Table 1 Characteristics of the ten cohort prospective studies included in meta-analysis
First author, publication year [ref] | Country | Enrolled study population (casea/total, baseline characteristics) | Gender (female /male) | Age (range, mean ± SD) | Hyper- uricemia (definition [mg/dl], prevalence [%]) | Comparison (SUA, mg/dl) | Follow-up (years, mean ± SD) | Outcome | HR (95%CI) | Calculation method | Adjusted covariates |
|---|---|---|---|---|---|---|---|---|---|---|---|
Ryu et al. 2007[31] | Korea | 708/4779 without MetS, without medication and without malignancy | 0/4779 | (30–39) 33.5 ± 2.5 | NG | Highest quintile vs. lowest quintile (≥6.5 vs. <5.5) | 3 | MetS | 1.41 (1.08–1.84) | Cox proportional hazards model | Age, GGT, FBG, BMI, HDL-C, TG, BP |
Sui et al. 2008[32] | USA | M: 1120/8429 without MetS, without CVD, without cancer, with normal cardiogram | 1260/8429 | M: HU(-): 43.6±9.2; HU(+): 43.5 ± 9.0 | M: >7, 18.9 | M: Highest tertile vs. lowest tertile (≥6.71 vs. <4.97) | 5.5 ± 4.7 | MetS | M(20–39yr): 1.54(1.10–2.14) M(40–49yr): 1.50(1.14–1.96) M(≥50yr): 1.80(1.28–2.54) | Multivariable logistic regression model | Age, examination year, BMI, current smoking, alcohol intake, number of baseline metabolic risk factors, family history of disease and treadmill test |
F: 44/1260 without MetS, without CVD, without cancer, with normal electro- cardiogram | F: HU(−): 44.2 ± 9.3; HU(+): 44.1 ± 9.2 | F: >6, 4.7 | F: Highest tertile vs. lowest tertile (≥4.6 vs. <3.8) | F(20–39yr): 5.12(0.57–46.07) F(40–49yr): 3.14(0.61–16.08) F(≥50yr): 1.16(0.36–3.75) | duration | ||||||
Yanget al. 2012[34] | Chinese Taiwan | M: 214/1748 without MetS | 2109/1748 | M: T1b: 44.44 ± 16.14; T2:38.85 ± 16.52; T3:39.61 ± 16.80 | M: ≥7.7, 33.8 | M: Highest tertile vs. lowest tertile (≥7.7 vs. <6.4) | 5.41 ± 0.36 | MetS | M: 1.38 (0.86–2.66) | Cox proportional hazards model | Age, variations of BP, TG, HDL-C, FBG and WC |
F: 262/2109 without MetS | F: T1:39.32 ± 13.67; T2:39.75 ± 15.13; T3:42.90 ± 14.63 | F: ≥6.6, 18.6 | F: Highest tertile vs. lowest tertile (≥5.8 vs. <4.7) | F: 3.18 (2.2–4.6) | |||||||
Goncalves et al. 2012[35] | Portugal | F: 237/1054 without MetS | 639/418 | 49.6 ± 14.7 | M: >7, F: > 6 17.6 | HU(+)/HU(−) (≥7 vs.<7 for men, ≥6 vs.<6 for women) | 5±3.33 | MetS | 1.73(1.08–2.76) | Poisson regression model | Age, sex and education, smoking, alcohol intake, protein, calories consumption and total physical activity, one or |
Per SD increase of UA level vs. before | 1.22(1.05–1.42) | two features of MetS at baseline | |||||||||
Zhang et al. 2013[19] | China | M:776/2181 without MetS | 4442/2957 | M: 51.1 ± 14.6 | M: >7,11.9 | M: HU(+)vs. lowest quartile (>7 vs. <5.3) | 3 | MetS | M: 1.78 (1.35–2.34) | Cox proportional hazards model | Age, BMI, smoking status, drinking status, habit of regular exercise, BP, LDL-C, TG, HDL-C and FBG |
F:749/3693 without MetS | F: 46.1 ± 14.0 | F: >6, 12.6 | F: HU(+) vs. lowest quartile (>6 vs. <4.1) | F: 1.55 (1.17–2.06) | |||||||
Nagahama et al. 2013[38] | Japan | M(T1):264/1056 without MetS | 2792/3144 | MT1:(20–42) | M(T1): ≥7,32.0 | M:HU(+)/HU(-) | 4 | MetS | M(T1): 1.8(1.3–2.6) | Multivariable logistic | Alcohol consumption, smoking status, WC,BP, |
M(T2):269/784 without MetS | MT2: (43–52) | M(T2): ≥7,31.0 | (≥7/<7) | M(T2): 1.6(1.1–2.2) | regression model | dyslipidemia, FBG,GFR and medication use for | |||||
M(T3):246/1035 without MetS | MT3: (53–65) | M(T3): ≥7,25.4 | M(T3): 1.4(1.0–2.0) | hypertension, dyslipidemia, diabetes | |||||||
F(T1):40/942 without MetS | FT1: (20–45) | F(T1): ≥6,5.9 | F: HU(+)/HU(−) | F(T1): 2.2(0.9–5.5) | |||||||
F(T2):44/910 without MetS | FT2: (46–53) | F(T2): ≥6,8.7 | (≥6/<6) | F(T2): 4.4(1.8–10.6) | |||||||
F(T3):81/940 without MetS | FT3: ≥ 54 | F(T3): ≥ 6,15.0 | F(T3): 1.5(0.8–2.8) | ||||||||
Oda et al. 2014[40] | Japan | M: 177/1606 without MetS | 953/1606 | M: 51.5 ± 9.6 | M: ≥7,23.8 | HU(+) vs. lowest quantile (≥7 vs.1.1–5.2) | 2.5 | MetS | 2.615 (1.918–3.566) | Cox proportional hazards models | Age, smoking, drinking, physical activity, medication for hypertension, hyperlipidemia, |
Per 1 SD increase of UA level vs. before | 1.282 (1.097–1.499) | and diabetes, histories of CHD and stroke, MetS components | |||||||||
Per 1 increase of UA level vs. before | 1.052 (0.895–1.236) | ||||||||||
F: 71/953 withoutMetS | F: 51.0 ± 9.7 | F: ≥6,25.2 | HU(+) vs. lowest quantile (≥6 vs.1.8–3.7) | 2.088 (1.04–4.19) | |||||||
Per 1 SD increase of UA level vs. before | 1.354 (1.041–1.762) | ||||||||||
Per 1 increase of UA level vs. before | 1.313 (0.857–2.013) | ||||||||||
Xu et al. 2010[42] | China | 813/6890 without NAFLD, alcohol abusers, hepatotoxic drugs medication and hepatitis) | 4492/2398 | 44.4 ± 12.7 | M: ≥7.0 F: ≥6.0 | Highest quintile vs. lowest quintile (M: ≥6.89 vs.<4.96, F: ≥ 5.03 vs. <3.45) | 3 | NAFLD | 1.62 (1.26–2.08) | Cox proportional hazards models | Age, gender, alcohol intake, BMI, waist circumference, BP, ALT, AST, GGT, TG, total cholesterol, HDL-C, LDL-C, FPG, creatinine and BUN |
Ryu et al. 2011[33] | Korea | 1717/5741 without NAFLD, alcohol abusers, ALT elevation, liver disease, medication, | 0/5741 | 36.7 ± 4.9 | ≥7.0, 14.1% | Highest quartile vs. lowest quartile (6.5–11.5 vs.0.8–5.1) | 4.9 | NAFLD | 1.34 (1.15–1.55) | Cox proportional hazards models | Age, BMI, smoking, alcohol intake, exercise, total cholesterol, HDL-C, TG, FPG, BP, insulin, hsCRP and the MetS presence |
malignancy, CVD and diabetes | HU(+) vs. HU(−) (≥7 vs. <7) | 1.21 (1.07–1.38) | |||||||||
Per 1 increase of UA level vs. before | 1.11(1.06–1.16) |
