Substance and claimed effects

Potassium-enriched salt substitutes (low-sodium salt, LSSS) are kitchen seasonings in which a fraction of the sodium chloride (NaCl) of ordinary table salt is replaced with potassium chloride (KCl). The dominant formulation studied in cardiovascular outcome trials is 75% NaCl / 25% KCl Neal 2021; consumer brands (LoSalt, Nu-Salt, Morton Lite Salt) sit between 50/50 and 70/30. Some products also include MgSO₄, amino-acid umami enhancers, or dried flavor concentrates Yuan 2023. The claimed effects in scope for this entry: reduction in 24-hour sodium intake; increase in 24-hour potassium intake; reduction in systolic and diastolic blood pressure; reduction in stroke, major adverse cardiovascular events (MACE), and all-cause mortality at the population level; and — the interaction the brief centers on — increased risk of hyperkalemia in users with impaired renal potassium excretion (CKD, especially eGFR < 30) or pharmacologically blunted aldosterone-driven kaliuresis (ACE inhibitors, ARBs, ARNi, steroidal MRAs like spironolactone/eplerenone, non-steroidal MRAs like finerenone, potassium-sparing diuretics, NSAIDs, calcineurin inhibitors). Out of scope: low-sodium broths, MSG, KCl supplements per se, and the sodium-reduction literature that doesn't involve KCl replacement.

Evidence by addressing question

Mechanism

Two simultaneous moves: less Na⁺, more K⁺. The blood-pressure-lowering effect is biologically additive — high Na⁺ intake raises extracellular fluid volume and vascular reactivity; high K⁺ intake promotes natriuresis, reduces sympathetic tone, and improves endothelium-dependent vasodilation. A factorial analysis of urinary 24-hour Na⁺ and K⁺ data from SSaSS estimated that roughly 60% of the systolic BP reduction was attributable to potassium increase and 40% to sodium reduction, an unusually clean partition for a real-world dietary intervention Neal 2021.

Hyperkalemia mechanism (the interaction at the heart of this entry): renal K⁺ excretion is the dominant route of potassium homeostasis (~90% of intake) and is controlled by aldosterone acting on the principal cells of the cortical collecting duct via the epithelial sodium channel (ENaC) and the renal outer medullary potassium channel (ROMK). ACE inhibitors and ARBs reduce angiotensin-II-driven aldosterone secretion; MRAs directly block the mineralocorticoid receptor; potassium-sparing diuretics block ENaC; CKD reduces functional nephron mass and tubular flow. Every one of these compresses the kidney's K⁺ excretory headroom. Add a sustained KCl load (a typical 75/25 substitute at 5 g/day adds ≈ 600–700 mg of elemental K⁺, roughly the potassium content of a medium banana) and the steady-state serum K⁺ shifts up. The Chronic Kidney Disease Prognosis Consortium documented the resulting clinical gradient: hyperkalemia events scale from ≈ 0.8 per 100 person-years in healthy people to ≈ 8.7 per 100 person-years in CKD on RAS inhibitors — a tenfold increase KDIGO 2024.

Evidence

The defining trial is SSaSS — a 5-year, open-label, cluster-randomized trial in 600 rural Chinese villages (n = 20,995), randomizing households to a 75/25 KCl-NaCl substitute or to regular salt Neal 2021. All participants had either a prior stroke or were ≥ 60 with uncontrolled hypertension; participants with severe CKD or on potassium-sparing diuretics were excluded. After mean 4.74 years: stroke RR 0.86 (95% CI 0.77–0.96, p = 0.006), MACE RR 0.87 (0.80–0.94, p < 0.001), all-cause mortality RR 0.88 (0.82–0.95, p < 0.001). Systolic BP fell 3.3 mm Hg vs. control. Critically for this entry's question, clinical hyperkalemia hospitalizations were not significantly different (RR 1.04, 95% CI 0.80–1.37) in this screened population. This is the strongest single piece of evidence; it also defines the exclusion zone.

DECIDE-Salt (Yuan et al., Nature Medicine 2023) was a 2×2 factorial cluster RCT in 48 Chinese eldercare facilities (n = 1,612, mean age 71). The substitute (62.5% NaCl / 25% KCl / 12.5% flavorings) lowered systolic BP by 7.1 mm Hg and cardiovascular events by HR 0.60 (0.38–0.96) vs. usual salt over 2 years Yuan 2023. The study reported a measurable increase in mean serum potassium and more frequent biochemical hyperkalemia, but no excess clinical adverse events. A post hoc normotensive sub-analysis showed substitute use reduced incident hypertension by adjusted HR 0.60 (0.39–0.92).

Bernabe-Ortiz et al. ran a stepped-wedge cluster RCT in 6 Peruvian villages (n = 2,376) with the same 75/25 formulation distributed at the household level. Population-wide systolic BP fell ≈ 1.3 mm Hg and incident hypertension by 51% (HR 0.49) in baseline-normotensive adults Bernabe-Ortiz 2020. The trial replicates the BP effect outside the East Asian, salt-heavy dietary context.

Meta-analytic synthesis. Hernandez et al. (Heart 2019) pooled 21 RCTs and found mean SBP/DBP reductions of −7.81 / −3.96 mm Hg Hernandez 2019. The post-SSaSS update by Yin et al. (Heart 2022, 21 trials, n = 31,949) gave pooled SBP −4.61 (95% CI −6.07 to −3.14) and DBP −1.61 (−2.42 to −0.79) mm Hg, with a 13% relative risk reduction in total mortality and 11% in major cardiovascular events Yin 2022. Evidence for hard outcomes is dominated by SSaSS; the other clinical-event signals come from DECIDE-Salt and the Chinese elderly-care literature feeding into the meta-analysis.

Protocol

The behavior is mechanical: dump the table-salt shaker, refill with a KCl-blended product, and use the same way for cooking and at the table. The 75/25 SSaSS formulation maps to roughly 1.5 g KCl per teaspoon, contributing ≈ 750 mg of elemental potassium per teaspoon. Typical home use of 3–6 g substitute per day delivers 450–900 mg of additional dietary K⁺. Commercial brands vary: LoSalt (UK) 33% NaCl / 66% KCl, Morton Lite Salt 50/50, Nu-Salt nearly pure KCl. Reading the label matters because total K⁺ load varies threefold across products. The KCl payload is bitter and metallic above ~30% replacement; the SSaSS and DECIDE-Salt formulations were palate-tuned with flavorings to stay below the taste-rejection threshold Yuan 2023. There is no need for restaurant or processed-food behavior change to capture the bulk of the BP benefit in trials, but the trial effect size scales with the share of dietary salt that's the discretionary table/cooking fraction — high in rural China (≈ 70%), lower in industrialized countries (≈ 20–25%).

Contraindications

This is the heart of the brief and the dossier's hard line. The patients who derive the most population benefit from sodium reduction — older adults with hypertension, CKD, heart failure — are also disproportionately on RAS inhibition. The hazard layers:

• CKD G4–G5 (eGFR < 30) and CKD G3b with prior hyperkalemia. Reduced excretory capacity; KCl-substitute loads of 600–900 mg/day are predictably enough to push serum K⁺ above 5.5 mmol/L in a meaningful fraction. KDIGO 2024 advises individualized counseling and limiting bioavailable K⁺ intake (including substitutes) in CKD G3–G5 with a history of hyperkalemia KDIGO 2024.
• Patients on MRAs (spironolactone, eplerenone, finerenone). The largest absolute hyperkalemia risk among RAS-axis drugs. RALES (n = 1,663) showed ≈ 2% incidence of severe hyperkalemia even in the trial's tightly monitored cohort; uncontrolled post-marketing use is worse Pitt 1999. Layering a KCl substitute on top is the high-risk combination.
• Dual RAAS blockade (ACEi + ARB, or with direct renin inhibitor). Generally contraindicated independent of salt substitutes; layering KCl multiplies the harm.
• Potassium-sparing diuretics (amiloride, triamterene) — same logic as MRAs; usually used in fixed combinations with thiazides.
• Trimethoprim, calcineurin inhibitors (tacrolimus, cyclosporine), heparin, NSAIDs. All blunt renal K⁺ excretion; the interaction with substitutes is less studied but mechanistically real.
• Hyporeninemic hypoaldosteronism (type 4 RTA), classically in diabetic kidney disease — endogenous parallel of the RAS-blocker hazard.
• Adrenal insufficiency, Addison's disease.
• Acute illness with volume depletion, AKI, or rhabdomyolysis — any state that drops eGFR acutely.

For uncomplicated hypertension on an ACE/ARB without CKD or prior hyperkalemia, the picture is different. The post-hoc SSaSS analysis (Circulation 2025) examining outcomes in the ≈ 8% of SSaSS participants reporting RAAS-inhibitor use found that the cardiovascular benefit was preserved and excess clinical hyperkalemia events were not detected at the group level — but the analysis was underpowered for rare hyperkalemia events and excluded the highest-risk groups by trial design from the start Yin 2025. The reasonable interpretation: an otherwise low-risk patient with normal renal function and a baseline K⁺ < 4.5 on monotherapy ACE/ARB can use a substitute with a check of serum K⁺ 2–4 weeks after starting and at routine monitoring intervals; the patient with CKD G3b+, on an MRA, or with prior K⁺ > 5.0 should not.

Misconceptions

• "It's just salt with less sodium." A 75/25 substitute delivers ~13 mmol K⁺ per teaspoon — pharmacologically meaningful, equivalent to a low-dose oral KCl supplement. Daily use is a low-dose K⁺ supplement disguised as a kitchen staple.
• "SSaSS proved it's safe." SSaSS excluded the high-risk groups: severe CKD, potassium-sparing diuretics. The hyperkalemia null result applies to screened populations, not to anyone with a salt shaker and a prescription Neal 2021.
• "If I'm on a 'water pill', I'm fine." The diuretic class matters. Thiazides (HCTZ, chlorthalidone, indapamide) and loop diuretics (furosemide, torsemide, bumetanide) lower serum K⁺ — adding a substitute may be appropriate. Potassium-sparing agents (amiloride, triamterene, spironolactone, eplerenone) and combinations like Maxzide/Dyazide raise it — adding a substitute is dangerous.
• "Hyperkalemia gives you warning symptoms." Mostly it doesn't, until it's severe and arrhythmogenic. The first sign is often a routine lab or an ECG with peaked T waves at K⁺ > 6.
• "It's contraindicated in everyone with kidney disease." Overstated for early CKD (G1–G2) with normal serum K⁺; KDIGO 2024 explicitly individualizes the call KDIGO 2024. The hard line is advanced CKD plus K⁺-retaining drugs, not all CKD.

Population variability

Effect size on BP scales with: (a) baseline sodium intake — bigger drop in high-Na populations (rural China > Mediterranean), (b) the discretionary salt fraction — bigger benefit where most salt comes from the cooking pot, not the supermarket, (c) baseline hypertension status — larger SBP fall in stage-2 hypertensives than normotensives, (d) age — older adults respond more Yuan 2023. Hyperkalemia risk varies with: baseline eGFR, RAAS-blockade intensity (monotherapy < dual blockade < MRA-containing combinations), diabetes status (type-4 RTA), heart failure with reduced ejection fraction (HFrEF, where MRA use is near-universal in GDMT). Ethnicity matters indirectly via background K⁺ diet (the DASH/Mediterranean baseline already runs high K⁺) and via the prevalence of CKD undetected at the population level.

Stakes

The stakes side splits in two. For low-risk hypertension: not switching to a substitute when you cook most of your own food is leaving a 3–5 mm Hg systolic improvement on the table — about the same as adding a second-line antihypertensive, but free and palatable. Over the next decade that translates, on the SSaSS effect size, to ≈ 12% fewer strokes and 14% fewer cardiovascular events at the individual level Neal 2021. For CKD / MRA / heart failure patients: the wrong stakes framing is the dominant clinical hazard. Severe hyperkalemia (> 6.5 mmol/L) can produce cardiac arrest with no warning prodrome; emergency K⁺-lowering protocols (calcium gluconate, insulin-dextrose, sodium zirconium cyclosilicate or patiromer) work but require ER admission. Reports of substitute-induced hyperkalemia from family members bringing "tasty salt" to a hospitalized CKD patient are now in the case-report literature.

Payoff

For the right patient — uncomplicated hypertension, normal renal function, cooks at home — payoff is fast, durable, and cheap. SBP starts dropping within 2–4 weeks (the timescale on which Na/K balance re-equilibrates); the full effect is in by 8 weeks. Over 5 years that's the SSaSS magnitude: 14% fewer strokes, 13% fewer cardiovascular events, 12% fewer deaths Neal 2021. The intervention costs cents per day (KCl is cheap; the substitute carries a small premium over table salt). For the high-risk patient, the payoff is the avoidance of a preventable harm — an iatrogenic hyperkalemia hospitalization, or worse, an MRA dose reduction that compromises heart-failure GDMT and shortens life.

Practicalities

Available globally — LoSalt, Selina Selena, Morton Lite Salt, Nu-Salt, Heart Salt, McCormick Salt Substitute, store-brand "lite" salts. Cost $4–12 per kg vs. $1–3 for table salt; daily delta is pennies. Cooking behavior is unchanged for moderate replacement; heavy KCl (>50%) gets bitter, especially in baking and brines. Iodine fortification — important globally for goiter prevention — needs checking on the substitute label; some brands carry iodized KCl, others do not. Restaurant and processed-food sodium is not addressed by switching the home shaker; in countries with high processed-food intake (US, UK), home-cooking switch captures only 20–30% of dietary Na.

Alternatives

Other paths to the same BP endpoint: (a) DASH eating pattern — same K⁺-up, Na⁺-down direction via whole foods, larger fiber/magnesium dividend, much higher effort; (b) plain sodium reduction without KCl — works but the effect size is smaller in trials because the K⁺ lift is missing Greer 2024; (c) industrial reformulation — voluntary or regulated salt reduction in packaged food, the WHO-preferred population strategy WHO 2023; (d) antihypertensive medication — second-line add-on or up-titration; same MAP delta, more side effects, more cost. The substitute is uniquely attractive because it's a single point-of-purchase swap that captures both Na and K levers.

Credibility range

Optimist case

SSaSS is one of the cleanest, largest dietary-intervention trials ever run with hard outcomes — randomized, cluster-design, 21,000 participants, 5-year follow-up, a 14% stroke reduction with a wide CI excluding zero, replicated mechanistically by DECIDE-Salt and population-wise by Bernabe-Ortiz. The intervention is a single behavioral change (swap the shaker), cheap, palatable in trial-grade formulation, and works whether the user does anything else right. Modeled out at population scale, universal substitution prevents ~460,000 strokes/year in China alone Neal 2021. Hyperkalemia concern is real but quantifiable — in SSaSS, the screened population showed no excess; in DECIDE-Salt, biochemical excess didn't produce clinical excess. Clinicians overuse the contraindication; the actual hazard zone is narrow (CKD G3b+, MRAs, prior K > 5) and screenable with a single serum-K measurement.

Skeptic case

SSaSS was conducted in rural China — a population with discretionary-salt fraction of ~70% (vs. ~20% in industrial diets), baseline Na intake far above Western norms, and limited access to RAS-blocker pharmacotherapy at SSaSS-era prescribing rates. Generalization to the US/UK/EU is a model-and-pray exercise. The hyperkalemia null result is in a population from which the high-risk groups were excluded by design — the trial cannot speak to safety in CKD, MRA users, or heart-failure GDMT patients, which are exactly the populations where the BP indication is strongest. The Circulation 2025 sub-analysis is suggestive but underpowered and observational within the trial. Case reports of severe hyperkalemia in CKD patients using substitutes exist; mechanistically, every nephrologist will tell you a chronic 500–800 mg/day K⁺ load is non-trivial in a damaged kidney Yin 2025. The commercial substitute market is unregulated for KCl content — labels read ambiguously, total K⁺ load varies threefold across brands. And the asymmetric harm structure (a missed BP drop is unobserved; an iatrogenic K of 7.2 is a code) means a population-level recommendation cannot be the individual-level call.

Author's call

Both cases are right where they apply. The substance is genuinely effective and genuinely safe for the bulk of the hypertensive population that doesn't have advanced CKD, isn't on an MRA, and doesn't have a baseline K⁺ > 5. It is genuinely dangerous in the population the brief flags — CKD G3b+, MRAs, dual RAAS blockade, type-4 RTA — and the harm is fast, asymptomatic, and lethal at the tail. The article must do two things: (1) tell the uncomplicated reader to switch, because the BP and stroke math is overwhelmingly favorable; (2) name the contraindication zone explicitly and concretely, not with a vague "talk to your doctor" wave. The framing is decision-support, not promotion. evidence rates as 5 (multi-trial RCT plus multiple meta-analyses, guideline-aligned); controversy rates as 2 — there's residual debate about generalization and about how aggressively to extend to higher-risk groups, but the core finding is not contested.

Stakeholder + incentive map

• Pushing for substitution: WHO and national CV-prevention bodies (population NaCl reduction is the cheapest cardiovascular intervention available); cardiology societies (AHA 2021 dietary guidance, 2025 ACC/AHA hypertension update endorse substitutes with a CKD caveat); George Institute for Global Health (ran SSaSS, advocates for population-scale rollout); a small commercial substitute industry (LoSalt etc., but margins are thin, not a major lobbying force).
• Pulling back: Nephrology societies (KDIGO 2024 — favorable to potassium-rich diet at population health but explicit on CKD individualization); pharmaceutical RAAS-inhibitor manufacturers (no commercial conflict — the drugs and the substitute are complements, not substitutes); processed-food industry (slow on reformulation but doesn't directly oppose substitutes); clinicians who have witnessed iatrogenic hyperkalemia (informal but powerful brake on individual prescribing).
• Misaligned: Public-health messaging in industrial countries struggles because the discretionary-salt fraction is small, so the apparent home effect is modest, while the same messaging that works for population-level cooking-pot users is dangerous for the multi-morbid drug-stack patient.

Population variability

BP-lowering effect modifiers: baseline SBP (larger drop at higher baseline), age (larger in older adults), ethnicity (slightly larger in Black populations in stand-alone potassium trials), discretionary-salt fraction (much larger where the home shaker dominates dietary Na). Hyperkalemia risk modifiers: eGFR (the dominant axis), RAAS blockade class and intensity (monotherapy < MRA < dual), diabetes (type-4 RTA), HFrEF (background MRA use), age, baseline K, NSAIDs, trimethoprim, calcineurin inhibitors. Pregnancy: not formally studied; KCl substitutes are presumed safe but the BP literature in pregnant women is thin. Children: no specific safety data beyond extrapolated KCl tolerance.

Knowledge gaps

• No dedicated RCT of salt substitute in CKD G3b–G5; the population most needing BP reduction is the one most excluded from the trials that proved the BP benefit. A small Australian trial (SUCCESS) and a pilot in dialysis patients exist but are not powered for outcomes.
• No head-to-head trial of substitute vs. DASH for BP / stroke; the two interventions are presumed additive but unproven in combination.
• Long-term (10+ year) outcome data outside SSaSS is absent; whether the stroke benefit persists, attenuates, or grows is unknown.
• Magnesium contribution — many substitutes carry MgCl₂ or MgSO₄; the BP contribution of Mg is poorly separated from K in trial reporting.
• Iodine — KCl-rich substitutes that aren't iodine-fortified risk reversing decades of population iodine sufficiency; surveillance has not caught up.
• Generalization to industrial diets where < 25% of Na is discretionary remains modeled, not measured.

Scope vs. brief. The brief named four consequences (BP, stroke risk, dietary K intake, hyperkalemia risk) and the four-population context (ACE/ARB/MRA/CKD). All four consequences land in the article: BP in mechanism, evidence, payoff; stroke risk in evidence, stakes; dietary K intake throughout (the "banana per teaspoon" anchor); hyperkalemia risk centered in contraindications and misconceptions. The action token decide rather than do is the deliberate framing — the entry's distinctive value is the screening logic, not a universal "switch the shaker" call.

Action choice (do vs. decide). Genuinely close. For the bulk of mild-to-moderate hypertensives the right action is do; for the brief's named subpopulation it's avoid. I picked decide because the entry's editorial center is the screening question — not because the typical reader can't act on it. A reviewer who'd push back toward do has a defensible argument; if site policy is to score the modal reader's action, change to do and reframe the highlights.

Rating difficulties.

• longevity at 4 not 5: SSaSS is one trial, and the effect size in industrial diets is plausibly smaller because the discretionary-salt fraction is ~25% vs. ~70% in rural China. A 12% all-cause mortality reduction is genuinely huge for a dietary intervention, but its replicability in Western diets is modeled, not measured. The 4-vs-5 call leans skeptic.
• health_short_term at 3: the 3–7 mm Hg systolic drop is "clear functional improvement" but mostly invisible to the reader unless they cuff at home. It's not transformative wellness — sat at 3 rather than 4 because the felt experience is mostly absent. The exception (SSaSS showed a measurable reduction in headache frequency) is a side note, not the dimension's center.
• beauty_cumulative at 0: vascular health does affect long-term appearance via skin perfusion, but attributing it specifically to a KCl-for-NaCl swap (vs. baseline BP control by any means) is over-reach. Better at 0.
• controversy at 2 not 3: the core BP and stroke findings are not contested by anyone running RCTs. The residual debate (CKD generalization, industrial-diet generalization) is real but quiet. If the field publishes a major trial that fails to replicate in a Western cohort, this should move to 3.

Contraindications token mapping. The closed list doesn't have a perfect token for "on a potassium-sparing diuretic or MRA," which is the highest-risk group. I used kidney-disease (captures CKD G3b+) and cardiac-condition (captures heart failure on MRAs as the canonical high-risk medication context). The MRA-without-heart-failure case (resistant hypertension, hormonal indications) is named in the contraindications body but isn't represented in the meta tokens; the schema should arguably gain a potassium-sparing-medication or raas-blockade token. Flagging for the schema-owner backlog.

Excluded on purpose.

• Dialysis-specific guidance — separate population, requires nephrology-level dosing, would dilute the entry's edge.
• Pregnancy and pediatrics — thin evidence base; the substitute is presumed safe in pregnancy but no outcome trials exist. Better to leave silent than to under-justify.
• Detailed Mg / amino-acid / iodine reformulation comparisons across brands — quickly obsoletes as products reformulate; the protocol callout's "25–50% KCl, check iodine" rule of thumb is the load-bearing fragment.
• The agronomy / supply-chain literature on KCl — out of scope; food-cost level is what matters to the reader.

Future-link candidates. When these entries exist, this one should cross-link them: DASH eating pattern, Home blood pressure monitoring, eGFR and routine kidney monitoring, ACE inhibitors and ARBs (overview entry), Mineralocorticoid receptor antagonists, Hyperkalemia recognition and emergency response, Iodine sufficiency. The related field is left empty for now; populate when these IDs are live.

Separate-entry candidates surfaced while writing. The MRA class itself (spironolactone / eplerenone / finerenone) warrants its own entry — life-extending in HFrEF, contested in resistant hypertension, the source of the highest-impact interaction risk here. Iodine sufficiency deserves a standalone entry given the substitute-driven iodine-deplete cooking risk.

The reassuring SSaSS RAAS sub-analysis (Yin 2025) is named in contraindications but the article holds the cautious line anyway, because the sub-analysis is underpowered for the rare-but-lethal event the entry is built around. If a Western-population RCT in CKD G3b+ or MRA users reports null, soften the language; until then, hold.