Diabetes is not a coronary risk equivalent, despite cardiovascular disease is the most common cause of death in diabetes. So, to identify diabetic patients at high cardiovascular risk is necessary. Coronary artery calcification score predicts major coronary events, and improves risk reclassification in asymptomatic diabetic patients. But, cornary artery calcification score is expensive and exposes patients to radiation. So, it cannot be used for large-scale screening. It could be interesting to identify the predictive factors of coronary artery calcification score.

Toe-brachial index is relevant in diabetic patients for the screening of peripheral arterial disease, and predicts cardiovascular events.

The aim of this study is to evaluate the association between toe-brachial index and coronary artery calcification score in asymptomatic patients with type 1 or 2 diabetes. The hypothesis is that toe-brachial index is associated with high coronary artery calcification score. It could be performed first to identify patients who require a coronary artery calcification score. It measurement is reliable, fully automated, repoducible ans cost-effectiveness.

This is a cross-sectional study, with restrospective data collection. All patients addressed to a one-day hospitalization to assess cardiovascular comorbidities are eligible.

Data are collected in patients'medical records. Clinical, biological and imaging data were collected previously during their one-day hospitalization

Cardiovascular disease is the most common cause of death in diabetes. Actually, systematic screening of asymptomatic diabetic patients for silent myocardial ischemia is highly controversial, and is recommended for selected high-risk patients.

Calcium artery calcification score predicts major coronary events, and improves risk reclassification in asymptomatic diabetic patients. The guidelines of the european society of cardiology published in 2013 recommend a screening for silent myocardial ischemia in patients with a high coronary score, without defining a cut-off value. But, assessing cardiovascular risk with calcium coronary score in all asymptomatic patients with diabetes is not feasible. In fact, calcium coronary score expose patients to radiation, is expensive, and is not easily available in health centres. It cannot be used to screen the 4 millions of diabetic patients in France. It could be interesting to identify the predictive factors of a high calcium coronary score, in order to perform coronary artery calcification score only in selected high-risk patients.

Ankle-brachial index is also a marker of cardiovascular risk. Several prospective studies revealed that a low ankle-brachial index predicts cardiovascular events and mortality, and all-cause mortality in diabetes. Nevertheless, a study involving 1343 patients with type 2 diabetes from MESA and Heinz Nixdorf Recall studies has showed that coronary artery calcification score provides better risk reclassification than ankle-brachial index.

Toe-brachial index is particularly relevant in diabetes for peripheral arterial disease screening.

The aim of this study is to evaluate the association between toe-brachial index and coronary artery calcification score in asymptomatic patients with type 1 and 2 diabetes.

The hypothesis is that toe-brachial index is associated with a high coronary artery calcification score. It could be performed first to identify patients who require a coronary artery calcification score. The measurement of toe-brachial index is fully automated, is reliable and reproducible and is cost-effectiveness. This technique is suitable for large-scale screening.

Secondary objectives are :

1. To assess the association between toe-brachial index and severe coronary artery calcification, and to determinate its performance in predicting severe coronary artery calcification
2. To assess the association between toe-brachial index and moderate coronary artery calcification, and to determinate its performance in predicting moderate coronary artery calcification
3. To assess the association between toe-brachial index and the absence of coronary artery calcification, and to determinate its performance in predicting the absence of coronary artery calcification
4. To assess the association between toe-brachial index and early coronary plaque, and to determinate its performance in predicting early coronary atheroma
5. To compare coronary artery calcification between patients with type 1 and type diabetes
6. To assess the association between toe-brachial index and coronary artery calcification in patients with type 1 diabetes
7. To assess the association between toe-brachial index and coronary artery calcification in patients with type 2 diabetes
8. To assess the association between toe-brachial index and an abnormal stress myocardial perfusion tomography, and to determinate its performance in predicting an abnormal stress myocardial perfusion tomography
9. To assess the association between toe-brachial index and an abnormal coronary angiography, and to determinate its performance in predicting an abnormal coronary angiography
10. To assess the association between coronary artery calcification score and an abnormal stress myocardial perfusion tomography, and to determinate its performance in predicting an abnormal stress myocardial perfusion tomography
11. To assess the association between coronary artery calcification score and an abnormal coronary angiography, and to determinate its performance in predicting an abnormal coronary angiography

This is a cross-sectional and single-centre study, with retrospective data collection. All patients addressed to a one-day hospitalization to assess cardiovascular comorbidities, between January 2014 and May 2017, in the diabetes department, in the Pitié-Salpêtrière hospital in Paris, are eligible.

Data are collected in patients' medical records. Clinical, biological and imaging data were collected previously during their one-day hospitalization.

Clinical data are age, sex, diabetes duration, type of diabetes, high blood pressure, dyslipidemia, smoking status, diabetes comorbidities and current medication. Physical examination data are weight, height, body mass index, blood pressure, orthostatic hypotension, symptoms of diabetic peripheral neuropathy, monofilament test, VibraTip and peripheral pulses.

Biological data are HbA1c, fasting blood glucose, HDL-cholesterol, LDL-cholesterol calculated using Friedewald equation, total cholesterol, triglycerides, estimated glomerular filtration rate (eGFR) by modification of diet in renal disease (MDRD), urinary albumin/creatinine ratio, ASAT, ALAT, fibromax protein C reactive and ferritin. Blood and urinary samples have been collected during the one-day hospitalization, and have been analyzed in biochemical department, in Pitié-Salpêtrière hospital.

A retinography bas been performed in patient with known retinopathy or with a mild nonproliferative retinopathy, without ophthalmologic examination since 1 year. Severe retinopathy is defined by severe nonproliferative retinopathy or proliferative retinopathy or retinopathy treated with laser.

Diabetic nephropathy is known or is defined by a urinary albumin/creatinine ratio up to 3 mg/mmol associated with a diabetic retinopathy or a peripheral neuropathy. Albuminuria stages are defined by the urinary albumin/creatinine ratio : no albuminuria if ratio is <3mg/mmol, microalbuminuria if ratio is ≥3 mg/mmol and <30mg/mmol and macroalbuminuria if ratio is ≥30 mg/mmol. Diabetic peripheral neuropathy is known or is defined by typical symptoms or abnormal monofilament test or abnormal ViBratip. Autonomic neuropathy is defined by gastroparesis, cardiovascular autonomic neuropathy, orthostatic hypotension, urinary autonomic dysfunction neuropathy and Charcot foot. Peripheral artery disease is known or is defined by a toe-brachial index <0.7 associated with 2 abnormal pulses on the same side or by a leg artery stenosis ≥ 70% on the ultrasound examination.

Carotid arteries have been studied using an echo-doppler. Intima-media thickness has been measured on longitudinal images, over a 1 cm plaque-free segment free of plaque, 1 cm proximal to the carotid artery bifurcation. Two measurement methods have been used to evaluate intima-media thickness: an automated method using a 3 to 8 MHz linear array transducer (Philips IE33, Koninklijke Philips N.V., Netherlands) and Philips Q-Lab version 8 software (Koninklijke Philips N.V., Netherlands), and a manual method using a 8 or a 4 to 9 MHz transducer (Acuson Sequoi ou Siemens Acuson, respectively). Endpoints are the highest intima-media thickness value between right and left side, and the mean intima-media thickness from right and left side. Plaque is defined as a stenosis <50%, using NASCET and ECST criteria. Carotid stenosis is defined by a stenosis ≥ 50%, using NASCET and ECST criteria.

Toe-brachial index measurement is described in "primary outcome measures". Coronary artery calcification score, stress myocardial perfusion tomography and coronary angiography are described in "secondary outcome measures". Stress myocardial perfusion tomography has been performed if coronary artery calcification score was >100. Coronary angiography has been performed if stress myocardial perfusion tomography was abnormal.

Inclusion Criteria:

• Age ≥ 18 ans
• Diabetes mellitus (type 1 or 2)
• Patients addressed to a one-day hospitalization to assess cardiovascular comorbidities, between January 2014 and May 2017, in the diabetes department, in the Pitié-Salpêtrière hospital
• Coronary artery calcium score available
• Toe-brachial index available

Exclusion Criteria:

• History of coronary artery disease
• History of Raynaud's syndrome
• Opposition of the patient

Patients treated in the diabetes department, in Pitié-Salpêtrière hospital, and adressed for a one-day hospitalization to assess cardiovascular comorbidities.