Salt and sugar

In 2019 a team went back to 2,974 American adults from a pair of blood pressure trials run between 1987 and 1995, people who hadn't been assigned to cut their salt.¹² Each had collected all their urine over a full day, three to seven times, and over a median of 24 years 272 of them died. Measured that way, sodium and death rose together in a straight line. Then the team kept only the concentration of sodium in the urine and ran it through three published formulas, the kind built to estimate a day's intake from a single spot sample. With the estimates in place of the measurements, every formula appeared to draw a J or a U, in which the people with the lowest estimates looked to be at higher risk.¹ Same people, same deaths, different ruler. This lesson is about salt and sugar, where an effect on the body is well shown and the argument is over what lies past it, and where part of the salt dispute turns on whether a curve came from the food or from the ruler.

Two disputes with one shape

Lesson 6 split "fat is bad for your heart" into claims of different strength. Salt and sugar split the same way.

Both sides of the salt dispute accept the first row, the dissent in its own words below. In the sugar dispute it's the critics' ground, and the harm side's claim lies elsewhere. Lesson 1's "instead of what?" does much of the sorting.

What salt does to blood pressure

Labels mix units, so start there. Salt is sodium chloride, and the evidence is about the sodium. WHO's sodium fact sheet, dated 11 May 2026 and read in full for this course, gives the conversion: "1 gram of salt is equivalent to 400 mg of sodium".³ Multiply grams of salt by 400 for milligrams of sodium, or divide the other way. WHO's adult target is "less than 2000 mg/day of sodium (equivalent to less than 5 g/day salt), or just under a teaspoon", and "The global mean intake of adults was 4278 mg/day sodium (equivalent to 11 g/day salt) in 2021".³ That's more than twice the target.

The cleanest test is a feeding trial. In DASH-Sodium, published by Sacks and colleagues in 2001, 412 people ate food provided for them, either a control diet "typical of intake in the United States" or the DASH diet, "rich in vegetables, fruits, and low-fat dairy products", at high, intermediate and low sodium "for 30 consecutive days each, in random order".⁴ This course read the abstract. Systolic blood pressure, the higher of the two numbers, measured in millimetres of mercury (mm Hg), fell by 2.1 mm Hg on the control diet from the high level to the intermediate one, and by a further 4.6 from intermediate to low. On DASH the falls were 1.3 and 1.7.⁴

So the effect didn't stop at the middle level, and sodium did less on a diet already rich in vegetables and fruit: what you cut salt from changes what cutting it does. Thirty days a level is a test of blood pressure, not of strokes.

For longer periods, He, Li and MacGregor pooled 34 randomised trials of at least four weeks, 3,230 people, in a 2013 Cochrane review in the BMJ; this course read the abstract.⁵ Cutting salt by an amount "equivalent to a reduction of 4.4 g/day salt" lowered systolic pressure by "-4.18 mm Hg (95% confidence interval -5.18 to -3.18, I(2)=75%)". The I², from lesson 1, says the trials differed a lot, and bigger cuts brought bigger falls.⁵

Blood pressure is the step between salt and stroke, as LDL was between saturated fat and heart disease in lesson 6. The review's conclusion: "These results support a reduction in population salt intake, which will lower population blood pressure and thereby reduce cardiovascular disease."⁵ The US National Academies (NASEM), in a 2019 report whose Summary this course read, found "moderate to high strength of evidence for both a causal relationship and an intake–response relationship between sodium and several interrelated chronic disease indicators: cardiovascular disease, hypertension, systolic blood pressure, and diastolic blood pressure."⁶ The dissent describes the guidelines as resting "on the premise that reductions in sodium intake, irrespective of the levels, will lower blood pressure, and, in turn, reduce cardiovascular disease occurrence."⁷ Its quarrel is with "irrespective of the levels". This course didn't read the physiology of how sodium raises blood pressure, so it doesn't teach it.

The trial that counted strokes

In 2020 O'Donnell's group, the main dissenters in this course's sources, wrote that low-sodium guidelines lacked "high-quality evidence that low sodium intake reduces cardiovascular events (compared with moderate intake)".⁷ The next year Neal and colleagues published SSaSS in the New England Journal of Medicine; this course read the abstract.⁸ It randomised 600 villages in rural China. Its 20,995 people "had a history of stroke or were 60 years of age or older and had high blood pressure". Some villages used "a salt substitute (75% sodium chloride and 25% potassium chloride by mass)"; the others kept regular salt. The mean age was 65.4, "72.6% had a history of stroke, and 88.4% a history of hypertension", and follow-up averaged 4.74 years.⁸

Results come as rates per 1,000 person-years, lesson 4's unit. Stroke: "29.14 events vs. 33.65 events per 1000 person-years; rate ratio, 0.86; 95% confidence interval [CI], 0.77 to 0.96; P = 0.006".⁸ A rate ratio is one rate divided by the other: 14 percent fewer strokes, the figure a headline prints.

The absolute figure takes a subtraction, and the arithmetic from here is this course's, not the paper's. 33.65 minus 29.14 is 4.51 fewer strokes per 1,000 person-years. Over an average 4.74 years, that's very roughly 21 fewer strokes per 1,000 people (4.51 × 4.74, treating the rate as steady, which it needn't have been). Now do the deaths: "39.28 events vs. 44.61 events per 1000 person-years; rate ratio, 0.88; 95% CI, 0.82 to 0.95; P<0.001".⁸

Major cardiovascular events ran 49.09 against 56.29 per 1,000 person-years, "rate ratio, 0.87", so 13 percent fewer and 7.20 fewer in absolute terms.⁸ The safety outcome, serious events put down to high blood potassium, hyperkalaemia, was "3.35 events vs. 3.30 events per 1000 person-years; rate ratio, 1.04; 95% CI, 0.80 to 1.37; P = 0.76".⁸ That's 0.05 more, with an interval wide both ways: no clear increase here, and no proof there's none. The abstract doesn't say who the trial left out, so this can't be carried to people with kidney disease or on drugs that raise potassium. The chart puts the four pairs side by side.

What the substitute changed

Before this becomes "cutting salt prevents strokes", three limits. The biggest: the substitute swapped a quarter of the sodium chloride for potassium chloride, lowering sodium and raising potassium at once, so SSaSS can't say how much each did. Instead of what? Regular salt was replaced by a salt with an active ingredient of its own. Second, the people were older, mostly past a stroke, nearly all with high blood pressure. Third, it was open-label: everybody knew which salt they had.

The second matters because, as lesson 6 showed with Steen's review, the same relative effect is worth less at lower risk. Healthy adults in a 2022 study, met below, had cardiovascular events at "5.9 per 1000 person-years", against SSaSS's 56.29.⁹ Events were defined differently, so this is rough, but if SSaSS's 13 percent held at 5.9, the gap would be about 0.8 per 1,000 person-years: this course's arithmetic, and only a hypothesis.

WHO now advises the swap: "If choosing to use table salt, WHO suggests replacing regular table salt with lower-sodium salt substitutes that contain potassium."³ This course couldn't open the 2025 guideline behind that and read a summary from the George Institute, which ran SSaSS, an interested party.¹⁰ After reporting no increased hyperkalaemia "amongst the general population", it adds: "There is, however, a rationale for increased risk of hyperkalaemia if lower-sodium salt substitutes that contain potassium are consumed in excess among people with chronic kidney disease, using other potassium supplements or potassium-sparing diuretics."¹⁰ That's one reason kidney disease is in the callout above, and whether a medicine you take rules the swap out is a question for a doctor or pharmacist.

Why the mainstream wants lower

The mainstream case starts with He, Li and MacGregor's review, which ends: "The current recommendations to reduce salt intake from 9-12 to 5-6 g/day will have a major effect on blood pressure, but a further reduction to 3 g/day will have a greater effect and should become the long term target for population salt intake."⁵ Three grams of salt is 1,200 mg of sodium, by this course's conversion, well under WHO's 2,000.

NASEM's 2019 report set sodium's adequate intake, lesson 5's AI, at 1,500 mg a day for adults, and created something new: "the report establishes the first Chronic Disease Risk Reduction Intake (CDRR) level".⁶ For adults the CDRR reads "Reduce intakes if above 2,300 mg/day". Why 2,300 and not lower: "The sodium CDRR is the lowest level of intake for which there was sufficient strength of evidence to characterize a chronic disease risk reduction. Further reductions in sodium intake below the CDRR have demonstrated a lowering effect on blood pressure, but the effect on chronic disease risk could not be characterized."⁶ And on harm: across the trials no deficiency symptoms were reported, and "there was insufficient evidence that low sodium intakes are associated with other potential harmful health effects."⁶ So NASEM stops at 2,300 because the evidence below wasn't good enough, and found too little evidence of harm from going lower to say there is any.

The Trials of Hypertension Prevention (TOHP), the pair from the opening, randomised people to cut sodium or not, and Cook, Appel and Whelton followed deaths for a median of 24 years; this course read the abstract.² Randomised, the sodium-cutting groups had "a nonsignificant 15% lower risk" of death (hazard ratio 0.85, 0.66 to 1.09). Among those not assigned to cut sodium, measured by repeated 24-hour collections, the hazard ratio was "1.12 per 1,000 mg/24 h (95% CI: 1.00 to 1.26; p = 0.05)", though by intake bands the trend was weaker, "p trend = 0.30", with the lowest band, under 2,300 mg, at a hazard ratio of 0.75 and no interval given. "There was no evidence of a J-shaped or nonlinear relationship." Their conclusion: a direct relationship with total mortality "even at the lowest levels of sodium intake."² Read at their own strength, the randomised result isn't significant and the trend's interval touches 1.00: supporting, not proving.

The PURE group's case

The dissent's evidence starts with a cohort called PURE. In a 2014 paper by O'Donnell and colleagues, 101,945 people in 17 countries gave a morning urine sample, from which the team "estimated 24-hour sodium and potassium excretion (used as a surrogate for intake)"; 3,317 died or had a major cardiovascular event over 3.7 years.¹¹ This course read the abstract.

Against a reference range of 4.00 to 5.99 g of sodium a day, 7 g or more went with more risk, an odds ratio of 1.15 (1.02 to 1.30). (An odds ratio is another way of comparing risk; for an outcome this uncommon it comes out close to a risk ratio, in this course's gloss.) So did intake below 3.00 g, an odds ratio of 1.27 (1.12 to 1.44). The conclusion: "an estimated sodium intake between 3 g per day and 6 g per day was associated with a lower risk of death and cardiovascular events than was either a higher or lower estimated level of intake."¹¹ Higher potassium went with lower risk, a finding that comes back below. Three grams of sodium is 7.5 g of salt, by this course's conversion, so on PURE's reading WHO's target sits inside the range that went with more risk.

In 2016 Mente and colleagues pooled PURE with three other studies, 133,118 people in 49 countries, and split them by blood pressure, treating the estimate, in their words, "as group-level measure of intake"; this course read the abstract.¹² Their data showed the blood pressure link, in systolic pressure: "2.08 mm Hg change per g sodium increase" with hypertension, "1.22 mm Hg change per g" without. With hypertension both ends of intake carried risk. Without it, high intake showed no association, "HR 0.90 [95% CI 0.76-1.08]", while low intake still did, "HR 1.26 [95% CI 1.10-1.45]". Their reading: lowering sodium "is best targeted at populations with hypertension who consume high sodium diets."¹²

The group's fullest statement is a 2020 review in the European Heart Journal, signed by 24 authors and titled "Salt and cardiovascular disease: insufficient evidence to recommend low sodium intake"; this course read the abstract.⁷ It says the guidelines "have been developed without effective interventions to achieve sustained low sodium intake in free-living individuals, without a feasible method to estimate sodium intake reliably in individuals, and without high-quality evidence that low sodium intake reduces cardiovascular events (compared with moderate intake)."

Only the third is about the curve; the first is about whether people can keep to a low target, and the second turns the measurement problem back on the guidelines. The review adds that a low target is "of uncertain effect on other dietary factors": "instead of what?", asked by the dissent. Their positive claim: "We contend that current evidence, despite methodological limitations, suggests that most of the world's population consume a moderate range of dietary sodium (2.3-4.6g/day; 1-2 teaspoons of salt) that is not associated with increased cardiovascular risk, and that the risk of cardiovascular disease increases when sodium intakes exceed 5 g/day."⁷ WHO's global mean of 4,278 mg sits inside that moderate range.

And then the concession, whole: "While current evidence has limitations, and there are differences of opinion in interpretation of existing evidence, it is reasonable, based upon observational studies, to suggest a population-level mean target of <5 g/day in populations with mean sodium intake of >5 g/day, while awaiting the results of large randomized controlled trials of sodium reduction on incidence of cardiovascular events and mortality."⁷ The dissent doesn't call salt harmless; it calls the case for going low unproven and possibly wrong.

How a urine sample can draw a J

PURE's J rests on sodium estimated from a single urine sample by formula. He and colleagues' 2019 analysis tested what such formulas do, on the TOHP people whose intake had also been measured properly; this course read the abstract.¹ It doesn't name PURE, and this course didn't read a reply from the PURE group, so the link between the two is this course's. PURE's own framing, met above, is a group-level measure, a claim about averages, while He 2019 tested each person's estimate. Whether that answers the critique is part of the dispute.

First, the estimates were wrong in a particular direction: "All estimated values, including those with constant sodium concentration, were systematically biased with overestimation at lower levels and underestimation at higher levels." Low eaters were put higher than they were, high eaters lower. Then the shape: "There was a significant linear association between the average measured sodium intake (ie, gold standard method) and mortality. This relationship was altered by using the estimated sodium intakes. There appeared to be a J- or U-shaped relationship for the average estimated sodium by all formulas."¹

The telling step is "constant sodium concentration". The team held the sodium reading fixed, let the formulas run anyway, and found that "all estimated values including those with constant sodium appeared to be inversely related to mortality at lower levels of sodium intake." Their conclusion: "These results demonstrate that inaccurate estimates of sodium cannot be used in association studies, particularly as the formulas per se seem to be related to mortality independent of sodium."¹

What that means, in this course's reasoning from the abstract: if the sodium reading is fixed and the estimate still varies, the formula must be built partly from other information about the person. The abstract doesn't say what, and this course didn't read the formulas. If that information is linked to health, the estimate carries a link to death that has nothing to do with salt, and the abstract says it shows at the low end. Lesson 2 drew the line that matters: error that's just noise weakens an association, and error tied to the outcome can make one.

What the dissent still has

Showing that a method can draw a J doesn't show that low intakes are safe or beneficial; it weakens one line of evidence for harm. The dissent's other points stand, and several come from mainstream sources. No trial in this course's sources compares low with moderate sodium: SSaSS tested a sodium and potassium swap in high-risk people. NASEM's line stops at 2,300 mg for want of evidence below it.⁶ The better cohorts have few people low down: in Ma's study only about one person in ten was below 2.1 g.⁹ The randomised TOHP result was null.² And feasibility, whether free-living people can keep to low targets, isn't about urine at all; DASH-Sodium's own authors wrote that long-term benefit "will depend on the ability of people to make long-lasting dietary changes".⁴

The mainstream's answer is Cook's direct relationship "even at the lowest levels of sodium intake", a trend whose banded version wasn't significant, and NASEM's finding of insufficient evidence of harm from low intakes in the trials.²⁶ What would settle it, in this course's summary of its research, is a trial of low sodium, under about 2 g, against moderate, about 3 to 4.5 g, counting strokes, heart attacks and deaths, with adherence checked by repeated full-day collections: the trial the dissenters say they're waiting for.

As of September 2026 the Dietary Guidelines for Americans, 2025–2030, which lesson 9 reads, put the general population "ages 14 and above" at "less than 2,300 mg per day", NASEM's CDRR.¹³

Sugar: two lines, and what each rests on

WHO's 2015 sugar guideline came with a news release, which this course read in full; the guideline itself wasn't retrieved.¹⁴ "A new WHO guideline recommends adults and children reduce their daily intake of free sugars to less than 10% of their total energy intake. A further reduction to below 5% or roughly 25 grams (6 teaspoons) per day would provide additional health benefits." Free sugars are those "added to foods and drinks by the manufacturer, cook or consumer, and sugars naturally present in honey, syrups, fruit juices and fruit juice concentrates", not those in whole fruit, vegetables or milk.¹⁴

The two lines carry different grades. The 10 percent line is "ranked by WHO as “strong”. This means they can be adopted as policy in most situations." The 5 percent line: "Given the nature of existing studies, the recommendation of reducing intake of free sugars to below 5% of total energy is presented as “conditional” in the WHO system for issuing evidence-based guidance."¹⁴

What each rests on is what people get wrong. Branca of WHO, quoted in the release, on the first: "We have solid evidence that keeping intake of free sugars to less than 10% of total energy intake reduces the risk of overweight, obesity and tooth decay". The second rests on "Only three national population-wide studies", "population-based ecological studies" from a time when sugar availability fell "from 15kg per person per year before the Second World War to a low of 0.2kg per person per year in 1946. This “natural experiment”, which demonstrated a reduction in dental caries, provides the basis for the recommendation".¹⁴ Caries is tooth decay. The release doesn't name the country, and this course didn't check.

So the 10 percent line is about weight and teeth, and the 5 percent line about teeth, from wartime data. Neither is a claim about diabetes or heart disease. Why is the second only conditional? This course's reading is "instead of what?" for a whole country: sugar fell during a war, and so did much else.

For scale, "A single can of sugar-sweetened soda contains up to 40 grams (around 10 teaspoons) of free sugars."¹⁴ If 5 percent is roughly 25 g, 10 percent is roughly 50 g, by this course's doubling of WHO's figure. A can at the top of that range takes most of the 10 percent line and all of the 5 percent one.

Sugar and weight: were calories held equal?

Te Morenga, Mallard and Mann's 2013 review in the BMJ pooled 30 trials and 38 cohort studies; this course read the abstract.¹⁵

Some trials were ad libitum, which the abstract defines as "with no strict control of food intake": people were told to eat less sugar, or more, and otherwise ate as they liked. In those trials in adults, "reduced intake of dietary sugars was associated with a decrease in body weight (0.80 kg, 95% confidence interval 0.39 to 1.21; P<0.001); increased sugars intake was associated with a comparable weight increase (0.75 kg, 0.30 to 1.19; P=0.001)."

Others were isoenergetic, also called isocaloric: sugar was swapped for other carbohydrate with calories held equal.

An isocaloric swap holds calories fixed, so whatever changes is down to sugar against the carbohydrate that replaced it. An ad libitum trial lets calories move, so it tests sugar and calories together, the way people eat. Both are useful, for different questions. So of any sugar trial ask: instead of what? Other carbohydrate at equal calories, or nothing?

Malik and colleagues pooled cohorts and trials of sugar-sweetened drinks in 2013; this course read the abstract.¹⁶ In adults, randomised trials that added the drinks found "increases in body weight when SSBs were added (random and fixed effects: 0.85 kg; 95% CI: 0.50, 1.20 kg)". (SSBs are sugar-sweetened beverages.) In children, trials that cut the drinks showed less BMI gain, but the random-effects interval crossed zero. A 2010 pooling of eleven cohorts by the same first author found "a 26% greater risk of developing type 2 diabetes" in the top drinkers, "most often 1-2 servings/day", a cohort association with lesson 1's caveats.¹⁷ Bray and Popkin, arguing the harm side of a 2014 debate, cite a trial in which "Drinking two 16-ounce SSBs per day for 6 months induced features of the metabolic syndrome and fatty liver". Metabolic syndrome, in this course's gloss, is a cluster of risk factors such as raised blood pressure, blood fats and blood sugar. This course read their abstract and not that trial.¹⁸

Is sugar special?

Here the agreed ground ends and the dispute begins. In this course's sources the case that sugar harms beyond its calories is made most directly by Lustig. This course read the abstract of his 2013 review, "Fructose: it's 'alcohol without the buzz'", and didn't open his popular book.¹⁹

His case, in his words. Table sugar and high fructose corn syrup "consist of 2 molecules, glucose and fructose." Sugar "has been vilified by nutritionists for ages as a source of "empty calories," no different from any other empty calorie. However, fructose is unlike glucose. In the hypercaloric glycogen-replete state, intermediary metabolites from fructose metabolism overwhelm hepatic mitochondrial capacity, which promotes de novo lipogenesis and leads to hepatic insulin resistance, which drives chronic metabolic disease."

In plain words, and this translation is this course's: when someone is already eating more than they burn and their liver's sugar stores are full, fructose overloads the liver's energy machinery, the liver makes new fat, and it stops responding well to insulin. He adds that fructose "promotes changes in the brain's reward system, which drives excessive consumption", and concludes: "Thus, fructose can exert detrimental health effects beyond its calories and in ways that mimic those of ethanol, its metabolic cousin."¹⁹

So the claim is about liver fat, insulin resistance and metabolic disease, not mainly weight, and it carries its own condition, "In the hypercaloric glycogen-replete state": in people eating more than they need.

His group tested it in a 2016 trial.²⁰ Its stated reason is a criticism of the evidence on the other side: intervention studies "are confounded by positive caloric balance, changes in adiposity, or artifactually high amounts". For nine days, 43 children with obesity and metabolic syndrome ate a diet in which "dietary sugar was reduced from 28% to 10% and substituted with starch", with "calories adjusted for weight maintenance". Blood pressure, blood fats and glucose tolerance improved, and the authors adjusted the blood results for weight change statistically. Weight still fell, "by 0.9 ± 0.2 kg"; in a post hoc look at the ten children who didn't lose weight, results were "directionally consistent". The conclusion: "Isocaloric fructose restriction improved surrogate metabolic parameters in children with obesity and metabolic syndrome irrespective of weight change."²⁰ This course read the abstract. Its limits are visible there: no control group, nine days, weight lost in a trial meant to hold it steady, and a weight-stable subgroup of ten, analysed afterwards.

A third paper, by Basu, Lustig and colleagues, compared 175 countries: "every 150 kcal/person/day increase in sugar availability (about one can of soda/day) was associated with increased diabetes prevalence by 1.1% (p <0.001)", with total calories and obesity controlled.²¹ Availability isn't what anybody ate, so this is ecological evidence, like lesson 6's Seven Countries Study.

The reply, in its own words

The other side of the 2014 debate was Kahn and Sievenpiper, whose abstract, read here, is the critics' best single statement:²² "we argue that there is no clear or convincing evidence that any dietary or added sugar has a unique or detrimental impact relative to any other source of calories on the development of obesity or diabetes. Sugar is purely a highly palatable source of energy; because it has no other property that appears to contribute to our nutritional well-being, it is not an essential food for most of us. For those who wish to reduce energy consumption, ingesting less sugar is a good place to start. However, doing so does not automatically portend any clinical benefit."

Their trial evidence is a 2012 review Sievenpiper led of fructose feeding trials; this course read the abstract.²³ It pooled trials of free fructose; trials of high fructose corn syrup "were excluded". "Fructose had no overall effect on body weight in isocaloric trials (mean difference, -0.14 kg [95% CI, -0.37 to 0.10 kg] for fructose compared with nonfructose carbohydrate)." In hypercaloric trials, which added large doses on top, weight rose by 0.53 kg. The trials "tended to be small (<15 participants), short (<12 weeks), and of low quality", and "The weight-increasing effect of fructose in hypercaloric trials may have been attributable to excess energy rather than fructose itself."²³

A different challenge came in 2017 from Erickson, Johnston and colleagues, who graded the evidence behind nine sugar guidelines: "Guidelines on dietary sugar do not meet criteria for trustworthy recommendations and are based on low-quality evidence."²⁴ This course read the abstract, which names its funder as the "North American branch of the International Life Sciences Institute" and says the authors "conducted the study independent of the funding source, which is primarily supported by the food and agriculture industry." Its senior author, Bradley Johnston, also led the 2019 guideline on meat that lesson 8 reads.

What each side has left

The weight question is the least disputed part: sugar eaten freely adds weight, and swapped for other carbohydrate at equal calories it doesn't add more. Lustig's claim was never mainly about weight, though. It concerns the liver and metabolic disease in people already eating more than they need, and the trials that would test it, long isocaloric trials measuring liver fat or new diabetes, are what this course's research names as missing, along with whether sugar in drinks behaves differently. Until they exist, Kahn and Sievenpiper can say there's "no clear or convincing evidence" of a unique harm, and Lustig can say the evidence they cite tests a different claim.

Sweeteners, and the cancer label

Are sweeteners the answer, then? In 2023 WHO said not, in a guideline that "recommends against the use of NSS to control body weight or reduce the risk of noncommunicable diseases (NCDs)." (NSS are non-sugar sweeteners.) The advice has a stated exception: it "applies to all people except individuals with pre-existing diabetes". If you have diabetes, the callout above applies. Its review "suggests that use of NSS does not confer any long-term benefit in reducing body fat in adults or children", and that there "may be potential undesirable effects from long-term use of NSS, such as an increased risk of type 2 diabetes, cardiovascular diseases, and mortality in adults."²⁵ This course read the release.

WHO grades its own caution: "Because the link observed in the evidence between NSS and disease outcomes might be confounded by baseline characteristics of study participants and complicated patterns of NSS use, the recommendation has been assessed as conditional". One candidate, in this course's reading rather than WHO's words, is that people already at risk switch to diet drinks.²⁵ So the advice amounts to not expecting sweeteners to help, which is different from calling them proven harmful.

The cancer question came in July 2023, when two WHO bodies reported on aspartame the same day: "Citing “limited evidence” for carcinogenicity in humans, IARC classified aspartame as possibly carcinogenic to humans (IARC Group 2B) and JECFA reaffirmed the acceptable daily intake of 40 mg/kg body weight."²⁶ (JECFA is the expert committee that sets that acceptable daily intake.) As Sleep lesson 4 taught, IARC classifications "reflect the strength of scientific evidence as to whether an agent can cause cancer in humans, but they do not reflect the risk of developing cancer at a given exposure level."²⁶

A second body answered the dose question. JECFA "concluded that the evidence of an association between aspartame consumption and cancer in humans is not convincing", and gave a worked example: "with a can of diet soft drink containing 200 or 300 mg of aspartame, an adult weighing 70kg would need to consume more than 9–14 cans per day to exceed the acceptable daily intake, assuming no other intake from other food sources."²⁶ Branca keeps both halves: "while safety is not a major concern at the doses which are commonly used, potential effects have been described that need to be investigated by more and better studies."²⁶

Who paid

This course read funding lines where an abstract printed them, not the disclosure statements behind them.

• Salt, mainstream. He and MacGregor, authors of He 2019 and Ma 2022, lead the advocacy group Action on Salt, a public stake.¹⁹ Ma 2022 was "Funded by the American Heart Association and the National Institutes of Health."⁹ SSaSS was "Funded by the National Health and Medical Research Council of Australia", and the George Institute ran it.⁸¹⁰
• Salt, dissent. PURE was "Funded by the Population Health Research Institute and others."¹¹ This course didn't read the funding or disclosures of Mente 2016 or O'Donnell 2020.
• Sugar. Erickson's review was funded through the International Life Sciences Institute, as above.²⁴ Sievenpiper's 2012 review names its primary funder as the "Canadian Institutes of Health Research."²³ A primary funder isn't the authors' own disclosures, and this course didn't read those for Sievenpiper's review, the 2014 debate or Lustig's papers.

What people get wrong

"Salt only matters if you have high blood pressure." In trials blood pressure fell in people without hypertension too, by less.⁴⁵ What's disputed is whether high salt raises events in that group, where Mente's estimated-sodium data found no link.¹²

"The low-salt advice has been debunked." The main dissenters accept a population target of under 5 g of sodium where intake is higher; their dispute is with targets near 2 g.⁷

"SSaSS proved that cutting sodium prevents strokes." It showed a sodium and potassium salt cut strokes and deaths in older, high-risk villagers, and it can't separate the two minerals.⁸

"Lower sodium is proven better all the way down." NASEM couldn't characterise the effect on disease below 2,300 mg, and the long-term randomised TOHP result was null at conventional significance.²⁶

"Sugar is uniquely fattening." Swapped for other carbohydrate at equal calories, it changed weight by 0.04 kg in Te Morenga's review.¹⁵

"Sugar is just calories, so it doesn't matter." Eaten freely it adds weight, WHO's tooth decay evidence is about it, and Lustig's metabolic claim is unsettled rather than refuted.¹⁴¹⁵¹⁹

"Sweeteners cause cancer." Aspartame is in IARC's Group 2B on limited evidence, and JECFA kept its acceptable intake.²⁶ The mirror, "sweeteners help you lose weight", isn't what WHO's review found either.²⁵

Practice

Connections

Back. Lesson 6 had the same shape, with blood pressure in LDL's place, and He 2019 showed lesson 2's warning that error tied to the outcome can create an association.

Forward. Lesson 8 asks whether "ultra-processed" is a useful category and reads the meat evidence. Lesson 9 returns to DASH as a pattern and reads the 2026 US guidelines.

Go deeper

• WHO's sodium fact sheet, 11 May 2026. Read here: the whole page.
• He and colleagues, 2019, Hypertension. Read here: the abstract.
• The 2014 Diabetes Care debate, Bray and Popkin against Kahn and Sievenpiper. Read here: both abstracts.

Sources

1. F. J. He and colleagues, "Formulas to Estimate Dietary Sodium Intake From Spot Urine Alter Sodium-Mortality Relationship", Hypertension, 2019, doi 10.1161/HYPERTENSIONAHA.119.13117. Read: the abstract.
2. N. R. Cook, L. J. Appel and P. K. Whelton, "Sodium Intake and All-Cause Mortality Over 20 Years in the Trials of Hypertension Prevention", JACC, 2016, doi 10.1016/j.jacc.2016.07.745. Read: the abstract.
3. World Health Organization, "Sodium reduction", fact sheet, 11 May 2026. Read: the whole page.
4. F. M. Sacks and colleagues, "Effects on blood pressure of reduced dietary sodium and the DASH diet", NEJM, 2001, doi 10.1056/NEJM200101043440101. Read: the abstract.
5. F. J. He, J. Li and G. A. MacGregor, "Effect of longer term modest salt reduction on blood pressure", BMJ, 2013, doi 10.1136/bmj.f1325. Read: the abstract.
6. National Academies of Sciences, Engineering, and Medicine, Dietary Reference Intakes for Sodium and Potassium, 2019, doi 10.17226/25353. Read: the Summary chapter, including Table S-2, and the preface passage on the CDRR.
7. M. O'Donnell and 23 co-authors, "Salt and cardiovascular disease", European Heart Journal, 2020, doi 10.1093/eurheartj/ehaa586. Read: the abstract.
8. B. Neal and colleagues, "Effect of Salt Substitution on Cardiovascular Events and Death", NEJM, 2021, doi 10.1056/NEJMoa2105675. Read: the abstract. The absolute differences and per-trial approximations are this course's arithmetic.
9. Y. Ma and colleagues, "24-Hour Urinary Sodium and Potassium Excretion and Cardiovascular Risk", NEJM, 2022, doi 10.1056/NEJMoa2109794. Read: the abstract. The comparison with SSaSS's rates is this course's, and rough.
10. The George Institute for Global Health, a summary for policy makers of WHO's Use of lower-sodium salt substitutes (2025), PDF. Read: the whole summary. The WHO guideline itself was not read.
11. M. O'Donnell and colleagues, "Urinary sodium and potassium excretion, mortality, and cardiovascular events", NEJM, 2014, doi 10.1056/NEJMoa1311889. Read: the abstract. The odds ratio gloss is this course's.
12. A. Mente and colleagues, "Associations of urinary sodium excretion with cardiovascular events", Lancet, 2016, doi 10.1016/S0140-6736(16)30467-6. Read: the abstract.
13. US Departments of Health and Human Services and of Agriculture, Dietary Guidelines for Americans, 2025–2030, 2026, PDF. Read: the whole document.
14. World Health Organization, news release on sugars intake, 4 March 2015. Read: the whole release. The guideline itself was not retrieved.
15. L. Te Morenga, S. Mallard and J. Mann, "Dietary sugars and body weight", BMJ, 2013, doi 10.1136/bmj.e7492. Read: the abstract.
16. V. S. Malik and colleagues, "Sugar-sweetened beverages and weight gain in children and adults", American Journal of Clinical Nutrition, 2013, doi 10.3945/ajcn.113.058362. Read: the abstract.
17. V. S. Malik and colleagues, "Sugar-sweetened beverages and risk of metabolic syndrome and type 2 diabetes", Diabetes Care, 2010, doi 10.2337/dc10-1079. Read: the abstract.
18. G. A. Bray and B. M. Popkin, "Dietary sugar and body weight ... health be damned! Pour on the sugar", Diabetes Care, 2014, doi 10.2337/dc13-2085. Read: the abstract. The trial it cites was not read.
19. R. H. Lustig, "Fructose: it's 'alcohol without the buzz'", Advances in Nutrition, 2013, doi 10.3945/an.112.002998. Read: the abstract. The plain-words translation is this course's.
20. R. H. Lustig and colleagues, "Isocaloric fructose restriction and metabolic improvement", Obesity, 2016, doi 10.1002/oby.21371. Read: the abstract. The limits listed are this course's reading of it.
21. S. Basu, P. Yoffe, N. Hills and R. H. Lustig, "The relationship of sugar to population-level diabetes prevalence", PLoS One, 2013, doi 10.1371/journal.pone.0057873. Read: the abstract.
22. R. Kahn and J. L. Sievenpiper, "Dietary sugar and body weight ... the pox on sugar is overwrought and overworked", Diabetes Care, 2014, doi 10.2337/dc13-2506. Read: the abstract. Disclosures not read.
23. J. L. Sievenpiper and colleagues, "Effect of fructose on body weight in controlled feeding trials", Annals of Internal Medicine, 2012, doi 10.7326/0003-4819-156-4-201202210-00007. Read: the abstract.
24. J. Erickson and colleagues, "The Scientific Basis of Guideline Recommendations on Sugar Intake", Annals of Internal Medicine, 2017, doi 10.7326/M16-2020. Read: the abstract.
25. World Health Organization, news release on non-sugar sweeteners, 15 May 2023. Read: the whole release.
26. WHO, IARC and JECFA, "Aspartame hazard and risk assessment results released", joint news release, 14 July 2023. Read: the whole release.