Association between cardiorespiratory fitness and impaired vascular function in type 2 diabetes

Abstract

BACKGROUND

Cardiorespiratory fitness (CRF) is inversely associated with the risk of cardiovascular disease, which is related to impaired vascular function. However, its relationship with vascular function remains unknown in patients with type 2 diabetes.

AIM

To assess the relationship of CRF with vascular function in type 2 diabetes.

METHODS

Patients with type 2 diabetes who were aged ≥ 18 years and underwent an incremental and symptom-limited exercise test were included. Vascular function was assessed by the construction of the vascular health index (VHI), which is defined as a composite score of ankle-brachial index, transcutaneous oxygen pressure, pulse wave velocity, and carotid intima-media thickness. Impaired vascular function is defined as a VHI of < 8 points. Linear and logistic regression analyses were used to assess the associations.

RESULTS

We included 343 patients with type 2 diabetes. CRF was positively correlated with VHI (β = 0.10, P = 0.047), particularly with ankle-brachial index and pulse wave velocity. The odds ratio (OR) of impaired vascular function was 0.44 [95% confidence interval (CI): 0.20-0.96] for the highest vs the lowest CRF category. For each one metabolic equivalent increase in CRF, the OR of impaired vascular function was 0.73 (95%CI: 0.57-0.93).

CONCLUSION

Higher CRF was associated with better vascular function and lower odds of impaired vascular function in patients with type 2 diabetes.

Key Words: Cardiorespiratory fitness; Type 2 diabetes; Vascular function; Ankle-brachial index; Pulse wave velocity

Core Tip: This study developed for the first time a new index for the assessment of vascular function, which incorporates measures related to microvascular function, macrovascular function, arterial stiffness, and vascular morphology. We found that higher cardiorespiratory fitness (CRF) was associated with better vascular health in type 2 diabetes, particularly with better macrovascular function. Our study provides evidence in support of the beneficial effect of exercise training, which is associated with improved CRF, in the management of cardiovascular diseases in patients with type 2 diabetes.

INTRODUCTION

Vascular function reflects the ability of efficient delivery of blood to peripheral organs and is key to the maintenance of vascular homeostasis[1]. Impairment in vascular function may contribute to the development of atherosclerosis via the mechanisms of impaired coronary perfusion, microvascular damage, and abnormal hemodynamics, which ultimately leads to increased risks of cardiovascular diseases[2]. Noninvasive measures, such as the ankle-brachial index (ABI), transcutaneous oxygen pressure (TcPO₂), and carotid intima-media thickness (cIMT), are commonly used in clinical practice to assess vascular function in patients with type 2 diabetes. However, these measures may not provide a comprehensive assessment of vascular function, as they mainly refer to an individual component of vascular function. For instance, ABI is mainly considered an indicator of macrovascular function[3-5], while TcPO₂ reflects microvascular function[6]. Moreover, despite a close correlation between ABI and TcPO₂, there is the evidence that some patients with type 2 diabetes might have an abnormal ABI but show a normal TcPO₂ in clinical practice. As a result, combining multiple indicators to develop a new index to enable a comprehensive assessment of vascular function is of clinical interest for patients with type 2 diabetes, who are at high risks for cardiovascular diseases[7].

Cardiorespiratory fitness (CRF) is an indicator of the ability of the circulatory and respiratory systems to deliver oxygen to the working muscles during exercise[8], and is proposed as the fifth vital sign in the assessment of overall health[9]. CRF is associated with increased risks of cardiovascular diseases in patients with type 2 diabetes. For example, studies showed that for every one-metabolic equivalent (MET) increase in CRF, the risks of coronary heart disease and heart failure were reduced by 21% and 20%, respectively[10,11]. However, to date there have been no studies examining the relationship between CRF and vascular function, which was shown to be implicated in the development of cardiovascular diseases[12], in patients with type 2 diabetes.

Therefore, the aim of this study was to assess the association between CRF and vascular function as well as its potential influencing factors based on the construction of a new index of vascular function by incorporating ABI, TcPO₂, pulse wave velocity (PWV), and cIMT together in patients with type 2 diabetes.

MATERIALS AND METHODS

Study design and population

This cross-sectional study enrolled patients with type 2 diabetes who were admitted to the Department of Endocrinology of Zhongda Hospital from 2015 to 2018. The study protocol was approved by the Ethics Review Committee of Zhongda Hospital (approval No. 2019ZDSYLL119-P01). Written informed consent was waived as the data analyzed in the present study were retrospectively collected.

In this study we included participants who were diagnosed with type 2 diabetes, aged ≥ 18 years, completed an incremental and symptom-limited exercise test, and had all measurements of ABI, TcPO₂, PWV, and cIMT. Participants were excluded if they: (1) Were diagnosed with type 1 diabetes or other specific types of diabetes; (2) Were pregnant women; and (3) Had missing information on any of the measures of vascular function. In total, 343 patients with type 2 diabetes were included in the study (Figure 1).

Figure 1  Study flowchart.

Measurement of CRF

An incremental and symptom-limited exercise test was conducted using a cycling ergometer to assess CRF under the supervision of an expert physician. Patients started pedaling on command at a cadence of 50-60 rpm with the work load set to be “25 W” for females and “50 W” for males. The work load was increased by 25 W every 3 minutes, and the test was terminated when symptoms were reported (e.g., fatigue, dyspnea, or others). Peak oxygen uptake was measured by the COSMED K4b2 system, and CRF was determined as the peak oxygen uptake/weight (kg)/3.5 and expressed in the unit of MET.

Covariates and blood biomarkers

General information including gender, age, and history of drinking and smoking was collected from the electronic medical records. Height, weight, systolic blood pressure (SBP), and diastolic blood pressure (DBP) were measured using standard protocols. Body mass index (BMI) was calculated as weight (kg)/height (m²). Hypertension was defined as SBP ≥ 140 mmHg, DBP ≥ 90 mmHg, the use of anti-hypertension medications, and/or the history of hypertension.

Fasting blood samples were obtained to measure glycated hemoglobin A1c (HbA1c), total cholesterol, triglycerides (TG), high-density lipoprotein-cholesterol (HDL), low-density lipoprotein-cholesterol (LDL), and creatine. The estimated glomerular filtration rate (eGFR) was calculated using the modification of diet in renal disease Chinese modified formula: 175 × creatine^-1.234 × age^{- 0.179} × (female × 0.79)[13].

Measurement of vascular function indicators

ABI: Patients were instructed to lie in the supine position. A pneumatic cuff was applied to the arms and ankles, and the SBPs were measured at the brachial, posterior tibial and dorsalis pedis arteries in both limbs using a hand-held continuous-wave Doppler probe. The ABI is defined as the ratio of ankle arterial pressure to brachial arterial pressure for each limb. In this study, the lower ABI value was selected for analysis. In case this was available for only one limb, the value from that limb was used for analysis.

TcPO₂: Patients were asked to lie in the supine position, and the measurement of TcPO₂ was performed at the dorsolateral site of both feet. The electrodes were applied to the skin and the measurements were obtained after calibration and preheating of the electrodes to approximately 44 °C. The lower TcPO₂ value was selected for analysis. If the value was available for only one foot, it was then used for analysis.

cIMT: CIMT was determined using high-resolution Doppler ultrasound scanning approximately 2 cm proximal and distal to the dilated common carotid artery. In this study the averages from both sides were calculated for analysis.

PWV: According to the Reference Values for Arterial Stiffness Collaboration[14], PWV was estimated using a validated formula based on age and mean blood pressure (MBP): 9.587 - 0.402 × age + 4.560 × 10^-3 × age² - 2.621 × 10^-5 × age² × MBP + 3.176 × 10^-3 × age × MBP - 1.832 × 10^-2 × MBP, in which MBP was calculated as DBP + 0.4 × (SBP - DBP). The formula was derived from the reference population (individuals presenting cardiovascular risk factors that included type 2 diabetes) and was validated in patients with type 2 diabetes due to its close relationship with all-cause mortality[15,16].

Calculation of vascular health index

With reference to the approach for the construction of the Healthy Aging Score[17], vascular function was assessed by the construction of the vascular health index (VHI), which is defined as a composite score of ABI, TcPO₂, PWV, and cIMT. For this, we assigned a score of 1 (worst) to 3 (best) points for each of the measures based on their tertiles. This composite score ranges from 4 to 16 points, with a higher score indicative of better vascular function. In this study, we defined a VHI of < 8 points as impaired vascular function. The details of the construction of the VHI are shown in Supplementary Table 1 and the performance of VHI over the individual vascular function indicators in identifying macrovascular dysfunction (defined as ABI < 1.0) and microvascular dysfunction (defined as TcPO₂ < 60 mmHg)[18,19] by the area under the curves is shown in Supplementary Figure 1.

Statistical analysis

Continuous data are presented as the median and interquartile range (25^th, 75^th percentile) or the means and standard deviations, based on the normality test. Categorical data are shown as numbers and percentages (%). The nonparametric test or χ² test was used to compare the differences between groups, where appropriate. The VHI was analyzed using two approaches: (1) Being categorized as impaired and normal vascular function groups; and (2) Being treated as a continuous variable. Similarly, CRF was also analyzed as a categorial variable (that is, lowest, middle, and highest tertiles) or a continuous variable. Linear regression analysis was used to assess the association between CRF and VHI and each parameter for vascular function. Additionally, logistic regression analysis was conducted to evaluate the relationship between CRF and the odds of impaired vascular function, which was expressed as odds ratios (ORs) and 95% confidence intervals (CIs). To account for the potential confounders, we employed three models: Model 1 was without adjustment, model 2 was adjusted for age and gender, and model 3 was additionally adjusted for history of smoking and drinking, SBP, BMI, eGFR, HDL, TG, and HbA1c. In this study, missing data such as TG, HDL, LDL, HbA1c, and eGFR were imputed by the Markov Chain Monte Carlo method.

Subgroup analysis was conducted to assess the influences of age (≥ 60 years vs < 60 years), gender (male vs female), duration of diabetes (< 1 years, 1-10 years, vs > 10 years), history of smoking (yes vs no), history of drinking (yes vs no), BMI (≥ 28 kg/m², 24-28 kg/m², vs < 24 kg/m²), and LDL (≥ 2.6 mmol/L vs < 2.6 mmol/L), on the association between CRF and the odds of impaired vascular function. Sensitivity analysis was performed by excluding participants with missing data and re-categorizing CRF from tertiles to quartiles. All the analyses were performed using Stata (version 14.0, College Station, TX, United States), and a 2-sided P < 0.05 was considered statistically significant.

RESULTS

Characteristics of study population

Table 1 shows the characteristics of the included 343 participants with type 2 diabetes. Their mean age was 53.3 ± 9.8 years and the majority of them were male (63.6%). Compared to patients with normal vascular function, those with impaired vascular function had higher age, SBP, and BMI (all P < 0.05) and lower eGFR (P < 0.05). Moreover, CRF was lower in patients with impaired vascular function [5.9 (5.0, 6.7) MET vs 5.2 (4.5, 6.0) MET, P < 0.001]. In contrast, participants with the highest tertile of CRF had the largest VHI (8.9 ± 1.4 vs 8.2 ± 1.6, P < 0.01; Supplementary Figure 2).

Table 1 Characteristics of participants stratified by vascular function, n (%).

Variables	Total	Normal vascular function	Impaired vascular function	P value
Sample size	343	255	88
Age (years)	54 (48, 59)	52 (45, 57)	59 (55, 65)	< 0.001
Gender				0.45
Male	218 (63.6)	165 (64.7)	53 (60.2)
Female	125 (36.4)	90 (35.3)	35 (39.8)
Smoking				0.64
Yes	128 (37.3)	97 (38.0)	31 (35.2)
No	215 (62.7)	158 (62.0)	57 (64.8)
Drinking				0.42
Yes	89 (25.9)	69 (27.1)	20 (22.7)
No	254 (74.1)	186 (72.9)	68 (77.3)
Hypertension				< 0.001
Yes	186 (54.2)	123 (48.2)	63 (71.6)
No	157 (45.8)	132 (51.8)	25 (28.4)
SBP (mmHg)	120 (120, 130)	120 (115, 126)	120 (112, 124)	< 0.001
DBP (mmHg)	80 (70, 80)	80 (70, 80)	80 (70, 80)	0.83
BMI (kg/m2)	25.0 (22.8, 30.1)	24.9 (22.5, 27.1)	25.9 (24.0, 28.6)	0.02
HbAlc (%)1	9.4 (7.8, 11.2)	9.4 (7.8, 11.3)	9.0 (7.5, 11.0)	0.25
TG (mmol/L)1	1.71 (1.06, 2.59)	1.79 (1.07, 2.68)	1.57 (1.05, 2.25)	0.23
TC (mmol/L)1	4.9 (4.2, 5.7)	5.0 (4.2, 5.7)	4.8 (4.3, 5.8)	0.54
HDL (mmol/L)1	1.2 (1.0, 1.4)	1.2 (1.0, 1.4)	1.2 (1.0, 1.4)	0.37
LDL (mmol/L)1	3.0 (2.5, 3.6)	3.0 (2.5, 3.5)	3.0 (2.6, 3.9)	0.49
eGFR (mL/minute/1.73 m2)1	90.9 (79.6, 104.0)	92.6 (81.3, 105.3)	87.9 (77.0, 98.7)	0.02
CRF (MET)	5.7 (4.8, 6.6)	5.9 (5.0, 6.7)	5.2 (4.5, 6.0)	< 0.001

¹There were 28 participants without information on estimated glomerular filtration rate, triglycerides and total cholesterol, 29 without high-density lipoprotein-cholesterol and low-density lipoprotein-cholesterol, and 24 without hemoglobin A1c.

SBP: Systolic blood pressure; DBP: Diastolic blood pressure; BMI: Body mass index; HbA1c: Hemoglobin A1c; TG: Triglycerides; TC: Total cholesterol; HDL: High-density lipoprotein-cholesterol; LDL: Low-density lipoprotein-cholesterol; eGFR: Estimated glomerular filtration rate; CRF: Cardiorespiratory fitness; MET: Metabolic equivalent.

Association between CRF and VHI

Table 2 shows the linear relationship between CRF and VHI. In the crude model (model 1), CRF was significantly and positively correlated with the VHI (β = 0.20, P < 0.001). The relationship remained significant after controlling for different variables (model 2 and model 3, both P < 0.05). Furthermore, after controlling for multivariable factors, CRF was positively correlated with TcPO₂ but inversely with PWV (β = 0.13 and -0.11, respectively; model 3). However, there was no significant association of CRF with ABI or cIMT (both P > 0.05).

Table 2 Association between cardiorespiratory fitness and vascular function.

Variables	Model 1		Model 2		Model 3
	Sβ	P value	Sβ	P value	Sβ	P value
VHI	0.20	< 0.001	0.10	0.045	0.10	0.047
ABI	0.07	0.20	0.04	0.53	0.08	0.18
TcPO2	0.16	0.004	0.16	0.01	0.13	0.03
PWV	-0.24	< 0.001	-0.08	0.02	-0.11	< 0.001
cIMT	-0.07	0.21	-0.04	0.53	-0.03	0.58

Model 1 was a crude model. Model 2 was adjusted for age and gender. Model 3 was adjusted for age, gender, history of smoking and drinking, systolic blood pressure, body mass index, estimated glomerular filtration rate, high-density lipoprotein-cholesterol, triglycerides, and hemoglobin A1c. VHI: Vascular health index; ABI: Ankle brachial index; TcPO₂: Transcutaneous oxygen pressure; PWV: Pulse wave velocity; cIMT: Carotid intima media thickness; Sβ: Standardized regression coefficient.

Association between CRF and impaired vascular function

Table 3 shows the association between CRF and impaired vascular function. In the crude model (model 1), the ORs for impaired vascular function in the middle and the highest tertiles of CRF vs the lowest tertile were 0.58 (95%CI: 0.33-1.01) and 0.34 (95%CI: 0.18-0.64), respectively. These associations were slightly changed after controlling for different variables (that is, model 2 and model 3). Dose-response analysis showed that there was no significant evidence of a departure from the non-linear relationship (P_{non-linearity} = 0.06) regarding the association between CRF and odds of impaired vascular function. Subsequent analysis showed that the OR for impaired vascular function was 0.73 (95%CI: 0.57-0.93) per one-MET increase in CRF (Table 3).

Table 3 Association between cardiorespiratory fitness and the odds of impaired vascular function.

Variable	No. of cases/total	Model 1		Model 2		Model 3
		OR (95%CI)	P value	OR (95%CI)	P value	OR (95%CI)	P value
CRF tertiles (MET)
Lowest (< 5.2)	42/117	1 (reference)		1 (reference)		1 (reference)
Middle (5.2-6.2)	29/119	0.58 (0.33-1.01)	0.06	0.73 (0.40-1.35)	0.32	0.68 (0.35-1.32)	0.25
Highest (> 6.2)	17/107	0.34 (0.18-0.64)	0.001	0.48 (0.24-0.98)	0.04	0.44 (0.20-0.96)	0.04
Per 1-MET increase	88/343	0.69 (0.57-0.84)	< 0.001	0.75 (0.61-0.94)	0.01	0.73 (0.57-0.93)	0.01

Model 1 was a crude model. Model 2 was adjusted for age and gender. Model 3 was adjusted for age, gender, history of smoking and drinking, systolic blood pressure, body mass index, estimated glomerular filtration rate, high-density lipoprotein-cholesterol, triglycerides, and hemoglobin A1c. CRF: Cardiorespiratory fitness; MET: Metabolic equivalent; OR: Odds ratio; CI: Confidence interval.

Subgroup and sensitivity analysis

Table 4 shows the subgroup analysis for the association between CRF and the odds of impaired vascular function. It appears that this association was more pronounced in some subgroups, e.g., among patients with older age, males, non-smokers, and non-drinkers, after multivariable-adjustment (all P < 0.05). However, none of the variables, which included age, gender, duration of diabetes, history of smoking, history of drinking, BMI, and LDL, showed any significant moderating effects (all P_interaction > 0.05). Sensitivity analysis showed that the associations remained comparable in general upon the exclusion of participants with missing data (Supplementary Table 2) or re-categorizing CRF into quartiles (Supplementary Table 3).

Table 4 Subgroup analysis of cardiorespiratory fitness and the odds of impaired vascular function.

Variables	Model 1		Model 2		Model 3		Pinteraction1
	OR (95%CI)	P value	OR (95%CI)	P value	OR (95%CI)	P value
Age (years)							0.68
< 60	0.77 (0.60-0.98)	0.04	0.76 (0.59-0.99)	0.04	0.74 (0.54-1.00)	0.05
≥ 60	0.67 (0.47-0.96)	0.03	0.66 (0.45-0.95)	0.03	0.58 (0.36-0.92)	0.02
Gender							0.30
Male	0.62 (0.47-0.82)	0.001	0.66 (0.50-0.88)	0.004	0.62 (0.45-0.87)	0.01
Female	0.79 (0.57-1.09)	0.15	0.84 (0.59-1.19)	0.32	0.80 (0.54-1.20)	0.28
Duration of diabetes (years)2							0.55
< 1	0.96 (0.60-1.55)	0.87	0.74 (0.52-1.04)	0.08	0.66 (0.46-0.95)	0.03
1-10	0.98 (0.58-1.68)	0.95	0.75 (0.51-1.09)	0.13	0.67 (0.45-0.99)	0.04
≥ 10	1.40 (0.54-3.61)	0.48	0.76 (0.46-1.26)	0.29	0.56 (0.36-0.88)	0.01
Smoking							0.82
Yes	0.75 (0.54-1.05)	0.09	0.76 (0.54-1.06)	0.11	0.75 (0.47-1.20)	0.23
No	0.66 (0.51-0.85)	0.001	0.70 (0.53-0.93)	0.01	0.63 (0.46-0.87)	0.003
Drinking							0.73
Yes	0.80 (0.55-1.16)	0.25	0.79 (0.53-1.18)	0.25	0.81 (0.50-1.31)	0.40
No	0.65 (0.51-0.83)	0.001	0.69 (0.53-0.89)	0.01	0.66 (0.50-0.88)	0.01
BMI (kg/m2)							0.19
< 24	0.66 (0.47-0.92)	0.01	0.67 (0.46-0.98)	0.04	0.57 (0.36-0.89)	0.01
24-28	0.85 (0.62-1.17)	0.33	0.91 (0.63-1.31)	0.61	0.86 (0.55-1.35)	0.52
≥ 28	0.56 (0.35-0.89)	0.01	0.67 (0.40-1.10)	0.12	0.61 (0.33-1.14)	0.12
LDL (mmol/L)							0.72
< 2.6	0.64 (0.45-0.93)	0.02	0.57 (0.37-0.89)	0.01	0.51 (0.30-0.85)	0.01
≥ 2.6	0.71 (0.56-0.90)	0.01	0.78 (0.61-1.00)	0.05	0.75 (0.56-1.00)	0.05

¹P_interaction values are derived from Model 3.

²There were 259 data available for analysis.

Model 1 was a crude model. Model 2 was adjusted for age and gender. Model 3 was adjusted for age, gender, history of smoking and drinking, systolic blood pressure, body mass index, estimated glomerular filtration rate, high-density lipoprotein-cholesterol, triglycerides, and hemoglobin A1c (except the variables used for subgrouping). HbA1c: Glycated hemoglobin A1c; OR: Odds ratio; CI: Confidence interval.

DISCUSSION

To our knowledge, this study is the first cross-sectional analysis to examine the relationship between CRF and vascular function in patients with type 2 diabetes by developing a new index. Our results showed a close relationship between CRF and vascular function, as indicated by the positive association of CRF with VHI as well as by the inverse association of higher CRF with lower odds of impaired vascular function in type 2 diabetes. This association was independent of differences in age, gender, history of hypertension, history of smoking, history of drinking, and BMI.

Previous studies have shown that increased physical activity is associated with favorable ABI and PWV; however, these significant associations do not remain consistent across different subgroups of populations[20-23]. Moreover, the relationship between physical activity and cIMT remains controversial, with some studies showing that increased physical activity was associated with decreases in cIMT, while others reported no significant association[24,25]. Notably, physical activity in these studies was mainly measured by self-reported questionnaires, which might be subject to recall or social desirability bias[26]. In contrast, CRF, which is closely related to physical activity, was objectively measured using an incremental and symptom-limited exercise test in the present study. Our subsequent analysis showed that CRF was positively correlated with TcPO₂, and negatively correlated with PWV in patients with type 2 diabetes. However, we did not find any significant association with ABI or cIMT.

Of note, in this study we constructed a new composite index of VHI for the evaluation of vascular function by integrating ABI (representative of macrovascular function), TcPO₂ (representative of microvascular function), PWV (representative of arterial stiffness), and cIMT (representative of vascular morphology) together. This is of clinical relevance, as it provides a comprehensive approach to assess vascular health from different aspects. Moreover, we found that this index was superior to all its components in identifying macrovascular and microvascular dysfunction as evidenced by its discriminability in the receiver operating characteristic curves (Supplementary Figure 1). Similar to our study, a previous study developed a grading system named “Beijing Vascular Health Stratification” for vascular health assessment by combining four vascular measures together, which included cardio-ankle vascular index, ABI, carotid femoral-PWV, and carotid radial-PWV[27,28]. However, this grading system did not include any measures that reflect microvascular function (that is, TcPO₂) or vascular morphology (that is, cIMT). Furthermore, it did not provide clear evidence that the grading system outperformed its component in predicting cardiovascular outcomes[27,28].

In this study, we found that CRF was positively correlated with VHI, and was associated with a decrease in the odds of impaired vascular function in type 2 diabetes dose-dependently. However, our categorial analysis showed that the middle tertile of CRF (5.2-6.2 METs) or the middle quartile (4.9-5.7 METs) was not associated with any decreased odds of impaired vascular function. Together with the evidence that the risk of cardiovascular diseases was significantly lower in individuals with a CRF of ≥ 7.9 METs than those with a CRF of < 7.9 METs[29], it appears likely that there may exist a threshold effect for the preventive effect of CRF. However, it should be noted that a lack of improvement in CRF following the exercise intervention does not necessarily entail a lack of improvement in vascular function among patients with type 2 diabetes[30].

Our study is the first to explore the association between CRF and various indicators of vascular function in patients with type 2 diabetes. Moreover, both the CRF and the indicators of vascular function including ABI, TcPO₂ and cIMT were measured directly and objectively, in particular, CRF was quantified by the gold standard method using a gas exchanger based on an incremental and symptom-limited exercise test. Furthermore, vascular function was analyzed by creating a new index with the integration of different items related to microvascular function, macrovascular function, morphology, and arterial stiffness.

Our study also has some limitations. First, due to the nature of the cross-sectional study, the causality of CRF with impaired vascular function cannot be inferred and the reverse causation cannot be ruled out either. Second, the sample size of our study was relatively small, but was larger than most of studies that focused on the analysis of TcPO₂[31,32]. Third, arterial stiffness was assessed by estimating PWV, and it is unclear whether CRF shows any association with flow-mediated dilatation in Chinese patients with type 2 diabetes, which is another measure which reflects vascular function, particularly arterial stiffness[33]. Fourth, although the formula for PWV estimation was validated in individuals with cardiovascular risk factors that included type 2 diabetes[15,16], the lack of a population-specific formula (e.g., for patients with type 2 diabetes only) might potentially affect its association with CRF. Fifth, despite the effort to control for multivariable factors, we could not completely exclude the possibility of unmeasured effects from some factors such as medication use. Finally, our study did not include patients who had contraindications in performing the incremental and symptom-limited exercise test, e.g., patients with diabetic ulcerations. This may potentially limit the generalization of our findings to a broader spectrum of patients with type 2 diabetes (e.g., those who cannot perform the exercise test).

CONCLUSION

Our study shows that higher CRF was associated with better vascular function as well as with lower odds of impaired vascular function in patients with type 2 diabetes. This is of clinical importance, as it could provide some evidence or explanations in support of the beneficial effect of exercise training, which is associated with improved CRF, in the management of cardiovascular diseases in patients with type 2 diabetes.