Women's Sexual Function and Dysfunction. Irwin Goldstein MD

Blood flow: heated electrodes

Roy J Levin

Introduction: measuring vaginal blood flow with heated electrodes

Vaginal blood flow assessed by the heated oxygen electrode

The heated oxygen electrode employed was a commercial electrode (E5250) and oxygen monitor (TCM1) manufactured by Radiometer (Copenhagen, Denmark) for use in intensive care units to measure the arterial partial pressure of oxygen (PO2) transcutaneously.* In essence, it is a Clarke-type oxygen electrode fitted with a temperature-measuring device and an electrical heater that could be set to heat up the skin to a depth of about 3 mm.1 The electrode is normally applied to the skin surface by a special adhesive holder. The temperature of the electrode can be set to temperatures above that of blood so that the tissue underneath becomes maximally vasodilated, allowing the fast-flowing capillary blood to be similar to that of arterial blood, and the oxygen that diffuses out of the blood onto the surrounding tissues closely mirrors that of the arterial partial pressure of oxygen. While this itself is a useful measure of the blood perfusion of the tissue, the real bonus of the electrode is that it can be set to operate at a particular temperature above that of blood (say, 41°C), and the electronic circuit of the device gives a readout of the electrical power (in milliwatts) needed to maintain the device at that set temperature. This power is a measure of the heat loss of the electrode, representing both (1) the heat loss by conduction into the tissue beneath the electrode, influenced mainly by the blood flow and (2) heat losses from the housing of the electrode to the environment by convection, conduction, and radiation. If the heat loss by conduction to the skin because of the blood flow is very much greater than the heat loss due to the other nonspecific process, then the actual power consumption of the electrode will be a reasonable indirect measure of the blood flow underneath the heated electrode. The attachment of the electrode to the relatively dry skin is by a special holder with an adhesive surface. This cannot be used in the vagina, as the walls of the organ are usually coated with a slippery, basal fluid secretion. Wagner and Levin2 designed a holder the electrode can be placed in that is then held onto the vaginal wall by a grooved suction rim (Fig. 10.4.1). The grooved rim is connected to a vacuum pump and the device is held onto the vaginal wall by the suction, which is sufficient to hold the electrode in contact against the vaginal epithelium even during the vaginal contractions of an orgasm. It is thus possible to follow the changing power consumption throughout sexual stimulation, including an orgasm.

* The original Radiometer heated oxygen electrode has now been replaced by a solid-state electrode, but the specifications are the same as the E5250. The present model, TCM4 monitor, still allows the power measure to be obtained, with the added advantage that the trend curve can be shown and printed out. Radiometer’s website is www.tc-monitoring.com.

The graphs in Fig. 10.4.2 show the changes in blood flow of the vaginal walls during arousal to orgasm (and after) by clitoral stimulation in three subjects. Two separate heated oxygen electrodes were employed attached to the lateral walls on either side of the vagina. The practically identical changes measured by both electrodes in all the subjects show that the increases in blood flow induced by the clitoral stimulation is not discrete but involves the whole vagina.

Obviously, to give the best index of the blood flow, the heat losses by the non-blood-flow processes need to be kept to the minimum. The ideal correction for these heat losses would be to attach the electrode to the vaginal wall and stop all blood flow; the heat lost (power consumption) would then be only that from non-blood-flow processes. Of course, while this can be achieved in an animal vagina by clamping of the arterial blood supply, it cannot be accomplished in the human vagina in situ. In humans, the best correction that can be made, if high accuracy is demanded and it is felt essential, is to attach the electrode to the volar surface of the subject’s arm and measure the power needed to keep the electrode at its vaginal fixed temperature with normal skin blood flow and after the inflation of a blood pressure cuff to approximately 200 mmHg to stop arterial inflow to the arm (Fig. 10.4.3). The amount of power used when the arterial blood flow to the arm is blocked off can be used as a correction for the non-blood-flow heat losses. It can be subtracted from the original readings of the power consumption (“calculated blood flow index” in Fig. 10.4.3). After the correction, the shape of the blood flow curve is identical to that previously obtained; the only difference is that the actual power consumption figures are now obviously much smaller and mirror more accurately the blood flow.

Figure 10.4.1. The heated oxygen electrode (model E5250, Radiometer, Copenhagen) mounted inside the suction capsule for attachment to the vaginal wall. The short tube opens into the groove around capsule's rim and is where the suction is applied. The scale beneath is in centimeters (reproduced with permission from Wagner and Levin2).

Another simple experiment confirms that the electrode loses heat to the vaginal wall and external milieu even if there is no blood flow underneath the electrode. An electrode was attached to the vaginal wall, and a subepithelial injection of adrenaline was given immediately underneath the electrode. This causes the contraction of the peripheral blood vessels underneath the electrode, as indicated by the sharp drop to zero of the surface partial pressure of oxygen (Fig. 10.4.4). There is still, however, a significant loss of heat by the electrode to the environment and to the underlying tissue by conduction.

For many practical purposes, the direct reading of the electrode’s power consumption without correction gives a reasonable index of the vaginal blood flow and congestion (pooling). As heat loss is proportional to the difference in temperature between the electrode’s surface and the tissue it is resting on, if there is blood pooling as well as blood flow, the loss of heat will be affected by the volume of fluid that the heat is being lost to, because a large volume of fluid can absorb a greater amount of heat for a smaller increase in its temperature. Thus, heat loss from any heat electrode attached to the vaginal wall will be influenced both by tissue blood flow and by blood pooling.

The heat electrode technique has been used in a number of laboratory studies. It was first employed in the vagina to measure the changes in oxygen tension of its epithelial surface,2 and in later work on vaginal wall blood flow during arousal to orgasm.3,3a In other studies, it was used: to ascertain which hemodynamic measure gives the best assessment of sexual arousal;4 to investigate the effect of atropine on vaginal blood flow, sexual arousal, and climax;5 to monitor simultaneously the vaginal hemodynamics by three independent methods during sexual arousal;6 to record the vaginal blood flow and the duration, intensity, and latency of orgasm in women in the laboratory;3 to evaluate the action of thyrotropin-releasing hormone on vaginal blood flow in conscious humans and anesthetized sheep;7 and to monitor the veracity of claimed subjective sexual arousal in a subject in the laboratory.8 Since its introduction, the heated oxygen electrode has been used by a number of other workers in a variety of studies on vaginal blood flow (see review by Levin9 for references). One group, overlooking the initial employment and subsequent early studies by Wagner and Levin, claimed, in 2001, that it was a new technique for assessing female arousal (see comment by Levin and Wagner10).

Advantages of the heat electrode

The advantages of using the heated oxygen electrode as a device to monitor vaginal blood flow are: (1) it can be employed throughout all phase of sexual arousal, even when orgasm is induced, as it is not affected by movement; (2) the data it produces are transferable; (3) the basis of its measurement (that is, heat loss to the tissue) is reasonably well understood; (4) although it is intrusive (residing in a body cavity), it is not invasive (does not go beneath the skin layer); (5) it is relatively safe; (6) it is commercially available and relatively cheap; (7) repeated measures can be obtained in the same subject by heating and allowing the electrode to cool and then reheating; (8) it can even be used during menstruation if the surface of the vagina is first wiped free of any blood (the photoplethysmograph cannot be used if there is free blood in the vagina) (see Chapters 10.1-10.3).

Figure 10.4.2. Power consumptions of two heated oxygen electrodes attached to the right and left hand sides of the vaginal walls in three subjects during sexual arousal by clitoral stimulation (C.S.) to orgasm (self-graded in the encircled number on a scale 1 = poor to 5 = excellent). In all the records, the decrease in electrode power consumption after orgasm is not very marked, and it slowly returns to the prearousal levels, suggesting that a single orgasm does not dissipate rapidly the vaginal blood flow/congestion. The first subject (SMD 1322878) induced a second orgasm 16 min after the first; note the very obvious decreased power response to the second arousal and orgasm compared with the first (Levin and Wagner, unpublished data), with the moderate decrease in subjective assessment of the second orgasm (from 4 to 3).

Disadvantages of heat electrode

Its main disadvantages are: (1) it needs to be applied to the vaginal surface by the experimenter, and this appears to be a drawback in some cultures, although having a female laboratory assistant or nurse do the application satisfies most ethical committees; (2) it should not be left stuck onto the vaginal wall for periods greater than 2-3 h); (3) its unit of measurement (milliwatts) is indirect rather than in direct terms of flow (ml/100 g tissue/min).

The clearance concept and the heat electrode

Measuring the blood flow in human organs in situ has always been a challenge, as it is usually impossible to measure the arterial inflow or to cannulate the venous drainage and directly collect all the flow coming from the organ. Because of this difficulty, various indirect methods of measuring blood flow have been developed, many using the concept of “clearance”. In essence, a substance is injected into the organ or applied to a tissue surface and the rate of its local disappearance monitored; this local rate is the clearance of the substance from the site, and it is practically exclusively removed by the blood.

Substances employed for the clearance have to be nontoxic and not metabolized, must equilibrate rapidly between tissue and blood, and must be easily and accurately estimated. Examples are radioactive sodium (Kety11) and xenon-133 (Lassen et al.12). Wagner and Ottesen13 were the first to employ the clearance of xenon-133 to estimate the vaginal blood flow in the quiescent, nonsexually aroused state (median 10, range 6-20 ml/100 g tissue per min, n = 7) and its increase during sexual arousal (median 29, range 22-45 ml/100 g tissue per min). However, because the technique is invasive (it uses an intraepithelial needle injection of the xenon-133 in saline) and uses a radioactive gas, it has remained a research technique and has not been employed by others for routine clinical studies on the vaginal blood flow.

Changes in the extraction of heat from a variety of probes has often been used to monitor tissue blood flow indirectly, but because heat fits all of the requirements for an ideal washout “substance”, it can be used to calculate the tissue’s blood flow in real terms of ml/100 g tissue per min as if it were an actual “substance”.

The heated oxygen electrode used for “heat washout” experiments

At the Panum Institute in Copenhagen, Middttun et al.14 developed a simple technique to measure the blood flow of the human skin with a modified Radiometer transcutaneous partial pressure of oxygen electrode. The disk electrode was attached to the skin and the initial skin temperature (Tb) measured. The electrode was heated until the underlying tissues reached the selected steady-state temperature (usually 37-45°C). The heat was then switched off, and the temperature, T, was recorded every 10 s until a stable baseline temperature, Tb, similar to that obtained before the heating, was obtained. On the forearm, this took 5-10 min, but on the pulp of the thumb (with a much higher blood flow), this could take as little as 1.5-3 min. When plotted semilogarithmically against time in seconds, the differential rate of cooling of the electrode (T = T - Tb), after the first few curvilinear readings presumably caused by heat equilibration between probe and tissue, yields a straight line, indicating an exponential heat washout. The blood flow (in ml/100 g tissue per min) was calculated by Kety’s11 formula, in which X, the tissue-to-blood partition coefficient for heat, was assumed, for simplification, to be 1 ml/g (its actual value is 0.954 ml/g) (Perl and Cucinell15), and the slope of the straight line graph k equal to loge 2 divided by the half-time of the change in T.

Figure 10.4.3. Electrode power consumption of a heated oxygen electrode (model E5250, Radiometer, Copenhagen) mounted on the volar skin surface of a female subject's forearm. On application of the pressure to 200 mmHg in the arterial pressure cuff around the arm (CUFF ON) for 5 min, there is a fall in pO2 on the skin surface of approximately 50% and a decrease in the power consumption of approximately 15 mW. Subtracting the power consumption that remains after the pressure is applied yields the curve of calculated "blood flow" index, which is of identical shape to uncorrected values (Levin and Wagner, unpublished data).

Figure 10.4.4. Vaginal surface pO2 and electrode heat power consumption (in milliwatts) before and after subepithelial injection of adrenaline under the electrode. Note that the surface pO2 falls within 1 min to zero, but that the power consumption initially falls in 1 min by approximately 30 mW and then slowly decreases to a 50-mW fall (Levin and Wagner, unpublished data).

Levin and Wagner16 adapted the technique to measure human vaginal blood flow. The Radiometer heated oxygen electrode was held against the vaginal surface by the same suction holder as that previously employed for oxygen measurements. The unheated, steady-state temperature of the vaginal surface was recorded. The electrode was usually heated to 41-43°C and then switched off. The cooling curve of the electrode’s temperature was followed every 5-10 s until the temperature reached the previously recorded unheated state. This usually took 2-3 min. The initial basal temperature was subtracted from each of the cooling temperatures and plotted semilogarithmically to obtain the slope of the linear portion of the cooling curve. An example of the results obtained is shown in Fig. 10.4.5, where the subject had her basal vaginal blood flow recorded initially, and then the flow was measured when she actively sexually fantasized. The much steeper slope of the latter graph shows the greater heat clearance induced by the mental sexual arousal. A feature of the technique that at first seems remarkable is that whatever temperature the electrode is initially set at (within the limits of 37-43°C) to obtain the blood flow by heat clearance, the same calculated flow is obtained.14 In one subject, for example, with the electrode set at 39°C, 41°C, and 43°C, the calculated vaginal blood flows in three separate serial measurements were approximately 80, 70, and 70 ml/100 g per min, respectively.

Figure 10.4.5. Heat washout responses of the heated oxygen electrode mounted on the vaginal wall of a subject in a basal, nonstimulated condition and when sexually aroused by fantasy. In the basal condition, her vaginal flow was 44 ml/100 g tissue per min, which increased to 173 ml/100 g per min by sexual fantasy (reproduced with permission from Levin and Wagner16).

One weakness of the method using the unmodified Radiometer electrode is the significant heat loss from the electrode that is not related to blood flow but is due to heat losses, especially to the surrounding environment from the back of the electrode. In order to reduce these losses as much as possible, a simplified Sheffield heat electrode was designed, developed, and manufactured at the Department of Biomedical Science, University of Sheffield, UK, by Tony Carter, one of the department’s electronic technicians, and myself (Fig. 10.4.6).

Figure 10.4.6. A diagrammatic representation of the electrical circuitry used in the Sheffield heat electrode (probe). See text for details (copyright RJ Levin a T Carter).

The simplified Sheffield heat electrode used for heat “washout”

The Sheffield heat electrode dispensed with the oxygenmeasuring device and was simply a gold disk (diameter approximately 18 mm) heated electrically by a surface-mounted, 47 Q, 3-W ceramic resistor from a 12-V, subject-isolated power supply with an analog output circuit to record its temperature. The disk, heater, and temperature sensor were housed in a ceramic (Makor) holder (diameter 25 mm, height 14 mm) that also had a second “shield” heater circuit behind the disk that was driven to the same temperature as that of the disk. This reduced heat losses from the back of the housing to the environment. The electrode could be used with the shield heater circuit switched on or off. The disk was held in contact against the vaginal wall by suction through a 1-2-mm-wide and -deep groove cut around the periphery of the front face of the holder connected to a portable clinical vacuum pump (Fig. 10.4.7).

The whole unit needed to be sealed very carefully to make it waterproof so that it could be routinely chemically sterilized. Good sealing is essential because the fluids used for sterilizing the electrode between subjects have searching qualities and can find the smallest of entry points into the electrode’s inner compartment and destroy the electrical circuits. A number of early prototype models were damaged beyond repair because of this problem. All electrodes were calibrated against an accurate glass thermometer and could read temperatures to 0.1°C.

The control unit consists of a heater digital display so that the temperature that the heating current creates in the disk can be set while a second digital display gives the actual temperature of the disk when the heater is switched off. Other switches control the shield heater, and one turns the whole unit off. A fail-safe circuit was built in so that the temperature of the gold disk can never rise beyond 47°C. Three electrodes were constructed and calibrated; each one loses a different amount of heat to the environment, so that when corrections are made for this heat loss, the actual electrode used must be known. The corrections for each electrode’s heat loss were obtained by applying the electrode to the volar surface of the subject’s arm and measuring the cooling of the electrode (initially run at the temperature used in the vagina) on the arm with an arterial cuff at 200 mmHg. The calculated “blood flow” from this cooling curve can then be subtracted from that obtained in the vagina, yielding a lower corrected vaginal blood flow.

Figure 10.4.7. The Sheffield heat electrode and the photoplethysmograph used in the study combining simultaneous photoplethysmography and heat electrode recordings. The two attachments to the heated electrode are the translucent suction tubing and the cabling for the electronics. The ballpoint pen is 140 mm long and 7 mm wide. See text for details.

Repeated measures of vaginal blood flow in the same subject can be obtained simply by allowing the electrode to cool down to the original vaginal temperature and then heating it again, switching it off, and repeating the cooling process. It is thus possible to measure vaginal blood flow many times in an hour in the same organ, a highly desirable feature if one wants to follow the action of a drug, hormone, or specific condition.

Final words

The control unit and the Sheffield electrodes have now been in service for approximately 6 years, and many measurements have been made in both the basal and sexually aroused conditions. We have recently finished a study, undertaken at the Sexual Physiology Laboratory at the Porterbrook Clinic, Sheffield, to compare, in a number of subjects, the vaginal flow measured by the Sheffield heat electrode in both basal and sexually aroused states simultaneously with photoplethysmographic recording of vaginal pulse amplitude. On the basis of this study, a large drug firm is manufacturing a number of the Sheffield electrodes for research.

The simultaneous technique (heat and vaginal pulse amplitude recording) allows one to calibrate the nontransportable vaginal pulse amplitude data (normally measured in arbitrary units) for each subject, and convert the beat-by-beat vaginal pulse amplitude signal into quantitative, transportable beat-bybeat data as ml/100 g tissue per min flow.

Because it needs some 3-5 min of temperature cooling recorded data, it is obvious that the heat electrode technique cannot follow rapid changes in vaginal blood flow, so it should not be used during an orgasm. However, if the double recording method as explained previously is used, it may be possible to use the vaginal pulse amplitude record obtained during orgasm to monitor the flow quantitatively. The great advantage of the electrode, however, is that it generates a quantitative measure of vaginal blood flow that is in ml/100 g tissue per min, and these data are completely transferable, unlike the vaginal pulse amplitude from the photoplethysmograph. The vaginal blood flows of different women, at different times of the month (menstrual cycle) or year, with different drugs and doses, can all be directly and quantitatively compared with one another. This method has yet to invade the clinical arena, but this is mainly because hardly any measurements of blood flow are undertaken by any method on women (or for that matter any other genital function) to characterize female sexual dysfunction, as opposed to the many that are in use for male genital dysfunction.17


Grateful acknowledgment is made to Tony Carter, who converted my simple thoughts and sketches by his expertise and skill into the working reality of the Sheffield heat electrode, and to Gorm Wagner and his Sexual Physiology Laboratory at Copenhagen, where the Sheffield electrode first came into contact with a human vagina.


1. Jaszczak P, Baumbach P. A combined tc-PO2 and skin blood flow sensar. Acta Anaesthesinl Scand. 198529: 623-8.

2. Wagner G, Levin RJ. Oxygen tension of the vaginal surface during sexual stimulation in the human. Fertil Steril 1978; 30: 50-3.

3. Levin RJ, Wagner G. Orgasm in women in the laboratory - quantitative studies on the duration, intensity, latency and vaginal blood flow. Arch Sex Behav 1985; 14: 439-49.

3a. Levin RJ, Wagner G. Haemodynamic changes of the human vagina during sexual arousal assessed by a heated oxygen electrode. J Physiol 1978; 275: 23P.

4. Levin RJ, Wagner G. Sexual arousal in women - which haemodynamic measure gives the best assessment? J Physiol 1980; 392: 22-3P.

5. Wagner G, Levin RJ. Effect of atropine and methyl atropine on human vaginal blood flow, sexual arousal and climax. Acta PharmaS  1980; 46: 321-5.

6. Levin RJ, Wagner G, Ottesen B. Simultaneous monitoring of human vaginal haemodynamics by three independent methods during sexual arousal. Sexology 1981; 114-20.

7. Levin RJ, Wagner G. TRH and vaginal blood flow - effects in conscious women and anaesthetized sheep. J Physiol 1986; 378: 83P.

8. Levin RJ, Wagner G. Self-reported central sexual arousal without vaginal arousal- duplicity or veracity revealed by objective measurement. J Sex Res 1987; 23: 540-4.

9. Levin RJ. The mechanisms of human sexual arousal. Annu Rev Sex Res 1992; 3: 1-48.

10. Levin RJ, Wagner G. Re-measurement of vaginal and minor labial oxygen tension for the evaluation of female sexual function. J Urol 2001; 166: 2324-5.

11. Kety SS. Measurement of regional circulation by the local clearance of radioactive sodium. Am Heart J 1949; 38: 321-8.

12. Lassen NA, Lindbjerg IF, Munck O. Measurement of blood flow through skeletal muscle by intra-muscular injection of xenon-133. Lancet 1964; ii: 686.

13. Wagner G, Ottesen B. Vaginal blood flow during sexual stimulation. Obstet Gynaecol 1980; 56: 621-3.

14. Midttun M, Sejrsen P, Colding-J0rgensen M. Heat-washout: a new method for measuring cutaneous blood flow rate in areas with and without arteriovenous anastomoses. Clin Physiol 1996; 16: 259-74.

15. Perl W, Cucinell SA. Local blood flow in human leg muscle measured by a transient response thermoelectric method. Biophys J 1965; 5: 211-30.

16. Levin RJ, Wagner G. Human vaginal blood flow - absolute assessment by a new, quantitative heat wash-out method. J Physiol 1997; 504P: 88P-9P.

17. Levin RJ. Measuring female genital functions - a research essential but still a clinical luxury? Sex Relatsh Ther 2004; 19: 191-200.