Sex-Specific Differences in the Association of Metabolically Healthy Obesity With Hyperuricemia and a Network Perspective in Analyzing Factors Related to Hyperuricemia

Abstract

Background: Although obesity is a well-known risk factor for hyperuricemia, it remains unclear whether obese subjects with metabolically healthy status have a decreased the risk of hyperuricemia and whether sex modifies the association of metabolically healthy obesity (MHO) with hyperuricemia risk. We aimed to investigate the sex-specific association between MHO and other obesity phenotypes and hyperuricemia, and to use Bayesian networks to determine and visualize the interactions among hyperuricemia and its related factors.

Methods: This study was conducted using data from the China Health and Nutrition Survey 2009. Hyperuricemia was defined as serum uric acid ≥ 420 μmol/L in men and ≥ 360 μmol/L in women according to the guidelines. Body mass index (BMI) was used to define normal weight, overweight, and obese status in subjects, and metabolic health state was defined by the Adult Treatment Panel (ATP)-III and Visceral Adiposity Index (VAI) criteria, respectively. Subjects were categorized into six phenotypes according to their metabolic health and BMI level status.

Results: Of the 7,364 Chinese adult individuals included, the prevalence of hyperuricemia among MHO women was only 8.5% (95% CI 4.8 to 14.3%), but increased to 30.7% among MUO women, whereas the highest prevalence among men was found in the MUOW phenotype (39.4%, 95% CI 35.4 to 43.6%), compared to 15.4% for male subjects with MHO. After adjusting for confounders, the MHO phenotype was significantly associated with an increased risk of hyperuricemia compared with their MHNW counterparts in women (OR: 1.95, 95% CI: 1.02-3.74) whereas a significant association was not found in men (OR: 1.46, 95% CI: 0.8-2.68). A complex network structure was established by BNs and then used to find connections between hyperuricemia and its related factors, as well as their interrelationships. By using BN reasoning, the probability of having hyperuricemia was 0.076 among MHO men, while it reached 0.124 in MHO women.

Conclusions: In conclusion, our results demonstrated that the MHO phenotype was significantly associated with the risk of hyperuricemia only in women, not in men. This sex-specific differences in the association may suggest a favorable condition of MHO for Chinese men with respect to hyperuricemia risk, meanwhile more attention should be paid to the increased risk of hyperuricemia among MHO women.

Keywords: Bayesian network (BN); hyperuricemia; metabolic health; obesity; sex-specific.

Figure 1

Distribution of ATP-III criteria-based metabolic obesity phenotypes according to by sex. ATP-III, the Adult Treatment Panel-III; MHNW, metabolically healthy normal-weight; MHOW, metabolically healthy overweight; MHO, metabolically healthy obese; MUNO, metabolically unhealthy non-obese; MHO, metabolically healthy obese; MUNW, metabolically unhealthy normal-weight; MUOW, metabolically unhealthy overweight; and MUO, metabolically unhealthy obese.

Figure 2

Adjusted odds ratios (OR) and 95% confidence intervals (CI) for hyperuricemia associated with Adult Treatment Panel-III metabolic components, body mass index categories, and metabolic status by sex. Horizontal bars are 95% CIs. The adjusted OR was obtained from model 2 which was adjusted for age, urban/rural resident, smoking status, alcohol status, and metabolic health-obesity phenotypes, white blood cell, total cholesterol, LDL-C, hs-CRP, and diabetes.

Figure 3

The DAG underlying the Bayesian network learned from the covariates and hyperuricemia. (A) Averaged DAG with strength of arcs greater than 0.5; (B) Simplified DAG derived from the averaged DAG after retaining arcs with a strength greater than 0.85. Abbreviations: DAG, directed acyclic graph; MetS, metabolic syndrome; TC, total cholesterol; LDL-C, low-density lipoprotein cholesterol; WaistCir, waist circumference.