Association between exposure to a mixture of organochlorine pesticides and hyperuricemia in U.S. adults: A comparison of four statistical models

Abstract

The association between the exposure of organochlorine pesticides (OCPs) and serum uric acid (UA) levels remained uncertain. In this study, to investigate the combined effects of OCP mixtures on hyperuricemia, we analyzed serum OCPs and UA levels in adults from the National Health and Nutrition Examination Survey (2005-2016). Four statistical models including weighted logistic regression, weighted quantile sum (WQS), quantile g-computation (QGC), and bayesian kernel machine regression (BKMR) were used to assess the relationship between mixed chemical exposures and hyperuricemia. Subgroup analyses were conducted to explore potential modifiers. Among 6,529 participants, the prevalence of hyperuricemia was 21.15%. Logistic regression revealed a significant association between both hexachlorobenzene (HCB) and trans-nonachlor and hyperuricemia in the fifth quintile (OR: 1.54, 95% CI: 1.08-2.19; OR: 1.58, 95% CI: 1.05-2.39, respectively), utilizing the first quintile as a reference. WQS and QGC analyses showed significant overall effects of OCPs on hyperuricemia, with an OR of 1.25 (95% CI: 1.09-1.44) and 1.20 (95% CI: 1.06-1.37), respectively. BKMR indicated a positive trend between mixed OCPs and hyperuricemia, with HCB having the largest weight in all three mixture analyses. Subgroup analyses revealed that females, individuals aged 50 years and above, and those with a low income were more vulnerable to mixed OCP exposure. These results highlight the urgent need to protect vulnerable populations from OCPs and to properly evaluate the health effects of multiple exposures on hyperuricemia using mutual validation approaches.

Keywords: Bayesian kernel machine regression; Hyperuricemia; NHANES; Organochlorine pesticide; Quantile g-computation; Weighted quantile sum.

Graphical abstract

Fig. 1

Association between five (organochlorine pesticide) exposures and hyperuricemia using weighted logistic regression. Adjusting age, sex, race, education level, BMI, PIR, serum cotinine concentration, alcohol consumption, physical activity, total frequency of seafood consumption, total calories taken, and medical history (diabetes, hypertension, CKD) (Model 3). BMI, body mass index; CI, confidence interval; CKD, chronic kidney disease; OR, odds ratio; PIR, poverty income ratio.

Fig. 2

WQS model regression index weights for hyperuricemia. Model was adjusted for age, sex, race, education level, BMI, PIR, serum cotinine concentration, alcohol consumption, physical activity, total frequency of seafood consumption, total calories taken, and medical history (diabetes, hypertension, chronic kidney disease). The dashed line represents the prespecified cut-off established to identify the most important OCPs of the mixture and set equal to the inverse of the number of the mixture components (0.20). eGFR, estimated glomerular filtration rate; HCB, hexachlorobenzene; β-HCH, β-hexachlorocyclohexane.

Fig. 3

The proportion of positive and negative effects of OCPs on hyperuricemia and the combined effects of mixed exposure to OCPs in the QGC model. (A) Proportion effects of OCPs on hyperuricemia. (B) Combined effects for hyperuricemia.

Fig. 4

Association between OCP's exposure and hyperuricemia by hierarchical BKMR model. h(z) can be interpreted as the relationship between OCPs and hyperuricemia. The model was adjusted for age, sex, race, BMI, education level, PIR, alcohol consumption, serum cotinine concentration, physical activity, total frequency of seafood consumption, total calories taken, and medical history (diabetes, hypertension, CKD). (A) Joint effects (95% CI) of mixed OCP's exposure on hyperuricemia estimated by comparing all OCPs at particular percentiles (from 25th to 75th) with all the OCPs at their 50th percentiles. (B) Univariate exposure-response functions (95% CI) between each OCP concentration and hyperuricemia with fixing the concentrations of other OCPs at their 50th percentiles.