Plethysmographic Curve Analysis and Response to Exercise in Normal Subjects, Hypertension, and Cardiac Failure Peter I. Woolfson, BSc, MBChB, MRCP, MD; Brian R. Pullan, BSc, PhD; Philip S. Lewis, BSc, MB, BS, FRCP Occlusion plethysmographic recordings were obtained on 26 subjects prior to and immediately following repeated venous occlusions. A simple method of approximating the curve shape by 2 straight lines is described. The results indicate that, following an initial occlusion, the height of subsequent curves is reduced and the angulation between the 2 lines approximating the curves changes in a way that indicates that the principal mechanism is venous shunting. The degree of shunting was quantified by taking the relative percentage change in shape of the 2 lines approximating the curve, ie, the percentage shunt. Venous shunting is shown to be much more marked after hand exercise than at rest in normal subjects and in those with heart failure or hypertension. (Biomedical Instrumentation & Technology 2002;36:173 179). The time course of venous occlusion plethysmographic curves consists of 2 parts: (1) an initial relatively steep linear upslope and (2) a later part with a smaller slope. Previously it has been assumed that the rate of volume increase decreases secondary to a rise in capillary and venous pressure distal to the cuff, which eventually exceeds the cuff pressure. 1,2 To obtain meaningful results, adequate venous outflow obstruction is essential. Therefore, the pressure in the cuff must occlude all the veins. 3 It is generally assumed that a cuff pressure not far below the diastolic blood pressure is effective (eg, 50 mm Hg can be high enough The Department of Medicine, Stepping Hill Hospital, Hazel Grove, Stockport, UK (PIW, BRP, PSL). Address correspondence and reprint requests to Dr Woolfson, Consultant Cardiologist, Trafford General Hospital, Moorside Road, Davyhulme, Manchester M41 5SL, UK (e-mail: peter@meem.demon.co.uk). to occlude all collapsible veins 4 ). The pressure in the collecting cuff must not alter arterial inflow. It is assumed that the effect of pressures in the part of the limb in the plethysmograph does not significantly affect the rate of blood flow in the first few seconds after cuff inflation. 5,6 Blood flow is calculated from the initial upslope of the curve, 7,8 and so the average rate of volume increase in the 3 4 seconds after cuff inflation is normally used. We have observed a gradual reduction in curve gradient (Y in Figure 2), suggesting that there may be a slow reopening of previously occluded veins; however, the redirection of venous blood to deeper veins unaffected by the venous occlusion cuff may also be occurring. After hand exercise, there is often a sudden change in curve gradient (Figure 3), suggesting a more dynamic change. Furthermore, after repeated venous occlusion, the overall height of the venous occlusion curve is reduced, again suggesting a dynamic change induced by venous occlusion. Arterial inflow still continues, as observation of the curves shows that volume pulsations resulting from pulsatile arterial flow can be seen superimposed on the general shape of the curve. Each part of the curve does not represent the absolute arterial blood flow but the net blood flow. Even with apparently effective venous occlusion, a point is reached when limb venous pressure increases to a level where there is an escape of venous blood proximal to the occluding cuff, this process accounting for the eventual establishment of the new baseline. 9 The shape of the entire curve, therefore, is relevant and may reflect more complex arterial and venous hemodynamics than previously suspected. Consequently, there is value in attempting to characterize these changes in curve shape. A simple measure of shape change would be useful if it correlated with disease state. Biomedical Instrumentation &Technology 267
Plethysmographic Curve Analysis and Response to Exercise Figure 1. The air plethysmograph with a subject s forearm and hand in situ. The manometer and air reservoir used with the hand exerciser (within the plethysmograph) are on the left and the calibration syringes are seen in front of the plethysmograph. Pressurized air containers used to achieve rapid cuff inflation to venous occlusion pressure are seen to the right. The computer on which the plethysmographic curves are displayed and later analyzed is seen behind the plethysmograph. The purpose of this study was to demonstrate a new method of analysis of plethysmographic curve shapes and to look for evidence of differences between normal subjects, hypertensive subjects, and patients with cardiac failure using an air plethysmograph. Methods Normal Subjects Eight male normal subjects were studied (mean age 47.6 years, range 29 66). None had any history of cardiovascular or respiratory disease, smoking, or diabetes and none were taking any medications. Hypertensive Subjects Twelve hypertensive subjects were studied. There were 10 males and 2 females (mean age 48.6 years, range 31 60). All had hypertension as found on repeated clinic measurements and confirmed with 24-hour ambulatory blood pressure monitoring. Investigation had not revealed any evidence of a secondary cause for their hypertension and there was no evidence of any other cardiovascular disease. Two subjects smoked cigarettes. All the hypertensive subjects were studied prior to the commencement of any antihypertensive therapy, and 6 were restudied after treatment for 2 weeks with verapamil SR at 240 mg daily. Cardiac Failure Subjects Six male subjects with cardiac failure were studied, mean age 62.8 years (range 55 75). All were in stable condition with no changes to therapy in the previous 3 months. None of the subjects had peripheral edema and all were 268 July/August 2002
Woolfson et al Figure 2. Plethysmographic venous occlusion curve showing (A) cuff artifact, (X) the initial upslope, (Y) the latter part of the curve with reduced gradient and prominent volume pulsations due to pulsatile arterial flow. in New York Heart Association (NYHA) functional class II. Echocardiography was performed on each subject and confirmed moderately severe or severe left ventricular (LV) dysfunction on 2-dimensional echocardiography, with LV end diastolic diameters (LVEDD) of 54 92 mm. Five subjects had cardiac failure secondary to ischemic heart disease and 1 subject had alcoholic cardiomyopathy. All were taking loop diuretics (1 was also taking a thiazide diuretic), 5 were taking an angiotensin-converting enzyme (ACE) inhibitor, 1 was taking an angiotensin II receptor antagonist, and 3 subjects were taking antianginal therapy (2 taking nitrates, 1 taking a potassium channel opener). All were in sinus rhythm, and 3 subjects were taking digoxin for its positive inotropic effect. The characteristics of each group are shown in Table 1. General Methods The study was approved by the local Ethics Committee and each subject gave written informed consent. The combined volume change of the right forearm and hand was measured using a validated air plethysmograph 10 by standard venous occlusion, with the upper arm occluding cuff pressure set at 75% of the diastolic blood pressure. The experimental apparatus and set-up is shown in Figure 1. Figure 3. Plethysmographic curves after hand exercise: (X) gradient of the initial upslope immediately after hand exercise, (Y) gradient of the curve immediately after hand exercise before the venous occlusion cuff is released, (X') gradient of the initial upslope 1 minute after hand exercise, (Y') gradient of the curve 1 minute after hand exercise before the venous occlusion cuff is released, (P P') change in height of the curve immediately after and 1 minute after hand exercise. Table 1. Baseline date in normal subjects, cardiac failure patients, and hypertensive patients* Cardiac Normal Failure Hypertensive Patients Patients Patients Sex M 8 6 10 F 0 0 2 Age (years) 47.6 ± 5.2 62.8 ± 3.0 48.6 ± 3.1 Body mass index 21.5 ± 1.3 27.2 ± 1.8 27.1 ± 1.2 MAP (mm Hg) 91.7 ± 3.6 91.9 ± 4.1 123.4 ± 2.8 Forearm volume (ml) 1441 ± 59 1567 ± 91 1656 ± 67 *All measurements ± SEM. Biomedical Instrumentation &Technology 269
Plethysmographic Curve Analysis and Response to Exercise and later parts of the curve do not simply reflect the arterial blood flow. Figure 4. Idealized plethysmographic venous occlusion curve showing the variables by which the curve can be described: t 1 = time interval used for blood flow calculation during the initial upslope; V1 = volume change during t 1 ; t 2 = time interval used for blood flow calculation during the second part of the curve; V2 = volume change during t 2 ; H = vertical height of the plethysmographic curve at the intersection of the 2 tangents used to determine t 1 and t 2. Measurements were made with each subject resting on a bed with the thorax and head supported at 45.The right forearm and hand were placed horizontally within the plethysmograph at heart level. The volume of the forearm and hand was measured by water displacement. Initial measurements were taken after 10 minutes of rest. Each subject then exercised by rhythmically squeezing an in situ sphygmomanometer bulb for 1 minute. This bulb was attached to a constant-leak device calibrated to give known workloads (0.59, 0.81, 1.34, and 1.70 Watts). At the end of 1 minute of any workload, limb blood flow was measured immediately and then every minute for up to 10 minutes. The subject then rested for a further 10 minutes prior to 1 minute at the next workload. This cycle was repeated up to 1.70 Watts, when the subject continued at this workload until exhaustion (unless exhaustion occurred at a lower level). The final period of exercise was taken to be the peak exercise. In each group, the initial part of the volume-change curve was used for analysis and values taken to be the indicated blood flow. This was because the shape of the volume-change curve varies at rest and after exercise Plethysmographic Curve Description The shape of venous occlusion plethysmographic curves can be divided into 2 main parts (seen in Figure 3): X, an initial steep upslope occurring immediately after inflation of the venous occlusion cuff, and Y, a subsequent, smaller gradient. The plethysmographic curve can be further described in 2 respects: (1) the degree of angulation between X and Y and (2) the change in height of the point of inflection between X and Y (Figure 3). We noted that, despite the reducing overall curve gradient, arterial pulses were present superimposed on the entire curve, representing arterial inflow. The instantaneous curve gradient represents the net blood flow (the indicated blood flow), ie, arterial inflow minus venous outflow. Therefore, venous outflow from the arm must be increasing during the latter part of the curve. The reduction of gradient Y compared to X indicates increased venous outflow and/or reduced arterial inflow. However, arterial pulsations do not tend to diminish during venous occlusion, suggesting that increased venous outflow (shunt) predominates. The sudden change in gradient seen after exercise suggests that venous blood may be rapidly shunted to deep venous channels, which are either unaffected by the venous occlusion cuff or open as a response to cuff occlusion after hand exercise. A Simple Curve Analysis Figure 4 shows an idealized plethysmographic curve and shows how its shape can be described using the 3 variables Flow 1, Flow 2, and H. If one considers blood flow over the initial steep part of the curve during time interval t 1 (Flow 1 ), then this can be expressed as Flow 1 = V 1, t 1 and if one considers blood flow over the second part of the curve during time interval t 2 (Flow 2 ), then this can be expressed as Flow 2 = V 2, t 2 where V 1 and V 2 are the changes in forearm/hand volume during time intervals t 1 and t 2, respectively. It 270 July/August 2002
Woolfson et al Figure 5. Plethysmographic venous occlusion curves from a single subject (a) at rest, (b) immediately after hand exercise, and (c) 1 minute after hand exercise. follows that the fractional change in blood flow (the fractional shunt) is Flow 1 Flow 2, Flow 1 The two tangents Flow 1 and Flow 2 were extrapolated to their point of intersection at point P (see Figure 4). Height H then represents the point of increased limb volume at which shunting commences. The shape of any plethysmographic curve can therefore be described in terms of the 3 variables Flow 1, Flow 2, and H. The change in height of the curve ( H) from immediately after hand exercise (time = 0 minutes) to 1-minute postexercise (time = 1 minute) was measured in absolute terms and as a percentage change of the total forearm/hand volume. Figure 5 shows 3 curves from 1 subject and demonstrates the typical shapes of the curves (a) at rest, (b) immediately after hand exercise, and (c) 1 minute after hand exercise. Figures 6 shows the sequence of curves at rest and at 1-minute intervals after hand exercise in normal subjects. The curves from 6 subjects are shown. Analysis of variance (ANOVA) was applied to compare results between the normal subjects and those with cardiac failure and hypertension (on and off treatment). A value of P of.05 was taken to be significant. Results Table 2 summarizes the results after minimum hand exercise. Shunt increased as a result of exercise in normal, cardiac failure, and hypertensive subjects before and after treatment (P <.0001) as a result of exercise. There were no significant differences in percentage shunt between the groups at rest or after exercise (P =.98). Curve height H 0 at time = 0 and H 1 at time = 1 minute postexercise in all the groups shows no overall Figure 6. Forearm/hand volume-change curves in 6 normal subjects at rest and after hand exercise. Biomedical Instrumentation &Technology 271
Plethysmographic Curve Analysis and Response to Exercise Table 2. Plethysmographic curve parameters after minimum hand exercise (± SEM) Normal Cardiac Failure Hypertensive Hypertensive Patients Patients Patients Patients (Verapamil) (n = 8) (n = 6) (n = 12) (n = 6) Resting shunt (%)* 48.6 ± 10.3 53.3 ± 14.1 45.1 ± 7.4 49.7 ± 7.8 Postexercise shunt (%) 92.3 ± 4.3 99.7 ± 3.6 87.8 ± 3.2 86.0 ± 5.6 H 0 (ml) 15.0 ± 1.7 10.9 ± 2.0 18.0 ± 3.0 18.3 ± 2.6 H 1 (ml) 10.9 ± 1.8 8.1 ± 2.2 14.9 ± 2.8 12.5 ± 2.9 % H 24.7 ± 12.1 5.0 ± 25.2 20.5 ± 6.4 31.4 ± 9.3 % Forearm/hand volume 0.28 ± 0.13 0.20 ± 0.21 0.19 ± 0.06 0.35 ± 0.10 *Resting shunt = (Flow 1 (rest) Flow 2 (rest) 100%) / Flow 1 (rest) = % of forearm/hand volume flowing through deep venous channels following venous occlusion at rest as a result of venous shunting. Postexercise shunt = (Flow 1 (exercise) Flow 2 (exercise) 100%) / Flow 1 (exercise) = % of forearm/hand volume flowing through deep venous channels following venous occlusion after hand exercise as a result of venous shunting. H 0 = volume of the plethysmographic curve immediately postexercise. H 1 = volume of the plethysmographic curve 1 minute postexercise. % H = H 0 H 1 100% / H 0. % Forearm/hand volume = H 100% / (forearm/hand volume). significant difference (P =.18) and no difference between the groups (P =.28). The same data after peak hand exercise are shown in Table 3. Percentage shunt increased in all groups (P <.0001) as a result of exercise. There were no significant differences in percentage shunt between the groups at rest or after exercise (P =.99). The difference in curve height H 0 at time = 0 and H 1 at time = 1 minute postexercise is significant in all groups (P <.001). There is also a significant difference in the difference between H 0 and H 1 between the groups (P =.05). Discussion Venous occlusion plethysmographic curves consist of an initial steep linear upslope and a later part with a smaller gradient. These phases become more prominent after hand exercise, often with a sharp angulation between them. The curve represents the net (arterial minus venous) blood flow. At rest, the gradual reduction in curve gradient (Figure 2) suggests that there may be a slow reopening of previously occluded veins or redirection of venous outflow to deeper veins unaffected by the occlusion cuff. After hand exercise, there is often a sudden change in curve gradient (Figure 3), suggesting a more dynamic change. Arterial inflow persists despite these venous changes, as observation of the curves shows that volume pulsations resulting from pulsatile arterial flow are superimposed on the general shape of the curve. These pulsations are still readily seen in the latter parts of the curve, indicating that arterial blood flow continues despite the increased venous pressure. After hand exercise, the pulsations become more prominent. Therefore, if arterial flow is continuing but there is a reduced rate of limb volume increase seen in the plethysmographic curve, then venous blood must be escaping proximal to the venous occlusion cuff. After hand exercise, this is most likely to be occurring by the opening of deep veins in the arm unaffected by the venous occlusion cuff. Blood is shunted to these deep veins or to venous channels in bone and flows proximal to the cuff, thus reducing the indicated blood flow, as reflected in the shape of the venous occlusion curve. This explanation of curve shape and the simple analysis of the curve shape using 2 straight lines as we have described allows calculation of the percentage shunt. It is also seen that there is a reduction in the height, H, of the plethysmographic curve between the curve recorded immediately after hand exercise and the curve recorded 1 minute later. This suggests that, after the shunt has been opened once, it will open more quickly on subsequent venous occlusions (ie, the shunting process becomes sensitized to the venous occlusion stimulus and responds more efficiently after its first use). 272 July/August 2002
Woolfson et al Table 3. Plethysmographic curve parameters after maximum hand exercise (± SEM) Normal Cardiac Failure Hypertensive Hypertensive Patients Patients Patients Patients (Verapamil) (n = 8) (n = 6) (n = 12) (n = 6) Resting shunt (%)* 48.6 ± 10.3 53.3 ± 14.1 45.1 ± 7.4 49.7 ± 7.8 Postexercise shunt (%) 82.8 ± 4.2 86.0 ± 5.2 89.2 ± 2.7 88.0 ± 2.6 H 0 (ml) 14.4 ± 1.2 10.8 ± 1.8 24.7 ± 2.2 19.4 ± 1.7 H 1 (ml) 9.1 ± 2.0 7.4 ± 1.1 19.6 ± 2.2 15.3 ± 1.7 % H 37.8 ± 10.8 22.6 ± 14.8 17.4 ± 7.7 20.7 ± 6.8 % Forearm/hand volume 0.37 ± 0.10 0.23 ± 0.10 0.33 ± 0.13 0.24 ± 0.08 *Resting shunt = (Flow 1 (rest) Flow 2 (rest) 100%) / Flow 1 (rest) = % of forearm/hand volume flowing through deep venous channels following venous occlusion at rest as a result of venous shunting. Postexercise shunt = (Flow 1 (exercise) Flow 2 (exercise) 100%) / Flow 1 (exercise) = % of forearm/hand volume flowing through deep venous channels following venous occlusion after hand exercise as a result of venous shunting. H 0 = volume of the plethysmographic curve immediately postexercise. H 1 = volume of the plethysmographic curve 1 minute postexercise. % H = H 0 H 1 100% / H 0. % Forearm/hand volume = H 100% / (forearm/hand volume). This measured reduction in height, therefore, gives a measure of shunt responsiveness. This is a new method of analysis of plethysmographic curves that has not been previously described, which allows hemodynamic information contained in the entire curve to be extracted from phenomena that are most likely to be reflex responses to the inflation of the venous occlusion cuff. The percentage shunt is significantly increased after both minimum and peak hand exercise in all the groups studied. After peak hand exercise, there is an overall significant reduction in curve height in the normal subjects and in those with cardiac failure or hypertension, with a significant difference between the groups. Conclusion The entire plethysmographic curve shape contains meaningful hemodynamic information. Arterial blood flow is not impeded by increased venous pressure, and the data is consistent with the theory that the curve shape reflects the shunting of venous blood from superficial to deep venous channels. A simple method of approximating the curve shape by 2 straight lines has been described and used to calculate the percentage shunt and shunt responsiveness to exercise. Venous shunting is much more marked after exercise than at rest in normal subjects and those with cardiac failure or hypertension. Further work is required in order to assess the use of these parameters as a means of assessing hemodynamic status. References 1. Hiatt WR, Huang SY, Regensteiner JG, et al. Venous occlusion plethysmography reduces arterial diameter and flow velocity. Appl Physiol. 1989;6:2239 2244. 2. Sumner DS. Volume plethysmography in vascular disease: an overview. In: Bernstein EF, ed. Vascular Diagnosis. 4th ed. St Louis, Mo: Mosby; 1993:186 187. 3. Greenfield ADM. Methods for the investigation of peripheral blood flow. Br Med Bull. 1963;19:101 105. 4. Whitney RJ. Circulatory changes in the forearm and hand of man with repeated exposures to heat. J Physiol. 1954;125:1 24. 5. Brodie TG, Russell AE. On the determination of the rate of blood flow through an organ. J Physiol. 1905;32:47P 49P. 6. Wilkins RW, Bradley SE. Changes in arterial and venous blood pressure and flow distal to a cuff inflated on the human arm. Am J Physiol. 1946;147:260 269. 7. Landowne M, Katz LN. A critique of the plethysmographic method of measuring blood flow in the extremities of man. Am Heart J. 1942;23:644 675. 8. Greenfield ADM. Venous occlusion plethysmography. Meth Med Res. 1960;8:293 301. 9. Greenfield ADM, Patterson GC. The effect of small degrees of venous distension in the apparent rate of blood inflow to the forearm. J Physiol. 1954;25:525 533. 10. Woolfson PI, Woolfson LAM, Pullan BR, Lewis PS. A new air plethysmograph. Poster presented at: Medical Research Society, November 1996. Biomedical Instrumentation &Technology 273