The Effect Of Sodium Balance On Sweat Sodium Secretion

Eur J Appl Physiol (1998) 78: 516 ± 521

Ó Springer-Verlag 1998

ORIGINAL ARTICLE

A.J. Allsopp á R. Sutherland á P. Wood á S.A. Wootton

The effect of sodium balance on sweat sodium secretion and plasma aldosterone concentration

Accepted: 22 May 1988

Abstract The e€ect of manipulating sodium intake upon sweat sodium secretion was investigated during heat acclimation. Twenty-®ve male subjects were con®ned to an environmental chamber at a temperature of 25°C for 3 days, and then acclimated to heat by a further 5 days at 40°C. The subjects' daily sodium intake was controlled throughout as follows: high (HNa), 348.4 (0.8) mmol á day)1, n ˆ 7; moderate (MNa), 174.1 (0.6) mmol á day)1, n ˆ 9; or low (LNa), 66.3 mmol á day)1, n ˆ 9. Sodium losses were estimated from urinary, faecal and sweat collections using a whole-body washdown method. Plasma aldosterone concentration was also measured from venous blood sampled each morning. Measurements of body temperature and heart rate during the heat exposure phase indicated a degree of heat acclimation. During this heat phase there was a reduction (P < 0.01) in sweat sodium secretion for all three conditions which was greatest for the LNa condition, although this ®nding was not signi®cant (P < 0.1). In the LNa condition, plasma aldosterone concentration increased (P < 0.05) prior to heat exposure, and the secretion of aldosterone was potentiated (P < 0.01) during the heat exposure in comparison with the MNa condition. In contrast, the HNa diet produced a fall (P < 0.05) in plasma aldosterone concentration prior to heat exposure and an attenuation of aldosterone secretion thereafter. These ®ndings are inconsistent with the hypothesis that retention of sweat sodium is dependent upon a net body sodium de®cit, but demonstrate that aldosterone secretion is potentiated under such conditions. A.J. Allsopp (&) Institute of Naval Medicine, Gosport, Hampshire, PO12 2DL, UK R. Sutherland á S.A. Wootton Institute of Human Nutrition, University of Southampton, Hampshire, SO16 6YD, UK P. Wood Department of Endocrinology, Southampton General Hospital, Hampshire, SO16 6YD, UK

Key words Heat acclimation á Sodium balance á Aldosterone á Sweat

Introduction The concentration of sodium in human sweat is dependent upon sodium intake, the rate of sweating, and heat acclimation status. Sweat sodium concentration falls over 2±3 days of dietary sodium restriction (McCance 1936; Taylor et al. 1943; Conn 1949; Robinson et al. 1950), as sweat production slows (Weiner and van Heyningen 1952; Robinson and Robinson 1954), and following several successive days of heat exposure (Weiner and van Heyningen 1952; Allan and Wilson 1971). Whereas the acute change in sweat composition with sweat rate is a peripheral phenomenon (Yoshimura 1964), a systemic factor is responsible for the decrease in sweat sodium in adaptation to body heat. The primary stimulus for this adaptation is generally held to be increased aldosterone secretion (Collins and Weiner 1968) in response to a negative sodium balance (Smiles and Robinson 1971). More recently, this accepted link between sodium de®ciency, aldosterone and sweat sodium reabsorption has been challenged (Davies et al. 1981). These authors attributed changes in sweat sodium secretion to the acute stressors of heat and exercise, as reported elsewhere (Finberg and Berlyne 1977), rather than to a sodium de®cit. Furthermore, it is apparent that heat acclimation persists for several weeks (Pandolf et al. 1977) whereas plasma aldosterone concentration falls over a 10-day heat acclimation period (Kirby and Convertino 1986). The primary aim of the present study was to examine how the manipulation of dietary sodium intake alters aldosterone secretion and consequent sodium balance, with the secondary aim of examining the e€ect of sodium balance upon sweat sodium secretion. This paper reports the latter aim. It was hypothesised that retention of sweat sodium occurs if the net sodium balance status of the body is negative, but not if this status is positive.

517

Methods Subjects The study was conducted at the Institute of Naval Medicine, Hampshire, UK, following the approval of the local ethical committee. All volunteers were informed about the experiment in accordance with the code of ethics of the World Medical Association and were required to pass a full medical examination prior to their participation in this investigation. Twenty-®ve male subjects (aged 18±40 years, body mass 64±102 kg and body mass index 20± 34 kg á m)2) participated. Protocol The subjects were con®ned to an environmental chamber at a temperature of 25°C for 3 days (control), followed by 5 days at 40°C from 0800 hours to 1800 hours (25°C from 1800 hours to 0800 hours). Relative humidity was maintained at 40% throughout, with a constant air velocity of 0.5 m á s)1. These experiments were conducted in the winter and spring months, and so subjects were assumed to be unacclimatised to heat. None of the subjects reported participation in regular strenuous physical training. Subjects wore shorts and sandals during the exercise period and long cotton underwear during the sweat washdown days, as described below for estimation of sweat sodium. Dietary conditions Dietary sodium intake was prescribed as follows: high (HNa), 348.4 (0.8) mmol á day)1, n ˆ 7; moderate (MNa) 174.1 (0.6) mmol á day)1, n ˆ 9; low (LNa), 66.3 mmol á day)1, n ˆ 9 (see Fig. 1). Energy requirements were estimated as 1.4 times the predicted basal metabolic rate calculated from age and body mass (Scho®eld 1985). Dietary manipulation was achieved by feeding standard pre-prepared food items that were selected using the manufacturer's nutrient analysis data which had been previously identi®ed as having a high accuracy with respect to nutrient content. The LNa intake was achieved by appropriate substitution with low-sodium food items, whereas sodium chloride tablets were included with meals in the HNa diet. Water was permitted ad libitum. Quantities of

Fig. 1 Schematic diagram of the dietary and environmental conditions. Low (LNa): 66.3 mmol á day)1 (n ˆ 9); moderate (MNa): 174.1 ‹ 0.6 mmol á day)1 (n ˆ 9); high (HNa): 348.4 ‹ 0.8 mmol á day)1 (n ˆ 7)

fruit juice, milk and tea were also permitted as part of the daily food allowance. Measures Sodium losses were estimated from daily 24-h urinary collections, faecal samples and whole-body washdowns on days 3, 4 and 8 using the procedure described by Collins et al. (1971). Sodium concentration was analysed by emission spectrophotometry (ICPAES, Perkin-Elmer Plasma 400) following sample preparation against certi®ed calibration standards (National Institute of Standards and Technology, Gaithersburg, Maryland, USA: NIST 2670). Blood was taken without stasis daily at 0700 hours before rising, and the samples were analysed for plasma aldosterone by radioimmunoassay (Maia procedure, Serono Diagnostics, Rome, Italy) against standards of 0±6.925 nmol á l)1 aldosterone. Relative changes in plasma volume were estimated from haematocrit and haemoglobin concentration (Dill and Costill 1974). Naked body mass was measured (to 5 g) at 0700 hours each day. The rise in aural temperature (Tau) was measured by aural thermistors and recorded using an electronic data logger (Grants, Cambridge, UK) whilst the subjects performed a light stepping exercise task [a metabolic heat production of 313 (7) J á s)1] for 1 h each day. Heart rate was monitored during the exercise period via a three-lead electrocardiogram (S & W Diascope, Sweden). Statistics A repeated measures analysis of variance was used to assess differences between the three dietary conditions, following appropriate transformation (natural logarithm for sweat and faecal sodium losses and aldosterone).

Results Summary results are given as the mean (SE). Unless stated otherwise, results are quoted at the 5% level of signi®cance.

518 Fig. 2 Urinary sodium excretion of the three dietary groups during the control and heat exposure phases: LNa (n ˆ 9); MNa (n ˆ 9); HNa (n ˆ 7)

The sodium intakes for the three dietary groups throughout the trial were: LNa 66.3 mmol á day)1; MNa 174.1 (0.6) mmol á day)1; HNa 348.4 (0.8) mmol á day)1. Urinary output during the control phase was: LNa 2094 (228) ml; MNa 2490 (513) ml; HNa 2625 (788) ml, falling (P < 0.01) to LNA 1156 (193) ml; MNa 1522 (400) ml; HNa 1387 (391) ml upon heat exposure. The mean ¯uid intake for the control phase was: LNa 2689 (261) g; MNa 3393 (584) g; HNa 3402 (763) g, rising (P < 0.05) to LNa 4346 (289) g; MNa 5177 (384) g; HNa 5223 (384) g upon heat exposure. Changes in urinary sodium excretion (UNa) are shown in Fig. 2. The di€erence in UNa between the groups (P < 0.01) during the control phase re¯ects the di€erences in dietary intake. There was a marked reduction (P < 0.01) in UNa on exposure to heat for each group. During the heat phase this divergence between the groups was maintained (P < 0.01). The initial fall in UNa was maintained until days 5 and 6, before rising (P < 0.01). Faecal sodium losses on day 4 and day 8 are given in Table 1. The values for day 4 were not statistically di€erent, whereas on day 8 faecal sodium for the HNa group was greater (P < 0.05) than for LNa. Table 1 Sodium intake, 24-h urinary sodium excretion, faecal sodium excretion, sweat sodium secretion during the 12-h periods (0800 hours±2000 hours), and sodium balance of the four dietary conditions on the ®rst (day 4) and last day (8) of heat exposure [mean (SE)]. (HNa High sodium diet, MNa moderate sodium diet, LNa low sodium diet)

day Sodium intake (mmol á day)1)

The average values of sweat sodium secretion for days 4 and 8 are presented in Table 1. On day 4 sweat sodium secretion was highest in the HNa condition and lowest in the LNa group, although this di€erence was not signi®cant. There was a reduction (P < 0.01) in sodium secretion in all three conditions from day 4 to day 8, which was greatest in the LNa condition. The in¯uence of diet upon this reduction in sweat sodium secretion approached signi®cance (P < 0.01). Following the heat exposure, sweat sodium secretion in the LNa condition was significantly (P < 0.05) lower than for either MNa or HNa. Overnight (12-h) values for day 3 were: LNa 4.9 (0.8) mmol; MNa 7.7 (0.9) mmol; HNa 14.6 (2.9) mmol, and on day 7 were unchanged: LNa 7.2 (1.5) mmol; MNa 7.8 (1.7) mmol; HNa 13.7 (1.3) mmol. Total [mean (SE)] sodium excretion on day 3, before heat exposure was: LNa 83 (7) mmol; MNa 162 (7) mmol; HNa 315 (11) mmol. All three values were signi®cantly (P < 0.05) di€erent from each other. Subjects were assumed to be in sodium balance on this 3rd day of constant sodium intake, in accordance with the theoretical model of Strauss (Strauss et al. 1958, Simpson 1988), prior to commencing the heat phase. The net HNa (n = 7)

MNa (n = 9)

LNa (n = 9)

348.4 (0.8)

174.1 (0.6)

66.3

24-h Urinary sodium excretion (mmol)

4

205.1 (20.3)**

66.4 (5.6)**

26.9 (4.8)

8

251.3 (20.9)**

92.3 (11.2)**

32.4 (6.8)

Faecal sodium excretion (mmol)

4

10.5 (3.2)

8

14.2 (6.0)*

5.4 (2.6)

1.4 (0.4)

Sweat sodium excretion (mmol)

4

78.8 (10.3)

63.8 (8.0)

53.5 (8.5)

8

52.6 (6.2)*

39.1 (4.0)*

24.8 (2.1)

Sodium balance (mmol)

4

40.0 (16.3)*

32.4 (6.1)*

23.3 (9.8)

8

14.7 (16.3)

29.1 (10.2)

0.2 (5.1)

3.8 (2.1)

Signi®cantly di€erent from LNa group *(P < 0.05) **(P < 0.01)

4.4 (2.2)

519

sodium balance on days 4 and 8 is given in Table 1. Analysis of variance indicated a signi®cant di€erence between the group average values (P < 0.01), but no signi®cant e€ect of diet upon the change in sodium balance over time. Body mass was reduced (P < 0.01) in the heat in all three conditions by 0.6 (0.1)%, relative to day 3, indicative of a negative ¯uid balance. Aural temperatures mid-way through the exercise period were: LNa 37.45 (0.09)°C; MNa 37.50 (0.10)°C; HNa 37.59 (0.06)°C on day 4, and 37.09 (0.11)°C; 37.24 (0.08)°C; 37.39 (0.04)°C on day 8. The within-group di€erences in temperature at this time point [LNa 0.34 (0.17)°C; MNa 0.26 (0.07)°C; HNa 0.29 (0.05)°C] indicated a signi®cant (P < 0.01) fall in Tau over the period of heat exposure. Exercise heart rate also declined (P < 0.01) for all subjects from a mid-exercise value of 126 (4.3) beats á min)1 on day 4, reduced to 107.3 (2.4) beats á min)1 on day 8. The mean increase in plasma volume over the 5 days of heat exposure was 7.9 (1.1)%, but there were no signi®cant di€erences between the three dietary conditions: LNa 10.1 (2.3)%; MNa 8.3 (1.2)%, HNa 4.6 (1.7)%. Plasma aldosterone concentration during the trial is shown in Fig. 3. In the MNa condition, plasma aldosterone was signi®cantly (P < 0.01) elevated during heat exposure, as expected. This response was signi®cantly (P < 0.01) attenuated in the HNa condition compared with the MNa diet: indeed, within this HNa group, there was a slight fall (P < 0.05) in concentration from day 1 [334 (19) pmol á l)1] to day 3 [266 (26) pmol á l)1] and the rise during the heat exposure on day 6 [351 (33) pmol á l)1] was not signi®cant compared to the control condition. In contrast, plasma aldosterone concentrations were signi®cantly higher (P < 0.01) during the heat exposure in the LNa condition compared to the MNa group, and the LNa plasma concentration of this hormone increased (P < 0.05) prior to heat exposure [day 1 412 (41) pmol á l)1; day 3 638 (70) pmol á l)1].

Fig. 3 Plasma aldosterone of the three dietary groups during the control, heat exposure and post exposure phases: LNa (n ˆ 9); MNa (n ˆ 9); HNa (n ˆ 7)

Discussion This study has demonstrated that ingestion of a lowsodium diet potentiates the aldosterone response associated with heat exposure, and that a high-sodium diet attenuates that response. Furthermore, plasma aldosterone concentrations were observed to change depending upon dietary sodium intake prior to the heat exposure. These ®ndings underline the importance of a controlled steady-state intake of sodium when conducting dietary balance studies. Sodium balance was achieved in this study, as judged from the values for total sodium excretion on day 3 in relation to the sodium intake of each of the three groups. In comparison to the present study, earlier investigations have either used a small number of subjects (e.g. Robinson et al. 1950) or have assumed a habitual sodium intake (e.g. Davies et al. 1981). The absence of an aldosterone response to heat exposure for the HNa diet indicates that the increased secretion of this hormone is dependent upon sodium balance, or possibly sodium availability; in the MNa group, although mean values indicated a net positive sodium balance, two subjects were in de®cit on day 3 and one of these was also negative on day 4. These data support the ®ndings of Smiles and Robinson (1971) who reported an increased urinary excretion of tetrahydroaldosterone, the major metabolite of aldosterone, during sodium deprivation. Hence the hypothesis that aldosterone secretion is sensitive to sodium balance (Collins and Weiner 1968) would appear to be correct. However, it would be incorrect to infer that this is the primary control mechanism: acute heat stress, exercise in the heat and dehydration have all been shown to stimulate aldosterone production via increased renin-angiotensin activity (Finberg and Berlyne 1977; Follenius et al. 1979; Francesconi et al. 1985). Rather, the aldosterone response to sodium de®ciency appears to be an important additional safeguard for the conservation of body sodi-

520

um. On re¯ection this is unsurprising since aldosterone secretion is more sensitive to changes in plasma potassium concentration, adrenocorticotropic hormone and angiotensin II than to sodium (Funder 1993; Verrey and Beron 1996). With regard to the question of whether a sodium de®cit is an essential pre-requisite of sodium conservation by the sweat glands, the apparent reduction of sweat sodium in the HNa and MNa conditions, which remained in positive balance with respect to total body sodium, would suggest that this is not the case. However, the LNa condition incurred a greater reduction in sweat sodium secretion (mean values: HNa 24.0 mmol; MNa 27.4 mmol; LNA 37.5 mmol). Thus the data presented here suggest that sodium conservation by the sweat glands still occurs, even in the absence of a sodium de®cit, but that this response is potentiated by a negative sodium balance, possibly via the increased aldosterone secretion. This ®nding is in contrast to that of other authors (e.g. Taylor et al. 1943) who reported an absence of sodium conservation in heat-acclimated subjects given a high salt diet. Unfortunately only sweat secretion data are given here; actual concentrations were impossible to assess over the 12-h washdown period since we did not have accurate values for sweat loss. It was not possible to calculate sweat losses accurately from mass losses over this period because the subjects were free-living and had access to the prescribed food and ¯uids ad libitum. It is possible that the fall in sweat sodium is a result of reduced sweat rate as acclimation progressed. Acclimation is assumed to have occurred, as indicated by the expansion of plasma volume, and the falls in body temperature and heart rate during exercise. At the environmental temperature used in this investigation (40°C), the observed reduction in body temperature cannot be ascribed to improved heat dissipation through mass ¯ow of heat from the body core to the shell and increased convective and radiative heat loss. It must be inferred therefore that sweat rate increased rather than decreased during the acclimation period, or that the eciency of sweat evaporation increased. Those previous studies in which sweat composition was investigated also have limitations: sweat loss has usually been measured over the short period of time over which sweat rate and composition is known to alter (Ladell et al. 1944; Weiner and van Heyningen 1952; Fox et al. 1963). Furthermore, many of these previous studies have employed the sweat bag collection method which in itself is likely to give confounding results due to the process of hydromeiosis (Peiss et al. 1956). This is the ®rst time that changes in sweat secretion, as measured by the whole-body washdown method, have been reported over such an extensive time period. In the study described here a long heat sojourn was used to initiate heat acclimation responses, and physical activity was kept to a minimum. The period of light exercise was intended to counter any fall in basal metabolic rate

which may have otherwise resulted from the chamber con®nement. In addition, dietary sodium intake was controlled at the same level prior to and during the heat exposure. This methodology contrasts with that which utilizes a combination of exercise and heat stress to produce heat acclimation (Francesconi et al. 1985), or experiments whereby sodium intake is reduced upon heat exposure (Armstrong et al. 1993; Francesconi et al. 1993). These design parameters are fundamental to the investigation of the e€ects of sodium balance upon aldosterone stimulation. It is concluded that aldosterone plays a facilitative rather than primary role in the regulation of sweat sodium reabsorption and sodium homeostasis. The retention of sodium by the sweat glands is not dependent upon a sodium de®cit, but should a de®cit occur, then aldosterone secretion will be potentiated, and this response will further assist sweat sodium reabsorption. Whilst recognising the methodological limitations of this study with regard to the measurement of sweat sodium concentration, the data here do not support the hypothesis that reabsorption of sodium by the sweat glands is dependent upon a net negative sodium balance. Acknowledgements The authors gratefully acknowledge the e€orts and advice of the following: the subjects, Mary Whetton, Dr. F. StC. Golden and Dr. K. Collins. This work was supported by the Ministry of Defence (Navy). The conduct of these experiments compiled with all current laws in the UK.

References Allan JR, Wilson CG (1971) In¯uence of acclimatization on sweat sodium concentration. J Appl Physiol 30:708±712 Armstrong LE, Hubbard RW, Askew EW, Francesconi RP (1993) Responses of soldiers to 4-gram and 8-gram NaCl diets during 10 days of heat acclimation. In: Marriott B M (ed) Nutritional needs in hot environments. National Academy Press, Washington DC, pp 247±258 Collins KJ, Weiner JS (1968) Endocrinological aspects of exposure to high environmental tempertures. Physiol Rev 48:785±839 Collins KJ, Eddy TP, Hibbs A, Stock AL (1971) Nutritional and environmental studies on an ocean-going oil tanker. 2. Heat acclimatization and nutrient balances. Br J Indust Med 28:246± 258 Conn JW (1949) The mechanism of acclimatisation to heat. Advan Intern Med 3:373±393 Davies JA, Harrison MH, Cochrane LA, Edwards RT, Gibson TM (1981) E€ect of saline loading during heat acclimatisation on adrenocortical hormone levels. J Appl Physiol 50:605±612 Dill DB, Costill DL (1974) Calculation of percentage changes in volumes of blood, plasma, and red cells in dehydration. J Appl Physiol 37:247±248 Finberg JPM, Berlyne GM (1977) Modi®cation of renin and aldosterone response to heat by acclimatisation in man. J Appl Physiol 42:554±558 Follenius M, Brandenberger G, Reinhardt B, Simeoni M (1979) Plasma aldosterone, renin activity, and cortisol responses to heat exposure in sodium depleted and repleted subjects. Eur J Appl Physiol 41:41±50 Fox RH, Goldsmith R, Kidd DJ, Lewis HE (1963) Acclimatisation to heat in man by controlled elevation of body temperature. J Physiol (Lond) 166:530±547

521 Francesconi RP, Sawka MN, Pandolf KB, Hubbard RW, Young AJ, Muza S (1985) Plasma hormonal responses at graded hypohydration levels during exercise heat stress. J Appl Physiol 59:1855±1860 Francesconi RP, Armstron LE, Leva NM, Moore RJ, Szlyk PC, Matthew WT, Curtis WC, Hubbard RW, Askew EW (1993) Endocrinological responses to dietary salt restriction during heat acclimation. In: Marriott BM (ed) Nutritional needs in hot environments. National Academy Press, Washington DC, pp 259±275 Funder JW (1993) Aldosterone action. Ann Rev Physiol 55:115± 130 Kirby CR, Convertino VA (1986) Plasma aldosterone and sweat sodium concentrations after exercise and heat acclimation. J Appl Physiol 61:967±970 Ladell WSS, Waterlow JC, Faulkner Hudson M (1944) Desert climate. Physiological and clinical observations. Lancet 247:491±497 & 527±531 McCance RA (1936) Medical problems in mineral metabolism. Lancet 876:825±830 Pandolf KB, Burse RL, Goldman RF (1977) Role of physical ®tness in heat acclimatisation, decay and reinduction. Ergonomics 20:399±408 Peiss CN, Randall WC, Hertzman AB (1956) Hydration of the skin and its e€ect on sweating and evaporative heat loss. J Invest Dermatol 26:459±470

Robinson S, Kincaid RK, Rhamy RK (1950) E€ect of salt de®ciency on the salt concentration in sweat. J Appl Physiol 3:55±62 Robinson S, Robinson AH (1954) Chemical composition of sweat. Physiol Rev 34:202±219 Scho®eld WN (1985) Predicting basal metabolic rate, new standards and a review of previous work. Human nutrition: Clinical Nutrition 39C [Suppl 1]: 5±41 Simpson FO (1988) Sodium intake, body sodium, and sodium excretion. Lancet 2(8601): 25±29 Jul 2 Smiles KA, Robinson S (1971) Sodium ion conservation during acclimatisation of men to work in the heat. J Appl Physiol 31:63±69 Strauss MB, Lamdin E, Smith WP, Bleifer DJ (1958) Surfeit and de®cit of sodium. A M A Arch Intern Med 102:527±536 Taylor HL, Henschel A, Mickelsen O, Keys A (1943) The e€ect of sodium chloride intake on the work performance of man during exposure to dry heat and experimental heat exhaustion. Am J Physiol 140:439±451 Verrey F, Beron J (1996) Activation and supply of channels and pumps by aldosterone. News Physiol Sci 11:126±133 Weiner JS, Van Heyninge RE (1952) Salt losses of men working in hot environments. Br J Industr Med 9:56±64 Yoshimura H (1964) Organ systems in adaptation: the skin. In: Dill DB, Adolph EF. Wilber CG (eds) Handbook of Physiology á Section 4. Adaptation to the environment. American Physiological Society, Washington, pp 109±131