Noninvasive Optical Characterization of Muscle Blood Flow, Oxygenation, and Metabolism in Women With Fibromyalgia

Yu Shang; Katelyn Gurley; Brock Symons; Douglas Long; Ratchakrit Srikuea; Leslie J Crofford; Charlotte A Peterson; Guoqiang Yu


Arthritis Res Ther. 2012;14(6) 

In This Article

Materials and Methods

Study Protocols

The study was reviewed and approved by the University of Kentucky Institutional Review Board. Thirty-seven women between the ages of 51 and 70 years participated in this study, including 14 women with FM and 23 healthy controls. All subjects gave signed consent prior to the study. FM was diagnosed in accordance with the 1990 criteria established by the American College of Rheumatology.[37] Subjects with FM and healthy controls were matched in age, height, weight, body mass index (BMI), physical activity and baseline maximal voluntary isometric contractions (MVIC). Physical activity was characterized by the international physical activity questionnaire (IPAQ)[38] while baseline MVIC was determined prior to fatigue exercise (described later in this section). The duration of FM (that is, the time interval between the diagnosis of FM and the start of the fatiguing exercise study) was 13.69 ± 1.90 years (mean ± standard error, SE), ranging from 3 to 24 years. Some subjects with FM were taking medications such as anti-inflammatory drugs, low-dose aspirin, and fish oil supplements. These medications were discontinued at least 3 days prior to exercise to minimize the impact on muscle hemodynamic responses.

Two experimental protocols were used to investigate muscle hemodynamic/metabolic responses, including fatiguing leg isometric contractions and arm cuff occlusion (muscle ischemic challenge). Each subject was asked to perform fatiguing leg exercise followed by cuff occlusion in the arm to minimize the potential interference between the two protocols in a single muscle, and to examine responses to both exercise and ischemia. Leg knee extensor muscles were evaluated for fatiguing exercise since these muscles are primarily used for daily activities, and the subjects were tolerant to the protocol. In addition, previous studies in FM[18,19] have utilized a cuff occlusion protocol using the forearm muscles (flexor carpi radialis) to investigate oxygen kinetics, allowing a basis for comparison.

At least 3 days prior to fatiguing exercise, each subject participated in a session to become familiar with the performance of MVIC on a Biodex multi-joint dynamometer (System 4, Biodex Medical Systems Inc., NY, USA) as follows: the subject was seated in an upright position with the seat tilted at an angle of 85°. The lateral femoral epicondyle was aligned to the center of the dynamometer shaft. Stabilization was provided by two shoulder straps, and a waist strap was used to minimize the use of skeletal muscles other than the knee extensor. Each subject's testing foot was secured by a strap to the dynamometer with a fixed knee angle of 90°. The subject was then instructed to perform leg isometric contractions (that is, kicking against the dynamometer lever arm) held for 3 to 4 seconds, while the maximal value of the torque generated during the isometric contraction was recorded. During MVIC, an electrical stimulation (ES) was used to noninvasively induce superimposed force to the muscle via surface electrodes over the proximal and distal portions of the thigh.[39] When applying ES to the muscle, all motor units are recruited. Any increment in force from ES would suggest incomplete activation of the muscle from MVIC. After performing MVIC measures, the central activation ratio (CAR) was quantified using the following equation:

The isometric fatiguing exercise was performed on the dynamometer and started with three MVIC trials with a 3-minute rest between MVICs. Following the MVIC tests, an optical sensor was secured by medical tape over the vastus lateralis (mid belly) at the mid thigh of the evaluated leg, and a 3-minute baseline measurement was recorded. Following the baseline measurement, the subject was instructed to perform six sets of twelve isometric muscle contractions at a 40% duty cycle (4-second contraction, 6-second rest) steadily increasing from an initial intensity of 20% MVIC and eventually reaching an intensity of 70% of the MVIC with a increment of 10% MVIC after each set. While performing exercise, the subject received visual feedback by looking at the targeted intensity level shown on a computer and was encouraged to achieve each set intensity (20 to 70% MVIC). Between sets of isometric muscle contractions, a single MVIC was performed as the primary measure of fatigue. There was no additional rest between sets. In total, 78 muscle contractions (12 × 6 contractions + 6 MVIC) were performed during the course of fatiguing exercise. Fatigue and pain during exercise were evaluated according to the visual analog scale (VAS); the subject was asked to indicate her pain/fatigue severity on a 100-mm scale. Details about the fatigue and pain questionnaires can be found in references[41] and.[42] A higher VAS score indicates more pain or fatigue. After completion of the exercise, the subject was asked to perform one MVIC to evaluate strength recovery at time points 3, 6, 9 and 12 minutes.

The cuff occlusion protocol in the forearm started approximately 10 minutes after the fatiguing leg exercise. The participant sat in an upright position and the right arm was extended resting on a horizontal support. A fast-inflating automatic tourniquet cuff (ATS 1000, Zimmer Inc., IN, USA) was placed on the upper arm and an optical sensor was secured by medical tape over the flexor carpi radialis muscle. After a 3-minute baseline measurement, arterial blood flow was occluded via the tourniquet on the upper arm at a pressure of 230 mmHg for 3 minutes. The pressure was then released and measurements continued for an additional 5-minute recovery period.

Hemodynamic/Metabolic Measurements

The optical sensor was connected to a hybrid diffuse optical instrument, which combined a commercial NIRS oximeter (Imagent, ISS Inc., IL, USA) for tissue oxygenation measurement and a custom-designed diffuse correlation spectroscopy (DCS) flowmeter for tissue blood flow measurement. The principle of the hybrid instrument has been described elsewhere.[23,43] Briefly, near-infrared laser light was delivered into the thigh or arm muscle, and the reflectance light was received by the photon detectors through source and detector fibers placed on the skin surface. The NIRS oximeter measured the amplitudes and phases of frequency-modulated light (110 MHz) at two wavelengths (830 and 690 nm) and four source-detector separations (2.0, 2.5, 3.0 and 3.5 cm) to extract tissue absorption and scattering coefficients.[23,44] Absolute values of [HbO2] and [Hb] were extracted from the measured tissue absorption coefficients at the two wavelengths.[44] Total hemoglobin concentration (THC) was then calculated as the sum of [HbO2] and [Hb], while tissue blood oxygen saturation (StO2) was calculated as 100% × [HbO2]/THC.

Preliminary data analyses indicated that the measured time courses of absolute tissue blood oxygenation, determined by the light amplitudes and phases from all four source-detector separations were too noisy to determine reliable time intervals for characterizing oxygen recovery, due to the unstable phase slopes over time. Thus, to evaluate tissue blood oxygenation recovery, we used the measured light amplitudes at the two wavelengths from a single source-detector separation (2.5 cm for arm muscles or 3.0 cm for leg muscles) to calculate the changes of [HbO2] and [Hb] (that is, Δ[HbO2] and Δ[Hb]) relative to their baselines (before physiological manipulations), based on the modified Beer-Lambert Law.[45]

Blood flow index was extracted by fitting the autocorrelation curve determined from the detected temporal fluctuation of light intensity measured by DCS.[23,27,30] The unit of the blood flow index is cm2/s. Although this unit is different from the classical blood flow unit in biological tissues (ml/min/100 ml), its percentage changes have been found to correlate well with the blood flow changes measured by many other established modalities.[32–36] The relative blood flow (rBF) was then calculated by normalizing/dividing the blood flow index to its baseline. As with the NIRS oximeter, the source-detector separation used for DCS measurement was 2.5 or 3.0 cm for arm or leg muscle, respectively. The distal tips of source and detector fibers for NIRS oximeter and DCS were embedded in a foam pad to form a hybrid optical sensor.[23,43] The NIRS oximeter and DCS flowmeter were operated alternately via triggers controlled by a computer. For both protocols of fatiguing exercise and cuff occlusion, muscle blood flow and blood oxygenation were continuously monitored by the hybrid optical instrument with a frame sampling time of approximately 3 seconds throughout the experiments.

Relative (normalized to baseline) oxygen extraction fraction (rOEF) and oxygen consumption rate (rVO2) during fatiguing exercise were calculated based on Fick's law using the measured blood flow and oxygenation data respectively:[27]

rOEF = 100% × (1-StO2)/(1-StO2baseline) and rVO2 = 100% × rBF × rOEF, where StO2baseline represents the absolute baseline value of tissue blood oxygen saturation before exercise. Similar to the rOEF and rVO2, relative oxygenation changes were calculated by normalizing their absolute values to baselines respectively, resulting in r[HbO2], r[Hb], rTHC and rStO2. Throughout this paper, 'r' represents the relative value normalized/divided by its baseline, and 'Δ' (for example, Δ[HbO2] and Δ[Hb]) represents the subtracted difference between the time course data and its baseline.

Data Analysis

Since the optical signals during exercise were easily contaminated by the muscle fiber motion during leg contractions,[26] hemodynamic/metabolic responses during fatiguing exercise were estimated by averaging the optical data over the 6 seconds immediately after fatiguing exercise. This very short post-exercise measurement period was selected as the most accurate reflection of the exercise state since rapid hemodynamic changes occur in muscle immediately after exercise.[46,47] Approximately one minute after exercise, most hemodynamic data became stable, allowing us to average a longer period of data acquisition to obtain a better signal-to-noise ratio. Therefore, hemodynamic data at the time points 3, 6, 9 and 12 minutes after exercise were quantified by averaging 30 seconds of data immediately preceding each time point. These data were then normalized to their pre-exercise baselines to evaluate responses during and post-exercise.

Hemodynamic recovery (that is, rBF, Δ[HbO2], Δ[Hb]) after fatiguing exercise or cuff occlusion was characterized by the recovery half-time,[18,19,48] which was defined as the time interval from the end of occlusion/exercise to the time by which tissue hemodynamics had recovered to the half-maximal value. We used Δ[HbO2] and Δ[Hb] from one single source-detector separation (2.5 cm for arm muscles or 3.0 cm for leg muscles) to determine the half-times of oxygenation recovery. Note that the rOEF and rVO2 data had similar noise levels as the absolute blood oxygenation data because they were calculated from the absolute StO2. Thus, the recovery half-times for THC, StO2, rOEF and rVO2 are not reported in this study.

Average hemodynamic/metabolic responses by group are presented as means ± SE (error bars) in figures. The Student's t-test was used to compare differences in hemodynamic/metabolic data between the FM and healthy control groups. Linear regression was used to investigate the correlations among demographic, physical activity, strength and optical data.