# Hemodynamic parameters and diabetes mellitus in community-dwelling middle-aged adults and elders: a community-based study

## TL;DR

This study finds that diabetes is linked to changes in blood flow in the carotid arteries, which could help identify people at higher risk and improve prevention strategies.

## Contribution

The study demonstrates that hemodynamic parameters, particularly EDV, significantly improve diabetes prediction models beyond conventional risk factors.

## Key findings

- Diabetes patients had lower PSV, EDV, and MFV and higher RI and PI compared to non-diabetic controls.
- Adding EDV to regression models significantly improved diabetes predictability, increasing AUROC by 1.00% and 0.80%.
- PSV and MFV also improved model predictability, though to a lesser extent than EDV.

## Abstract

Hemodynamic parameters have been correlated with stroke, hypertension, and arterial stenosis. While only a few small studies have examined the link between hemodynamics and diabetes mellitus (DM). This case-control study enrolled 417 DM patients and 3475 non-DM controls from a community-based cohort. Peak systolic velocity (PSV), end-diastolic velocity (EDV), blood flow velocity (MFV), pulsatility index (PI), and the resistance index (RI) of the common carotid arteries were measured by color Doppler ultrasonography. Generalized linear regression analyses showed that as compared to the non-DM controls, the age-sex-adjusted means of PSV, EDV, and MFV were − 3.28 cm/sec, − 1.94 cm/sec, and − 2.38 cm/sec, respectively, lower and the age-sex-adjusted means of RI and PI were 0.013 and 0.0061, respectively, higher for the DM cases (all p-values < 0.0005). As compared to the lowest quartiles, the multivariable-adjusted ORs of DM for the highest quartiles of PSV, EDV, MFV, RI, and PI were 0.59 (95% confidence interval [CI] 0.41–0.83), 0.45 (95% CI 0.31–0.66), 0.53 (95% CI 0.37–0.77), 1.61 (95% CI 1.15–2.25), and 1.58 (95% CI 1.12–2.23), respectively. More importantly, the additions of EDV significantly improved the predictabilities of the regression models on DM. As compared to the model contained conventional CVD risk factors alone, the area under the receiver operating curve (AUROC) increased by 1.00% (95% CI 0.29–1.73%; p = 0.0059) and 0.80% (95% CI 0.15–1.46%; p = 0.017) for models that added EDV in continuous and quartile scales, respectively. Additionally, the additions of PSV and MFV also significantly improved the predictabilities of the regression models (all 0.01 < p-value < 0.05). This study reveals a significant correlation between DM and altered hemodynamic parameters. Understanding this relationship could help identify individuals at higher risk of DM and facilitate targeted preventive strategies to reduce cardiovascular complications in DM patients.