How Does Dietary Potassium Influence BP?

Abstract

Hypertension is a global burden and a major contributor to cardiovascular disease and premature death. Public health guidelines for the management of high BP contain numerous dietary recommendations, of which one is a reduction in salt (NaCl) intake. In addition, the modern diet is also characterized by low potassium (K⁺) content and recent guidelines propose increasing K⁺ intake as an alternative or complementary measure to reducing salt intake for lowering of BP. Most beneficial effects of K⁺ supplementation on BP involve a homeostatic response of the kidney to dietary-induced changes in extracellular K⁺ concentrations, particularly decreased reabsorption of NaCl in the distal convoluted tubule. However, the effects of greater K⁺ intake on BP are not linear, and the ideal K⁺ supplementation or intake for management of BP remains unclear. This article covers the mechanisms in the kidney by which changes in K⁺ translate to alterations in BP, the effects of altered K⁺ intake in animal models and human populations, and the importance of concurrent salt intake and what constitutes K⁺ supplementation.

Introduction

Hypertension is a global burden. Public health guidelines for the prevention and management of hypertension, such as those from the World Health Organization (WHO), emphasize general lifestyle changes and particularly dietary interventions, such as reducing sodium (Na⁺) intake.¹ In 2016, the WHO released the Surveillance, Harness industry, Adopt standards for labeling and marketing, Knowledge, Environment package aimed at reducing Na⁺ intake by 30% by 2025.² However, the midterm evaluation showed a minimal effect of the initiatives,^3,4 and complementary dietary strategies are necessary for effective BP management. Recent guidelines emphasize greater potassium (K⁺) intake, or a higher K⁺ to Na⁺ ratio, as an approach to lower BP.^5,6 In paleolithic times, diets were composed primarily of K⁺ rich foods⁷ that provided around 30 mmol/d (690 mg) of Na⁺ and 280 mmol/d (11 g) of K⁺, with a K⁺ to Na⁺ ratio of 16 to 1.⁸ By contrast, in industrialized societies, current average daily intakes are estimated to be approximately 11 g of NaCl and approximately 2 g of K⁺, well off the WHO recommendations of <5 g of NaCl and at least 3.5 g of K⁺ per day in adults (70–80 mmol/d).^1,9 As in the usual diet, most Na⁺ is consumed as NaCl, and as high Na⁺ intake provided as nonchloride salts has not been linked to an increase in BP, this review focuses on Na⁺ from NaCl.^10–12

With a modern diet, the population attributable risk percentages for hypertension are similar for high Na⁺ intake and for low K⁺ intake.¹³ Supporting this, multiple animal and human studies suggest that a higher K⁺ intake, or a higher K⁺-to-Na⁺ ratio, lowers BP or limits BP increases.^14–17 However, recommendations for K⁺ supplementation are often incomplete and inconsistent.³ It is also unclear if K⁺ alone has beneficial effects on BP. What is the importance of concurrent Na⁺ intake: the absolute amount of Na⁺ and K⁺ or their ratio? In addition, K⁺ supplementation appears to have different effects on BP depending on existing hypertensive status. Recently, the American Society of Nephrology published a new health guidance on potassium and phosphorus food additives¹⁸ that underlines the risks of electrolyte alteration for people with CKD. We point out that patients with CKD require individual dietary guidelines that in general focus on overall healthy eating patterns. Furthermore, although K⁺-induced relaxation of vascular smooth muscle cells and vasodilation^19,20 may influence BP, and a role for the gut in modulating extracellular fluid (ECF) K⁺ is established,^19,21–24 it is not discussed here.

Mechanisms of Kidney K⁺ Handling and BP Control

The kidney is responsible for 90% of K⁺ excretion,²⁵ helping maintain ECF K⁺ concentrations within 3.5–5 mmol/L. Our daily intake of K⁺ is approximately equal to the total amount of ECF K⁺, and most ingested K⁺ is sequestered in skeletal muscle.²⁶ The remaining K⁺ is freely filtered by the glomeruli.²⁷ In total, 50%–60% of this filtered K⁺ is reabsorbed in the proximal tubule,²⁸ primarily through the paracellular pathway driven by net fluid reabsorption across the proximal tubule and solvent drag. The K⁺ that enters the thin descending limbs of the loop of Henle is largely conserved in the inner medulla through counter-current exchange with the thin ascending limbs.^29,30 A further approximately 30% of the initial K⁺ filtrate is reabsorbed along with Na⁺ in the thick ascending limb through the Na-K-2Cl cotransporter NKCC2 and renal outer medullary potassium channel (ROMK) feedback loop.³¹ The importance of this for K⁺ and Na⁺ homeostasis is illustrated by the polyuria, K⁺ wasting, and activation of the renin-angiotensin-aldosterone system observed in Bartter syndrome type 1 due to mutations in the NKCC2 gene SLC12A1^32,33 or in Bartter syndrome type 2 due to mutations of the ROMK gene KCNJ1.^34,35

The majority of K⁺ secretion occurs in the late distal convoluted tubule (DCT), connecting tubule, and collecting duct through ROMK and big potassium channel (BK) channels (also known as large conductance calcium-activated potassium channels, or Maxi-K).^36,37 In the early DCT, electroneutral Na⁺ reabsorption occurs via the sodium-chloride cotransporter NCC,³⁸ activity of which determines delivery of Na⁺ to K⁺ secretory tubule segments. Furthermore, in the late DCT, NCC-mediated Na⁺ transport is directly linked to electrogenic Na⁺ and K⁺ transport through the epithelial Na⁺ channel (ENaC) and ROMK.³⁹ NCC function is impaired in Gitelman syndrome,^40,41 resulting in high Na⁺ delivery to the connecting tubule and collecting duct, K⁺ secretion through ROMK and hypokalemia.^41–43 By contrast, NCC activity is increased in pseudohypoaldosteronism type 2, despite low or normal levels of aldosterone, resulting in hypertension and hyperkalemia.^44,45

NCC transport activity is increased by phosphorylation at conserved amino-terminal residues (Figure 1), controlled by a complex signaling response centered on the inwardly rectifying K⁺ channel Kir4.1/5.1, the with no lysine kinases (WNK), and the Ste20p-related proline alanine-rich kinase (SPAK) pathway, which is covered in detail in other reviews.^46–50 In addition to direct effects of ECF K⁺, multiple NCC regulatory mechanisms exist that act at various levels of the Kir-WNK-SPAK cascade or independently of the pathway.^51–59

Figure 1.

The Kir-WNK-SPAK-NCC cascade under low and high dietary K⁺ intakes. The inwardly rectifying K⁺ channels Kir4.1 and Kir5.1 form heterotetramers at the basolateral membrane of DCT cells. During low dietary K⁺ intake, plasma K⁺ (ECF K⁺) decreases, and K⁺ passes out of the cell through Kir4.1/5.1 leading to membrane hyperpolarization. This increases the driving force for Cl⁻ efflux through CLC-Kb channels. As Cl⁻ ions inhibit WNK kinases by stabilizing an autoinhibited conformation, lower intracellular Cl⁻ (estimated at approximately 12 mM) results in WNK4 activation, phosphorylation of the terminal signaling kinase SPAK, and enhanced NCC phosphorylation. During high dietary K⁺ intake, plasma K⁺ (ECF K⁺) is higher and the DCT cell membrane is depolarized. Subsequently, intracellular Cl⁻ levels remain high and the WNK-SPAK axis is inhibited. Concurrently, PP1A becomes dissociated from its inhibitory subunit I-1, driving NCC dephosphorylation. High ECF K⁺ promotes NCC ubiquitylation, resulting in NCC degradation. High K⁺ also stimulates neddylation and activation of the Cullin family of ubiquitin E3 ligases resulting in WNK degradation. Created in BioRender. DCT, distal convoluted tubule; ECF, extracellular fluid; NCC, sodium-chloride cotransporter; PP1A, protein phosphatase 1A; SPAK, Ste20p-related proline alanine-rich kinase; WNK, with no lysine kinases.

The mechanisms that link small changes in plasma K⁺ to NCC have been referred to as the “K⁺ switch,” with ECF K⁺ flicking the switch to alter NCC activity. BP measurements in rodents have demonstrated that the negative correlation between plasma K⁺ levels and NCC activity translates to alterations in systolic BP, although this relationship is most pronounced at lower K⁺ concentrations.^54,60,61 Importantly, evidence that a functional kidney K⁺ switch mechanism occurs in humans has been uncovered by tracking NCC in urinary extracellular vesicles.^62–64

During normal K⁺ intake, principal cells and intercalated cells of the connecting tubule and collecting duct are generally responsible for K⁺ secretion and reabsorption, respectively.⁶⁵ In principal cells, Na⁺ is reabsorbed across the apical membrane through ENaC, creating a negative electrical potential in the lumen that drives K⁺ secretion through ROMK.⁶⁶ This process is enhanced in the late connecting tubule and collecting duct when aldosterone levels are high due to a rise in plasma K⁺.⁶⁷ However, under baseline conditions, when aldosterone levels are low, secretion of K⁺ appears to occur principally in the late DCT and early connecting tubule.⁶⁸ ENaC activity and K⁺ secretion in these regions appears to be independent of aldosterone and driven by stimulation of mineralocorticoid receptors by glucocorticoids.⁶⁹ In collecting duct intercalated cells, an apical H⁺/K⁺-ATPase reabsorbs K⁺ in exchange for H⁺ ions, and expression and activity of this H⁺/K⁺-ATPase can be modified by K⁺ intake.^70–72 Alterations in K⁺-dependent acid secretion may contribute to BP control. In addition, BK channels^36,37 are present in the connecting tubule and collecting duct.⁷³ These channels are activated by elevated luminal flow rate⁷⁴ and, therefore, contribute to K⁺ secretion when tubule fluid K⁺ content is high enough to induce an osmotic diuresis.

Kidney K⁺ Handling and BP Responses during Low Dietary K⁺ Intake

The ability of a low dietary K⁺ intake to increase BP is consistent across many studies. On a low K⁺ intake, K⁺ reabsorption by proximal tubules is increased⁷⁵ and K⁺ secretion in the distal tubules and collecting duct is decreased to conserve K^+76–78 (Figure 2). Anatomically, this translates into reversible kidney enlargement involving most nephron segments apart from the thick ascending limb,⁷⁹ where hypokalemia reduces NKCC2 protein abundance.⁸⁰ Collecting duct principal cells and intercalated cells appear particularly affected by hypertrophy and hyperplasia during low dietary K⁺, consistent with an increase in Na⁺ reabsorption in the collecting duct and enhanced apical nongastric hydrogen potassium ATPase in intercalated cells for K⁺ retention.^81,82 These changes are most apparent in the medulla, where there is a large increase in the quantity of α-intercalated cells and a decrease in β-intercalated cells.⁸³ Hypertrophy of the proximal tubule also occurs after K⁺ depletion, resulting in a greater quantity of Na⁺ transporters.⁸⁴ The molecular signaling underlying proximal tubule sensitivity to low K⁺ appears to center on the inwardly rectifying K⁺ channel Kir4.2, as following K⁺ depletion, Kir4.2 knockout (KO) mice suffer from K⁺ wasting and renal distal tubular acidosis, with proximal tubule hypertrophy and kidney injury reduced.^85,86

Figure 2.

Potassium and sodium handling in the kidney tubule during low K⁺ intake. Simplified representation of K⁺ (blue arrows) and Na⁺ (brown arrows) flux in the different segments (numbered 1–5) of the kidney tubule under a low K⁺ intake. Small black arrows indicate hypertrophy of the segment. Black ± signs, and arrow color intensity and border type indicate an increase (full line)/decrease (dotted line) in transporter number and activity and/or transport. (1) After being freely filtered from the glomerulus into the PT, K⁺ and Na⁺ are reabsorbed in increased quantity. Na⁺ tr.=Na⁺ transporters. (2) In the LoH, K is conserved through counter-current exchange between descending and ascending limbs. (3) In the TAL, hypokalemia reduces NKCC2. (4) In the DCT, low plasma K⁺ leads to a large increase in Na⁺ reabsorption, segment hypertrophy, NCC abundance, and NCC phosphorylation (activity; red P dots). K⁺ excretion through ROMK is reduced. (5) As less Na⁺ is delivered to the CD, less is reabsorbed by ENaC, and K⁺ secretion through ROMK is reduced. The segment also shows hypertrophy. CD, collecting duct; DCT1, early DCT; DCT2, late DCT; LoH, loop of Henle; PT, proximal tubule; ROMK, renal outer medullary potassium channel; TAL, thick ascending limb.

While multiple factors and tubule segments probably contribute to increased BP under low K⁺ intake, the main culprit appears to be the K⁺ switch response and enhanced NCC activity. Studies in rodents showed higher BP during K⁺ deficiency that was greater when fed high salt.^{54,84,87–89} The increased BP correlates with higher NCC and phosphorylated NCC levels and presumably enhanced Na⁺ reabsorption in the DCT.^87,90–92 Marked increases in systolic BP under low K⁺ diets correlated with plasma K⁺: a decrease ranging from 4.5 to 3 mM led to a corresponding systolic BP increase of up to 8 mm Hg, accompanied by increased phosphorylated NCC.⁵⁴ The most convincing evidence linking enhanced NCC activity to higher BP during low dietary K⁺ comes from studies in NCC KO mice, where BP was unchanged when mice were challenged with a K⁺-deficient high salt diet, whereas BP of control mice increased.¹⁵ The low K⁺-induced high Na⁺ reabsorption in the DCT decreases delivery of Na⁺ to the connecting tubule and collecting duct, leading to reduced K⁺ secretion by electrogenic Na⁺/K⁺ exchange. Low K⁺ intake also reduces ENaC^80,92 and ROMK expression at the mRNA and protein level.^93,94

Kidney K⁺ Handling and BP Responses during High Dietary K⁺ Intake

Healthy kidneys have a large capacity to secrete K⁺ to maintain ECF K⁺ homeostasis, with high K⁺ decreasing fluid and Na⁺ reabsorption in proximal tubules and increasing delivery to distal segments⁹⁵ (Figure 3). However, the increased Na⁺ delivery to the DCT does not result in increased Na⁺ reabsorption in this segment, as high ECF K⁺ inhibits NCC activity. Studies in rodents have demonstrated that a dietary K⁺ load (2% K⁺ intake) induces a rapid dephosphorylation of NCC within 15 minutes, leading to natriuresis and kaliuresis.⁹⁶ The decrease in NCC is concurrent with an increase in plasma K⁺, but aldosterone levels were only increased at a later time point. Directly increasing plasma K⁺ by intravenous injection induces the same effect,⁹⁷ suggesting that a gut-derived factor is not essential for the response. The mechanism/s behind the rapid effects on NCC are multifactorial (Figure 1). Higher ECF K⁺ inhibits Kir activity and hence the WNK-SPAK pathway.⁹⁸ Furthermore, high K⁺ stimulates neddylation and activation of the Cullin family of ubiquitin E3 ligases that degrade with no lysine kinases kinases,⁹⁹ further reducing NCC phosphorylation. High K⁺ also promotes expression of the Ppp1Ca gene encoding protein phosphatase 1A and reduces expression of its negative regulator Ppp1r1a (I-1), ultimately enhancing NCC dephosphorylation.⁵⁴ Finally, high K⁺, potentially also through alterations of protein phosphatase 1A activity, enhances chaperone-mediated and ubiquitin-dependent degradation of NCC.¹⁰⁰

Figure 3.

Potassium and sodium handling in the kidney tubule with high K⁺ intake. Simplified representation of K⁺ (blue arrows) and Na⁺ (brown arrows) flux in the different segments (numbered 1–5) of the kidney tubule under a high K⁺ intake. Black ± signs, and arrow color intensity and border type indicate an increase (full line)/decrease (dotted line) in transporter number and activity and/or transport. (1) After being freely filtered from the glomerulus into the PT, K⁺ and Na⁺ are reabsorbed in decreased quantity. Na⁺ tr.=Na⁺ transporters. (2) In the LoH, K⁺ is conserved through counter-current exchange between descending and ascending limbs. (3) In the TAL, Na⁺ is reabsorbed through NKCC2. (4) In the DCT, high plasma K⁺ leads to a sharp decrease in Na⁺ reabsorption by reducing NCC abundance and phosphorylation. K⁺ excretion through ROMK is increased. (5) As more Na⁺ is delivered to the CD, aldosterone-driven increases in ENaC and ROMK facilitate electrogenic K⁺ secretion. Aldo, aldosterone.

In collecting duct principal cells, K⁺ itself acts rapidly through the type 2 mTOR complex to activate SGK1, which stimulates ENaC to enhance K⁺ secretion.^101,102 Subsequent aldosterone-driven increases in ENaC and the Na-K-ATPase in the distal nephron and collecting duct,^103–105 coupled to enhanced Na⁺ delivery (due to NCC inhibition), further promotes K⁺ secretion. K⁺ secretion is also enhanced with long-term high K⁺ intake, after plasma aldosterone has been balanced, which is partly due to an increased secretory capacity of the distal nephron through enhanced flow.⁷⁴ With high K⁺ intake and increased luminal flow rate due to osmotic diuresis, intercalated cells can secrete K⁺ through Ca²⁺-activated BK channels. This mechanism is mediated by the mechanosensor Piezo1, which modulates intracellular Ca²⁺.¹⁰⁶ BK channels are also activated in principal cells; however, the role of BK appears limited, as net K⁺ excretion was unchanged in BK-KO mice under high K⁺ diet, most likely compensated by higher ROMK activity.¹⁰⁷

Despite animal studies of BP responses often using very high K⁺ intake levels that are not easily achieved in humans, up to 5% of the total diet, they do offer mechanistic insights (Table 1). A 5% K⁺ intake reduces reabsorption of both Na⁺ and K⁺ in the proximal tubule, whereas ROMK abundance and apical localization in early DCT, connecting tubule, and collecting duct are all increased.¹¹⁰ With normal salt intake, short-term feeding (4 days) of 5% K⁺ to mice, independent of the accompanying anion (chloride or citrate), consistently reduced NCC phosphorylation and systolic BP.^54,60,87 Similar effects after 4 days, although dampened, were seen using more moderate K⁺ intakes (2%–2.5% K⁺).^54,60 However, a 5% K⁺ dietary intake for longer than 10 days (up to 3 weeks) sees BP rise compared with control K⁺ intake, although NCC is dephosphorylated.^84,87 In these studies, the elevated plasma K⁺ is accompanied by an elevated plasma aldosterone and increased ENaC, and a positive correlation between plasma K⁺ and αENaC has been observed.⁹² Whether these alterations in ENaC activity drive the increase in BP observed with chronically high extreme dietary K intakes is unclear, but the BP effects do appear to be independent of the accompanying anion.^54,84,87 Unfortunately, very few animal studies have looked at the effects of concurrent moderately increased K⁺ and high salt,⁶⁰ and it is therefore difficult to draw conclusions on their combined effects. However, a recent study suggests that moderately increased dietary K⁺ is most beneficial for BP when Na⁺ intake is high.¹¹¹

Table 1.

Summary of animal studies examining BP responses during various sodium (Na⁺) and potassium (K⁺) manipulated diets

Diet Type and Composition (as Listed in Paper)	Altered Transport Proteins	Plasma K Measured and Δ (Compared with Control)	Effects on RAAS	Effects on BP	Ref
K deplete, normal salt	NCC up	Lower	Renin and aldo unchanged or down	Up (3 wk)	87
0% K+, 0.3% Na+	pNCC up	3.21 mM, Δ=−1.25 mM		Unchanged (10 d)	84
		2.88 mM, Δ=−1.18 mM		Up (4 d)	54
K deplete, high salt	NCC up	Lower	Aldo down	Up ++	15, 84, 87–89
0% K+, 1.57% Na+	pNCC up	2.37 mM, Δ=−1.01 mM
0% K+, 1% NaCl		3.9 mM, Δ=−1.3 mM
0% K+, 2.4% Na+		2.2 mM, Δ=−1.69 mM
0% K+, 3% NaCl
High K, normal salt	NCC down or unchanged	Normal or slightly increased	Renin slightly down (ns)	Down (4 d)	54, 60
2.5% K+, 0.3% Na+	pNCC down	4.4 mM, Δ=+0.7 mM	Aldo up
2.1% K+ (KCl), 0.3% Na+	ENaC up
Very high K, normal salt	NCC down or unchanged	Increased	Renin down, aldo up	Up (3 wk)	87
5% K+ (KCl or KCit), 0.3% Na+	pNCC down	5.23 mM, Δ=+0.77 mM		Up (10 d)	84
5% K+ (1:1:1 KCl, KCit, KHCO3), 0.3% Na+	ENaC up	5.34 mM, Δ=+1.28 mM		Up (7 d)	87
5% K+ (KCl or KHCO3), 0.3% Na+		4.43 mM, Δ=+0.64 mM			54, 60, 87
5.2% K+ (KCl), 0.74% NaCl		5.1 mM, Δ= +1.4 mM		Down (4 d)
Normal K, high salt	NCC down	Normal	Renin and aldo down	Up	60, 87
0.8% K+ (KCl), 4% NaCl	ENaC down or unchanged	4.63 mM, Δ=+0.16mM		Unchanged	89
1% K+, 1.57% Na+		3.89 mM, Δ=? (No control salt)
0.93% K+ (KCl), 3% NaCl
High K, high salt	NCC down or unchanged	Normal to increased	Renin down	Down (4 d; control diet high salt)	60
2.1% K+ (KCl), 4% NaCl	pNCC down	4.4 mM, Δ=+0.7 mM	Aldo down
	ENaC up
Very high K, high salt	NCC down or unchanged	Normal to increased	Aldo up	Up (3 wk)	87
5% K+, 1.57% Na+	pNCC down	4.59 mM, Δ=+0.13 mM		Up ++ (10 d)	89
5% K+ (KCl), 3% NaCl		5.1 mM, Δ= +1.4 mM		Unchanged (10 d)	84
5% K+ (1:1:1 KCl, KCit, Kcarb), 1% NaCl	ENaC up	4.26 mM, Δ=+0.37 mM		Down (4 d)	60
5.2% K+ (KCl), 4% NaCl

The relatively small range of plasma K⁺ values despite wide changes in K⁺ intake highlight the capacity of the kidney to maintain homeostasis and limit hyperkalemia, which is only diminished if K⁺ intake swings suddenly from depleted to high K⁺.¹⁰⁸ However, this small magnitude of change should be enough to switch sodium-chloride cotransporter from maximally phosphorylated to dephosphorylated.¹⁰⁹ NCC, sodium-chloride cotransporter; RAAS, renin-angiotensin aldosterone system.

The potential negative effects of K⁺ on BP are intriguing. Higher K⁺ intake increases aldosterone secretion, which when chronically high can detrimentally affect the cardiovascular system.^112–114 Patients with primary aldosteronism have higher incidence of left ventricular hypertrophy compared with patients with primary hypertension.^115–117 However, a high Na⁺ intake seems to be required,^118–120 as aldosterone on its own, independently of hypertensive status, does not appear to have adverse effects on the cardiovascular system.^121–126 Supporting this, primary aldosteronism patients consuming a high Na⁺ diet have a higher risk of cardiovascular events, whereas natives from Papua New Guinea or the Yanomami tribe of the Amazon consuming a diet low in Na⁺ and high in K⁺ do not suffer from high BP or cardiovascular damage despite high aldosterone.¹²⁷

Effects of Dietary K⁺ Interventions in Humans

Observational, epidemiologic, and interventional studies examining the effects of dietary K⁺ intake on BP have great differences in outcomes. For example, in normotensive individuals, 10 mmol/d K⁺ supplementation raised systolic BP,¹²⁸ approximately 20–30 mmol K⁺/d lowered BP,^129,130 while approximately 80–150 mmol K⁺/d had no effect.^{21,64,122,125} This variability may result from low participant numbers, their hypertensive status (also stratification of “normotensive” or “hypertensive”), age, sex, study duration, K⁺ supplement (10–120 mmol/d), the accompanying anion, or the use of urinary K⁺ as a proxy for dietary intake. However, even in relatively similar studies, BP outcomes can be different. For example, in hypertensive elderly people supplemented with 48 mmol K⁺/d for 4 months a significant reduction in BP was observed,¹³¹ but in other studies using similar K⁺ amounts and duration, no major beneficial effects on BP were observed.^132,133 Here, we have compiled studies where K⁺ intake alone is changed (Table 2), grouping normotensive and hypertensive status of participants for clarity.

Table 2.

Summary of studies assessing different potassium interventions on BP in normotensive and hypertensive individuals

Author	Study Design	Participants	Participants' Initial BP Status (mm Hg)	Intervention/Duration	Main Results
Normotensive participants
Krishna et al.128	Randomized Cross-over	No: 10 Sex (M/F): 10/0 Age (yr): 20–40 Ethnicity: White	BP <140/90	KCl (10 or 90 mmol/d) Duration: 9 d Washout: 4–8 wk	10 mmol: SBP+DBP ↑ 90 mmol: SBP+DBP: NC
Naismith and Braschi130	Double-blinded Randomized Parallel	No: 30(KCl)/29(placebo) Sex (F): 53%/35% Age (yr): 44.5±2.1/41.7±2.2 Ethnicity: European 83/83, Middle Eastern 7/7, East Asian 7/3, South Asian 3/7	SBP 118.2±2.6 (KCl)/115.7±2.4 (placebo)	KCl (24 mmol K/d) or placebo tablets Duration: 6 wk	SBP+DBP ↓
Braschi and Naismith129	Double-blinded Randomized Parallel	No: 34(placebo)/26(KCl)/30KCit Sex (F): 58.8%/80.8%/56.7% Age (yr): 33.8/36.9/36.2 Ethnicity: Caucasian 67.6/76.9/86.7, Middle Eastern 14.7/7.7/6.7, East Asian 14.7/7.7/6.7, Afro-Caribbean 2.9/7.7/0	SBP 107–114 DBP 66–70	KCl (30 mmol K/d), KCit (30 mmol K/d), or placebo tablets Duration: 6 wk	KCl: SBP+DBP ↓ KCit: SBP+DBP ↓
Dreier et al.125	Double-blinded Randomized Cross-over	No: 25 Sex (M/F): 25/0 Age (yr): 20–55 Ethnicity: ND	SBP <140, DBP <90	KCl (90 mmol K/d) or placebo tablets Duration: 4 wk Washout: 2 wk	SBP+DBP: NC
Matthesen et al.122	Randomized Cross-over	No: 21 Sex (M/F): 9/12 Age (yr): 18–40 Ethnicity: Caucasian	SBP <140, DBP <90	KCl (100 mmol K/d) or placebo tablets Duration: 28 d Washout: 2 wk (Postintervention participants ingested a standardized diet for 4 d before measurements)	SBP+DBP: NC
Wu et al.64	Double-blinded Randomized Cross-over	No: 28 Sex (M/F): 8/20 Age (yr): 22–62 Ethnicity: ND	SBP <140, DBP <90	KCl (132 mmol K/d) or placebo tablets Duration: 5 d Washout: 2 d (All participants kept on a high Na [200 mmol/d])	SBP+DBP: NC
Blanch et al.21	Single-blinded Randomized Cross-over	No: 35 Sex (M/F): 9/26 Age (yr): 31±11 Ethnicity: ND	SBP <130, DBP <90	High K (150 mmol/d from fruit and vegetables) or low K (80 mmol/d from fruit and vegetables) Duration: 6 d Washout: 24 h	SBP+DBP: NC
Hypertensive participants
Berry et al.121	Double-blinded Randomized Cross-over	No: 48 Sex (M/F): 23/25 Age (yr): 44.8±8.2(F)/45.5±10.6(M) Ethnicity: European 29, Asian 9, African 10	Mild DBP >80 and <100	K (20 mmol/d from fruit and veggies), K (40 mmol/d from fruit and veggies), KCit (40 mmol K/d from tablets), or placebo tablets Duration: 6 wk Washout: 5 wk	SBP+DBP: NC
Franzoni et al.134	Parallel	No: 104 Sex (M/F): 65/39 Age (yr): 53±12 Ethnicity: ND	Mild to moderate SBP/DBP 154.2/96.2±10.8/5.4	K-Asp (30 mmol K/d) or none Duration: 4 wk	SBP+DBP ↓
Vongpatanasin et al.135	Double-blinded Randomized Cross-over	No: 30 Sex (M/F): 14/16 Age (yr): 54±12 Ethnicity: 12 Black, 18 White	Mild SPB 120–159 DBP 80–99	KCl (40 mmol K/d), KCit (40 mmol K/d), or KMgCit (40 mmol K, 20 mmol mg, 74 mmol citrate/d), or placebo powder Duration: 4 wk Washout: 1 wk	KCl: SBP ↓, DBP: NC KCit: SBP+DBP: NC KMgCit: SBP+DBP: NC
Gu et al.136	Double-blinded Randomized Parallel	No: 75(KCl)/75(placebo) Sex (M): 28/32 Age (yr): 56.9±7.4/55±7.6 Ethnicity: Chinese	Mild SBP 130–159 and/or DBP 80–94	KCl (60 mmol K/d) or placebo tablets Duration: 12 wk	SBP ↓, DBP: NC
Fotherby and Potter131	Double-blinded Randomized Cross-over	No: 8 Sex (M/F): 1/7 Age (yr): 68–79 Ethnicity: ND	SBP 160 and/or DBP 90	KCl (60 mmol K/d) or placebo tablets Duration: 4 wk then KCl (48 mmol K/d) or placebo tablets Duration: 4 mo	4 wk: SBP ↓, DBP: NC 4 mo: SBP ↓, DBP: NC
Kawano et al.133	Randomized Crossover	No: 55 Sex (M/F): 26/29 Age (yr): 36–77 Ethnicity: Japanese	Mild to moderate SBP >140 or DBP >90	KCl (64 mmol K/d) tablets or none Duration: 4 wk	SBP+DBP ↓
Graham. et al.123	Double-blinded Randomized Cross-over	No: 40 Sex (M/F): 32/8 Age (yr): 40–70 Ethnicity: ND	Average SBP 140.6±3	KCl (64 mmol K/d) or placebo tablets Duration: 6 wk Washout: 6 wk Washout of K-sparring diuretics and other antihypertensives before trial	SBP ↓, DBP: NC
He et al.132	Double-blinded Randomized Cross-over	No: 42 Sex (M/F): 30/12 Age (yr): 51±11 Ethnicity: 29 White, 10 Black, 3 Asian	SBP 140–170 and/or DBP 90–105	KCl (64 mmol K/d), KHCO3 (64 mmol K/d), or placebo tablets Duration: 4 wk	KCl: SBP+DBP: NC KHCO3: SBP+DBP: NC
Smith et al.137	Double-blinded Randomized Cross-over	No: 20 Sex (M/F): 11/9 Age (yr): 30–66 Ethnicity: White 18, Black 2	Mild or moderate DBP 90–119	KCl (64 mmol/d) or placebo tablets Duration: 1 mo (Prior restriction of sodium intake to app 70 mmol/d)	SBP+DBP: NC
Gijsbers et al.126	Double-blinded Randomized Cross-over	No: 36 Sex (M/F): ND Age (yr): 47–80 Ethnicity: ND	Untreated elevated BP average BP of 145/81, 69% SPB above 140	3 g added potassium (77 mmol), 3 g added sodium (130.4 mmol), or placebo/d Background diet of 2 g Na and 2 g K	SBP+DBP: NC
Krishna and Kapoor138	Double-blinded Randomized	No: 12 Sex (M/F): 10/2 Age (yr): 23–58 Ethnicity: 7 Black, 5 White	DBP >90	KCl (80 mmol K/d) or placebo tablets Duration: 10 d Washout: 4–8 wk	SBP+DBP ↓
Turban et al.139	Randomized Cross-over	No: 25 Sex (F): 58.6% Age (yr): 67.2±11.6 Ethnicity: Black 69%	Stage 3 CKD SBP 120–159 DBP <100	K (100 mmol/d through diet) or K (40 mmol K/d through diets) Duration: 4 wk Washout: 3–4 wk	SBP+DBP: NC
Svetkey et al.140	Double-blinded Randomized Parallel	No: 59(KCl)/57(placebo) Sex (M): 76%/72% Age (yr): 51.3/50.8±12.3 Ethnicity: 89%/83% White	Mild DBP 90–105	KCl (120 mmol K/d) tablets or placebo Duration: 8 wk	SBP+ DBP ↓

d, days; DBP, diastolic BP; M/F, males/females; mo, months; NC, no change; ND, no data available; SBP, systolic BP; wk, weeks; yr, years.

In hypertensive individuals, effects of K⁺ are more consistent (Table 2). For example, over 100 years ago, Addison et al.¹⁴¹ demonstrated that K⁺ administration to hypertensives reduced BP, confirmed in numerous other studies,^{123,131,133–136,138,140} although others show no change in BP after K⁺ supplementation.^{121,126,132,137,139} Meta-analysis of these effects in normotensive and hypertensive individuals suggests a range of K⁺ intake exists that lowers BP, under and above which BP either rises or does not change.^14,64,125 Other meta-analysis studies in general support that K⁺ supplementation can reduce BP, predominantly in hypertensive individuals. For example, in a study with 2609 participants,¹⁴² systolic BP trended lower in 81% of the trials and 73% for diastolic BP (significant in 34% and 30%, respectively). Pooled data from this study yielded robust evidence for the intervention's efficacy with systolic BP and diastolic BP reductions of 3.1 and 2.0 mm Hg, respectively, with a tendency for greater effects on systolic BP in hypertensives. Other meta-analysis studies of randomized supplementation trials draw similar conclusions; the ability of higher K⁺ intake to reduce systolic BP and diastolic BP is greater in people with existing high BP,^143,144 and these effects may be influenced by duration of the supplement.^142,144 Furthermore, the Dietary Approaches to Stop Hypertension (DASH) diet trial showed substantially greater BP reducing effects in hypertensive individuals.¹⁴⁵

The Intersalt study¹⁴⁶ highlighted that the relationship between the urinary Na⁺/K⁺ ratio and BP was stronger than that of urinary Na⁺ excretion alone, with a significant negative correlation between K⁺ excretion and BP even when correcting for Na⁺. This association between high Na⁺ intake and hypertension has long been observed^16,147,148 and was termed salt-sensitive hypertension by Dahl et al.¹⁴⁷ Dahl also observed that salt-sensitive rats on a high Na⁺ diet had lower BP when K⁺ intake was high.¹⁴⁹ Various meta-analyses also suggest that the effects of K⁺ to reduce BP are greater when Na⁺ excretion is high (>165 mmol/d),¹⁴² and K⁺-driven BP reductions are positively correlated with baseline urinary Na:K ratios.¹³⁰ This relationship between K⁺ and Na⁺ on BP is strengthened by the lack of effect of dietary K⁺ in mild to moderate hypertensive patients when Na⁺ intake was already restricted.^137,150,151 Background Na⁺ intake appears crucial to the observed effects, as BP tends to rise with high K⁺ intake in prehypertensive patients on a restricted Na⁺ diet.^126,137,152 Randomized crossover studies also support that Na⁺ effects on BP differ depending on K⁺ intake, with high Na⁺ increasing BP with a low K⁺ intake but not a normal K⁺ intake.^128,153 Furthermore, participants on a control K⁺ diet (45 mmol/d) with high Na⁺ (150 mmol/d) had the biggest reduction in BP when placed on the DASH diet (120 mmol/d K⁺) and a low sodium intake (50 mmol/d), compared with the DASH or low Na⁺ diet alone,¹⁵⁴ effects that again were greater in hypertensive participants.

Strong evidence that Na⁺ and K⁺ effects on BP are interdependent comes from the Salt Substitute and Stroke Study¹⁵⁵ that assessed the impact of long-term salt substitution (25% of NaCl substituted by KCl) in a large cohort in rural China. Those with salt substitute showed significantly lower rates of stroke, major cardiovascular events, and death, although participants were restricted to individuals older than 60 years and nearly 90% had hypertension and/or a history of cardiovascular disease. Reanalysis of the Salt Substitute and Stroke Study data showed that salt substitution significantly reduced cumulative and traditional systolic BP and mean arterial pressure measurements, but not diastolic BP.¹⁵⁶ A comparable strategy in elderly care facilities showed a similar reduction in BP and cardiovascular outcomes, while only reducing salt had no effect.¹⁵⁷

At first glance, it appears that the effects of K⁺ supplementation may be influenced by ethnicity. For example, Black populations are particularly sensitive to the BP lowering effects of high K⁺.^139,142 K⁺ interventional studies (approximately 60 mmol/d) in Asian populations with mild hypertension, who traditionally consume low K⁺ and high Na⁺ diets, also showed significant reductions in systolic BP.^133,136 However, as Black and Asian populations in general have higher Na⁺ intake, the K⁺ effect might not be genetic but linked to dietary habits. For example, Kurtz et al.¹⁵⁸ showed that BP effects of changing salt intake in Black versus White individuals were only apparent when K⁺ intake was low. Furthermore, as skeletal muscle stores approximately 80% of intracellular K⁺ and men on average have a greater muscle mass than females that reduces over time,^159,160 age-dependent sex differences may exist in how K⁺ influences BP. To the best of our knowledge, no clinical studies have been conducted on females alone, but interestingly, in five of eight studies showing no changes in BP with K⁺ intervention, most of the participants were female.^{21,121,122,125,136,139}

The modern diet has also resulted in reversed ratios of HCO₃⁻ to Cl⁻, with fruit and vegetables rich in K⁺ and HCO₃-yielding precursors such as citrate, but containing very little Cl⁻.¹⁶¹ These alternative anions may play a role in perceived “K⁺ effects,” especially as Cl⁻ may be crucial for the hypertensive response to Na⁺,^162–164 and a pressor effect of Cl⁻ may antagonize the beneficial effect of K⁺ on BP.¹⁶⁵ However, KCl is more beneficial that KHCO₃ for systolic BP reduction,¹³² and KCl has similar¹²⁹ or superior effects to K-citrate in lowering systolic BP.¹³⁵ Interestingly, when K⁺ was administered as K-aspartate to hypertensive individuals,¹³⁴ a dose of approximately 30 mmol/d lowered systolic BP and diastolic BP (−8.8 and −5.4 mm Hg, respectively) to a greater extent than the average effects of approximately 75 mmol/d of KCl (−4.1 and −2.4, respectively).¹⁴²

Summary and Knowledge Gaps

In animal studies, low K⁺ intake or K⁺ deficiency consistently increases BP, whereas a higher K⁺ intake appears to have a transient BP-lowering effect, explained by reduced NCC-mediated NaCl reabsorption. However, these beneficial effects disappear when K⁺ intake is either extreme, or intake is sustained for an extended period. Although these bad effects have been attributable to aldosterone-driven increases in ENaC,¹¹¹ it is not clear whether aldosterone-mediated alterations in vascular function contribute, whether there is aldosterone-driven inflammation in the kidney or other tissues, or whether combining moderately increased K⁺ with a high salt intake can suppress aldosterone secretion to a level where the K⁺ intake remains beneficial for BP over a longer period. Further studies in animals using moderate alterations in K⁺ intake are warranted to address these issues. Furthermore, whether effects of K⁺ on BP and the molecular responses are similar between animals of different sexes, genetic background, age, body mass, or hypertensive status are lacking and could provide useful parameters to guide clinical studies.

In humans, there is consensus that K⁺ supplementation is beneficial. However, as modern diets are typically low in K⁺, supplementation may simply restore K⁺ levels to the recommended baseline rather than providing an excess. This raises the question of whether such interventions are true supplementations or a correction of deficiency. It is feasible that the “U-shape relationship” between BP and long-term K⁺ supplementation suggested by meta-analysis¹⁶⁶ holds true, but the “sweet spot” for consistently lowering BP when modifying only K⁺, that is, the bottom of the curve, will be difficult to find due to uncertainties about base diet. Overall, normotensive individuals tend to be less responsive to K⁺ supplementation, while hypertensive individuals show greater effects. Notably, there is a lack of clear data comparing the average baseline K⁺ intake between normotensive and hypertensive individuals. Furthermore, BP effects of K⁺ intake appear closely linked to Na⁺ intake, with effects most beneficial when salt intake is high. As high salt intake leads to raised BP, at least at the population level,¹⁶⁷ this then exacerbates K⁺ intake effects.

Disclosures

Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/JSN/F571.

Author Contributions

Conceptualization: Adrienne M. Assmus, Robert A. Fenton, Lena K. Rosenbaek.

Data curation: Adrienne M. Assmus, Lena K Rosenbaek.

Formal analysis: Adrienne M. Assmus.

Writing – original draft: Adrienne M. Assmus, Robert A. Fenton, Lena K Rosenbaek.

Writing – review & editing: Adrienne M. Assmus, Robert A. Fenton, Lena K Rosenbaek.

Funding

F.A. Fenton: Novo Nordisk Fonden (NNF21OC0067647) and Danmarks Frie Forskningsfond (0134-00018B and 3101-00136B). A.M. Assmus: HORIZON EUROPE Marie Sklodowska-Curie Actions (101105368).