Not So Sweet – AHA Review of the Potential Risks and Benefits of Consuming Low-Calorie Sweetened Beverages

In the last decade evidence has strengthened for the adverse health impacts of excessive added sugars intake, particularly in terms of body weight and cardiometabolic health. A consistent alignment of national guidelines has emerged recommending limits on, and reductions of, added sugars, particularly in the context of sugar-sweetened beverages (SSBs). A common follow-up question that arises is whether low calorie sweetened (LCS) beverages should be recommended as an approach or strategy to support reductions in SSBs and added sugars. A 2018 Science Advisory from the American Heart Association has provided an updated overview and interpretation of the available evidence on this topic. Eight specific LCS were considered, including six high-intensity sweeteners currently approved by the FDA, and two extracts from plants. The main conclusion drawn is that it currently appears prudent to advise against regular and long-term consumption of LCS beverages, particularly in children; rather, the use of other alternatives to SSBs, with a focus on water, is encouraged. An important caveat to this statement is that in general the available evidence is limited, making it difficult to draw firm conclusions.

The AHA reviewers included a helpful summary of current guidelines from five governmental and non-governmental health agencies: US Dietary Guidelines for Americans, American Medical Association, American Association of Physicians, American Diabetes Association, and the Academy of Nutrition and Dietetics. Across the organizations there was consistency for presenting the plausibility of benefit from LCS beverages if and when these were used to replace or displace SSB’s and resulted in decreased sugar, carbohydrate, and energy intake that further led to helping individuals with weight or blood glucose control. However, the plausibility of benefit was balanced by multiple concerns for possible deleterious effects that included: fostering a taste preference for sweet foods and beverages, making naturally-sweetened foods less appealing, adversely altering feelings of hunger and fullness, reducing awareness of calorie intake, adversely altering gut microbiota, increasing glucose intolerance, substituting for healthier beverages, contributing to the perception that individuals can consume more calories from other foods, and contributing to the possible intake of other ingredients in LCS beverages that could be of concern such as caffeine and artificial colors. With plausible pros and cons to consider, a succinct overview of the available evidence, and types of evidence, was provided.

Three categories of scientific evidence were considered. First, a search for the type of evidence with the strongest causal inference came up empty—no long-term randomized trials with clinical outcomes (RCTs) were identified. Second, in a review of short term RCTs with outcomes related to weight and blood glucose control the authors concluded there is some evidence LCS beverages may help in the management of overweight and obesity. However, this conclusion was tempered by concerns for compensatory behaviors that in the long-term could diminish or negate short-term effects; an example would be an individual who chooses a diet soda over a regular soda with lunch, and later, after dinner, chooses to consume a piece of cake that they wouldn’t have consumed otherwise if they hadn’t had the diet soda at lunch. The third type of evidence involved examinations of clinical outcome data from observational studies. Multiple studies of this type have been published, however, this type of study is unable to establish causal inference; only association, not causation. In this case, the associations observed with LCS beverage intake were all in the direction of adverse health outcomes, including increased risk of Type 2 Diabetes, stroke, and CVD. However, for the observational data there were methodological concerns with regard to unmeasured or residual confounding, or reverse causality. Therefore, the well-balanced conclusion of this review of different types of available evidence was that there was plausibility of both benefit and harm, but inadequate evidence to strongly support or refute either position.

Consumption trend data (NHANES 1999-2014) were also reported in the science advisory, and these provide an important perspective. The data show a steady drop in SSB intake over this 15-year period, from ~2 to ~1 serving/day, in both adults and children; consistent with successful public health efforts to promote the reduction of SSB intake. If this success were to be even partly attributable to substitution with LCS beverages, then an inverse association of intake would be expected. To the contrary, while the NHANES data show a transient but negligible rise in LCS beverage intake in adults from 0.6 to 0.7 servings/day between 1999/2000 and 2005/2006, from 2005/2006 to 2013/2014 there is a steady trend of a decrease in LCS beverage intake from 0.7 to 0.5 servings/day, parallel with the SSB intake decrease. The trend in children and adolescents is similar for LCS beverage intake, but the number of servings is considerably smaller at all time points, in the range of 0.1 servings/day. A previous AHA position statement on this topic that is cited and referred to in the current Science Advisory included USDA availability data of regular and diet soda intake from 1980 to 2000, with diet soda increasing from ~5 to ~10 gallons/year/capita, and regular soda increasing in parallel from ~30 to ~38 gallons/year/capita. If the primary health benefit of LCS beverages is to help displace SSBs, then the trends for intake of these two beverages should be inverse. However, they are not, the trends are parallel. This does not refute the possibility that some individuals can successfully use to LCS beverages to help the cut back on SSB intake. But the national trend data suggest that this is not a strategy used successfully by most Americans.

In summary, the Science Advisory writers have effectively examined the potential effects of LCS beverages on cardiometabolic health and concluded from the totality of available evidence, albeit limited, it is prudent to advise against the prolonged consumption of LCS beverages. It is notable that the reviewers never consider or suggest that LCS beverages are healthy choices in and of themselves. Rather the only possible health benefit would come from substituting for something even less healthy, SSBs. Trend data suggest this is not how LCS beverages have been used at the national level, and compensatory behaviors provide a plausible explanation as to why there would not be an overall cardiometabolic benefit even if LCS beverages were used to substitute for SSBs. Adding concern to what is already a very weak case for possible benefit, is the set of possible plausible adverse dietary and cardiometabolic effects of LCS beverage intake presented, making these appear to be not such a sweet choice after all.