Postural vasoconstriction in women during the normal menstrual cycle

Transcription

1 Clinical Science (1 99) Postural vasoconstriction in women during the normal menstrual cycle AHMAD A. K. HASSAN, G. CARTER*'AND J. E. TOOKE Departments of Physiology and *Chemical Pathology, Charing Cross and Westminster Medical School, London (Received February/3 June 1989; accepted 4 July 1989) SUMMARY 1. Postural vasoconstriction in the foot was examined in women during the menstrual, follicular and luteal phases of the menstrual cycle, and in 13 age-matched men on two separate occasions, in a constant-temperature environment (22 C). 2. Skin blood flow was measured using laser Doppler flowmetry with the subject lying down, first with the foot maintained at heart level, then with the foot lowered passively 5 cm below the heart. In six of the women, at the time of experiment, serum oestradiol and progesterone were determined by radioimmunoassay. In four women and three men, foot swelling rate was also measured in the dependent foot using a strain gauge plethysmograph in addition to the postural changes in flow. At each visit, in all subjects, arterial blood pressure, heart rate, body temperature, foot skin temperature and body weight were also recorded. 3. The men showed no significant changes in all the variables assessed. In contrast, in women during the luteal phase diastolic and mean arterial pressures were significantly reduced, whereas heart rate, body temperature, foot skin temperature and body weight were significantly increased, as compared with the follicular and menstrual phases of the cycle. 4. During the follicular phase, when oestradiol concentration was high, there were significant reductions in dependent flow and foot swelling rate associated with a significantly augmented postural fall in flow, whereas during the luteal phase, when both oestradiol and progesterone levels were high, there were significant increases in dependent flow and foot swelling rate associated with a significantly impaired postural fall in flow. Four women who reported premenstrual ankle oedema showed significantly higher flow values during the luteal phase than the rest of women. Correspondence: Dr J. E. Tooke, Postgraduate Medical School, Royal Devon and Exeter Hospital, Barrack Road, Exeter EX2 5DW, U.K. 5. These results confirm the modulating influence of female sex hormones on peripheral blood flow and vascular tone. The partially impaired postural vasoconstrictor response during the luteal phase of the cycle might be partly implicated in the pathogenesis of premenstrual oedema in some women. Key words: foot, laser Doppler flowmetry, menstrual cycle, posture, premenstrual oedema, skin blood flow, vasoconstriction. INTRODUCTION The different phases of the normal menstrual cycle have been shown to be associated with variations in peripheral blood flow [l, 21 and peripheral venous distensibility [2,3] as well as fluctuation in limb volume [4]. Moreover, female sex hormones (oestrogens and progesterone) are reported to have profound effects on peripheral blood flow in man [5], and a strong modulating influence on vascular tone in both humans [6,7] and animals [8-lo]. However, few data are available concerning the reaction of the peripheral vasculature in women to changes in limb posture at different stages of the reproductive cycle. In normal subjects assumption of the upright posture or lowering the extremity below heart level is associated with a marked reduction in foot blood flow mediated by precapillary vasoconstriction in response to the increased hydrostatic pressure in the dependent position [ 111. The postural increase in precapillary resistance, that is thought to be largely mediated by a local sympathetic axon reflex [12] or a myogenic response [13], limits the expected rise in capillary hydrostatic pressure [ 141, thereby reducing capillary filtration. Moreover, as filtration continues, the markedly reduced blood flow in the dependent extremity perfits a build-up of plasma Oncotic pressure [l51. The resulting limited rise in capillary hydrostatic pressure and

2 4 A. A. K. Hassan et al. the increase in plasma oncotic pressure associated with the postural vasoconstrictor response thus act as important oedema prevention mechanisms in the dependent position [16]. Cyclical oedema [17] is one of the characteristic features of the premenstrual syndrome [ 181. Although it has been suggested that premenstrual oedema is related to fluctuations in the female sex hormones [19], or an abnormality in the vasopressin and/or the reninangiotensin-aldosterone systems [ 181, the exact mechanism behind the excessive tissue fluid accumulation remains unclear. Nevertheless, the observation that oedema was precipitated by upright posture [2] and improved by the use of vasoconstrictor drugs [21] suggests that some form of a postural defect might exist in these women resulting in an undue pooling of blood in the dependent extremities [2 11. We have previously shown that the use of combined oral contraceptive steroids is associated with a marked attenuation of the postural vasoconstrictor response in the foot, particularly in women with premenstrual oedema [22]. The aim of the present study was to examine such responses in women at different stages of their normal menstrual cycles and relate the postural changes in flow to the swelling rate of the dependent foot and the blood levels of female sex hormones. METHODS Subjects Fifteen women aged 2-39 years were studied during the menstrual (days 1-2), follicular (days 12-) and luteal (days 21-26) phases of the menstrual cycle. Three of the women were studied on similar days of two consecutive cycles and six were studied on similar days of two non-consecutive cycles. All cycles (n = ), except two, were ovulatory cycles as judged by a characteristic biphasic basal body temperature [23] and/or changes in serum progesterone concentration. The cycles were regular ranging in length between 28 and 34 days. None of the women was taking oral contraceptive pills or any medication throughout the cycle under study. Observations were made from one menstrual period to the next, thus excluding the possibility of amenorrhoea or pregnancy. From the menstrual history taken at the end of the study, four of the women reported bilateral ankle oedema recurring monthly 4-7 days before the onset of menstruation. Physical examination was carried out in all women on each visit, and pitting oedema of the ankles was confirmed during the luteal phase in those subjects who complained of premenstrual oedema, but there was no clinical evidence of any cardiac, renal or hepatic disease. Thirteen men aged years were also studied during the same period on two separate occasions at least 1 week apart. All subjects gave their informed consent to the study which was approved by the local Ethical Committee. Experiments were performed after 2-3 min of adaptation in a temperature-controlled room (22"C), at least 2 h after the last food, drink or smoking. Each subject wore similar clothing on repeat visits. Measurement of the postural changes in foot skin blood flow The postural vasoconstrictor response in the foot was assessed in all subjects by measuring the postural changes in foot skin blood flow using a laser Doppler flowmeter (Periflux MKVII; Perimed, Stockholm, Sweden) as described in detail elsewhere []. In brief, this noninvasive technique utilizes the Doppler shift of laser light backscattered from the moving erythrocytes in the cutaneous microcirculation to produce an output voltage flow signal linearly related to the product of erythrocyte concentration and velocity [25]. Repeated measurement of flow from the same skin area using this instrument produced coefficients of variation of %. On each visit the flow (expressed in arbitrary units, as mv) was measured with the subject lying down, first with the foot maintained at heart level for 2 min (H), then with the foot lowered passively 5 cm below the heart for 4 min (D) with the rest of the body remaining in the horizontal position. The postural fall in flow (YO) was calculated as (H- D) x 1/H. Recordings were made on two different skin areas: (i) the dorsum of the foot (with no arteriovenous anastomoses), and (ii) the pulp of the big toe (with numerous anastomoses) [26]. During each experiment arterial blood pressure (using auscultation), heart rate (radial pulse) and body temperature (using a sublingual mercury clinical thermometer) were measured after at least 1 min of supine rest. Foot skin temperature was continuously recorded using a sensitive thermocouple (Comark Electronics Ltd, Littlehampton, Sussex, U.K.) during blood flow measurement. Body weight was also measured at the end of the experiment using an ordinary scale. Determination of serum oestradiol and progesterone concentrations In six women (seven cycles) venous blood samples were withdrawn from an arm vein, just before the start of the experiment for the analysis of female sex hormones on the observational days mentioned above. Serum oestradiol concentration was measured by a direct radioimmunoassay method (Steranti Research Ltd, St Albans, Herts, U.K.). Serum progesterone concentration was measured by a direct in-house radioimmunoassay using an antiserum raised against progesterone-1 1 a-hemisuccinate (Guildhay Antisera, Guildford, Surrey, U.K.). Measurement of foot swelling rate Towards the end of the study it appeared that there might be an association between the magnitude of the postural vasoconstrictor response and the degree of foot swelling on dependency. Therefore, postural vasoconstriction and foot swelling rate were simultaneously assessed in the last four women and three men studied. These were not the four women who complained of premenstrual ankle oedema, although one of these final four women did have clinically manifest oedema. Changes in

3 Menstrual cycle and postural vasoconstriction 41 foot circumference were measured on dependency of the foot ion the experimental days described above) using a strain gauge plethysmograph (Janssen Scientific Instruments, Beerse, Belgium) [27] at the time of the measurement of the postural changes in flow. The gauge encircled the foot midway between the ankle and toes. Foot swelling rate (expressed as ml min- 1 ml- foot) was derived from the slope of the steady-state foot swelling curve obtained during min 1-2 with the foot in the dependent position, thus avoiding the phase of changes in foot volume due to changes in vascular volume which often lasts for the initial 5 min [28]. The number of women and cycles involved in the various study protocols are presented in Table 1. Statistical analysis The results are expressed as means k SEM, except the flow values measured in the toe pulp which are expressed as medians and ranges. Differences between the mean values of each variable (except percentages and toe pulp flows) obtained at different stages of the cycle were assessed by one-way analysis of variance. Differences between percentages and the toe pulp flows were assessed using Friedman s test (a non-parametric analysis of variance). Paired data were compared using Student s t- test or Wilcoxon s signed-rank test, Unpaired data were compared used Student s t-test for independent samples or Wilcoxon s rank sum test. Throughout the text the use of menstrual, follicular or luteal phase refers to days 1-2, 12- and of the cycle, respectively. RESULTS Comparison between men and women In the men there was no significant difference in any of the variables assessed on the two visits (Table 2). In the women, analysis of variance showed no significant differences in systolic blood pressure, heart rate or foot skin blood flow measured in the horizontal position on the dorsum of the foot at different stages of the cycle, whereas the analysis of variance revealed significant differences in diastolic and mean arterial pressures (P<.1, P<.9, body temperature (P<.1), body weight (P<.5), foot skin temperature (dorsum: P<.1; big toe: P=.9), blood flow measured in the horizontal position on the toe pulp (P=.99), blood flow measured in the dependent position (dorsum: WO.1; big toe: P=.9) and the postural fall in flow (dorsum: P<.8; big toe: P<:.9). During the luteal phase, diastolic and mean arterial pressures were significantly reduced, whereas heart rate, body temperature, foot skin temperature and the body weight were significantly increased as compared with both the follicular and menstrual phases of the cycle (Table 2). Female sex hormones In five ovulatory cycles oestradiol concentrations were 13.8 f 28.3,684.3 k 35.9 and f 8.9 pmol/l, and Table 1. Number of women and cycles involved in the various protocols of the study Parameter to be assessed No. of women No. of cycles Postural vasoconstriction Foot skin temp. Arterial blood pressure Heart rate Body temp. Body wt. Female sex hormones Foot swelling rate progesterone concentrations were < 5, < 5 and 35.5 k 8.6 nmol/l during the menstrual, follicular and luteal phases of the cycle, respectively (Fig. la). In two anovulatory cycles, during the late phase, there was no change in basal body temperature or serum progesterone concentration which remained at < 5 nmol/l throughout each of the two cycles, whereas serum oestradiol concentrations were 231,264,211 and 377,567 and 611 pmol/l during the three phases of the two cycles, respectively. Postural changes in flow during the cycle In Fig. 1, the postural changes in foot skin blood flow in the five ovulatory cycles are presented as related to serum oestradiol and progesterone. During the follicular phase, when oestradiol concentrations were high, there was a sigificant reduction in skin blood flow measured in the dependent foot associated with a significantly greater postural fall in flow as compared with the menstrual phase. During the luteal phase, however, when both oestradiol and progesterone levels were high, there was a significant increase in dependent flow associated with a significantly attenuated postural fall in flow as compared with both the follicular and menstrual phases (Fig. 1 and Table 2), i.e. the postural vasoconstrictor response was preserved and even partially augmented during the follicular phase but markedly impaired during the luteal phase. The pattern of the response was quite similar in the two skin areas tested, but the flow values were much greater in the toe pulp than in the dorsum of the foot. In addition, the women showed significantly lower dependent flov measured on the toe pulp (P<O.Ol) during the follicular phase, and significantly higher dependent flow measured on the dorsum of the foot (P<O.OOl) during the luteal phase as compared with the corresponding values measured in the men at either visit (Table 2). Also the percentage fall in flow on dependency was significantly lower in women during both the menstrual (dorsum: P<.1; big toe: P<.5) and luteal (dorsum: P<O.OOl; big toe: P<O.OOl) phases, as compared with the corresponding values obtained in the men at either visit (Table 2). There was also a positive correlation between serum oestradiol concentrations and the blood flow values measured at heart level on the dorsum of the foot (r=.51, P<.5). Moreover, the three women studied at different stages of two consecutive ovulatory cycles showed highly reproducible experimental results (Table 3). Unlike ovulatory cycles, during the late phase of

4 .p.. N Table 2. Postural changes in skin blood flow measured on the dorsum of the foot and the pulp of the big toe in women at different phases of 22 ovulatory cycles, and in 13 men at two separate visits The changes in arterial blood pressure, heart rate, body temperature, foot skin temperature and body weight are also shown. H, With the foot horizontal; D, during min 4 with the foot dependent. The postural fall in flow was calculated as (H-D) x loa/h. Values are given as means ± SEM, except the blood flow values measured in the pulp of the big toe which are expressed as median and range. Statistical significance: *P<.5, **P<O.2, ***P<O.Ol, ****P<O.OOI compared with menstrual phase; t P<.1, ttp<.1 compared with follicular phase. Women with normal cycles Male subjects Arterial blood pressure (mmhg) Menstrual phase Follicular phase Luteal phase 1st visit 2nd visit?> > Systolic 19.6± ± ± ± ± 1.2 Diastolic 7.2 ± ± ±.9***tt 75.2 ± ± 1.1 ::z:: p) Mean 83.3± ±.8 84 ±.8**t 89.3± ± 1.1 Heart rate (beats/min) 67A± ± ± 1.3t 67A± ± 1.1 Body temp. (OC) 36.62± ± ±.3****tt (ll 36.79± ±.5... Body wt, (kg) 55.29± ± ± 3.42"****tt 7.35 ± ±~.18 ~ Dorsum of the foot Skin temp. (OC) 27.7±OA 27A± ±.3***tt 29.9± ±.3 Skin blood flow (H) (my) 11.8 ± ± ± ± ± 1.5 Skin blood flow (D) (mv) 3.5±O ±.2*** 64 ±.6****tt 3.1±OA 3.1 ±.3 Postural fall in flow (%) 66.1 ± ± 2.9*** 55.7±3.9*t 78.3± ± 1.7 Pulp of the big toe Skin temp. (OC).2±.8.1 ± ±.5***t 25.8± ±.8 Skin blood flow (H) (mv) 25.8{ ) 37.5(6.1-8.) 23.6( ) 45.1( ) 37.5( ) Skin blood flow (D) (mv) 5.4(3.9-9.) 2.5(.5-8.8)*** 11.l{ )**t 7.8( ) 8.5( ) Postural fall in flow (%) 75.8± ± 1.6*** 59.1 ±3.9*t 85.± ±2.8 p) ::3

5 Table 3. Arterial blood pressure, heart rate, body temperature, foot skin temperature, body weight and the postural changes in skin blood flow measured in three women during the menstrual, follicular and luteal phases of two consecutive ovulatory cycles Values are given as means ± SEM, except the blood flow values measured in the pulp of the big toe which are expressed as median and range. H, With the foot horizontal; D, during min 4 with the foot dependent. The postural fall in flow was calculated as (H-D) x 1/H. Abbreviation: CV,mean coefficient of variation. ~ ~ ::s Menstrual phase Follicular phase Luteal phase a 2 ~ Cyclel Cycle 2 CV(%) Cycle 1 Cycle 2 CV(%) Cycle 1 Cycle 2 CV(%) o Arterial blood pressure (mmhg) Systolic 18.3 ± ±.7 I.S 1S.3±2,4 los.7± ± ±2. 3. ::s.. Diastolic 67.7 ± ± ± ± S.3± ± "' Mean 81.2± ±O ± ± ± ± [I) Heart rate (beats/min) 71.3±2.9 7.± ± ± ± ± 2,4 3.2 E" "1 Body temp. (C) 36.63±.O ±O ±.O ± ±.3 37.±O.6.1 e:.. Body wt. (kg) 5.2±4.9 SO.± ±4.8 SO.2±S ± ±5..7 -e ~ [I) Dorsum of thefoot (') Skin temp. (OC) 27.2±.7 27.S± S.9± ± ±.9 28.S± Skin blood flow (H) (mv) 1.9± ± ± ± ±4.4 IS.1 ±S.O 13.8 ~ = Skin blood flow (D)(mV) 3.2 ± ± ± ± ± ± :::1. (') Postural fall in flow (%) 64.4 ± ± ± ± ± ±.S o' ::s Pulp of the big toe Skin temp. (C).1 ± 1.9.5± ±O ± ± ± Skin blood flow (H) (mv) 26.2( ) 2S.4( ) ( S) 41.S( ) ( S.6) 44.1( S).6 Skin blood flow (D) (mv) 6.3( ) 5.6(5.1-9.) ( ) 3.9( ) 13.2 IS.2(11.1-6S.3) 14.5( ) 1.3 Postural fall in flow (%) 74.± ± ± ± ±1.9 SO.S± ~ (') ~ t.:l.p... w

6 44 A. A. K. Hassan et al. the two anovulatory cycles, the postural vasoconstrictor response was maintained (Table 4). The four women who reported premenstrual ankle oedema showed significantly higher flow values particularly in the dependent position, as well as a significantly higher increment in body weight during the luteal phase, than the rest of the women (Table 5). Although the postural fall in flow appeared to be more attenuated in women Menstruation Follicular Luteal - I P <.2, - r P <.19r-P <.21 T with oedema than in women without oedema, the difference was not statistically significant (Table 5). Foot swelling rate as related to postural vasoconstriction In three men the simultaneously measured postural fall in flow and foot swelling rate were very consistent at the two visits (78.7 k 1.% vs 79.6 f 1.O/o,.62 f.7 vs.59f.4 ml min-l 1 ml-' foot; Fig. 2). In contrast, four women with ovulatory cycles showed an inverse relationship between the postural fall in flow and foot swelling rate. Thus, as compared with the values obtained during the menstrual phase (71.4 k 2.4%,.57i. mi min-' 1 m1-l foot), there was a significantly greater postural fall in flow during the follicular phase associated with a reduction in foot swelling rate (85.2 f 1.9%,.3i.4 ml min-l 1 ml-i foot), whereas during the luteal phase, there was a sigificantly impaired postural fall in flow associated with a significantly greater foot swelling rate (49. k 5.9%,.85 k.12 ml min-' 1 ml-' foot) (Fig. 2). Furthermore, a significant positive correlation was found between the absolute values of dependent flow and foot swelling rate measured in women at different stages of the cycle (Fig. 3). Menstruation Follicular Luteal Fig. 1. Postural changes in skin blood flow (means f SEM) measured on the dorsum of the foot at heart level () and at 5 cm below the heart (B)(b) in five women at different stages of their ovulatory cycles, as related to the blood levels (individual values) of oestradiol () and progesterone (*) (a) measured at the same visit before the start of the experiment. DISCUSSION This study has demonstrated significant physiological variations in the postural vasoconstrictor response and the swelling rate of the dependent fool during the normal menstrual cycle. These variations appear to be related, in some way, to the physiological fluctuations in the blood levels of the female sex hormones. Most important is the observation that postural vasoconstriction is markedly Table 4. Postural changes in foot skin blood flow measured during the menstrual, follicular and late phases of two women with anovulatory cycles Individual values are presented. H, With the foot horizontal; D, during min 4 with the foot dependent. The postural fall in flow was calculated as (H-D) X 1/H. The changes in body temperature, foot skin temperature and body weight are also shown. Body temp. ("C) Body wt. (kg) Subject 2 Subject 2 Menstrual Follicular Late Menstrual Follicular Late phase phase phase phase phase phase Dorsum of the foot Skin temp. ("C) Skin blood flow (H) (mv) Skin blood flow (D)(mV) Postural fall in flow (%) Pulp of the big toe Skin temp. ("C) Skin blood flow (H)(mV) Skin blood flow (D)(mV) Postural fall in flow (%)

7 Menstrual cycle and postural vasoconstriction 45 Table 5. Postural changes in skin blood flow measured during the luteal phase of the menstrual cycle in four women who reported premenstrual oedema as compared with the corresponding changes in 11 women without oedema The increase in body weight from the late follicular to mid-luteal phase is also shown. H, With the foot horizontal; D, during min 4 with the foot dependent. The postural fall in flow was calculated as (H-D) X 1/H. Values are given as means If: SEM, except the blood flow values measured in the pulp of the big toe which are expressed as median and range. Abbreviation: NS, not significant. Women without Women with Statistical oedema (n = 1 1 ) oedema (n = 4) significance Dorsum of the foot Skin blood flow (H)(mV) 12.3 f f 1.4 NS Skin blood flow (D)(mV) 3.9 f f.7 P<.1 Postural fall in flow (%) 64. f f 3. NS Pulp of the big toe Skin blood flow (H)(mV) 18.5( ) 163.4( ) P<.5 Skin blood flow (D)(mV) 74( ) 65.4( ) P<.5 Postural fall in flow (%) 66. f f 8. NS Increase in body wt. (kg).41 f fo.1 P<. 1 impaired during the luteal phase of the cycle, particularly in women with premenstrual oedema. The partially augmented postural vasoconstrictor response during the follicular (oestrogenic) phase of the cycle might be ascribed to an oestrogen-induced potentiation of vascular smooth muscle contraction [8, lo]. On the other hand, the marked impairment of the response during the luteal phase of the cycle in skin areas with and without arteriovenous anastomoses is unlikely to be wholly attributable to an overriding reflex thermoregulatory vasodilatdtion due to increased body temperature, since we have previously shown that during indirect heating in man the postural response is impaired in the toe pulp (with anastomoses), but preserved in the dorsum ** ** T (4 Menstruation Follicular Luted 1st visit 2nd visit phase phase phase Fig. 2. Postural fall in skin blood flow measured on the dorsum of the foot (a) in four women at different stages of the menstrual cycle and (b) in three men at two separate visits, as related to foot swelling rate measured between min 1 and min 2 with the foot in the dependent position. Statistical significance: 'P<.5, "P<.2 compared with the menstrual phase. of the foot (without anastomoses) []. Although the exact mechanism of the impaired postural vasoconstriction during the luteal phase of the cycle remains unclear, it has been shown that female sex hormones not only strongly influence vascular smooth muscle reactivity to several vasoactive agonists [7,9], but also modify the mechanisms involved in the synthesis, release or disposition of noradrenaline as well as the process of excitationcontraction coupling of the vascular smooth muscle [6]. This explanation is also supported by our previous observation that the use of combined oral contraceptive steroids is associated with markedly impaired postural vasoconstriction [22]. The significantly reduced diastolic and mean arterial pressures in women during the luteal phase of the cycle observed under carefully controlled experimental conditions in the present study might reflect a reduced resting peripheral vascular tone, and thus a reduced resting vascular resistance. This could provide further indirect evidence as to the mechanism by which the postural increase in precapillary resistance was markedly attenuated during the luteal phase. The absence of any impairment of the postural response during the luteal phase of anovulatory cycles, when there was no increase in serum progesterone concentration, might indicate that progesterone is critically involved in the mechanisms leading to inhibition of this response. Indeed, in a recent study, progesterone concentrations were found to be significantly higher during the post-ovulatory phase in patients with premenstrual syndrome than in normal women [29]. The association between an attenuated postural vasoconstrictor response (with greater dependent flow) and increased fluid filtration (foot swelling rate) and body weight in women during the luteal phase of the cycle adds further support to the view that the postural increase in precapillary resistance might be involved in the protection against the development of dependent oedema [16]. It could also strengthen the supposition that female sex

8 I.1- -z I e.d 4.6- Q - v M 5 c.2- g o A o A A A 5 1 Skin blood flow measured 5 cm below the heart (mv) Fig. 3. Relationship between skin blood flow recorded on the dorsum of the foot and foot swelling rate measured during min 1-2 after the foot was lowered 5 cm below the heart in four women with normal ovulatory menstrual cycles studied during the menstrual (A), follicular () and luteal () phases of their cycles. r=.828; P<O.OOl (n= 12). hormones are involved directly or indirectly in the pathogenesis of premenstrual cyclical oedema [ 17-19]. Although premenstrual swelling has been previously suggested to be related to changes in capillary permeability [3] or alterations in the vasopressin and/or the renin-angiotensin-aldosterone systems [ 181, the actual mechanism of this condition is not yet completely understood. However, the significantly higher dependent flow shown in our previous [22] and present study in women with premenstrual oedema during the latter phase of the cycle suggests that sex hormone-induced changes in peripheral vascular reflexes could be an important factor contributing to the complex pathogenesis of this condition. This hypothesis is supported by the observation of increased capillary filtration coefficient during the premenstrual period in women with premenstrual syndrome [31], excessive increase in leg volume on standing in patients with idiopathic oedema [2] and the marked improvement of periodic orthostatic oedema when the patients were treated with vasoconstrictor drugs [21]. It might be speculated that the attenuated postural vasoconstriction observed during the luteal phase of ovulatory cycles could result in excessive pooling of blood in the dependent extremities. As a consequence of the greater reduction in the central blood volume activation of the vasopressin and/or renin-angiotensin-aldosterone system, with a subsequent decrease in urinary volume and urinary sodium excretion [32] might ensue, thus leading to salt and water retention and an aggravation of the oedema state [2]. In conclusion, this study confirms that female sex hormones have a modulating influence on peripheral blood flow and vascular tone. The partially impaired postural vasoconstriction during the luteal phase of the A. A. K. Hassan et al. menstrual cycle might partly explain the incidence of premenstrual oedema in some women. ACKNOWLEDGMENTS We thank professor L. H. Smaje for his continuous support, and Mrs Alison Allen for typing the manuscript. A.A.K.H. was supported by an Egyptian Government Scholarship, and J.E.T. by the Wellcome Trust. REFERENCES 1. Edwards, E.A. & Duntley, S.Q. Cutaneous vascular changes in women in reference to the menstrual cycle and ovariectomy. Am. J. Obstet. Gynecol. 1949; 57, Keates, J.S. & Fitzgerald, D.E. Limb volume and blood flow changes during the menstrual cycle. 11. Changes in blood flow and venous distensibility during the menstrual cycle. Angiology 1969; 2, McCausland, A.M., Holmes, F. & Trotter, A.D., Jr. Venous distensibility during the menstrual cycle. Am. J. Obstet. Gynecol. 86, Keates, J.S. & Fitzgerald, D.E. Limb volume and blood flow changes during the menstrual cycle. I. Limb volume changes during the menstrual cycle. Angiology 1969; 2, Lim, Y.L. & Walters, W.A.W. Oestrogens and the maternal circulation. Aust. N.Z. J. Obstet. Gynaecol. 197; 1, Altura, B.M. & Altura, B.T. Influence of sex hormones, oral contraceptives and pregnancy on vascular muscle and its reactivity. In: Carrier,. & Shibata, S. eds. Factors influencing vascular reactivity. Tokyo, New York Igaku- Shoin, 1977: Eccles, N.K. & Leathar, H.L. 17/?-Oestradiol and progesterone on human vascular reactivity. In: Rose, C. ed. Migraine, Proceedings of the Fifth International Migraine Symposium. Basel: Karger, 1985: Lloyd, S. The vascular responses of the rat during the reproductive cycle. J. Physiol. (London) 1959; 148, Hettiaratchi, E.S.G. & Pickford, M. The effect of oestrogen and progesterone on the pressor action of angiotensin in the rat. J. Physiol. (London) 1968; 196, Altura, B.M. Sex and oestrogens and responsiveness of terminal arterioles to neurohypophyseal hormones and catecholamines. J. Pharmacol. Exp. Ther. 1975; 193, Beaconsfield, P. & Ginsburg, J. Effect of changes in limb posture on peripheral blood flow. Circ. pes. 1955; 3, Henriksen,. & Sejrsen, P. Local reflex in microcirculation in human cutaneous tissue. Acta Physiol. Scand. 1976; 98, Mellander, S., Oberg, B. & Odelram, H. Vascular adjustments to increased transmural pressure in cat and man with special reference to shifts in capillary fluid transfer. Acta Physiol. Scand. 1964; 61, Levick, J.R. & Michel, C.C. The effects of position and skin temperature on the capillary pressures in the fingers and toes. J. Physiol. (London) 1978; 274, Noddeland, H., Aukland, K. & Nicolaysen, G. Plasma colloid osmotic pressure in venous blood from the human foot in orthostasis. Acta Physiol. Scand. 1981; 113, Sejrsen, P., Henriksen,. & Paaske, W.P. Effect of orthostatic blood pressure changes upon capillary filtra- tion-absorption rate in the human cafi. Acta Physidl. Scand. 1981: 111, Thom, G:W. Cyclical oedema. Am. J. Med. 1957; 23, Reid, R.L. & Yen, S.S.C. Premenstrual syndrome. Am. J. Obstet. Gynecol. 1981; 139,85-14.

9 Menstrual cycle and postural vasoconstriction Thorn, G.W., Nelson, K.R. & Thorn, D.W. A study of the mechanism of oedema associated with menstruation. Endocrinology 1938; 22, Kuchel, O., Horky, K., Gregorova, I., Marek, J., Kopecka, J. & Kobilkova, J. Inappropriate response to upright posture: a precipitating factor in the pathogenesis of idiopathic oedema. Ann. Int. Med. 197; 73, Streeten, D.H.P. & Conn, J.W. Studies on the pathogenesis of idiopathic oedema. J. Lab. Clin. Med. 1959; 54, Hassan, A.A.K. & Tooke, J.E. Effect of combined oral contraceptive therapy on the postural regulation of skin blood flow in the foot. Int. J. Microcirc.: Clin. Exp. 1986; 5, Martin, P.L. Detection of ovulation by the basal temperature curve with correlating endometrical studies. Am. J. Obstet. Gynecol. 1943; 46, Hassan, A.A.K., Rayman, G. & Tooke, J.E. Effect of indirect heating on the postural control of skin blood flow in the human foot. Clin. Sci. 1986; 7, Nilsson, G.E., Tenland, T. & Oberg, P.A. Evaluation of a laser Doppler flowmeter for measurement of tissue blood flow. IEEE Trans. Biomed. Eng. 198; 27, Grant, R.T. & Bland, E.F. Observations on arteriovenous anastomoses in human skin and in the bird's foot with special reference to the reaction to cold. Heart 1931;, Whitney, R.J. The measurement of volume changes in human limbs. J. Physiol. (London) 1953; 121, Michel, C.C. & Moyses, C. The measurement of fluid filtration in human limbs. In: Tooke, J.E. & Smaje, L.H. eds. Clinical investigation of the microcirculation. Boston: Martinus Nijhoff Publishers, 1987: Brien, P.M.S., Selby, B. & Symonds, E.M. Progesterone, fluid and electrolytes in premenstrual syndrome. Br. Med. J. 198; i, Jones, E.M., Fox, R.H., Verow, P.W. & Asscher, A.W. Variations in capillary permeability to plasma proteins during the menstrual cycle. Br. J. Obstet. Gynaecol. 1966; 73, Wong, W.H., Freedman, R.I., Levan, N.E., Hyman, C. & Guilligan, E.J. Changes in capillary filtration coefficient of cutaneous vessels in women with premenstrual tension. Am. J. Obstet. Gynecol. 1972; 114, Edwards, O.M. & Bayliss, R.I.S. Urinary excretion of water and electrolytes in normal females during the follicular and luteal phases of the menstrual cycle: the effect of posture. Clin. Sci. 1973; 45,