Objective To explore the relationship between brachial-ankle pulse wave velocity (baPWV) and retinal arteriosclerosis in hypertensive population.

Methods  A total of 2631 hypertensive patients admitted to Tianshui Wulin Street Community Health Service Center, Gongshu District, Hangzhou from June 2021 to June 2024 were enrolled. All patients underwent baPWV detection and retinal arteriosclerosis grading based on binocular fundus images. Among them, 1131 patients had hypertension combined with retinal arteriosclerosis (Scheie classification: 475 cases of grade 1, 396 cases of grade 2, 203 cases of grade 3, 57 cases of grade 4), and 1500 patients had hypertension without retinal arteriosclerosis. Stepwise test combined with Bootstrap method was used to analyze the mediating factors of the correlation between baPWV and retinal arteriosclerosis (mediation analysis). A multivariate logistic regression model with interaction terms was used to analyze the interaction effect of baPWV with diabetes mellitus and dyslipidemia on retinal arteriosclerosis.

Results  Compared with patients without retinal arteriosclerosis, the baPWV in patients with grade 1-4 retinal arteriosclerosis gradually increased (1354±176 vs 1367±178, 1400±182, 1495±194, 1558±203 cm/s, F=89.76, P＜0.001). Taking retinal arteriosclerosis grade as the dependent variable, baseline baPWV as the independent variable, and high-sensitivity C-reactive protein (hsCRP) as the mediating variable, the analysis showed that baPWV indirectly promoted retinal arteriosclerosis by increasing hsCRP (β=0.12, 95%CI: 0.05 to 0.19) (OR=1.41, 95%CI: 1.12 to 1.78), with a mediating effect value of 0.042 (Bootstrap 95%CI: 0.023 to 0.089, P=0.002). Taking baseline baPWV as the dependent variable, retinal arteriosclerosis grade as the independent variable, and peripheral vascular resistance index (PVRI) as the mediating variable, the analysis showed that the increase in retinal arteriosclerosis grade could indirectly lead to elevated baPWV (β=12.3, 95%CI: 8.5 to 16.1) by increasing PVRI (β=0.85, 95%CI: 0.32 to 1.38), with a mediating effect value of 10.455 (Bootstrap 95%CI: 3.21 to 17.70), and the difference was statistically significant. The results of the interaction test of the multivariate logistic regression model showed that compared with the population with diabetes mellitus, the association between baPWV and hypertensive retinal sclerosis was more significant in the population without diabetes mellitus (interaction P=0.003); compared with the population with dyslipidemia, the association between baPWV and hypertensive retinal sclerosis was more significant in the population without dyslipidemia (interaction P=0.004).

Conclusions  There is a bidirectional causal pathway between baPWV and retinal arteriosclerosis. baPWV may promote retinal arteriosclerosis through hsCRP, while retinal arteriosclerosis may reversely increase baPWV through PVRI. The association is more significant in populations without diabetes mellitus and without dyslipidemia.