Skip to main content
NCBI home page
The Journal of Clinical Endocrinology and Metabolism logoLink to The Journal of Clinical Endocrinology and Metabolism
. 2008 Mar 18;93(6):2072–2078. doi: 10.1210/jc.2007-2336

Osmotic and Nonosmotic Regulation of Arginine Vasopressin during Prolonged Endurance Exercise

Tamara Hew-Butler 1, Esme Jordaan 1, Kristin J Stuempfle 1, Dale B Speedy 1, Arthur J Siegel 1, Timothy D Noakes 1, Steven J Soldin 1, Joseph G Verbalis 1
PMCID: PMC2435641  PMID: 18349067

Abstract

Context: Although the primary cause of exercise-associated hyponatremia (EAH) is relative overconsumption of fluids beyond the kidneys’ ability to excrete excess fluid, the mechanisms limiting maximum renal excretory ability during exercise remain to be elucidated.

Objective: The objective of the study was to: 1) perform a comprehensive evaluation of the endocrine secretion of pituitary, natriuretic and adrenal steroid hormones, and cytokines immediately before and after running an ultramarathon; and 2) evaluate the relationship between osmotic and nonosmotic stimuli to arginine vasopressin (AVP) secretion within the overall context of assessing the hormonal regulation of fluid balance during prolonged endurance exercise.

Design: This was an observational study.

Setting: The study setting was a 56-km ultramarathon.

Participants: Eighty-two runners participated in the study.

Interventions: There were no interventions.

Main Outcome Measures: Plasma sodium concentration [Na+] and plasma volume [(AVP)p] were measured.

Results: Fluid homeostasis during exercise (356 ± 4 min) was maintained with ad libitum fluid intakes. [Na+] was maintained from before the race (139.3 ± 0.3 mmol/liter) to after the race (138.1 ± 0.4 mmol/liter) with a significant decrease in plasma volume (−8.5 ± 0.1%, P < 0.01). Increases in the plasma (AVP)p (3.9-fold), oxytocin (1.9-fold), brain natriuretic peptide (4.5-fold), and IL-6 (12.5-fold) were highly significant (P < 0.0001). Changes in brain natriuretic peptide, oxytocin, and corticosterone were associated with 47% of the variance noted in (AVP)p and 13% of the variance in plasma [Na+] in pathway analyses.

Conclusions: (AVP)p was markedly elevated after the ultramarathon despite unchanged plasma [Na+]. Therefore, an inability to maximally suppress (AVP)P during exercise as a result of nonosmotic stimulation of AVP secretion may contribute to the pathogenesis of exercise-associated hyponatremia if voluntary fluid intake were to exceed fluid output.

This study documents non-osmotic AVP secretion in runners participating in a 56-km foot race. The likely non-osmotic stimulus is due to significant plasma volume contraction combined with the possible influence of other endocrine factors such as oxytocin, brain natriuretic peptide, and corticosterone.

There is currently a paucity of data describing the endocrine regulation of fluid and electrolyte balance during prolonged endurance exercise. Most of the available field-based data have been collected in studies of small (n < 10) groups (1,2,3) under varying states of hydration (4) and with considerable time delays in blood sampling that often exceed the half-lives of many of the regulatory hormones being evaluated (3,5,6). Each of these factors markedly limits the strength of the scientific knowledge obtained.

Exercise-associated hyponatremia (EAH) has recently emerged as the most common life-threatening complication of endurance exercise (7). Four otherwise healthy female and one male marathon runner have died from EAH since 1993, and 23–27% of Ironman triathletes have finished triathlons with documented hyponatremia (serum sodium concentrations < 135 mmol/liter) (8). Although the primary cause of EAH is a relative overconsumption of fluids beyond the ability of the kidneys to excrete excess fluid, the mechanisms that limit maximum renal excretory ability during exercise remain to be elucidated.

Inappropriate secretion of the antidiuretic hormone, arginine vasopressin (AVP), has been implicated as the main causative factor in the pathogenesis of EAH (9). This hypothesis suggests that nonosmotic stimuli to AVP secretion can occur normally during prolonged endurance exercise. This normal nonosmotic stimulation of AVP during exercise would then become pathological if hypoosmolality develops as a result of voluntary fluid intake that exceeds urinary and sweat water losses.

The aim of this study was to perform a comprehensive evaluation of the endocrine secretion of pituitary, natriuretic, and adrenal steroid hormones, as well as cytokines, in well-trained endurance athletes immediately before and after running a 56-km ultramarathon. More specifically, we wanted to evaluate the relationship between osmotic and nonosmotic stimuli to AVP secretion within the overall context of assessing the hormonal regulation of fluid and electrolyte balance during prolonged endurance exercise. We hypothesized that nonosmotic factors might contribute significantly to AVP secretion during an ultramarathon. This would explain why EAH could occur as a result of a failure to suppress AVP secretion despite profound hypotonicity and whole body fluid overload.

Subjects and Methods

Subjects and sample collection

Informed written consent was obtained from 82 runners competing in the Two Oceans 56-km ultramarathon, held in Cape Town, South Africa, in 2005. This study was approved by both the Ethics Committee of the University of Cape Town and the Georgetown University Institutional Review Board.

Baseline body weight, blood, and urine samples were obtained within 60 min of the start of the race. Postrace body weight, blood, and urine samples were obtained immediately upon race completion; the average time interval between race finish to postrace blood draw was 6.9 ± 0.5 min. All pre- and postrace blood samples were immediately placed on ice and centrifuged within 10 min at 3000 rpm; separated plasma was stored on dry ice until the samples were frozen to −80 C. All samples remained frozen until further analysis was performed. Body weight was measured with athletes in racing attire without shoes on calibrated Adamlab JPS electronic scales placed on a hard, flat surface (Scales; Brackenfell, South Africa). Food and fluid intake were allowed ad libitum during the race, and estimated total fluid intake was self-reported immediately after the race via a questionnaire.

Analytical measurements

Changes in plasma volume were estimated by comparing pre- and postrace measurements of plasma protein using a clinical refractometer (Schuco Clinical Refractometer 5711–2020, Japan). Plasma and urine sodium (Na+) and potassium (K+) concentrations were measured using ion-selective electrodes (Synchron EI-ISE; Beckman, Fullerton, CA). Plasma osmolality was measured using a vapor pressure osmometer (VAPRO 5520; WESCOR, Logan, UT).

Hormone measurements

Plasma levels of arginine vasopressin [(AVP)P] and oxytocin [(OT)P] were measured by specific RIA after acetone-ether extraction as described previously (10). The standard curve for AVP is linear between 0.5 and 10.0 pg/tube with the use of a synthetic AVP standard (PerkinElmer Life Sciences Inc., Boston MA). The minimum detectable concentration of AVP in extracted plasma was 0.5 pg/ml. The AVP antiserum (R-4) displayed less than 1% cross-reactivity with oxytocin (OT). The standard curve of the OT assay was linear between 0.25 and 5.0 pg/tube with the use of a synthetic OT standard (PerkinElmer Life Sciences). The minimum detectable concentration of OT in extracted plasma was 0.25 pg/ml. The OT antiserum (Pitt-Ab2) displayed less than 1% cross-reactivity with AVP.

Eleven adrenal steroid hormones (cortisol, 11-deoxycortisol, aldosterone, corticosterone, dehydroepiandrosterone, dehydroepiandrosterone sulfate, testosterone, androstenedione, 17-hydroxyprogesterone, progesterone and 25-hydroxyvitamin D3) were measured using a liquid chromatography-tandem mass spectrometer in conjunction with an atmospheric pressure photoionization source, using methodology described previously (11).

Brain natriuretic peptide (BNP), as assessed via measurement of the more stable cleaved inactive fragment, N-terminal pro-BNP [(NT-pro-BNP)P], using the automated RxL Dimension (Dade Behring, Newark, DE), as described previously (12). (IL-6)P was measured using a chemiluminescence method with a commercial kit and an automatic chemiluminassay analyzer (Immulite 1000 system; Diagnostic Products Corp., Los Angeles, CA). The minimal detectable limit of the assay was 5 pg/ml.

Statistical analysis

Differences were calculated as postrace values minus prerace values and presented as means ± sem, together with the range of values. Paired t tests were used to assess significant differences between pre- and postrace values.

Potential models for the associations among the fluid balance markers, peptide hormones, cytokine changes, and adrenal steroid hormone changes were presented as path diagrams, and path analyses were performed on the variance-covariance matrix, using the maximum-likelihood method of parameter estimation. Goodness of fit for each model was assessed via the χ2 analysis, the normed fit index (NFI), the nonnormed fit index (NNFI), and the comparative fit index (CFI). Indices over 0.9 indicated an acceptable fit between model and data. All reported values represent standardized estimates, unless specifically indicated as an unstandardized estimate. Individual path coefficients were significant at the P < 0.05 level when t was greater than 1.96. The R2 value reflects the percent of variance of the endogenous variable that was accounted for by their direct antecedents.

Multiple linear regression analyses were performed to determine which endocrine variables best correlated with fluid balance parameters. Statistical significance was accepted when P < 0.05.

Results

Overall, a total of 5472 men completed the Two Oceans race with a mean finishing time of 5:35 (hour:minutes), and 1502 women (22%) completed the event with a mean finishing time of 5:55. The maximum temperature on race day was 23 C and the minimum temperature at race start was 15 C. Coke and sachets containing 100 ml of water and Powerade [(Na+) = 10 mmol/liter, 8% carbohydrate] were available at fluid stations located every 3 km along the 56-km course.

All 82 runners who consented to participate in the research trial completed the Two Oceans Ultramarathon, including 58 men (71%) and 24 women (29%), with a combined mean finishing time of 5:56. The mean age of the cohort was 43.7 ± 1.1 yr. The mean prerace body weight of the subjects was 74.3 ± 1.4 kg with the mean body mass index of 23.5 ± 0.3 kg/cm2. The subjects’ mean estimated total fluid intake during the race was 3.2 ± 1.7 liters.

Physiological markers of fluid balance homeostasis were assessed by changes (Δ = postrace minus prerace values) in body weight, plasma (Na+), plasma volume, urine (Na+), and urine osmolality (Table 1). The changes in body weight, plasma volume, urine (Na+), and urine osmolality were statistically significant from before to after the race, whereas the changes in plasma (Na+) were not. Significant increases (P < 0.0001) from pre- to postrace values were observed for (AVP)P, (OT)P, (NT-proBNP)P, and (IL-6)P (Table 1) and in nine of the 11 adrenal steroid hormones measured (Fig. 1).

Table 1.

Physiological markers of fluid balance, peptide hormone, and cytokine changes from pre- to postrace values

Variable Mean ± sem Minimum Maximum
Body weight Δ (%) (n = 79) −3.8 ± 0.1a −6.4 −0.4
Plasma (Na+) Δ (mmol/liter) (n = 78) −1.3 ± 0.5 −9.9 15.4
Plasma volume Δ (%) (n = 79) −8.5 ± 0.1a −21.0 9.0
Urine (Na+) Δ (mmol/liter) (n = 64) −25.5 ± 6.4b −163.0 116.5
Urine osmolality Δ (mOsmol/kg H2O) (n = 72) 122.3 ± 33.3a −557.0 644.0
AVP Δ (pg/ml) (n = 78) 5.7 ± 0.9c −2.4 43.2
OT Δ (pg/ml) (n = 79) 1.8 ± 0.2c −1.7 6.4
NT-pro-BNP (pg/ml) (n = 76) 130.9 ± 15.5c −123.0 622.0
IL-6 Δ (pg/ml) (n = 61) 48.5 ± 3.8c 13.5 197.0
Open in a new tab
a

P < 0.01 from pre- to postrace values. 

b

P < 0.001 from pre- to postrace values. 

c

P < 0.0001 from pre- to postrace values. 

Figure 1.

Figure 1

Open in a new tab

Changes in adrenal steroid hormones from before to after the race within the pathway of steroid biosynthesis from cholesterol. The adrenal steroids measured in this study are noted inside each rectangular box. Mean differences (mean ± sem) between pre- and postrace concentrations are noted directly underneath each plasma steroid, respectively. The asterisks denote statistical significance from before to after the race. DHEA, Dehydroepiandrosterone; DHEAS, dehydroepiandrosterone sulfate; 17OH-pregnenolone, 17-hydroxypregnenolone; 17OH-progesterone, 17-hydroxyprogesterone.

Figure 2 depicts the pathway (model fit: CFI = 1.0; NFI = 0.98; NNFI = 1.04) between postrace (IL-6)P, (cortisol)P Δ, (corticosterone)P Δ, (AVP)P Δ, and plasma (Na+) Δ. Overall, (AVP)P Δ was associated with13% of the variance seen in plasma (Na+) Δ, whereas postrace (IL-6)P, (cortisol)P Δ, and (corticosterone)P Δ was associated with 21% of the variance seen in (AVP)P Δ. The pathway coefficient between (corticosterone)P Δ and (AVP)P Δ was statistically significant (t > 1.96); for every 1 U increase in (corticosterone)P Δ, (AVP)P Δ increased by 2.7 (unstandardized) units and holding constant the effects of the other independent variables. This effect was at least 8 times stronger than the other effects on AVP. Similarly, the pathway coefficient between (AVP)P Δ and plasma (Na+) Δ was significant; for every 1 U increase in (AVP)P Δ, plasma (Na+) Δ increased by 0.2 (unstandardized) units.

Figure 2.

Figure 2

Open in a new tab

Pathway diagram showing a significant mathematically predicted association of AVP Δ on plasma (Na+) Δ and corticosterone Δ on AVP Δ for the overall cohort (P < 0.05). Single and underlined values represent the entire cohort. Separate male (M) and female (F) coefficients are listed below the overall values. All reported values are standardized estimates, with unstandardized estimates in brackets. The double-headed arrows indicate the correlation estimates between the endogenous/independent variables. R2 percent of the variation is explained by the direct antecedent(s).

When analyzed by sex, no significant changes in any of the overall pathway models were noted in the male cohort, primarily because the number of males dominated the analyses (71%). When males and females were analyzed as separate groups, however, the pathway coefficient for females in Fig. 2 between (corticosterone)P Δ and (AVP)P Δ became insignificant. This resulted in a much higher R2 value for males (32%), compared with females (2%).

Figure 3 depicts the pathway (model fit; CFI = 0.99; NFI = 0.98; NNFI = 0.99) between (NT-proBNP)P Δ, (OT)P Δ, (corticosterone)P Δ, (AVP)P Δ, and plasma (Na+) Δ. Overall, (AVP)P Δ was associated with 10% of the variance seen in plasma (Na+) Δ, whereas (NT-proBNP)P Δ, (OT)P Δ, and (corticosterone)P Δ were associated with 47% of the variance seen in (AVP)P Δ. The overall pathway coefficient between (OT)P Δ and (AVP)P Δ was statistically significant; for every 1U increase in (OT)P Δ, (AVP)P Δ increased by 3 (unstandardized) units. This effect was almost 4 times stronger than the other effects on AVP.

Figure 3.

Figure 3

Open in a new tab

Overall pathway diagram showing a significant mathematically predicted association of OT Δ on AVP Δ (P < 0.05) for the overall cohort. Separate male (M) and female (F) coefficients are listed below the overall values. All reported values are standardized estimates, with unstandardized estimates in brackets. R2 percent of the variation is explained by the direct antecedent(s).

When males and females were analyzed separately, the pathway coefficient depicted in Fig. 3 between both (corticosterone)P Δ and (AVP)p Δ with plasma (Na+) Δ became statistically significant in the female cohort (path coefficient = 0.44 for females). This resulted in a much higher R2 value for plasma (Na+) Δ (∼7-fold higher) in females, compared with males. The overall pathway models remained statistically significant for the female cohort when analyzed separately (Figs. 2 and 3).

Figure 4 depicts the pathway (model fit: CFI = 1.0; NFI = 0.99; NNFI = 1.45) between (NT-proBNP)P Δ, (OT)P Δ, (corticosterone)P Δ, (AVP)P Δ, and plasma (Na+) postrace values for the female cohort only. (OT)pΔ and (NT-proBNP)p Δ accounted for 43% of the variance noted in (AVP)p Δ. The pathway coefficient between (OT)P Δ and (AVP)P Δ was statistically significant; where for every 1 unstandardized unit increase in (OT)P Δ, there was a corresponding 2.0 U increase in (AVP)p Δ. In the female cohort, (AVP)p Δ, (corticosterone)p Δ and (OT)p Δ was associated with 43% of the variance noted in postrace plasma (Na+). (The R2 for the entire cohort was 13%; 4% for males.) The statistically significant pathway coefficient between (OT)pΔ and postrace plasma (Na+) was roughly 3- to 6-fold higher than the nonsignificant pathway coefficients between (corticosterone)P Δ and postrace plasma (Na+) and between (AVP)p Δ and postrace plasma (Na+), respectively.

Figure 4.

Figure 4

Open in a new tab

Pathway diagram for the female cohort only (n = 24) showing a significant mathematically predicted association of OT Δ on plasma (Na+) postrace values and AVP Δ (P < 0.05). All reported values are standardized estimates, with unstandardized estimates in brackets. R2 percent of the variation is explained by the direct antecedent(s).

Multiple linear regression analysis revealed significant correlations between postrace urine (Na+) and (AVP)p Δ (R2 = 19.7; P < 0.001), urine (Na+) Δ with both (OT)P Δ and (aldosterone)P Δ (R2 = 14.1; P < 0.05 for both variables), and % plasma volume Δ with (aldosterone)P Δ (R2 = 4.9; P < 0.05).

Simple linear regression analysis revealed significant correlations between (AVP)P Δ with both plasma (Na+) Δ and (OT)P Δ (Fig. 5). However, there was wide variability (∼20 mmol/liter) in the response of plasma (Na+) Δ to an approximately 6 pg/ml (mean) change in (AVP)P Δ.

Figure 5.

Figure 5

Open in a new tab

Linear relationships between AVP Δ with both plasma (Na+) Δ and OT Δ. The boxed area indicates a 20 mmol/liter range of plasma (Na+) Δ values associated with a corresponding 6 pg/ml (mean) change in (AVP)P.

Discussion

Body fluid homeostasis during prolonged endurance exercise (356 ± 4 min) was well maintained in this cohort of experienced athletes with ad libitum food and fluid intakes. Plasma (Na+) was maintained from before the race (139.3 ± 0.3 mmol/liter) to after the race (138.1 ± 0.4 mmol/liter) despite significant decreases in body weight (−3.8 ± 0.1%, P < 0.01) and plasma volume (−8.5 ± 0.1%, P < 0.01). Urine (Na+) decreased, whereas urine osmolality increased before to after the race, indicating sodium and water retention, also contributing to the successful regulation of plasma (Na+) and osmolality during this period of heightened physical stress (Table 1). These findings suggest that the physiological mechanisms responsible for body fluid homeostasis primarily function primarily to preserve plasma (Na+) and osmolality.

The endocrine hormone classically responsible for the regulation of plasma osmolality and sodium concentration is AVP (13). Plasma levels of AVP were significantly elevated 3.9-fold at the end of the ultramarathon despite unchanged plasma (Na+). This finding confirms the presence of nonosmotic stimuli to AVP secretion during prolonged endurance exercise. Nonosmotic stimuli during exercise potentially include plasma volume contraction (14), nausea and/or vomiting (15), hypoglycemia (13), elevated body temperature (16), and elevated IL-6 concentrations (17). However, this should not be interpreted to mean that osmotic stimulation of AVP secretion does not also occur under these conditions, which is demonstrated by the positive correlation between plasma AVP concentrations and plasma (Na+) (Figs. 2 and 5).

Osmotic stimulation of AVP secretion during endurance exercise has been documented in other field studies, such as a study involving 16 elite runners completing a 42.2-km marathon in 164 min. In that study, a 4-fold increase in postrace (AVP)P was documented in conjunction with a plasma (Na+) increase of 6 mmol/liter, a body weight loss of 5%, and a decline in plasma volume of 12% (18). The differences between those findings and our results likely reflect greater evaporative fluid losses with less fluid replacement in more highly trained runners, compared with the more moderately trained but better hydrated runners who participated in our study.

The estimated 8.5 ± 0.1% plasma volume contraction from pre- to postrace values was sufficient to stimulate volume-sensing baroreceptors, as confirmed by the significant linear relationship between exercise-induced plasma volume contraction and (aldosterone)PΔ. The plasma volume contraction likely represents a combination of increased hydrostatic forces (19) with (sweat) sodium losses which accrued over approximately 6 h of running. The concomitant stimulation of AVP by baroreceptors may have inhibited osmotically induced suppression of AVP secretion, as documented in those runners demonstrating the greatest decrease in plasma (Na+) (Fig. 5). Therefore, it may be reasonable to speculate that nonosmotic volume-dependent stimulation of AVP secretion may be a pathophysiological factor contributing to the development of exercise-associated hyponatremia. Volume-dependent AVP stimulation may impair osmotic AVP suppression if hypotonicity and whole-body fluid overload occurred during marathon running. Although an inverse correlation between plasma volume and (AVP)P was not apparent from these data, the magnitude of plasma volume contraction may have been sufficient to stimulate AVP secretion but, unlike aldosterone secretion, this relationship was not dose dependent.

In addition to nonosmotic AVP stimulation via activation of baroreceptors during prolonged endurance exercise, pronounced elevations in (corticosterone)P, (NT-proBNP)P, and (OT)P in pathway analyses was associated with 47% of the variance noted in (AVP)P (Fig. 3). Confirmation that these three endocrine stimuli were statistically associated with almost half of the increase seen in (AVP)P suggests that other endocrine interactions may impact AVP secretion in the coordinated maintenance of plasma (Na+) during exercise. The statistical associations between (IL-6)P, (cortisol)P, and (NT-proBNP)P on (AVP)P were insignificant as individual relations but combined made a significant contribution in the overall mathematical pathway model predicted for the endocrine regulation of plasma (Na+) (Figs. 2 and 3).

Significant increases in (IL-6)P after a marathon foot race have been reported previously (20) because actively contracting muscles (21) but not circulating monocytes (22) produce IL-6 (21) for reasons that are physiologically unclear. Because IL-6 stimulates AVP production in nonexercising humans (17), IL-6 has been implicated in the pathogenesis of EAH as a variant of syndrome of inappropriate secretion of antidiuretic hormone (23). Our data do not support this hypothesis, however, because the statistical association between (IL-6)P and (AVP)P was minimal (and inverse); for every one-unit increase in (IL-6)P there was a predicted 0.03 (unstandardized) unit decrease in (AVP)P Δ (Fig. 2).

The 1.7-fold increase in (cortisol)P is a well-documented response to marathon running (24), secondary to stress-induced hypothalamic-pituitary axis stimulation to facilitate gluconeogenesis and counteract inflammation (25). Furthermore, a moderate inhibitory association between cortisol secretion and (AVP)P would be expected (26) and was statistically confirmed by the results of this study (Fig. 2).

The mild and inverse statistical association between the 4.5-fold increase in (NT-proBNP)P and (AVP)P Δ remained a curious, but not a novel, phenomenon (Table 1 and Fig. 3). Similar increases in (NT-pro-BNP)P have been documented after prolonged endurance exercise (27), whereas an inverse relationship between BNP and AVP has been previously verified in rats (28). The impressive elevation in the natriuretic peptide that was deemed most sensitive to fluid overload conditions (29) appears paradoxical, however, because plasma volume was significantly decreased after the ultramarathon (Table 1). Alternative stimuli to NT-proBNP from lipolysis (30), systemic inflammation (31) or other unknown factors that might occur during prolonged endurance exercise warrants further investigation.

The significant statistical association between the 3.3-fold increase in (corticosterone)P and (AVP)P was an intriguing finding because corticosterone has no specific physiological role in humans (32). The pathway coefficient between (corticosterone)P and (AVP)P was quite robust (0.74; P < 0.05) when the mathematical model included (IL-6)P and (cortisol)P (Fig. 2). However, when (NT-proBNP)P and (OT)P entered the hypothetical pathway (Fig. 3), the pathway coefficient between (corticosterone)P and (AVP)P diminished significantly (0.14; P = NS). Nonetheless, an apparent interrelationship among AVP, aldosterone, and corticosterone (aldosterone precursor) may cooperatively exist to regulate fluid homeostasis during exercise.

Plasma concentrations of OT were significantly elevated in runners immediately after completion of the Two Oceans 56-km race, although the increase was approximately half (1.9-fold) of the relative increase seen in AVP. This is the first study to document significant associations between (OT)P with plasma (Na+), urine (Na+), and (AVP)P in humans during exercise (Figs. 3–5). Osmotic stimulation of OT and its role in stimulating natriuresis and inhibiting sodium appetite has been well described in rats (33) but remains an equivocal finding in humans (34,35).

It is interesting to note that in all significant pathway analyses (reported and unreported), the percentage of variance (R2) accounted for by the various antecedents in each model was consistently higher for males, compared with females, with regard to (AVP)p. This sex predilection was reversed with regard to plasma (Na+), whereas the R2 for females was up to 7-fold higher than the percentage of variance noted for males (Fig. 3). When pathway analyses were performed for plasma (Na+) after the race, the significant pathway coefficient between (OT)p and postrace plasma (Na+) superseded the previously significant effects of both (corticosterone)p and (AVP)p on plasma (Na+) Δ for the female cohort only (Fig. 3). Furthermore, there was a statistically significant mathematical influence of (OT)p on both (AVP)p and postrace plasma (Na+) noted in females (Fig. 4) but not males (data not shown). The potential influence of oxytocin on both AVP and plasma (Na+) in females during prolonged endurance exercise is intriguing because female sex is a risk factor in the development of hyponatremic encephalopathy (7,36), and in some studies, the relative risk of death or permanent neurological dysfunction in hyponatremic patients is approximately 30 times higher in females, compared with males (37).

These pathway diagrams were designed to flow in only one direction. However, the relationship between (AVP)P and plasma (Na+) should be represented as a loop in which associations between cause and effect are expected to be circular and continuous. The interrelationship between stimulus and response was supported by the positive relationship between postrace urine (Na+) and (AVP)P Δ in regression analysis (R2 = 19.7; P < 0.001). The mathematical pathway models were not significant, however, when the direction of the entire path was reversed. This would further support the hypothesis that nonosmotic stimuli to AVP secretion may occur normally during prolonged endurance running, mainly from stimulation of volume-sensing baroreceptors during exercise combined with potential interactions from other endocrine factors concomitantly stimulated during marathon running.

Conclusion

Plasma AVP concentrations were markedly elevated after a 56-km ultramarathon despite unchanged plasma (Na+). Mathematical pathway modeling suggests that nonosmotic stimulation of (AVP)P, most likely from decreased plasma volume with potential influence from BNP, oxytocin, and corticosterone, may contribute to 47% of the increase observed in postrace (AVP)P. Therefore, it would seem reasonable to predict that an inability to maximally suppress (AVP)P during exercise as a result of nonosmotic stimulation of AVP secretion could potentially contribute to the pathogenesis of EAH under conditions in which voluntary fluid intake exceeds urinary and sweat water losses.

Acknowledgments

The authors thank Willem Meeuwisse, Steve Reid, Carlos Ayus, Louise Weschler, Jonathan Dugas, Christopher Almond, Yoshihisa Sugimura, Ying Tian, Lara Dugas, Ross Tucker, Bianca Meurer, Judy Belonje, Liane Berretta, George Mokone, Robert Lamberts, Nelleke Langerak, and Bill Butler for their technical support and Artisha Polk and Helaine Resnick for their preliminary statistical advice. Special thanks go to our 82 Two Oceans Runners who enthusiastically supported this project in selfless support of the safety of other runners.

Footnotes

This work was supported by a research grant from Astellas Pharma US, Inc. (Deerfield, IL) and Grant M01RR-023942-01 from the National Center for Research Resources (NCRR), a component of the National Institutes of Health (NIH). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of NCRR or NIH. Secondary financial assistance was received from Gettysburg College and the Medical Research Council/University of Cape Town Research Unit for Exercise Science and Sports Medicine.

First Published Online March 18, 2008

Abbreviations: AVP, Arginine vasopressin; (AVP)P, plasma level of AVP; BNP, brain natriuretic peptide; CFI, comparative fit index; EAH, exercise-associated hyponatremia; (Na+), urine sodium; NFI, normed fit index; NNFI, nonnormed fit index; (NT-pro-BNP)P, plasma level of N-terminal pro-BNP; OT, oxytocin; (OT)P, plasma level of oxytocin.

References

  1. Wade CE, Dressendorfer RH, O'Brien JC, Claybaugh JR 1981 Renal function, aldosterone, and vasopressin excretion following repeated long-distance running. J Appl Physiol 50:709–712 [DOI] [PubMed] [Google Scholar]
  2. Gastmann U, Dimeo F, Huonker M, Bocker J, Steinacker JM, Peterson KG, Wieland H, Keul J and Lehmann M 1998 Ultra-triathlon-related blood-chemical and endocrinological responses in nine athletes. J Sports Med Phys Fitness 38:18–23 [PubMed] [Google Scholar]
  3. Dessypris A, Wagar G, Fyhrquist F, Makinen T, Welin MG, Lamberg BA 1980 Marathon run: effects on blood cortisol—ACTH, iodothyronines—TSH and vasopressin. Acta Endocrinol (Copenh) 95:151–157 [DOI] [PubMed] [Google Scholar]
  4. Mudambo KS, Coutie W, Rennie MJ 1997 Plasma arginine vasopressin, atrial natriuretic peptide and brain natriuretic peptide responses to long-term field training in the heat: effects of fluid ingestion and acclimatization. Eur J Appl Physiol Occup Physiol 75:219–225 [DOI] [PubMed] [Google Scholar]
  5. Nelson PB, Ellis D, Fu F, Bloom MD, O'Malley J 1989 Fluid and electrolyte balance during a cool weather marathon. Am J Sports Med 17:770–772 [DOI] [PubMed] [Google Scholar]
  6. Freund BJ, Claybaugh JR, Hashiro GM, Buono M, Chrisney S 1990 Exaggerated ANF response to exercise in middle-aged vs. young runners. J Appl Physiol 69:1607–1614 [DOI] [PubMed] [Google Scholar]
  7. Hew-Butler TD, Almond CS, Ayus JC, Dugas J, Meeuwisse W, Noakes T, Reid S, Siegel A, Speedy D, Stuempfle K, Verbalis J, Weschler L 2005 Consensus Statement of the 1st International Exercise-Associated Hyponatremia Consensus Development Conference, Cape Town, South Africa, 2005. Clin J Sport Med 15:208–213 [DOI] [PubMed] [Google Scholar]
  8. Speedy DB, Campbell R, Mulligan G, Robinson DJ, Walker C, Gallagher P, Arts JH 1997 Weight changes and serum sodium concentrations after an ultradistance multisport triathlon. Clin J Sport Med 7:100–103 [DOI] [PubMed] [Google Scholar]
  9. Siegel AJ, Verbalis JG, Clement S, Mendelson JH, Mello NK, Adner M, Shirey T, Glowacki J, Lee-Lewandrowski E, Lewandrowski KB 2007 Hyponatremia in marathon runners due to inappropriate arginine vasopressin secretion. Am J Med 120:461–467 [DOI] [PubMed] [Google Scholar]
  10. Verbalis JG, McHale CM, Gardiner TW, Stricker EM 1986 Oxytocin and vasopressin secretion in response to stimuli producing learned taste aversion in rats. Behav Neurosci 100:466–475 [DOI] [PubMed] [Google Scholar]
  11. Guo T, Taylor RL, Singh RJ, Soldin SJ 2006 Simultaneous determination of 12 steroids by isotope dilution liquid chromatography-photospray ionization tandem mass spectrometry. Clin Chim Acta 372:76–82 [DOI] [PubMed] [Google Scholar]
  12. Soldin SJ, Soldin OP, Boyajian AJ, Taskier MS 2006 Pediatric brain natriuretic peptide and N-terminal pro-brain natriuretic peptide reference intervals. Clin Chim Acta 366:304–308 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Verbalis JG 2003 Disorders of body water homeostasis. Best Pract Res Clin Endocrinol Metab 17:471–503 [DOI] [PubMed] [Google Scholar]
  14. Robertson GL 1995 Posterior pituitary. In: Felig P, Frohman LA, eds. Endocrinology and metabolism. New York: McGraw-Hill, 385–432 [Google Scholar]
  15. Rowe JW, Shelton RL, Helderman JH, Vestal RE, Robertson GL 1979 Influence of the emetic reflex on vasopressin release in man. Kidney Int 16:729–735 [DOI] [PubMed] [Google Scholar]
  16. Takamata A, Mack GW, Stachenfeld NS, Nadel ER 1995 Body temperature modification of osmotically induced vasopressin secretion and thirst in humans. Am J Physiol 269(4 Pt 2):R874–R880 [DOI] [PubMed] [Google Scholar]
  17. Gionis D, Ilias I, Moustaki M, Mantzos E, Papadatos I, Koutras DA, Mastorakos G 2003 Hypothalamic-pituitary-adrenal axis and interleukin-6 activity in children with head trauma and syndrome of inappropriate secretion of antidiuretic hormone. J Pediatr Endocrinol Metab 16:49–54 [DOI] [PubMed] [Google Scholar]
  18. Rocker L, Kirsch KA, Heyduck B, Altenkirch HU 1989 Influence of prolonged physical exercise on plasma volume, plasma proteins, electrolytes, and fluid-regulating hormones. Int J Sports Med 10:270–274 [DOI] [PubMed] [Google Scholar]
  19. Myhre LG, Hartung GH, Nunneley SA, Tucker DM 1985 Plasma volume changes in middle-aged male and female subjects during marathon running. J Appl Physiol 59:559–563 [DOI] [PubMed] [Google Scholar]
  20. Ostrowski K, Schjerling P, Pedersen BK 2000 Physical activity and plasma interleukin-6 in humans—effect of intensity of exercise. Eur J Appl Physiol 83:512–515 [DOI] [PubMed] [Google Scholar]
  21. Steensberg A, van Hall G, Osada T, Sacchetti M, Saltin B, Klarlund PB 2000 Production of interleukin-6 in contracting human skeletal muscles can account for the exercise-induced increase in plasma interleukin-6. J Physiol 529(Pt 1):237–242 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Starkie RL, Rolland J, Angus DJ, Anderson MJ, Febbraio MA 2001 Circulating monocytes are not the source of elevations in plasma IL-6 and TNF-α levels after prolonged running. Am J Physiol Cell Physiol 280:C769–C774 [DOI] [PubMed] [Google Scholar]
  23. Siegel AJ 2006 Exercise-associated hyponatremia: role of cytokines. Am J Med 119(7 Suppl 1):S74–S78 [DOI] [PubMed] [Google Scholar]
  24. Maron MB, Horvath SM, Wilkerson JE 1975 Acute blood biochemical alterations in response to marathon running. Eur J Appl Physiol Occup Physiol 34:173–181 [DOI] [PubMed] [Google Scholar]
  25. Duclos M, Corcuff JB, Pehourcq F, Tabarin A 2001 Decreased pituitary sensitivity to glucocorticoids in endurance-trained men. Eur J Endocrinol 144:363–368 [DOI] [PubMed] [Google Scholar]
  26. Dingman JF, Despointes RH 1960 Adrenal steroid inhibition of vasopressin release from the neurohypophysis of normal subjects and patients with Addison’s disease. J Clin Invest 39:1851–1863 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Scharhag J, Herrmann M, Urhausen A, Haschke M, Herrmann W, Kindermann W 2005 Independent elevations of N-terminal pro-brain natriuretic peptide and cardiac troponins in endurance athletes after prolonged strenuous exercise. Am Heart J 150:1128–1134 [DOI] [PubMed] [Google Scholar]
  28. Makino S, Hashimoto K, Hirasawa R, Hattori T, Ota Z 1992 Central interaction between endothelin and brain natriuretic peptide on vasopressin secretion. J Hypertens 10:25–28 [DOI] [PubMed] [Google Scholar]
  29. Huang WS, Lee MS, Perng HW, Yang SP, Kuo SW, Chang HD 2002 Circulating brain natriuretic peptide values in healthy men before and after exercise. Metabolism 51:1423–1426 [DOI] [PubMed] [Google Scholar]
  30. Moro C, Polak J, Hejnova J, Klimcakova E, Crampes F, Stich V, Lafontan M, Berlan M 2006 Atrial natriuretic peptide stimulates lipid mobilization during repeated bouts of endurance exercise. Am J Physiol Endocrinol Metab 290:E864–E869 [DOI] [PubMed] [Google Scholar]
  31. Shor R, Rozenman Y, Bolshinsky A, Harpaz D, Tilis Y, Matas Z, Fux A, Boaz M, Halabe A 2006 BNP in septic patients without systolic myocardial dysfunction. Eur J Intern Med 17:536–540 [DOI] [PubMed] [Google Scholar]
  32. Raubenheimer PJ, Young EA, Andrew R, Seckl JR 2006 The role of corticosterone in human hypothalamic-pituitary-adrenal axis feedback. Clin Endocrinol (Oxf) 65:22–26 [DOI] [PubMed] [Google Scholar]
  33. Verbalis JG, Mangione MP, Stricker EM 1991 Oxytocin produces natriuresis in rats at physiological plasma concentrations. Endocrinology 128:1317–1322 [DOI] [PubMed] [Google Scholar]
  34. Williams TD, Abel DC, King CM, Jelley RY, Lightman SL 1986 Vasopressin and oxytocin responses to acute and chronic osmotic stimuli in man. J Endocrinol 108:163–168 [DOI] [PubMed] [Google Scholar]
  35. Kostoglou-Athanassiou I, Treacher DF, Forsling ML 1994 Is oxytocin natriuretic in man? J Endocrinol 143(Suppl O):39 [Google Scholar]
  36. Moritz ML, Ayus JC 2007 Hospital-acquired hyponatremia—why are hypotonic parenteral fluids still being used? Nat Clin Pract Nephrol 3:374–382 [DOI] [PubMed] [Google Scholar]
  37. Ayus JC, Wheeler JM, Arieff AI 1992 Postoperative hyponatremic encephalopathy in menstruant women. Ann Intern Med 117:891–897 [DOI] [PubMed] [Google Scholar]

Articles from The Journal of Clinical Endocrinology and Metabolism are provided here courtesy of The Endocrine Society

RESOURCES