Introduction

Serum uric acid (UA) is the final enzymatic product when the body breaks down purine1. Increased production or decreased excretion of UA causes hyperuricemia2. Previous studies have indicated that hyperuricemia is associated with cardiovascular diseases3 and metabolic diseases such as diabetes4, hypertension5 and dyslipidemia6. In past decades, the prevalence of hyperuricemia has increased to 21% and 13% in the US and Chinese general populations, respectively. Although this trend may be related to the increasing prevalence of adiposity and hypertension1, 4, environmental factors cannot be ignored.

Cadmium is a toxic metal with negative effects on health7. Occupational exposure is mainly from industrial processes. Smoking tobacco and contaminated food such as vegetables and rice are the main sources of general cadmium exposure7. Blood cadmium (CdB) levels vary by region, age and ethnicity7. Previous studies have confirmed the pathogenic role of cadmium exposure in renal damage7, bone destruction8 and cancer9, 10. Recent research has focused on the role of cadmium as an important environmental endocrine disruptor11. Epidemiological studies have linked cadmium exposure to metabolic diseases such as diabetes12, obesity13 and thyroid disease14, although the results have not been consistent15, 16. We also found a relationship between cadmium exposure and prediabetes in our previous work17. However, the relationship between cadmium exposure and hyperuricemia remains unknown.

Due to its long biological half-life18, cadmium mainly accumulates in the kidney and liver of human bodies, which may lead to elevated plasma uric acid levels according to several animal studies19, 20. Furthermore, as mentioned above, cadmium exposure is associated with metabolic diseases and thus may prompt the occurrence of hyperuricemia and gout21, 22. Epidemiological evidence of a relationship between cadmium exposure and hyperuricemia is scarce. A cross-sectional study from the National Health and Nutrition Examination Survey (NHANES) showed no relationship between them2. Nevertheless, our previous study revealed that the CdB level (median of 1.70 μg/L) was much higher than that reported in other countries17, and approximately 17% of subjects still had a CdB level higher than 5.0 µg/L. No study has ever explored this association in the Chinese population at the current CdB level, which differs from the level in the US. Hence, using data from a population-based investigation called the Survey on Prevalence in East China for Metabolic Diseases and Risk Factors (SPECT-China) in 2014, we aimed to explore the relationships between the CdB and serum UA levels and hyperuricemia in the general Chinese population.

Results

Characteristics of participants by hyperuricemia status

The characteristics of the study population, categorized by sex and hyperuricemia status, are provided in Table 1. Participants with hyperuricemia were more likely to have comorbid conditions such as obesity, hypertension, dyslipidemia and reduced renal function in both genders (P < 0.05). The median CdB level was 2.40 (0.68–4.61) μg/L higher in men with hyperuricemia than in men without hyperuricemia (P < 0.05), but the CdB levels in women showed no significant difference between individuals with and without hyperuricemia. Additionally, the median blood lead (PbB) levels were comparable between different serum UA levels.

Table 1 Characteristics of the participants categorized by hyperuricemia status.
Full size table

Association of serum UA level with CdB by linear regression

Linear regression modeling of the data showed that a higher CdB level was associated with a higher serum UA concentration (B = 2.963, p < 0.05) in men after adjusting for PbB (Table 2). This positive correlation remained even after the data were adjusted for estimated glomerular filtration rate (eGFR), current smoking status, diabetes, dyslipidemia, hypertension and body mass index (BMI) (B = 2.718, p < 0.05). After we excluded participants with renal impairment (eGFR ≤ 60 mL/min per 1.73 m2) and smokers, the serum UA level remained positively associated with CdB (B = 2.595, p < 0.05 and B = 2.771, p < 0.05, separately). However, no relation between the CdB and serum UA levels was observed in women in either the crude or the fully adjusted model. Furthermore, PbB was analyzed as an independent variable, and we found no correlation between PbB and serum UA levels in either gender.

Table 2 Association of CdB level (independent variable) with serum urate level (dependent variable).
Full size table

Association of CdB quartiles with hyperuricemia by logistic regression analyses

In multivariate-adjusted logistic regression analyses (Table 3), the CdB levels were divided by quartile (Q1: ≤0.60; Q2: 0.61–2.09; Q3: 2.10–4.29; Q4: ≥4.30). The Q1 of CdB was used as the reference. Male participants in the highest quartile of CdB had an OR of 1.50 (95% CI, 1.00 to 2.24) for hyperuricemia after adjusting for age, smoking and PbB (P for trend < 0.05). After adjusting for eGFR, current smoking status, PbB, diabetes, dyslipidemia, hypertension and BMI, the ORs of Q3 and Q4 CdB for hyperuricemia were 1.82 (95% CI, 1.18, 2.79) and 1.61 (95% CI, 1.04, 2.49), respectively (P for trend < 0.05). Higher ORs for the CdB levels in Q3 (OR = 1.99, 95% CI, 1.26, 3.15) and Q4 (OR = 1.77, 95% CI, 1.11, 2.80) were observed in participants with relatively normal renal function (eGFR > 60 mL/min per 1.73 m2, P for trend < 0.01). After we excluded smoking participants, a marginal significance for CdB as a risk factor for hyperuricemia remained (P for trend = 0.08). In women, the CdB levels were still not related to hyperuricemia (P for trend > 0.05).

Table 3 Association of blood cadmium level quartiles with hyperuricemia.
Full size table

Discussion

We explored the association between CdB and UA in Chinese adults. Our study revealed that CdB was positively associated with serum UA levels and hyperuricemia in Chinese men but not in women. This association was independent of PbB, eGFR, current smoking status, diabetes, dyslipidemia, hypertension and BMI. Furthermore, in male participants with relatively normal renal function (eGFR > 60 mL/min per 1.73 m2), a positive relationship between CdB and hyperuricemia remained.

Cadmium exposure has been linked to numerous human health problems11. Cadmium has been found to target the kidneys and induce proximal tubular reabsorptive dysfunction7. Prolonged exposure to high cadmium levels has given rise to osteomalacia as well as osteoporosis7. In particular, various studies have demonstrated the possible role of cadmium as an endocrine disruptor11, 14, 15. Cadmium can accumulate in the thyroid gland. Colloid cystic goiter, diffuse parafollicular cells, nodular hyperplasia and hypertrophy are often found in chronic cadmium toxicity14. Both animal studies and epidemiology studies have revealed that cadmium alters various blood sex hormone levels, such as luteinizing hormone, progesterone and testosterone11, 23. Moreover, cadmium can exert an estrogenic effect both in vivo and in vitro 24. Cadmium has been found to accumulate in the pancreas and exhibit detrimental effects on β cell function25. Both NHANES and our previous study showed that CdB level was associated with prediabetes12, 17.

The CdB levels of our participants were higher than in those in developed countries18, 26,27,28, which may be attributed to the economic boom and industrialization29. Industrial uses have led to the widespread dispersion of Cd at trace levels into the air, water, and soil and thus into foods18. Atmospheric Cd emissions from non-ferrous metal smelting and coal combustion in China increased by approximately 4.6 fold from 1990 to 201029, 30. Another explanation is dietary habits. As in other Asian countries such as Bangladeshi and Korea26, 31, our staple foods are rice and vegetables, which are more likely to be contaminated by cadmium pollution7. Furthermore, participants living in areas with low economic status had higher CdB than participants in high-economic-status areas17. Industrial factories prefer to build sites in low-economic-status areas because of the low prices of land and labor. Poor infrastructure construction and environmental supervision systems combined with a lack of water-quality monitoring together led to water cadmium contamination17.

Little is known about the association of cadmium exposure with UA. The NHANES data for 2005–2008 revealed no relationship between cadmium and gout in the USA2. However, the CdB levels in Americans were much lower than those in Chinese adults17. Previous animal studies have established models of renal toxicity with decreased eGFR upon cadmium administration, which resulted in elevated serum UA levels32, 33. Nevertheless, in our research, CdB was positively correlated with hyperuricemia after adjustment for eGFR and in participants with an eGFR > 60 mL/min per 1.73 m2, suggesting other mechanisms beyond a decreased eGFR.

UA is primarily produced in the liver by xanthine oxidoreductase34 and then undergoes glomerular filtration, tubular reabsorption and excretion by the kidneys35. The excretion of UA consists of a basolateral uptake step mediated by an organic anion transporter36, followed by an efflux step mediated by multidrug resistance protein 4 and the urate transporter37. Cadmium-related renal damage begins with proximal renal tubular injury38 before glomerular injury. Tubular organic anion uptake transporters may be a target for cadmium19, 33 because sub-chronic cadmium intoxication results in a loss of basolateral invaginations and the down-regulation of organic anion transporters and organic cation transporters, which may lead to decreased urate secretion from the tubular cells. Cadmium toxicity may lead to impaired p-aminohippurate excretion due to a loss of organic anion carriers in the proximal tubular basolateral membranes20. Therefore, we hypothesized that the early renal damage by cadmium exposure might lead to a defect in urate excretion and give rise to hyperuricemia.

Oxidative stress is among the important mechanisms of cadmium toxicity, and the liver is a critical target organ39. There is an increased conversion of xanthine oxidoreductase from xanthene dehydrogenase to xanthine oxidase in the cadmium-treated liver40. The transition from purine to UA, mediated by xanthine oxidase, leads to the production of reactive oxygen species, which may be accompanied by increased UA production. Furthermore, previous studies have suggested that serum UA is an antioxidant41. Hence, elevated serum UA may be a protective mechanism against oxidative stress from cadmium exposure.

The gender-specific association between CdB and UA level is inconclusive. Sex hormones may be involved. CdB was found in a previous study to negatively correlate with total testosterone and sex hormone binding globulin in Chinese men23. Conversely, the data from NHANES 2011–2012 show significantly positive associations between CdB and serum testosterone in men42. Moreover, estrogen-induced increases in the fractional excretion of UA were associated with lower levels of UA in male-to-female transsexuals43. A previous study showed that women and men differed in their pathogenic factors and treatment monitoring because female patients had greater co-morbidities and received the appropriate treatment more often44. Knowledge on this gender-specific association is thus rather limited.

Cadmium and lead (Pb) are two toxic metals that are widely distributed in the environment. They share similar population exposure routes45, 46. Concurrent exposure to both metals is very common46, 47. Epidemiological evidence has shown that CdB is positively related to PbB17, 48 and that the two metals have interactive effects in certain diseases45, 49. Lead toxicity (>80 μg/dL) is associated with hyperuricemia and gout47, 50, 51. Moreover, there is still a link between relatively lower PbB and hyperuricemia2, 52. Thus, we regarded PbB as an important confounding factor. Moreover, we evaluated the relationship between PbB and UA levels in our participants, but there was no significant relationship in either men or women.

This study is the first exploration of the relationship between CdB and hyperuricemia in different genders in the Chinese population. Homogeneity and strict quality control were guaranteed because the same trained staff was used. Furthermore, we considered PbB to be a confounding factor when exploring the association between CdB and UA levels.

There are some limitations of this study. First, using a cystatin C-based formula to adjust for the GFR estimates is required in healthy populations with normal renal function, which was not available to us, but the CKD-EPI equation applied in our study was confirmed to be more accurate than the Modification of Diet in Renal Disease Study equation, particularly for censoring numerical estimates greater than 60 mL/min per 1.73 m2  53. Second, we used the blood cadmium levels rather than urinary cadmium. Urinary cadmium reflects lifetime cadmium exposure, but for relatively low cadmium exposure levels, blood cadmium levels may be more appropriate38. It would have been ideal if we could detect both. Third, this study did not include information on food intake. A high serum UA level is usually associated with an intake of large amounts of food that is high in purines2. It is reasonable that the CdB levels are parallel with the serum UA levels in participants with large daily food intakes. Furthermore, we could not determine the causal relationship between CdB and hyperuricemia in this cross-sectional study.

In conclusion, CdB was positively associated with serum UA levels and hyperuricemia in Chinese men but not in women. This study indicated that cadmium exposure may confer a risk for hyperuricemia, which was not attributed solely to cadmium toxicity-induced renal dysfunction. However, in cases of relatively normal renal function, the CdB level was still positively related to serum UA. Further study is needed to demonstrate causality and elucidate the underlying mechanisms. In addition, efforts to reduce cadmium exposure in adults are warranted.

Methods

Study population

Our data (n = 6899) were from the SPECT-China study54, 55. The sampling method was described in detail in our previous study23. A total of 2996 subjects were enrolled in our final study after excluding participants with missing values for UA (n = 3429) and CdB (n = 474). Before the data collection, written informed consent was provided by all participants. All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008. The study protocol was approved by the Ethics Committee of Shanghai Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine.

Measurements

The questionnaires about demographic characteristics, medical history and lifestyle risk factors and anthropometric data were constructed by the same trained staff as previously described54, 55. Body weight, height and the calculation of BMI were calculated consistently with the previous study23. Waist circumference and blood pressure were measured by strict adherence to the standard procedure21. Current smoking was defined as having smoked at least 100 cigarettes in one’s lifetime and currently smoking cigarettes17.

Venous blood samples were drawn, processed and shipped as previously described54, 55. Serum UA levels were measured using the uricase method with a Beckman Coulter AU 680 (Germany). The coefficient of variation was between 1.2% and 2.7%. Serum creatinine (Scr) was measured using a kinetic-rate Jaffe method with a Beckman Coulter AU 680 (Germany), and we converted the Scr levels to the estimated glomerular filtration rate (eGFR) according to the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation: (1) for a woman with Scr ≤0.7 mg/dL, eGFR = 144 × (Scr/0.7)−0.329 × (0.993)age; (2) for a woman with Scr >0.7 mg/dL, eGFR = 144 × (Scr/0.7)−1.209 × (0.993)age; (3) for a man with Scr ≤0.9 mg/dL, eGFR = 141 × (Scr/0.9)−0.411 × (0.993)age; and (4) for a man with Scr >0.9 mg/dL, eGFR = 141 × (Scr/0.9)−1.209 × (0.993)age. The value of eGFR is reported in units of mL/min per 1.73 m2 of body surface area17.

Cadmium and lead levels in blood samples were tested using graphite furnace atomic absorption spectrometry17. Standard curves were established, and quality control materials were tested before the samples were measured. Two quality control personnel participated in the process control. Outliers were detected by duplicate runs. The detection limits for blood cadmium and lead were 0.01 µg/L and 0.1 µg/L, respectively. The inter-assay coefficient of variation for cadmium was 10%.

Fasting plasma glucose (FPG), total cholesterol (TC), triglycerides (TG), high-density lipoprotein (HDL), low-density lipoprotein (LDL), insulin and glycated hemoglobin (HbA1c) were assessed with the methods used previously17.

Definition of variables

Hyperuricemia was defined as a serum UA concentration ≥416.4 μmol/L and ≥356.9 μmol/L for men and women, respectively2. The definitions of overweight, obese, diabetic and hypertensive in this study have been previously described21. Dyslipidemia was defined as described previously56.

Statistical analyses

The IBM SPSS Statistics software, version 22 (IBM Corporation, Armonk, NY, USA), was used for data analysis. Analyses were performed separately for men and women due to major gender differences in serum UA concentrations. A P value < 0.05 for a two-tailed test indicated a significant difference. The specific statistical methods for continuous variables and categorical variables were described in detail in a previous study17.

The association of CdB (an independent variable) with serum UA levels (a dependent variable) was assessed by linear regression analysis. The results were expressed as unstandardized coefficients (B) and standard errors. The full model included PbB, eGFR (which incorporates age and serum creatinine level), current smoking status, diabetes, dyslipidemia, hypertension and BMI.

To consider the association of CdB with hyperuricemia, logistic regression analyses were used. CdB was divided into quartiles, with the first quartile representing the lowest levels and the fourth quartile the highest. The full model included eGFR (which incorporates age and serum creatinine levels), current smoking status, PbB, diabetes, dyslipidemia, hypertension and BMI. PbB, eGFR, and BMI were entered as continuous measures. Data were expressed as odds ratios (ORs) (95% confidence interval (CI).

Subgroup analyses

Because hyperuricemia is known to be associated with kidney dysfunction and the kidneys are the most important target organs for cadmium exposure, we performed subgroup analyses that excluded participants with an eGFR of 60 mL/min per 1.73 m2 or less2. Moreover, smokers are at high risk of cadmium exposure17, and previous studies have indicated an association between smoking and increased purine catabolism57. Thus, we performed another subgroup analysis excluding current smokers. The regressions were performed by the same strategy as in the above analyses.