Abstract
PURPOSE: The albumin-globulin ratio (AGR) influences the development of prostate cancer; however, the relationship between AGR and prostate-specific antigen (PSA) has not been reported. METHODS: This cross-sectional investigation used comprehensive AGR versus PSA data from men with 40 years of age and older, who participated in the National Health and Nutrition Examination Survey (NHANES) from 2003 to 2010, spanning 4 investigation cycles, as only these cycles contained complete PSA data. To evaluate the nonlinear relationship between the ARG and PSA level, a regression utilizing smoothed curve fitting (penalized spline approach) and a generalized additive model (GAM) were employed. A two-segment linear regression model was used to conduct threshold effect evaluations. Lastly, subgroup analyses were carried out along with interaction tests. RESULTS: This study included 5,376 subjects, whose total serum PSA (mean ± standard deviation) was 1.83 ± 3.34, and its level decreased roughly with increasing quartiles of AGR. In the fully-adjusted model, AGR was negatively correlated with the likelihood of PSA, and this relationship persisted across subgroups (trend > 0.05). The PSA was characterized by an "L"-shaped curve with an inflection point. On the left side of the inflection point (K = 1.32), there was a negative relationship between AGR and PSA. CONCLUSION: In the United States, among men over 40 years of age without prostate diseases, AGR demonstrated a nonlinear relationship with PSA, negatively correlating when AGR was below 1.32. CLINICAL TRIAL NUMBER: Not applicable.