Background
As the end product of the metabolic breakdown of purine nucleotides [
1], an increasing amount of research has indicated that the role of uric acid (UA) in metabolic syndrome has changed from innocent bystander to central player [
2] and UA plays a key role in the development of hypertension [
3,
4], hyperglycemia [
5], hyperlipidemia [
6], and also obesity [
7]. Obesity, a multiple organ-system disease with underlying metabolic abnormalities, is a public health crisis and results in a huge economic burden [
8,
9]. A series of cross-sectional studies suggested a positive association between serum uric acid (SUA) and body mass index (BMI) [
10‐
13], which was demonstrated in a population-based longitudinal study [
7]. However, limited data is available in the Chinese population.
In addition, several laboratory studies have suggested discrimination in the SUA-BMI association among different BMI levels [
3,
14‐
16]. Our previous study also found that SUA and blood pressure (BP) might have a nonlinear, instead a simple linear, relationship [
3]. Therefore, in this study, we used data from a large cross-sectional study in China to better understand the relationship between SUA and BMI, and additionally explore whether other metabolic factors may affect the relationship between SUA and BMI.
Methods
Study population
This is a population-based cross-sectional study among people in the coastal region of China, from August 2004 to December 2014. Participants were excluded if one of the following criteria were met: (I) medication use to lower weight, serum lipids, uric acid, blood sugar or blood pressure; (II) history of liver, severe renal, or heart diseases. A total of 14,039 participants (6098 males and 7941 females) met the criteria for enrollment in this study. The study protocol was approved by the ethics committee of the Affiliated Hospital of Qingdao University and informed consent was obtained from each participant.
Data collection and measures
The participants’ demographic and lifestyle information were collected using a standard questionnaire by in-person interview, including current smoking status (smoker, never, past), alcohol consumption (never, moderate, heavy, past), and occupation type (light, moderate, heavy physical). BMI was calculated as weight (kg) per height squared (m2) and categorized as: under-weight (BMI < 18.5 kg/m2), normal weight (18.5 kg/m2 ≤ BMI < 24.9 kg/m2), overweight (25 kg/m2 ≤ BMI < 29.9 kg/m2), and obesity (BMI ≥ 30 kg/m2). For alcohol consumption in past 6 months, different from that used in the US and EU, heavy drinking was defined as equal as or greater than one time per week; and moderate drinking was defined as drinking at holiday and festival days, averagely one time per month. For occupation types, the “light physical” jobs referred to those with sedentary/desk job, such as official staffs, teachers and light physical houseworkers. Moderate physical work included students, gym teachers, and light physical farmworkers. And heavy physical works referred to the porter (workers who employed to help carry, ship or move luggage or other loads), construction workers, athletes, and so on. BP was measured with a standard mercury sphygmomanometer, and subjects were required to rest for at least 15 min before BP measurement. All measurements of height, weight, and BP were carried out by the same group of seven professional physicians.
Elbow venous blood (5 mL) was extracted from all participants after fasting for at least 12 h. The fasting blood glucose (FBG), triglyceride (TG), total cholesterol (TC), creatinine, high/low-density lipoprotein cholesterol (H/LDL), and SUA levels of all blood specimens were examined by automatic Sysmex Chemix-180 biochemical analyzer (Nanchang Micare Medical Equipment Co., LTD, Jiangxi, China). Estimated glomerular filtration rate (eGFR) was calculated by the following formula: eGFR = 175 × (creatinine/88.4)−1.234 × age−0.179 × (0.79 for females). Hyperuricemia was diagnosed if SUA levels were higher than 420 µmol/L for men and postmenopausal women, and higher than 357 µmol/L for premenopausal women.
Statistical analyses
Previous studies suggested large differences in BMI and SUA among men and women; therefore, all analyses were separately applied to men and women. We calculated mean ± standard deviation (SD) and median (interquartile range) for frequency of participant characteristics, t-tests for normal distributions, Kruskal–Wallis tests for non-normal distributions, and Chi square tests were used to compare characteristic differences among men and women. We evaluated the possible linear and nonlinear relationships between SUA and BMI by multivariate linear regression models and two-piece piecewise regression models adjusted for age, current alcohol consumption status, current smoking status, occupation type, systolic blood pressure (SBP), diastolic blood pressure (DBP), FBS (log transformed), eGFR, low density lipoprotein (LDL), and TC, among men and women. We further conducted stratified and interaction analyses to explore the potential modifier and interaction effects on the SUA-BMI association. Before these, covariate screening was also performed among all variables included in Table
1 using univariate analysis. 93.7% of the participants enrolled in our study had complete data, and only those participants with complete data were included in the analysis.
Table 1Characteristics of 7941 women and 6098 men included in this study
n | 6098 | 7941 | 14,039 | |
Hyperuricemia | 17.96% | 4.04% | 10.18% | < 0.001 |
Age (years) | 47.70 ± 14.28 | 48.65 ± 13.65 | 48.24 ± 13.94 | < 0.001 |
Body mass index (kg/m2) | 24.54 ± 3.51 | 24.58 ± 3.74 | 24.56 ± 3.64 | 0.535 |
Occupation types | < 0.001 |
Light physical | 3266 (53.56) | 5606 (70.60) | 8872 (63.20) | |
Moderate physical | 2123 (34.81) | 1901 (23.94) | 4024 (28.66) | |
Heavy physical | 709 (11.63) | 434 (5.47) | 1143 (8.14) | |
Serum uric acid (μmol/L) | 349.62 ± 85.30 | 267.37 ± 70.79 | 303.09 ± 87.50 | < 0.001 |
Smoking status in last 6 months (%) | | | | < 0.001 |
Smoking | 2804 (45.98) | 7769 (97.83) | 10,573 (75.31) | |
Never | 2972 (48.74) | 152 (1.91) | 3124 (22.25) | |
Past | 322 (5.28) | 20 (0.25) | 342 (2.44) | |
Alcohol drinking status in last 6 months (%) | < 0.001 |
Never | 2659 (43.60) | 7698 (96.94) | 10,357(73.77) | |
Moderate | 2204 (36.14) | 180 (2.27) | 2384 (16.98) | |
Heavy | 1161 (19.04) | 62 (0.78) | 1223 (8.71) | |
Quit | 74 (1.21) | 1 (0.01) | 75 (0.53) | |
Systolic blood pressure (mmHg) | 131.76 ± 19.38 | 129.73 ± 22.13 | 130.61 ± 21.01 | < 0.001 |
Diastolic blood pressure (mmHg) | 85.68 ± 12.07 | 82.49 ± 11.82 | 83.87 ± 12.03 | < 0.001 |
Fasting blood glucose (mmol/L) | 5.10 (4.42-5.70) | 5.08 (4.50-5.68) | 5.09 (4.48–5.70) | 0.362 |
Triglyceride (mmol/L) | 1.20 (0.82–1.90) | 1.10 (0.74–1.66) | 1.14 (0.77–1.76) | < 0.001 |
High-density lipoprotein cholesterol (mmol/L) | 1.31 ± 0.42 | 1.40 ± 0.38 | 1.36 ± 0.40 | < 0.001 |
Low-density lipoprotein cholesterol (mmol/L) | 2.76 ± 0.82 | 2.74 ± 0.84 | 2.75 ± 0.83 | 0.330 |
Total cholesterol (mmol/L) | 4.86 ± 1.04 | 4.88 ± 1.09 | 4.87 ± 1.07 | 0.277 |
Creatinine (μmol/L) | 82.98 ± 21.77 | 70.11 ± 21.07 | 75.70 ± 22.31 | < 0.001 |
Estimated glomerular filtration rate (mL/min/1.73 m2) | 90.54 (76.98–126.17) | 85.45 (73.57–127.39) | 89.15 (75.28–136.30) | < 0.001 |
All statistical analyses were performed using Empower Stats software (X&Y Solutions, Inc., Boston, USA). P < 0.05 was considered statistically significant.
Discussion
Positive association between SUA and BMI was found among men and women. Using the two-piece piece-wise regression model, we found a U-shaped SUA-BMI relationship for both men and women. Although a positive association was maintained when the BMI was higher than 20 kg/m2, a negative association was found when the BMI was lower than 20 kg/m2. Our results suggested that SUA level might be a good index for BMI, depending on whether or not the participants were underweight.
Accumulating evidence suggests that elevated SUA levels are common comorbidities of obesity and are accompanied by gradually increasing BMI [
17]. Dr. Rathmann and colleagues used the data from the Coronary Artery Risk Development in Young Adults (CARDIA) Study (1249 male and 1362 female black and white subjects aged 17–35 years with a 10-year follow-up) to evaluate changes in SUA with changes in components of the metabolic syndrome in young adults [
18]. They found that BMI had a significant independent linear association with UA in all race-sex-groups. Another Chinese study of 2962 patients with type 2 diabetes also observed a linear association between the prevalence of obesity and increasing SUA levels [
19]. However, these studies only focused on the linear relationship between SUA and BMI, while our results indicated a strong U-shaped relationship for both men and women. Previous mechanism studies suggested that chronic inflammation contributed to the pathogenesis of obesity [
20,
21]; the infiltration and accumulation of macrophages in adipose tissue was demonstrated to be associated with increased tumor necrosis factor-α and interleukin 6 secretion [
22,
23]. Similarly, macrophages also play an important role in the inflammation associated with gout/hyperuricemia [
24‐
27]. Of note, physiological concentrations of SUA displayed anti-inflammatory effects both in vitro and in vivo [
14] and thus might partly explain the U-shaped relationship between SUA and BMI.
Meanwhile, our results also showed a similar U-shaped SUA-BMI association and nadir point of BMI, around 20 kg/m
2, for men and women. The strength of the association was significantly stronger for men than women. A prospective study on 3857 Chinese participants with normal metabolic function [
28] found that baseline SUA levels and metabolic syndrome during a mean follow-up of 5.41 years were more closely related in women than in men. A study among Mexican-origin infants, youth and adults [
10] also found a stronger association between salivary UA levels and BMI for females than males. Although the mechanisms underlying these observations are still unclear, several studies indicated that differences in sex hormone levels may partially explain the effects. An analysis study on the data from 7662 women aged 20 years and older in the Third National Health and Nutrition Examination Survey (1988–1994) found that menopause was independently associated with higher SUA levels, and postmenopausal hormone use was associated with lower UA levels among postmenopausal women [
29]. A study on 128 obese patients who underwent laparoscopic sleeve gastrectomy found that increased estradiol levels, decreased total testosterone levels, and increased estradiol/total testosterone ratios in obese female patients 6 months post-surgery might be related to SUA improvement [
30]. Additionally, mechanism-related evidence also suggested that estradiol can affect fat metabolism and distribution in women.
In addition, we found an interaction effect of TG levels on the association between BMI and SUA; BMI had a stronger effect on SUA at higher TG levels. Similar interdependent relationships have been reported among SUA, C-reactive protein and interleukin 10 levels, which related to early hepatic damage [
25]. As the hepatic manifestation of obesity and metabolic dysfunction, nonalcoholic fatty liver disease could result in hepatic damage. However, these assumptions need further investigation in additional well-designed studies.
Our study has several limitations. First, due to the inherent nature of cross-sectional designs, our results could not make a causality conclusion. Second, the participants were limited to coastal areas and had special diet features with more marine food products, such as sea-fish, shrimp, and shellfishes. Therefore the extension of these conclusions should be prudent. Furthermore, additional large-scale studies with representative populations are warranted to validate our conclusions.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.