Salt, Sugar, and Sweeteners

TL;DR. Most of the salt you eat was already in the food when you bought it, so the salt cellar is not the lever. All sugar is metabolically the same regardless of source: honey, agave, coconut sugar, and raw cane sugar are sugar with trace minerals and better marketing. Non-nutritive sweeteners are, on the weight of evidence, safe at realistic intakes and are better than sugar for teeth and blood glucose; the WHO's 2023 recommendation against using them for weight control was about their failure to help, not about toxicity. Erythritol is the one with a genuinely unresolved cardiovascular signal.

Salt

1. What it is

Sodium chloride. The physiologically active part is sodium; chloride matters less. Conversion:

$$\text{salt (g)} \times 0.4 = \text{sodium (g)} \qquad\qquad \text{sodium (g)} \times 2.5 = \text{salt (g)}$$

So 6 g of salt is 2.4 g of sodium. Labels use one or the other and rarely both, which causes constant confusion.

Requirement is around 500 mg of sodium a day. WHO recommends under 2,000 mg (5 g salt). Actual intakes average 3,000 to 5,000 mg in most countries.

2. Where it actually comes from

Roughly 70 to 80 percent of sodium intake in industrialised countries comes from processed and restaurant food, not from cooking or the table.

The largest single contributor in the UK and US is usually bread, not because it is salty but because of quantity. Then processed meat, cheese, sauces, soups, and ready meals.

This changes what works. Removing the salt cellar addresses perhaps 5 percent of intake. Reformulation by manufacturers, and cooking from raw ingredients, address the 70 to 80 percent. The UK's voluntary salt reduction programme, which gradually lowered targets across food categories from 2003, achieved a population intake reduction of around 15 percent and is associated with falls in blood pressure and cardiovascular mortality. It is one of the better public health nutrition case studies.

3. Blood pressure, and the honest state of the argument

Well established: reducing sodium lowers blood pressure. The DASH-Sodium trial, which fed participants controlled diets at three sodium levels, showed a clear dose-response. Meta-analyses of trials find larger effects in hypertensive people and in those who are salt-sensitive (more common with age, in Black populations, and in people with kidney disease or diabetes).

Population evidence: Finland's decades-long salt reduction programme is associated with substantial falls in stroke and coronary mortality. Japan's earlier reduction similarly.

The contested part: some prospective cohorts, notably PURE, report a J-shaped curve, where very low sodium intake also associates with higher mortality. The methodological criticisms are substantial: single spot urine samples estimate intake poorly, and reverse causation (sick people being told to cut salt) is hard to exclude. The mainstream position, that most populations consume far more sodium than needed and that reducing it lowers blood pressure and cardiovascular events, survives this challenge.

Salt substitutes are the most promising intervention: replacing part of the sodium chloride with potassium chloride tackles both sides of the sodium-potassium ratio at once. The SSaSS trial in rural China (over 20,000 people) found a substituted salt reduced stroke, cardiovascular events, and total mortality. Contraindicated in advanced kidney disease and with potassium-sparing drugs, where the potassium load is dangerous.

4. The salt varieties question

Sea salt, Himalayan pink, kosher, fleur de sel, and table salt are all essentially sodium chloride. Trace mineral differences are nutritionally negligible: the iron in pink salt would require eating a toxic quantity of salt to matter.

Two things do genuinely differ:

• Crystal size and shape change how salt dissolves and how it tastes on the tongue, which is why flaky finishing salts are worth using at the table. By volume they are less salty than table salt (a teaspoon of flaky salt contains less sodium than a teaspoon of fine salt), which matters in recipes.
• Iodine. Table salt is often iodised; speciality salts almost never are. Switching from iodised table salt to pink Himalayan salt for perceived health benefits removes an iodine source while delivering identical sodium, and this is a genuine and growing problem in several countries (Chapter 16).

5. Practical salt reduction

• Cook from raw ingredients. This alone addresses the majority of intake.
• Read labels: over 1.5 g salt (0.6 g sodium) per 100 g is high; under 0.3 g (0.1 g sodium) is low.
• Use acid, heat, and aromatics to replace salt: lemon, vinegar, chilli, garlic, herbs, and black pepper all raise perceived intensity (Chapter 49).
• Salt taste adapts over roughly 8 to 12 weeks. Food that tastes bland at first stops doing so, and previously normal food starts tasting too salty.
• Salt at the end of cooking, on the surface, gives more perceived saltiness per gram than salt dissolved throughout.
• Consider a potassium salt substitute, unless you have kidney disease.

Sugar

6. All sugar is the same sugar

Sucrose is glucose plus fructose, and your gut splits it in seconds regardless of where it came from. The metabolic consequences depend on the dose and what it arrives with, not on the source (Chapter 11).

Free sugars is the category that matters: added sugars plus those in honey, syrups, and fruit juice, excluding sugars inside intact fruit and in milk. WHO recommends under 10 percent of energy, with a conditional suggestion of under 5 percent, which is about 25 g (six teaspoons) a day for an average adult. One standard can of soft drink exceeds that.

7. What excess sugar actually does

• Dental caries. The most direct and least disputed harm, and the reason WHO's 5 percent suggestion exists. Frequency matters as much as amount, because each exposure produces a 20 to 30 minute acid attack.
• Sugar-sweetened beverages are the specific problem in the epidemiology, consistently associated with weight gain, type 2 diabetes, and cardiovascular disease. Liquid calories are poorly compensated for at subsequent meals.
• Liver fat and triglycerides, at high fructose intakes, through unregulated de novo lipogenesis.
• Displacement of more nutritious food.

Sugar does not cause hyperactivity in children. This is one of the best-tested and most thoroughly negative findings in nutrition: double-blind trials, including a widely cited meta-analysis by Wolraich and colleagues, find no effect on behaviour. Parents who were told their child had received sugar rated their behaviour as more hyperactive even when they had not. The belief is robust and wrong.

"Sugar is addictive" is contested. Animal models of intermittent sugar access show addiction-like behaviours; in humans, the evidence for sugar as a substance of addiction is weak, and the better-supported framing is that highly palatable food combinations (sugar plus fat plus salt) drive overconsumption.

Sweeteners

8. The non-nutritive ones

Sugar alcohols (polyols): xylitol, erythritol, sorbitol, maltitol, isomalt. Partially absorbed, lower in calories, do not feed oral bacteria, and cause osmotic diarrhoea and bloating in quantity. Xylitol is extremely toxic to dogs, causing hypoglycaemia and liver failure at small doses.

9. The safety question, honestly

Aspartame is among the most tested food additives in existence. Regulatory reviews by EFSA (2013, a full re-evaluation), the FDA, and others have consistently concluded it is safe at the ADI. In July 2023, IARC classified aspartame as Group 2B, "possibly carcinogenic to humans", while the joint FAO/WHO expert committee JECFA reaffirmed the acceptable daily intake of 40 mg/kg on the same day.

Both statements are consistent, and the pairing is instructive. IARC assesses hazard: whether something can cause cancer under some conditions, and Group 2B is the weakest positive category, containing aloe vera extract and pickled vegetables. JECFA assesses risk at realistic exposures. A 70 kg adult would need to drink roughly 9 to 14 cans of diet drink a day to reach the ADI (Chapter 97).

Erythritol is the genuine open question. A 2023 Nature Medicine study by Witkowski and colleagues at the Cleveland Clinic found high circulating erythritol associated with major adverse cardiovascular events in three cohorts, and showed in laboratory work that erythritol enhanced platelet aggregation and thrombosis. Erythritol is also produced endogenously, which complicates interpretation, and the observational component cannot establish causation. It is being actively investigated and it is currently the sweetener with the least reassuring evidence.

The WHO's 2023 guideline recommended against using non-sugar sweeteners for weight control. This was widely misreported as a safety warning. The reasoning was that they do not produce long-term weight loss and that observational data associate long-term use with higher risk of type 2 diabetes and cardiovascular disease, with the guideline explicitly noting very low certainty and the strong possibility of reverse causation, since people at higher risk choose diet drinks. The recommendation was labelled "conditional."

Sweeteners and the microbiome is an active area. Some studies, notably Suez and colleagues in 2014 and 2022, found that certain sweeteners altered gut microbiota and glucose tolerance in some individuals. Effects are inconsistent, individual, and not established as clinically meaningful.

10. Where this leaves you

• Water, unsweetened tea, and coffee are the best default. Nobody disputes this.
• For someone drinking several sugary drinks a day, switching to diet versions is a clear improvement, for teeth, for calories, and for blood glucose. That is the comparison that matters, not sweetener versus water.
• Do not expect sweeteners to produce weight loss on their own. They remove calories from one place; behaviour determines whether those calories return elsewhere.
• Xylitol is genuinely good for teeth and is used in dental chewing gum for that reason.
• If you have phenylketonuria, avoid aspartame absolutely.
• Sugar alcohols in quantity cause diarrhoea. Sugar-free sweets carry warnings for a reason.

11. The bottom line

• Most of your salt was already in the food. Reformulation and home cooking are the levers; the salt cellar is not. Speciality salts are chemically the same and are not iodised.
• Salt substitutes containing potassium chloride reduced stroke and mortality in a large randomised trial, and are dangerous in advanced kidney disease.
• All sugars are metabolically equivalent. Honey, agave, coconut sugar, and raw sugar are sugar with better marketing; agave is the highest in fructose. Aim for under 25 to 50 g of free sugars a day.
• Sugar does not cause hyperactivity in children. That has been tested repeatedly and negatively.
• Non-nutritive sweeteners are safe at realistic intakes on the weight of regulatory evidence. The 2023 aspartame headline was a hazard classification, not a change in the safe intake.
• Erythritol has a real, unresolved cardiovascular signal and is the one to watch.
• Switching from sugary drinks to sweetened ones is an improvement; switching to water is better.

Sources and notes

Sodium sources and the 70 to 80 percent from processed food follow national dietary surveys and Mattes and Donnelly's discretionary salt work. DASH-Sodium is Sacks et al., NEJM, 2001. The UK salt reduction programme's outcomes follow He, Pombo-Rodrigues, and MacGregor, BMJ Open, 2014. Finland's programme follows Karppanen and Mervaala's account. The J-curve controversy follows PURE (Mente et al.) and its critiques regarding spot urine estimation. SSaSS is Neal et al., NEJM, 2021. Iodisation of table salt versus speciality salts follows Iodine Global Network reporting. WHO free sugar guidance is the 2015 guideline. Sugar and hyperactivity follows Wolraich, Wilson, and White's meta-analysis, JAMA, 1995. Sweetener ADIs and safety assessments follow EFSA's 2013 aspartame re-evaluation and JECFA evaluations. The July 2023 IARC Group 2B aspartame classification and the simultaneous JECFA reaffirmation of the 40 mg/kg ADI were published together by WHO. The saccharin rat bladder mechanism and 2000 delisting follow the US National Toxicology Program. Erythritol and cardiovascular events is Witkowski et al., Nature Medicine, 2023. The WHO 2023 non-sugar sweetener guideline, and its explicit very-low-certainty caveat, is published by WHO. Sweeteners and glucose tolerance follow Suez et al., Nature, 2014, and Cell, 2022.

Open questions. The sodium J-curve and the erythritol signal are the two genuinely unresolved questions here, and both turn on whether observational associations reflect causation or reverse causation. Sweetener effects on the microbiome are inconsistent between studies and individuals.

👉 Next: ultra-processed food, the category argument that has dominated nutrition for a decade.