I thank Drs Yan, Chan and Louie for their Correspondence (Evidence does not support the reduction of added sugar intake from all food sources. Nat. Rev. Cardiol. https://doi.org/10.1038/s41569-022-00791-w (2022))1 on my Comment article (Guidelines to lower intake of added sugar are necessary and justified. Nat. Rev. Cardiol. 19, 569–570 (2022))2. The authors feel that their views were misrepresented and the purpose of their letter was to clarify them.

They state that the intention of their review3 was not to exonerate added sugars as being harmless to health. My Comment article did not suggest that this was their intention. However, on the basis of the final sentence of their abstract — “We argue the current public health recommendations to encourage the reduction of both solid and liquid forms of free sugar intake (e.g., sugar reformulation programs) should be revised due to the overextrapolation of results from [sugar-sweetened beverages] SSBs studies”3 — I did conclude that their review, and the other two articles that I discussed, sent the implicit message that consumers should be sceptical about the recommendations to reduce their consumption of added sugar. I stand by that conclusion and also by my firm belief that such a message is harmful to consumers and likely to exacerbate the incidence of dental caries and the epidemic of metabolic disease.

As I previously stated2, there are three clinical dietary intervention studies conducted by Reiser and Hallfrish that demonstrated that added sugar in solid food increases cardiometabolic risk factors compared with a carefully matched starch diet. These studies have generated ten publications. They are among the most well-controlled diet studies that have ever been conducted, with participants consuming half of their standardized, eucaloric meals while being monitored at a clinical research centre. The most likely reason that the findings from these studies have not been replicated is expense. Clinical research centres with dietary kitchens are no longer subsidized by the NIH, so these studies are now far more expensive to conduct. However, the research groups of John Bantle and Arne Astrup have each conducted two studies that demonstrated detrimental health effects of dietary interventions that were high in added sugar from both solid food and beverages. Also, five recent dietary intervention studies have demonstrated that the restriction of free and added sugar from both solid food and beverages reduced the levels of liver fat and other risk factors. Would these same results have been generated if only SSBs had been restricted and the participants had been allowed to eat their usual amount of ice cream, cake, cookies, candy and pastry? Given that sugars in solid food make up about 58% of the average added sugar intake in the USA, perhaps not. Indeed, it is interesting to compare the results of the latest of these sugar-restriction studies4 with those of a study in which the intervention consisted solely of reducing the consumption of SSBs5. In both studies, the interventions were 12 weeks long and the participants were adults who were overweight or obese and had non-alcoholic fatty liver disease. For the trial that focused only on the reduction of SSBs, the participants had a decrease in liver lipid levels compared with the control group, but there were no other significant differences between the groups5. For the trial that focused on the reduction of both sources of free sugar, the intervention group had reduced steatosis and fibrosis score, HOMA-IR (Homeostatic Model Assessment for Insulin Resistance) score, and circulating triglyceride, total cholesterol, C-reactive protein and tumour necrosis factor levels compared with the control group4.

In summary, there are clinical data to show that the consumption of added sugar in solid food increases cardiometabolic risk factors. There are also data to show that increased consumption of added sugar from both solid and liquid sources increases risk factors and that decreased consumption of added sugar from both solid and liquid sources decreases risk factors. Therefore, it is more than my “best guess” that the risk of harm would be greater than the likelihood of benefit if the WHO and US dietary guidelines were revised to recommend solely a reduction in free and added sugar from SSBs. Such a revision might better meet the principles of evidence-based medicine, but I do not believe that it would meet the ‘do no harm’ axiom of the health-care profession.

The authors do suggest potential negative consequences of reducing free and added sugar consumption to less than 10% of daily energy, and these concerns consist of increased exposure to artificial sweeteners and nutrient dilution. The data suggesting that exposure to artificial sweeteners might promote obesity and metabolic diseases in humans are almost exclusively from epidemiology studies. The conclusions from these studies might be subject to the ‘reverse causality’ phenomenon, whereby associations exist because individuals with prediabetes or diabetes mellitus or who are overweight or obese might choose to consume artificial sweeteners to reduce their blood glucose level or their risk of weight gain, and not because artificial sweeteners increase blood glucose level or cause weight gain6. Of note, the various artificial and non-nutritive sweeteners have different chemical structures that lead to different digestion and uptake patterns in the gut and so are likely to have different metabolic effects in the human body7. Therefore, long-term randomized controlled trials will have to be conducted on the individual sweeteners because the results for one sweetener cannot be extrapolated to others. These studies are mainly lacking, but those that have been performed suggest that the consumption of artificial or non-nutritive sweeteners might promote body weight loss and do not elevate blood glucose levels8,9. Clearly the evidence does not support revising the WHO and US dietary guidelines to avoid increased exposure to artificial sweeteners.

As reviewed by Louie and colleagues, the majority of the studies do not support the suggestion that reduced consumption of added sugar leads to nutrient dilution; instead, they show the opposite10. However, the paper cited by the authors1 suggests that holding free sugar to <5% of daily energy is associated with nutrient dilution compared with the 5–10% level. This observation is not surprising because the group consuming <5% of daily energy as free sugar had a significantly higher proportion of dieters than the other groups and consumed the least energy. When adjusted for energy intake, this group had higher intake of fibre, vitamin A, vitamin E, magnesium, potassium and zinc than the other five groups10. Therefore, I do not agree that nutrient dilution is a potential negative consequence of reducing the consumption of added sugar that justifies revising the WHO and US dietary guidelines.