Seasonal differences in finger skin temperature and microvascular blood flow in healthy men and women are exaggerated in women with primary Raynaud's phenomenon

Abstract

Aims

Patients with primary Raynaud's phenomenon (PRP) have more severe symptoms in the winter. The aetiology of this is more complex than simply increased vasoconstriction in response to the immediate ambient temperature. The aim of this study was to investigate differences in skin temperature (Tsk), microvascular blood flow and responses to endothelium-dependent and independent vasodilators in healthy controls, and women with PRP under identical environmental temperatures but in different seasons.

Methods

Ten women with PRP were compared with age matched women (10) and men (10). Finger skin responses were recorded immediately on arrival, after stabilizing in a temperature regulated laboratory at 22–24 °C, and at matched warm (35 °C) and cold (15 °C) Tsk in the winter and summer. Baseline red blood cell flux (r.b.c. flux), and the change in flux in response to iontophoresis of acetylcholine (ACh) and sodium nitroprusside (SNP) were recorded by laser Doppler fluxmetry at the warm and cold Tsk.

Results

Arrival Tsk were significantly cooler for all subjects during the winter (mean seasonal difference −2.6 °C, P < 0.0001), and markedly colder in subjects with PRP (mean seasonal difference −3.5 °C, P < 0.0005). Statistically significant seasonal differences persisted in all subjects at stable Tsk despite an identical laboratory temperature (mean difference 1.3 °C, P < 0.0001). To achieve comparable controlled finger Tsk a significantly colder local environment was required for male controls (mean of −2.1 °C, P < 0.0001), and a significantly warmer environment for subjects with PRP (mean of +2.4 °C, P < 0.0001) compared with female controls. This needed to be warmer in the winter, by a mean of 2.4 °C, than the summer for all subjects. Vasodilatation in response to ACh, but not SNP, was significantly smaller (P < 0.0001) in the PRP group compared with the female controls for all visits, with most of this difference arising in the winter visits (P < 0.01).

Conclusions

There is a seasonal and persistent influence on finger Tsk, and microvascular blood flow in healthy men and women, which modifies the observed responses to immediate changes in finger Tsk. The seasonal differences are greater in women than men, and are further exaggerated in women with PRP, in whom this is associated with reduced endothelium-dependent vasodilatation.

Introduction

Primary Raynaud's phenomenon (PRP) is a common complaint, predominantly of young women [1], characterized by vasospasm of the microvasculature of the fingers induced by the cold [2]. People with PRP also have persistently colder hands even in warm situations, such as when in bed [3], and have more vasospasm in the winter [4].

The pathophysiology of PRP continues to be elusive, despite ongoing investigation of both local endothelial abnormalities [5, 6], and investigation of heightened neural sensitivity [7, 8]. The hypothesis that an imbalance in the control of vascular endothelial responses, in favour of vasoconstriction, might underlie the increased vasospasm has led to investigation of a number of different endothelium-dependent responses in Raynaud's. The endothelium-dependent vasocontrictor endothelin-1 (ET-1) is increased with cold exposure [8, 9], and a number of investigators have identified increased levels of ET-1 in subjects with PRP [10–13] but without identifying a consistent rise with cold challenge. On the other side of the balance, reduced levels of endothelium-dependent vasodilators have been found, including bradykinin [14], in response to acetylcholine both in the finger microcirculation [5] and brachial artery [6], but not by all groups [15–17], and not consistently.

Leppart et al. [4] have demonstrated seasonal variation in circulating venous cGMP levels in subjects with PRP, an indirect measure of endothelium-dependent vasodilatation [18, 19]. They found reduced venous cGMP production in response to cold in the winter, and increased baseline levels in the summer. Seasonal variation in microvascular control may therefore be critical to interpreting responses in PRP. Seasonal change in healthy people has not previously been described. In this report we examine differences in finger skin temperature (Tsk) and microvascular blood flow in healthy men and women across the seasons, and compare the responses to women with PRP. We wished to determine if an underlying, seasonal alteration of vascular control influenced the observed response to the immediate environmental temperature, and to determine any difference in these responses between the different subject groups.

Methods

Subjects

Local Ethics Committee approval was obtained for these studies, and all subjects gave written informed consent prior to taking part. Ten women with PRP fulfilling the criteria of Allen & Brown [20] and with a minimum of one attack of vasospasm a week throughout the year were recruited. All had a normal full blood count, erythrocyte sedimentation rate, thyroid function tests, total immunoglobulins, and negative autoantibody screen, including rheumatoid factor, and antibodies to nuclear and extractable nuclear antigen. Age matched male controls (mean age 28.3 years ±8.3 sd.) and female controls (mean age 28.6 years ±9.0 sd.) were recruited (PRP mean age 28.5 years ±9.5 sd.). No women were postmenopausal, taking the oral contraceptive pill, and all women were in the follicular phase of the menstrual cycle. Controls were matched for smoking habit (one PRP subject, female control and male control smoked) and also had normal results for the investigations above. Raynaud's patients taking nifedipine stopped this 2 weeks before the laboratory visits. No other subject was taking any medication likely to influence vascular tone, or was hypertensive, hypercholesterolaemic or diabetic.

Protocol

Each subject attended for four visits, two in the winter months (November–February) and two the following summer (June–August). The two visits each season were designed to measure the responses to local cooling and heating of the hand. All subjects abstained from coffee, alcohol and tobacco for the 12 h preceding laboratory visits, and attended at 09.00 h. All subjects refrained from eating for 2 h prior to the laboratory visit, and had only a light breakfast on the day of the visit. No special instructions were given to any subjects about appropriate clothing to wear whilst travelling to the laboratory, and the PRP subjects were noted to be wearing more clothing on arrival, especially in the winter. All subjects wore light clothing in the laboratory. The mean daily 09.00 h outdoor temperatures for the visits are given in Table 1 (courtesy of Mr J. Hodgson, British Geological Survey, Keyworth, Nottingham). All experiments took place in a temperature regulated laboratory controlled between 22 and 24 °C for all visits.

Table 1.

Mean 09.00 outdoor temperature and mean finger skin temperatures (s.e.mean) on arrival at the laboratory, and after reaching a stable skin temperature in the constant environmental temperature, and the time to achieve this temperature (excluding the 6 min of recorded stable Tsk).

*S P < 0.002 compared with the summer visit, *C P < 0.002 compared with the male and female controls.

Outline of measurements

The measurements at each visit were as follows: Tsk was recorded immediately on arrival from the uncontrolled environmental temperature, after stabilizing in a controlled laboratory temperature, and finally in a hand box where the temperature was varied to achieve controlled finger Tsk. Baseline finger skin red blood cell flux (r.b.c. flux) was recorded after Tsk stabilized in the controlled laboratory temperature, and at controlled skin temperatures. Finally the changes in finger skin r.b.c. flux in response to iontophoresis of vehicle, acetylcholine and sodium nitroprusside were recorded at the controlled Tsk.

Measurements of Tsk

Tsk was recorded by thermocouples attached to the distal phalanx of both index fingers, and a control site at the interscapulae region of the back, immediately on arrival from the ambient temperature. The subject then put on light clothing, and rested semirecumbent on a couch with both hands resting on a support at the level of the left atrium. Temperature was recorded from these three sites every 2 min throughout the experiment. The temperature of the finger skin of the nondominant hand was documented when it became stable. A stable Tsk was defined as the temperature which remained constant (± 0.3 °C) for a minimum of 6 min after arrival in the temperature controlled laboratory.

The nondominant hand was then placed in a temperature-regulated unit designed to control Tsk. The subject sat in the same position, with hands at the same height as before. For one visit in each season the Tsk was warmed (target 35 °C), and for the other visit it was cooled (target 15 °C). The order of these visits was randomised.

The temperature-regulated unit was prewarmed to 35 °C. The temperature of the unit was then altered to control Tsk close to 35 °C, as precisely as possible, and maintained at this temperature. On the other visit the nondominant hand was placed in the temperature-regulated unit set at 0 °C for controls, or 8 °C for the PRP subjects because of concern of causing vasospasm. The hand was cooled to a controlled finger Tsk as close to 15 °C as possible, and the unit temperature modified to maintain this Tsk. The temperature-regulated unit had two thermistors at different sites to ensure constant temperature throughout, and a small fan to distribute the air evenly, which was shielded from the subject so as to prevent cold air being directed onto the hand. No attacks of Raynaud's were induced in the PRP subjects at the cold temperatures, although several subjects developed vasospasm as their hands were removed from the cold chamber.

Recording of finger skin r.b.c. flux and iontophoresis

Blood flow in the microcirculation was recorded by laser Doppler fluxmetry (MBF2, Moor Instruments, Devon, U.K.) and recorded as r.b.c. flux in volts. Acetylcholine and sodium nitroprusside were supplied by Sigma (Poole, Dorset), and freshly prepared as 0.1% solutions in 0.9% saline (vehicle). The finger iontophoresis chamber (Medical Physics Department, University Hospital, Nottingham) was attached, in a random order, to the dorsum of the index and middle fingers of the nondominant hand and recordings made from both fingers consecutively. The laser Doppler probe was incorporated in the centre of the chamber and recorded r.b.c. flux from the area to which iontophoresis was applied. Vehicle alone was randomised to be iontophoresed from either the anode or cathode, and from either finger. ACh and SNP were given by iontophoresis, in a random order to the index and middle fingers. Doses were used as determined from a previous study; 0.1% acetylcholine was given at 150 µA for 60 s with the iontophoresis chamber acting as the anode, and 0.1% sodium nitroprusside was given as 150 µA for 60 s with the iontophoresis chamber acting as the cathode [21]. R.b.c. flux was recorded for 1 min before iontophoresis, and 2 min after iontophoresis. Baseline and the maximum r.b.c. flux were measured, for each period of iontophoresis for all three substances.

Calculations and statistical analysis

Skin blood flow was measured as r.b.c. flux, and expressed as volts. All values are expressed as mean values ±s.e.mean. Statistical analysis was made using unbalanced repeated measures analysis of variance (BMDP programme 5v (release 7) 1992).

Results

Temperature measurements

a. Outdoor temperatures

(Table 1) The mean 09.00 h outdoor temperature for the summer visits was approximately 10 °C warmer than for the winter visits (P < 0.001). There was no significant difference between the outdoor temperatures within season for the visits of the three subject groups (Table 1).

b. Skin temperatures

(i) Temperatures on arrival (Table 1)

The Tsk of both hands immediately on arrival in the laboratory were cooler for all subjects during the winter visits (P < 0.0001). The PRP group had significantly colder Tsk on arrival in both seasons than the male or female controls (P < 0.0005). There was no difference between the Tsk of the female and male controls on arrival within season, or between the temperature of the two hands within any of the groups (data not shown).

(ii) Stable skin temperatures in the controlled environmental temperature (Table 1)

For all subjects the stable Tsk were warmer than the arrival temperatures (P < 0.0001) and did not change significantly after reaching stability. The stable Tsk of both hands reached by the PRP group in the laboratory environmental temperature of 23 °C remained significantly cooler than the hand Tsk for the two other groups (P < 0.002). Despite the identical laboratory temperature all groups had stable Tsk which were cooler during the winter than the summer (P < 0.002).

The time taken for the male and female controls to reach stable Tsk was comparable for all visits, but the PRP subjects took a significantly longer time to stabilize than the controls (P < 0.0001); this was longest in the winter.

(iii) Skin temperatures recorded at controlled hand temperatures (Table 2)

Table 2.

Mean skin temperatures in °C (s.e. mean) of different body sites, and mean final box temperature (°C) during the time the nondominant hand temperature is controlled in the box.

*H P < 0.01 compared with hot box visit, *F P < 0.05 compared with FC, *S P < 0.0001 compared with summer visits.

The temperatures of the nondominant hand in the temperature-regulated unit were maintained at the desired hot and cold Tsk with no significant difference between the temperatures of the three groups, or between season (Table 2), or between fingers (data not shown). These mean temperatures were maintained throughout the time in the temperature-regulated unit with no significant difference between the start and end temperatures (data not shown).

Having fixed the Tsk of the nondominant hand, differences in the Tsk of the dominant hand, which remained in room temperature, were assessed. The male controls had significantly warmer dominant hand Tsk than the female controls by a mean of 1.6 °C over the four visits (P < 0.01). In contrast the PRP group had a significantly cooler dominant hand Tsk than the female controls, by a mean of 1.34 °C (P < 0.05) over the four visits. The groups had cooler dominant hand Tsk whilst the nondominant hand was held at the cold compared with the warm Tsk (P < 0.01). The dominant hand Tsk was also significantly cooler overall for the winter visit (P < 0.0001), with most of this difference in Tsk identified during the hot box visits.

The recorded back temperature was 0.16 °C higher during the cold box visits than the hot box visits (P < 0.005). There was no significant difference in back temperatures between the different groups or between seasons.

c. Box temperatures

Different final box temperatures (Tbx) were required to maintain the constant finger Tsk in the different groups. The final Tbx, to achieve the equivalent Tsk, was colder for the male controls than the female controls, by a mean of 2.13 °C over all visits (P < 0.0001). The final Tbx, to achieve comparable Tsk, was warmer by 1.57 °C (P < 0.001) for the PRP than the female controls, and this difference was predominantly during the cold box visits.

There was a seasonal difference in the final cold Tbx required to achieve comparable cold Tsk, which was higher in the winter, by 0.90 °C (P < 0.0001) for all subjects.

Finger skin r.b.c. flux

(i) Baseline finger skin r.b.c. flux at controlled skin temperatures

(Table 3) The mean r.b.c. flux was significantly greater at controlled warm Tsk than at the colder Tsk for all subjects by a mean of 0.74 V (P < 0.0001). Despite controlled finger Tsk, there was significantly lower baseline finger skin r.b.c. flux under identical conditions in the winter in comparison with the summer, P < 0.0001. In the winter, at the cold Tsk there was significantly lower baseline finger skin r.b.c. flux in both female groups in comparison with the male controls (P < 0.05).

Table 3.

Mean finger skin temperatures and baseline flux (s.e.mean) at the hot and cold controlled skin temperatures demonstrating controlled skin temperatures were achieved. Within season controlling skin temperature controlled flux in the different groups, but different flux was identified at the same skin temperature between season.

*H P < 0.0001 from hot box visits, *S P < 0.0001 from the summer visits, *M P < 0.05 from the male controls.

Change in r.b.c. flux with iontophoresis

(i) Change in r.b.c. flux in response to vehicle iontophoresis

Iontophoresis of saline, from either anode or cathode, did not significantly change baseline r.b.c. flux in any of the subject groups at any Tsk (Table 4).

Table 4.

Baseline and change in flux with saline vehicle, with iontophoresis current from the anode and cathode. Flux in volts (s.e.mean).

		Saline (anode)		Saline (cathode)
Visit	Subjects	Baseline flux (volts)	Change in flux	Baseline flux (volts)	Change in flux
Summer	Male controls	2.63 (0.15)	0.02 (0.04)	2.50 (0.12)	0.01 (0.00)
hotbox	Female controls	2.53 (0.11)	0.03 (0.10)	2.48 (0.19)	0.05 (0.04)
	Raynaud's	2.43 (0.19)	0.06 (0.14)	2.35 (0.09)	0.05 (0.12)
Winter	Male controls	2.35 (0.20)	−0.04 (0.08)	2.46 (0.11)	0.01 (0.02)
hotbox	Female controls	2.20 (0.15)	0.02 (0.05)	2.68 (0.10)	0.05 (0.04)
	Raynaud's	2.27 (0.19)	0.04 (0.09)	2.29 (0.17)	0.03 (0.10)
Summer	Male controls	1.15 (0.09)	−0.01 (0.02)	1.25 (0.08)	0.00 (0.01)
coldbox	Female controls	1.17 (0.10)	0.01 (0.04)	1.20 (0.12).	0.03 (0.00)
	Raynaud's	1.15 (0.06)	0.02 (0.01)	1.10 (0.10)	−0.01 (0.01)
Winter	Male controls	0.64 (0.10)	0.00 (0.00)	1.26 (0.14)	0.01 (0.01)
coldbox	Female controls	0.52 (0.06)	−0.02 (0.02)	0.55 (0.09)	0.00 (0.01)
	Raynaud's	0.72 (0.13)	0.01 (0.02)	0.59 (0.10)	0.02 (0.03)

(ii) Response to iontophoresis of ACh and SNP at controlled Tsk

(Table 5) The lower Tsk in the coldbox was associated with lower baseline r.b.c. flux and a smaller vasodilatation in response to both ACh and SNP across the groups than the responses in the hotbox (P < 0.0001).

Table 5.

Mean baseline and change in flux with acetylcholine and sodium nitroprusside at the finger at controlled temperature. Flux in volts (s.e.mean).

*H P < 0.0001 from Hotbox, *F P < 0.0001 from female controls, *S P < 0.01 from summer visits.

Differences in the vasodilatation to ACh were seen between the groups. The change in r.b.c. flux with ACh was smaller by a mean of 0.87 volts (P < 0.0001), in the PRP group compared with the female controls for all visits, with most of this difference arising in the winter visits (P < 0.01). There was no significant difference in the PRP group compared with the female controls in the responses to SNP.

Discussion

The main finding in this study is that all subjects, healthy men and women as well as subjects with PRP, demonstrated seasonal differences in finger Tsk and r.b.c. flux which were consistently lower in the winter despite controlled experimental environmental temperatures. These seasonal responses appear to be persistent, and independent of transient but dramatic variations in environmental temperature, or the temperature of the opposite hand. Seasonal variation in finger Tsk and r.b.c. flux has not been described in healthy people before, but has been observed in animals. Deer digital arteries demonstrate increased vasoconstriction in the winter, associated with reduced endothelium-dependent responses [22], but these animals have increased exposure to the external environment compared with human beings. In man seasonal differences in blood pressure, serum lipids and fibrinogen, and mortality from cardiovascular disease are recognized [23, 24]. These findings all support a seasonal change in cardiovascular responses occurring in human beings despite our predominantly indoor existence.

We observed sex differences in Tsk and r.b.c. flux. Men had the warmest finger Tsk on arrival and at stable temperatures, and required a cooler box temperature to achieve equivalent finger Tsk than healthy women. These patterns fit in well with previously described sex differences in Tsk [25, 26]. However, there were no observed seasonal differences between the responses in healthy men and women.

The observation that women had colder finger Tsk than the men under most conditions was further exaggerated in the women with PRP, who had significantly cooler hands than the control women under all conditions. In addition the PRP women took longer to warm to a stable temperature although the final Tsk eventually achieved were still cooler.

Seasonal differences in Tsk were observed in the PRP subjects, as in the control subjects, but these Tsk were significantly colder in the winter in the PRP subjects. In the PRP subjects these seasonal differences were associated with altered endothelium-dependent vasodilatation, unlike the controls. Reduced vasodilatation in response to endothelium-dependent ACh, but not endothelium-independent SNP was seen in the PRP women compared to the control women, with most of this difference arising during the winter visits. This supports the hypothesis that seasonal variation in endothelium-dependent vasodilatation may contribute to the more severe responses to cold characteristic of subjects with PRP in the winter. Further support is added by Leppert et al. [4] who described the same pattern, demonstrating reduced venous cGMP levels, a measure of endothelium-derived nitric oxide, in Raynaud's phenomenon in the winter. They found these levels did not rise in response to cold challenge in the winter, but responded normally in the summer. We also found the smallest responses to ACh in subjects with PRP in the winter, and at the coldest skin temperatures. The vasodilatation in the summer in response to ACh in subjects with PRP was closer to that of female controls, but was still significantly reduced.

Like Leppert et al. [4], who found no seasonal difference in cGMP production in their healthy female controls, we also found no seasonal difference in the change of flux in response to ACh in the two control groups. This suggests nitric oxide dependent vasodilator mechanisms are not involved in the seasonal variation in Tsk observed in healthy men and women.

It was important to observe the response to the vasodilators starting from comparable Tsk and baseline r.b.c. flux, as variation in these altered the observed responses. Lower baseline flux, but also a smaller change in flux in response to ACh and SNP were observed at the colder finger Tsk for all groups, demonstrating the importance of controlled finger Tsk when comparing responses in different subject groups. This was particularly important in PRP subjects who have persistently colder hands in the same environment as demonstrated here, and by others [3]. These subjects require comparable Tsk in order to compare responses. Different Tsk, and hence different baseline vascular tone will alter the balance of vasoactive agents acting on peripheral vascular tone. This may contribute to some of the differences in endothelium-dependent responses reported by different groups [5, 6, 8].

In healthy men and women there is a seasonal influence on finger Tsk and microvascular blood flow responses to altered local environmental temperature. The seasonal influence is greater in women than men, and is further exaggerated in women with PRP. Reduced endothelium-dependent responses are demonstrated in PRP, and these are smallest in the winter, but endothelium-dependent vasodilator responses do not appear to contribute to the seasonal variation in finger Tsk and r.b.c. flux we observed in healthy controls.