Dietary Sugars: Not as Sour as They Are Made Out to Be?
61 min read
/
Sugars may be Mother Nature’s most essential nutrient since, in the course of evolution, humans have retained the ability to synthesize every sugar needed for metabolic functions while, at the same time, humans lost the ability to make 9 amino acids and 2 fatty acids required for life. During the same evolutionary period, Mother Nature iterated to preferring glucose as the almost sole fuel of the brain and the fetus. Human infants are born with sweet taste receptors, and sugars constitute approximately one-third of the energy supply of human milk. Moreover, complex carbohydrates, half or more of the energy content of human diets, are assimilated only after being broken down to simple sugars whose absorption is enhanced by other fuels serving as the energy source for enterocyte metabolism. If sugars are as unhealthful as many postulate, has Mother Nature been trying to fool us all these eons?
More recently, observational studies have reported repeatedly that consumption of dietary sugars is associated with various adverse consequences, including an increased risk of cardiovascular diseases and mortality [1]. It is important to realize that such observational studies can only find associations (i.e., uncover new hypotheses, not answers) and can never, no matter how extensive, prove causality [2]. “Big data” techniques have now demonstrated unequivocally that there are a huge number of
correlation interdependencies among unmeasured and uncontrolled “exposome” variables that make it impossible to attribute causality to individual dietary components themselves (e.g., sugars) [3]. Nonetheless, causal implications of this kind are prevalent throughout both the scientific literature as well as media reports. Importantly, a significant number of such studies fail to control for equivalent total energy intakes. In fact, a meta-analysis of sugar studies commissioned by the WHO concluded that the apparent effects of sugar intake were likely the consequence of increased energy intake [4]. Additionally, the bulk of observational studies have been graded of low quality (including those used by the WHO for its recommendations). Furthermore, many of the reported small-effect sizes are, arguably, within the propagated “noise” level of the observational methods themselves.
Relatively consistently, the detrimental consequences of increased sugar intake have been limited to the consumption of sugar-sweetened beverages (SSB), not to the consumption of food sugars. In this case, one must question whether the apparent uniqueness of SSB is due to methodological rather than biological distinction. For instance, is the apparent significance due to the fact that one can tally cans or bottles of sodas more accurately and precisely than one can estimate food sugars by
dietary intake methods, leading to a statistically significant effect size for the former due merely to less error of the estimate? Additionally, “artificially sweetened” beverages containing no sugar (i.e., the null-sugar control) also often show an apparent, albeit usually less strong, association with detrimental noncommunicable disease outcomes. These findings are often interpreted by creatively constructed “plausible” causal metabolic mechanisms while the most parsimonious explanations deal with recognition
of the role of “reverse causality” and acknowledgement of the fact that the “negative sugar control” (i.e., the non-SSB) effect size is simply an indication of the extent of the methodological “noise” of dietary intake methods.
On the other hand, with several notable exceptions, an extensive body of meta-analytical data of randomized controlled sugar reports (including those of sucrose and fructose) that controlled for total energy consumption have failed to confirm the causal implications of the observational data. Likewise, reports of authoritative committees (e.g., the UK Scientific Committee on Nutrition) that evaluated this issue comprehensively over many years also failed to confirm the widely postulated detrimental effects of sugar consumption per se [5]. These points will be discussed.
References
1. Seidelmann SB, Claggett B, Cheng S, et al: Dietary carbohydrate intake and mortality: a prospective cohort study and meta-analysis. Lancet Online 2018:3.e419 e428.
2. Trepanowski JF, Ioannidis JPA: Perspective: limiting dependence on nonrandomized studies and improving randomized trials in human nutrition research: why and how. Adv Nutr 2018;9:367–377.
3. Patel CJ: Analytic complexity and challenges in identifying mixtures of exposures associated with phenotypes in the exposome era. Curr Epidemiol Rep 2017;4:22–30.
4. Te Morenga L, Mallard S, Mann J: Dietary sugars and body weight: systematic and meta-analyses of randomised controlled trials and cohort studies. BMJ 2012;346:e7492.
5. SACN (Scientific Advisory Committee on Nutrition): Carbohydrates and Health. London, The Stationery Office, 2015.
Abstract
Over the course of evolution, Mother Nature preserved the ability of humans to make every sugar they need for metabolic functions. Glucose is the almost exclusive fuel preferred by the human brain. Human infants are born with sweet taste receptors, sugars are a significant energy source in human milk, and mammals have a direct gut-to-brain sugar-sensing system that enhances development of a preference for sugars. If sugars are as toxic as many postulate, what species advantage was conferred by this evolutionary progression? Observational studies have reported that sugar consumption is associated with various adverse health risks. However, observational studies can never prove causality, dietary intake records are known to be highly problematic, and the huge number of correlation interdependencies among environmental “exposome” variables makes it impossible to attribute causality to individual dietary components. Additionally, these studies overall have been graded as low quality, and many reported the small effect sizes are likely within the propagated methodological “noise.” With several exceptions, data from randomized controlled trials that ensured isocaloric energy intakes have failed to confirm the causal implications of the observational data. Likewise, the comprehensive UK Scientific Committee on Nutrition Report on Carbohydrates and Health also failed to confirm the vast majority of widely postulated detrimental effects of sugar consumption per se. Current data on intakes of sugar-sweetened beverages and on the risks associated with high intakes of dietary fructose remain under debate.Historical Background
For more than 50 years, nutritionists and physicians have debated the optimal macronutrient composition of one’s diet that will promote health and minimize the risk of developing chronic diseases. Over time, the pendulum has swung from higher/lower percentages of dietary fats to higher/lower percentages of dietary carbohydrates (for one cannot change one without correspondingly changing the other in the opposite direction, if one wants to maintain energy balance) and back again. At present, prevailing opinions place fats in the ascendency and carbohydrates in the descendancy. Over the approximately 6 million years of human evolution, our species has evolved the metabolic “machinery” to live well on whatever local macronutrientsare available in sufficient amounts to supply our biological needs. Thus, we are omnivores, and the individual members of our species consume varying diets of diverse composition worldwide. Some diets are very dependent on carbohydrates
as their primary source of energy. Others are principally dependent on fats to supply energy. Despite these very wide differences, when one excludes obvious reasons for accelerated mortality like wars, famines, and the like, humans tend to live similar, relatively long lives irrespective of their differences in macronutrient intakes. For sure, there are differences in the lifespan that can be associated with components of the human diet [1]. The implication of the associations is, of course, that they reflect causality. Nonetheless, given the inadequacy of the methods to determine actual food intakes and the innumerable, unmeasured, unaccounted for environmental variables that constitute covarying interdependencies present within our vast “exosome,” it
is impossible to calculate confidently which and/or how much of the risk differences
are due to diets alone or to the macronutrient contents of those diets. In fact, it is difficult to know for sure if any of the differences are due to diet alone. Nonetheless, this has not prevented expert groups from trying [1] although the nonlinear nature of the associations further compounds the difficulty and confounds the conclusions [2]. Recognizing these limitations, in western societies reflected by the UK, the lowest allcause mortality risks appear to be associated with dietary intakes of about 40–55% energy from carbohydrates and on the order of 30–40% energy from dietary fats [3]; good news since, overall, these are the ranges in most commonly consumed western diets. Now, as difficult as it is to ascribe causal longterm risk conclusions to dietary
macronutrients, the discussion following will attempt to focus on the constituent components of one of the macronutrient classes, dietary sugars. Needless to say, related discussion about individual fats constituting dietary fat intakes is also taking place elsewhere [4].
Axioms
Before doing so, however, it is critical to lay out the axioms underlying the theorem under development. First, sugars are not, per se, defined as dietary essential nutrientsin the traditional nutrition sense since the human body can make every sugar it needs. In fact, I might use the latter observation as an argument that Mother Nature considers sugars as the truly essential nutrients because She did not leave their adequate internal supply to the vagaries of the available external food supply. Furthermore, in directly related sense, dietary sugars do contain the essential external nutrient, energy, a critical essential nutrient for sustaining human life in many societies worldwide. However, relatively recently, consumption of excess energy has been a leading contribution to the development of human obesity worldwide. So, the first axiom in my thesis below is that sugars should not be consumed in amounts that contribute to excess energy intake. This axiom applies to dietary fat or protein intakes as well since they, too, are sources of energy. In fact, dietary fats contribute considerably more energy per gram than dietary sugars. For example, Mozaffarian et al. [5] showed that increased consumption of red and processed meat contributes as much, or more, to weight gain over a 4year period than consumption of sugarsweetened beverages (SSBs). The second axiom for agreement a priori is that the consumption of dietary
sugars should not be in an amount that interferes with the consumption of the 40 or so essential dietary nutrients necessary to satisfy nutritional adequacy. No single food is an adequate source of all essential nutrients, and dietary sugars should not be consumed in quantities that reduce intakes of the variety of foods necessary to establish and maintain nutritional sufficiency. Although the fraction of dietary energy contributed by sugars in some individuals in western societies now appears to approach the level necessary to convey this risk, the amount of sugars consumed by most people still does not. Thus, given the 2 axiomatic conditions, the following exposition will focus on the risks of sugars per se on human health risks. In other words, I will focus on sugars themselves and not on human misuse of dietary sugar consumption. I do so because much of the literature either confuses the two, or because one direction of the literature argues aggressively for the fact that sugars per se are detrimental to humans.
Arguments from an Evolutionary Perspective
Table 1 summarizes various observations that support a thesis that human evolutionarydevelopment is inconsistent with the implication that sugars are toxic substances in themselves. Of course, no arguments of this kind prove causality,
but taken together they provide a strong body of evidence that speaks against a proposition that sugars are toxic substances in their own right. First, as mentioned above, humans can synthesize every sugar they need while they must obtain 2 fatty acids and 8 or 9 amino acids from external dietary sources. It would appear entirely counterintuitive that evolutionary progression would proceed in this direction if sugars were toxic per se. Secondly, glucose is the only significant fuel for the mammalian fetus. This fact has been established conclusively by decades of experimental studies and clinical observations. Moreover, there are specific placental glucose transporters and other fetoplacental substrate metabolic exchange mechanisms that optimize the delivery of glucose to the fetus. It defies credulity that evolution of the mammalian species progressed in a manner that established glucose as the principal, almost exclusive, fetal fuel and simultaneously developed fetoplacental mechanisms to ensure an adequate fetal glucose supply were glucose is a toxic substance. It is well established that human infants are born with sweet taste receptors. One must seriously consider why this would be an evolutionary advantage if developing a taste for sweetness was detrimental to infant survival, health, and development. Furthermore, the sugar content of human milk is considerably higher than that of other mammalian species. Approximately one third of the energy supplied by human milk comes from the disaccharide lactose. One must again question the evolutionary advantage of progression to a species milk supply high in sugar if, in fact, the sugar was detrimental to the species. Even more importantly, mammals are born with a sugar sensing system in the gut that “speaks” directly to the brain, promoting a behavioral, developmental pathway for sugar preference [6]. What possible species advantage is conferred by evolving such a mechanism if the end result proved toxic to the species?
Finally, the human brain prefers, almost exclusively, glucose as the fuel for its activities. The brain consumes and oxidizes approximately two thirds of the glucose produced daily by the liver. I believe that few would argue with the “philosophical” thesis that the human brain represents the highest evolutionary achievement in biology as we know it. If so, what is a possible, even remotely plausible, reason for the brain’s choice of a simple sugar as its preferred (almost only) source of fuel for its activities, if this sugar, itself, were a toxic substance? Additionally, in the context of dietary sugar consumption, we must consider the following mass balance issue. Brain glucose oxidation results in irreversible conversion of glucose carbon to expired carbon dioxide. In other words, net glucose carbon leaves the body forever. This is no trivial loss of carbon given the fact that the brain oxidizes the majority of glucose produced daily by the liver. Although
the liver can make new glucose via gluconeogenesis, the only new internal carbons for net new glucose synthesis come from amino acid carbons stored as muscle proteins. Carbons made into glucose via the Cori cycle are carbons recycled via lactate and pyruvate. As such, they are not new carbons and cannot replace carbons irreversibly lost via CO2. For this reason, if there is no dietary source of glucose or of gluconeogenic carbons available from other ingested macronutrients, the body must break down muscle protein to feed the brain. Depending on age or feeding state, this carbon requirement can amount to > 100 g of “glucose equivalents” daily. This is clearly not a desirable circumstance over the long term. In the only controlled human experiment of its kind of which I am aware [7], we know that the need for external carbon sources to make new glucose fuel for the brain is eventually reduced by reduction in brain energy needs and by ketone body replacement for glucose as the fuel source for the brain.
Finally, the human brain prefers, almost exclusively, glucose as the fuel for its activities. The brain consumes and oxidizes approximately two thirds of the glucose produced daily by the liver. I believe that few would argue with the “philosophical” thesis that the human brain represents the highest evolutionary achievement in biology as we know it. If so, what is a possible, even remotely plausible, reason for the brain’s choice of a simple sugar as its preferred (almost only) source of fuel for its activities, if this sugar, itself, were a toxic substance? Additionally, in the context of dietary sugar consumption, we must consider the following mass balance issue. Brain glucose oxidation results in irreversible conversion of glucose carbon to expired carbon dioxide. In other words, net glucose carbon leaves the body forever. This is no trivial loss of carbon given the fact that the brain oxidizes the majority of glucose produced daily by the liver. Although
the liver can make new glucose via gluconeogenesis, the only new internal carbons for net new glucose synthesis come from amino acid carbons stored as muscle proteins. Carbons made into glucose via the Cori cycle are carbons recycled via lactate and pyruvate. As such, they are not new carbons and cannot replace carbons irreversibly lost via CO2. For this reason, if there is no dietary source of glucose or of gluconeogenic carbons available from other ingested macronutrients, the body must break down muscle protein to feed the brain. Depending on age or feeding state, this carbon requirement can amount to > 100 g of “glucose equivalents” daily. This is clearly not a desirable circumstance over the long term. In the only controlled human experiment of its kind of which I am aware [7], we know that the need for external carbon sources to make new glucose fuel for the brain is eventually reduced by reduction in brain energy needs and by ketone body replacement for glucose as the fuel source for the brain.
Arguments from the Perspective of Metabolism in Childhood
Forty to 50 years ago, there was a concerted effort among pediatric endocrinologists to find a means of identifying in families having a child with type 1 diabetes other children in the same family who might be at increased risk for future development of diabetes. For this reason, the oral glucose tolerance tests (OGTTs) were performed in hundreds of children at a time when most children were lean. Unfortunately, most of these data have vanished from the present view. The standard test dose was 1.75 g of glucose per kilogram ideal body weight up to a maximum dose of 100 g of glucose. A 12oz can of soda contains about 40 g of sugar, approximately half as glucose and half as fructose. Since fructose has essentially no effect on the post ingestion circulating glucose level, the maxi
mum glucose dose during an OGTT is equivalent to the glucose dose in about 5 cans of soda. Despite the magnitude of the oral glucose load during an OGTT, approximately 10% of children show no clear peak circulating glucose level above the fasting level in the hours immediately following glucose ingestion. In other words, the insulin secretory capacity and the peripheral insulin sensitivity in these children are such that they are able to dispose of an oral glucose dose very rapidly without any clearly discernable rise in the circulating glucose level. In OGTTs carried out in 200 children by Guthrie et al. [8], 50% of the children had blood glucose levels < 131 mg/dL 30 min after glucose ingestion, and 50% had blood glucose levels < 110 mg/dL 1 h after consumption of the glucose dose (Fig. 1) [8]. Knopf et al. [9] performed OGGTs in 100 children and found almost identical results. In this case, the vast majority of children had a blood glucose
level < 120 mg/dL by 60 min after ingestion, when the mean blood glucose level was approximately 100 mg/dL, not much higher than the fasting blood glucose value. In other words, healthy lean children have a very high capacity to deal with a glucose load. Nearly 2 decades ago, Sunehag et al. [10–12] and Treuth et al. [13] engaged in a series of studies confirming that lean children and adolescents were capable of dealing rapidly to alterations in dietary macronutrient content. The group conducted a series of dietary studies in which the subjects consumed isocaloric, isonitrogenous diets for 7 days that were either high in fat (55% energy) or high in carbohydrates (60% energy). As such, to maintain a constant energy intake, both diets correspondingly contained 30% energy from carbohydrates or 25% energy from fat, respectively [10–14]. In some of the subjects, the dietary carbohydrate content was constituted so that fructose was responsible for 6 or 24% of dietary energy [12]. The subjects’ energy balance was assessed using room respiration calorimetry, the proportion of macronutrient fuels oxidized was calculated from the respiratory quotient, glucose production rates, gluconeogenic rates, and lipolytic rates were quantified from stable, isotopically labeled substrate infusions, insulin secretory dynamics and insulin sensitivity were measured
using stable isotopically labeled intravenous or oral glucose tolerance tests with accepted minimal modeling approaches [10–14]. These studies showed clearly that healthy, lean children are fully capable of appropriately adjusting their energy expenditure and the proportions of dietary carbohydrate and/or fat oxidized to match wide changes in the macronutrient contents of their diets [13]. Substrate fuel kinetics were minimally affected [10]. Because of their high levels of insulin sensitivity, prepubertal children were able to adapt to a high carbohydrate intake with essentially no change in insulin secretion while adolescents did so with an appropriate increase in both insulin secretion and peripheral insulin sensitivity [10]. Dietary fructose content ranging from 6 to 24% energy intake has no effect on any of these parameters [10, 12, 13]. Again, healthy lean children are perfectly capable of adjusting their internal metabolic milieu and its regulatory “machinery” to accommodate wide changes in dietary macronutrient composition. Once a child becomes obese, however, his or her ability to improve insulin sensitivity when challenged with a high carbohydrate diet
is impaired [11, 12]. Correspondingly, these obese individuals had to increase their insulin secretory rates more than twofold to maintain circulating glucose homeostasis [11, 12]. Nevertheless, ingestion of dietary fructose at intake levels responsible for 6–24% total energy (approximately 45–180 g fructose daily) had no effect on insulin secretion or insulin sensitivity (Fig. 2). Nor was there any effect on glucose kinetics or lipolysis. It is important to recognize here that the situation might be different in adults with established obesity since very carefully controlled SSB trials by Stanhope et al. [15, 16] were able to demonstrate that dietary fructose decreased insulin sensitivity in overweight/obese adults [15] and increased cardiovascular risk factors [15, 16], the latter not measured in the pediatric studies of Sunehag et al. [10–12, 14] and Treuth et al. [13]. Moreover, Stanhope and her colleagues from multiple academic institutions have recently provided a rationale for fructose as a causal agent in the development of
hepatic insulin resistance [17].
level < 120 mg/dL by 60 min after ingestion, when the mean blood glucose level was approximately 100 mg/dL, not much higher than the fasting blood glucose value. In other words, healthy lean children have a very high capacity to deal with a glucose load. Nearly 2 decades ago, Sunehag et al. [10–12] and Treuth et al. [13] engaged in a series of studies confirming that lean children and adolescents were capable of dealing rapidly to alterations in dietary macronutrient content. The group conducted a series of dietary studies in which the subjects consumed isocaloric, isonitrogenous diets for 7 days that were either high in fat (55% energy) or high in carbohydrates (60% energy). As such, to maintain a constant energy intake, both diets correspondingly contained 30% energy from carbohydrates or 25% energy from fat, respectively [10–14]. In some of the subjects, the dietary carbohydrate content was constituted so that fructose was responsible for 6 or 24% of dietary energy [12]. The subjects’ energy balance was assessed using room respiration calorimetry, the proportion of macronutrient fuels oxidized was calculated from the respiratory quotient, glucose production rates, gluconeogenic rates, and lipolytic rates were quantified from stable, isotopically labeled substrate infusions, insulin secretory dynamics and insulin sensitivity were measured
using stable isotopically labeled intravenous or oral glucose tolerance tests with accepted minimal modeling approaches [10–14]. These studies showed clearly that healthy, lean children are fully capable of appropriately adjusting their energy expenditure and the proportions of dietary carbohydrate and/or fat oxidized to match wide changes in the macronutrient contents of their diets [13]. Substrate fuel kinetics were minimally affected [10]. Because of their high levels of insulin sensitivity, prepubertal children were able to adapt to a high carbohydrate intake with essentially no change in insulin secretion while adolescents did so with an appropriate increase in both insulin secretion and peripheral insulin sensitivity [10]. Dietary fructose content ranging from 6 to 24% energy intake has no effect on any of these parameters [10, 12, 13]. Again, healthy lean children are perfectly capable of adjusting their internal metabolic milieu and its regulatory “machinery” to accommodate wide changes in dietary macronutrient composition. Once a child becomes obese, however, his or her ability to improve insulin sensitivity when challenged with a high carbohydrate diet
is impaired [11, 12]. Correspondingly, these obese individuals had to increase their insulin secretory rates more than twofold to maintain circulating glucose homeostasis [11, 12]. Nevertheless, ingestion of dietary fructose at intake levels responsible for 6–24% total energy (approximately 45–180 g fructose daily) had no effect on insulin secretion or insulin sensitivity (Fig. 2). Nor was there any effect on glucose kinetics or lipolysis. It is important to recognize here that the situation might be different in adults with established obesity since very carefully controlled SSB trials by Stanhope et al. [15, 16] were able to demonstrate that dietary fructose decreased insulin sensitivity in overweight/obese adults [15] and increased cardiovascular risk factors [15, 16], the latter not measured in the pediatric studies of Sunehag et al. [10–12, 14] and Treuth et al. [13]. Moreover, Stanhope and her colleagues from multiple academic institutions have recently provided a rationale for fructose as a causal agent in the development of
hepatic insulin resistance [17].
Arguments from the Perspective of Estimating Adult Chronic Disease Risks
Today, a significant fraction of criticism leveled at the consumption of dietary sugars focuses on the role of sugars as causative agents that increase adult chronic disease risk, particularly the risk for cardiovascular disease (CVD). Necessarily, most of the studies assessing these risks are observational in nature. In this context, it is critical to recognize the limitations of the methods widely used in such studies, and how they contribute to the confidence one can assign to the estimated increased relative risks, particularly to relative risk changes that are not much higher than 1.0 [18–20]. These limitations are widely ignored by authors who not only uncritically report small increases in relative risks, but also frequently use causal wording and/or imply causality when, in fact, an observational study can never prove causality, no matter how good or large it
might be. For example, a recent observational study [21] on the risks of meat
consumption reported that increased absolute al lcause mortality risks were < 2% over a 30year period but claimed that higher intakes of processed meat and unprocessed red meat were statistically associated with a 3–7% increase in incident CVD and all cause mortality. Given the propagated known errors associated with food intake methods and food composition tables, and the uncertainties introduced by unmeasured unaccounted for “exosome” confounders and their interdependencies, it is inconceivable that risk differences of this magnitude reflect detection of a true causal signal outside the level of methodological “noise” [21].
might be. For example, a recent observational study [21] on the risks of meat
consumption reported that increased absolute al lcause mortality risks were < 2% over a 30year period but claimed that higher intakes of processed meat and unprocessed red meat were statistically associated with a 3–7% increase in incident CVD and all cause mortality. Given the propagated known errors associated with food intake methods and food composition tables, and the uncertainties introduced by unmeasured unaccounted for “exosome” confounders and their interdependencies, it is inconceivable that risk differences of this magnitude reflect detection of a true causal signal outside the level of methodological “noise” [21].
In 2015, the UK Scientific Advisory Committee on Nutrition (UKSACN) published its extensive report on carbohydrates and health [22]. This 384page report was 9 years in the making, and its analyses were documented with 2,765 pages of supporting data. It is highly unlikely that another report as comprehensive as this one will be published anytime in the foreseeable future. Moreover, despite more recent additional publications on the relationships among dietary sugar intake and chronic disease risks, none of these, in my estimation, changes the findings of UKSACN. Tables 2–4 present a summary of the principal conclusions of the UKSACN report [22]. UKSACN found no association of dietary sugar intake with coronary events, blood pressure, total or LDL cholesterol, fasting triglycerides, fasting glucose or insulin, or the risk of type 2 diabetes mellitus (T2DM) (Table 2) [22], nor did it find a relationship of SSB intake and BMI, body fat, or colon cancer (Table 2) [22]. In cohort studies (Table 3), there was insufficient evidence to conclude that a relationship existed among mono and disaccharide intakes and CVD events, or sugar rich foods or SSB intakes and coronary events, blood pressure, body fat, fat distribution, and weight gain, or sugar rich food consumption and T2DM, or SSB consumption and hypertension, energy intake, glycemia, insulinemia,
and insulin sensitivity (Table 3) [22]. From reported randomized controlled trials
(Table 4), there was insufficient evidence to conclude that relationships existed among sugar intake and vascular compliance, CRP, impaired glucose tolerance or HbA1C, and caries in mixed/permanent teeth [22]. The report was also unable to conclude from randomized controlled trials that relationships existed among SSB intakes and fasting blood lipids, glycemia, insulinemia, insulin sensitivity, body weight, weight gain, and energy intake [22]. On the other hand, UKSACN was able to conclude that there were “fairly consistent” data showing that sugar intake can result in higher energy intakes,
even though this effect appeared to be the result of increased energy intake. Only
1 trial provided diets that were “designed to be isoenergetic and differ in the type of CHO provided.” Moreover, this study was confounded by some subjects losing weight, some subjects gaining weight on lower energy intakes, and on “a spontaneous increase in energy intake on the high fat intervention relative to the three high carbohydrate interventions” [22]. Finally, although UKSACN concluded that an effect was demonstrated in observational studies “between sugar sweetened beverage consumption and higher incidence of type 2 diabetes mellitus,” confidence in this finding was weakened by the lack of association between sugar consumption and T2DM, the failure to demonstrate an association between SSB intake and BMI or body fat, and the inability to exclude confounding [22]. Naturally, additional studies have appeared in the literature since the publication of the UKSACN report 5 years ago. One of these [23], a systematic review and dose response metaanalysis of 24 prospective cohort studies in 624,128 individuals, concluded that “current evidence supports a threshold of harm for intakes of total sugars, added sugars, and fructose at higher exposure and lack of harm for sucrose independent of form for CVD mortality” although the quality of evidence for that conclusion was low [23]. The thresholds for harm were
133 g of total sugars (26% energy), 58 g of fructose (11% energy), and 65 g of added sugars (13% energy). One must interpret these data cautiously. They failed to show a relationship among sugar intakes and the incidence of CVD, nor was an association found between added sugar intake and CVD mortality across the whole dietary intake range. Paradoxically, sucrose intake was associated with decreased CVD mortality risk [23]. Moreover, the increased relative risk of CVD mortality was on the order of 8–9%, a change likely within the propagated “noise level” of observational methods [23]. A related analysis [24] was unable to find an association among total sugars or fructose intakes and T2DM while sucrose intake, again, was associated with a decreased risk of T2DM. Finally, in 6 cohorts totaling 240,506 subjects, every additional SSB consumed from none to ≥1 daily increased the incident hypertension risk by 8.2% [25].
Conclusion
It is almost impossible today to read an article, either in the nutritional literature or in the public media, that does not conclude that dietary sugars are harmful. However, there are a series of arguments based on evolutionary preservation of favorable traits, the ability of lean individuals, particularly non obese children, to readily metabolize dietary sugars, and the limited degree of confidence to which one can ascribe causal implications from observational studies reporting small changes in relative risks that should give one pause in believing that sugars in the diet are as sour as they are commonly made out to be.
Conflict of Interest Statement
In the last 3 years, Dr. Bier has received consultant and/or lecture fees and/or reimbursements for travel, hotel, and other expenses from the International Life Sciences Institute; the International Council on Amino Acid Science; Nutrition, and Growth Solutions, Inc.; Ajinomoto, Co.; the Lorenzini Foundation; the Nutrition Coalition, the CrossFit Foundation; the International Glutamate Technical Committee; Nestlé S.A.; Ferrero SpA; Indiana University; the National Institutes of Health; Mallinckrodt Pharmaceuticals; the Infant Nutrition Council of America; and the Israel Institute.
References
1. GBD 2017 Diet Collaborators: Health effects of dietary risks in 195 countries, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet 2019; 393: 1958–1972.
2. Ho FK, Gray SR, Welsh P, et al: Associations of fat and carbohydrate intake with cardiovascular disease and mortality: prospective cohort study of UK Biobank participants. BMJ 2020; 368:m688.
3. Seidelmann SB, Claggett B, Cheng S, et al: Dietary carbohydrate intake and mortality: a prospective cohort study and meta-analysis. Lancet Online 2018; 3:e419–e428. www.thelancet.com/publichealth.
4. Astrup A, Magkos F, Bier DM, et al: Saturated fats and health: a reassessment of evidence and proposal for food-based recommendations. J Am Coll Cardiol 2020; 76. https://www.onlinejacc.org/ content/accj/76/7/844.full.pdf.
5. Mozaffarian D, Hao T, Rimm EB, et al: Changes in diet and lifestyle and long-term weight gain in women and men. N Engl J Med 2011; 364: 2392– 2404.
6. Tan H-E, Sisti AC, Jin H, et al: The gut-brain axis mediates sugar preference. Nature 2020; 580: 511– 516.
7. Owen OE, Morgan AP, Kemp HG, et al: Brain metabolism during fasting. J Clin Invest 1967; 46: 1589–1595.
8. Guthrie RA, Guthrie DW, Murthy DYN, et al: Standardization of the oral glucose tolerance test and the criteria for diagnosis of chemical diabetes in children. Metabolism 1973; 22: 275–282.
9. Knopf CF, Cresto JC, Dujovne IL, et al: Oral glucose tolerance test in 100 normal children. Acta Diabetol Lat 1977; 14: 95–103.
10. Sunehag AL, Toffolo G, Treuth MS, et al: Effects of dietary macronutrient content on glucose metabolism in children. J Clin Endocrinol Metab 2002; 87: 5168–5178.
11. Sunehag AL, Toffolo G, Campioni M, et al: Effects of dietary macronutrient intake on insulin sensitivity and secretion and glucose and lipid metabolism in healthy, obese adolescents. J Clin Endocrinol Metab 2005; 90: 4496–4502.
12. Sunehag AL, Toffolo G, Campioni M, et al: Shortterm high dietary fructose intake had no effects on insulin sensitivity and secretion or glucose and lipid metabolism in healthy, obese adolescents. J Pediatr Endocr Metab 2008; 21: 225–235.
13. Treuth MS, Sunehag AL, Trautwein LM, et al: Metabolic adaptation to high-fat and high-carbohydrate diets in children and adolescents. Am J Clin Nutr 2003; 77: 479–489.
14. Sunehag AL, Dalla Man C, Toffolo G, et al: β-Cell function and insulin sensitivity in adolescents from an OGTT. Obesity 2008; 17: 233–239.
15. Stanhope KL, Schwarz JM, Keim NL, et al: Consuming fructose-sweetened, not glucose-sweetened, beverages increases visceral adiposity and lipids and decreases insulin sensitivity in overweight/ obese humans. J Clin Invest 2009; 119: 1322–1334.
16. Stanhope KL, Medici V, Bremer AA, et al: A doseresponse study of consuming high-fructose corn syrup-sweetened beverages on lipid-lipoprotein risk factors for cardiovascular disease in young adults. Am J Clin Nutr 2015; 101: 1144–1154.
17. Softic S, Stanhope KL, Boucher J, et al: Fructose and hepatic insulin resistance. Crit Rev Clin Lab Sci 2020; 14: 1–15.
18. Trepanowski JF, Ioannidis JPA: Perspective: limiting dependence on nonrandomized studies and improving randomized trials in human nutrition research: why and how. Adv Nutr 2018; 9: 367– 377.
19. Archer E, Lavie CJ, Hill JO: The failure to measure dietary intake engendered a fictional discourse on diet-disease relations. Front Nutr 2018; 5: 105.
20. Archer E, Marlow ML, Lavie CJ: Controversy and debate: memory-based methods paper 2: the fatal flaws of food frequency questionnaires and other memory-based dietary assessment methods. J Clin Epidemiol 2018; 104: 113–124.
21. Zhong VW, Van Horn L, Greenland P, et al: Associations of processed meat, unprocessed red meat, poultry, or fish intake with incident cardiovascular disease and all-cause mortality. JAMA Intern Med 2020; 180: 503–512.
22. SACN (Scientific Advisory Committee on Nutrition): Carbohydrates and Health. London, The Stationery Office, 2015.
23. Khan TA, Mobusha T, Agarwal A, et al: Relation of total sugars, sucrose, fructose, and added sugars with the risk of cardiovascular disease: a systematic review and dose-response meta-analysis of prospective cohort studies. Mayo Clin Proc
2019; 94: 2399–2414.
24. Tsilas CS, de Souza RJ, Blanco Mejia S, et al: Relation of total sugars, fructose and sucrose with incident type 2 diabetes: a systematic review and meta-analysis of prospective cohort studies. CMAJ 2017; 189:E711–E720.
25. Jayalath VH, de Souza RJ, Ha V, et al: Sugarsweetened beverage consumption and incident hypertension: a systematic review and meta-analysis of prospective cohorts. Am J Clin Nutr 2015; 102: 914–921.
and insulin sensitivity (Table 3) [22]. From reported randomized controlled trials
(Table 4), there was insufficient evidence to conclude that relationships existed among sugar intake and vascular compliance, CRP, impaired glucose tolerance or HbA1C, and caries in mixed/permanent teeth [22]. The report was also unable to conclude from randomized controlled trials that relationships existed among SSB intakes and fasting blood lipids, glycemia, insulinemia, insulin sensitivity, body weight, weight gain, and energy intake [22]. On the other hand, UKSACN was able to conclude that there were “fairly consistent” data showing that sugar intake can result in higher energy intakes,
1 trial provided diets that were “designed to be isoenergetic and differ in the type of CHO provided.” Moreover, this study was confounded by some subjects losing weight, some subjects gaining weight on lower energy intakes, and on “a spontaneous increase in energy intake on the high fat intervention relative to the three high carbohydrate interventions” [22]. Finally, although UKSACN concluded that an effect was demonstrated in observational studies “between sugar sweetened beverage consumption and higher incidence of type 2 diabetes mellitus,” confidence in this finding was weakened by the lack of association between sugar consumption and T2DM, the failure to demonstrate an association between SSB intake and BMI or body fat, and the inability to exclude confounding [22]. Naturally, additional studies have appeared in the literature since the publication of the UKSACN report 5 years ago. One of these [23], a systematic review and dose response metaanalysis of 24 prospective cohort studies in 624,128 individuals, concluded that “current evidence supports a threshold of harm for intakes of total sugars, added sugars, and fructose at higher exposure and lack of harm for sucrose independent of form for CVD mortality” although the quality of evidence for that conclusion was low [23]. The thresholds for harm were
133 g of total sugars (26% energy), 58 g of fructose (11% energy), and 65 g of added sugars (13% energy). One must interpret these data cautiously. They failed to show a relationship among sugar intakes and the incidence of CVD, nor was an association found between added sugar intake and CVD mortality across the whole dietary intake range. Paradoxically, sucrose intake was associated with decreased CVD mortality risk [23]. Moreover, the increased relative risk of CVD mortality was on the order of 8–9%, a change likely within the propagated “noise level” of observational methods [23]. A related analysis [24] was unable to find an association among total sugars or fructose intakes and T2DM while sucrose intake, again, was associated with a decreased risk of T2DM. Finally, in 6 cohorts totaling 240,506 subjects, every additional SSB consumed from none to ≥1 daily increased the incident hypertension risk by 8.2% [25].
Conclusion
It is almost impossible today to read an article, either in the nutritional literature or in the public media, that does not conclude that dietary sugars are harmful. However, there are a series of arguments based on evolutionary preservation of favorable traits, the ability of lean individuals, particularly non obese children, to readily metabolize dietary sugars, and the limited degree of confidence to which one can ascribe causal implications from observational studies reporting small changes in relative risks that should give one pause in believing that sugars in the diet are as sour as they are commonly made out to be.
Conflict of Interest Statement
In the last 3 years, Dr. Bier has received consultant and/or lecture fees and/or reimbursements for travel, hotel, and other expenses from the International Life Sciences Institute; the International Council on Amino Acid Science; Nutrition, and Growth Solutions, Inc.; Ajinomoto, Co.; the Lorenzini Foundation; the Nutrition Coalition, the CrossFit Foundation; the International Glutamate Technical Committee; Nestlé S.A.; Ferrero SpA; Indiana University; the National Institutes of Health; Mallinckrodt Pharmaceuticals; the Infant Nutrition Council of America; and the Israel Institute.
References
1. GBD 2017 Diet Collaborators: Health effects of dietary risks in 195 countries, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet 2019; 393: 1958–1972.
2. Ho FK, Gray SR, Welsh P, et al: Associations of fat and carbohydrate intake with cardiovascular disease and mortality: prospective cohort study of UK Biobank participants. BMJ 2020; 368:m688.
3. Seidelmann SB, Claggett B, Cheng S, et al: Dietary carbohydrate intake and mortality: a prospective cohort study and meta-analysis. Lancet Online 2018; 3:e419–e428. www.thelancet.com/publichealth.
4. Astrup A, Magkos F, Bier DM, et al: Saturated fats and health: a reassessment of evidence and proposal for food-based recommendations. J Am Coll Cardiol 2020; 76. https://www.onlinejacc.org/ content/accj/76/7/844.full.pdf.
5. Mozaffarian D, Hao T, Rimm EB, et al: Changes in diet and lifestyle and long-term weight gain in women and men. N Engl J Med 2011; 364: 2392– 2404.
6. Tan H-E, Sisti AC, Jin H, et al: The gut-brain axis mediates sugar preference. Nature 2020; 580: 511– 516.
7. Owen OE, Morgan AP, Kemp HG, et al: Brain metabolism during fasting. J Clin Invest 1967; 46: 1589–1595.
8. Guthrie RA, Guthrie DW, Murthy DYN, et al: Standardization of the oral glucose tolerance test and the criteria for diagnosis of chemical diabetes in children. Metabolism 1973; 22: 275–282.
9. Knopf CF, Cresto JC, Dujovne IL, et al: Oral glucose tolerance test in 100 normal children. Acta Diabetol Lat 1977; 14: 95–103.
10. Sunehag AL, Toffolo G, Treuth MS, et al: Effects of dietary macronutrient content on glucose metabolism in children. J Clin Endocrinol Metab 2002; 87: 5168–5178.
11. Sunehag AL, Toffolo G, Campioni M, et al: Effects of dietary macronutrient intake on insulin sensitivity and secretion and glucose and lipid metabolism in healthy, obese adolescents. J Clin Endocrinol Metab 2005; 90: 4496–4502.
12. Sunehag AL, Toffolo G, Campioni M, et al: Shortterm high dietary fructose intake had no effects on insulin sensitivity and secretion or glucose and lipid metabolism in healthy, obese adolescents. J Pediatr Endocr Metab 2008; 21: 225–235.
13. Treuth MS, Sunehag AL, Trautwein LM, et al: Metabolic adaptation to high-fat and high-carbohydrate diets in children and adolescents. Am J Clin Nutr 2003; 77: 479–489.
14. Sunehag AL, Dalla Man C, Toffolo G, et al: β-Cell function and insulin sensitivity in adolescents from an OGTT. Obesity 2008; 17: 233–239.
15. Stanhope KL, Schwarz JM, Keim NL, et al: Consuming fructose-sweetened, not glucose-sweetened, beverages increases visceral adiposity and lipids and decreases insulin sensitivity in overweight/ obese humans. J Clin Invest 2009; 119: 1322–1334.
16. Stanhope KL, Medici V, Bremer AA, et al: A doseresponse study of consuming high-fructose corn syrup-sweetened beverages on lipid-lipoprotein risk factors for cardiovascular disease in young adults. Am J Clin Nutr 2015; 101: 1144–1154.
17. Softic S, Stanhope KL, Boucher J, et al: Fructose and hepatic insulin resistance. Crit Rev Clin Lab Sci 2020; 14: 1–15.
18. Trepanowski JF, Ioannidis JPA: Perspective: limiting dependence on nonrandomized studies and improving randomized trials in human nutrition research: why and how. Adv Nutr 2018; 9: 367– 377.
19. Archer E, Lavie CJ, Hill JO: The failure to measure dietary intake engendered a fictional discourse on diet-disease relations. Front Nutr 2018; 5: 105.
20. Archer E, Marlow ML, Lavie CJ: Controversy and debate: memory-based methods paper 2: the fatal flaws of food frequency questionnaires and other memory-based dietary assessment methods. J Clin Epidemiol 2018; 104: 113–124.
21. Zhong VW, Van Horn L, Greenland P, et al: Associations of processed meat, unprocessed red meat, poultry, or fish intake with incident cardiovascular disease and all-cause mortality. JAMA Intern Med 2020; 180: 503–512.
22. SACN (Scientific Advisory Committee on Nutrition): Carbohydrates and Health. London, The Stationery Office, 2015.
23. Khan TA, Mobusha T, Agarwal A, et al: Relation of total sugars, sucrose, fructose, and added sugars with the risk of cardiovascular disease: a systematic review and dose-response meta-analysis of prospective cohort studies. Mayo Clin Proc
2019; 94: 2399–2414.
24. Tsilas CS, de Souza RJ, Blanco Mejia S, et al: Relation of total sugars, fructose and sucrose with incident type 2 diabetes: a systematic review and meta-analysis of prospective cohort studies. CMAJ 2017; 189:E711–E720.
25. Jayalath VH, de Souza RJ, Ha V, et al: Sugarsweetened beverage consumption and incident hypertension: a systematic review and meta-analysis of prospective cohorts. Am J Clin Nutr 2015; 102: 914–921.
