d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 205 (2023) 110919

Contents lists available at ScienceDirect

Diabetes Research and Clinical Practice

journal homepage: www.journals.elsevier.com/diabetes-research-and-clinical-practice

Novel pathological implications of serum uric acid with cardiovascular

disease risk in obesity
Dai Wakabayashi a, b, Sayaka Kato a, c, Masashi Tanaka a, d, *, Hajime Yamakage a, Hisashi Kato a,
Toru Kusakabe a, Naoki Ozu a, b, Shu Kasama b, Masato Kasahara b, Noriko Satoh-Asahara a, e, *,
The Japan Obesity Metabolic Syndrome Study (JOMS) Group
a
Department of Endocrinology, Metabolism and Hypertension Research, Clinical Research Institute, National Hospital Organization Kyoto Medical Center, 1-1 Fukakusa
Mukaihata-cho, Fushimi-ku, Kyoto 612-8555, Japan
b
Department of Clinical and Translational Science, Nara Medical University, Kashihara, Nara 634-8521, Japan
c
Department of Endocrinology and Metabolism, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kawaramachi-Hirokoji Kajii-cho,
Kamigyo-ku, Kyoto 602-8566, Japan
d
Department of Rehabilitation, Health Science University, 7187 Kodachi, Fujikawaguchiko-machi, Minamitsuru-gun, Yamanashi 401-0380, Japan
e
Department of Metabolic Syndrome and Nutritional Science, Research Institute of Environmental Medicine, Nagoya University, Aichi 464-8601, Japan

A R T I C L E I N F O A B S T R A C T

Keywords: Aims: This cohort study intended to elucidate the association between serum uric acid (SUA) levels and car­
Obesity diovascular disease events in Japanese patients with obesity.
Uric acid Methods: Altogether, 450 obese Japanese outpatients were enrolled in a multicenter prospective cohort Japan,
Cardiovascular disease
the Japan Obesity and Metabolic Syndrome Study. Primary analysis regarding the measurements of cardiovas­
U-shaped relationship
cular risk factors, including SUA levels, and the occurrence of macrovascular complications was based on
Effective predictor
following the participants over a 5-year period.
Results: Of the eligible patients, 335 (74.4%) were followed into the fifth year. During the study period, 15
coronary heart disease, 7 stroke, and 6 arteriosclerosis obliterans events occurred in 39 patients. The CVD
incidence rate was 15.8 per 1000 person-years. In the analysis of adjusted models for traditional risk factors,
hyperuricemia was a significant factor for the incidence of CVD events, especially in female obese patients.
Additionally, we estimated the association between SUA levels and CVD events using cubic spline models, which
showed a U-shaped association in both male and female patients.
Conclusions: SUA is an effective predictor of CVD events in female obese patients and a risk factor for CVD
incident in obese patients.

1. Introduction Uric acid (UA) is the final product of purine metabolism and has
various bioactivities, including dual effects of pro-oxidant and antioxi­
The prevalence of obesity is increasing worldwide [1]. Obesity, dant in vivo [4,5]. UA is produced in the liver and vascular endothelium
especially visceral obesity, is a risk factor of metabolic syndrome (MetS) through xanthine oxidoreductase (XOR)-related pathways [5]. Adipose
that is closely implicated in the development and progression of car­ tissues also produced UA and obesity promoted UA production by
diovascular disease (CVD) [1–3]. Obesity is also accompanied by hy­ elevating the XOR activity [7]. The serum UA (SUA) levels are positively
peruricemia, and recent extensive studies have focused on the potential associated with storage of visceral and hepatic fat in humans [8].
roles of hyperuricemia in CVD pathogenesis [4–6]; however, the rela­ Regarding the pathological significance of SUA, hyperuricemia has been
tionship between hyperuricemia and a risk for CVD in patients with implicated in various health issues, including gout, metabolic diseases,
obesity has not been fully elucidated. cardiometabolic diseases, and kidney and liver dysfunction [4–6,9–16];

* Correspondence authors at: Department of Endocrinology, Metabolism, and Hypertension Research, Clinical Research Institute, National Hospital Organization
Kyoto Medical Center, 1-1 Fukakusa Mukaihata-cho, Fushimi-ku, Kyoto 612-8555, Japan (N. Satoh-Asahara); Department of Rehabilitation, Health Science Uni­
versity, 7187 Kodachi, Fujikawaguchiko-machi, Minamitsuru-gun, Yamanashi 401-0380, Japan (M. Tanaka).
E-mail addresses: [email protected] (M. Tanaka), [email protected] (N. Satoh-Asahara).

https://doi.org/10.1016/j.diabres.2023.110919
Received 15 June 2023; Received in revised form 10 September 2023; Accepted 21 September 2023
Available online 22 September 2023
0168-8227/© 2023 Elsevier B.V. All rights reserved.
D. Wakabayashi et al. Diabetes Research and Clinical Practice 205 (2023) 110919

however, whether SUA levels would be an independent risk factor for 2.2. Data collection and laboratory measurements
future incident CVD events remain unclear. SUA levels were reported to
be not associated with incident coronary heart disease (CHD) and all- At enrollment and after 3 months in the JOMS study, the MetS-
cause and CVD mortality in the general population [17]. A recent related [BMI, waist circumference, and systolic and diastolic blood
study also reported no significant association between SUA levels and pressures (SBP and DBP)] and blood parameters were measured. Fasting
all-cause and CVD mortality in a community-based obese population blood samples were drawn in the absence of prescription drugs, and
[18]. Conversely, another general population-based study revealed that were analyzed for fasting plasma glucose (FPG), hemoglobin A1c
hyperuricemia was related to incident CVD events in women and obese (HbA1c), total cholesterol (TC), low-density lipoprotein cholesterol
patients [19]. Furthermore, a significant association of higher SUA (LDL-C), high-density lipoprotein cholesterol (HDL-C), triglycerides
levels with increased risk of all-cause and CVD mortality in patients with (TG), and serum C-reactive protein (CRP) level.
diabetes was shown by a recent epidemiological study [20]. Accord­ Participants with SUA > 7.0 mg/dL [24] and those with SUA ≤ 7.0
ingly, these findings suggest the need to conduct a cohort study mg/dL but taking UA drugs were defined as hyperuricemia in this study.
involving obese patients to better understand the pathological signifi­ The MetS score was defined as the number of risk factors that cor­
cance of SUA levels in incident CVD events in patients with obesity. responded to one of the following 1–4 MetS diagnostic criteria (Japanese
Since SUA levels are higher in men than in women, sex-specific analyses standard) [25]. The risk factors and cut-off values are as follows: 1.
are required for research on SUA [17]. Therefore, a cohort study abdominal circumference: men ≥85 cm, women ≥90 cm; 2. neutral fat:
addressing these issues would provide novel insights into the relation­ ≥150 mg/dL and/or HDL-C: <40 mg/dL; 3. blood pressure: SBP ≥130
ship between SUA and a risk for CVD events in patients with obesity. mmHg and/or DBP ≥85 mmHg; and 4. fasting blood glucose level:
We previously showed evidence of the pathological roles of obesity ≥110 mg/dL.
in CVD development and progression, using a database of a National
Hospital Organization cohort comprising patients with obesity and/or 2.3. Study endpoints
diabetes. Our multicenter prospective cohort study (Japan Obesity and
Metabolic Syndrome study: JOMS) demonstrated the utility of cardio- The primary outcomes of the study were the occurrence of CVD:
ankle vascular index, an index of arterial stiffness, as an effective pre­ macrovascular complication events defined as follows [21]. Macro­
dictor for CVD events in obese patients [21]. Moreover, urinary cystatin angiopathy endpoints included the occurrence of definite CHD (angina
C was found to be a CVD and chronic kidney disease risk factor in pa­ pectoris that is of arteriosclerotic origin and requiring percutaneous
tients with obesity and MetS [22]. In the present study, we conducted a coronary intervention or myocardial infarction), stroke (brain infarc­
5-year longitudinal study to elucidate the relationship between SUA tion, atherosclerosis-originated brain hemorrhage, or transient ischemic
levels and incident CVD events in patients with obesity without a CVD attack), or arteriosclerosis obliterans. The diagnosis of angina pectoris
history who underwent guideline-based diet and/or exercise therapy, and myocardial infarction was according to the criteria defined by the
using a cohort comprising patients with obesity. WHO/MONICA (Multinational Monitoring of Trends and Determinants
in Cardiovascular Disease) project, and the stroke diagnosis was ac­
2. Methods cording to the guidelines defined by the Ministry of Health, Labour and
Welfare of Japan [26–28]. Information on each patient’s primary
2.1. Study participants and design outcome was collected through an annual report from each physician.
To verify the endpoints, medical records for events were further
Study approval was obtained from the Central Ethics Committee for confirmed by the researchers who were independent of the physician.
Clinical Research at the National Hospital Organization headquarters
(approval number 05-27,14-034). The study was performed in accor­ 2.4. Statistical analysis
dance with the Declaration of Helsinki and Ethical Guidelines for
Medical and Health Research Involving Human Subjects. Written Data are shown as the mean ± standard deviation or median with
informed consent was obtained from participants. The JOMS study was interquartile range (IQR). Two-sided p values (at a significance level of
registered in the University Hospital Medical Information Network 0.05) are provided in this article.
(UMIN) system (UMIN Study ID: 000000559). Linear and U-shaped correlations between SUA levels and baseline
Altogether, 450 obese Japanese outpatients [body mass index (BMI) characteristics were evaluated by performing tests of linear and
≥ 25 kg/m2] were enrolled in the JOMS study that involved five Na­ quadratic contrasts in a general linear model. A trend test for the mean
tional Hospital Organization hospitals (Kyoto, Tokyo, and Nagoya SUA level stratified by the number of MetS components was performed
Medical Centers and Kokura and Mie Hospitals) and the Ooishi Clinic in in a test of linear contrast in a general linear model adjusted for age,
Japan as part of a study conducted by the Policy Based Medical Service BMI, and anti-hyperuricemia medication. Adjusted UA-estimated means
Network for Endocrine and Metabolic Diseases from October 2005 to and standard errors at each MetS score were calculated.
March 2007 [23]. The study was designed to assess the characteristics of The study endpoints were analyzed as time-to-event variables, and
MetS and CVD risk factors, and the onset of CVD events for 5 years in the Cox proportional hazard model was used to calculate the unadjusted
obese Japanese patients. During the follow-up period, the patients were and adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) for
recommended optimal treatments, especially weight-reduction therapy risk factors. The SUA values were divided into quartiles by sex. The
through lifestyle modifications to reduce energy intake, and an increase reference standard for HR was the SUA range with the lowest risk for
in physical activity [21,22]. All patients undergoing weight-reduction each sex. HRs were calculated by crude and adjusted analyses by age,
therapy were instructed to maintain the same levels of energy intake BMI, and anti-hyperuricemia. Restricted cubic splines with knots were
and physical activity for the entire period, as recommended by the Japan used to further explore the shape of the dose–response relationship be­
Atherosclerosis Society’s “Guidelines for the diagnosis and treatment of tween the SUA level and HR of CVD onset.
atherosclerotic cardiovascular disease” [21,22]. The dietary therapy All statistical analyses were performed using SPSS 24.0 for Windows
comprised 25 kcal/kg of ideal body weight per day. The participants (SPSS Inc., Chicago, IL, USA).
consumed 60% of the total energy as carbohydrates, 20%–25% as fat,
and 15%–20% as protein.

2
D. Wakabayashi et al. Diabetes Research and Clinical Practice 205 (2023) 110919

3. Results men with higher SUA (p = 0.002 and p < 0.001, respectively). Waist
circumference and hs-CRP tended to be higher in women with higher
3.1. Clinical variables and their changes SUA level (p = 0.014 and p = 0.002, respectively), and HbA1c level and
antidiabetic rates tended to be lower in women with higher SUA level (p
The patients’ baseline clinical characteristics are shown in Table 1. = 0.043 and p < 0.001, respectively).
The median follow-up time was 5.0 years and the total person-years The association between the number of MetS component and SUA
were 1862.4. Of the eligible patients, 74.4% were followed into the level by sex is shown in Fig. 2. In women, the more MetS score gained,
fifth year. The dropout rate, which is the proportion of patients who the statistically significantly higher UA levels were observed (p = 0.023
were lost to follow-up until the fifth year, was 25.6%. During follow-up, for trend). Contrarily, no significant association was observed in men (p
one patient died of cancer and one died of old age. Of the 450 enrolled = 0.089 for trend).
obese patients, 115 were lost to follow-up, and observations were dis­
continued for the following reasons: 41 patients stopped visiting the
3.2. Endpoint analysis
outpatient clinics (21 patients moved residence, 20 patients had a
change in their work schedule), 56 patients moved to other clinics since
During follow-up, 22 CHD, 11 stroke, and 6 arteriosclerosis obliter­
they successfully lost weight, 11 patients stopped consultation visits for
ans events occurred. Among all CHD events, 63.6% (n = 14) were angina
unknown reasons, and seven patients withdrew participation (Fig. 1).
pectoris and 36.3% (n = 8) were myocardial infarction, and among all
There were no significant differences in the baseline characteristics
stroke events, 72.7% (n = 8) were brain infarction, 18.2% (n = 2) were
between patients who completed the study and those who did not,
atherosclerosis-originated brain hemorrhage, and 9.1% (n = 1) were
except for BMI (31.1 ± 5.9 vs. 32.5 ± 7.8 kg/m2, p = 0.045). The median
transient ischemic attack. Therefore, all events (CHD, stroke, and arte­
SUA level at baseline for the entire cohort was 5.7 mg/dL (IQR: 4.7–6.7);
riosclerosis obliterans) occurred in 39 patients, and the CVD incidence
it was higher in men [6.5 mg/dL (IQR: 5.5–7.5)] than in women [5.2
rate was 20.9 per 1,000 person-years (CHD, stroke, and arteriosclerosis
mg/dL (IQR: 4.3–6.0) (Table 1). The prevalence of hyperuricemia was
obliterans: 11.8, 5.9, and 3.2 per 1,000 person-years, respectively).
38.6% and 7.3% in men and women, respectively.
The SUA levels were divided into sex-specific quartile groups and
In Table 2, the baseline characteristics of obese patients are shown
their association with CVD events was examined (Table 3), with 2nd
for each group divided by quartiles of SUA level (men: Q1, 1.9–5.4 mg/
quartile (5.5–6.4) in men and 3rd quartile (5.2–5.9) in women set as
dL; Q2, 5.5–6.4 mg/dL; Q3, 6.5–7.5 mg/dL; Q4, 7.6–10.7 mg/dL;
references. Using the Cox regression model, there was no significant
women: Q1, 0.6–4.2 mg/dL; Q2, 4.3–5.1 mg/dL; Q3, 5.2–5.9 mg/dL; Q4,
difference among the male groups but, in the female groups, high SUA
6.0–11.0 mg/dL). BMI tended to be higher with higher SUA level in both
level (4th quartile) showed significant difference compared to the
men and women (p = 0.027 and p = 0.001, respectively, in linear), but
reference (HR: 4.71, 95%CI: 1.03–21.50, p = 0.045). Even after multi­
the FPG level and eGFR tended to be lower with higher SUA level in both
variable adjustment by age, BMI, and anti-hyperuricemia medication,
men and women (FPG: p = 0.015 and p = 0.003, respectively; eGFR: p =
the results were similar to those before the adjustment.
0.011 and p = 0.004, respectively). Contrarily, the TC and LDL-C levels
Additionally, cubic spline models were used to estimate the associ­
tended to be higher in men with higher SUA levels (p = 0.004 and p =
ation between SUA levels and CVD events in more detail (Fig. 3). Both
0.026, respectively), and age and antidiabetic rates tended to be lower in
men and women showed a U-shaped association between SUA levels and

Table 1
Baseline characteristics of the study obese patients stratified by sex.
Valuables Total Men Women

N (men/women) 450 (202 / 248) 202 248

Age (years) 51.5 ± 14.2 49.7 ± 14.4 53.0 ± 13.8
BMI (kg/m2) 31.1 ± 5.9 31.3 ± 6.5 31.0 ± 5.4
SBP (mmHg) 140.4 ± 19.0 140.6 ± 18.1 140.2 ± 19.7
DBP (mmHg) 84.1 ± 11.7 85.5 ± 12.5 82.9 ± 10.9
FPG (mg/dL) 123.6 ± 52.0 121.0 ± 43.5 125.8 ± 58.0
HbA1c (%) 6.3 [5.8, 7.4] 6.1 [5.7, 7.3] 6.3 [5.8, 7.4]
IRI (μU/mL) 14.6 [8.3, 26.9] 15.5 [9.1, 28.6] 13.8 [7.7, 24.9]
Total cholesterol (mg/dL) 211.1 ± 58.2 205.9 ± 37.3 215.3 ± 70.7
Triglyceride (mg/dL) 145.0 [103.0, 220.3] 150.5 [107.0, 256.5] 139.5 [101.3, 203.0]
HDL-cholesterol (mg/dL) 54.7 ± 14.0 50.0 ± 12.9 58.5 ± 13.8
LDL-cholesterol (mg/dL) 126.1 ± 31.5 125.8 ± 32.9 126.4 ± 30.4
eGFR 82.8 ± 24.4 83.0 ± 24.7 82.7 ± 24.2
UA (mg/dL) 5.7 [4.7, 6.7] 6.5 [5.5, 7.5] 5.2 [4.3, 6.0]
hs-CRP (μg/mL) 0.81 [0.43, 1.92] 0.79 [0.44, 1.89] 0.83 [0.41, 1.99]

Proportion of
hypertension (n, %) 285 , 63.3 122 , 60.4 163 , 65.7
dyslipidemia (n, %) 344 , 76.4 156 , 77.2 188 , 75.8
diabetes (n, %) 212 , 47.1 94 , 46.5 118 , 47.6
hyperuricemia (n, %) 96 , 21.3 78 , 38.6 18 , 7.3
taking calcium antagonist (n, %) 103 , 22.9 44 , 21.8 59 , 23.8
taking ACEi/ARB (n, %) 126 , 28.0 59 , 29.2 67 , 27.0
taking statins (n, %) 102 , 22.7 30 , 14.9 72 , 29.0
taking antidiabetic medications (n, %) 142 , 31.6 56 , 27.7 86 , 34.7
taking UA drug (n, %) 6 , 1.3 4 , 2.0 2 , 0.8
current smoking (n, %) 77 , 17.1 45 , 22.3 32 , 12.9

Data are expressed as mean ± standard deviation, median [interquartile range], or the number and percentage of patients. BMI: Body mass index, SBP: systolic blood
pressure, DBP: diastolic blood pressure, FPG: fasting plasma glucose, HbA1c: hemoglobin A1c, IRI: immunoreactive insulin, TG; triglycerides, HDL: high-density li­
poprotein, LDL: low-density lipoprotein, eGFR: estimated glomerular filtration rate, UA: uric acid, hs-CRP: high sensitive C-reactive protein, ACEi: angiotensin-
converting enzyme inhibitor, ARB: angiotensin II type 1 receptor blocker.

3
D. Wakabayashi et al. Diabetes Research and Clinical Practice 205 (2023) 110919

Fig. 1. Flow diagram of the JOMS. CVD: cardio vascular disease.

CVD events (Fig. 3A and B). The SUA level in the bottom value of the HR effects of elevated SUA levels. UA is produced in various tissues [5] and
was 6.6 and 5.2 mg/dL for men and women, respectively. In women, the obesity increases UA production in adipose tissue by elevating the XOR
SUA level of >7.0 mg/dL confirmed that the lower 95% CI limit of the activity [7]. Elevated SUA levels would exert deleterious effects on
HR was >1.0. oxidative stress, inflammation, and endothelial function, thereby lead­
ing to an increased CVD risk [4], potentially in both men and women.
4. Discussion Conversely, SUA is mechanistically maintained at lower levels in women
than in men in physiological settings through various mechanisms, such
This is the first study to show that hyperuricemia is a novel marker to as estrogen-related pathways [32]. Accordingly, women may be more
independently predict incident CVD events in women with obesity, a vulnerable to elevation-induced pathological effects of SUA than men,
finding obtained by a longitudinal multicenter cohort study on obese thereby leading to increased risk for CVD events in women in conjunc­
patients without a CVD history over a 5-year follow-up period. tion with obesity and/or decrease in estrogen levels. Supporting these
Furthermore, a U-shaped relationship between SUA levels and HR of possibilities, SUA levels were positively associated with exacerbation of
CVD events was found in both sexes, suggesting that lower and higher inflammation and MetS scores, which are risk factors for atherosclerosis,
SUA levels are risk factors for incident CVD events in these patients. in women with obesity in this study. Alternatively, estrogen has po­
These findings highlight the novel clinical significance of SUA in pre­ tential protective effects against CVD [33], so that a decrease in estrogen
venting incident CVD events in patients with obesity. levels and an increase in SUA levels would synergistically increase a risk
We found a sex difference in the association of hyperuricemia with a for CVD events.
risk for CVD events in obese patients. Similar female-specific relation­ Another possibility concerning the sex difference is the potential
ships have been reported between SUA levels and a high risk for CVD masking effects of glucose metabolism. Diabetes is a risk factor for CVD
events in a general population [19], arteriosclerotic CVD in patients [34], and higher SUA levels were correlated with improved glucose
with type 2 diabetes (T2D) [29], or all-cause mortality in obese or metabolism in this study, thereby leading to no significant relationship
overweight patients with pre-existing T2D and/or CVD [30]. Since between higher SUA levels and CVD events in men. Our multivariate
fructose is a carbohydrate that generates UA [31] and metabolic activity analysis also revealed that age, which was negatively correlated with
for fructose is suggested to differ between sexes [30], SUA elevation in SUA, was not a confounding factor affecting the relationship of SUA with
women might reflect more severe metabolic derangements [30]; how­ CVD events in men. Thus, the significant relationship between SUA and
ever, the underlying mechanisms remain unclear. CVD in women obtained in this study would highlight the more direct
Our findings suggest that the observed sex difference might be implications of SUA in CVD in women compared with men. Further­
ascribed to the potential difference in vulnerability to detrimental more, androgen exhibits protective effects on the cardiovascular system

4
D. Wakabayashi et al. Diabetes Research and Clinical Practice 205 (2023) 110919

Table 2
Correlation between serum uric acid level and baseline characteristics stratified by sex.
Serum uric acid concentration quartiles p p

Q1 Q2 Q3 Q4 linear U-shape

Men
n 48 50 54 50
UA (mg/dL) 1.9 – 5.4 5.5 – 6.4 6.5 – 7.5 7.6 – 10.7
Age (years) 54.7 ± 14.5 51.1 ± 12.2 46.4 ± 14.4 47.1 ± 15.2 0.002 0.277
BMI (kg/m2) 29.4 ± 5.1 31.2 ± 6.8 32.6 ± 7.3 32.0 ± 6.4 0.027 0.214
Waist circumference (cm) 100.4 ± 11.1 101.8 ± 12.7 105.9 ± 17.0 103.8 ± 16.4 0.127 0.392
SBP (mmHg) 139.7 ± 16.5 141.3 ± 17.7 141.2 ± 19.3 140.2 ± 18.9 0.917 0.602
DBP (mmHg) 83.6 ± 9.7 86.1 ± 9.2 86.8 ± 15.2 85.5 ± 14.4 0.423 0.277
FPG (mg/dL) 127.0 ± 49.8 129.4 ± 48.2 120.2 ± 40.1 107.7 ± 32.3 0.015 0.220
HbA1c (%) 6.6 ± 1.7 6.3 ± 1.5 6.3 ± 1.4 6.0 ± 1.3 0.080 0.950
TC (mg/dL) 195.1 ± 36.1 201.4 ± 34.2 211.8 ± 37.4 214.3 ± 38.9 0.004 0.714
TG (mg/dL) 164.9 ± 92.0 220.6 ± 251.9 201.4 ± 149.0 211.2 ± 117.0 0.255 0.323
HDL-C (mg/dL) 48.5 ± 12.5 48.5 ± 11.6 53.1 ± 15.5 49.4 ± 11.0 0.366 0.309
LDL-C (mg/dL) 120.0 ± 35.6 120.4 ± 33.3 129.9 ± 28.2 132.4 ± 33.7 0.026 0.828
eGFR 86.4 ± 19.8 87.1 ± 28.7 84.8 ± 22.8 73.6 ± 24.5 0.011 0.089
hs-CRP (μg/mL) 6.6 ± 1.1 6.7 ± 1.4 7.0 ± 1.2 6.9 ± 1.1 0.125 0.507

Medications
antihypertension 24 , 50.0 13 , 26.0 21 , 38.9 15 , 30.0 0.122 0.262
antidiabetic 23 , 47.9 12 , 24.0 15 , 27.8 6 , 12.0 <0.001 0.506
antidyslipidemia 21 , 43.8 18 , 36.0 21 , 38.9 18 , 36.0 0.516 0.725

Women
n 57 64 60 67
UA (mg/dL) 0.6 – 4.2 4.3 – 5.1 5.2 – 5.9 6.0 – 11.0
Age (years) 53.1 ± 13.5 55.6 ± 12.2 53.8 ± 12.9 49.8 ± 15.8 0.134 0.063
BMI (kg/m2) 28.9 ± 3.5 31.1 ± 5.9 31.0 ± 6.1 32.6 ± 5.1 0.001 0.599
Waist circumference 96.1 ± 9.7 100.4 ± 14.3 99.3 ± 13.3 102.3 ± 11.7 0.014 0.667
SBP (mmHg) 139.2 ± 17.1 140.4 ± 17.8 140.8 ± 19.1 140.5 ± 24.0 0.972 0.772
DBP (mmHg) 83.1 ± 9.7 81.0 ± 9.6 82.8 ± 12.3 84.8 ± 11.6 0.259 0.132
FPG (mg/dL) 140.9 ± 73.2 133.6 ± 65.0 116.0 ± 36.8 114.1 ± 48.0 0.003 0.713
HbA1c (%) 6.6 ± 1.4 6.6 ± 1.6 6.4 ± 1.3 6.1 ± 1.1 0.043 0.355
TC (mg/dL) 204.7 ± 28.9 212.5 ± 36.5 217.0 ± 34.6 225.5 ± 124.2 0.098 0.971
TG (mg/dL) 158.9 ± 124.7 176.2 ± 105.3 147.7 ± 78.2 175.7 ± 80.3 0.696 0.671
HDL-C (mg/dL) 60.4 ± 13.7 57.8 ± 13.4 61.8 ± 13.7 54.8 ± 13.5 0.101 0.201
LDL-C (mg/dL) 119.3 ± 26.2 127.5 ± 29.3 131.0 ± 34.1 127.3 ± 31.0 0.111 0.123
eGFR 88.6 ± 23.8 85.1 ± 19.7 81.2 ± 23.9 76.6 ± 27.5 0.004 0.866
hs-CRP (μg/mL) 6.2 ± 1.1 6.9 ± 1.2 6.9 ± 1.2 7.0 ± 1.2 0.002 0.068

Medications
antihypertension 21 , 36.8 20 , 31.3 26 , 43.3 24 , 35.8 0.744 0.876
antidiabetic 29 , 50.9 25 , 39.1 20 , 33.3 12 , 17.9 <0.001 0.760
antidyslipidemia 26 , 45.6 30 , 46.9 26 , 43.3 32 , 47.8 0.919 0.804

Data are expressed as mean ± standard deviation, or the number and percentage of patients. Abbreviations used in this table are the same as in Table 1.
Linear and U-shaped correlations between serum uric acid levels and baseline characteristics were evaluated by tests of first and second order contrasts in a general
linear model.

[35], which would also contribute to no significant relationship between for CVD. Reportedly, exacerbation of glucose metabolism resulted in the
SUA and CVD in men. Although additional studies are required, our decrease in SUA levels [39]. Our study also revealed a negative associ­
findings suggest the increased need to focus on hyperuricemia in women ation of SUA levels with FPG and antidiabetic usage rate. Thus, reduc­
with obesity, as compared to men, to reduce the risk of incident CVD tion of UA-related beneficial activities and/or aggravation of glucose
events. metabolism in hypouricemia would be implicated in a high CVD risk in
In this study, a U-shaped relationship was observed between the SUA obese patients.
levels and incident CVD events in both sexes, suggesting the patholog­ As assumed by the U-shaped relationships, the pathological impacts
ical roles of hypouricemia and hyperuricemia in CVD events develop­ of SUA levels would change according to its concentrations, thereby
ment in obese patients. Although a similar U-shaped association of SUA suggesting the need to control SUA levels within an appropriate range to
with a CVD risk has been reported in various populations [36–38], the prevent CVD events. In this respect, we found that the optimal target
mechanistic details have not been fully understood. Regarding hyper­ window of SUA levels to reduce the risk of CVD events differs between
uricemia, detrimental effects of elevated SUA levels would be implicated male and female obese patients. The SUA values corresponding to the
in the increased risk of CVD in obesity, as discussed above. Conversely, lowest risk of incident CVD events were lower in women (5.2 mg/dL)
the pathological roles of hypouricemia would be caused by the attenu­ than in men (6.6 mg/dL) with obesity in this study, suggesting that the
ation of beneficial effects of UA. Since UA is one of the major endoge­ optimal SUA values for male obese patients rather increased the CVD
nous antioxidants in humans [4], hypouricemia would result in risk for female obese patients, as previously reported in a general pop­
exacerbation of oxidative stress and subsequent vascular dysfunction, ulation [19]. Similarly, the appropriate SUA values for women would
thereby leading to the increased risk for CVD [37]. Another possibility is not be applicable to men with obesity. These findings highlight the need
that lower SUA levels reflect aggravation of hyperglycemia, a risk factor to revisit the guideline recommendations for reference values of SUA

5
D. Wakabayashi et al. Diabetes Research and Clinical Practice 205 (2023) 110919

Table 3
Adjusted hazard ratio for cardiovascular disease onset by Cox regression models
according to serum uric acid level by sex.
Serum uric acid concentration quartiles

Q1 Q2 Q3 Q4

Men
N 50 48 54 50
UA (mg/ 1.9–5.4 5.5–6.4 6.5–7.5 7.6–10.7
dL)
HR for CVD (HR [95%CI], p)
Crude 2.91 [0.77, reference 1.04 [0.21, 1.40 [0.31,
model 10.98], 0.114 5.15], 0.962 6.26], 0.659
Adjusted 2.37 [0.62, reference 1.15 [0.23, 1.52 [0.34,
model 9.02], 0.205 5.67], 0.868 6.78], 0.586

Women
N 64 57 60 67
UA (mg/ 0.6–4.2 4.3–5.1 5.2–5.9 6.0–11.0
dL)
HR for CVD (HR [95%CI], p)
Crude 2.20 [0.40, 1.90 [0.35, reference 4.71 [1.03,
model 12.02], 0.362 10.38], 0.459 21.50], 0.045
Adjusted 2.24 [0.41, 1.81 [0.33, reference 5.05 [1.10,
model 12.25], 0.351 9.88], 0.495 23.08], 0.037

HR: hazard ratio, 95%CI: 95% confidence interval, other abbreviations used in
this table are the same as in Table 1. Adjusted model: adjusted for age, BMI, and
anti-hyperuricemia medication.

underlying the association of SUA levels with CVD risk in obesity.

Finally, racial and ethnic differences would exist among obese patients.
Future studies across diverse groups would allow a comprehensive un­
derstanding of the pathological significance of SUA levels in incident
CVD events in patients with obesity.
In conclusion, this study provided the first evidence that hyperuri­
cemia is an independent predictive marker for incident CVD events in
female obese patients without a CVD history. Accordingly, measuring
the SUA levels would allow the identification of patients with an
increased risk for CVD events, thereby indicating the need to reduce the
CVD risk by intensive treatments in these patients. The U-shaped rela­
tionship suggests that hypouricemia and hyperuricemia would be
implicated in an increased risk of CVD in both male and female obese
patients. We further found a sex difference between the optimal win­
dows of SUA levels to reduce the risk of incident CVD events. These
findings would be helpful for developing novel strategies for predicting
and preventing incident CVD events in patients with obesity.
Fig. 2. Severity of metabolic syndrome (MetS) risk factors and SUA level Funding
among men (A) and women (B). Association between the number of MetS This work was supported in part by Grant-in-Aid for Scientific
components and SUA levels is shown for men (p for trend = 0.089) and women Research (C) to HY (JP22K11720), and (B) to NS-A (21H02835), and by
(p for trend = 0.023) adjusted for age, BMI and anti-hyperuricemia medication. Grant-in-Aid for Exploratory Research to NS-A (JP22K19723) from
n indicates the number of patients included in the group with the cumulative Japan Society for the Promotion of Science. This study was also sup­
MetS components. Bars show adjusted SUA-estimated means and standard er­
ported in part by a grant from the National Hospital Organization for
rors at each MetS score.
collaborative clinical research to NS-A (H26-NHO-02). The funders had
no role in the design of the study; in the collection, analyses, or inter­
that are the same between sexes, thereby contributing to the develop­ pretation of data; in the writing of the manuscript, or in the decision to
ment of effective diagnostic guidelines and therapeutics for preventing publish the results.
incident CVD events.
The main strength of this study is that this is the well-characterized, CRediT authorship contribution statement
prospective longitudinal cohort study on patients with obesity of both
sexes; this allowed analyses specialized for pathological significance of Dai Wakabayashi: Conceptualization, Validation, Investigation,
SUA in a risk of CVD in obesity. Nevertheless, this study has some lim­ Data curation, Writing – original draft, Writing – review & editing.
itations. First, the effects of potential confounding factors (e.g., alcohol Sayaka Kato: Conceptualization, Validation, Investigation, Data cura­
consumption) were not investigated due to the limited sample size. A tion, Writing – original draft, Writing – review & editing. Masashi
longitudinal cohort study with a larger sample size and a longer follow- Tanaka: Conceptualization, Writing – original draft, Writing – review &
up period is required to corroborate our findings. Second, we did not editing. Hajime Yamakage: Validation, Formal analysis, Investigation,
measure the sex hormone levels and XOR activity that would affect the Writing – original draft, Writing – review & editing. Hisashi Kato:
SUA activity. Potential effects of renal functions on the relationship Writing – review & editing. Toru Kusakabe: Writing – review & editing.
between SUA levels and CVD risk also remain unclear. Additional Naoki Ozu: Writing – review & editing. Shu Kasama: Writing – review
studies addressing these issues would elucidate the mechanisms

6
D. Wakabayashi et al. Diabetes Research and Clinical Practice 205 (2023) 110919

Fig. 3. Cubic spline models for the association between serum uric acid (SUA) level and hazard ratios (HRs) for cardiovascular disease (CVD) events among men (A)
and women (B). The reference standard for the hazard ratio was the median SUA. The dashed line indicates the 95% confidence interval (CI). 95%CI+: the upper
limit of the 95% confidence interval, 95%-: the lower limit of the 95% confidence interval. The SUA level corresponding to the lowest hazard ratio of CVD events is
6.6 mg/dL for men and 5.2 mg/dL for women.

& editing. Masato Kasahara: Writing – review & editing. Noriko Declaration of Competing Interest
Satoh-Asahara: Conceptualization, Validation, Investigation, Re­
sources, Data curation, Writing – original draft, Writing – review & The authors declare that they have no known competing financial
editing. interests or personal relationships that could have appeared to influence
the work reported in this paper.

7
D. Wakabayashi et al. Diabetes Research and Clinical Practice 205 (2023) 110919

Acknowledgments [19] Katsiki N, Kouvari M, Panagiotakos DB, Borghi C, Chrysohoou C, Mikhailidis DP,
et al. The association between serum uric acid levels and 10-year cardiovascular
disease incidence: results from the ATTICA prospective study. Rev Cardiovasc Med
Special thanks go to Kazunori Koyama, Yasuhisa Kato, Tsutomu 2021;22:991–1001.
Yamada, Taiichiro Okajima, Makito Tanabe, Rika Aaraki, and Mariko [20] Li B, Chen L, Hu X, Tan T, Yang J, Bao W, et al. Association of serum uric acid with
Ooishi for their contribution to patient recruitment. We would like to all-cause and cardiovascular mortality in diabetes. Diabetes Care 2023;46:425–33.
[21] Satoh-Asahara N, Kotani K, Yamakage H, Yamada T, Araki R, Okajima T, et al.
thank Satista Co., Ltd. (www.satista.jp) for supporting medical statistics. Cardio-ankle vascular index predicts for the incidence of cardiovascular events in
We also thank the members of the Satoh-Asahara laboratories for their obese patients: a multicenter prospective cohort study (Japan Obesity and
helpful discussions. We would like to thank Enago (www.enago.jp) for Metabolic Syndrome Study: JOMS). Atherosclerosis 2015;242:461–8.
[22] Satoh-Asahara N, Suganami T, Majima T, Kotani K, Kato Y, Araki R, et al. Urinary
the English language review. cystatin C as a potential risk marker for cardiovascular disease and chronic kidney
disease in patients with obesity and metabolic syndrome. Clin J Am Soc Nephrol
References 2011;6:265–73.
[23] Satoh N, Shimatsu A, Kato Y, Araki R, Koyama K, Okajima T, et al. Evaluation of
the cardio-ankle vascular index, a new indicator of arterial stiffness independent of
[1] Muzurovic EM, Volcansek S, Tomsic KZ, Janez A, Mikhailidis DP, Rizzo M, et al.
blood pressure, in obesity and metabolic syndrome. Hypertens Res 2008;31:
Glucagon-Like peptide-1 receptor agonists and dual glucose-dependent
1921–30.
insulinotropic polypeptide/glucagon-like peptide-1 receptor agonists in the
[24] Hisatome I, Ichida K, Mineo I, Ohtahara A, Ogino K, Kuwabara M, et al. Japanese
treatment of obesity/metabolic syndrome, prediabetes/diabetes and non-alcoholic
Society of Gout and Uric & Nucleic Acids 2019 guidelines for management of
fatty liver disease-current evidence. J Cardiovasc Pharmacol Ther. 2022;27:
hyperuricemia and Gout 3rd edition. Gout Uric Nucl Acids 2020;44:sp-1–sp-40.
10742484221146371.
[25] [Definition and the diagnostic standard for metabolic syndrome–Committee to
[2] Park S, Lee S, Kim Y, Lee Y, Kang MW, Han K, et al. Altered risk for cardiovascular
Evaluate Diagnostic Standards for Metabolic Syndrome]. Nippon Naika Gakkai
events with changes in the metabolic syndrome status: a nationwide population-
Zasshi 2005;94:794–809.
based study of approximately 10 million persons. Ann Intern Med 2019;171:
[26] Peltonen M, Stegmayr B, Asplund K. Time trends in long-term survival after stroke:
875–84.
the Northern Sweden Multinational Monitoring of Trends and Determinants in
[3] Lombardo M, Feraco A, Bellia C, Prisco L, D’Ippolito I, Padua E, et al. Influence of
Cardiovascular Disease (MONICA) study, 1985–1994. Stroke 1998;29:1358–65.
nutritional status and physical exercise on immune response in metabolic
[27] Sone H, Katagiri A, Ishibashi S, Abe R, Saito Y, Murase T, et al. Effects of lifestyle
syndrome. Nutrients 2022;14.
modifications on patients with type 2 diabetes: the Japan Diabetes Complications
[4] Yu W, Cheng JD. Uric acid and cardiovascular disease: an update from molecular
Study (JDCS) study design, baseline analysis and three year-interim report. Horm
mechanism to clinical perspective. Front Pharmacol 2020;11:582680.
Metab Res 2002;34:509–15.
[5] Yanai H, Adachi H, Hakoshima M, Katsuyama H. Molecular biological and clinical
[28] Sone H, Tanaka S, Iimuro S, Tanaka S, Oida K, Yamasaki Y, et al. Long-term
understanding of the pathophysiology and treatments of hyperuricemia and its
lifestyle intervention lowers the incidence of stroke in Japanese patients with type
association with metabolic syndrome, cardiovascular diseases and chronic kidney
2 diabetes: a nationwide multicentre randomised controlled trial (the Japan
disease. Int J Mol Sci 2021;22.
Diabetes Complications Study). Diabetologia 2010;53:419–28.
[6] Crawley WT, Jungels CG, Stenmark KR, Fini MA. U-shaped association of uric acid
[29] Yang H, Gao J, Li S, Xia H, Chen Z, Zhu S, et al. Gender differences in the
to overall-cause mortality and its impact on clinical management of hyperuricemia.
association between serum uric acid and arteriosclerotic cardiovascular risk among
Redox Biol 2022;51:102271.
Chinese Type 2 diabetes mellitus patients. Int J Gen Med 2021;14:687–95.
[7] Tsushima Y, Nishizawa H, Tochino Y, Nakatsuji H, Sekimoto R, Nagao H, et al. Uric
[30] Skak-Nielsen H, Torp-Pedersen C, Finer N, Caterson ID, Van Gaal L, James WP,
acid secretion from adipose tissue and its increase in obesity. J Biol Chem 2013;
et al. Uric acid as a risk factor for cardiovascular disease and mortality in
288:27138–49.
overweight/obese individuals. PLoS One 2013;8:e59121.
[8] Rospleszcz S, Dermyshi D, Muller-Peltzer K, Strauch K, Bamberg F, Peters A.
[31] Russo E, Leoncini G, Esposito P, Garibotto G, Pontremoli R, Viazzi F. Fructose and
Association of serum uric acid with visceral, subcutaneous and hepatic fat
Uric acid: major mediators of cardiovascular disease risk starting at pediatric age.
quantified by magnetic resonance imaging. Sci Rep 2020;10:442.
Int J Mol Sci 2020;21.
[9] Ali N, Mahmood S, Islam F, Rahman S, Haque T, Islam S, et al. Relationship
[32] Halperin Kuhns VL, Woodward OM. Sex differences in urate handling. Int J Mol Sci
between serum uric acid and hypertension: a cross-sectional study in Bangladeshi
2020;21.
adults. Sci Rep 2019;9:9061.
[33] Arosio B, Corbi G, Davinelli S, Giordano V, Liccardo D, Rapacciuolo A, et al. Sex
[10] Haque T, Rahman S, Islam S, Molla NH, Ali N. Assessment of the relationship
differences in cardiovascular diseases: a matter of estrogens, ceramides, and
between serum uric acid and glucose levels in healthy, prediabetic and diabetic
sphingosine 1-phosphate. Int J Mol Sci 2022;23.
individuals. Diabetol Metab Syndr 2019;11:49.
[34] Wang H, Zu Q, Tang H, Lu M, Chen R, Yang Z. Long-term cardiovascular outcomes
[11] Ali N, Miah R, Hasan M, Barman Z, Mou AD, Hafsa JM, et al. Association between
of biodegradable polymer drug eluting stents in patients with diabetes versus non-
serum uric acid and metabolic syndrome: a cross-sectional study in Bangladeshi
diabetes mellitus: a meta-analysis. Cardiovasc Diabetol 2023;22:228.
adults. Sci Rep 2020;10:7841.
[35] Ciarambino T, Crispino P, Giordano M. Hyperuricemia and endothelial function: is
[12] Borghi C, Agabiti-Rosei E, Johnson RJ, Kielstein JT, Lurbe E, Mancia G, et al.
it a simple association or do gender differences play a role in this binomial?
Hyperuricaemia and gout in cardiovascular, metabolic and kidney disease. Eur J
Biomedicines 2022;10.
Intern Med 2020;80:1–11.
[36] Kuo CF, See LC, Yu KH, Chou IJ, Chiou MJ, Luo SF. Significance of serum uric acid
[13] Shahin L, Patel KM, Heydari MK, Kesselman MM. Hyperuricemia and
levels on the risk of all-cause and cardiovascular mortality. Rheumatology (Oxford)
Cardiovascular Risk Cureus 2021;13:e14855.
2013;52:127–34.
[14] Molla NH, Kathak RR, Sumon AH, Barman Z, Mou AD, Hasan A, et al. Assessment
[37] Mannarino MR, Pirro M, Gigante B, Savonen K, Kurl S, Giral P, et al. Association
of the relationship between serum uric acid levels and liver enzymes activity in
between uric acid, carotid intima-media thickness, and cardiovascular events:
Bangladeshi adults. Sci Rep 2021;11:20114.
prospective results from the IMPROVE Study. J Am Heart Assoc 2021;10:e020419.
[15] Barman Z, Hasan M, Miah R, Mou AD, Hafsa JM, Trisha AD, et al. Association
[38] Cang Y, Xu S, Zhang J, Ju J, Chen Z, Wang K, et al. Serum uric acid revealed a U-
between hyperuricemia and chronic kidney disease: a cross-sectional study in
shaped relationship with all-cause mortality and cardiovascular mortality in high
Bangladeshi adults. BMC Endocr Disord 2023;23:45.
atherosclerosis risk patients: the ASSURE study. Front Cardiovasc Med 2021;8:
[16] Burnier M. Gout and hyperuricaemia: modifiable cardiovascular risk factors? Front
641513.
Cardiovasc Med 2023;10:1190069.
[39] Choi HK, Ford ES. Haemoglobin A1c, fasting glucose, serum C-peptide and insulin
[17] Culleton BF, Larson MG, Kannel WB, Levy D. Serum uric acid and risk for
resistance in relation to serum uric acid levels–the Third National Health and
cardiovascular disease and death: the Framingham Heart Study. Ann Intern Med
Nutrition Examination Survey. Rheumatology (Oxford) 2008;47:713–7.
1999;131:7–13.
[18] Li J, Huang JY, Liu L, Lo K, Sun S, Chen CL, et al. Relationship between serum uric
acid level and all-cause and cardiovascular mortality in population with obesity.
Postgrad Med J 2020;96:660–5.