In this study, we found that TG was the best lipid index associated with hyperuricemia. Following TG, non-HDL-C was also significantly associated with hyperuricemia in both genders. Previous studies have examined the associations of lipid profiles such as TG, TC, LDL-C and HDL-C with hyperuricemia [13, 14]. Recently, non-HDL-C was demonstrated to be more powerful in predicting CVD events  and highlighted as a key secondary target of therapy in patients with an increased TG [15, 16], which implied that there may be a stronger association between non-HDL-C and CVD risks compared with other lipids. Some studies [17, 18] have showed that non-HDL-C is associated with other risk factors of CVD including hypertension, diabetes, and overweight. However, no study has compared non-HDL-C with other lipid profiles in relation to hyperuricemia to date. Therefore, in addition to the relationship between non-HDL-C and hyperuricemia, we compared the discriminatory values of these lipid profiles in relation to hyperuricemia. We found that TG and Non-HDL-C had larger AUCs related to hyperuricemia compared with LDL-C and TC in males and females.
Our findings are consistent with other studies [19, 20] that have reported a close link between serum TG and SUA in some groups. Conen and colleagues conducted a study in a population ongoing healthy examination, which demonstrated that TG was more strongly associated with SUA than HDL-C and TC . Russo and colleagues reported that TG had a more closely relation to SUA than other lipids even in the healthy people . The correlation coefficient between TG and SUA was higher than that between TC or LDL-C and SUA in the Verona Young men, even after adjustment for other potential covariates .
Non-HDL-C includes all lipoproteins that contain apolipoprotein B i.e. very low density lipoprotein, LDL-C, intermediate-density lipoprotein, chylomicrons and lipoprotein (a). Erdogan’s study showed that correlation coefficient of non-HDL-C to SUA (r = 0.382, P < 0.001) was slightly higher than that of TG (r = 0.379, P < 0.001) and obviously higher than those of TC, HDL-C and LDL-C in the healthy adults . It is worthwhile to note that non-HDL-C can be accurately calculated based on all routine lipid profiles, fasting or non-fasting, at no further expense to the patient. Importantly, high non-HDL-C is treatable and has already been assimilated into current treatment guidelines.
Some previous studies found that treatment for hypertriglyceridemia with fenofibrate or atorvastatin reduced SUA levels [25, 26]. In addition, an animal experiment study demonstrated that lowering uric acid with allopurinol prophylactically prevented fructose-induced hyperuricemia and hypertriglyceridemia . All these studies indicated that TG and SUA might share a common metabolic mechanism so that some drugs can reduce TG as well as SUA levels. Keenan et al.’s study also suggested that increased uric acid via increased systematic inflammation predict insulin resistance, followed by dyslipidemia and hepatic steatosis . In addition, some investigators tend to conclude that both hyperuricemia and dyslipidemia reflect the common lifestyles and dietary habits such as excessive alcohol consumption, high fat diet and physical inactivity [21, 29]. Reduction of TG intake in diet has been shown to increase uric acid elimination. However, the effect was minimal and transient . Thus, a combination of anti-hyperlipidemia and anti-hyperuricemia agents is a good option for the treatment of hyperuricemic patients with hypertriglyceridemia . Some remedy drugs with dual effects of lowering both TG and SUA levels are needed to control the combined syndrome with hyperuricemia and hyperlipidemia.
This study has some strength that should be mentioned. This is a large-scale survey research in a middle-aged and elderly Chinese population and we firstly compared various blood lipid profiles including non-HDL-C in relation to hyperuricemia. Since men had a higher prevalence of SUA compared with women in current study, we studied the relations of blood lipids to hyperuricemia in men and women separately. The important co-variables such as hepatorenal indexes were also measured and evaluated in our study [32, 33]. To some extent, these could control potential influence of some confounding factors including hepatorenal indexes on the relations of blood lipid profiles to hyperuricemia. However, the results of ROC analysis suggested that lipid indexes including TG and non-HDL-C were not good discriminate predictors for hyperuricemia in this study. Therefore, some additional factors besides lipid indexes should be considered in treating and preventing hyperuricemia, such as control of purine-rich foods, obesity or drinking .
There were several limitations in this study. First, this was a cross-sectional study and therefore a causal relationship between blood lipid profiles including non-HDL-C and risk of hyperuricemia could not be established. Secondly, data on other confounding factors such as diet, alcohol consumption and physical activity were not considered in this study, although the hepatorenal indexes can reflect accumulation of diet and alcohol consumption to some extent.
In summary, our results show that serum TG and non-HDL-C have stronger relations to hyperuricemia compared with other lipid indexes. Considering the increasing prevalence of hyperuricemia and its potential links with blood lipids, it is necessary to pay more attention to TG and non-HDL-C levels in these people with hyperuricemia.