0001 18.95 0.79 < 0.0001 17.63 0.82 < 0.0001 > 6.5 −13.18 0.89 < 0.0001 −12.98 0.88 < 0.0001 −14.84 0.83 < 0.0001 −16.66 0.91 < 0.0001 Note. Model 1 was an unadjusted model. Model 2 was adjusted for age. Model 3 was adjusted for age, body mass index, low-density lipoprotein cholesterol, systolic blood pressure, residence, education level, current drinking status, current smoking status, physical activity, and glomerular filtration rate. Model 4 was adjusted for age, body mass index, low-density lipoprotein cholesterol, systolic blood pressure, residence, education level, current drinking status, current smoking status, physical activity, glomerular filtration rate, and fasting insulin level. Abbreviations: FPG, fasting plasma glucose; 2h-PG, 2-hour postload glucose; HbA1c, glycated hemoglobin; SUA, serum uric acid. Table 2. Association of glycemic indices and SUA by sex
Inverted U-Shaped Associations between Glycemic Indices and Serum Uric Acid Levels in the General Chinese Population: Findings from the China Cardiometabolic Disease and Cancer Cohort (4C) Study
doi: 10.3967/bes2021.003
- Received Date: 2020-07-22
- Accepted Date: 2020-11-09
-
Key words:
- Cross-sectional study /
- Serum uric acid /
- Glycemic index /
- Glycemic status
Abstract:
Citation: | ZHU Yuan Yue, ZHENG Rui Zhi, WANG Gui Xia, CHEN Li, SHI Li Xin, SU Qing, XU Min, XU Yu, CHEN Yu Hong, YU Xue Feng, YAN Li, WANG Tian Ge, ZHAO Zhi Yun, QIN Gui Jun, WAN Qin, CHEN Gang, GAO Zheng Nan, SHEN Fei Xia, LUO Zuo Jie, QIN Ying Fen, HUO Ya Nan, LI Qiang, YE Zhen, ZHANG Yin Fei, LIU Chao, WANG You Min, WU Sheng Li, YANG Tao, DENG Hua Cong, ZHAO Jia Jun, CHEN Lu Lu, MU Yi Ming, TANG Xu Lei, HU Ru Ying, WANG Wei Qing, NING Guang, LI Mian, LU Jie Li, BI Yu Fang. Inverted U-Shaped Associations between Glycemic Indices and Serum Uric Acid Levels in the General Chinese Population: Findings from the China Cardiometabolic Disease and Cancer Cohort (4C) Study[J]. Biomedical and Environmental Sciences, 2021, 34(1): 9-18. doi: 10.3967/bes2021.003 |