Artificial sweeteners and heart health: 37 studies pooled

A 2026 meta-analysis of 37 studies (3,625,613 participants) found that people in the highest category of artificial sweetener intake had a higher hazard of type 2 diabetes (hazard ratio 1.26) and heart failure (1.26) than those in the lowest. These are associations, and the abstract reports relative differences only, with no absolute rates.

What did the 2026 meta-analysis find?

Kim et al., Nutrition Research and Practice 2026 searched PubMed and Embase to 8 October 2025 and pooled hazard ratios comparing the highest with the lowest sweetener intake category, using a random-effects model. Higher intake was associated with type 2 diabetes (HR 1.26, 95% CI 1.14 to 1.40), hypertension (1.13, 1.09 to 1.18), heart failure (1.26, 1.14 to 1.39), stroke (1.12, 1.05 to 1.19), coronary artery disease (1.11, 1.00 to 1.23), cardiovascular mortality (1.18, 1.03 to 1.35) and all-cause mortality (1.09, 1.01 to 1.17).

The association with atrial fibrillation was not significant (HR 1.13, 95% CI 0.94 to 1.37). The abstract reports significant heterogeneity for all endpoints except stroke and heart failure.

The abstract does not list the designs of the included studies. It assessed their quality with the Newcastle-Ottawa Scale, a checklist for non-randomised studies, and it words its findings as association, not proof that sweeteners caused any outcome.

Why is the link hard to read?

A pooled hazard ratio compares groups defined by sweetener intake. It cannot separate sweeteners from other differences between those groups that may also affect outcomes, a problem called confounding.

The World Health Organization reached a similar caution in 2023. Its news release on a guideline for non-sugar sweeteners, based on its own systematic review, said the evidence showed no long-term benefit in reducing body fat in adults or children. It also said cohort studies suggested potential undesirable effects, including a higher risk of type 2 diabetes, cardiovascular disease and mortality in adults.

WHO added that the link between sweeteners and disease outcomes “might be confounded by baseline characteristics of study participants and complicated patterns of NSS use”, and rated its recommendation as conditional.

Processed food more broadly has its own evidence base, summarised in the Lancet series evidence on ultra-processed foods.

The bottom line

• A 2026 meta-analysis of 37 studies associated the highest sweetener intake category with type 2 diabetes (HR 1.26) and heart failure (HR 1.26) versus the lowest, but not with atrial fibrillation.
• The abstract gives no study designs or absolute rates and reports significant heterogeneity for most outcomes.
• WHO’s 2023 review flagged possible confounding in the cohort findings and made a conditional recommendation.