Borderline ankle–brachial index is associated with increased prevalence of micro- and macrovascular complications in type 2 diabetes: A cross-sectional analysis of 12,772 patients from the Joint Asia Diabetes Evaluation Program

Abstract

Borderline ankle–brachial index is increasingly recognised as a marker of cardiovascular risk. We evaluated the impact of borderline ankle–brachial index in 12,772 Chinese type 2 diabetes patients from the Joint Asia Diabetes Evaluation Program between 2007 and 2012. Cardiovascular risk factors, complications and health-related quality of life were compared between patients with normal ankle–brachial index (1.0–1.4), borderline ankle–brachial index (0.90–0.99) and peripheral arterial disease (ankle–brachial index < 0.9). The prevalence of peripheral arterial disease and borderline ankle–brachial index was 4.6% and 9.6%, respectively. Borderline ankle–brachial index patients were older, more likely to be smokers and hypertensive, had longer duration of diabetes, poorer kidney function and poorer health-related quality of life than patients with normal ankle–brachial index. After adjustment for traditional cardiovascular risk factors, borderline ankle–brachial index was an independent predictor of diabetes-related micro- and macrovascular complications including retinopathy (odd ratios: 1.19 (95% confidence interval: 1.04–1.37)), macroalbuminuria (1.31 (1.10–1.56)), chronic kidney disease (1.22 (1.00–1.50)) and stroke (1.31 (1.05–1.64)). These findings suggest that patients with diabetes and borderline ankle–brachial index are at increased cardiovascular risk and may benefit from more intensive management.

Keywords

Ankle–brachial index peripheral arterial disease type 2 diabetes Chinese micro- and macrovascular complications Joint Asia Diabetes Evaluation Program

Introduction

Asia is currently facing a burgeoning epidemic of type 2 diabetes.^1,2 Cardiovascular (CV) diseases remain the leading causes of morbidity and pre-mature death in the diabetic population, but the assessment of CV risk is often inadequate.^1,3 The measurement of ankle–brachial index (ABI) is a simple and useful method for estimating CV risk and is recommended by the American Diabetes Association for all diabetic patients older than 50 years.⁴ An ABI value ⩽0.9 is conventionally used as the threshold to define peripheral arterial disease (PAD) and high CV risk.^5,6 Recent studies have reported that ABI value of 0.91–0.99 is also associated with increased risk of CV diseases (CVDs) in non-diabetic patients.^7,8 The 2011 American College of Cardiology Foundation (ACCF) and American Heart Association (AHA) guidelines for the management of PAD have recommended ABI values of 0.90–0.99 as ‘borderline’ rather than normal.⁹ However, there is a paucity of data on the impact of borderline ABI in diabetic patients. In this study, we aim to evaluate the prevalence of borderline ABI and its associations with diabetes-related CV complications in a large cohort of Chinese patients with type 2 diabetes.

Methods

Study subjects and setting

Between November 2007 and July 2012, Chinese patients with type 2 diabetes aged 18 years or above enrolled into the Joint Asia Diabetes Evaluation (JADE) Program from three public hospitals and three community-based out-patient clinics in Hong Kong were studied. Exclusion criteria included patients with type 1 diabetes as defined by presentation with ketoacidosis, unprovoked ketosis or requirement of insulin within 1 year of diagnosis; ABI > 1.4 indicating non-compressible arteries; and missing data on ABI. The design, implementation and adoption of the JADE Program have been published.^10–12 In brief, the JADE Program is a web-based programme that uses information technology to facilitate healthcare professionals to create a diabetes registry and to deliver evidence-based care tailored to patients’ risk profiles. All enrolled patients gave written informed consent to undergo comprehensive assessment, which includes clinical and demographic information and laboratory investigations at enrolment and at regular intervals thereafter. At each review visit, key parameters as recommended by international guidelines are captured to document clinical progress, risk factor control and complications since last visit. The JADE Program is governed by the Asia Diabetes Foundation, which is a non-profit making organisation incorporated under the Chinese University of Hong Kong Foundation.

Clinical assessments

All patients recruited in Hong Kong JADE Program undertook comprehensive assessment including medical history taking, physical examination, laboratory testing and quality of life assessment using a structured protocol. History of complications, comorbidities and medication use were cross-checked with the Hong Kong Hospital Authority (HA) Clinical Management System (CMS), which captured all admissions to public health facilities used by over 90% of Hong Kong residents. Physical examination included standardised anthropometric measurements of body weight, height and waist circumference, as well as blood pressure (BP). Fundoscopy or retinal photography was used to assess the presence of retinopathy. Peripheral sensory neuropathy was assessed using monofilament and graduated tuning fork. All patients underwent routine ABI assessment during annual assessments regardless of PAD status or symptoms suggestive of PAD. ABI was measured by a trained nurse using an 8-mHz hand-held Doppler device (Summit Doppler System, Trumbull, CT) and a sphygmomanometer. Systolic BP from both brachial arteries and dorsalis pedis arteries were measured with the patient in the supine position. ABI was calculated for each leg by taking the higher pressure of the two arteries at the ankle, divided by the higher of the two arm brachial arterial systolic pressure. The lower ABI of each patient was used for analysis. ABI was classified into one of the three categories: PAD (ABI ⩽ 0.90), borderline ABI (0.91–0.99) and normal ABI (1.00–1.40).

Fasting blood samples after at least an 8-h fast overnight were collected for measurement of fasting plasma glucose, glycated haemoglobin (HbA_1c), lipid profile [total cholesterol, high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C) and triglycerides] and renal function. The abbreviated Modification of Diet in Renal Disease study formula recalibrated for the Chinese population was used to determine the estimated glomerular filtration rate (eGFR).¹³ A random spot urine sample was collected to measure the albumin-to-creatinine ratio (ACR).

Health-related quality of life (HRQoL) was assessed using the EuroQoL 5-dimension (EQ-5D) tool.^14,15 Patients were asked to rank their health status of self-care, usual activities, pain/discomfort, mobility and anxiety/depression using three levels: ‘no problems’, ‘moderate problems’ and ‘extreme problems’. The UK tariff was used to calculate the index.¹⁶ ‘Moderate problems’ and ‘extreme problems’ were classified as ‘having problems’ in analysis.

Definitions of complications

Macroalbuminuria was defined as urinary ACR > 25.0 mg/mmol. Chronic kidney disease (CKD) was defined as eGFR < 60 mL/min/1.73 m². History of CVD [coronary heart disease (CHD), congestive heart failure or stroke] was based on medical history confirmed by the referring doctor and/or Hong Kong HA record. Peripheral sensory neuropathy was defined by meeting two of three criteria of abnormal sensation over lower limb, reduced sensation to monofilament over any part of the sole and a score of ⩽6 of 8 (aged >65 years) or ⩽4 of 8 (aged ⩾65 years) using a graduated tuning fork. Retinopathy was defined by typical retinal changes including vitrectomy, haemorrhages and exudates on fundoscopy. General obesity was defined as body mass index (BMI) ⩾27.5 kg/m² while central obesity was defined as waist circumference ⩾90 cm in men or ⩾80 cm in women. Hypertension was defined as BP ⩾ 140/90 mmHg and/or treatment with anti-hypertensive drugs. Dyslipidaemia was defined as a plasma LDL-C ⩾ 2.6 mmol/L, HDL-C < 1 mmol/L, triglycerides ⩾ 2.3 mmol/L¹⁷ or treatment with any lipid-regulating drugs. The ‘ABC’ treatment goals were defined as (A) HbA_1c < 7.0%, (B) BP < 130/80 mmHg and (C) LDL-C < 2.6 mmol/L.¹⁸ We also calculated the 5-year risk of CHD and stroke using the equations derived from the Hong Kong Diabetes Registry.^19,20 We then used 5-year CHD risk ⩾20% and 5-year stroke risk ⩾10% to identify high-risk subjects for future CV events in those without prior history of CHD or stroke.

Statistical analysis

All data were reported as number (percentages), mean ± standard deviation (SD) or median [interquartile range (IQR)] as appropriate. χ²-test, Kruskal–Wallis H test and one-way analysis of variance (ANOVA) were used for comparisons as appropriate. Multiple logistic regression models were used to examine the associations between ABI and microvascular (retinopathy, CKD, sensory neuropathy) and macrovascular complications (stroke, CHD) using normal ABI as the reference and expressed as odds ratios (OR) with 95% confidence interval (CI). The covariates entered into the model were based on clinical a priori knowledge, including age, sex, duration of diabetes, high school education, smoking, waist circumference, hypertension, dyslipidaemia and HbA_1c. Statistical analysis was performed using the SPSS (version 20.0, IBM). A p-value < 0.05 was considered to be statistically significant. Bonferroni correction was used for post hoc comparisons between normal, borderline ABI and PAD.

Results

Of the 13,228 patients with diabetes enrolled into the JADE Program, we excluded 280 patients with type 1 diabetes, 112 patients with missing data for ABI and 64 (0.5%) patients with an ABI > 1.4. Of the remaining 12,772 patients (mean age: 58.8 ± 11.9 years), 54.6% (n = 6972) were men and median duration of diabetes was 7 (IQR: 2–14) years. Anthropomorphically, 63.6% of patients were centrally obese, 63.6% had hypertension and 87.4% had dyslipidaemia. In these subjects, 9.2% had CKD, 6.4% had history of stroke and 10.3% had known CHD.

Overall, 9.6% had borderline ABI and 4.6% had PAD with higher rates in women (borderline ABI 11.6% and PAD 5.4%) than men (borderline ABI 8.0%, PAD 4.0% and both p < 0.001). Compared to patients with normal ABI, those with borderline ABI had lower education and were more likely to be active smokers (Table 1). They also had longer duration of diabetes, higher systolic BP and worse renal function but had comparable BMI, waist circumference, HbA_1c, LDL-C and triglyceride. Fewer patients in the borderline ABI group attained all three ‘ABC’ targets than normal ABI group (7.5% vs 10.0%, p = 0.002, Table 2). Among the three groups, patients with PAD had the worst metabolic profile with the highest HbA_1c, systolic BP and waist circumference, although waist circumference was only significantly higher in women with PAD. Overall, 21.8% of patients were treated with insulin with a progressive increase in insulin usage across normal ABI, borderline ABI and PAD groups. While patients with borderline ABI were more likely than the normal ABI group to receive anti-hypertensive drugs, they were less likely to be treated with statins and renin–angiotensin system (RAS) inhibitors than those with PAD (both p < 0.001) (Table 2).

Table 1.

Demographic and clinical characteristics of Chinese patients with type 2 diabetes categorised according to ankle–brachial index (ABI).

	Normal, n = 10,952 (85.8%)	Borderline, n = 1230 (9.6%)	PAD, n = 590 (4.6%)	p-trend	p-value (borderline vs normal)
Demographic
Age, years	58.2 ± 11.3	60.4 ± 14.2	68.3 ± 13.3	<0.001	<0.001
Female	4813 (43.9)	675 (54.9)	312 (52.9)	<0.001	<0.001
Diabetes duration, years	7 (2, 13)	8 (3, 15)	12 (5, 20)	<0.001	<0.001
Education				<0.001	<0.001
Illiterate/primary school	4188 (38.4)	620 (50.9)	369 (63.5)
Middle school	4447 (40.8)	378 (31.1)	142 (24.4)
High school and above	2258 (20.7)	219 (18.0)	70 (12.0)
Smoking				<0.001	0.019
Ex	2169 (19.8)	215 (17.5)	153 (25.9)
Current	1216 (11.1)	164 (13.3)	72 (12.2)
Physical activity ⩾3 times/week	5139 (49.6)	541 (47.6)	259 (48.4)	0.258	0.201
Clinical measurements
ABI	1.12 ± 0.08	0.96 ± 0.03	0.75 ± 0.13	<0.001	<0.001
Body mass index, kg/m²	25.9 ± 4.2	25.8 ± 4.9	25.3 ± 4.3	0.001	0.569
Waist circumference (female), cm	86.4 ± 10.8	86.9 ± 12.3	88.0 ± 10.7	0.012	0.492
Waist circumference (male), cm	91.9 ± 10.7	91.3 ± 10.9	91.3 ± 10.3	0.408	0.500
Systolic BP, mmHg	134 ± 18	140 ± 21	149 ± 22	<0.001	<0.001
Diastolic BP, mmHg	79 ± 11	78 ± 11	78 ± 12	0.016	0.032
Laboratory assessments
HbA_1c, %	7.46 ± 1.53	7.52 ± 1.61	7.68 ± 1.72	0.001	0.506
HbA_1c, mmol/mol	58.1 ± 16.8	58.1 ± 17.6	60.4 ± 18.8	<0.001	0.506
Total cholesterol, mmol/L	4.57 ± 0.98	4.57 ± 1.00	4.50 ± 1.10	0.081	1.000
Triglycerides, mmol/L	1.3 (0.9, 1.9)	1.3 (0.9, 1.9)	1.4 (1.0, 2.0)	0.100	0.998
HDL-C, mmol/L	1.03 ± 0.35	1.03 ± 0.35	1.25 ± 0.35	0.003	0.888
LDL-C, mmol/L	2.57 ± 0.81	2.58 ± 0.87	2.51 ± 0.92	0.113	1.000
eGFR (mL/min/1.73 m²)	106 ± 36	101 ± 40	81 ± 37	<0.001	<0.001
Urine ACR (mg/mmol)	1.4 (0.6, 5.7)	2.1 (0.7, 12.7)	6.6 (1.5, 45.3)	<0.001	<0.001
Cardiometabolic risk factors
General obesity	3311 (30.3)	356 (29.0)	163 (28.1)	0.376	0.358
Central obesity	6921 (63.3)	791 (64.4)	401 (68.3)	0.038	0.456
Hypertension	6796 (62.1)	838 (67.1)	487 (82.5)	<0.001	<0.001
Dyslipidaemia	9494 (87.0)	1090 (88.8)	547 (92.7)	<0.001	0.066

ACR: albumin:creatinine ratio; ABI: ankle–brachial index; BP: blood pressure; eGFR: estimated glomerular filtration rate; PAD: peripheral arterial disease; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol.

Data are shown as mean ± SD, number (percentage) or median (interquartile range).

Table 2.

Diabetes-related complications, treatment and achievement of therapeutic targets in type 2 diabetic patients according to ABI category.

	Normal, n = 10,952 (85.8%)	Borderline, n = 1230 (9.6%)	PAD, n = 590 (4.6%)	p-trend	p-value (borderline vs normal)
Diabetes-related complications
Diabetic retinopathy	2752 (25.1)	375 (30.5)	219 (37.4)	<0.001	<0.001
Sensory neuropathy	421 (3.8)	70 (5.7)	90 (15.3)	<0.001	0.002
Macroalbuminuria	1249 (11.5)	202 (16.6)	177 (30.6)	<0.001	<0.001
Chronic kidney disease	822 (7.5)	159 (12.9)	196 (33.4)	<0.001	<0.001
Stroke	585 (5.3)	108 (8.8)	125 (21.2)	<0.001	<0.001
Coronary heart disease	1017 (9.3)	147 (12.0)	158 (26.8)	<0.001	0.002
Treatments
Insulin	2257 (20.6)	323 (26.3)	209 (35.4)	0.001	0.002
Antihypertensive drugs	6423 (58.6)	790 (64.2)	457 (77.5)	<0.001	<0.001
Lipid lowering drugs	5031 (45.9)	602 (48.9)	360 (61.0)	<0.001	0.045
Statin	4633 (42.3)	564 (45.9)	340 (57.6)	<0.001	0.017
RAS inhibitor	4954 (45.2)	596 (48.5)	349 (59.2)	<0.001	0.031
Target achievements and risk scores
HbA_1c < 7.0% (53 mmol/mol)	4890 (44.8)	553 (45.0)	234 (39.7)	0.195	0.851
BP < 130/80 mmHg	3932 (35.9)	351 (28.5)	101 (17.1)	<0.001	<0.001
LDL-C < 2.6 mmol/L	5579 (51.9)	641 (53.0)	355 (61.0)	<0.001	0.477
At least one target	8861 (81.7)	980 (80.6)	459 (78.6)	0.071	0.339
At least two targets	4444 (41.0)	473 (38.7)	205 (34.9)	0.003	0.128
All three targets	1096 (10.0)	92 (7.5)	26 (4.4)	<0.001	0.002
5-year CHD risk (%)	10.0 (7.2,20.3)	11.4 (7.6, 24.8)	18.3 (11.4, 32.4)	<0.001	<0.001
5-year stroke risk (%)	2.7 (1.6, 5.2)	3.5 (1.6, 8.5)	9.8 (3.9, 18.4)	<0.001	<0.001
5-year CHD risk ⩾ 20%	859 (10.6)	181 (19.4)	164 (43.5)	<0.001	<0.001
5-year stroke risk ⩾ 10%	991 (9.7)	222 (20.1)	224 (49.2)	<0.001	<0.001

ABI: ankle–brachial index; CHD: coronary heart disease; RAS: renin–angiotensin system; BP: blood pressure; LDL-C: low-density lipoprotein cholesterol.

Data are shown as mean ± SD, number (percentage) or median (interquartile range).

Based on the risk equations derived from the Hong Kong Diabetes Registry,^19,20 the 5-year risk of stroke and CHD increased with declining ABI measurements (p-trend < 0.001) (Table 2). Additionally, more patients with borderline ABI had higher frequencies of CHD, stroke, CVD, macroalbuminuria and CKD than those with normal ABI. They also had higher rates of retinopathy and sensory neuropathy (Table 2). After adjusting for traditional CV risk factors, including age, sex, disease duration, higher school education, smoking, waist circumference, hypertension, dyslipidaemia and HbA_1c, borderline ABI remained independently associated with 1.2–1.3 increased odds of retinopathy, macroalbuminuria, CKD and stroke (Figure 1).

Figure 1.

Adjusted odds ratios for micro- and macrovascular complications in patients with borderline ankle–brachial index (ABI) compared to those with normal ABI.

Patients with PAD and borderline ABI reported lower HRQoL than patients with normal ABI as evidenced by lower EQ-5D health utility index. Across the five dimensions of EQ-5D evaluated, patients with borderline ABI were more likely to experience problems in mobility, self-care and usual activity, but not anxiety/depression or pain/discomfort than those with normal ABI (Figure 2).

Figure 2.

Prevalence of reported problems in EQ-5D by ABI groups.

Discussion

To our best knowledge, this is the first report on the prevalence of borderline ABI and its associations with micro- and macrovascular complications in a large cohort of Chinese patients with type 2 diabetes. We demonstrated that twice as many diabetic patients had borderline ABI (9.6%) than PAD (4.6%). Apart from having poorer HRQoL, patients with borderline ABI had 20%–30% increased odds of having micro- and macrovascular complications compared to those with normal ABI. However, these high-risk subjects were less likely to be prescribed disease-modifying drugs such as RAS inhibitors and statins than those with PAD. These findings suggest that borderline ABI identifies a significant population of high-risk type 2 diabetic patients for whom intensive management may reduce future risk of CV–renal complications. PAD has been reported in 6%–30% of Chinese type 2 diabetes patients,^21–23 but the prevalence of borderline ABI remained less studied. In our cohort, the prevalence of borderline ABI (9.6%) was comparable to reported rates of 8.9% and 8.7% in a recent Japanese study of 3981 diabetic patients and in US National Health and Nutrition Examination Survey (NHANES, 1999–2002) involving 4895 subjects aged ⩾40 years in the community, respectively.^8,24 Women in our cohort were more likely to have borderline ABI (11.6%) than men (8.0%). Similarly, in the NHANES (1999–2002) and the Multi-Ethnic Study of Atherosclerosis (MESA), the prevalence of borderline ABI nearly doubled in women (11.7% and 10.6%) than men (6.0% and 4.3%).^7,8 The increased prevalence of borderline ABI in women may be due to older age compared to men.

Borderline ABI is increasingly recognised as a marker of CV risk. In the MESA study, patients with borderline ABI were more likely to have increased carotid artery intima-media thickness and a coronary artery calcium score >20 than those with normal ABI.⁷ In a meta-analysis, patients with borderline ABI had a hazard ratio (HR) of 1.43 in men and 1.53 in women for major adverse coronary events.²⁵ In a recent study of Japanese diabetic patients, the risks for all-cause death in patients with PAD (HR = 2.16; 95% CI: 1.46–3.14) and borderline ABI (HR: 1.78; 95% CI: 1.14–2.70) were significantly higher than in those with normal ABI.²⁴ In our study, 5-year CHD and stroke risk were increased by 14%–30% in patients with borderline ABI. The latter was also independently associated with retinopathy and macroalbuminuria, which predicted CV and all-cause death in both general²⁶ and diabetic populations.²⁷

The association of impaired physical function and HRQoL with PAD, with or without symptoms, has been reported.^28,29 In our study, patients with borderline ABI had worse HRQoL and lower scores in all EQ-5D (mobility, self-care and usual activity) except for anxiety/depression and pain/discomfort compared to those with normal ABI. Other researchers have also reported poorer HRQoL in individuals with borderline ABI than in those with normal ABI.³⁰ In a 5-year prospective study, borderline ABI was associated with a HR of 3.07 (95% CI: 1.21–7.84) for loss of mobility compared to those with normal ABI.³¹ These findings highlighted that in addition to being at risk for clinical events, patients with borderline ABI also had functional impairment and suboptimal quality of life.

Akin to the prognostic significance of ‘pre-hypertension’ (systolic BP: 120–139 mmHg and/or diastolic BP: 80–89 mmHg),³² our findings strongly support the use of borderline ABI to identify high-risk diabetic patients for early intervention. In the NHANES cohort, subjects with borderline ABI were more likely to have worse modifiable cardio-metabolic risk factors such as hypertension and abdominal obesity.⁸ Clairotte et al.³³ found that an ABI threshold of 0.90 had a high specificity (96.8%) but a medium sensitivity (54.4%) for two-dimensional ultrasound-identified arterial stenosis in diabetic patients compared to a sensitivity of 98.3% and specificity of 72.6% in non-diabetic subjects. However, using a cut-off value of 1.0–1.1 had better performance and identified >80% of diabetic patients with PAD.³³ In support of these findings, our results also suggested the use of a higher cut-off value for PAD screening in diabetic patients.

In both randomised clinical trials and observational cohorts, optimal control of multiple risk factors reduced micro- and macrovascular complications and mortality in type 2 diabetes.^2,18 However, less than one-third of our patients with borderline ABI attained BP target (<130/80 mmHg), and only half were at LDL-C goal (<2.6 mmol/L). Compared to patients with PAD, they were also less likely to be treated with statins or RAS inhibitors. Given the increased risk of complications in those with borderline ABI, using a higher cut-off value of ABI may identify high-risk subjects for intensive management.

Several limitations of this study need to be acknowledged. First, this study only involved Chinese patients and therefore might not apply to other ethnic populations. Second, we did not repeat ABI measurements since all nurses had undergone training. Besides, measurement bias in this large population may be assumed to be random and are not likely to affect the conclusions. Third, despite the relatively large sample size and careful clinical assessment, the cross-sectional nature of our study does not allow causal inference of the associations between borderline PAD and diabetic complications. Further research is required to validate the predictive value of borderline ABI for risk stratification and clinical outcomes in type 2 diabetes.

In Chinese patients with type 2 diabetes, twice as many subjects had borderline ABI as those with PAD which was associated with increased prevalence of micro- and macrovascular complications and decreased HRQoL. Diabetic patients with not only abnormal ABI < 0.9 but also borderline ABI (0.91–0.99) could be at high risk of CV complications. We proposed using a higher cut-off value of ABI (<1.0) to screen for CV risk, potentially identifying an additional 10% of high-risk diabetic patients for intensive treatment.

Footnotes

Acknowledgements

We thank all participating doctors, nurses and patients for contributing to the JADE Program. Special thanks are extended to Ms Harriet Chung, Ms Rebecca Wong, Ms Grace Lin, Ms Maggie Lau and Ms Marina Cheung, for coordinating the diabetes comprehensive assessment programmes. B.P.Y., Y.Z., A.P.S.K. and J.C.N.C. participated in the study design and conception. B.P.Y., Y.Z., A.P.S.K., A.O.Y.L., R.O., R.Y., P.C.Y.T., C.T., K.L., T.W., G.L., W.S., R.C.W.M., F.C.C.C. and J.C.N.C. performed the statistical analysis. B.P.Y., Y.Z., A.O.Y.L., R.O., R.Y., P.C.Y.T., C.T., K.L., W.S., R.C.W.M., F.C.C.C. and J.C.N.C. performed the data interpretation. B.P.Y. and Y.Z. drafted the manuscript. B.P.Y., Y.Z., A.P.S.K., A.O.Y.L., R.O., R.Y., P.C.Y.T., C.T., K.L., T.W., G.L., W.S., R.C.W.M., F.C.C.C. and J.C.N.C. revised the manuscript. A.P.S.K. and J.C.N.C. participated in the data acquisition. All authors read and approved the final manuscript. B.P.Y. and Y.Z. have equal contributions.

Declaration of conflicting interests

Dr Juliana CN Chan has received honorarium for consultancy or giving lectures from Bayer, Boehringer Ingelheim, Daiichi-Sankyo, Eli-Lilly, GlaxoSmithKline, Merck Sharp & Dohme, Merck Serono, Pfizer, Astra Zeneca, Sanofi, Novo-nordisk and/or Bristol-Myers Squibb. Troels Wolthers and Greg Lyubomirsky are also employees of Merck Sharp & Dohme. The rest of authors declare that they have no conflicts of interest.

Funding

This work was partially supported by the Asia Diabetes Foundation, a charitable organisation under the auspices of the Chinese University of Hong Kong Foundation, and an unrestricted grant from Merck Sharpe & Dohme and Liao Wun Yuk Diabetes Memorial Fund, the latter under the Chinese University of Hong Kong.

References

Emerging Risk Factors Collaboration; Seshasai

Kaptoge

Thompson

. Diabetes mellitus, fasting glucose, and risk of cause-specific death. N Engl J Med 2011; 364: 829–841.

Inzucchi

Bergenstal

Buse

. Management of hyperglycemia in type 2 diabetes: a patient-centered approach: position statement of the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetes Care 2012; 35: 1364–1379.

Chan

Gagliardino

Baik

. Multifaceted determinants for achieving glycemic control: the International Diabetes Management Practice Study (IDMPS). Diabetes Care 2009; 32: 227–233.

American Diabetes Association. Standards of medical care in diabetes – 2013. Diabetes Care 2013; 36: S11–S66.

Zheng

Sharrett

Chambless

. Associations of ankle-brachial index with clinical coronary heart disease, stroke and preclinical carotid and popliteal atherosclerosis: the Atherosclerosis Risk in Communities (ARIC) Study. Atherosclerosis 1997; 131: 115–125.

Resnick

Lindsay

McDermott

. Relationship of high and low ankle brachial index to all-cause and cardiovascular disease mortality: the Strong Heart Study. Circulation 2004; 109: 733–739.

McDermott

Liu

Criqui

. Ankle-brachial index and subclinical cardiac and carotid disease: the multi-ethnic study of atherosclerosis. Am J Epidemiol 2005; 162: 33–41.

Menke

Muntner

Wildman

. Relation of borderline peripheral arterial disease to cardiovascular disease risk. Am J Cardiol 2006; 98: 1226–1230.

American College of Cardiology Foundation, American Heart Association Task Force, Society for Cardiovascular Angiography and Interventions, . 2011 ACCF/AHA focused update of the guideline for the management of patients with peripheral artery disease (updating the 2005 guideline). Vasc Med 2011; 16: 452–476.

10.

Tong

. From design to implementation – the Joint Asia Diabetes Evaluation (JADE) program: a descriptive report of an electronic web-based diabetes management program. BMC Med Inform Decis Mak 2010; 10: 26.

11.

Chan

. The Joint Asia Diabetes Evaluation (JADE) Program: a web-based program to translate evidence to clinical practice in type 2 diabetes. Diabet Med 2009; 26: 693–699.

12.

Raboca

Sobrepena

. Comprehensive risk assessments of diabetic patients from seven Asian countries: the Joint Asia Diabetes Evaluation (JADE) program. J Diabetes 2011; 3: 109–118.

13.

Zuo

Chen

. Modified glomerular filtration rate estimating equation for Chinese patients with chronic kidney disease. J Am Soc Nephrol 2006; 17: 2937–2944.

14.

EuroQol Group. EuroQol – a new facility for the measurement of health-related quality of life. Health Policy 1990; 16: 199–208.

15.

Grandy

Fox

. EQ-5D visual analog scale and utility index values in individuals with diabetes and at risk for diabetes: findings from the study to help improve early evaluation and management of risk factors leading to diabetes (SHIELD). Health Qual Life Outcomes 2008; 6: 18.

16.

Dolan

. Modeling valuations for EuroQol health states. Med Care 1997; 35: 1095–1108.

17.

Asian-Pacific Type 2 Diabetes Policy Group. Type 2 diabetes – practical targets and treatments. 4th ed. Melbourne, VIC, Australia: The International Diabetes Institute (IDI); Singapore: In Vivo Communications (Asia) Pte Limited, 2005.

18.

Kong

Yang

. Effects of treatment targets on subsequent cardiovascular events in Chinese patients with type 2 diabetes. Diabetes Care 2007; 30: 953–959.

19.

Yang

Kong

. Development and validation of a total coronary heart disease risk score in type 2 diabetes mellitus. Am J Cardiol 2008; 101: 596–601.

20.

Yang

Kong

. Development and validation of stroke risk equation for Hong Kong Chinese patients with type 2 diabetes: the Hong Kong Diabetes Registry. Diabetes Care 2007; 30: 65–70.

21.

Tseng

. Prevalence and risk factors of peripheral arterial obstructive disease in Taiwanese type 2 diabetic patients. Angiology 2003; 54: 331–338.

22.

Rhee

Guan

Liu

. Multi-country study on the prevalence and clinical features of peripheral arterial disease in Asian type 2 diabetes patients at high risk of atherosclerosis. Diabetes Res Clin Pract 2007; 76: 82–92.

23.

Thomas

Critchley

Tomlinson

. Peripheral vascular disease in type 2 diabetic Chinese patients: associations with metabolic indices, concomitant vascular disease and genetic factors. Diabet Med 2003; 20: 988–995.

24.

Natsuaki

Inoguchi

Maeda

. Association of borderline ankle-brachial index with mortality and the incidence of peripheral artery disease in diabetic patients. Atherosclerosis 2014; 234: 360–365.

25.

Fowkes

Murray

Butcher

. Ankle brachial index combined with Framingham Risk Score to predict cardiovascular events and mortality: a meta-analysis. JAMA 2008; 300: 197–208.

26.

Chronic Kidney Disease Prognosis Consortium; Matsushita

van der Velde

Astor

. Association of estimated glomerular filtration rate and albuminuria with all-cause and cardiovascular mortality in general population cohorts: a collaborative meta-analysis. Lancet 2010; 375: 2073–2081.

27.

Tong

Kong

. Interactive effect of retinopathy and macroalbuminuria on all-cause mortality, cardiovascular and renal end points in Chinese patients with type 2 diabetes mellitus. Diabet Med 2007; 24: 741–746.

28.

McDermott

Greenland

Liu

. Leg symptoms in peripheral arterial disease: associated clinical characteristics and functional impairment. JAMA 2001; 286: 1599–1606.

29.

Maksimovic

Vlajinac

Marinkovic

. Health-related quality of life among patients with peripheral arterial disease. Angiology 2014; 65: 501–506.

30.

Korhonen

Seppala

Kautiainen

. Ankle-brachial index and health-related quality of life. Eur J Prev Cardiol 2012; 19: 901–907.

31.

McDermott

Guralnik

Tian

. Associations of borderline and low normal ankle-brachial index values with functional decline at 5-year follow-up: the WALCS (Walking and Leg Circulation Study). J Am Coll Cardiol 2009; 53: 1056–1062.

32.

Lenfant

Chobanian

Jones

. Seventh report of the Joint National Committee on the Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC 7): resetting the hypertension sails. Hypertension 2003; 41: 1178–1179.

33.

Clairotte

Retout

Potier

. Automated ankle-brachial pressure index measurement by clinical staff for peripheral arterial disease diagnosis in nondiabetic and diabetic patients. Diabetes Care 2009; 32: 1231–1236.