Serum Uric Acid and Mortality Form Cardiovascular Disease: EPOCH-JAPAN Study

Abstract

Aim: To investigate the relationship between serum uric acid levels and cardiovascular disease in Asians.

Methods: We examined the above relationship using the data of Evidence for Cardiovascular Prevention from Observational Cohorts in Japan (EPOCH-JAPAN Study). The data of 36,313 subjects (15,628 men and 20,685 women aged 35–89 years without histories of stroke, coronary heart disease, or cancer at baseline) were used for the analyses.

Sex-specific hazard ratios (HRs) of mortality from cardiovascular disease were estimated according to the quintiles of serum uric acid using Cox hazard models stratified by cohorts.

Results: During 441,771 person-years of follow-up, we documented 1,288 cardiovascular deaths. A J- or U-shaped relationship between serum uric acid level and cardiovascular disease mortality was observed. Compared with the lowest quintile of serum uric acid levels, the highest quintile was associated with an increased cardiovascular disease mortality in men [HR: 1.28; 95% confidence interval (CI): 1.01–1.63] and women (HR: 1.51; 95% CI: 1.14–1.99). However, there was no significant association with mortality from stroke, coronary heart disease or heart failure in both men and women.

Conclusion: This large pooled analysis in Japan suggested a J- or U-shaped relationship between serum uric acid levels and cardiovascular mortality. The highest quintile of serum uric acid levels was associated with increased cardiovascular disease mortality in both Japanese men and women.

See editorial vol. 23: 671–672

Introduction

Although high serum uric acid level has been associated with an increased risk of cardiovascular disease¹⁾, there are some findings suggesting its protective effect on oxidative stress. The loss of urate oxidase activity leading to high serum uric acid levels has been hypothesized to protect the body from oxidative damage and the prolonged lifespan of hominoids^{2, 3)}. In the last decades, a number of epidemiologic studies showed conflicting results^4–15).

The Framingham Study was the first to show an independent association between serum uric acid levels and risk of cardiovascular outcomes in the general population under the careful measurement of known cardiovascular risk factors in a 23-year follow-up period⁴⁾. However, most of the participants in that study were whites, and we do not know whether the results might apply to non-white populations. A 14-year follow-up study of 8,172 Japanese men and women showed that uric acid levels were not associated with mortality from cardiovascular disease or stroke after adjustment for known cardiovascular risk¹⁰⁾. Another 8-year follow-up study of 90,393 Taiwanese men and women indicated that hyperuricemia was an independent risk factor of mortality from cardiovascular disease¹¹⁾. Two recent cross-sectional surveys showed that serum uric acid level was significantly associated with various metabolic indicators and elevated carotid intima-media thickness in the Asian Mongolian as well as middle-aged and elderly Chinese subjects, respectively^{12, 13)}. Serum uric acid levels predict the incidence of coronary heart disease but not stroke among atomic bomb survivors in Nagasaki or is a risk factor for ultrasonographically determined carotid arterial intima-media thickness in Japanese elderly persons (≥ 74 years)^{14, 15)}.

Accordingly, a large-scale prospective study on the association between serum uric acid levels and risk of mortality from cardiovascular disease in Asian adults was still required. For this purpose, the present study has been conducted.

Study Population

The Evidence for Cardiovascular Prevention from Observation Cohorts in Japan (EPOCH-JAPAN) is the pooling project of a number of wellqualified cohort studies, which investigated the relationship between health examination measures (laboratory measures and lifestyle factors) and mortality in the Japanese population. The EPOCH-JAPAN comprises 13 cohort studies in Japan with an average of 10-year follow-up periods. The year range of baseline survey in the cohort was between 1977 and 1995. The details of these projects have been previously described^16–23). A total of 90,528, of which the end-point was death owing to cardiovascular disease were included in the study. Serum uric acid was measured using a colorimetric phosphotungstic acid procedure.

Subjects were excluded if they reported a history of stroke, coronary heart disease, or cancer (n = 4,144) at baseline or if they were unable to provide data for serum uric acid levels; in addition, those who were older than 90 years or younger than 35 years (total: n = 50,071) were also excluded. Data from the remaining 36,313 subjects (15,628 men and 20,685 women) were used for the analyses.

Endpoints

The primary endpoints for this analysis were deaths from cardiovascular disease (CVD) (the International Classification of Disease, 9th revision, codes 390–459 and 10th revision, codes I01–I99), which was further divided into hemorrhagic stroke (430–431 and I60–I61) and ischemic stroke (433–434 and I63) as well as coronary heart disease (CHD, 410–414 and I20–I25) and heart failure (428 and I50).

Statistical Methods

The data of serum uric acid were divided into sex-specified quintiles. The median level of serum uric acid of each quintile was 4.0 mg/dl, 4.9 mg/dl, 5.5 mg/dl, 6.2 mg/dl, and 7.3 mg/dl for men and 3.0 mg/dl, 3.6 mg/dl, 4.1 mg/dl, 4.7 mg/dl, and 5.7 mg/dl for women. Hazard ratios (HRs) for mortalities were estimated in both men and women by Cox hazard models, which were stratified by cohorts. The adjustment variables included age (continuous), smoking status (never, past, 1–20/day, or ≥ 21/day), drinking status (drinkers, ex-drinkers, or never-drinkers), body mass index (quartiles), triglycerides (quartiles), total cholesterol (quartiles), and systolic blood pressure (continuous).

Two fitting curves of association between serum uric acid and cardiovascular mortality in both sexes were plotted to clearly demonstrate the results (Fig. 1 & 2).

Fig. 1.

Fitting curve of association between serum uric acid and cardiovascular mortality in men.

Fig. 2.

Fitting curve of association between serum uric acid and cardiovascular mortality in women.

All statistical analyses for two-tailed tests were conducted using SAS version 9.13 (SAS Institute Inc., Cary). P values of < 0.05 were regarded as statistically significant.

Results

During 441,771 person-years of follow-up, we documented 1,288 deaths from CVD (649 in men and 639 in women) including 301 total strokes, 131 coronary heart diseases, and 116 heart failures in men as well as 293 total strokes and 136 heart failures in women.

Table 1 shows the age-adjusted mean values and prevalence of cardiovascular risk factors at baseline according to the quintiles of serum uric acid levels. Men with higher uric acid were younger, whereas women with higher uric acid levels were older. Compared with men and women in the lowest quintile of uric acid levels, those in the higher quintiles were likely to be overweight and to have higher levels of total cholesterol, systolic and diastolic blood pressures, and triglycerides. Men with higher uric acid levels drank less but the opposite trend was observed in women. In addition, uric acid levels were inversely associated with high density cholesterol levels for both men and women.

Table 1. Baseline characteristics according to quintiles of serum uric acid levels.

	Quintile of serum uric acid levels					P for trend
	1 (low)	2	3	4	5 (high)
Men
No. at risk	3042	3353	2939	3052	3242
Median uric acid (mg/dl)	4.0	4.9	5.5	6.2	7.3
Range of uric acid (mg/dl)	0.6–4.6	4.7–5.2	5.3–5.8	5.9–6.6	6.7–16.0
Mean age (years)	54.6	53.3	52.6	52.0	52.9	< 0.001
Mean body mass index (kg/m2)	21.9	22.2	22.7	23.1	23.8	< 0.001
Mean total cholestrerol (mg/dl)	187.8	189.8	192.8	196.4	201.2	< 0.001
Mean HDL cholesterol (mg/dl)	53.3	52.7	51.3	50.5	50.0	< 0.001
Mean systolic blood pressure (mmHg)	131.0	131.2	132.1	133.3	137.2	< 0.001
Mean diastolic blood pressure (mmHg)	79.1	79.7	80.8	81.6	84.4	< 0.001
Median triglycerides (mg/dl)	93.0	99.0	108.0	115.0	135.0	< 0.001
Current smokers (%)	63.8	68.7	70.6	72.6	74.6	< 0.001
Current drinkers (%)	59.2	58.8	56.2	54.8	50.6	< 0.001
Women
No. at risk	4388	3933	4386	3628	4350
Median uric acid (mg/dl)	3.0	3.6	4.1	4.7	5.7
Range of Uric acid (mg/dl)	0.4–3.3	3.4–3.8	3.9–4.3	4.4–5.0	5.1–10.8
Mean age (years)	51.3	52.3	53.3	55.1	58.2	< 0.001
Mean body mass index (kg/m2)	22.1	22.5	22.8	23.3	24.2	< 0.001
Mean total cholestrerol (mg/dl)	193.9	199.5	202.5	208.5	214.2	< 0.001
Mean HDL cholesterol (mg/dl)	58.2	57.1	56.7	55.9	51.6	< 0.001
Mean systolic blood pressure (mmHg)	126.9	128.4	129.5	133.2	138.3	< 0.001
Mean diastolic blood pressure (mmHg)	75.8	76.9	77.7	79.5	81.6	< 0.001
Median triglycerides (mg/dl)	82.0	87.0	92.0	104.0	120.0	< 0.001
Current smokers (%)	5.2	5.4	6.1	7.4	8.2	< 0.001
Current drinkers (%)	13.0	15.0	16.0	18.2	16.2	< 0.001

Table 2 shows sex-specified HRs of stroke, coronary heart disease, heart failure, and total CVD according to the quintiles of serum uric acid levels. In multivariable models, we observe a J- or U-shaped relationship between serum uric acid level and total or cause-specific cardiovascular mortality. We did not observe a linear increase in mortality from almost all causes of death associated with the increase in serum uric acid levels. In most causes of death, the 5th quintile of uric acid showed highest mortality in both men and women; however, few of them reached to statistical significance. For both men and women, the highest uric acid levels (= > 6.7 mg/dl for men and = > 5.1 mg/dl for women) were associated with increased mortality owing to total CVD; we also observed statistical significance for trend test in the association between uric acid quintile and total cardiovascular mortality and in the relationship between uric acid quintile and mortality because of stroke in women.

Table 2. Sex-specific hazard ratios (95% CI) of mortality from stroke, coronary heart disease, heart failure and total cardiovascular diseases according to quintiles of serum uric acid levels.

	Quintiles of serum uric acid levels					P for trend
	1 (low)	2	3	4	5 (high)
Men
No. at risk	3042	3353	2939	3052	3242
Person Year	34762	40456	34398	38145	40323
Total stroke
No. of death	63	59	46	48	85
Age-adjusted HR	1.00	0.81 (0.56–1.15)	0.76 (0.52–1.12)	0.75 (0.52–1.10)	1.14 (0.82–1.59)	0.291
Multivariable HR†	1.00	0.83 (0.58–1.18)	0.77 (0.52–1.13)	0.77 (0.52–1.13)	1.19 (0.84–1.68)	0.258
Ischemic stroke
No. of death	35	34	26	31	47
Age-adjusted HR	1.00	0.83 (0.52–1.34)	0.76 (0.45–1.26)	0.90 (0.55–1.47)	1.19 (0.76–1.85)	0.286
Multivariable HR†	1.00	0.87 (0.54–1.40)	0.75 (0.45–1.26)	0.91 (0.55–1.50)	1.19 (0.75–1.90)	0.350
Hemorrhagic stroke
No. of death	17	18	16	15	29
Age-adjusted HR	1.00	0.89 (0.46–1.73)	1.04 (0.52–2.06)	0.81 (0.41–1.64)	1.37 (0.75–2.51)	0.248
Multivariable HR†	1.00	0.90 (0.46–1.77)	1.07 (0.54–2.14)	0.83 (0.41–1.68)	1.41 (0.75–2.65)	0.252
Coronary heart disease
No. of death	24	27	18	27	35
Age-adjusted HR	1.00	0.99 (0.57–1.72)	0.83 (0.45–1.53)	1.15 (0.66–2.00)	1.29 (0.76–2.18)	0.235
Multivariable HR†	1.00	0.98 (0.57–1.71)	0.75 (0.40–1.39)	1.02 (0.58–1.79)	1.12 (0.65–1.93)	0.600
Heart failure
No. of death	19	23	24	18	32
Age-adjusted HR	1.00	1.06 (0.57–1.95)	1.38 (0.76–2.54)	0.94 (0.49–1.80)	1.45 (0.82–2.58)	0.229
Multivariable HR†	1.00	1.09 (0.59–2.03)	1.46 (0.79–2.69)	1.05 (0.54–2.03)	1.76 (0.97–3.18)	0.066
Total cardiovascular disease
No. of death	126	127	102	111	183
Age-adjusted HR	1.00	0.87 (0.68–1.19)	0.86 (0.66–1.20)	0.87 (0.67–1.12)	1.24 (0.98–1.55)	0.028
Multivariable HR†	1.00	0.89 (0.70–1.14)	0.86 (0.66–1.12)	0.88 (0.68–1.14)	1.28 (1.01–1.63)	0.022
Women
No. at risk	4388	3933	4386	3628	4350
Person Year	51098	46069	55526	44685	56309
Total stroke
No. of death	33	42	45	51	122
Age-adjusted HR	1.00	1.22 (0.77–1.93)	0.93 (0.59–1.46)	1.07 (0.69–1.67)	1.45 (0.98–2.15)	0.024
Multivariable HR†	1.00	1.27 (0.90–2.01)	0.98 (0.62–1.54)	1.05 (0.67–1.64)	1.46 (0.98–2.19)	0.036
Ischemic stroke
No. of death	15	23	18	30	57
Age-adjusted HR	1.00	1.38 (0.71–2.63)	0.77 (0.39–1.54)	1.22 (0.65–2.29)	1.33 (0.75–2.37)	0.285
Multivariable HR†	1.00	1.42 (0.74–2.74)	0.80 (0.40–1.61)	1.22 (0.65–2.30)	1.35 (0.75–2.44)	0.314
Hemorrhagic stroke
No. of death	11	14	20	14	38
Age-adjusted HR	1.00	1.41 (0.64–3.11)	1.32 (0.63–2.76)	1.12 (0.51–2.50)	1.55 (0.78–3.07)	0.269
Multivariable HR†	1.00	1.41 (0.64–3.13)	1.33 (0.63–2.80)	1.09 (0.48–2.43)	1.54 (0.76–3.10)	0.301
Coronary heart disease
No. of death	10	13	18	23	50
Age-adjusted HR	1.00	1.28 (0.56–2.93)	1.15 (0.53–2.50)	1.60 (0.75–3.38)	1.83 (0.92–3.64)	0.032
Multivariable HR†	1.00	1.29 (0.56–2.96)	1.20 (0.55–2.61)	1.49 (0.70–3.18)	1.75 (0.87–3.54)	0.067
Heart failure
No. of death	18	13	23	26	56
Age-adjusted HR	1.00	0.64 (0.31–1.31)	0.78 (0.42–1.46)	0.86 (0.47–1.57)	1.09 (0.63–1.88)	0.200
Multivariable HR†	1.00	0.69 (0.34–1.41)	0.84 (0.45–1.56)	0.96 (0.52–1.77)	1.29 (0.74–2.25)	0.071
Total cardiovascular diseases
No. of death	68	82	106	119	264
Age-adjusted HR	1.00	1.15 (0.83–1.58)	1.04 (0.76–1.41)	1.18 (0.87–1.59)	1.46 (1.11–1.91)	< 0.001
Multivariable HR†	1.00	1.18 (0.86–1.64)	1.09 (0.80–1.48)	1.17 (0.86–1.58)	1.51 (1.14–1.99)	< 0.001

^†Adjusted further for body mass index, smoking status, ethanol intake, systolic blood pressure and total cholesterol.

To check whether there was any reverse causal bias, a sub-analysis was conducted to estimate the associations censoring the first 3 years. The results were shown in Supplementary Table and Supplementary Figs. In multivariable models, a J- or U-shaped relationship between serum uric acid level and total or cause-specific cardiovascular mortality was also observed. The whole results were quite similar to the ones shown in Table 2. Therefore, we considered that there was no or few reverse causal bias.

Supplementary Table. Sex-specific hazard ratios (95% CI) of mortality from stroke, coronary heart disease, heart failure and total cardiovascular diseases according to quintiles of serum uric acid levels censoring the first 3 years.

	Quintiles of serum uric acid levels					P for trend
	1 (low)	2	3	4	5 (high)
Men
No. at risk	2944	3270	2860	2984	3149
Person Year	34600	40315	34266	38029	40172
Total stroke
No. of death	56	53	41	44	78
Age-adjusted HR	1.00	0.81 (0.56–1.18)	0.77 (0.51–1.16)	0.77 (0.52–1.15)	1.19 (0.84–1.68)	0.211
Multivariable HR†	1.00	0.82 (0.51–1.20)	0.77 (0.51–1.16)	0.78 (0.52–1.17)	1.21 (0.84–1.74)	0.220
Ischemic stroke
No. of death	29	31	23	29	46
Age-adjusted HR	1.00	0.91 (0.55–1.51)	0.82 (0.46–1.40)	1.01 (0.60–1.70)	1.40 (0.87–2.24)	0.082
Multivariable HR†	1.00	0.94 (0.56–1.58)	0.79 (0.45–1.37)	1.01 (0.59–1.71)	1.37 (0.83–2.24)	0.142
Hemorrhagic stroke
No. of death	16	16	16	14	24
Age-adjusted HR	1.00	0.91 (0.52–1.60)	0.82 (0.44–1.51)	1.06 (0.60–1.86)	1.14 (0.66–1.94)	0.484
Multivariable HR†	1.00	0.85 (0.42–1.71)	1.19 (0.59–2.41)	0.83 (0.40–1.73)	1.29 (0.66–2.51)	0.406
Coronary heart disease
No. of death	24	25	18	25	31
Age-adjusted HR	1.00	0.99 (0.57–1.72)	0.83 (0.45–1.53)	1.15 (0.66–2.00)	1.29 (0.76–2.18)	0.485
Multivariable HR†	1.00	0.90 (0.51–1.59)	0.74 (0.40–1.37)	0.92 (0.52–1.64)	0.97 (0.55–1.70)	0.992
Heart failure
No. of death	15	20	19	16	28
Age-adjusted HR	1.00	1.13 (0.58–2.21)	1.35 (0.68–2.66)	1.01 (0.50–2.06)	1.55 (0.83–2.92)	0.193
Multivariable HR†	1.00	1.17 (0.60–2.30)	1.39 (0.70–2.77)	1.09 (0.53–2.23)	1.86 (0.97–3.57)	0.068
Total cardiovascular disease
No. of death	110	114	90	103	164
Age-adjusted HR	1.00	0.89 (0.68–1.15)	0.87 (0.66–1.15)	0.91 (0.70–1.19)	1.26 (0.99–1.61)	0.023
Multivariable HR†	1.00	0.90 (0.69–1.17)	0.86 (0.65–1.14)	0.91 (0.69–1.19)	1.28 (0.99–1.65)	0.029
Women
No. at risk	4320	3867	4329	3555	4247
Person Year	50989	45970	55440	44575	56142
Total stroke
No. of death	30	35	44	47	103
Age-adjusted HR	1.00	1.12 (0.69–1.83)	0.99 (0.62–1.58)	1.08 (0.68–1.72)	1.35 (0.89–2.04)	0.079
Multivariable HR†	1.00	1.16 (0.71–1.89)	1.05 (0.65–1.67)	1.06 (0.67–1.69)	1.35 (0.88–2.06)	0.121
Ischemic stroke
No. of death	15	21	18	29	49
Age-adjusted HR	1.00	1.25 (0.64–2.41)	0.77 (0.38–1.53)	1.18 (0.63–2.23)	1.17 (0.65–2.10)	0.562
Multivariable HR†	1.00	1.31 (0.67–2.55)	0.80 (0.40–1.60)	1.19 (0.63–2.25)	1.19 (0.65–2.18)	0.588
Hemorrhagic stroke
No. of death	10	10	20	11	34
Age-adjusted HR	1.00	1.11 (0.46–2.68)	1.44 (0.67–3.10)	0.97 (0.41–2.31)	1.53 (0.74–3.14)	0.252
Multivariable HR†	1.00	1.09 (0.45–2.64)	1.45 (0.67–3.12)	0.90 (0.37–2.15)	1.44 (0.69–3.00)	0.360
Coronary heart disease
No. of death	9	13	17	21	43
Age-adjusted HR	1.00	1.41 (0.60–3.30)	1.16 (0.51–2.61)	1.58 (0.72–3.48)	1.68 (0.81–3.48)	0.122
Multivariable HR†	1.00	1.41 (0.60–3.33)	1.20 (0.53–2.72)	1.49 (0.67–3.30)	1.61 (0.76–3.39)	0.206
Heart failure
No. of death	17	11	22	24	47
Age-adjusted HR	1.00	0.56 (0.26–1.19)	0.76 (0.40–1.43)	0.80 (0.43–1.50)	0.92 (0.52–1.62)	0.548
Multivariable HR†	1.00	0.61 (0.28–1.31)	0.81 (0.43–1.55)	0.91 (0.48–1.73)	1.11 (0.62–1.99)	0.236
Total cardiovascular diseases
No. of death	62	72	102	108	226
Age-adjusted HR	1.00	1.10 (0.78–1.54)	1.07 (0.78–1.47)	1.16 (0.84–1.59)	1.35 (1.01–1.80)	0.012
Multivariable HR†	1.00	1.13 (0.81–1.59)	1.12 (0.82–1.55)	1.16 (0.84–1.59)	1.40 (1.04–1.88)	0.010

^†Adjusted further for body mass index, smoking status, ethanol intake, systolic blood pressure and total cholesterol.

Supplementary Fig. 1.

Fitting curve of association between serum uric acid and cardiovascular mortality in men censoring the first three years

Supplementary Fig. 2.

Fitting curve of association between serum uric acid and cardiovascular mortality in women censoring the first three years

Discussion

In this large pooled analysis of Japanese cohort studies with a median follow-up of 10 years, we found a J- or U-shaped relationship between serum uric acid levels and cardiovascular mortality. Compared with the lowest quintile of serum uric acid levels, the highest quintile was associated with an increased cardiovascular mortality in both men and women; however, none of the cause-specific death was significantly associated with serum uric acid levels.

Elevated serum uric acid has been recognized as an independent risk factor for heart failure^{13, 24, 25)}. In the Framingham Offspring Study, a longitudinal observational study of 4,912 children from the original Framingham cohort participants and their spouses showed that the serum uric acid levels were associated with the incidence of heart failure. The hazard ratio (95% CI) for the highest versus lowest quintiles of serum uric acid (> 6.3 mg/dl and < 3.4 mg/dl, respectively) was 2.10 (1.04–4.22) after adjusting for sex, age, smoking, body mass index, renal dysfunction, diuretics, systolic blood pressure, valvular heart disease, diabetes, alcohol, and use of antihypertension medications²⁴⁾. Although the risk ratio did not reach statistical significance level, the magnitude of hazard ratio for both men and women in the highest uric acid quintile in the present study was almost similar to that of the Framingham Offspring Study. The MJ Health Screening Cohort conducted in Taiwan¹³⁾ failed to show significant positive associations between serum uric acid levels and mortality from heart failure in men and women but showed a weak positive association for both sexes combined; the multivariable hazard ratio (95% CI) for serum uric acid level of > 7.0 mg/dl compared with that of ≤ 7.0 mg/dl was 1.13 (1.07–1.19); p value for trend was < 0.01¹³⁾.

The associations between hyperuricemia and risk of stroke have been investigated in several studies, but the results also have been controversial^8–14). The NIP-PON DATA 80 performed in 1980 at baseline was the first cohort study to examine the association between hyperuricemia and deaths because of CVDs in a representative sample of Japanese adults¹⁰⁾. In that study, uric acid levels were not significantly associated with mortality because of CVD or stroke in either age-adjusted or multivariable models, probably because of the limited number of deaths¹⁰⁾. The present large-cohort study showed no significant association between serum uric acid levels and stroke mortality, although increased risk with borderline statistical significance was observed in the highest quintile of serum uric acid levels in women. MJ Health Screening Cohort reported increased mortality from ischemic stroke with higher serum uric acid levels in women only¹³⁾. A meta-analysis comprising 16 cohort studies and 238,449 adults showed that persons with hyperuricemia had higher age-adjusted risk for stroke: stroke incidence (6 studies; risk ratio, 1.41; 95% CI: 1.05–1.76) and mortality (6 studies; risk ratio 1.36, 95% CI: 1.03–1.69)²⁶⁾. The subgroup analysis of this study adjusting for known risk of factors including age, hypertension, diabetes mellitus, and cholesterol still showed an increased risk of stroke in persons with hyperuricemia. However, the abovementioned study was a traditional meta-analysis based on bibliographic information, which is not a pooled analysis based on individual data.

The present study showed an excess risk with borderline statistical significance for mortality because of coronary heart disease in women, but not in men. The NIPPON DATA 80¹⁰⁾ and the MJ Health Screening Cohort¹³⁾ showed no significant association between serum uric acid levels and mortality from coronary heart disease in either sex. A recent metaanalysis including 26 studies and 402,997 adults showed that hyperuricemia was associated with increased risk of coronary heart disease: an adjusted risk ratio for incidence (95% CI) of 1.09 (1.01–1.16) and an adjusted risk ratio for mortality (95% CI) of 1.16 (1.01–1.30)²⁷⁾. However, the sex-specific analysis also showed a significant association of hyperuricemia with the incidence of or mortality from coronary heart disease in women, but not in men. Furthermore, their study is not a pooled analysis based on individual data.

In the present study, the overall mortality rate from CVD in women was lower than that in men, particularly at low serum uric acid levels (Fig. 3). Therefore, the hazard ratio for the highest versus lowest quintiles of serum uric acid was higher in women and was more statistically significant than that in men. Several previous studies also showed that elevated serum uric acid levels in women were associated with a higher cardiovascular hazard ratio than that in men^{9, 13, 28)}. This may be because of estrogen that probably plays a cardioprotective role in women^{8, 13, 29)}, whereas hyperuricemia in women could possibly be a hallmark of escape from estrogen protection¹³⁾. In addition, another possible reason could be that both the mean age level and the mean glucose level were higher in the higher uric acid level (data not shown) groups of women than of men. Uric acid is also associated with diabetes or glucose intolerance and is a risk factor that confers greater relative risk for CVD in women^{8, 30, 31)}.

Fig. 3.

Sex-specific age–adjusted cardiovascular disease (CVD) mortality rates according to quintiles of serum uric acid levels.

Hyperuricemia induces endothelial dysfunction³²⁾, which may stimulate glucose assimilation in oxidative process in adipocytes³³⁾. In a populationbased study of 783 men, hyperuricemia was associated with increased renal tubular sodium re-absorption, which may provide a link with hyperinsulinemia and hypertension³⁴⁾. In a recent randomized, placebo-controlled, crossover trial of hyperuricemia involving individuals with newly diagnosed hypertension, compared with placebo (n = 15), casual and 24-hour ambulatory blood pressure levels was reduced to a greater extent with allopurinol treatment (n = 15)³⁵⁾. On the other hand, uric acid has been reported to be an antioxidant that may prevent stress-induced cell transformation and oxidant-induced cardiac and renal toxicity³⁶⁾. An in-vivo study showed that uric acid protected cultured rat hippocampal neurons against cell death induced by glutamate and NaCN insults, which are relevant to the pathogenesis of cerebral ischemia³⁷⁾. A clinical study also supported a potential neuroprotective role of the exogenous uric acid administration in stoke patients treated by thrombolysis³⁸⁾. These studies suggested that the effect of hyperuricemia on CVDs was strongly associated with other cardiovascular risk factors such as hypertension and insulin resistance. These findings may explain J- or U-shaped relationship between serum uric acid levels and cardiovascular mortality in the present study. In other words, a certain level of serum uric acid may provide benefit on cardiovascular disease to some extent; however, among the subjects with other cardiovascular risk factors, such as high glucose level or high blood pressure, higher serum uric acid levels would be associated with elevated risk of cardiovascular mortality.

The strengths of the present study include its large population-based individual data from all over Japan as well as its prospective design. To our knowledge, EPOCH-JAPAN is the largest-scale pooled data to examine the associations between uric acid level and risk of CVD in Japan. However, there were several limitations in the present study. First, the pooled data for most of the cohorts were from participants in municipal health examinations, but not the representative samples in communities. Second, we did not adjust for medication use for hypertension or diabetes as well as for glucose or creatinine level because the information of more than 30% participants was missing. However, among the participants who had the information of antihypertensive medication use, we essentially obtained the same estimation for HRs of mortality from CVD according to the quintiles of serum uric acid, adjusting further for medication use for hypertension (data not shown).

In conclusion, the results of our large pooled analysis indicated a J- or U-shaped relationship between serum uric acid level and cardiovascular mortality and also showed that the highest quintile of serum uric acid levels compared with the lowest quintiles were associated with an increased CVD mortality in both Japanese men and women.

Appendix

The Evidence for Cardiovascular Prevention from Observational Cohorts in Japan (EPOCH–JAPAN) Research Group is composed of the following investigators. Chairperson: Hirotsugu Ueshima (Shiga University of Medical Science); Co–Chairperson: Tomonori Okamura (Keio University);

Executive committee: Hirotsugu Ueshima (Shiga University of Medical Science), Yutaka Imai (Tohoku University Graduate School of Pharmaceutical Sciences), Takayoshi Ohkubo (Teikyo University School of Medicine), Fujiko Irie (Ibaraki Prefecture), Hiroyasu Iso, Akihiko Kitamura (Osaka University Graduate School of Medicine), Yutaka Kiyohara (Kyushu University Graduate School of Medicine), Katsuyuki Miura (Shiga University of Medical Science), Yoshitaka Murakami (Toho University), Hideaki Nakagawa (Kanazawa Medical University), Takeo Nakayama (Kyoto University School of Public Health), Tomonori Okamura (Keio University), Akira Okayama (Research Institute of Strategy for Prevention), Toshimi Sairenchi (Dokkyo Medical University), Shigeyuki Saitoh (Sapporo Medical University), Kiyomi Sakata (Iwate Medical University), Akiko Tamakoshi (Hokkaido University Graduate School of Medicine), Ichiro Tsuji (Tohoku University Graduate School of Medicine), Michiko Yamada (Radiation Effects Research Foundation), Masahiko Kiyama (Osaka Center for Cancer and Cardiovascular Disease Prevention), Yoshihiro Miyamoto (National Cerebral and Cardiovascular Center), Shizukiyo Ishikawa (Jichi Medical University) and Hiroshi Yatsuya (Fujita Health University).

Funding Sources

This research was supported by a grant–in–aid from the Ministry of Health, Labour and Welfare, Health and Labor Sciences research grants, Japan (Research on Health Services: H17–Kenkou–007; Comprehensive Research on Cardiovascular Disease and Life–Related Disease: H18–Junkankitou [Seishuu]–Ippan–012; Comprehensive Research on Cardiovascular Disease and Life–Related Disease: H19–Junkankitou [Seishuu]–Ippan–012; Comprehensive Research on Cardiovascular and Life–Style Related Diseases: H20–Junkankitou [Seishuu]–Ippan–013; Comprehensive Research on Cardiovascular and Life–Style Related Diseases: H23–Junkankitou [Seishuu]–Ippan–005), and an Intramural Research Fund (22-4-5) for Cardiovascular Diseases of National Cerebral and Cardiovascular Center; and Comprehensive Research on Cardiovascular and Life-Style Related Diseases: H26–Junkankitou [Seisaku]-Ippan-001.

Disclosures

None.