Effects of electrical stimulation on edema formation in different strains of rats.

Key Words: Edema, Electrical stimulation, High-voltage pulsed current, Model, Rat strain.

Electrical stimulation (ES), typically high-voltage pulsed current (HVPC), is commonly used along with ice, elevation, and compression to reduce the formation of posttraumatic edema and joint effusion.[1] Neither HVPC alone nor HPVC in combination with other modalities, however, has been demonstrated to be effective in curbing edema formation. Failure to document such treatment effects is attributable, in part, to the difficulty in conducting controlled clinical studies. Injuries causing edema, although common in humans, vary considerably in severity and anatomical location. Other variables, such as age, gender, overall health, time since injury, and medications, would need to be controlled in clinical trials.

To avoid these and other problems inherent in clinical studies, investigators have turned to animal models. The use of such models allows investigators to control the severity and locality of injury as well as other variables in order to determine the effects of individual modalities, such as ES, on edema formation.[2-11]

In the late 1980s, members of our group used a species of frog as a model in experiments designed to determine whether HVPC had any effect on edema formation after acute injury.[2-4,6-10] This group documented, for the first time, that cathodal HVPC at amplitudes less than those required to induce muscle activity exerted a treatment effect and implied that similar results might be expected with humans.[2-4,8] This expectation was supported by the finding of a similar outcome when HVPC was applied to acute injuries induced in two strains of rats (Sprague-Dawley and Zucker-Lean).[5] In subsequent research, however, which we have not published, we found no treatment effect in Sprague-Dawley rats, even when the protocol was only slightly different from that used in our successful trial.

Other investigators[12-14] examining the effects of ES on edema used Sprague-Dawley rats as their model species. These investigators[12-14] reported ES to be ineffective in controlling developing edema or reducing preexisting edema. These studies differed in many aspects, such as when ES was applied (immediately after injury versus 24 to 72 hours after injury),[12] the form of ES applied, and how edema was induced (impact injury versus fluid injection).[13,14] These variables may well account for the different outcomes. Based on our observations, however, another factor may have contributed to the different outcomes; that is, the effect depends on the strain of rat used as a model. We reexamined our only successful rat trial and found that only 1 of the 6 Sprague-Dawley rats included in that trial had clearly responded to HVPC, whereas virtually all of the 14 Zucker-Lean rats did so.[5] The response of the Zucker-Lean rats to HVPC and their larger numbers in our sample masked the dichotomous result. These observations led us to hypothesize that different strains of rats may respond differently to ES. If we and other investigators want to continue to use rats as a model for human responses to ES, then it requires us to demonstrate the viability of this model.

The purpose of our study was to determine whether different strains of rats respond differently to HVPC as a means of curbing edema formation. The three strains examined were Sprague-Dawley, Zucker-Lean, and Brown Norway. As noted, HVPC has been demonstrated to curb edema formation in Zucker-Lean rats and seemingly to have little or no effect in Sprague-Dawley rats.[5] We wanted to confirm those findings by duplicating our earlier study of Sprague-Dawley and Zucker-Lean rats. Anticipating that we would reproduce our previous findings, both positive and negative, for Zucker-Lean and Sprague-Dawley rats, respectively, we included Brown Norway rats in an attempt to discern the more common response of the species.

The protocol used in this experiment was similar to that previously applied by Mendel et al.[5] Both hind limbs were traumatized; one limb was randomly selected to be treated with HVPC, and the other limb served as a control. The limbs were maintained in a dependent position throughout the experiments. The HVPC treatment was started immediately after injury in an attempt to curb the formation of edema.

Method

Subjects

Nineteen Zucker-Lean rats weighing 180 to 280 g ([bar] X = 224, SD = 30),14 Brown Norway rats weighing 210 to 233 g ([bar] X =2 21, SD = 8), and 11 Sprague-Dawley rats weighing 213 to 258 g ([bar] X = 234, SD = 15) were used in this study. Animals were obtained from Harlan Sprague-Dawley Inc(*) and provided with food and water ad libitum until used. The method of anesthesia and handling procedures, including the methods of traumatizing hind limbs and sacrificing the animals, were approved by the Institutional Laboratory Animal Care Committee of the State University of New York at Buffalo.

Instrumentation

Impact injury was induced by a procedure similar to that described by Mendel et al.[5] This procedure consisted of dropping a steel rod weighing 85.5 g through a vertical tube from a height of 30 cm onto the planter aspect of each foot just distal to the malleoli. A rectangular piece of plastic (2 x 2 x 0.05 cm) was interposed between the foot and the tube to distribute the force of the impact. This method of inducing trauma resulted in changes in limb volume that were attributable to edema formation and not to bleeding (de, it caused tissue damage without rupturing major vessels). The skin of these animals is nearly translucent, and we observed no change in color throughout these procedures. The protocol was used on all animals except for five Sprague-Dawley rats to which trauma was induced by dropping the weight from 50 cm. This change in protocol was used to determine whether ES would be more beneficial for a more severe injury. Again, we observed no change in the color of the traumatized feet.

Limb volume was measured by immersing a hind limb and measuring the amount of water displaced in a manner similar to that reported by Mendel et al[5] (Fig. 1). The immersion vessel was 2 cm in diameter and 12 cm long. The inferior end tapered and was attached via a three-way stopcock to a large syringe filled with water. By use of the stopcock and the syringe, the immersion vessel could be filled quickly and efficiently. A 23-gauge stainless steel tube, 3 cm long and bent 90 degrees in the middle, was fixed with epoxy to the inside wall so that 1.5 cm of the tube extended into the immersion vessel. The end of the tube outside the vessel was attached via polyethylene tubing and a 23-gauge needle to a 5-cc syringe. Drawing water into this syringe lowered the water in the immersion vessel to the level of the mouth of the bent tubing affixed to the wall (de, the same level each time). At this level, a piece of 2-0 suture was fixed with white plastic tape to the outside surface of the immersion vessel.

[Figure 1 ILLUSTRATION OMITTED]

Animals, suspended in cloth slings, were prepared for limb volume measurements by painting lines at the level of the malleoli. Animals were then lowered by a motor-driven boom until the line painted on the hind limb was aligned with the thread fixed to the immersion vessel (at the level of the tip of the stainless steel tube). Water displaced by immersion of a foot was drawn into the 5-cc syringe, which was then disconnected from the needle hub and weighed on a microbalance (Fisher Scientific model S-300D).[dagger] The weight of the fluid collected (less the tared weight of the syringe) was equivalent to the animal's limb volume on the basis of 1 g being equal to 1 mL.

Before our experiment, we established the reliability and validity of the volume measurement system by determining the volume displacement of a small aluminum cylinder (1.270-cm diameter, 3.160-cm length). The use of an inanimate object to determine reliability and validity was deemed more appropriate than the use of volume measurements from animals because it has been shown that, in the dependent position, volumes of nontraumatized limbs change over time.[15] The mean cylinder volume, as determined from 13 consecutive measurements, was 3.949 mL (SE = 0.007, range = 3.9193.982). When the cylinder volume was measured 13 more times on another day, the mean was 3.940 mL (SE = 0.001, range = 3.934-3.947). The volume of the cylinder calculated from physical dimensions was 4.003 mL. This cylinder volume is 0.054 mL and 0.063 mL larger than the volumes of the cylinder measured by our method on the two respective occasions. This difference represents an underestimation of cylinder volume by our measuring method of 1.57% or less. We believe that our measurements of volume displacement are highly reliable because of the small standard error for repeated measurements. We believe that they are also valid because our method produced volume measurements that were very close to that determined from the physical dimensions of the cylinder.

High-voltage pulsed current was applied by Intelect 500S stimulators[double dagger] as described previously by Mendel et al.[5] Pulses, delivered at 120 pulses per second (pps), consisted of a unidirectional twin-spike pulsed current with an interphase interval of 75 microseconds. A carbon-rubber electrode[sections] (approximately 9 x 3.6 cm), which functioned as the anode, was coated with electrode gel and applied to an animal's shaved abdomen, where it was held in place by the cloth sling in which the animal was suspended. A carbon-rubber electrode was immersed in each of two 100-mL beakers, one beaker for each hind limb, so that water in the beakers served as a cathode. Current flow (+ to -) was proximal to distal through the sites of injury.

Because anesthesia may cause body temperature to fall,[16] body temperature was regulated throughout these experiments. A rectal probe was inserted about 6 cm and connected to a YSI Tele-thermometer[parallel] to allow body temperature to be monitored throughout an experiment. Body temperature was regulated at 37 [degrees] [+ or -] 1 [degrees] C by directing a 60-W lamp on or off an animal. The temperature of the water serving as the cathode was also regulated at approximately what we considered the thermoneutral temperature for hind limbs (32 [degrees] [+ or -] 2 [degrees] C). Water temperature was monitored continuously with YSI Tele-thermometers.

Procedure

Each rat was anesthetized by an intraperitoneal injection of sodium pentobarbital (60 mg/kg of body weight). The anesthesia was supplemented over the course of the 4-hour experiments, as needed, with 0.05-mL doses. After the animals were anesthetized, their hind limbs and abdomens were shaved and the animals were placed in cloth slings and suspended at 45 degrees (caudal end down) with both hind limbs fully exposed and in the dependent position. Lines were then painted at the level of the malleoli, and rectal probes were inserted.

After the animals had been suspended for approximately 20 minutes, the volume of each hind limb was determined (pretrauma). Each hind limb of each rat then was injured by dropping a steel rod onto the planter aspect just distal to the malleoli. Volumes of hind limbs were measured again (posttrauma, before the first treatment) and, within 5 minutes after injury, each hind limb was immersed in its own 100-mL beaker. Voltage output from the stimulator then was increased slowly until a visible contraction of digital muscles was observed in both feet. Voltage was noted and reduced until the contraction disappeared. One limb was randomly selected to receive cathodal HVPC at 120 pps and at an amplitude 10% below that needed to induce a visible motor threshold. The control limb was maintained in its beaker during the treatment phase, but stimulation was turned off for the remainder of the experiment. Treatment continued for 30 minutes and was followed immediately by determination of the volume of each limb. After removal from water-filled beakers and before volume measurement, shaved limbs were dabbed with tissue paper to remove adherent water and to minimize evaporative cooling. Limbs were not rubbed or squeezed during drying. Posttreatment limb measurements were followed by a 30-minute rest period, during which limbs were suspended above the beaker. A rest period was followed by another measurement of limb volumes. This sequence of events continued for three more cycles so that at the conclusion of a session, each animal had received four treatments and four rests. Times selected for measuring limb volumes matched those in previous studies but ultimately were arbitrary.

Data Analysis

To minimize the effects of size on the amount of swelling, data were expressed as changes from pretrauma hind-limb volumes per kilogram of body weight. An analysis of variance for repeated measures was used to test the null hypothesis that HVPC treatment would not influence posttraumatic limb volumes in a particular strain of rat. A level of significance of .05 was selected. Tukey's post hoc t tests were used to further assess differences between treated and untreated limb volume changes in Zucker-Lean, Brown Norway, and Sprague-Dawley rats at selected times postinjury.

Results

Sprague-Dawley Rots

Volumes of treated limbs were not different from those of untreated limbs at any time throughout the study, regardless of whether trauma was induced by a weight dropped from a height of 30 or 50 cm (P = .9757; Fig. 2). Hind limbs of Sprague-Dawley rats did swell, as there was an increase (P [is greater than] .0001) in the volumes (about 1 mL/kg) of both limbs over time.

[Figure 2 ILLUSTRATION OMITTED]

Zucker-Lean Rats

Volumes of treated limbs ([bar] X = 0.586 mL/kg) were smaller than volumes of untreated limbs ([bar] X = 0.874 mL/kg) after the second treatment with HVPC (Fig. 3). Volumes increased over time for both limbs, although the increase for the untreated limbs ([bar] X = 1.112 mL/kg) was considerably larger than that for the treated limbs ([bar] X = 0.767 mL/kg) (de, HVPC curbed edema formation but did not preclude swelling).

[Figure 3 ILLUSTRATION OMITTED]

Brown Norway Rats

Volumes of treated limbs ([bar] X = 0.057 mL/kg) were smaller than volumes of untreated limbs ([bar] X = 0.352 mL/kg) after the first HVPC treatment (Fig. 4). Over time, both limbs increased in volume, although the increase was larger in the untreated limbs ([bar] X = 0.797 mL/kg versus [bar] X = 0.425 mL/kg for treated limbs). The initial HVPC treatment returned limb volume to nearly the initial pretrauma value, and each subsequent treatment reduced limb volume in comparison with the preceding limb volume measured after a 30-minute rest (no treatment) period. Volumes of control limbs also decreased during treatment periods, albeit not as much as those of treated limbs. Therefore, HVPC curbed but did not prevent edema formation.

[Figure 4 ILLUSTRATION OMITTED]

Discussion

Cathodal HVPC at 10% less than that needed to induce the visible motor threshold curbed edema formation in Zucker-Lean and Brown Norway rats after impact injury. This result occurred even though limbs were in the dependent position throughout the experiment. In contrast, HVPC had no effect on edema formation in Sprague-Dawley rats after a similar impact injury (a weight dropped from a height of 30 cm) and a more extreme impact injury (a weight dropped from a height of 50 cm). The treatment effects confirm the findings of Mendel et al[5] for Zucker-Lean rats and are in keeping with previous findings obtained with a frog model.[2-4,8]

The lack of any treatment effect in the Sprague-Dawley rats parallels results from unpublished studies in our laboratory and prompts us to suggest that the physiology of Sprague-Dawley rats may be different somehow from that of other strains of rats. Even among strains that responded to HVPC, there were differences. Both Zucker-Lean and Brown Norway rats responded to HVPC, but Zucker-Lean rats did so only after the second treatment (as they had previously[5]), whereas Brown Norway rats responded after the first treatment. Untreated limbs of Zucker-Lean rats examined here swelled to virtually the same total extent as had untreated limbs in our previous examination of rats of this strain,5 but limb volumes during the first 3 hours were higher in the present study than they were previously. Volumes of treated limbs were similarly larger in the present study than in our previous examination. Volumes of treated limbs in the present study were different from, but closely paralleled, those of untreated limbs. That is, volumes of treated and control limbs were statistically different, but both tracked upward over time. In contrast, volumes of treated and untreated limbs diverged over time in our earlier effort. We did not monitor or regulate body temperature or beaker water temperature in our previous study, but we did so in the present study. Because untreated limbs of Zucker-Lean rats in the present and previous studies swelled to approximately the same extent over 4 hours, we conclude that body temperature probably was not a significant factor. We reason that if body temperature were markedly lower in one group, then total swelling should be less extensive in that group than in the group in which body temperature was maintained at nearly normal levels.

Volumes during the first 3 hours of the present Zucker-Lean rat experiment were higher than in our earlier study.[5] This finding suggests that water temperature in the "treatment" beakers may have been important. Application of HVPC clearly affected limb volumes in both studies; volumes of treated limbs were smaller than those of untreated limbs. Volumes of treated and untreated limbs were larger during the first 3 hours when beaker water was maintained near thermoneutral temperature than when that water was somewhat cooler (at room temperature). That is, the divergence of volumes of treated and untreated limbs observed in our earlier study[5] may have resulted from the combined effects of HVPC and relatively cool water.

Volumes of both treated and untreated limbs of all three strains tended to increase, sometimes quite markedly, during rest periods and to level off or even decrease during treatment periods (Figs. 2-4). This tendency was most obvious with Brown Norway rats (Fig. 4) but was evident with the other strains as well. This tendency was less obvious with the treated limbs in the first Zucker-Lean rat study,[5] but those limbs swelled very little. Why should increases in the volume of untreated limbs slow, stop, or even reverse when the contralateral limbs are being treated.? Mohr et al,[12] like Suckert[17] and Haussmann and Kohnlein,[18] suggested that trauma induced in one limb caused an increase in the volume of the contralateral (uninjured) limb and implied that treatment of injured limbs might influence volumes of uninjured limbs (in our case, contralateral injured limbs). Although other researchers[19-21] have observed changes in contralateral limbs after unilateral injuries, we remain unconvinced that a true crossover effect exists. We believe that the effect observed by us as well as by Suckert[17] and Haussmann and Kohnlein[18] might have resulted from suspending the animals during volume determinations. This factor could have caused the formation of edema in both dependent limbs. That is, the swelling in uninjured limbs was gravity induced.

We have demonstrated that uninjured frog limbs contralateral to injured limbs swell over time, but no more than limbs of uninjured animals suspended with limbs in the dependent position.[15] Levine et al[19] had to injure limbs for 3 consecutive days before observing any changes in contralateral limbs. Other researchers[20,21] measured effects on contralateral limbs days after injury. We injured both limbs of rats only once and sacrificed the rats approximately 4 hours posttrauma. If repeated traumas or long periods after trauma are required to observe changes in contralateral limbs, then such conditions did not pertain in our study.

We suspect that the apparent treatment effect in the untreated limbs during treatment of contralateral limbs and the simultaneous increases in volumes of treated and untreated limbs during rest periods (Figs. 2-4) stem from immersion in water of both feet during treatment periods and withdrawal from water during rest periods. Submersing both limbs during treatment periods would subject both limbs to slight pressure, which could resist swelling. Conversely, withdrawing both feet from water during rest periods would remove the pressure and allow both feet to swell. If the latter were true, then HVPC, as applied here, and even minimal pressure (calculated to be about 4.5 g/[cm.sup.2] at a depth of 4.5 cm) would seem to have additive effects and to support the common clinical practice of simultaneously applying these two modalities. By leaving both limbs submerged during rest periods, we should, in future experiments, be able to distinguish how much HVPC adds to compression of about 4.5 g/[cm.sup.2] in curbing acute edema.

That different strains of rats respond differently to a putative therapy raises several questions and suggests some possibilities. An obvious question is: Which strain of rat is a predictor of how HVPC will affect edema formation in humans? Stated another way, are rats useful as models for humans with regard to the effect of HVPC on trauma-induced edema formation? If two of three rat strains respond positively and the third strain does not respond, it seems that the more common response is more likely to be typical of the species. However, we sampled only 3 of the 20 to 30 common rat strains used in research. The results of our study, therefore, do not allow us to identify the typical response for rats or to determine whether rats are a viable model for humans with regard to the effects of ES on an aspect of the inflammatory response. However, when our results are combined with the results of related studies on other species of rodents and even on frogs, we believe that the findings from this study shed light on these questions. One species of frog[2-4,8] (class: Amphibia) and members of at least two genera of rodents[5,22,23] (class: Mammalia) respond positively to HVPC. This finding suggests that the underlying physiology most likely is derived from a common ancestor rather than that it evolved separately and independently in several lineages. If this suggestion is true, then Sprague-Dawley rats may not be suitable as models, but rats in general probably are.

Other groups investigating the effects of ES on edema all have used Sprague-Dawley rats. How and when various investigators applied ES and even the form of ES varied from study to study.[12-14] These variations may well account for the lack of a treatment effect (de, curbing or reduction of edema). Indeed, we have argued, on theoretical grounds,[24] that the expectation that HVPC at least will reduce existing edema is not likely to be met. Even so, it is also possible that the strain of rat used may have contributed to the negative results reported by these investigators.

Our finding that edema is curbed by HVPC in some strains of rats but not in others suggests that an underlying difference in physiology may exist between strains. Understanding this difference may aid in determining the site and mechanism of action of ES. Building on the initial suggestion of Reed,[22] we currently believe that HVPC curbs edema by influencing the behavior of the endothelial lining of postcapillary venules, the site at which most protein and water exchange typically occurs between the vasculature and the interstitium. We speculate that HVPC alters the permeability of these microvessels by somehow prohibiting individual endothelial cells from actively or passively changing their shape from flat to more rounded, as they typically do under the influence of inflammatory mediators such as histamine, serotonin, bradykinin, and other autacoids released after injury. This rounding up creates gaps between adjacent endothelial cells and allows proteins and other formed elements and their attendant water to escape the vascular tree, resulting in edema. In a previous study,[23] we found that fewer labeled protein molecules leak from capillaries of cheek pouches of hamsters superfused with physiologic levels of histamine when those pouches are stimulated with HVPC similar to that used in the present study. We infer that HVPC acts on the endothelium as noted earlier, but we cannot observe the behavior of individual endothelial cells in an intact and functioning animal. However, we can potentially culture cells from both "responding" and "nonresponding" strains of rats and begin to look for morphologic and biochemical differences that may account for their different behaviors. Such experiments could reveal a great deal about how ES affects edema formation.

More practically, results of our study serve to alert researchers to the dangers of using animal models to shed light on treatments used for humans. Although humans share a great deal of their genomes with fellow vertebrates, especially mammals, there are differences. In the current state of ignorance, it is difficult to judge a priori whether a particular aspect of the anatomy or physiology of some model species is an appropriate model for humans. Similarities among closely related taxa typically exceed differences, but even small differences can result in major changes in behavior and physiology.

Conclusions

We believe that rats in general probably serve as useful models for humans with regard to aspects of the inflammatory response that we explored. If we are correct, then our results augment those of previous research suggesting that ES, in particular cathodal HVPC at 10% less than that needed to induce visible muscle contraction, can influence edema formation if applied immediately after injury (de, when edema is still forming). The importance of these findings derives from the fact that they are among the few[2-5,8,22,23] that suggest that the clinical practice of applying HVPC to acute injuries to control edema formation may have merit. To date, no clinical evidence exists that HVPC or any other form of ES positively affects edema formation. Only the work of Griffin et al[25] suggests, but does not demonstrate, that ES may affect chronic posttraumatic edema.

Acknowledgments

We thank Sanjay Patel and Linda Gottstein-Yerke for their invaluable help in collecting some of the data presented here. We also thank John Nyquist for the artwork.

(*) Harlan Sprague-Dawley Inc. PO Box 29176. Indianapolis, IN 46229-0176.

([dagger]) Fisher Scientific, 711 Forbes Ave. Pittsburgh. PA 15219.

([double dagger]) Chattanooga Corp, PO Box 4287, Chattanooga, TN 37405.

([sections]) Tenszcare, 3M Corp, Bldg 225-5S-01, 3M Center, St Paul, MN 55144.

([parallel]) Yellow Springs Instrument Co Inc, 1725 Brannum Ln, Yellow Springs, OH 45387.

References

[1] Newton RA. High-voltage pulsed current: theoretical bases and clinical applications. In: Nelson RM, Currier DP, eds. Clinical Electro-therapy. East Norwalk, Conn: Appleton & Lange; 1991:201-220.

[2] Bettany JA, Fish DR, Mendel FC. Influence of cathodal high voltage pulsed current on acute edema. J Clin Electrophysiol. 1990;2:5-8.

[3] Bettany JA, Fish DR, Mendel FC. High-voltage pulsed direct current: effect on edema formation after hyperflexion injury. Arch Phys Med Rehabil. 1990;71:677-681.

[4] Bettany JA, Fish DR, Mendel FC. Influence of high voltage pulsed direct current on edema formation following impact injury. Phys Ther. 1990;70:219-224.

[5] Mendel FC, Wylegala JA, Fish DR. Influence of high voltage pulsed current on edema formation following impact injury in rats. Phys 7 hen 1992;72:668-673.

[6] Mendel FC, Caputi CD, Karnes JL, Fish DR. High voltage pulsed current using surface electrodes: effects on acute edema formation after hyperflexion injuries in frogs. J Orthop Sports Phys Ther. 1992;16: 140-144.

[7] Taylor K, Fish DR, Mendel FC, Burton HW. Effect of electrically induced muscle contractions on posttraumatic edema formation in frog hind limbs. Phys Ther. 1992;72:127-132.

[8] Taylor K, Fish DR, Mendel FC, Burton HW. Effect of a single 30-minute treatment of high voltage pulsed current on edema formation in frog hind limbs. Phys Ther. 1992;72:63-68.

[9] Fish DR, Mendel FC, Schultz AM, Gottstein-Yerke LM. Effect of anodal high voltage pulsed current on edema formation in frog hind limbs. Phys Ther. 1991;71:724-730.

[10] Karnes JL, Mendel FC, Fish DR. Effects of low voltage pulsed current on edema formation in frog hind limbs following impact injury. Phys Ther. 1992;72:273-278.

[11] Karnes JL, Mendel FC, Fish DR, Burton HW. High-voltage pulsed current: its influence on diameters of histamine-dilated arterioles in hamster cheek pouches. Arch Phys Med Rehabil. 1995;76:381-386.

[12] Mohr TM, Akers TK, Landry RG. Effect of high voltage stimulation on edema reduction in the rat hind limb. Phys Ther. 1987;67:1703-1707.

[13] Cosgrove KA, Alon G, Bell SF, et al. The electrical effect of two commonly used clinical stimulators on traumatic edema in rats. Phys Ther. 1992;72:227-233.

[14] Cook HA, Morales ML. La Rosa EM, et al. Effects of electrical stimulation on lymphatic flow and limb volume in the rat. Phys Ther. 1994:74: 1040-1046.

[15] Fish DR, Mendel FC, Bettany JA. Author response to letter to the editor [Research design considerations in studying edema]. Phys !Ther 1990;70:584-586.

[16] Green CJ. General principles. In: Green CJ. ed. Animal Anesthesia. London, England: Laboratory Animals Ltd; 1982:9-16.

[17] Suckert VR. Experimentelle Modelle fur traumatische Rattenpfotenodeme. Int J Exp Med. 1967;13:43-50.

[18] Haussmann P, Kohnlein HE. Animal experimental examinations of medicinal influence of post-traumatic rat paw edema. Fortschr Med. 1974;92:301-303.

[19] Levine JD, Dardick SJ. Basbaum AI, Scipio E. Reflex neurogenic inflammation, I: contribution of the peripheral nervous system to spatially remote inflammatory responses that follow injury. J Neurosci. 1985;5:1380-1386.

[20] Kleinrensink GJ, Stoeckart R, Meulstee J, et al. Lowered motor conduction velocity of the peroneal nerve after inversion trauma. Med Sri Sports Exerc. 1994;26:877-883.

[21] Bullock-Saxton JE. Local sensation changes and altered hip muscle function following severe ankle sprain. Phys Ther. 1994;74:17-28.

[22] Reed BV. Effect of high voltage pulsed electrical stimulation on microvascular permeability to plasma proteins: a possible mechanism in minimizing edema. Phys Ther. 1988;68:491-495.

[23] Taylor K, Mendel FC, Fish DR, et al. Effect of high-voltage pulsed current and alternating current on macromolecular leakage in hamster cheek pouch microcirculation. Phys Ther. 1997;77:1729-1740.

[24] Mendel FC, Fish DR. New perspectives in edema control via electrical stimulation. Journal of Athletic Training. 1993;28:63-64, 66-72, 74.

[25] Griffin JW, Newsome LS, Stralka SW, Wright PK. Reduction of chronic posttraumatic hand edema: a comparison of high voltage pulsed current, intermittent pneumatic compression, and placebo treatments. Phys Ther. 1990;70:279-286.

RM Thornton, PhD, PT, is Rehabilitation Therapist, Rehabilitation and Research Center, Medical College of Virginia Hospitals, Richmond, Va.

FC Mendel, PhD, is associate Professor, Department of Anatomy and (tell Biology, State University of New York at Buffalo, 317 Farber Hall, Buffalo, NY 14214 (USA) ([e-mail]). Address all correspondence to Dr Mendel.

DR Fish, PhD, PT, is Associate Professor, Department of Physical Therapy, Exercise, and Nutrition Sciences, State University of New York at Buffalo.

This study was approved by the Institutional Laboratory, Animal Care Committee of the State University of New York at Buffalo.

Some of the work presented here was sponsored by Grant 2T32 HD0742306 from the National Institutes of Health, National Institute of Child Health and Human Development.

This article was submitted March 26, 1997, and was accepted October 10, 1997.