Obesity and decreased vibration perception associated with premature cardiovascular mortality in a single center prospective study of people with diabetes

Abstract Background and aims The aim of this prospective, single center study was to define factors associated with early cardiovascular mortality in diabetes. Materials and methods 1345 patients under age 75 were included who were undergoing assessment of their diabetes between January 2008 and May 2010 as part of standard practice in a specialist clinic at a regional teaching hospital in Serbia. Peripheral artery disease (PAD) was assessed by audible Doppler waveform and ABPI with cut-offs >1.4 and <0.9. Peripheral neuropathy was assessed by vibration perception threshold (VPT, using a semi-quantitative tuning fork: abnormal if ≤5), ankle reflexes (AR) and sudomotor function of the foot. Diabetic retinopathy (DR) was assessed by fundoscopy. Evidence of vascular disease: thromboendarterectomy and/or cerebrovascular insult (TEA/CVI), myocardial infarction (MI), heart failure (HF), diabetic foot ulcer (DFU), minor amputations (sAMP) and major amputations (mAMP) was also collected. Outcome was determined in 2021 and baseline characteristics were compared between those who had and had not suffered cardiovascular death under age 75 years within 10 years of review in two casually selected cohorts. Results Those who died (n2=70) were more frequently male (60 vs. 45.3%, p=0.08), younger (66.4±7.4 vs. 79.9±3.4, p<0.000), had a shorter period of follow-up (3.6±2.3 vs. 11.2±1.7 years, p<0.000) when compared to those still alive (n1=75). Those who died were also significantly (p<0.01) more likely to have had TEA/CVI (34.3 vs. 10.7%), HF (21.4 vs. 1.3%), MI (44.3 vs. 20%), PAD (48.6 vs. 9.3%), DFU (25.7 vs. 9.3%), mAMP (17.1 vs. 1.3%) at baseline. There were no differences in proliferative DR (17.1 vs. 8%, p=0.10) and laser photocoagulation (25.7 vs. 13.3%, p=0.06) Following multivariable logistic regression analysis significant differences between groups remained for only creatinine (123±45 vs. 88.9±16.9 mmol/L, p<0.003) and VPT <5 (7.8 [95% CI: 3.7–16.4)], p=0.008), estimated maximum lifetime BMI (3.4 [95% CI: 1.7–6.8)], p<0.000), alcohol usage (4.7 [95% CI: 1.5–14.7)], p=0.005), smoking habit (2.2 [95% CI: 1.1–4.3)], p<0.03) and earlier age of diabetes onset (43.4±12.5 vs. 49.2±9.9, p=0.0029). When the 72 patients with impaired vibration sense were compared with 73 with VPT>6, there were significant differences in TEA/CVI (4.2 [95% CI: 1.6–10.9)], p=0.003), PAD (3.9 [95% CI: 1.8–8.8)], p<0.001) and estimated maximum lifetime BMI (9.4 [95% CI: 3.4–25.7), p<0.000). Finally, when those who had had a previous MI at baseline (n=46) were compared with those who hadn't (n=99), MI was associated with increased death rate (3.2 [95% CI: 1.5–6.6)], p=0.002), as was PAD (2.9 [(1.3–6.1)], p=0.007). Conclusion Decreased VPT, the presence of PAD on clinical testing and higher maximum estimated lifetime BMI are strongly associated with premature cardiovascular death. These measures may be independent markers of greater risk of reduced life expectancy. FUNDunding Acknowledgement Type of funding sources: None.