Fig. 3.

Restricted cubic splines were utilized to evaluate the hypothesis of potential nonlinear relationships between NHHR and all-cause (A) and cardiovascular (B) mortality in participants with diabetes. The NHHRs of 2.55 for A and 2.50 for B were chosen as reference estimates for each hazard ratio (HR). Similarly, restricted cubic splines were utilized to evaluate the hypothesis of potential nonlinear relationships between NHHR and all-cause (C) and cardiovascular (D) mortality in participants with prediabetes. Using NHHR of 3.13 for C and 2.83 for D as reference estimates for each HR, the analysis was adjusted for sex, age, race, education, marital status, poverty income ratio, BMI, smoking status, drinking status, physical activity, hypertension, hyperlipidemia, antihyperglycemic treatment, lipid-lowering treatment, and antihypertensive treatment. The Cox proportional hazards model was used to fit the data, with solid lines representing the estimated values and shaded areas representing the corresponding 95% CIs. The black dashed line represents the location of the inflection point in the curve. BMI, body mass index; CVD, cardiovascular disease