A pilot study of regional perfusion and oxygenation in calf muscles of individuals with diabetes with a noninvasive measure

Abstract

Objective

To assess alterations in the regional perfusion and oxygenation of the calf muscles in individuals with diabetes.

Methods

Age-matched individuals with (n = 5) and without diabetes (n = 6) were investigated. Skeletal muscle perfusion, oxygen extraction fraction, and oxygen consumption rate were measured by newly developed noncontrast magnetic resonance imaging (MRI) techniques. The subjects lay supine on the MRI table with their foot firmly strapped to a custom-built isometric exercise device. The measurements were performed at rest and during an isometric plantar flexion muscle contraction.

Results

Individuals without diabetes had up to a 10-fold increase in muscle perfusion, 25% elevation in muscle oxygen extraction fraction, and a 12-fold increase in oxygen consumption rate in the calf during the plantar flexion isometric contraction. In patients with diabetes, the increases in these parameters were only up to sixfold, 2%, and sixfold, respectively. Exercise oxygen consumption rate was inversely associated with blood HbA1c levels (r² = .91).

Conclusions

This is the first study to quantify regional skeletal muscle oxygenation in patients with diabetes using noncontrast MRI and warrants additional study. Attenuation of perfusion and oxygenation during exercise may have implications for understanding diabetic complications in the lower extremities.

Complications affecting the lower extremity, such as ulceration, neuropathy, and structural changes of the foot, are a major cause of morbidity associated with diabetes mellitus (DM). Circulatory compromise to the extremity is known to have a major impact on the development of these complications.¹ However, the effects of DM on the circulatory system and the impact of those changes on the delivery of blood to the tissues of the leg are only crudely understood. It is well recognized that individuals with DM are prone to relatively distinct clinical manifestations of dysfunction of both the macrocirculation and microcirculation. These DM-related changes include calcific atherosclerosis, particularly of the tibial vessels, reduced capillary size, altered endothelial function, and thickening of the capillary basement membranes.² However, the underlying mechanisms, and in particular, the relationship between the microvascular disease, macrovascular disease, and the lower extremity complications have not been well defined. In part, this has not been explored because of the lack of reproducible quantifiable measures of end-organ perfusion.

Traditionally, the macrocirculation was assessed by ankle/brachial index, contrast imaging of the vessels, and ultrasound duplex.³ These assessments are reproducible and clinically valuable to evaluate and plan treatment for stenotic/occlusive disease in the larger vessels. Unfortunately, there does not yet exist a similar clinically valuable technique to assess the microcirculatory changes that may also have an important effect on the functional perfusion of the tissues. A number of noninvasive imaging techniques have been used to assess local skin perfusion or oxygenation (<3 mm depth), such as capillaroscopy,^4,5 thermography,⁶ laser Doppler flowmetry,⁷ laser Doppler imaging,⁸ transcutaneous O₂ tension,⁹ or orthogonal polarization spectral imaging.¹⁰ These imaging techniques, however, have relatively low spatial resolution, reliability, reproducibility, and sensitivity, and none has been widely adopted into clinical practice. The techniques also are limited to the skin, and thus do not provide information on regional muscle perfusion or oxygenation, which may have an impact on the function of the leg and foot.¹¹ Another technique that can assess peripheral microcirculation is contrast-enhanced ultrasound.^12,13 However, it usually provides semiquantitative measurements without any information about oxygen utility.

Magnetic resonance imaging (MRI) is a noninvasive imaging modality that provides excellent soft tissue contrast and has the capability for detailed delineation of anatomy, perfusion, and metabolism in skeletal muscle.^14–18 We anticipate that delineation of the lower extremity perfusion with this modality could lead to important insights into the impact of microvascular changes on the development of DM-related complications in the limb. Recently, we have developed a new MRI method to assess skeletal muscle perfusion, also called skeletal muscle blood flow (SMBF), and oxygen extraction fraction (SMOEF). The latter is defined as ([O₂]_artery − [O₂]_vein)/[O₂]_artery and represents the relationship between skeletal muscle oxygen supply and demand. A lower number represents lower oxygen extraction by the muscle. The feasibility of the measurement was demonstrated in healthy young volunteers.¹⁹ In this pilot study, feasibility for the assessment of regional muscle perfusion and oxygenation is demonstrated in DM patients without known macrovascular disease, in comparison with age-matched healthy volunteers. It is hypothesized that the microvascular changes related to DM result in impairment of the regional skeletal muscle perfusion and oxygenation.

METHODS

Patients

Six healthy volunteers (age, 70 ± 3 years old; body mass index, 32.3 ± 8.9 kg/m²) and five nonsmoking DM patients (age, 66 ± 5 years old; body mass index, 37.9 ± 7.1 kg/m²; HbA1c, 7.7% ± 2.0%; type 1, n =1; type 2, n =4) were recruited for the measurement of perfusion and oxygenation in the calf muscle. The healthy volunteers were nonsmokers, free of cardiovascular, metabolic, and musculoskeletal diseases, and did not have a history of DM or peripheral neuropathy. These volunteers were screened by a questionnaire regarding their history and symptoms. For the DM patients, the mean duration of DM was 7.4 ± 7.3 years. Three DM patients (type 1, n =1; type 2, n =2) had peripheral neuropathy, determined by the inability to sense the 5.07 Semmes-Weinstein mono-fllament and no vibration perception <25 volts measured at the plantar great toe. None of them had a history or current plantar ulcer. All of the DM subjects had a history of cardiovascular disease (4 with hypertension, 2 with cardiac artery bypass graft, and 2 with a myocardial infarction), but none of the subjects had a documented history of peripheral vascular disease. Activity level and exercise capacity were not measured in these groups of subjects but based on previous study of a similar subject population; we expect that the DM subjects had lower physical performance compared with the healthy controls.²⁹ The local human study committee approved this study, and signed consent forms were received from all volunteers prior to the imaging sessions.

Data collection

Subjects were instructed to not consume alcohol or perform any moderate to heavy exercise 24 hours prior to the imaging session. Each subject was positioned supine on the MRI table with his or her right foot firmly strapped to a pedal of a custom-built isometric exercise device.¹⁹ The resistance of the pedal to depression was adjusted on an individual basis (mean resistance force, 67.7 ± 6.5 N) to allow the subject to completely depress the pedal for the duration of the scan (up to 6 minutes).

Prior to the exercise study, a phase-array cardiac coil was placed between the knee and the heel to cover the lower legs. A three-dimensional noncontrast MR angiography was then acquired,²⁰ which was used to detect any hemodynamically significant stenosis (≥50%) in large peripheral vessels (anterior tibial artery, posterior tibial artery, and peroneal artery). This approach was established recently with superior sensitivity and specificity.²¹ The MRI perfusion and oxygenation measurements were subsequently performed at rest and during a sustained isometric contraction of the plantar flexor muscles, which included the gastrocnemius and soleus muscles. Specifically, each measurement started at 2 minutes after the start of contraction, resulting in a total contraction time of approximately 3 minutes for a perfusion measurement and 6 minutes for an oxygenation measurement (see Imaging Methods section). There were few reports about the steady-state time for muscle perfusion and oxygenation during an isometric contraction. It was estimated that this steady-state time for leg muscle perfusion can be reached after 1 to 2 minutes of muscle contraction.²² There was a 5-minute rest interval after exercise measurements.

Imaging methods

All images were acquired on a Siemens 3T Trio whole-body scanner (Siemens Health-care, Malvern, Pa). A study using the methods here was reported recently in healthy young human subjects.³¹ To measure SMBF, an arterial spin-labeling method was adapted for skeletal muscle imaging.²³ The two-dimensional arterial spin-labeling sequence parameters included: gradient-echo acquisition; repetition time (TR)/ echo time (TE), 2.8/1.2 ms; 10 T₁-weighted images for each T₁ measurement; fiip angle, 5°; field of view (FOV), 160 × 112 mm²; data acquisition matrix, 128 × 90; data average, 3; acquisition time, 50 seconds. A single transverse slice was scanned in the middle of the calf muscle.

Skeletal muscle oxygenation is represented by both SMOEF and the oxygen consumption rate (SMVO₂). The latter can be determined using Fick’s law: SMVO₂ = [O₂]_a × SMOEF × SMBF. The constant [O₂]_a is defined as the total oxygen content of arterial blood, and a value of 8.61 μmol × mL⁻¹ was used for this pilot study.²⁴ The SMVO₂ provides an accurate measure of total oxygen metabolism in skeletal muscle. The MRI method for SMOEF measurement is based on a method developed for the brain, which relies on the magnetic susceptibility effect of intravascular deoxyhemoglobin.^25,26 A multi-slice two-dimensional triple-echo asymmetric spin-echo sequence was employed to acquire source images for SMOEF measurements.³⁶ The imaging parameters are: TR, 4 sec; TE₁/TE₂/TE₃, 44/62/80 ms; FOV, 160 × 140 mm²; data acquisition matrix size, 64 × 56 and interpolated to 128 × 112; single slice, slice thickness, 8 mm; total acquisition time, 3 minutes 48 seconds.

The three-dimensional noncontrast angiography uses images acquired during systole and diastole to remove the venous signals so that arteries can be clearly depicted. The typical imaging parameters for calf arterial imaging are: true fast imaging with steady-state precession TR/TE, 3.8 ms/1.63 ms; fiip angle, 500; segmentation number, 40; generalized autocalibrating partially parallel acquisitions acceleration factor, 2; FOV, 390 × 390; matrix, 432 × 432; 80 slices; isotropic resolution, 0.9 × 0.9 × 0.9 mm³; total acquisition, 6 to 7 minutes.

Image analysis

SMOEF and SMBF maps were created using a custom software written in Matlab (MathWorks, Natick, Mass).^19,24 Because the plantar flexion isometric contraction primarily affects the soleus and gastrocnemius muscles, region-of-interest measurements were performed in these two muscle regions (Fig 1).¹⁹ In addition, the two muscle regions were positioned closest to the surface coil, resulting in the highest SNR enhancing the accuracy of our measurement. The deep plantar flexion compartment muscles (ie, tibialis posterior, flexor halluces longus, and flexor digitorum longus) were excluded from analysis due to relatively lower signal-to-noise ratio. SMVO₂ data were then subsequently calculated. The hemodynamic reserves (ratio of exercise value to the resting value) from three parameters, SMBF, SMOEF, and SMVO₂, were calculated to gauge the exercise performance.

Fig 1.

Examples of skeletal muscle blood flow (SMBF) maps, at rest and during isometric contraction, from a healthy volunteer (A and B) and an age-matched diabetes mellitus (DM) patient (C and D). The yellow and red regions of interest in the SMBF map of (A) indicate the soleus and gastrocnemius muscle regions, respectively. Reduced SMBF in DM from the contraction can be easily seen in the subtracted map (D), compared with the map of a healthy volunteer in (B). Color bar range, 0 to 150 mL/100 g/min.

Statistical analysis

Data are reported as mean ± standard deviation. The comparison between the healthy and DM groups was made using independent two-sided Student t-tests with two samples and equal variance. The comparison between rest and isometric exercise was made using paired, two-sided Student t-tests. The relationship between quantitative MRI parameters and HbA1c was determined using an empirical exponential curve. Significance for all statistical tests was defined as P < .05.

RESULTS

All subjects completed the imaging protocol without difficulty. No apparent motion artifacts were noticed in any MR source images. None of the subjects had significant peripheral arterial stenosis as assessed from the non-contrast three-dimensional MR angiography.

Table I lists the average values of SMBF, SMOEF, and SMVO₂ for the gastrocnemius and soleus muscles. In healthy volunteers, the resting values of SMBF, SMOEF, and SMVO₂ in the soleus muscle were 7.3 ± .9 mL/ 100 g/min, .36 ± .04, and .49 ± .1 mL/100 g/min, respectively. These hemodynamic parameters all increased significantly during the isometric exercise to 57.8 ± 15.8 mL/100 g/min (P < .001), .43 ± .08 (P < .05), and 4.64 ± 1.4 mL/100 g/min (P < .01), respectively. The gastrocnemius muscle in healthy volunteers had similar resting and exercise values for these parameters.

Table I. A.

Hemodynamics measured by noncontrast magnetic resonance imaging (MRI) in soleus muscle

	SMBF (mL/100 g/min)		SMOEF		SMVO2 (mL/100 g/min)
Subjects	Rest	Contraction	Rest	Contraction	Rest	Contraction
Healthy (6)	7.3 ± .9	57.8 ± 12.1a	.36 ± .04	.43 ± .09a	.49 ± .10	4.62 ± 1.41a
DM (5)	5.8 ± .8b	34.7 ± 3.1a,b	.48 ± .15	.43 ± .10	.53 ± .21	2.76 ± .72a,b

DM, Diabetes mellitus; SMBF, skeletal muscle blood flow; SMOEF, skeletal muscle oxygen extraction fraction; SMVO₂, skeletal muscle oxygen consumption rate.

^aP < .05, contraction vs rest.

^bP < .05, DM vs healthy.

In the soleus muscle of the DM group (Table IA), resting SMBF was significantly lower than that in the healthy group (5.8 ± .8 mL/100 g/min vs 7.3 ± .9 mL/100 g/min in healthy; P =.03), but resting SMOEF and SMVO₂ were not significantly different from those found in the healthy volunteers. The exercise SMBF and SMVO₂ in the DM group were significantly lower than those in the healthy group (SMBF, 34.7 ± 3.1 mL/100 g/min vs 57.8 ± 12.1 mL/100 g/min in healthy; P =.003; SMVO₂, 2.76 ± .72 mL/100 g/min vs 4.62 ± 1.41 mL/100 g/min in healthy; P =.02). However, exercise SMOEF in the DM group was not significantly different from that in the healthy group.

In the gastrocnemius muscle of the DM group (Table IB), all resting SMBF, SMOEF, and SMVO₂ were not significantly different from those found in the healthy group. The exercise SMBF and SMVO₂ in the DM group were significantly lower than those in the healthy group (SMBF, 33.2 ± 12.5 mL/100 g/min vs 57.4 ± 11.6 mL/100 g/min in healthy; P < .01; SMVO₂, 2.98 ± 1.89 mL/100 g/min vs 5.45 ± 1.50 mL/100 g/ min in healthy; P =.04). Like in the soleus muscle, the exercise SMOEF in the gastrocnemius muscle of the DM group was the same as resting SMOEF, but not significantly different from the exercise SMOEF in the healthy group.

Table I. B.

Hemodynamics measured by noncontrast magnetic resonance imaging (MRI) in gastrocnemius muscle

	SMBF (mL/100 g/min)		SMOEF		SMVO2 (mL/100 g/min)
Subjects	Rest	Contraction	Rest	Contraction	Rest	Contraction
Healthy (6)	6.0 ± .8	57.4 ± 11.6a	.41 ± .05	.52 ± .09a	.45 ± .07	5.45 ± 1.50a
DM (5)	6.4 ± .5	33.2 ± 12.5a,b	.46 ± .20	.46 ± .16	.54 ± .20	2.98 ± 1.89a,b

DM, Diabetes mellitus; SMBF, skeletal muscle blood flow; SMOEF, skeletal muscle oxygen extraction fraction; SMVO₂, skeletal muscle oxygen consumption rate.

^aP < .05, contraction vs rest.

^bP < .05, DM vs healthy.

Table II show the reserves of each hemodynamic parameter. There were significant decreases in SMVO₂ reserve in the soleus and gastrocnemius muscles in the DM group compared with the healthy group (soleus, 5.72 ± 2.34 vs 9.59 ± 2.86 in healthy; P =.04; gastrocnemius, 5.66 ± 2.59 vs 12.56 ± 4.95 in healthy; P =.02). It is noticed, however, that there are regional differences between soleus and gastrocnemius muscles in terms of SMBF and SMOEF reserves. Both SMBF and SMOEF reserves in the gastrocnemius region of the DM group were significantly lower than those in the healthy group (SMBF, 5.54 ± 2.54 in DM vs 9.76 ± 2.77 in healthy; P =.02; SMOEF, 1.02 ± .07 in DM vs 1.27 ± .19 in healthy; P =.03), but not for the soleus region. Figs 1, A and B and 2 show maps of SMBF and SMOEF from one healthy and one type 2 DM patient with an HbA1c level of 6.4, respectively.

Table II. A.

Hemodynamic reserves measured by noncontrast magnetic resonance imaging (MRI) in soleus muscle

Subjects	SMBF reserve	SMOEF reserve	SMVO2 reserve
Healthy (6)	8.03 ± 1.71	1.19 ± .18	9.59 ± 2.86
DM (5)	6.05 ± 1.12a	.92 ± .22a	5.72 ± 2.34b

DM, Diabetes mellitus; SMBF, skeletal muscle blood flow; SMOEF, skeletal muscle oxygen extraction fraction; SMVO₂, skeletal muscle oxygen consumption rate.

^aP =.05, DM vs healthy.

^bP < .05, DM vs healthy.

Fig 2.

Skeletal muscle oxygen extraction fraction (SMOEF) maps, at rest and during isometric contraction, from the same subjects in Fig 1: a healthy volunteer (A) and an age-matched diabetes mellitus (DM) patient (B). Elevated SMOEF in a healthy volunteer during the contraction can be observed. In contrast, the SMOEF map of contraction in DM patient shows similar intensity as the map of rest. Color bar range, 0.1 to 0.8.

Fig 3 shows the relationships between HbA1c and exercise SMVO₂ and SMBF in the DM group. With increased HbA1c, exercise SMVO₂ decreased exponentially with a correlation coefficient r² of 0.91 for both muscle regions. However, the gastrocnemius muscle appeared to decay more quickly than the soleus muscle with a decay constant of .316 vs .135. It is noted that exponential function used for fitting the curves are included in this pilot study only to demonstrate the trend of the data. In a similar fashion, exercise SMBF decreased appreciably in the gastrocnemius muscle, but it was unchanged with increasing HbA1c level in the soleus muscle. There is no association between HbA1c and exercise SMOEF or any resting hemodynamic parameters.

Fig 3.

Relationships of exercise skeletal muscle blood flow (SMBF) and skeletal muscle oxygen consumption rate (SMVO₂) with HbA1c in patients with diabetes mellitus (DM). The fast-twitch gastrocnemius muscle shows more decreased values with increased HbA1c than the slow-twitch soleus muscle during the isometric contraction. The exercise SMBF is less sensitive to the changes in HbA1c than the exercise SMVO₂. The empirical exponential fitting curves are used to show the trend of the data.

DISCUSSION

The development of complications of the legs and feet in patients with DM is understood to be the result of a number of factors that are likely interrelated. Some of the most prominent features include neuropathy, immune dysfunction, and circulatory compromise. This is the first study to quantify regional skeletal muscle oxygenation in patients with DM using a noncontrast MRI approach. During an isometric exercise, significant increases in SMBF, SMOEF, and SMVO₂ were observed in all healthy volunteers, both in the soleus and gastrocnemius muscles. In contrast, patients with DM had attenuated elevations of these parameters, reflected by 20% to 55% reductions in the reserves (Table IIA and B). Furthermore, exercise SMBF and SMVO₂ in the gastrocnemius muscle decreased with elevated HbA1c levels in the DM group. Similar reduction was also observed in the soleus muscle, but to a lesser degree.

Table II. B.

Hemodynamic reserves measured by noncontrast magnetic resonance imaging (MRI) in gastrocnemius muscle

Subjects	SMBF reserve	SMOEF reserve	SMVO2 reserve
Healthy (6)	9.76 ± 2.77	1.27 ± .19	12.56 ± 4.95
DM (5)	5.54 ± 2.54b	1.02 ± .07b	5.66 ± 2.59b

DM, Diabetes mellitus; SMBF, skeletal muscle blood flow; SMOEF, skeletal muscle oxygen extraction fraction; SMVO₂, skeletal muscle oxygen consumption rate.

^aP =.05, DM vs healthy.

^bP < .05, DM vs healthy.

Compared with an eight- to 10-fold increase in exercise SMBF in the healthy group, patients with DM showed only a five- to sixfold increase. There is compiling evidence to suggest limb arterial blood flow is impaired in patients with DM.^27,28 Lalande et al found that reduced exercise femoral arterial blood flow in patients with type 2 diabetes is caused by impaired vascular function and is independent of cardiac output.²⁹ Reduced perfusion reserve was clearly observed in patients with DM without peripheral arterial disease, using ^99mTc-MIBI.³⁰ Our SMBF data are in general agreement with these findings. Furthermore, we have observed that the gastrocnemius muscle appeared to have greater reduction in SMBF reserve than the soleus muscle (43% vs 25% reduction) in patients with DM. There are two possible explanations for the differences measured in the gastrocnemius compared with the soleus. The first is that previous research has reported that the gastrocnemius shows two times the fat infiltration compared with the soleus muscle in persons with diabetes and peripheral neuropathy.³¹ The second is that the gastrocnemius is a fast-twitch muscle, while the soleus is a slow-twitch muscle, and there is a lower capillary density in fast-twitch muscle compared with slow-twitch muscle. However, the underlying mechanisms of vascular or muscle changes in DM is unknown at this stage.

For muscle oxygenation, there were significant increases in SMOEF (20% to 27%; P < .05) and SMVO₂ (nine- to 10-fold; P < .01) from rest to exercise in the healthy group. In patients with DM, exercise-induced changes in SMOEF (no significant change), and SMVO₂ (five- to sixfold; P < .05) were significantly attenuated. This is in accordance with evidence that found diabetes results in slowed oxygen uptake kinetics and reduced exercise capacity.^32,33 In addition, there were some noticeable differences observed between the gastrocnemius and soleus muscles. In the healthy group, the SMOEF and SMVO₂ appeared to have a greater increase from rest to exercise in gastrocnemius muscle than soleus muscle. This is consistent with the finding in rats where a greater O₂ extraction fraction was demonstrated in muscle comprised of fast-twitch fiber compared with their slow-twitch counterparts during a twitch contraction.³⁴ In the DM group, such regional differences in SMOEF and SMVO₂ remained similar to the healthy group.

It is noted that resting SMBF of soleus muscle in the DM group vs the healthy group was significantly reduced. However, there was no significant difference in resting SMBF of gastrocnemius muscle between DM and healthy groups. This appears to be consistent with a recent report that a resting MRI relaxation time T2* (effective transverse relaxation time in inhomogeneous magnetic fields) in the soleus muscle of DMpatients is significantly reduced (as deoxyhemoglobin concentration is increased due to less SMBF), compared with healthy controls.³⁵ Although this observation is not fully understood, it was explained by the greater reduction in capillaries in the soleus muscle of DM patients due to its slow-twitch fibers and oxidative metabolism.³³

The reduced exercise SMBF and SMVO₂ with in- creasing HbA1c levels (Fig 2) may indicate hyperglycemia-induced vasoactive dysfunction.³⁶ Many studies have suggested that insulin directly acts on the endothelium of the vessel wall by stimulation of nitric oxide production, specifically augmentation of nitric oxide synthase via the phosphatidylinositol-3 kinase Akt pathway.^37–39 Such reduced flow response in DM is associated with dysfunction of the oxygen transport pathway (eg, reduced total oxygen extraction, oxygen consumption, and submaximal exercise capacity).^40–42 However, all of these studies used total body VO₂ max to represent the total oxygen consumption. This study confirmed the findings of reduced oxygenation in patients with DM and is the first one to demonstrate regional differences in the oxygenation of different muscle groups. The hemodynamic responses of the gastrocnemius muscle appear to be affected by the HbA1c level more than the soleus muscle. This is possibly due to the disparity of muscle fiber type (fast-twitch vs slow-twitch), capillary density, oxygen metabolism (anaerobic vs oxidative metabolism), and/or greater fat infiltration in the gastrocnemius compared with the soleus in patients with DM.^29,30 The exact underlying mechanism is unknown, and future study needs to isolate each factor to elucidate the pathway of oxidation dysfunction in different muscle groups of patients with DM.

We believe that direct evaluation and measurement of the regional perfusion of the lower extremities holds great potential to advance the care and management of diabetic patients. An optimal test of regional perfusion could herald a number of changes in clinical management including: (1) provide a more accurate screening prior to invasive arterial evaluation; (2) direct interventional strategies to improve perfusion; (3) accelerate the institution of novel interventions such as local angiogenic therapy; and (4) accelerate definitive decisions for diabetes foot ulcers when healing potential appears to be insufficient. In this pilot study, the DM patients had a variety of cardiovascular diseases. However, it is unknown whether reduced exercise perfusion in the DM patients is related to the cardiovascular function and/or sympathetic activity. In a small study of type 2 DM patients without cardiovascular disease, reduced skeletal muscle perfusion was independent of cardiac output during a submaximal exercise.²⁹ We suspected that both impaired central cardiovascular function and reduced local blood flow affect the exercise capacity in the DM patients. This needs a large scale of clinical study to validate.

There are several limitations in this pilot study. As this is a technical development study, rather than a hypothesis-testing study, we had a small number of participants with both type 1 and 2 diabetes in the DM group. These hemodynamic parameters may be different between type 1 and type 2 patients. Although the subjects had no clinical manifestation or history of peripheral vascular disease, complete vascular assessment and exercise capacity testing were not conducted. A large-scale study with complete vascular and exercise capacity testing will be needed to evaluate these important factors. In addition, it was noticed that one DM patient with a duration of DM of 20 years had the lowest exercise SMOEF and SMVO₂, but his exercise SMBF was comparable with other DM patients. His HbA1c was also the highest (10.1). Again, due to the small number of patients, no conclusion can be drawn in this pilot study regarding the role of duration of DM on the skeletal muscle microcirculation. Another limitation of this study is the inability to evaluate the anterior or lateral muscle compartments due to the low signal intensity observed using our custom-made surface coil. This problem could be partially solved using a larger surface coil that is under development in our laboratory. Finally, although resistance to plantar flexion was individualized to allow successful task completion, we did not attempt to measure each individual’s maximal voluntary contraction and standardize the resistance. Nevertheless, the standard deviation of the resistance pressure was small (6.5 N), indicating similar physical resistance among DM patients and healthy volunteers. The significant attenuation in our exercise data in DM patients indicated our conclusions about this group of patients should remain valid.

CONCLUSIONS

In summary, skeletal muscle perfusion and oxygenation can be quantified in healthy and diabetic subjects. Isometric muscle contraction induced significant increases in both perfusion and oxygenation in healthy subjects, but such increases were markedly attenuated in patients with DM. This may have implications for early detection of compromised hemodynamics in the lower extremities of patients with DM. The exercise oxygen consumption rate appears to be associated with blood serum HbA1c levels. Heterogeneity of hemodynamic responses in gastrocnemius and soleus muscles may reflect differences in the muscle fiber type, capillary volume, and oxidation. The technical advancement described in this pilot study may provide a useful noninvasive tool for understanding of the mechanism of hemodynamic response to stress in skeletal muscle of patients with DM. It also may help guide therapeutic strategies to target local angiogenic factors so that the incidence of abnormal cardiovascular outcome can be reduced and even avoided.