Baseline clinical characteristics

After exclusion, a total of 1593 patients with HF from the NHANES were enrolled in our final analysis. The baseline demographic, clinical and biochemical characteristics of both men and women stratified by gender-specific SUA tertiles were illustrated in Tables 1 and 2. The participants of our investigation included 897 (56.3%) men and 696 (43.7%) women. For the demographic characteristics, the age distribution by SUA tertiles varied between men and women patients. The positive correlation of SUA tertiles and age was only observed in women (p < 0.001) rather in men (p = 0.326). Furthermore, the proportion of patients with diabetes increased with SUA tertiles only in women (p = 0.005) but not in men (p = 0.681), while the proportion of patients with hypertension increased with SUA tertiles in both genders. For both men and women, a higher SUA tertile was associated with increased BMI. The gender differences were also observed in laboratory tests. Serum HbA1c, glucose, and potassium increased with SUA tertiles only in women (HbA1c, p < 0.001; serum glucose, p = 0.001; and serum potassium, p < 0.001, respectively) but not in men. In contrast, higher tertiles of SUA correlated with decreased hemoglobin and decreased eGFR in both male and female patients (p < 0.05, both). For the medication, the prescription rate of ACEI/ARB, betablockers, MRA, and diuretics were higher in patients with higher SUA tertiles, for both male and female patients.

Table 1 Baseline demographic and clinical characteristics in menTable 2 Baseline demographic and clinical characteristics in womenSUA levels and clinical outcomes in different genders

As shown by histogram (Fig. 1), the median level of SUA in women was significantly lower than that in men (5.9 mg/dL, IQR 4.8-7.1 mg/dL vs. 6.6 mg/dL, IQR 5.4-7.9 mg/dL, p < 0.001). During a median follow-up of 127months (95% CI 120–134 months), there were 853 all-cause deaths (493 events in men, 360 events in women) and 361 cardiovascular deaths (206 events in men, 155 events in women). There was no significantly gender difference regarding to all-cause and cardiovascular mortality. The occurrence of all-cause death (55.0% vs. 51.7%, p = 0.199) and cardiovascular death (23.0% vs. 22.3%, p = 0.742) had no statistical difference between men and women. When analysis of long-term outcomes according to SUA tertiles and gender, the all-cause mortality was higher in patients with a higher SUA tertile in both genders (men, p = 0.001; women, p = 0.031). While the positive association between SUA and cardiovascular death was only observed in men (p = 0.001), but not in women (p = 0.691) (Table 3).

Table 3 Proportion of all-cause and cardiovascular events in patients according to gender and SUA tertilesFig. 1

Histogram of SUA in male and female HF patients

Role of SUA in prediction of long-term outcomes in different genders

Kaplan-Meier analysis was conducted to explore the gender differences in the association between SUA and the long-term prognosis in patients with HF. There was no significantly gender difference for SUA in the prediction of long-term all-cause mortality. In the whole population and both genders, individuals with SUA in tertile 3 had significantly higher all-cause mortality than those with SUA in tertile 1–2 (Log-rank p < 0.010, both, Fig. 2A-C). Furthermore, for patients with SUA in tertile 1 and 2, SUA showed poor discrimination ability for all-cause mortality in both men (Log-rank p = 0.898) and women (Log-rank p = 0.065). When it comes to cardiovascular mortality, gender difference emerged. As shown in Fig. 2D-F, male patients with SUA in tertile 3 was associated with highest cumulative incidence of cardiovascular deaths (Log-rank p < 0.001). While for female patients, SUA performed poorly in prediction of cardiovascular mortality (Log-rank p = 0.150).

Fig. 2

Kaplan-Meier curves for all-cause mortality and cardiovascular mortality according to gender-specific SUA tertiles

The prognostic value of SUA on long-term outcomes of CHF was also analyzed by Cox regression analysis. On univariable Cox regression analysis, SUA was significantly associated with all-cause mortality in the whole population (HR 1.15, 95% CI 1.11-1.19, p < 0.001) and both genders (HR 1.14, 95% CI 1.09-1.20, p < 0.001, in men; HR 1.16, 95% CI 1.10-1.23, p < 0.001, in women) (Table 4). However, after adjusted by covariates via multivariable Cox regression analysis, SUA (1 mg/dL) was a significant predictor of overall mortality only in men (HR 1.11, 95% CI 1.05-1.18, p < 0.001) but not in women (HR 1.05, 95% CI 0.98-1.12, p = 0.186).

Table 4 Association of SUA and all-cause mortality via Cox regression analysis

For cardiovascular mortality, similar gender difference was also observed. On crude Cox analysis, SUA was an importantly prognostic factor for the incidence of cardiovascular deaths in overall patients (HR 1.19, 95% CI 1.13-1.25, p < 0.001) and both genders (HR 1.21, 95% CI 1.13-1.29, p < 0.001, in men; HR 1.15, 95% CI 1.06-1.26, p = 0.001, in women) (Table 5). After adjusted by multivariate Cox analysis, the association of SUA and cardiovascular mortality was only detected in men (HR 1.18, 95% CI 1.09-1.28, p < 0.001) but not in women (HR 1.01, 95% CI 0.91-1.12, p = 0.902).

Table 5 Association of SUA and cardiovascular mortality via Cox regression analysisRCS and threshold analysis

As shown in Fig. 3, RCS analysis of fully adjusted model revealed that SUA was associated with all-cause mortality and cardiovascular mortality in the whole population (p < 0.001, both). Furthermore, SUA had a non-linear relationship with all-cause mortality (p for nonlinearity = 0.006) as well as cardiovascular mortality for over-all population (p for nonlinearity = 0.003). When analyzed according to gender, the nonlinear relationship between SUA and all-cause mortality was only observed in male patients (p < 0.001, p for nonlinearity = 0.030), but not in female patients. For male patients, the risk of all-cause death showed an ascending trend as SUA levels increased, especially when SUA was higher than 7.67 mg/dL., Furthermore, there was a linear positive relationshipbetween SUA with cardiovascular mortality (p for nonlinearity = 0.123) for male patients. In contrast, no statistically significant relationship between SUA and all-cause and cardiovascular mortality was detected in female patients (p = 0.101, for all-cause mortality; p = 0.139, for cardiovascular mortality).

Fig. 3

Dose-response relationship between SUA and all-cause mortality and cardiovascular mortality stratified by gender. (For all-cause mortality, covariates including, age, race, diabetes, stroke, drinking history, BMI, diastolic BP, WBC, hemoglobin, albumin, total cholesterol, eGFR, serum potassium, serum sodium, serum chloride, prescription of beta-blockers, prescription of MRA, prescription of diuretics, and prescription of urate-lowering agent, were adjusted. For cardiovascular mortality, covariates including age, race, stroke, BMI, diastolic BP, hemoglobin, triglyceride, albumin, eGFR, serum sodium, serum potassium, serum chloride, prescription of beta-blockers, prescription of MRA, prescription of diuretics, and prescription of urate-lowering agent, were adjusted.)

Competitive risk analysis via Fine-Gray competing risk model

To further explore the gender differences for the prognostic implication of SUA on cardiovascular mortality of CHF, a competing risk analysis via Fine-Gray competing risk model was performed, and non-cardiovascular death was used as a competing event. As illustrated in Fig. 4, Fine-Gray competing risk model test revealed that male patients with highest SUA tertile had highest cumulative risk of cardiovascular death than that of the other categories (Fine-Gray p < 0.001, Fig. 4B), while no significantly statistical difference was observed in female patients (Fine-Gray p = 0.506, Fig. 4C).

The results of the sub-distribution hazard function in the Fine-Gray model were shown in Table 6. For the whole population, univariable and multivariable analysis both demonstrated that SUA were significantly associated with cardiovascular mortality (SHR 1.14, 95% CI 1.08-1.21, p < 0.001, for univariable analysis; SHR 1.10, 95% CI 1.03-1.17, p = 0.006, for multivariable analysis). Furthermore, SUA performed well in prognostic discrimination of cardiovascular mortality for male patients (SHR 1.17, 95% CI 1.09-1.26, p < 0.001, for univariable analysis; SHR 1.17, 95% CI 1.08-1.27, p < 0.001, for multivariable analysis) while not for female patients (SHR 1.10, 95% CI 1.00-1.20, p = 0.044, for univariable analysis; SHR 0.98, 95% CI 0.87 − 1.10, p = 0.690, for multivariable analysis).

Table 6 Association of SUA and cardiovascular mortality via fine-gray competing risk modelFig. 4

Association between SUA and cardiovascular deaths with non-cardiovascular death as a competing risk (Fine-Gray competing risk model)