Cardiovascular autonomic neuropathy (CAN) in people with diabetes is independently associated with vascular dysfunction and mortality (1). The prevalence of CAN varies but is approximately 20% and can reach 60%, depending on the definition used for diagnosis as well as population characteristics, including age and metabolic control (2,3). Observational studies report a higher prevalence of CAN in those with diabetes-related foot ulceration (DFU), and evidence from prospective cohort studies suggest that CAN increases a person’s risk of developing DFUs (4). Therefore, CAN may precede DFUs and be a risk factor for lower-extremity complications. Using a prospective cohort study design, we investigated if CAN is associated with an increased risk of first DFU or lower-extremity amputation.
The cohort consisted of individuals with type 1 (n = 478) or type 2 (n = 584) diabetes, aged 20–80 years and without existing cardiac arrythmia, who were screened for CAN between May 2013 and February 2015 at the Steno Diabetes Centre Copenhagen (SDCC). CAN diagnosis was based on heart rate response to cardiovascular reflex tests (CARTs): the lying-to-standing test (30:15 ratio), the deep breathing test (E:I ratio), and the Valsalva maneuver (5). Two or more abnormal CARTs indicated definite CAN. Data on incident foot ulceration (ICD-10 code L979) or lower-extremity amputation (ICD-10 codes Z894, Z895, or Z896) were gathered from patient records at the SDCC and the Danish National Patient Register. HbA1c, lipid levels, systolic and diastolic blood pressure, height, and weight were recorded in conjunction with CAN testing. Vibration perception threshold was measured using a Biothesiometer on the apex of the hallux bilaterally. A detection threshold >25 V was considered as peripheral neuropathy.
Statistical methods used included Poisson regression for split-time follow-up (along calendar year in 0.5-year intervals) to model incidence rate ratios (IRR). Baseline covariates were included in the models in three steps: model 1 - age, sex, diabetes duration, and diabetes type; model 2 - addition of systolic blood pressure, LDL cholesterol, triglycerides, HbA1c, and smoking; model 3 - addition of vibration sensation status. Age and diabetes duration were included as natural splines with 5 knots. A modifying effect of sex and diabetes type on associations was tested. Likelihood ratio tests were used to compare models with and without interactions.
We followed 1,062 individuals without previous DFU for a mean (SD) of 7.3 (1.2) years until first outcome, emigration, death, or last update of medical records (1 January 2021). During follow-up, 14 participants emigrated and 80 participants died. Data on DFU status were available for all individuals until they died or emigrated. Those with CAN (n = 142, 13%) had significantly (P < 0.05) higher HbA1c, BMI, triglycerides, and urinary albumin-to-creatinine ratio, smoked more frequently, and had lower physical activity levels than those without CAN. The median (interquartile range) age for both groups was 57 (48, 66) years. Overall, 57 (5%) people developed a DFU, resulting in a crude incident rate of 7.6 per 1,000 person-years; 25 per 1,000 person-years for those with CAN vs. 5 per 1,000 person-years for those without CAN. For all models, the IRR of DFUs was higher in those with CAN than in those without. For model 1, the IRR was 4.3 (95% CI 2.5–7.4, P < 0.001), and for model 2 it was 4.0 (95% CI 2.1–7.3, P < 0.001). The inclusion of bilateral absent vibration sensation in model 3 reduced the effect (IRR 2.99, 95% CI 1.5–5.5; P < 0.01), yet it was still significant. Diabetes type and sex had no modifying effect (P > 0.05) on the association. The IRR remained higher (>2) in those with CAN at any assessed age (Fig. 1). However, risk was greater for those who were younger, with the highest risk being among those aged 45 years and the lowest risk being among those aged 80 years. Only 7 (0.7%) individuals underwent a lower-extremity amputation. Their IRRs could not be modeled due to small numbers.
This graph illustrates the predicted age-specific risk of developing a first foot ulcer for individuals with CAN compared with individuals without CAN. Data are for individuals who are female, who have type 2 diabetes, a diabetes duration of 12 years, systolic blood pressure 132 mmHg, LDL 2.4 mmol/L, triglyceride 1.8 mmol/L, HbA1c 55 mmol/mol, have absent vibration sensation and who smoke. The shaded areas indicate 95% CI.
This graph illustrates the predicted age-specific risk of developing a first foot ulcer for individuals with CAN compared with individuals without CAN. Data are for individuals who are female, who have type 2 diabetes, a diabetes duration of 12 years, systolic blood pressure 132 mmHg, LDL 2.4 mmol/L, triglyceride 1.8 mmol/L, HbA1c 55 mmol/mol, have absent vibration sensation and who smoke. The shaded areas indicate 95% CI.
In conclusion, we found CAN to be an independent risk factor for DFUs, with a three-times-higher incidence rate of DFUs when compared to those without CAN. We are the first to show this association independently of peripheral neuropathy (4). This study has some limitations. First, the broad 95% CI suggest a degree of uncertainty around the results. Despite this, we find CAN is an independent DFU risk factor. Second, CAN status was assessed at baseline, meaning people may have developed CAN during the follow-up period, potentially leading to underestimation of the impact of CAN on ulceration.
We conclude that our results demonstrate that CAN is associated with a threefold increased risk of DFUs. This suggests that CAN is a potential marker for early screening of individuals with diabetes who are at high risk of developing foot complications and thus contribute to prevention of lower-extremity complications in people with diabetes.
Data and Resource Availability
The data that support the findings of this study are available from Statistics Denmark, but restrictions apply to the availability of these data, which were used under license for the current study and, therefore, are not publicly available.
Article Information
Acknowledgments. We acknowledge all patients who have participated in the screening study that the study is based on. The authors thank Bendix Carstensen, senior statistician, SDCC, for his advice on statistical analysis.
Funding. J.A.P. was supported by Health Research Board Ireland (CDA-19-007). T.S.A. is supported by a Novo Nordisk Foundation grant (NNF23OC0084081; Map D-Foot).
Author Contributions. J.A.P. contributed to the formal analysis, investigation, and writing the original draft. C.S.H. contributed to conceptualization of the study and to study supervision and reviewed and edited the manuscript. P.F.R. contributed the formal analysis and reviewed and edited the manuscript. T.S.A., S.F.D., C.M.B., and P.M.K. contributed to supervision of the work and reviewed and edited the manuscript. A.R., J.F., and M.E.J. reviewed and edited the manuscript. J.A.P. and C.S.H. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Duality of Interest. J.F. is the coinventor of the vagus device that was used in this study. T.S.A. and P.F.R. holds stocks with Novo Nordisk A/S. No other potential conflicts of interest relevant to this article were reported.
Prior Presentation. Portions of this study were presented at the Neurodiab 34th Annual Conference, Rome, Italy, 5–8 September 2024.
Handling Editors. The journal editors responsible for overseeing the review of the manuscript were Elizabeth Selvin and Justin B. Echouffo-Tcheugui.
