Association between the serum uric acid-to-high-density lipoprotein cholesterol ratio and gynecological cancer history

Serum uric acid (sUA) and high-density lipoprotein cholesterol (HDL-C) reflect complementary aspects of metabolic and inflammatory status. The uric acid-to-HDL-C ratio (UHR) has been investigated in cardiometabolic disorders, but evidence regarding gynecological cancer remains limited. We examined the association between UHR and a self-reported history of cervical, ovarian, or uterine cancer among U.S. women. We analyzed cross-sectional data from 7 NHANES cycles (2003-2004 through 2015-2016). The complete-case sample comprised 13,775 women, including 369 who reported a history of cervical, ovarian, or uterine cancer. UHR was calculated as [sUA (mg/dL)/HDL-C (mg/dL)] × 100. Survey-weighted logistic regression was used to estimate prevalence odds ratios (ORs) for UHR modeled continuously and by quartiles. Model 2 adjusted for age and race/ethnicity, and model 3 additionally adjusted for educational attainment, poverty-income ratio, alcohol use, smoking status, body mass index, oral contraceptive use, pregnancy history, and marital status. Restricted cubic spline (RCS), exploratory subgroup, receiver operating characteristic (ROC), and E-value sensitivity analyses were also performed. Women with a self-reported history of gynecological cancer had descriptively higher UHR values than those without such a history. In the fully adjusted model, each 1-percentage-point increase in UHR was associated with 3% higher prevalence odds (OR 1.03, 95% CI: 1.00-1.05; P = .03). However, none of the estimates for Q2-Q4 differed significantly from that for Q1 in model 3, although the P value for trend was < .05. RCS analysis provided no evidence of nonlinearity. The AUCs were 0.560 for UHR, 0.538 for sUA, and 0.554 for HDL-C, indicating poor discrimination. The E-value was 1.20 for the point estimate and 1.00 for the confidence-limit closest to the null, indicating limited robustness to unmeasured confounding. Higher UHR was weakly associated with greater prevalence odds of a self-reported history of gynecological cancer. The small effect size, nonsignificant fully adjusted quartile estimates, poor discriminatory performance, and susceptibility to residual confounding do not support the use of UHR for screening, early detection, or individual risk prediction. Prospective studies with validated outcomes and more comprehensive control of confounding are warranted.