Addressing reverse causality in observational research is critical to establish reliable associations between low/no calorie sweeteners and cardiometabolic health

Results of a systematic review and meta-analysis of prospective cohort studies that model intake of sweeteners as change or substitutions to mitigate the influence of reverse causality

 

Highlights:

  • Reverse causality is a major source of bias in observational studies investigating associations between low/no calorie sweeteners and health outcomes
  • By including prospective cohort studies that model sweetener intake as change or substitutions, a new systematic review and meta-analysis provides more consistent and robust associations, partly mitigating the influence of reverse causality
  • The study results showed that low/no calorie sweetened beverages may be associated with cardiometabolic benefit, rather than risk, when used to replace sugar-sweetened beverages

 

Reverse causality is a common and important problem in epidemiological research exploring the relation between low/no calorie sweeteners and health. By reverse causality, we refer to the phenomenon that describes the association of two variables differently than expected; for example, a positive association between low/no calorie sweeteners’ consumption and higher body weight may be the consequence of and not the reason for overweight and obesity.1 Therefore, in order to identify more robust and credible associations between low/no calorie sweeteners and health outcomes that could inform nutrition guidelines, it is essential to address the issue of reverse causality in observational studies.

With the objective to reliably inform their Clinical Practice Guidelines for Nutrition Therapy, the Diabetes and Nutrition Study Group (DNSG) of the European Association for the Study of Diabetes (EASD) conducted a systematic review and meta-analysis of prospective cohort studies that model intake of low/no calorie sweeteners as change or substitutions to partly mitigate the influence of reverse causality.2

Low/no calorie sweetened beverages may be associated with cardiometabolic benefit, when used in place of sugar-sweetened beverages

The systematic review and meta-analysis by Lee et al included fourteen prospective cohort studies with 416,830 adults.2 The study restricted the analyses to cohort comparisons where investigators modelled the exposure as either change in intake or substitution of sugar-sweetened beverages (SSBs) with low/no calorie sweetened beverages (i.e., the “intended substitution”), low/no calorie sweetened beverages with water, or SSBs with water. The authors stressed that the assessment of changes in exposure over time rather than baseline or prevalent exposure, and further modelling of the intended substitution of SSBs with low/no calorie sweetened alternatives appear to provide more consistent, robust, and biologically plausible associations.

The study results showed that the intended substitution of SSBs with low/no calorie sweetened beverages was associated with lower weight and reduced risk of obesity, coronary heart disease, total cardiovascular disease (CVD) mortality, and total mortality, without an adverse association with any other cardiometabolic outcomes. Substitution of SSBs with water was associated with lower weight and reduced risk of obesity and type 2 diabetes, while the substitution of low/no calorie sweetened beverages with water was not associated with changes in any outcomes indicating similar effects.

Why is addressing the issue of reverse causality in observational research key to draw credible conclusions about low/no calorie sweeteners’ cardiometabolic health effects?

It is well recognised that residual confounding and reverse causality are major sources of bias in observational studies.3 This is particularly relevant to low/no calorie sweeteners’ research because people who are at high risk for obesity or cardiometabolic disease frequently turn to low/no calorie sweeteners as a risk reduction strategy. Indeed, higher consumers of low/no calorie sweeteners may choose these products because they are at a higher risk for adverse cardiometabolic outcomes and not the other way around.4 This is also acknowledged in research supported by the World Health Organization (WHO).5

Therefore, it is critical to address the issue of reverse causality in epidemiological research involving health effects of low/no calorie sweeteners to obtain credible associations. In a consensus workshop aiming to address future research needs, among other objectives, experts emphasised the need for population cohort studies to model changes in health outcomes in the context of changes in low/no calorie sweeteners’ consumption over time (with repeated measures), and also stressed that studies should include substitution analysis and adjustment for adiposity.6 An important strength of the systematic review by Lee et al is that it analysed studies with change and substitutions analysis to address sources of bias in epidemiological research, mainly reverse causation, for the update of the EASD clinical practice guidelines for nutrition therapy.2

So, whilst other reviews of observational studies relying largely on baseline intake data claim to show a positive association between low/no calorie sweeteners intake and cardiometabolic risk5, the current systematic review of prospective cohort studies that model the exposure as change or substitutions found the opposite: low/no calorie sweetened beverages in their intended substitution for SSBs are not associated with cardiometabolic harm and, in fact, may be associated with cardiometabolic benefit, comparable with the standard of care, water.6 Importantly, the Lee et al study results are also in line with evidence from systematic reviews and meta-analyses of randomised controlled trials (RCTs) that overall provide higher certainty evidence.7 Evidence from RCTs confirms no adverse effect of low/no calorie sweeteners on intermediate cardiometabolic risk factors including blood glucose and lipid levels, blood pressure, etc, and in fact indicates some benefits on body weight and liver fat when low/no calorie sweeteners are used to replace sugars in the diet.5,7

  1. Lohner S, Toews I, Meerpohl JJ. Health outcomes of non-nutritive sweeteners: analysis of the research landscape. Nutr J. 2017 Sep 8;16(1):55
  2. Lee JJ, Khan TA, McGlynn N, Malik VS, Hill JO, Leiter LA, Jeppesen PB, Rahelić D, Kahleová H, Salas-Salvadó J, Kendall CWC, Sievenpiper JL. Relation of Change or Substitution of Low- and No-Calorie Sweetened Beverages With Cardiometabolic Outcomes: A Systematic Review and Meta-analysis of Prospective Cohort Studies. Diabetes Care. 2022 Aug 1;45(8):1917-1930
  3. Mela DJ, McLaughlin J, Rogers PJ. Perspective: Standards for Research and Reporting on Low-Energy (“Artificial”) Sweeteners. Adv Nutr. 2020 May 1;11(3):484-491
  4. Sievenpiper JL, Khan TA, Ha V, Viguiliouk E, Auyeung R. The importance of study design in the assessment of non-nutritive sweeteners and cardiometabolic health. CMAJ. 2017 Nov 20;189(46):E1424-E1425
  5. Rios-Leyvraz & Montez, 2022 Rios-Leyvraz M, Montez J (World Health Organization)‎. Health effects of the use of non-sugar sweeteners: a systematic review and meta-analysis. World Health Organization (WHO) 2022.https://apps.who.int/iris/handle/10665/353064. License: CC BY-NC-SA 3.0 IGO
  6. Ashwell M, Gibson S, Bellisle F, Buttriss J, Drewnowski A, Fantino M, Gallagher AM, de Graaf K, Goscinny S, Hardman CA, Laviada-Molina H, López-García R, Magnuson B, Mellor D, Rogers PJ, Rowland I, Russell W, Sievenpiper JL, la Vecchia C. Expert consensus on low-calorie sweeteners: facts, research gaps and suggested actions. Nutr Res Rev. 2020 Jun;33(1):145-154
  7. McGlynn ND, Khan TA, Wang L, Zhang R, Chiavaroli L, Au-Yeung F, Lee JJ, Noronha JC, Comelli EM, Blanco Mejia S, Ahmed A, Malik VS, Hill JO, Leiter LA, Agarwal A, Jeppesen PB, Rahelic D, Kahleová H, Salas-Salvadó J, Kendall CWC, Sievenpiper JL. Association of Low- and No-Calorie Sweetened Beverages as a Replacement for Sugar-Sweetened Beverages With Body Weight and Cardiometabolic Risk: A Systematic Review and Meta-analysis. JAMA Network Open 2022 Mar 1;5(3):e222092