Transient expression of myofibroblast-like cells in rat rib fracture callus.

BACKGROUND AND PURPOSE: We have previously shown that early fracture callus of rat rib has viscoelastic and contractile properties resembling those of smooth muscle. The cells responsible for this contractility have been hypothesized to be myofibroblast-like in nature. In soft-tissue healing, force generated by contraction of myofibroblasts promotes healing. Accordingly, we tried to identify myofibroblast-like cells in early fibrous callus. ANIMALS AND METHODS: Calluses from rat rib fractures were removed 7, 14, and 21 days after fracture and unfractured ribs acted as controls. All tissues were analyzed using qPCR and immunohistochemistry. We analyzed expression of smooth muscle- and myofibroblast-associated genes and proteins including alpha smooth muscle actin (αSMA), non-muscle myosin, fibronectin extra domain A variant (EDA-fibronectin), OB-cadherin, connexin-43, basic calponin (h1CaP), and h-caldesmon. RESULTS: In calluses at 7 days post-fracture, there were statistically significant increases in expression of αSMA mRNA (2.5 fold), h1CaP mRNA (2.1 fold), EDA-fibronectin mRNA (14 fold), and connexin-43 mRNA (1.8 fold) compared to unfractured ribs, and by 21 days post-fracture mRNA expression in calluses had decreased to levels approaching those in unfractured rib. Immunohistochemistry of 7 day fibrous callus localized calponin, EDA-fibronectin and co-immunolabeling of OB-cadherin and αSMA (thus confirming a myofibroblastic phenotype) within various cell populations. INTERPRETATION: This study provides further evidence that early rat rib callus is not only smooth muscle-like in nature but also contains a notable population of cells that have a distinct myofibroblastic phenotype. The presence of these cells indicates that in vivo contraction of early callus is a mechanism that may occur in fractures so as to facilitate healing, as it does in soft tissue wound repair.

Lowering of sensory, motor, and pain-tolerance thresholds with burst duration using kilohertz-frequency alternating current electric stimulation: part II.

OBJECTIVE: To determine the optimum burst duration for discrimination between sensory, motor, and pain tolerance thresholds using 20-Hz bursts of kilohertz-frequency sinusoidal alternating current (AC) applied transcutaneously to human participants. DESIGN: A within-subject, repeated-measures trial. SETTING: A research laboratory. PARTICIPANTS: Healthy young adults (N=20). INTERVENTIONS: Bursts of AC electric stimulation at frequencies of 1 and 4 kHz. The burst frequency was 20 Hz. Burst durations ranged from 250 microseconds (for 1 cycle of 4-kHz AC) and 1 millisecond (for 1 cycle of 1-kHz AC) to 50 milliseconds (continuous AC). MAIN OUTCOME MEASURES: Measurement of sensory, motor, and pain-tolerance thresholds. RESULTS: Thresholds decreased to a minimum with increasing burst duration. The minimum threshold identified the utilization time over which summation of subthreshold stimuli occurs. Utilization times were different for sensory (approximately 20 ms), motor (approximately 30 ms), and pain (>50 ms) and were much higher than found in a previous study that used a higher burst frequency (50 Hz). As with the previous study, relative thresholds were found to vary with burst duration. Despite the very different utilization times, maximum separation between sensory, motor, and pain thresholds was found to occur with bursts in the range of 1 to 4 milliseconds, the same range found in the previous study. CONCLUSIONS: Our conclusions concur with those reported previously and support the contention that short-duration kilohertz-frequency AC bursts (1-4 ms) have a more useful role in rehabilitation than the long-duration kilohertz-frequency bursts that characterize Russian and interferential currents.

Comparison of the hypoalgesic efficacy of low-frequency and burst-modulated kilohertz frequency currents.

BACKGROUND AND PURPOSE: A within-subject randomized controlled trial was conducted to compare monophasic pulsed current (PC) with a frequency of 50 Hz and a phase duration of 500 microseconds with burst-modulated alternating current (BMAC) (1-kHz alternating current, burst modulated at a frequency of 50 Hz with a 20% duty cycle) to establish whether there were differences in hypoalgesic efficacy as assessed by cold pain threshold measurements. SUBJECTS: Twenty-two young adults who were healthy and drawn from a population of students of La Trobe University volunteered to participate in the study. Nineteen subjects (7 male, 12 female) met the inclusion criteria. METHOD: Each participant experienced monophasic PC and BMAC. Time to cold pain threshold was measured before, during, and after the electrical stimulation intervention. RESULTS: Both PC and BMAC currents were effective at elevating the cold pain threshold, although there was no statistically significant difference between the 2 currents during stimulation. DISCUSSION AND CONCLUSION: Pulsed current and BMAC appear to be equally effective at elevating the cold pain threshold. Analysis indicated that if any real difference exists, it would only be apparent with large (100) subject numbers. Any differences in hypoalgesic efficacy thus are not likely to be clinically important.

Wrist extensor torque production and discomfort associated with low-frequency and burst-modulated kilohertz-frequency currents.

BACKGROUND AND PURPOSE: A randomized controlled trial to compare 2 forms of monophasic pulsed currents with 2 forms of burst-modulated, kilohertz-frequency alternating current ("Russian current" and "Aussie current") was conducted to establish whether different amounts of wrist extensor torque were produced and whether discomfort varied with stimulus type. SUBJECTS: The 32 subjects were adults who were healthy and were drawn from a population of staff and students at La Trobe University. METHODS: Each subject received all 4 currents. Maximal electrically induced torque (MEIT) of the wrist extensors was measured for each stimulus type. Relative discomfort of stimulation also was assessed. RESULTS: Russian current elicited lower mean torque than those elicited by Aussie current and monophasic pulsed currents. The Russian and Aussie currents elicited significantly less discomfort than the 2 monophasic pulsed currents. DISCUSSION AND CONCLUSION: When force production and relative discomfort were jointly used as the criteria, Aussie current was found to be more effective than either of the monophasic pulsed currents or Russian current stimulation.

Russian electrical stimulation: the early experiments.

Russian forms of electrical stimulation became popular to a large extent as a result of the activities of Kots, who claimed force gains of up to 40% in elite athletes as a result of what was then a new form of stimulation. He did not provide details of his published work, nor did he give references. Russian electrical stimulation became popular despite the lack of research in the English-language literature. No studies published in English examined whether the "10/50/10" treatment regimen (10 seconds of stimulation followed by 50 seconds rest, repeated for 10 minutes) advocated by Kots is optimal, and only one study addressed whether maximum muscle torque was produced at an alternating current frequency of 2.5 kHz. The few studies that compared low-frequency monophasic pulsed current and Russian electrical stimulation are inconclusive. This article reviews and provides details of the original studies by Kots and co-workers. The authors contend that these studies laid the foundations for the use of Russian forms of electrical stimulation in physical therapy. The authors conclude that there are data in the Russian-language literature that support the use of Russian electrical stimulation but that some questions remain unanswered.

The effect of duty cycle and frequency on muscle torque production using kilohertz frequency range alternating current.

We investigated the frequency and duty cycle dependence of maximal electrically induced torque (MEIT) of the wrist extensors. Fifty hertz burst modulated sinusoidal alternating current (AC) in the frequency range 0.5-20 kHz was used, with duty cycles ranging from a minimum (one cycle) to maximum (continuous AC). MEITs were similar at low frequencies but decreased markedly above 2.5 kHz. MEITs also decreased markedly above a 20% duty cycle. Subjective reports of discomfort were fewest at 4 kHz and at duty cycles in the range 20-25%. Our conclusion is that for maximum torque production, a frequency of 1 kHz and a duty cycle of 20% are indicated. When comfort is a major consideration, a frequency of 2.5 kHz provides an acceptable trade-off between MEIT and comfort. The findings also suggest that low duty cycle, burst modulated AC stimulation may be more effective than stimulation using conventional low-frequency pulsed current.

Vastus medialis electrical stimulation to improve lower extremity function following a lateral patellar retinacular release.

STUDY DESIGN: A single-case study design. OBJECTIVES: To examine the effect of electrical stimulation of the vastus medialis muscle on stiffness, pain and function for a patient with delayed functional progress following a lateral patellar retinacular release. BACKGROUND: Five months after an arthroscopic lateral patellar retinacular release, the patient, although highly motivated, had made little progress using routine exercises and taping. METHODS AND MEASURES: An electrical stimulation program producing approximately 300 contractions daily of the vastus medialis muscle was implemented. The electrical stimulation applied for 33 of the 36 days was a rectangular and balanced biphasic pulse of 625-micros duration, 70-Hz frequency, 8-second peak on-time, 3-second off-time, 1-second ramp-up, and 0.5-second ramp-down. Objective measures of stair climbing and hopping, together with the subjective measure of therapist-palpated superomedial patella displacement force, were recorded for each treatment visit. Other subjective measures were the patient's daily recordings of knee pain and stiffness. RESULTS: Patient-reported stiffness reduced rapidly as the actual and cumulative number of daily contractions of the vastus medialis muscle increased. After 8 days of electrical stimulation, the patient was able to ascend stairs unassisted and after another 21 days to hop unsupported. CONCLUSIONS: Stiffness rapidly reduced and function started to improve once the electrical stimulation program was implemented. Recovery during the 36 days of treatment with electrical stimulation was greater than during the previous 5 months using other methods. Compliance was not an issue, nor was muscle soreness.

Thymosin ß4 administration enhances fracture healing in mice.

Thymosin ß4 (Tß4 ) is a regenerative peptide that we hypothesized would promote healing of fractured bone. Mice received a bilateral fibular osteotomy and were given i.p. injections of either Tß4 (6 mg/kg) or saline. Calluses from saline- and Tß4 -treated mice were analyzed for: (1) biomechanical properties and (2) composition using micro-computed tomography (µCT) and histomorphometry. Biomechanical analysis showed that Tß4 -treated calluses had a 41% increase in peak force to failure (p < 0.01) and were approximately 25% stiffer (p < 0.05) than saline-treated controls. µCT analysis at 21 days post-fracture showed that the fractional volume of new mineralized tissue and new highly mineralized tissue were respectively 18% and 26% greater in calluses from Tß4 -treated mice compared to controls (p < 0.01; p < 0.05, respectively). Histomorphometry complemented the µCT data; at 21 days post-fracture, Tß4 -treated calluses were almost 23% smaller (p < 0.05), had nearly 47% less old cortical bone (p < 0.05) and had a 31% increase in new trabecular bone area/total callus area fraction compared with controls (p < 0.05). Our finding of enhanced biomechanical properties of fractures in mice treated with Tß4 provides novel evidence of the therapeutic potential of this peptide for treating bone fractures.

The influence of stimulus phase duration on discomfort and electrically induced torque of quadriceps femoris.

BACKGROUND: Although a number of studies have compared the influence of different electrical pulse parameters on maximum electrically induced torque (MEIT) and discomfort, the role of phase duration has been poorly investigated. OBJECTIVE: To examine the variation in muscle torque and discomfort produced when electrically stimulating quadriceps femoris using pulsed current with three different phase durations in order to establish whether there are any advantages or disadvantages in varying the phase duration over the range examined. METHOD: This is a two repeated-measures, within-subject study conducted in a research laboratory. The study was divided into 2 parts with 19 healthy young adults in each part.In part 1, MEIT was determined for each phase duration (400, 700, and 1000 µs), using a biphasic pulsed current at a frequency of 50 Hz. In part 2, stimulus amplitude was increased until the contractions reached 40% of maximum voluntary isometric contraction (MVIC) and the associated discomfort produced by each phase duration was measured. RESULTS: In part 1 of the study, we found that the average MEITs generated with each phase duration (400, 700, and 1000 µs) were 55.0, 56.3, and 58.0% of MVIC respectively, but the differences were not statistically significant (p=.45). In part 2, we found a statistically significant increase in discomfort over the same range of phase durations. The results indicate that, for a given level of torque production, discomfort increases with increasing phase duration (p=.008). CONCLUSIONS: Greater muscle torque cannot be produced by increasing the stimulus phase duration over the range examined. Greater discomfort is produced by increasing the stimulus phase duration.

Sensory transcutaneous electrical stimulation fails to decrease discomfort associated with neuromuscular electrical stimulation in healthy individuals.

OBJECTIVE: Sensory transcutaneous electrical nerve stimulation (TENS) is frequently used for pain modulation. Neuromuscular electrical stimulation used to induce strong muscle contractions is often limited by muscular discomfort and by discomfort associated with the electrical current. The objective of this study was to determine whether the application of TENS can reduce the discomfort associated with neuromuscular electrical stimulation, leading to stronger maximal electrically induced contractions. DESIGN: In a double-blind randomized controlled trial, 24 young adults with no known impairment received a TENS and a sham treatment before and after the assessment of the maximal electrically induced contractions of the quadriceps muscle. RESULTS: Sensory TENS had no significant effect on maximal electrically induced contractions (normalized to maximal voluntary isometric contractions), on current amplitude, on the discomfort associated with the muscle contraction, and on the discomfort associated with the current amplitude (the latter two rated using a numerical rating scale). The discomfort described as a pulling/tearing sensation of the muscle (mean [SD], 5.7 [1.6]) was significantly higher than the discomfort associated with the current amplitude (mean [SD], 4.6 [2.3]; P = 0.02). CONCLUSIONS: Sensory TENS was not effective in modulating the discomfort associated with neuromuscular electrical stimulation in individuals with no known impairment. During periods of maximal electrically induced contractions, the pulling sensation of the muscle is generally more disconcerting than the sensation of the electrical current. Further studies are necessary to determine the effect of different stimulation parameters in patients with pathologic conditions.

α(1) adrenergic receptor agonist, phenylephrine, actively contracts early rat rib fracture callus ex vivo.

Early, soft fracture callus that links fracture ends together is smooth muscle-like in nature. We aimed to determine if early fracture callus could be induced to contract and relax ex vivo by similar pathways to smooth muscle, that is, contraction via α(1) adrenergic receptor (α(1) AR) activation with phenylephrine (PE) and relaxation via ß(2) adrenergic receptor (ß(2) AR) stimulation with terbutaline. A sensitive force transducer quantified 7 day rat rib fracture callus responses in modified Krebs-Henseliet (KH) solutions. Unfractured ribs along with 7, 14, and 21 day fracture calluses were analyzed for both α(1) AR and ß(2) AR gene expression using qPCR, whilst 7 day fracture callus was examined via immunohistochemistry for both α(1) AR and ß(2) AR- immunoreactivity. In 7 day callus, PE (10(-6) M) significantly induced an increase in force that was greater than passive force generated in calcium-free KH (n = 8, mean 51% increase, 95% CI: 26-76%). PE-induced contractions in calluses were attenuated by the α(1) AR antagonist, prazosin (10(-6) M; n = 7, mean 5% increase, 95% CI: 2-11%). Terbutaline did not relax callus. Gene expression of α(1) ARs was constant throughout fracture healing; however, ß(2) AR expression was down-regulated at 7 days compared to unfractured rib (p < 0.01). Furthermore, osteoprogenitor cells of early fibrous callus displayed considerable α(1) AR-like immunoreactivity but not ß(2) AR-like immunoreactivity. Here, we demonstrate for the first time that early fracture callus can be pharmacologically induced to contract. We propose that increased concentrations of α(1) AR agonists such as noradrenaline may tonically contract callus in vivo to promote osteogenesis.

Electrical stimulation using kilohertz-frequency alternating current.

Transcutaneous electrical stimulation using kilohertz-frequency alternating current (AC) became popular in the 1950s with the introduction of "interferential currents," promoted as a means of producing depth-efficient stimulation of nerve and muscle. Later, "Russian current" was adopted as a means of muscle strengthening. This article reviews some clinically relevant, laboratory-based studies that offer an insight into the mechanism of action of kilohertz-frequency AC. It provides some answers to the question: "What are the optimal stimulus parameters for eliciting forceful, yet comfortable, electrically induced muscle contractions?" It is concluded that the stimulation parameters commonly used clinically (Russian and interferential currents) are suboptimal for achieving their stated goals and that greater benefit would be obtained using short-duration (2-4 millisecond), rectangular bursts of kilohertz-frequency AC with a frequency chosen to maximize the desired outcome.

A comparison of the analgesic efficacy of medium-frequency alternating current and TENS.

OBJECTIVE: To compare the analgesic efficacy of burst-modulated medium-frequency alternating current (BMAC) and transcutaneous electrical nerve stimulation (TENS) using an experimental cold pain model. DESIGN: Within-group crossover study. SETTING: A university research laboratory. PARTICIPANTS: Twenty healthy subjects. INTERVENTIONS: BMAC (4-kHz AC applied in 4-millisecond bursts at 50Hz) and TENS (125-microsecond phase duration applied at a frequency of 50Hz) administered to each participant on separate occasions. MAIN OUTCOME MEASURE: Time to cold pain threshold. RESULTS: The mean time to cold pain threshold with the BMAC intervention was no different than with TENS. Statistical analysis showed that both interventions elevated the cold pain threshold significantly [BMAC: increase=15.2seconds, 97.5% confidence interval (CI) 3.1 to 27.2, P=0.01; TENS: increase=15.4seconds, 97.5%CI 2.5 to 28.4, P=0.02], and the difference between interventions was not simply insignificant but the intervention effects were 'significantly the same' (mean difference=0.3seconds, 95%CI -15.3 to 15.9, P=0.97). CONCLUSIONS: BMAC is as effective as TENS in increasing cold pain thresholds in healthy subjects. Since BMAC has been shown to be more comfortable than TENS in previous studies and is likely to be better accepted and tolerated by patients, clinical investigation is warranted.

Early fracture callus displays smooth muscle-like viscoelastic properties ex vivo: implications for fracture healing.

Cells of early, fibrous callus in bone fractures possess much alpha smooth muscle actin. This callus contracts and relaxes; however, active and passive components of its force production have yet to be defined. We aimed to establish whether passive viscoelastic properties of early soft fracture callus are smooth muscle-like in nature. Under anesthesia one rib was fractured in rats and calluses removed 7 days later for analysis. Urinary bladder detrusor muscle and Achilles tendon were also resected and analyzed. Force production in these tissues was measured using a force transducer when preparations were immersed in calcium-free Krebs-Henseleit solution (pH 7.4, 22 degrees C). Viscoelastic responses were measured in each preparation in response to 50 microN increases and decreases in force after achieving basal tissue tension by preconditioning. Callus, bladder, and tendon all displayed varying, reproducible degrees of stress relaxation (SR) and reverse stress relaxation (RSR) (n = 7 for all groups). Hysteresis was observed in callus, with the first SR response significantly larger than that produced in subsequent stretches (p < 0.05). Callus SR responses were greater than tendon (p < 0.001) but less than bladder (p < 0.001). Callus RSR responses were greater than tendon (p < 0.001), but no significant difference was seen between RSR of callus and bladder. We concluded that early, soft callus displayed significant SR and RSR phenomena similar to smooth muscle tissue, and SR and RSR may be important in maintenance of static tension in early callus by promoting osteogenesis and fracture healing.

The influence of stimulus phase duration on discomfort and electrically induced torque of quadriceps femoris

BACKGROUND: Although a number of studies have compared the influence of different electrical pulse parameters on maximum electrically induced torque (MEIT) and discomfort, the role of phase duration has been poorly investigated. OBJECTIVE: To examine the variation in muscle torque and discomfort produced when electrically stimulating quadriceps femoris using pulsed current with three different phase durations in order to establish whether there are any advantages or disadvantages in varying the phase duration over the range examined. METHOD: This is a two repeated-measures, within-subject study conducted in a research laboratory. The study was divided into 2 parts with 19 healthy young adults in each part.In part 1, MEIT was determined for each phase duration (400, 700, and 1000 µs), using a biphasic pulsed current at a frequency of 50 Hz. In part 2, stimulus amplitude was increased until the contractions reached 40% of maximum voluntary isometric contraction (MVIC) and the associated discomfort produced by each phase duration was measured. RESULTS: In part 1 of the study, we found that the average MEITs generated with each phase duration (400, 700, and 1000 µs) were 55.0, 56.3, and 58.0% of MVIC respectively, but the differences were not statistically significant (p=.45). In part 2, we found a statistically significant increase in discomfort over the same range of phase durations. The results indicate that, for a given level of torque production, discomfort increases with increasing phase duration (p=.008). CONCLUSIONS: Greater muscle torque cannot be produced by increasing the stimulus phase duration over the range examined. Greater discomfort is produced by increasing the stimulus phase duration. .

Lowering of sensory, motor, and pain-tolerance thresholds with burst duration using kilohertz-frequency alternating current electric stimulation.

OBJECTIVE: To determine the optimum burst duration for discrimination between sensory, motor, and pain-tolerance thresholds using 50-Hz bursts of kilohertz-frequency sinusoidal alternating current (AC) applied transcutaneously to human subjects. DESIGN: A repeated-measures randomized controlled trial. SETTING: A research laboratory. PARTICIPANTS: Twenty-six healthy young adults. INTERVENTIONS: Bursts of AC electric stimulation at frequencies of 1 and 4kHz. Burst durations ranged from 250micros (for 1 cycle of 4kHz AC, ie, a single biphasic pulse) to 20ms (continuous AC). MAIN OUTCOME MEASURES: We measured sensory, motor, and pain-tolerance thresholds at frequencies of 1 and 4kHz. RESULTS: We found that threshold voltages decreased to a minimum with increasing burst duration. The minimum threshold identified the "utilization time" over which summation of subthreshold stimuli occurs. Above the utilization time, thresholds increased. Estimated utilization times differed for sensory ( approximately 7ms), motor (>10ms), and pain-tolerance (>or=20ms). As a consequence, relative thresholds varied with burst duration. A maximum separation between sensory, motor, and pain-tolerance thresholds was found to occur with bursts in the range 1 to 4ms. CONCLUSIONS: Short-duration kilohertz-frequency AC bursts might have a more useful role in rehabilitation than either pulsed current or the long duration bursts that characterize Russian and interferential currents. Further clinical studies are needed.

The effect of heat on tissue extensibility: a comparison of deep and superficial heating.

OBJECTIVE: To compare the effects of deep heating (shortwave diathermy [SWD]) and superficial heating (hydrocollator packs) on tissue extensibility. DESIGN: A double-blind, repeated-measures study. Possible effects of sex and intervention order were controlled. SETTING: A clinical laboratory. PARTICIPANTS: Twenty-four subjects with no neurologic or musculoskeletal pathologies affecting their lower limbs. INTERVENTIONS: Three intervention conditions: deep heating (SWD), superficial heating (hot packs), and no heating were applied in preallocated order to each subject at least 36 hours apart. MAIN OUTCOME MEASURES: Ankle dorsiflexion in weight bearing was measured by using an inclinometer to ascertain changes in the extensibility of the calf muscles and associated soft tissues. RESULTS: Deep heating increased the range of ankle dorsiflexion by 1.8 degrees +/-1.9 degrees . The change in ankle dorsiflexion after superficial and no heating was 0.7 degrees +/-1.5 degrees and -0.1 degrees +/-1.0 degrees , respectively. CONCLUSIONS: Deep heating, in the absence of stretching, increases tissue extensibility more than superficial heating or no heating. Superficial heating is more effective than no heating, but the difference was not statistically significant.

A comparison of true and premodulated interferential currents.

OBJECTIVE: To compare true and premodulated interferential currents (IFCs) in terms of sensory, motor, and pain thresholds; maximum electrically induced torque (MEIT); and comfort. DESIGN: Repeated-measures design. SETTING: Laboratory setting. PARTICIPANTS: University student and staff volunteers. INTERVENTIONS: Participants were exposed to 4 different conditions, chosen to evaluate 2 fundamental differences between true and premodulated IFCs. The conditions were different combinations of (1) premodulated or constant-amplitude currents applied at the skin and (2) crossed or parallel current paths. MAIN OUTCOME MEASURES: Sensory, motor, and pain thresholds; MEIT; and subjective reports of relative discomfort were recorded for each of the 4 conditions. Motor to sensory threshold ratios were subsequently calculated to assess depth efficiency of stimulation. RESULTS: The major findings were that crossed currents (true IFC) had no advantage over parallel currents (premodulated IFC) in terms of motor to sensory threshold ratio, MEIT, or comfort, and that premodulated currents produced higher torque values and less discomfort than constant-amplitude currents (true IFC). These results contradict the claimed superiority of true IFC. CONCLUSIONS: The findings indicate that premodulated IFC, delivered via 2 large electrodes, may be clinically more effective than the traditional true IFC arrangement in terms of depth efficiency, torque production, and patient comfort.

Optimal frequencies for electric stimulation using medium-frequency alternating current.

OBJECTIVE: To determine the effect of single cycles of alternating current of various frequencies on sensory and motor thresholds and on relative thresholds (motor threshold/sensory threshold). DESIGN: Repeated-measures design. SETTING: Laboratory setting. PARTICIPANTS: University student and staff volunteers (N=16; mean age, 34y). INTERVENTIONS: Single cycles of sinewave frequencies between 1 and 35kHz were delivered at 50Hz. The frequencies were applied in a random order. MAIN OUTCOME MEASURES: The motor and sensory thresholds were recorded at each applied frequency. RESULTS: Both sensory and motor thresholds showed a smooth decrease to a minimum at approximately 3kHz. The relative threshold reached a minimum close to 9kHz. Comparison with previous studies showed that although absolute thresholds reach a minimum at a frequency that depends on electrode size, the frequency at which the relative threshold is a minimum was independent of electrode size and independent of whether the stimulus was applied as single pulses or in burst mode. CONCLUSIONS: The optimal frequency for transcutaneous stimulation using medium-frequency alternating current depends on the outcome measure used. It would therefore be desirable for clinical stimulators to provide a selection of carrier frequencies.