Artificially sweetened beverages vs. water for weight loss and cardiometabolic health Original paper

In this randomized controlled trial, nonnutritive-sweetened beverages and water had similar effects on body weight and cardiometabolic risk factors.

The effect of nonnutritive-sweetened (NNS) beverages, compared to water, on cardiometabolic risk factors.

The primary outcome was change in body weight. The secondary outcomes were changes in other anthropometrics (waist and hip circumference), blood pressure, blood lipids, fasting glucose and insulin, HbA1c, liver enzymes, hunger, and sugar consumption (assessed using a questionnaire).

493 participants (average age of 45, 70% women, 30% men) with overweight or obesity (BMI of 27–35) living in or around the county of Merseyside, England.

In this 12-week randomized controlled trial, the participants were assigned to consume at least 2 servings (330 milliliters) of either water or NNS beverages daily. The participants received deliveries of their assigned beverages throughout the study. In the NNS group, the participants could select from a range of 20 different beverages containing fewer than 5 calories per 8 ounces. The participants in the water group were asked to refrain from drinking NNS beverages during the trial.

To monitor intervention adherence, the participants completed online daily beverage logs, submitted photos of the empty packaging from consumed beverages to the research staff, and completed periodic 24-hour dietary recalls.

During the trial, participants took part in a weight loss program. This program involved weekly education classes led by dietitians and psychologists, during which the participants received dietary advice, physical activity recommendations, and behavioral strategies.

Compared to baseline, body weight decreased in both groups (−5.8 kg in the NNS group vs. −5.6 kg in the water group), with no differences between groups. Similar reductions were found for waist and hip circumference and sugar consumption.

Nearly all of the examined cardiometabolic risk factors improved compared to baseline in both groups, with only a small number of exceptions (e.g., the liver enzymes alanine transaminase and aspartate transaminase). The only difference between groups was for HbA1c, which decreased more in the water group than the NNS group (–0.9 vs. –0.3 mmol/mol).

The interest in NNS beverages for weight loss and cardiometabolic health stems from the fact that regular sugar-sweetened beverage (SSB) consumption is associated with an increased risk of obesity and obesity-related cardiometabolic disease (e.g., type 2 diabetes, cardiovascular disease) in observational studies, and adding SSBs to the diet has been found to increase body weight in randomized controlled trials.^[1]

So, the hypothesis is that replacing SSBs with NNS beverages can assist with weight management by reducing energy intake, and this will improve cardiometabolic health because caloric restriction and subsequent weight loss has been shown to improve a variety of cardiometabolic biomarkers, including measures of glycemic control (e.g., HbA1c,^[2][3][4] blood pressure,^[5] and blood lipids, along with other risk factors for atherosclerosis.^[6][4]

Furthermore, NNS have been the subject of significant interest because, in contrast to what would be expected based on the above, meta-analyses of observational studies have reported that a higher intake of NNS is associated with an increased risk of obesity and cardiometabolic disease.^[7] However, these findings seem likely to be a consequence of reverse causation — that is, a higher consumption of NNS does not cause weight gain, leading to obesity and related cardiometabolic disease, but rather, people who experience weight gain and related cardiometabolic disease consume higher amounts of NNS, likely in an attempt to reduce their body weight and improve their health condition.

Indeed, meta-analyses of randomized controlled trials indicate that consuming NNS beverages in place of SSBs reduces energy intake and body weight.^[8][9]

Evidently, if the replacement of SSBs in the diet with NNS beverages reduces energy intake and results in a decrease in body weight, then the consumption of NNS beverages can have a net positive effect on health. However, this doesn’t mean that replacing SSB with NNS beverages is the ideal strategy. Replacing SSBs with other low-calorie or non-calorie-containing beverages — namely, water — could be better for weight management and improving cardiometabolic risk factors.

One theory for why replacing SSBs with water is better than replacing them with NNS beverages is that the consumption of NNS beverages might increase appetite and the desire for sweet, energy-containing food items.^[10] However, the results of the summarized study do not support this hypothesis: sugar consumption decreased compared to baseline in both groups, with no difference between groups, and hunger decreased compared to baseline in the NNS group only, although there was not a significant difference between groups.

The evidence at large is in agreement with these findings: there is no clear and consistent relationship between the exposure to sweet foods and the desire for sweet foods.^[11] In fact, most acute trials and some longer-term trials indicate that greater exposure to sweet foods reduces the desire for sweet foods.^[11] This aligns with the research pertaining to sensory-specific satiety, which shows that exposure to a particular sensory attribute, such as sweetness, can lead to short-term reductions in the apparent pleasantness and choice of foods with the same attribute.^[12]

Considering that the intake of sweet foods does not reliably increase the desire for sweet foods — which are often energy-dense and not very satiating, contributing to excess calorie intake — it is perhaps unsurprising that there was no difference in weight loss between groups in the summarized study. Other research on the topic echoes this finding.^[8]

However, there is also evidence that the consumption of NNS beverages is superior to water for weight loss. In the second largest randomized controlled trial on the topic (after the summarized study), 303 participants consumed either 24 ounces of water or NNS beverages daily as part of a 12-week weight loss intervention (the same program used in the current trial) and a 40-week weight loss maintenance intervention.^[13] It was found that weight loss at 12 weeks and 1 year were greater in the NNS beverage group.

NNS beverages vs. water for weight loss

Adapted from Peters et al., 2016, Obesity.

One factor that might influence the effectiveness of NNS beverages, compared to water, on weight loss is the individual’s baseline beverage intake. In the above study, the participants were regular NNS beverage drinkers at baseline. The researchers speculated that assignment to the water group — and thus the elimination of the NNS beverages that the participants typically consumed — led to a desire to seek out sweetness from other aspects of the diet, resulting in the consumption of more sweet foods and subsequently greater energy intake and less weight loss.

In comparison, the inclusion criteria in the summarized study did not require regular NNS consumption, only that the participants regularly drank more than 3 cold beverages per week (water, NNS, or SSB) and that NNS and SSB consumption had to be less than 2 liters per day. However, in contrast to the above study, subgroup analysis indicated that whether the participants regularly consumed NNS beverages at baseline (defined as NNS beverages making up more than 26% of drink choices in the past 5 years) did not influence weight loss.

Other research suggests that in people who regularly consume SSBs, replacement with NNS beverages may be more effective than water. In one study of participants who consumed at least 280 kcal per day from caloric beverages, the odds of achieving clinically significant weight loss at 6 months was higher in participants who replaced caloric beverages with NNS beverages compared to participants who replaced them with water.^[14] The researchers speculated that participants may have an easier time replacing SSBs with NNS beverages due to their similar taste and properties (e.g., caffeine content).

Another reason why replacing SSBs with water may be superior to NNS beverages is that NNS beverages negatively alter the gut microbiome, resulting in downstream consequences on glycemic control. The inspiration for this hypothesis stems from the results of rodent studies.^[15]

Studies in humans, most of which have the participants consume sucralose or aspartame, do not demonstrate any clear and consistent effects of NNS consumption on measures of glycemic control.^{[16][17][18][19]}

Although these data indicate that NNS may not negatively affect glycemic control, the summarized study found that HbA1c decreased more in the water group than the NNS group, although there were no differences between groups for changes in fasting glucose and insulin. Could water be better than NNS beverages for improving some measures of glycemic control?

In support of the findings of the summarized study, a 2022 meta-analysis indicated that replacing NNS beverages with water did not affect fasting glucose or insulin.^[9]

Studies examining changes in HbA1c are scarce. There have been two 24-week studies (both by the same research team) that had women with overweight or obesity and who were regular NNS beverage consumers follow a weight-loss intervention and either continue drinking NNS beverages or replace them with water.^[20][21] In both of these studies, HbA1c decreased in both groups, with no differences between groups.

However, a crucial limitation of these studies is the dose of NNS beverages. In the groups that continued drinking NNS beverages, the participants consumed one 250-milliliter serving after lunch on 5 days per week. In comparison, the summarized study had participants consume at least two 330-milliliter servings daily. As such, the dose of NNS beverages may have been insufficient for differences in HbA1c to manifest between groups. Further research is needed to determine whether replacing SSBs with water results in superior improvements in HbA1c compared to NNS beverages, and if so, whether this difference is of clinical significance.

The researchers did not adjust for multiple comparisons, which increases the risk of false-positive results. Thus, the secondary outcomes (e.g., the greater reduction in HbA1c with water) should be interpreted with caution.