Artificial Sweeteners

What it is

Artificial sweeteners, also called non-nutritive or non-sugar sweeteners, provide sweet taste with few or no calories. Common ones include aspartame, sucralose, saccharin, acesulfame potassium and neotame; steviol glycosides (stevia) and monk fruit are plant-derived and often grouped with them [1][2][3]. They are used in diet soft drinks, tabletop sweeteners, yogurts and other foods [4]. Regulators set acceptable daily intakes, and trials have mostly used doses within them [5].

How it might work

Replacing sugary foods and drinks with sweetened versions cuts calories, which is the simplest explanation for the small weight loss seen in trials [5][6]. Whether sweet taste changes appetite is debated: some trials show no effect on appetite hormones [3], while an acute crossover found the energy saved was fully eaten back at later meals [7]. A 2-week randomised trial found saccharin and sucralose altered gut bacteria and impaired glucose responses in some people, in a pattern that transferred to mice [8]; what this means for health is not known.

What the research says

Replacing sugar gives a small weight benefit [1]. A 2026 meta-analysis of 19 randomised trials found that swapping caloric sugars for non-nutritive sweeteners reduced body weight by 0.79 kg on average, with the effect concentrated in trials under 18 weeks; longer trials showed little difference [1]. A network meta-analysis of 17 trials in 1,733 adults with overweight or obesity found replacing sugar-sweetened drinks with low- and no-calorie sweetened drinks cut weight by 1.06 kg, body fat by 0.60% and liver fat, with moderate certainty; many authors had food- and sweetener-industry ties [9]. An earlier meta-analysis of 20 trials (2,914 people) found the advantage appeared only against sucrose: sweeteners did nothing versus water, placebo or no intervention [10]. A 2026 consensus review of 14 trials found 0.73 kg of weight loss and 0.72% body-fat reduction when sweetened drinks replaced sugar-sweetened drinks, and recommended them only as a short-term substitution [6]. In children, sweeteners produced slightly less BMI gain than sugar (moderate certainty), while the adult BMI evidence was rated low certainty [5].

Water does just as well [10]. Compared with water, sweeteners did not produce extra weight loss in pooled trials [10], and a review of 9 trials (1,457 people) found no differences in weight, waist, glucose, insulin resistance, lipids or blood pressure between sweetened and unsweetened drinks [11]. A 12-week trial in 493 adults found weight loss was equivalent (5.8 kg with sweetened drinks versus 5.6 kg with water) [12]; the same programme's 52-week results favoured sweetened drinks by 1.4 kg, a difference the authors called not clinically significant, and the trial was funded by the American Beverage Association [13]. One 1-year trial (303 people, all regular sweetener users) found -6.21 kg with sweetened drinks versus -2.45 kg with water [14].

Appetite and metabolism results are mixed [15]. Pooled trials found hunger slightly lower during the weight-loss phase with sweeteners but no difference during maintenance, and insulin fell less than with controls [15]. A crossover trial found appetite, ghrelin and GLP-1 unchanged after sweetened biscuits, with lower post-meal insulin and glucose [3], while an acute crossover found the calories saved from a sweetened drink were fully eaten back at the next meal [7]. Daily stevia for 12 weeks did not change glucose or insulin responses, and the stevia group held weight steady while controls gained (small, open-label trial) [16]. A 2-week randomised trial found saccharin and sucralose impaired glucose responses in some people, tied to individual gut microbiome changes; the clinical meaning is unclear [8]. Cochrane reviewers found inconclusive, very-low-certainty evidence on HbA1c and weight in diabetes, with side effects similar to placebo [17].

Long-term disease links come from cohort studies [18]. Umbrella reviews of cohort studies link habitual sweetened-drink intake to higher risks of obesity, type 2 diabetes, all-cause mortality, hypertension and cardiovascular disease [18][19]. In NutriNet-Sante (105,588 people, median 9.1 years), above-median consumers had a 69% higher risk of type 2 diabetes [20]; the same cohort linked higher intake to cardiovascular disease (HR 1.09) and cerebrovascular events (1.18) [21]. Across cohorts, each 250 mL/day of artificially sweetened drinks was associated with 21% higher obesity risk, 15% higher type 2 diabetes risk, 8% higher hypertension risk and 6% higher mortality [22]. These studies cannot show cause and effect, and the authors acknowledged possible reverse causation [20]. Notably, when an umbrella review re-analysed cohorts with substitution methods, sweeteners were associated with lower obesity, coronary heart disease and all-cause mortality instead of higher; the authors had extensive industry funding, and cohort certainty was very low [23]. In adults with type 2 diabetes, sweetened-drink intake was not linked to higher mortality, and replacing sugar-sweetened drinks with them was linked to lower mortality [24]. An early meta-analysis found trials did not clearly support weight benefits while cohorts pointed to weight gain and cardiometabolic risk [25].

Cancer evidence is conflicting [4]. Two large cohorts reported small increases in overall cancer risk with higher artificial sweetener or aspartame intake, without a clear dose-response [4][26]. Other large studies found no links: nearly 30 years of the Nurses' Health Studies found no aspartame-breast cancer association [27], a meta-analysis of 25 prospective studies (3.7 million people) found no overall association with cancer incidence or mortality [28], and a Spanish case-control study found no overall associations [29]. Reviews reached opposite conclusions, partly along funding lines: one funded by the beverage industry found no consistent associations [30], and another with a sweetener-industry grant concluded there is no evidence of cancer risk [31]. Reviews of the World Health Organization's 2023 guidance note that the benefit evidence is weak and long-term harm cannot be ruled out [32][33]. Sugary drinks themselves are linked to weight gain and heart disease, which is why replacing them is worth studying [34].

Doses used in studies

Trials have used 710 mL to 1.75 L per day of sweetened drinks, sachets of individual sweeteners at doses below the acceptable daily intake, and sweetened biscuits, for 2 weeks to 1 year [14][2][8][3]. Weight-management programmes lasted 12-52 weeks with daily sweetened drinks [12][13]; diabetes trials ran 4-10 months [17]; stevia was taken daily for 12 weeks [16]. These are doses used in research, not recommendations.

Safety and interactions

Short trials report side effects at rates similar to control groups, and most sweeteners are considered safe within the acceptable daily intakes set by regulators, which the trials used [17][5]. Aspartame must be avoided by people with phenylketonuria (PKU), who cannot break down phenylalanine; one review also advises caution for people with seizures or other neurological conditions and suggests reducing or avoiding aspartame during pregnancy [35]. In pregnancy, an analysis of 11 studies (138,007 women) linked daily artificial sweetener use to a slightly higher risk of preterm delivery, with low certainty [36]. No trial data were found on breastfeeding, and the diabetes trials did not report medicine-interaction data [17]. For people with diabetes, Cochrane reviewers found the evidence on HbA1c and weight inconclusive [17]. A 2-week trial found saccharin and sucralose changed the gut microbiome and impaired glucose responses in some people, but the health meaning is unknown [8].

Caveats and open questions

Most weight trials were short and unblinded, heterogeneity was high (I2 up to 84%), and the benefit over sugar was absent in trials longer than 18 weeks [1][5]. Cohort studies cannot show cause and effect and may suffer reverse causation: people who are already gaining weight or at risk of diabetes often switch to diet drinks [20]. A bias-adjusted re-analysis that used substitution methods found lower obesity, coronary heart disease and mortality with sweeteners instead of higher, but rated the cohort evidence very low certainty [23]. Industry funding runs through this field on both sides: a trial and a review funded by the American Beverage Association, a review with a grant from the International Sweeteners Association, and an umbrella review whose authors disclosed extensive industry honoraria [13][30][31][23]. Other key reviews declared no competing interests [19][1]. Sweeteners differ from one another: saccharin raised weight in one trial while sucralose lowered it [2]. The trials so far have been too short to show whether sweeteners change rates of diabetes, heart attacks or cancer over years [5]; a small trial of sucralose and saccharin on blood sugar and gut bacteria is due to finish in 2027 (NCT07124585).