Fructose: A New Variable to Consider in SIADH and the Hyponatremia Associated With Long-Distance Running?

Abstract

Fructose has recently been proposed to stimulate vasopressin secretion in humans. Fructose-induced vasopressin secretion is not only postulated to result from ingestion of fructose-containing drinks, but may also occur from endogenous fructose production via activation of the polyol pathway. This raises the question of whether fructose might be involved in some cases of vasopressin-induced hyponatremia, especially in situations where the cause is not fully known, such as in the syndrome of inappropriate secretion of diuretic hormone (SIADH) and exercise-associated hyponatremia that has been observed in marathon runners. Here we discuss the new science of fructose and vasopressin, and how it may play a role in some of these conditions, as well as in the complications associated with rapid treatment (such as the osmotic demyelination syndrome). Studies to test the role of fructose could provide new pathophysiologic insights as well as novel potential treatment strategies for these common conditions.

Nephrologists are well aware of the role of hyperosmolarity, low extracellular volume, or low effective arterial volume, in stimulating vasopressin production, yet only recently has it been appreciated that another stimulus is fructose. Herein we review these recent findings and discuss its potential relevance to several common clinical disorders.

Fructose as a Stimulus for Vasopressin Secretion

The unique ability of fructose to increase vasopressin levels in the blood was first shown by Wolf et al. thirty years ago in which an intravenous infusion of hypertonic fructose in humans stimulated an acute rise in vasopressin that was not observed with hypertonic glucose¹. Subsequently it was shown that oral administration of fructose could also cause a rise in vasopressin concentrations in both experimental animals² and humans^3,4, that it could potentiate the effect of heat and dehydration³, and that it can occur rapidly (as early as 15 minutes)⁴. These studies raise the possibility that drinks containing sugar (sucrose) or high fructose corn syrup (HFCS), both which contain fructose, could be stimuli that might increase serum vasopressin levels.

Additional studies demonstrated that endogenously produced fructose can also stimulate vasopressin production (Figure 1). The primary way fructose is made in mammals is from a series of chemical reactions known as the polyol pathway in which glucose is first converted to sorbitol by aldose reductase (AR), and then to fructose by sorbitol dehydrogenase (SDH). The rate-limiting enzyme in the reaction is AR, and it can be stimulated by hyperosmolality, ischemia, and other mechanisms^5–7. Interestingly we found that acute dehydration-associated hyperosmolarity in rats was associated with activation of AR in the hypothalamus, leading to local fructose production and an increase vasopressin mRNA in the supraoptic nucleus.⁸ When acute dehydration was performed in mice who were unable to metabolize fructose (the fructokinase knockout mice, KHK KO), this increase in vasopressin mRNA was completely attenuated, and there was also reduced secretion of vasopressin from the pituitary and lower serum copeptin concentrations (a stable analogue of vasopressin).⁸ Although fructose metabolism appeared to have only a relatively modest effect on serum copeptin concentrations with acute dehydration, in chronically dehydrated mice the rise in serum copeptin levels was completely blocked in KHK KO mice.⁸ This suggests that endogenous generation and metabolism of fructose may be important in the vasopressin response to acute and chronic dehydration in the mouse. However, whether this is also true for humans is not known.

Figure 1. Fructose Stimulates Vasopressin Synthesis.

Both dietary and endogenously produced fructose can stimulate the synthesis and release of vasopressin (straight arrow). In turn, vasopressin can induce an upregulation of the polyol (aldose reductase)-fructokinase pathway in the kidney (curved arrows). In turn, the metabolism of fructose in the proximal tubule can lead to tubular dysfunction with an increased excretion of urate, sodium and phosphates. Created with BioRender.com

These latter studies suggest that dehydration-associated hyperosmolality stimulates endogenous fructose production that then stimulates vasopressin synthesis. This finding naturally raises the question of whether dietary fructose might also stimulate vasopressin secretion, especially since sugary beverages often have high osmolality. Indeed, the study by Wolf et al administered a hypertonic solution of fructose.¹ Additionally, the two human studies that reported an increase in serum copeptin concentrations with fructose-containing beverages administered either apple juice or soft drinks, both of which were also hypertonic.^3,4 In fact, in all three studies there was an increase in plasma osmolality. While this could be due to the tonicity of the fructose-containing solutions, orally administered hypertonic solutions, especially those that primarily contain fructose, may cause some intestinal sequestration of fluids that can lead to water loss from the extracellular and intracellular compartments with a secondary rise in plasma osmolality⁹. In this case, one might argue that soft drinks and fructose are simply working via the known mechanism for hyperosmolality to stimulate vasopressin secretion.

However, there is evidence from animal studies that the stimulation of vasopressin by fructose may occur independently of changes in plasma osmolality. For example, while an increase in vasopressin secretion is observed ex vivo when hypothalamic slices are incubated with hypertonic fructose from healthy mice, vasopressin secretion is not observed when hypertonic concentrations of fructose are incubated with hypothalamic slices from KHK-knockout mice that are unable to metabolize fructose⁸. This finding suggests that it is the metabolism of fructose, and not its osmolality, that stimulates vasopressin release. Likewise, in the study by Wolf et al, the infusion of hypertonic fructose was associated with an acute mild increase in serum vasopressin levels at 15 minutes along with a rise in plasma osmolality. Plasma osmolality and serum vasopressin levels then normalized. However, at one hour a dramatic rise in serum vasopressin occurred despite no change in plasma osmolality. The rise at one hour would be consistent with an effect related to fructose metabolism, as blood levels of fructose were normal at that time.

Nevertheless, the observation that oral fructose ingestion is associated acutely with an increase in plasma osmolality suggests that osmolality may also be important in stimulating the fructose response. One possible explanation, as noted above, is that the rise in plasma osmolality causes fluid sequestration in the intestine, leading to a fall in plasma volume^3,9,10 However, studies in rats also preliminarily suggest that the rise in plasma osmolality observed with fructose may represent an acute shift of water into the intracellular space^11,12. Indeed, we believe that fructose-induced enhancement of copeptin in heat-stressed rats was associated with an acute increase in intracellular water and a fall in extracellular/plasma volume by bioimpedance^11,12, which we postulate may result from the rapid fructose-dependent accumulation of glycogen that has a strong avidity to adsorb water. Clearly, more studies are needed to better understand the role of osmotic versus non-osmotic fructose-dependent stimulation of vasopressin.

The possible stimulation of fructose by vasopressin may shed new light on the function of vasopressin in disease. Studies found that the effect of fructose to induce obesity and metabolic syndrome was dependent on vasopressin and is mediated by engagement of the V1b receptor². The teleological reason appears to be that fat is a source of metabolic water, and stimulation of fat production is yet another way vasopressin can help prevent dehydration¹³. This observation explains why serum vasopressin (copeptin) levels are commonly elevated in participants with obesity^14,15. It also explains the association of obesity with higher serum osmolality, more signs of dehydration, and reduced water yet higher salt (and sugar) intake^16,17. Thus, we speculate that vasopressin is not just a water hormone, it is a fat hormone.

Fructose and SIADH

The syndrome of inappropriate secretion of diuretic hormone (SIADH) is a type of hyponatremia with elevated vasopressin levels associated with a variety of conditions (Box 1) and for which the cause is likely multifactorial. A similar condition is hyponatremia associated with renal salt wasting (RSW)^18,19. Controversy currently exists as to how frequent RSW occurs, whether it reflects part of the spectrum of SIADH in which salt wasting is more severe, or whether it is its own clinical entity^20–23.

Box 1. Typical Causes of SIADH.

• Cerebral Disease: Infections, Cancer, Strokes, Bleeds

• Pulmonary Disease:

• Other Cancers

• Chronic liver disease

• Post-surgical

• Marathon Runners

• Drugs

• Unknown

Could fructose potentially have a role in these conditions? Studies suggest that while fructose may stimulate vasopressin secretion and urinary concentration, that it may increase serum sodium acutely as opposed to lowering it^4,24. As mentioned,, we speculate that this might be due to a shift of fluid into the intracellular space due to the rapid synthesis of glycogen that takes up water (Figure 2)^11,12. Once glycogen stores are full, then the increased retention of water from persistent vasopressin secretion would be expected to cause hyponatremia in the setting of continued water intake. Therefore, one might see a transition from a low extracellular volume state to one in which the patient may appear euvolemic. Moreover, it is possible that fructose could play a role in either SIADH and/or RSW.

Figure 2. Fructose Intake may Mimic Renal Salt Wasting by Increasing Water Reabsorption while Maintaining a Low Extracellular Volume.

We found that the administration of fructose (such as in a soft drink) to laboratory rats following heat stress is associated with remarkable stimulation of vasopressin (noted by serum copeptin) compared to intake of water¹². The increased vasopressin is associated with an amplification in urinary concentration compared to water hydration. However, serum osmolality fails to decrease because the water retained through the action of vasopressin shifts into the cell, resulting in a persistently contracted extracellular volume with increased intracellular volume (measured by bioimpedance, and reported in¹¹). This is likely because glycogen is being rapidly made, and glycogen incorporates water into its lattice structure. Reprinted from Johnson et al¹¹ with permission of the copyright holder (John Wiley & Sons); original graphic © 2019 the Association for the Publication of the Journal of Internal Medicine.

Unlike true extracellular volume depletion in which serum uric acid is typically elevated, in SIADH and RSW the serum uric acid is typically low (<4mg/dL) due to an increased fractional excretion of urate (FEurate typically >10%) in the urine²⁵. The mechanism for the hypouricemia is not well understood, as acute administration of vasopressin does not increase urinary uric acid excretion, but rather reduces it^26,27. However, the classic explanation for why SIADH is associated with hypouricemia is that SIADH may reflect a state of volume expansion from increased total body water, and that this may result in increased uric acid excretion. However, acute volume expansion with either isotonic or hypotonic fluids has only a modest increase in fractional excretion of uric acid and does not significantly reduce serum uric acid levels²⁸.

A more likely explanation for the hypouricemia in SIADH has recently been identified. There is evidence that the increased uric acid excretion in SIADH and RSW is associated with tubular injury, as biomarkers for kidney damage such as urinary neutrophil gelatinase-associated lipocalcin (NGAL) are elevated in the urine of subjects with SIADH and RSW²⁹. Given that uric acid reabsorption and excretion is almost exclusively via the proximal tubule, it also suggests the proximal tubule as the site of injury. This is supported by the finding that elevated fractional excretion of phosphate can be seen in RSW and possibly SIADH, and rare cases of Fanconi syndrome have also been reported^29,30. Likewise, studies of RSW have found that the site of increased sodium excretion is also the proximal tubule (based on endogenous lithium clearance)³¹. One study suggests that the natriuresis might be due to a specific protein (haptoglobin related protein without signal peptide) that is elevated in the blood of subjects with RSW³².

Interestingly, the proximal tubule is a major site in the kidney where fructose is metabolized, and this can lead to tubular injury³³. While low levels of fructose metabolism may lead to salt retention and hypertension³⁴, higher levels can lead to a partial Fanconi-like syndrome. One example is in hereditary fructose intolerance (HFI) which has been shown to be driven by enhanced activity of the fructokinase pathway.³⁵ Participants with HFI show markedly increased urinary urate excretion, along with increased phosphate and sodium excretion; these are all enhanced in response to fructose.^36,37

While isotonic volume expansion does not lead to significant uricosuria in people²⁸, there are reports that large infusions of hypertonic saline can induce marked uricosuria with high fractional excretion of uric acid and sodium³⁸. Hypertonic saline might be expected to stimulate aldose reductase and fructokinase in the proximal tubule, similar to what has been shown with hypertonic radiocontrast³⁹ and hence would support fructose metabolism as the driver of the uricosuric response.

While it is too early to know if fructose is involved in SIADH, it is interesting that one study found high urinary fructose levels in subjects with SIADH and RSW²⁹. We suggest a potential pathway for how fructose could mediate some cases of SIADH in Figure 1.

Fructose and Marathon Runner’s Hyponatremia

Hyponatremia can occur in long distance runners, and rarely it can be fatal.⁴⁰ It is thought that excessive sweating with systemic salt depletion triggers vasopressin release, followed by the ingestion of large amounts of hypotonic fluids that causes the serum sodium to fall⁴¹. Subjects who develop hyponatremia following marathon runners often gain several kilograms over the marathon, consistent with large fluid intake and retention.

While a decrease in plasma or extracellular volume may be an initial trigger for the stimulation of vasopressin in participants with marathon hyponatremia, there are well described cases in which the administration of isotonic saline led to a lowering of serum sodium levels⁴¹. This inability to correct serum sodium with isotonic saline resuscitation suggests that extracellular volume depletion cannot fully account for the hyponatremia. Therefore, there may be other stimuli driving vasopressin levels besides increased osmolarity or extracellular volume depletion, suggesting some similarity to SIADH.

Fructose is commonly present in sports drinks and soft drinks. There is evidence that the vasopressin response to dehydration is significantly enhanced when fructose is administered, both in animals^12,24 and humans³. Because fructose may be initially associated with a reduction in the extracellular fluid with expansion of the intracellular fluid (Figure 2), it may initially stimulate more thirst that might lead to increased fluid intake. Sugary beverages may also encourage increased fluid intake due to excess sweetness. In addition, a fructose-dependent nonosmotic stimulation of vasopressin could explain why isotonic saline may not correct the hyponatremia.⁴¹ In addition, while fructose would stimulate glycogen accumulation, the effect of exercise is to burn the glycogen, and once glycogen is depleted, the water sequestered by the glycogen would be released and vasopressin levels would still be high, and so the serum sodium might be expected to fall rapidly. Thus, the effects of fructose might be delayed, increasing its risk with the duration of the exercise. Indeed, the risk for hyponatremia appears to be much greater in longer endurance runs such as the ultramarathons⁴².

Osmotic Demyelination Syndrome

The treatment of SIADH often requires the administration of hypertonic saline, which has to be done carefully as rapid correction can result in the osmotic demyelination syndrome (ODS). While ODS may have many manifestations, central pontine myelinolysis (CPM) is especially common. Patients most at risk are those with alcoholism, malnutrition, liver disease and liver transplant recipients, as well as individuals with cancer⁴³. ODS can also occur in other conditions associated with an acute increase in plasma osmolality such as diabetic ketoacidosis and hypernatremia. While CPM is most commonly associated with ODS, CPM can also occur in alcoholics even in the absence of hyponatremia⁴⁴. CPM can also occur in pregnancy (hyperemesis gravidarum) in which it has been linked with thiamine deficiency (Box 2).

Box 2. Osmotic Demyelination Syndrome.

• Associations with Osmolar Shifts or Hyperosmolarity

• Rapid correction of Hyponatremia

• Rapid Onset Hypernatremia

• Salt Intoxication⁵⁷

• Following Acute Kidney Injury⁵⁸

• Associated with hyperglycemia and hepatic encephalopathy⁵⁹

• Poorly Controlled Diabetes

  • Hyperosmolar Hyperglycemia^60,61
  • Associated with Diabetic Ketoacidosis⁶² with hypokalemia⁶³
  • Soft Drink Binge in Diabetic with NAFLD⁵⁶ Acute Ethylene Glycol Intoxication with Hyperosmolarity⁶⁴
• Possible Thiamine Deficiency Syndromes

  • Alcoholism in absence of Hyponatremia⁴⁴
  • Associated with hyperemesis gravidarum^65–67
• Other Associations with Low Intracellular Energy

  • Hypophosphatemia and Refeeding Syndrome⁶⁸

The mechanisms underlying osmotic demyelination are currently not well understood, but are thought to result from a rapid increase in serum osmolarity that causes an acute shrinkage of neurons in the brain leading to cell death. The initial cell targeted appears to be the astrocyte^45,46, which has a major supportive role for neighboring neurons⁴⁷ and oligodendrocytes (that produce myelin)⁴⁸. This is because the astrocyte provides both glucose (from the breakdown of glycogen) and lactate (generated from glycolysis) to the neighboring cells which can be converted to pyruvate and then acetyl CoA that can enter the citric cycle and eventually stimulate mitochondrial oxidative phosphorylation.

Interestingly, the astrocyte can express both aldose reductase⁴⁹ and fructokinase⁵⁰, and fructokinase activity is high throughout the brain.⁵¹ Fructose production increases markedly in the brains of humans when glucose infusions are administered to maintain a serum glucose of approximately 220 mg/dl.⁵² While studies in the brain are lacking, in diabetic mice in which aldose reductase and fructokinase have been induced in the kidney, the acute administration of a high osmotic load (radiocontrast) results in an acute burst of oxidative stress associated with tubular injury and inflammation.³⁹ One of the key mechanisms appears to be mitochondrial oxidative stress with a suppression of ATP production¹¹ (Figure 3). While the effects of sudden increases in brain fructose levels from acute increases in serum osmolality is not known, fructose is toxic to neurons in culture as it inhibits mitochondrial oxidative phosphorylation⁵³, while cultured astrocytes are also sensitive to the effects of fructose.⁵⁴ Indeed, if fructose is injected into the hypothalamus, there is an acute drop in intracellular ATP levels that if severe could lead to neuron death.⁵³ Thus we suggest that the administration of hypertonic saline to correct severe hyponatremia could potentially have an adverse side effect of rapidly increasing fructose in the brain, thereby triggering oxidative stress and ATP depletion that could cause demyelination and neuronal loss.

Figure 3. Proposed Pathogenesis of Osmotic Demyelination Syndrome.

An acute increase in osmolarity will activate the polyol-fructose pathway, resulting in the endogenous production and rapid metabolism of fructose in astrocytes, leading to oxidative stress and mitochondrial injury that may lead to death of the astrocyte. In so doing, nutritional support to the neighboring oligodendrocyte is affected, leading to demyelination.

A mechanism involving ATP depletion might also explain why alcoholism, and those with coexistent thiamine deficiency, are at risk for CPM, as thiamine pyrophosphate has a critical role in the conversion of pyruvate to acetyl CoA, and so thiamine deficiency results in acute ATP depletion and lactate generation⁵⁵. This might also explain the rare observation that marked intake of sugary beverages can also precipitate CPM⁵⁶.

Limitations and Summary

The potential role of fructose in hyponatremia remains unknown, but the evidence is suggestive enough to warrant investigation of its role both in the etiology and complications of hyponatremia. At this stage there are many unanswered questions, and it seems likely that if fructose is involved, that it is would have a role in only a subset of patients, or perhaps is more of a contributory factor than a causal factor. Nevertheless, it is a new area of investigation, and if it does have a role in these disorders, the new insights into the pathophysiology of hyponatremia could lead to changes in management and treatment. Given what we know at this stage, we suggest considering restricting not just fluid intake, but specifically restricting soft drinks in subjects with presumed SIADH, and consider thiamine as part of the management plan. Most importantly, we recommend more studies. Could fructose have a role in SIADH? Preposterous, one might say, but perhaps just possible!