Muscle oxygenation during dynamic plantar flexion exercise: combining BOLD MRI with traditional physiological measurements

Abstract

Blood-oxygen-level-dependent magnetic resonance imaging (BOLD MRI) has the potential to quantify skeletal muscle oxygenation with high temporal and high spatial resolution. The purpose of this study was to characterize skeletal muscle BOLD responses during steady-state plantar flexion exercise (i.e., during the brief rest periods between muscle contraction). We used three different imaging modalities (ultrasound of the popliteal artery, BOLD MRI, and near-infrared spectroscopy [NIRS]) and two different exercise intensities (2 and 6 kg). Six healthy men underwent three separate protocols of dynamic plantar flexion exercise on separate days and acute physiological responses were measured. Ultrasound studies showed the percent change in popliteal velocity from baseline to the end of exercise was 151 ± 24% during 2 kg and 589 ± 145% during 6 kg. MRI studies showed an abrupt decrease in BOLD signal intensity at the onset of 2 kg exercise, indicating deoxygenation. The BOLD signal was further reduced during 6 kg exercise (compared to 2 kg) at 1 min (-4.3 ± 0.7 vs. -1.2 ± 0.4%, P < 0.001). Similarly, the change in the NIRS muscle oxygen saturation in the medial gastrocnemius was -11 ± 4% at 2 kg and -38 ± 11% with 6 kg (P = 0.041). In conclusion, we demonstrate that BOLD signal intensity decreases during plantar flexion and this effect is augmented at higher exercise workloads.

Keywords: Blood pressure; heart rate; sympathetic nervous system.

Figure 1

Magnetic resonance imaging (MRI) compatible leg exercise device. The subjects were positioned supine with the largest part of the lower leg in the coil. Sand bags were added to the weight pan to increase the workload. A personal computer in the control room was connected via an MRI compatible cord to monitor the range of motion. This same device was also used for ultrasound studies (Protocols 1 and 3).

Figure 2

Original Doppler ultrasound recordings of popliteal blood flow velocity recorded at baseline (left panel) and during single‐leg dynamic plantar flexion exercise at 6 kg (right panel) in a 62‐year‐old healthy man. Note that in the left panel, velocity waveforms are triphasic whereas in the right panel the velocity waveforms are not triphasic and can only be measured in between exercise (Ex) contractions.

Figure 3

Systemic hemodynamic responses to single‐leg dynamic plantar flexion exercise at 2 kg (black diamonds) and 6 kg (white squares) in Experiment 1. All parameters significantly increased over time. The magnitude of difference between workloads was most noticeable for popliteal mean velocity (bottom panel), which was consistent with the purpose of our study (i.e., influencing local blood flow regulation while not significantly raising sympathetic tone). Data are M ± SEM.

Figure 4

Magnetic resonance imaging of the lower leg from one subject (same individual as Fig. 2). (A) T2‐weighted image to orient the reader. (B) Echo planar image with regions of interest that was used for analysis: tibialis anterior (TA), soleus (S), medial gastrocnemius (MG), and lateral gastrocnemius (LG). (C) Unfiltered blood‐oxygen‐level‐dependent (BOLD) signal intensity data during 6 kg exercise shown as a percent change from baseline. (D) The same BOLD signal intensity data after wavelet‐based filtering.

Figure 5

Time‐course of blood‐oxygen‐level‐dependent (BOLD) signal intensity in the medial gastrocnemius in response to plantar flexion at 2 kg (solid line) and 6 kg (dashed line). All 400 data points from the 20 min scan are presented, but only some of the error bars are shown to improve clarity. A greater negative BOLD response was found during exercise with 6 kg, indicating lower deoxyhemoglobin concentration. Data are M ± SEM. Planned comparisons were conducted between 2 and 6 kg at specific time points (please see the text for details). *indicates significant difference between workloads (P < 0.05).

Figure 6

Individual recordings of muscle oxygen saturation obtained from near‐infrared spectroscopy during 2 and 6 kg plantar flexion. These data were obtained from the same individual shown in Figures 2 and 4.