Background
High intake of free sugars has been linked to overweight and obesity, which affects nearly 40% of the global adult population and millions of children, and, in turn, diet-related noncommunicable diseases (NCDs), which are the leading causes of death worldwide. In response, the World Health Organization (WHO) has issued recommendations to reduce the intake of free sugars. Various measures are being taken to reduce consumption of free sugars as part of global efforts to address the epidemic of obesity and associated diseases.
Non-sugar sweeteners (NSS)1 are low- or no-calorie alternatives to free sugars that are generally marketed as aiding weight loss or maintenance of healthy weight, and are frequently recommended as a means of controlling blood glucose in individuals with diabetes. Individual sweeteners undergo toxicological assessment to establish safe levels of intake (i.e. acceptable daily intake, or ADI). However, there is no clear consensus on whether NSS are effective for long-term weight control or if they are linked to other long-term health effects at habitual intakes within the ADI.
Since the release of updated WHO guidance on free sugars intake in 2015, interest in the potential utility of NSS in reducing sugars intake has increased. Therefore, it was considered necessary to review the evidence in a systematic manner, and issue WHO guidance on NSS use through the WHO guideline development process.
Objective, scope and methods
The objective of this guideline is to provide guidance on the use of NSS to be used by policy-makers, programme managers, health professionals and other stakeholders in efforts to reduce free sugars intake, promote healthy diets, and prevent unhealthy weight gain and diet-related NCDs. Because the WHO Nutrition Guidance Expert Advisory Group (NUGAG) Subgroup on Diet and Health focuses on providing guidance on the prevention of unhealthy weight gain and diet-related NCDs, providing guidance on the management of diabetes in individuals with pre-existing diabetes is beyond the scope of this guideline. Therefore, the guidance in the guideline may not be relevant for individuals with existing diabetes. The guidance is based on evidence of health effects of NSS use at levels already considered safe (i.e. within the ADI), and is not intended to provide updated or alternative guidance on safe or maximal levels of intake.2
The guideline was developed following the WHO guideline development process, as outlined in the WHO handbook for guideline development. This process includes a review of systematically gathered evidence by an international, multidisciplinary group of experts; assessment of the certainty in (i.e. quality of) that evidence via the Grading of Recommendations Assessment, Development and Evaluation (GRADE) framework; and consideration of additional, potentially mitigating factors3 when translating the evidence into recommendations.
The evidence
Evidence from a recent systematic review and meta-analyses of randomized controlled trials (RCTs) and prospective observational studies found that higher NSS consumption by adults led to lower body weight and body mass index (BMI), compared with not consuming NSS or consuming lower amounts of NSS, when assessed in short-term RCTs, but was associated with increased BMI and risk of incident obesity in long-term prospective observational studies. Effects on body weight and BMI from RCTs are observed only when intake of NSS is compared with intake of free sugars, and are likely mediated at least in part by a reduction in energy intake. No other significant effects or associations on measures of body fatness were observed in either RCTs or prospective cohort studies.
Long-term NSS use was associated with increased risk of type 2 diabetes, cardiovascular diseases (CVDs) and mortality in prospective cohort studies conducted in adults. However, significant effects were not observed on intermediate markers of disease such as fasting glucose, fasting insulin or blood lipids when assessed in short-term RCTs.
Evidence from studies conducted in children and pregnant women was more limited than that identified for adults. One RCT conducted in children reported a reduction in several measures of body fatness when sugar-sweetened beverages were replaced with beverages containing NSS; however, no effect was observed when results for BMI z-score4 were combined with those from a second trial. Results from prospective observational studies did not suggest any significant associations between NSS use and measures of body fatness. Two RCTs conducted in children reported lower indicators of dental caries with use of the NSS stevia. All other identified studies reported no significant associations between NSS use and prioritized health outcomes in children.
Meta-analysis of three prospective observational studies found an increased risk of preterm birth with higher NSS use during pregnancy, but associations observed between birth weight or weight of offspring later in life and NSS use during pregnancy were inconsistent. Single prospective observational studies reported associations between NSS use during pregnancy and outcomes in offspring, including increased risk of asthma and allergies, and poorer cognitive function.
Recommendation and supporting information
This recommendation should be considered in the context of WHO recommendations to reduce free sugars intake and other guidance promoting healthy diets, including WHO guidelines on carbohydrates, total fat, saturated and trans-fatty acids, polyunsaturated fatty acids, sodium and potassium.
WHO recommendation
WHO suggests that non-sugar sweeteners not be used as a means of achieving weight control or reducing the risk of noncommunicable diseases (conditional recommendation).
Rationale
- ▶
The recommendation is based on evidence of low certainty overall, from a systematic review that assessed the health effects of higher compared with lower intake of NSS.5 The systematic review found no evidence of long-term benefit on measures of body fatness in adults or children, and potential undesirable effects from long-term use in the form of increased risk of type 2 diabetes, CVDs and mortality in adults. Limited evidence suggests potential undesirable effects in the form of increased risk of preterm birth with NSS use during pregnancy.
- ▶
Specific findings from the systematic review supporting this recommendation are as follows.
Adults
Evidence from randomized controlled trials (RCTs) was as follows.
NSS use in any manner
6 resulted in reduced sugars and energy intake, lower body weight and lower BMI in short-term RCTs (all
low certainty evidence), the majority of which lasted 3 months or less. NSS use did not significantly affect other measures of body fatness or intermediate markers of cardiometabolic health, including glucose, insulin or blood lipids (
very low to
moderate certainty evidence). Evidence from a small number of longer-term trials lasting 6–18 months did not suggest an effect on body weight but was difficult to interpret because of many differences in how these trials were conducted and results reported.
When intake of NSS was directly compared with intake of free sugars (i.e. one group in a trial received NSS, and another group received free sugars), those receiving NSS had lower body weight and BMI, similar in magnitude to the results when NSS was used in any manner. However, most of these trials provided foods and beverages containing NSS or free sugars in addition to existing diets and therefore did not directly measure the effects of replacing free sugars with NSS. When NSS were compared with nothing/placebo or water (i.e. one group in a trial received NSS, and another group received nothing/placebo or water), no effects on body weight or BMI were observed.
When NSS were assessed specifically as replacements for free sugars in a small number of RCTs (i.e. habitual consumers of foods or beverages containing free sugars were asked to switch to versions containing NSS in place of free sugars), the effect on body weight was significantly weakened relative to that observed for NSS used in any manner, and an effect on BMI was no longer observed.
Evidence from prospective observational studies, with up to 10 years of follow-up, was as follows.
Higher intakes of NSS were associated with higher BMI and increased risk of incident obesity, but not other measures of body fatness (very low to low certainty evidence).
Higher intakes of NSS were associated with increased risk of type 2 diabetes, CVDs and CVD mortality, and all-cause mortality in long-term prospective observational studies with average follow-up of 13 years (very low to low certainty evidence), but were not associated with differences in overall cancer incidence or mortality (very low certainty evidence).
Use of NSS (predominantly saccharin) was associated with increased risk of bladder cancer as assessed in case–control studies (very low certainty evidence).
Children
One RCT conducted in children reported a reduction in several measures of body fatness when sugar-sweetened beverages were replaced with those containing NSS (moderate certainty evidence). However, when results for BMI z-score were combined with those from a second trial, no effect was observed (moderate certainty evidence), and results from prospective observational studies did not suggest any significant associations between NSS use and measures of body fatness (very low certainty evidence). All other identified studies reported no significant associations between NSS use and prioritized health outcomes in children.
Pregnant women
Meta-analysis of three prospective observational studies found an increased risk of preterm birth with higher NSS use during pregnancy (low certainty evidence), but associations between birth weight or weight of offspring later in life and NSS use during pregnancy were inconsistent (very low certainty evidence). Other individual prospective observational studies reported associations between NSS use during pregnancy and outcomes in offspring, including increased risk of asthma and allergies, and poorer cognitive function (very low certainty evidence). No associations were observed between NSS use and risk of gestational diabetes.
- ▶
The lack of evidence for long-term benefit of NSS use on measures of body fatness assessed in RCTs and potential long-term effects of NSS use observed for adults in prospective observational studies were considered to be relevant for women during pregnancy, and were reasonably expected to be relevant for children and adolescents as well. Therefore, in addition to the limited direct evidence for children and pregnant women, the evidence from RCTs and observational studies in adults was extrapolated to children, adolescents and pregnant women without downgrading for indirectness.
- ▶
In reviewing the evidence and formulating the recommendation, the NUGAG Subgroup on Diet and Health noted the following.
Because the primary role of NSS use is presumably to reduce free sugars intake (and consequently risk of unhealthy weight gain and disease associated with excess free sugars intake), the currently available evidence on which to base a recommendation on NSS is largely indirect – that is, most RCTs comparing intake of NSS with intake of free sugars did not explicitly assess the replacement of free sugars with NSS.
Because weight loss and maintenance of a healthy weight must be sustained over the long term
7 to have a meaningful impact on health, evidence of minor weight loss or reduced BMI over several months or less, as observed in the RCTs, without additional evidence of long-term impact, does not represent a health benefit.
The discordant results between the RCTs and prospective cohort studies suggest that the small amount of weight loss resulting from NSS use in short-term experimental settings may not be relevant to the effects of long-term NSS use in the general population.
In addition, the NUGAG Subgroup on Diet and Health noted that:
there were no identified undesirable effects or other mitigating factors that would argue against not using NSS;
NSS are not essential dietary factors and have no nutritional value; and
use of NSS is not the only way to achieve a reduction in free sugars intake; viable alternatives exist that are compatible with features of a healthy diet including consumption of foods with naturally occurring sugars, such as fruit, and unsweetened foods and beverages.
Based on the evidence and other considerations noted above, the NUGAG Subgroup on Diet and Health concluded that the lack of evidence to suggest that NSS use is beneficial for body weight or other measures of body fatness over the long term, together with possible long-term undesirable effects in the form of increased risk of NCDs and death, outweighed any potential short-term health effects resulting from the small reductions in body weight and BMI observed in RCTs.
- ▶
Because of lack of certainty about the overall balance of desirable and undesirable effects associated with long-term NSS use for reducing NCD risk, including the possibility that reverse causation8 may have contributed to one or more of the associations observed between long-term NSS use and risk of disease in prospective observational studies, a conservative approach was taken, leading to a conditional recommendation.
Remarks
- ▶
With the exception of individuals with diabetes (as noted below), this recommendation is relevant for everyone: children and adults of any age, including pregnant and lactating women.
- ▶
The objective of this guideline is to provide guidance on the use of NSS in efforts to prevent unhealthy weight gain and diet-related NCDs, in the context of reducing free sugars intake. Assessing the health effects of NSS on individuals with pre-existing diabetes with the aim of providing guidance on disease management was beyond the scope of the guideline. Consequently, in the evidence reviewed, studies conducted exclusively in individuals with pre-existing diabetes were excluded, and in studies with mixed populations, diabetes was often controlled for as a potential confounding characteristic. Therefore, although individuals with diabetes can also reduce free sugars intake without the need for NSS, the recommendation does not apply to individuals with existing diabetes.
- ▶
The recommendation is relevant for all NSS, which are defined in this guideline as all synthetic and naturally occurring or modified non-nutritive sweeteners that are not classified as sugars. Common NSS include acesulfame K, aspartame, advantame, cyclamates, neotame, saccharin, sucralose, stevia and stevia derivatives. Because low-calorie sugars and sugar alcohols (polyols) are sugars or sugar derivatives containing calories, they are not considered NSS, and therefore the recommendation does not apply to these sweeteners.
- ▶
In this recommendation, “use” of NSS means consumption of foods or beverages that contain NSS, or the addition of NSS to food or beverages by the consumer.
- ▶
Many medications, and personal care and hygiene products contain NSS in small amounts to make them more palatable. The recommendation in this guideline does not apply to such products.
- ▶
“Weight control” in this recommendation refers to weight loss in cases of existing overweight or obesity, and preventing unhealthy weight gain by maintaining a healthy weight.
- ▶
The Joint FAO/WHO Expert Committee on Food Additives (JECFA) has set acceptable daily intakes (ADIs) for most commercially used NSS. Evidence supporting this WHO recommendation comes from a systematic review of studies in which NSS were consumed in amounts within the ADI set by JECFA, either because this was explicitly stated in the study or it was reasonably inferred that the ADI was not being exceeded.9
- ▶
The recommendation in this guideline was made based on evidence that suggests that there may be health effects associated with NSS use irrespective of which NSS is being used – that is, NSS as a class of compounds, despite individual NSS having different chemical structures, may have an impact on health. It is recognized that NSS are not a homogeneous class of compounds: each has a unique chemical structure. As a result, individual NSS have different sweetness intensities and organoleptic properties, and are processed differently by the body. Although limited evidence suggests that individual NSS may also differ in some of their physiological effects in humans, the evidence is currently insufficient to make recommendations for individual NSS.
- ▶
Efforts to reduce free sugars intake should be implemented in the context of achieving and maintaining a healthy diet. Because free sugars are often found in highly processed foods and beverages with undesirable nutritional profiles, simply replacing free sugars with NSS results means that the overall quality of the diet is largely unaffected. Replacing free sugars in the diet with sources of naturally occurring sweetness, such as fruits, as well as minimally processed unsweetened foods and beverages, will help to improve dietary quality, and should be the preferred alternatives to foods and beverages containing free sugars.
- 1
For the purposes of this guideline, NSS are defined as all synthetic and naturally occurring or modified non-nutritive sweeteners that are not classified as sugars. Sugar alcohols and low-calorie sugars are not considered to be NSS.
- 2
- 3
These include desirable and undesirable effects of the intervention, priority of the problem that the recommendation addresses, values and preferences related to the recommendation in different settings, the cost of the options available to public health officials and programme managers in different settings, feasibility and acceptability of implementing the recommendation in different settings, and the potential impact on equity and human rights.
- 4
BMI z-scores are adjusted for sex and age relative to standardized reference values.
- 5
Many RCTs compared use of NSS with no use of NSS, whereas prospective observational studies compared different levels of NSS use. To maintain consistency in comparing results across study designs, results are therefore generally reported for effects of higher compared with lower intake, noting that, in most trials, “lower intake” may in fact be no intake.
- 6
NSS were consumed by the participants in the RCTs in a variety of ways, including in pre-mixed beverages, powders or drops to be added to beverages by the participants themselves, solid foods, and capsules. To test for inherent properties of NSS, all forms of NSS were combined in the main analysis regardless of how they were consumed. Additional analyses assessed the individual ways of consuming NSS separately.
- 7
Ideally, healthy body weight is maintained throughout the life course.
- 8
A phenomenon sometimes observed in prospective cohort studies whereby those already in a pre-disease state or with increased risk of disease increase their exposure to the risk factor of interest, erroneously leading to the conclusion that increased exposure to the risk factor of interest leads to increased risk of disease.
- 9
For prospective cohort studies, it was generally not possible to determine the absolute highest intakes because the highest quantile was generally a specified amount or more (e.g. ≥2 servings per day). Although it is possible that some adults may have exceeded the ADI in some of these studies, the number doing so would probably have been an extremely small percentage of the entire group. The likelihood that children exceed the ADI is greater given their lower body weight; however, it is still expected to be a small percentage in most populations.
