OUP user menu

Beverages and body weight: challenges in the evidence-based review process of the Carbohydrate Subcommittee from the 2010 Dietary Guidelines Advisory Committee

Joanne Slavin
DOI: http://dx.doi.org/10.1111/j.1753-4887.2012.00537.x S111-S120 First published online: 1 November 2012

Abstract

Concern about the role of beverages, especially those containing sugar, in the obesity epidemic continues to escalate. Bans on sugar-sweetened beverages and chocolate milk have expanded from the school cafeteria to the ballpark and convenience store. This review describes the experience of the 2010 Dietary Guidelines Advisory Committee (DGAC) in conducting an evidence-based review of dietary exposure and health outcomes. The following four topics relevant to fluids and body weight were reviewed: added sugar, noncaloric sweeteners, food form and body weight, and macronutrients and satiety. There were limited and conflicting data on how liquids and solids affect energy intake and body weight. Fluid intake is typically not tracked in prospective, cohort longitudinal studies; thus, data are not available on fluid intake and health status from studies using the strongest epidemiologic designs. Despite public perception that beverages are linked to increased body weight compared with whole foods, evidence-based reviews of this topic do not support that liquid calories are processed differently in the body. The practical recommendation to replace caloric beverages with water as an aid to control weight is based on calorie reduction, rather than a link between added-sugar intake and obesity.

  • beverages
  • body weight
  • satiety
  • sugar
  • water

INTRODUCTION

In 1980, Nutrition and Your Health: Dietary Guidelines for Americans was issued in response to the public's desire for authoritative, consistent guidelines on diet and health. Public Law 101-445, Section 3, requires publication of the Dietary Guidelines for Americans (DGA) at least every 5 years. They represent federal nutrition policy established jointly by the US Department of Agriculture (USDA) and the US Department of Health and Human Services (DHHS). They are designed to provide science-based advice for individuals aged ≥2 years to help prevent chronic diseases and promote health. They lay the foundation for federal nutrition programs and nutrition education programs and serve as a basis for research gaps and priorities.

The DGA have evolved over time to be more detailed and more prescriptive. The 2010 Report of the Dietary Guidelines Advisory Committee on the Dietary Guidelines for Americans is 445 pages long,1 while the 2010 Dietary Guidelines for Americans is a 95-page document.2 In contrast, the 1980 Dietary Guidelines for Americans was a short consumer brochure containing the following seven general recommendations: “1) eat a variety of foods; 2) maintain ideal weight; 3) avoid too much fat, saturated fat, and cholesterol; 4) eat foods with adequate starch and fiber; 5) avoid too much sugar; 6) avoid too much sodium; and 7) if you drink alcohol, do so in moderation.”3

The need for “numbers” in the DGA is driven by the relationship between the guidelines and federal nutrition programs.4 Nutrition assistance programs such as the National School Lunch Program and the Special Supplemental Nutrition Program for Women, Infants, and Children are required to base regulations on the most recent scientific knowledge (e.g., the DGA in the United States).

The process of developing the DGA is an open one, with the ability to provide input available to any person or organization who wants to contribute. The 13-member 2010 Dietary Guidelines Advisory Committee (DGAC) was composed of scientists with a broad range of expertise representing nutrition, physical activity, food behavior, and nutritional changes through the life cycle. The DGAC met six times publicly to agree on questions to examine in order to set nutrition policy. The DGAC report is prepared and then presented to the Secretaries of the USDA and the DHHS; for the latest version, this occurred in June 2010. Once the report has been presented, the DGAC is dismissed and has no other input into the DGA. The USDA and DHHS write the policy document.

ADDRESSING QUESTIONS AND EVALUATING EVIDENCE

The DGAC worked in subcommittees to address questions of diet and disease risk. Subcommittees in 2010 included energy balance, carbohydrates and protein, fats, nutrient adequacy, sodium and fluids, and food safety. I served as the chair of the carbohydrate and protein committee and also served as a member of the energy balance committee and the nutrient adequacy committee.

How exactly do the DGAC and the subcommittees go about addressing the agreed-upon questions on the relationships of diet to health outcomes? The 2010 DGAC used an evidence-based review process with a hierarchy of evidence. The strongest evidence was considered to come from randomized controlled trials, preferably double-blinded. Of course, food studies suffer in this arena because it is difficult or impossible to blind food treatments; subjects know that they are consuming an apple or apple juice. These types of trials can work with nutrients, as nutrients can be added to food or drinks without the knowledge of the participants or investigators (double-blind). The next strongest studies are prospective cohort studies, in which a group or cohort of subjects is studied over time. Food frequency instruments are often used to collect dietary information before any diagnosis of disease, making these studies more reliable than cross-sectional studies. These studies are limited by their ability to assess intake of foods or beverages, especially quantitative estimates. Case-control studies, animal research, or in vitro studies were generally excluded in the 2010 DGAC review; cross-sectional studies were only included if no stronger prospective studies were available for review.

The “body of evidence” for each question was then examined and a conclusion statement was crafted. Conclusions were deemed strong, moderate, limited, or lacking data to support them. It was possible for there to be strong evidence for no relationship. For example, the 2010 DGAC found strong evidence of no relationship between the glycemic index and disease outcomes.

Agreeing on the strength of the relationship was difficult because for each question, different types of studies have been published. For each question that the 2010 DGAC addressed in the evidence-based report, the search criteria, inclusion and exclusion criteria for studies, the range of dates searched, and other information used in the review were noted. All studies included in the review are listed in the Nutrition Evidence Library (http://www.nutritionevidencelibrary.com), which will be a resource for future DGA work. The transparency used in an evidence-based approach should minimize bias.5

CHALLENGES IN EVALUATING DIET AND DISEASE RELATIONSHIPS

Inconsistencies in the DGAC report exist, often because of differences in inclusion criteria for studies. For example, limited evidence was found for a relationship between intake of sugar-sweetened beverages and body weight in adults in the review conducted for the carbohydrate chapter of the 2010 DGA, from which cross-sectional studies were excluded.1 In contrast, strong evidence was found between intake of sugar-sweetened beverages and body weight in children when cross-sectional studies were included in the review conducted by the Energy Balance Committee.1 Hite et al.6 reviewed issues with contradictory evidence in the DGAC 2010 report.

Dietary recommendations to prevent chronic diseases have always been controversial. Alfred E. Harper, in his paper, “Killer French Fries: The Misguided Drive to Improve the American Diet,” describes our ways of learning about nutrient deficiencies and how such a model will not work for chronic diseases such as heart disease and cancer.7 He also points out the misinformation included in the early DGA reports. For example, fruits are listed as a source of “complex carbohydrates,” when in reality they are mostly a source of sugar and are poor sources of nutrients including protein and certain vitamins and minerals. The high-protein quality and quantity of animal products has been lost in our translation of dietary guidance for public health. As Harper suggests, clinical advice to change diet based on the need to lower serum cholesterol is much different than public health advice to suggest that all Americans should consume plant foods of low protein quality.

AN EVIDENCE-BASED APPROACH

Although the recommendations of the DGA have not changed significantly since the 1980s, the development of the DGA policy has become more open and science based. Questions on the relationship between dietary exposure and disease outcome are challenging and contentious. The following four topics relevant to the discussion of fluids and body weight were reviewed by the 2010 DGAC and are discussed below: 1) added sugars, 2) noncaloric sweeteners, 3) food form and body weight, and 4) macronutrients and satiety. The questions, conclusions, and implications related to each topic are presented as they appear in the 2010 DGAC report.1

Sugars

Sugars are found naturally in fruits, milk, and milk products. They also are added to foods during processing, preparation, or at the table.8 These added sugars sweeten the flavor of foods and beverages and improve their palatability. They are also used in food preservation and to confer functional attributes, such as viscosity, texture, body, and browning capacity. They provide calories but insignificant amounts of vitamins, minerals, or other essential nutrients. The Nutrition Facts Label provides information on total sugars per serving, but does not distinguish between sugars naturally present in foods and added sugars. There is no method to measure added sugars; added-sugar values are calculated values.

Recommended intakes of sugars

In its 2002 report titled Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids,9 the Institute of Medicine (IOM) established a recommended dietary allowance for carbohydrate of 130 g/day. This value is based on the amount of sugars and starches required to provide the brain with an adequate supply of glucose. The IOM set an acceptable macronutrient distribution range (AMDR) for carbohydrate of 45–65% of total calories. The Dietary Reference Intake Committee concluded that evidence was insufficient to set a tolerable upper intake level for carbohydrates9; however, the panel suggested a maximal intake level of ≤25% of total calories from added sugars. This suggestion is based on trends indicating that people with diets at or above this level of added sugars are more likely to have poorer intakes of important essential nutrients.9

Adults should consume 45–65% of their total calories from carbohydrates, except for younger children who need a somewhat higher proportion of fat in their diets. Vegetables, fruits, whole grains, milk, and milk products are the major food sources of carbohydrates. Grains and certain vegetables including corn, potatoes, and dried beans and peas are rich in starch, whereas sweet potatoes are mostly sucrose, not starch. Fruits and dark green vegetables contain little or no starch. Regular soft drinks, sugar/sweets, sweetened grains, and regular fruitades/drinks comprise 72% of the intake of added sugar.10

Marriott et al.10 examined the intake of added sugars and selected nutrients from the 2003–2006 National Health and Nutrition Examination Survey data. Thirteen percent of the population had added-sugar intake of =25% of calories. The predominant issue of concern was the overall high calorie content and low quality of the US diet, not added sugars.

SELECTED CARBOHYDRATES AND HEALTH OUTCOMES: QUESTIONS FROM THE 2010 DGAC

The following questions were asked regarding selected carbohydrates and health outcomes. 1) In adults, what are the associations between intake of sugar-sweetened beverages and energy intake and body weight? 2) How are noncaloric sweeteners related to body weight? 3) What is the impact of liquids versus solid foods on energy intake and body weight? 4) How do carbohydrates, fiber, protein, fat, and foods influence satiety?

All human study designs were originally included in the literature searches, but cross-sectional studies were later excluded from the review if there was sufficient evidence from studies with stronger study designs. Well-conducted systematic reviews and meta-analyses of original research articles were considered by the 2010 DGAC.

The original search for studies of added sugars was broad and included terms such as added sugars, dietary sucrose, and candy, as well as various terms for sugar-sweetened beverages. However, few studies were identified that looked at added sugars other than by counting sugar-sweetened beverage consumption in food frequency instruments; thus, sugar-sweetened beverages were the focus of the review.

To determine the effect of food form on energy intake, studies that compared a liquid to a solid or a semi-solid form were searched. Furthermore, only articles that considered energy intake and/or body weight were reviewed. Additional research on food form and appetite, hunger, and related outcomes was available, but these outcomes were not included in the 2010 DGAC review. The following question was considered by the Carbohydrate Subcommittee of the 2010 Dietary Guidelines Advisory Committee.

DGAC question.  “In adults, what are the associations between intake of sugar-sweetened beverages and energy intake and body weight?”1

DGAC conclusion statement.  Limited evidence shows that intake of sugar-sweetened beverages is linked to higher energy intake in adults. A moderate body of epidemiologic evidence suggests that greater consumption of sugar-sweetened beverages is associated with increased body weight in adults. A moderate body of evidence suggests that under isocaloric controlled conditions, added sugars, including sugar-sweetened beverages, are no more likely to cause weight gain than any other source of energy.1

Implications.  “Added sugars are not different than other extra calories in the diet for energy intake and body weight. Thus, reducing intake of added sugars is a recommended strategy to reduce calorie intake in Americans. Intake of caloric beverages, including sugar-sweetened beverages, sweetened coffee and tea, energy drinks, and other drinks high in calories and low in nutrients should be reduced in consumers needing to reduce body weight.”1

Review of the evidence

The role of dietary carbohydrates in the current obesity epidemic is much debated. Saris11 concluded that fat content of the diet is the most important contributor to overconsumption of calories and that carbohydrate content, regardless of carbohydrate type, is relatively benign, with little evidence for direct negative effects of dietary sugar on body weight. Van Baak and Astrup12 determined there is insufficient evidence to conclude that an exchange of sugar for nonsugar carbohydrates in the context of a reduced-fat ad libitum diet or energy-restricted diet results in lower body weight. They also noted that observational studies suggest, a possible relationship between consumption of sugar-sweetened beverages and body weight, but that current supporting evidence from randomized controlled trials of sufficient size and duration was insufficient to support a difference between liquid and solid sugar intake in body weight control.

Most reviews have investigated whether or not intake of sugar-sweetened beverages is linked to obesity. The prevalence of obesity has increased in the past 30 years, and at the same time, consumption of soft drinks has increased.13 Olsen and Heitmann reviewed the literature on calorically sweetened beverages and obesity, relative to adjustment for energy intake, and no cross-sectional studies were included. They concluded that a high intake of calorically sweetened beverages can be regarded as a determinant for obesity. However, there seems to be no support for an association between intake of calorically sweetened beverages and obesity, as mediated through increased energy intake, which suggests that alternative biological explanations should be explored. Other studies that examined obesity risk and intake of sugar-sweetened beverages in adults in the United States found no associations.14

Methodologic challenges

Sugar is a ubiquitous term, but one that is not easy to define and measure. Analytical methods measure total sugar in foods and nutrient databases and Nutrition Facts labels include values for total sugars.8 Added sugars are typically calculated values and can be estimated using dietary assessment tools in nutrition studies. Exact definitions of sugar are often omitted from studies, making it difficult to determine exactly what was under investigation.15 This hinders the ability to compare studies. Studies can report specific sugars such as sucrose, glucose, or fructose, or use the term sugar to mean mono- and disaccharides. The term “total sugars” means all dietary sugars, whether added or naturally occurring, and “sugar-containing” is thought to mean foods and beverages that contain sugar. In epidemiologic studies, it is often easier to assess intake of sugar-sweetened beverages as these can be counted in food frequency instruments. This tends to be nonspecific because fruitades, fruit punches, sport drinks, energy drinks, and juices may or may not be counted in these systems.

Two studies in the United Kingdom used nonmilk extrinsic sugars and reported an inverse relationship between nonmilk extrinsic sugars and body mass index (BMI),16,17 although no such relationship was found between body weight status and sugar intake in a New Zealand population.18 Thus, the assessment of added sugars or extrinsic sugars is challenging because no analytical methods exist to measure sugars added to foods. In addition, studies use different techniques to assess intake of added sugars. Reliable and standardized measures of exposure to added sugars are necessary to draw meaningful conclusions. Currently, the best assessments involve counting the frequency of intake of sugar-sweetened beverages in epidemiologic studies.

Are sugar-sweetened beverages and energy intake linked?

To answer this question, the 2010 DGAC reviewed one meta-analysis19 and three trials2022 published since 1990. Vartanian et al.19 conducted a meta-analysis that examined the association between soft drink consumption and various health outcomes, including energy intake. This analysis included some unpublished data as well as cross-sectional studies. Vartanian et al.19 conducted separate analyses based on study design and outcomes. Of the 88 studies in the review, 3 longitudinal studies and 11 experimental studies examined the relationship between soft drink consumption and energy intake in adults. Although the effect size was small, the authors concluded there was a positive association between soft drink intake and energy intake.

Two additional primary studies also support a relationship between the intake of sugar-sweetened beverages and increased energy intake. Flood et al.20 examined the impact of beverage type (cola, diet cola, or water) and size (12 or 18 fl oz) on intake at an ad libitum lunch. Energy intake from food consumed at lunch did not differ across conditions. However, when the energy from beverages was added to the energy consumed from food, the mean total energy intake at lunch was greater when regular cola was served compared with the other beverages, regardless of portion size.

Reid et al.21 compared the effects of supplementary soft drinks sweetened with sucrose or aspartame, added to the diet over 4 weeks, on dietary intake in normal-weight women. Participants consumed four 250-mL bottles of drink per day. Sucrose supplements provided 430 kcal/day and aspartame supplements provided <20 kcal/day. For those consuming the sucrose drink, daily energy intake was higher during the intervention phase than at baseline; women consuming the sugar-sweetened beverages consumed about 200 kcal more energy each day.

Soenen and Westerterp-Plantenga22 examined the satiating effects of high-fructose corn syrup (HFCS) and sucrose in comparison with milk and a diet drink. In this trial, participants completed four test sessions that included an ad libitum meal served after one of four beverages: one containing sucrose, one containing HFCS, one containing milk, and a diet drink. All four drinks were isovolumetric (800 mL). The energy drinks were isocaloric. Test meal energy intake was lower after consumption of preloads containing sucrose, HFCS, or the milk preload (with no differences between the energy-containing preloads) than after the diet drink preload. Total energy intake (preload + meal) with the energy-containing preloads was significantly higher than total energy intake with the diet drink preload. During the meal, energy intake from the beverage was partly compensated. However, compensation for energy intake from the preloads containing sucrose, HFCS, or milk did not differ significantly and ranged from 30% to 45%. This study indicated that although energy intake was higher following the drinks sweetened with HFCS and sucrose compared to a diet drink preload, energy intakes were not different from the milk preload, indicating that the added sugar did not have a unique effect on energy intake.

Are sugar-sweetened beverages and body weight linked?

The 2010 DGAC addressed this question by reviewing four systematic reviews,1524 four trials,2127 and four prospective observational studies.2831 However, the systematic reviews did not use consistent methods to evaluate added sugars. Typical search terms included the following: soft drinks, sugar-sweetened beverages, liquid sugar, and soda. The systematic reviews used different criteria to review the literature, and three reviews1924 included cross-sectional studies, which are considered a weaker epidemiologic design than prospective, cohort studies. Malik et al.24 attempted a meta-analysis, but the degree of heterogeneity among study designs made a more qualitative assessment necessary. Vartanian et al.19 attempted to separate the effects of different study designs. Studies with experimental designs (five studies) showed no association with added-sugar intake for body weight for adults. Significant relationships were found among longitudinal studies (three studies) of the relationship between added-sugar intake and body weight, although the effect size was small. Similarly, Malik et al.24 concluded that epidemiologic and experimental data indicated that increased consumption of sugar-sweetened beverages is associated with weight gain and obesity.

In contrast, Gibson23 reviewed six longitudinal and one intervention study with adults and concluded that sugar-sweetened beverages are a source of energy, but that little evidence showed that they are any more obesogenic than any other source of energy. Ruxton et al.15 concluded that the evidence does not suggest a positive association between BMI and sugar intake. However, some studies that specifically involve sweetened beverages highlight a potential concern in relation to obesity risk. The methods used for these systematic reviews varied and may explain the discrepancies.

The four trials included in the 2010 DGAC review varied greatly in design. In general, with calorie intake controlled, there were no differences in weight gain when participants consumed diets with a higher (versus lower) percentage of calories from added sugars.2527 When energy intake was not controlled, Reid et al.21 found a nonsignificant trend for weight gain among normal-weight women consuming four regular soft drinks per day, compared with those consuming diet soft drinks. In a trial by Stanhope et al.26 that included 25% of energy from beverages sweetened with glucose or fructose, weight gain was observed when participants consumed self-selected diets in an outpatient setting.

The 2010 DGAC also reviewed four prospective studies. Lower consumption of soft drinks was linked to weight loss in the PREMIER study.28 A reduction in sugar-sweetened beverage intake of one serving per day was associated with a weight loss of approximately 0.5 kg at 6 months and 18 months, and a significant dose-response trend between change in body weight and change in sugar-sweetened beverage intake was also observed. Over a mean follow-up of 4 years in the Framingham Heart Study,29 consumption of ≥1 soft drinks per day was associated with increased odds of developing obesity and increased waist circumference compared to drinking none. Stookey et al.31 compared four weight-loss diets and predicted that replacing sweetened caloric beverages with water would save 200 kcal per day over 12 months. Although weight loss might be expected due to lower energy intake, the study by Stookey et al.31 was not an intervention trial and thus did not measure change in body weight.

Randomized controlled trials report that added sugars are not different from other calories in increasing energy intake or body weight. Prospective studies report some relationship with sugar-sweetened beverages and weight gain, but it is not possible to determine if these relationships are merely linked to additional calories, as opposed to added sugars per se. The systematic reviews in this area are also inconsistent, probably based on different methods used to determine added intake of sugars or of sugar-sweetened beverages. Weed et al.32 conducted a systematic evaluation of reviews on sugar-sweetened beverages and health outcomes; they found that reviews of sugar-sweetened beverages and health outcomes received moderately low-quality scores by the assessment of multiple systematic reviews rating system. The following question was considered by the Carbohydrate Subcommittee of the 2010 Dietary Guidelines Advisory Committee.

DGAC question.  “How are noncaloric sweeteners related to energy intake and body weight?”1

DGAC concluding statement.  “Moderate evidence shows that using noncaloric sweeteners will affect energy intake only if they are substituted for higher-calorie foods and beverages. A few observational studies reported that individuals who use noncaloric sweeteners are more likely to gain weight or be heavier. This does not suggest that noncaloric sweeteners cause weight gain rather that they are more likely to be consumed by overweight and obese individuals.”1

Implications.  “The replacement of sugar-sweetened foods and beverages with sugar-free products should theoretically reduce body weight. Yet many questions remain, as epidemiologic studies show a positive link with use of nonnutritive sweeteners and BMI, whereas animal studies suggest that inclusion of nonnutritive sweeteners in the diet promotes energy intake and contributes to obesity.”1

Review of the evidence

Replacing sugar with low-calorie sweeteners is a common strategy to facilitate weight control.33 Intense sweeteners help lower the energy density of beverages and foods, which should result in lower energy intakes. Mattes and Popkin34 estimate that 15% of the US population ingests nonnutritive sweeteners, but that percentage is increasing. Concern about the negative effects of diet soft drink consumption on energy intake came from animal studies that suggested that increased food intake and weight gain follow prolonged exposure to saccharin-sweetened yogurt.35 This animal research suggested that artificial sweeteners “uncouple” the relationship between sweet taste and energy, which promoted rats to consume more food and gain weight.

The use of noncaloric sweeteners has increased greatly over the past 3 decades while the incidence of obesity has also risen. Thus, cross-sectional studies suggest that intake of noncaloric sweeteners is positively associated with increased obesity. If noncaloric sweeteners are used as substitutes for higher-energy-yielding sweeteners, they have the potential to aid in weight management, but whether they will be effective in this regard is not found in the existing literature.

Noncaloric sweeteners, energy intake, and body weight conclusion

If noncaloric sweeteners are substituted for higher-calorie sweeteners in foods or beverages, they are associated with weight loss. However, observational studies find that individuals who use noncaloric sweeteners are more likely to gain weight or be heavier. This seems paradoxical, but rather than indicating that noncaloric sweeteners cause weight gain, the findings reveal that noncaloric sweeteners are more likely to be used by overweight and obese individuals. This conclusion is based on one meta-analysis,36 one randomized crossover study,20 and one prospective cohort study37 published since 2006.

The meta-analysis performed by de la Hunty et al.36 supports a significant reduction in energy intakes with aspartame compared with control diets, except when aspartame was compared with nonsucrose controls, such as water. For body weight, the analysis was conducted in three stages: 1) used all weight outcomes including follow-up weights, 2) excluded studies in which the control group gained weight, and 3) excluded follow-up periods. A significant reduction in weight was seen for all three analyses. The combined effect was approximately a 3% reduction in body weight. The authors concluded that using foods and drinks sweetened with aspartame instead of sucrose resulted in a significant reduction in both energy intakes and body weight. Furthermore, use of foods and drinks sweetened with aspartame instead of those sweetened with sucrose was an effective way to maintain and lose weight.

In a prospective cohort study, Fowler et al.37 reported a significant positive dose-response relationship between baseline artificially sweetened beverage consumption and incidence of overweight/obesity, incidence of obesity, and BMI change; however, this association does not establish causality.

Flood et al.20 examined the impact of beverage type (cola, diet cola, or water) and size (12 or 18 fl oz) on intake during an ad libitum lunch. Participants consumed significantly more energy at lunch when cola was provided versus diet cola or water. The following question was considered by the Carbohydrate Subcommittee of the 2010 Dietary Guidelines Advisory Committee.

DGAC question.  “What is the impact of liquid versus solid foods on energy intake and body weight?”1

DGAC concluding statement.  “A limited body of evidence shows conflicting results about whether liquid and solid foods differ in their effects on energy intake and body weight, except that liquids in the form of soup may lead to decreased energy intake and body weight.”1

Implications.  “In general, if total calorie content is held constant, there is little support for any effects on energy intake and body weight due to the calories consumed either as a liquid or solid. Some studies suggest that whole foods may be more satiating than liquid foods. Food structure, specifically a whole food (apple, carrots), plays a role in satiety and decreasing food intake at subsequent meals, yet fiber added to a drink is not effective in reducing food intake at subsequent meals. Soup as a preload decreases food intake at a subsequent meal. Thus, Americans are advised to pay attention to the calorie content of the food or beverage consumed, regardless of whether it is a liquid or solid. Calories are the issue in either case.”1

Review of the evidence

Twelve studies with no consistent experimental design were included in the DGAC review of the impact of liquid versus solid foods on energy intake and body weight.. One study, the PREMIER trial, compared intake of liquid calories to solid calories.28 Six of the studies were crossover trials that investigated the impact of a preload before breakfast38 or lunch3943 on ad libitum intake of a meal. An additional crossover trial44 examined the intake of carrots in various forms with a meal rather than as a preload. DiMeglio and Mattes45 conducted a crossover trial that included two 4-week interventions with daily consumption of liquid (caffeine-free soda) or solid (jelly beans) food. Finally, three studies4648 examined soup as the liquid form.

No standard protocol has been established to answer this question, and information on food form and consumption of liquid is not collected in prospective cohort trials. Most of the available evidence to answer this question is from preload studies, in which meals are controlled for total calories and macronutrient content, and then satiety is measured for 3 h after the meal. Subsequent food intake is then measured by consumption of a buffet lunch, with food intake for 24 h then calculated.

Chen et al.28 examined beverage consumption in the prospective PREMIER study at baseline, 6 months, and 18 months. Analyses considered changes in volume, calorie intake, and percentage of calories from beverages both overall and from seven categories (sugar-sweetened beverages, diet drinks, milk, 100% juices, coffee and tea with sugar, coffee and tea without sugar or with artificial sweeteners, and alcoholic beverages). A reduction of 100 kcal/day in liquid calorie intake was associated with an approximate 0.25 kg weight loss at 6 and 18 months. In comparison, a reduction in solid calorie intake by 100 kcal/day was associated with a <0.1 kg weight loss at 6 and 18 months. Reductions in liquid calorie intake had a stronger effect on weight loss than did a reduction in solid calorie intake, but the difference was only statistically significant at 6 months. A significant dose-response trend between change in body weight and change in liquid calorie intake was observed at 6 and 18 months.

Consumption of solid food compared to juice in a controlled caloric load may decrease energy intake at a subsequent meal. Flood-Obbagy and Rolls40 examined how consuming preloads of apples in different forms (apple, applesauce, and apple juice with and without added fiber) influenced energy intake of a meal. Study participants consumed fewer calories at lunch after consuming apples compared to equal calories as applesauce, apple juice, or apple juice with added fiber. In a similar study, whole carrots were associated with less calorie intake for the remainder of the day compared to carrot juice or a carrot juice cocktail that contained all the nutrients in carrots.44

Mourao et al.42 investigated the independent effect of food form on appetite and energy intake in lean and obese adults using high-carbohydrate, -fat, or -protein food stimuli. Experimental treatments involved matched beverage and solid food forms, including high carbohydrate (watermelon and watermelon juice), high protein (cheese and milk), and high fat (coconut meat and coconut milk). Participants consumed the entire test food as part of an ad libitum meal. Regardless of the predominant energy source, the beverage form elicited a weaker compensatory dietary response than the matched solid food form. The inclusion of a caloric beverage in a lunch meal led to greater daily energy intake, compared to customary intake or days in which a solid version of the same food was ingested. This occurred regardless of the primary energy source, and there was no clear indication that lean and obese individuals differ in this regard.

Stull et al.38 assessed the effect of liquid versus solid meal replacements on appetite and subsequent food intake in healthy older adults. After an overnight fast, subjects consumed meal replacement products as either a liquid or as a solid (bar) followed by ad libitum oatmeal. Subjects consumed more calories from oatmeal after the liquid versus solid meal-replacement product.

DiMeglio and Mattes45 examined the effects of matched liquid (soda) and solid (jelly beans) carbohydrate loads on diet and body weight. Participants were assigned to one of two dietary load conditions (solid: 450 kcal serving of jelly beans; liquid: 450 kcal serving of caffeine-free soda) for 4 weeks, followed by a 4-week washout period and subsequent participation in the other condition for 4 weeks. During the solid load condition, subjects compensated for some of the energy in the test foods by reducing free-feeding intake such that the overall compensation score was 118%. However, when the liquid load was included in the diet, no compensation was observed, resulting in a compensation score of −17%. The authors concluded that liquid carbohydrate promotes positive energy balance, whereas a comparable solid carbohydrate elicits dietary compensation; furthermore, body weight and BMI increased only with the liquid load.

Other studies that were well controlled for macronutrients and calories found no differences in subsequent food intake when solid foods were compared to liquids. Mattes and Campbell41 assessed the effects of apple food form (apple, applesauce, apple juice) and timing of eating events (meal or snack) on appetite and daily energy intake. There were no treatment effects on daily energy intake. Almiron-Roig et al.39 compared the impact on energy intakes of equal-energy preloads (300 kcal) of regular cola or fat-free cookies presented either 2 h or 20 min before a tray lunch. Liquid or solid form had no impact on energy intakes during the test meal. Similarly, physical form had no effect when the sum of the energy intake of breakfast, preload, and lunch was considered.

In another crossover trial,43 participants consumed 200-kcal preloads of semi-solid peach yogurt with peach pieces, peach yogurt homogenized to liquid form, peach syrup and water, or a milk-based peach and apricot beverage followed by an ad libitum lunch. No significant differences in energy intakes were detected across the four conditions, either for lunch alone or for total energy consumed from breakfast, preload, and lunch.

Liquids in soup may have different effects as studies find that adults who consume soup daily have lower energy intakes than those who consume little soup48; furthermore, soup preloads reduce food intake at a subsequent meal.47 Rolls et al.46 tested the effect on weight loss of a diet incorporating one or two servings per day of foods equal in energy but differing in energy density. Participants followed an energy-restricted diet in a 1-year trial (6-month weight loss and 6-month weight maintenance) in which they were randomized to one of four intervention groups. Participants were instructed to consume the following daily: one serving of soup, two servings of soup, or two servings of dry snack foods. Participants in the fourth group were not provided with any specific food to consume (comparison group). There were no significant differences in reported energy intake among the intervention groups at any time points. All four groups showed significant weight loss at 6 months that was well maintained at 12 months. The magnitude of weight loss, however, differed by group. At 1 year, weight loss in the comparison (8.1 ± 1.1 kg) and two-soup (7.2 ± 0.9 kg) groups was significantly greater than that in the two-snack group (4.8 ± 0.7 kg); weight loss in the one-soup group (6.1 ± 1.1 kg) did not differ significantly from other groups. The authors concluded that on an energy-restricted diet, consuming two servings of low-calorie soup daily led to 50% greater weight loss than consuming the same amount of energy as high energy-dense snack food.

When the macronutrient content of a liquid food and a solid food is balanced, there are little data that food form affects energy intake. These studies are difficult to design and conduct because the form of the food cannot be blinded (i.e., participants know that they are eating apples or drinking apple juice). In the acute studies of food intake, efforts are made to control variables, including the time allowed to consume the test food, but it is difficult to generalize these results to the eating environments of real life.

Food structure may play a role in food intake. Whole foods such as apples and carrots play a role in satiety and decrease food intake at a subsequent meal. When a nonviscous fiber was added to apple juice, the fiber-enriched apple juice was not as effective as an apple in reducing food intake at a subsequent meal.40 Thus, factors in addition to the fiber in whole foods may affect energy intake, including food structure and chewing.

The data with soup as a preload apparently conflict with other data on liquid calories. In a 1-year weight-loss trial,48 consumption of two servings of soup per day led to greater weight loss than consuming the same amount of energy from two snack foods. Soup preload significantly reduced test meal and total meal energy intake in one study.47 Thus, the studies with soup as a liquid calorie source suggest that liquid calories can be an aid to weight loss and that liquid calories from soup result in reduced intake at a subsequent meal.

The following question was considered by the Carbohydrate Subcommittee of the 2010 Dietary Guidelines Advisory Committee.

DGAC question.  “What is the role of carbohydrate, fiber, protein, fat, and food form on satiety?”1

DGAC conclusion statement.  “Many factors affect satiety and most studies are conducted in laboratory settings to control for variables. Thus results may not be generalized to the more complicated eating environment of the outside world. Foods high in dietary fiber generally are more satiating than low fiber foods, although some fibers added to drinks have little impact on satiety. Overall, small changes in the macronutrient content of the diet do not significantly alter satiety.”1

Implications.  “Intakes of caloric preloads, whether carbohydrate, protein, or fat, typically increase satiety. Protein and carbohydrate may be more satiating than fat, although studies are not consistent. Dietary fiber, especially from whole foods, appears to enhance satiety in studies. Not all fibers added to beverages or foods are equally satiating. In fact, some functional fibers show no effect on satiety.”1

CONCLUSION

Limited data exist which indicate that added-sugar intake is linked to adverse health outcomes. Generally, intake of all types of carbohydrates is linked to lower body weight in prospective cohort studies.49 Since it is difficult to measure added sugar in epidemiologic studies, intake of sugar-sweetened beverages is the proxy for added-sugar intake. In prospective cohort studies there are little data to indicate that intake of sugar-sweetened beverages is linked to higher energy intake or body weight in adults.

Food modeling is a process in which USDA food patterns are used to describe the types and amounts of foods that can be consumed to achieve nutritional adequacy. In these examples, foods and beverages with added sugar removed are more nutrient dense. Although these examples work well in computer-based food modeling, it is likely that such interventions will not improve food choices.

The cereal industry has worked diligently to remove added sugar from breakfast cereals; however, in many cases, refined starches have replaced added sugars.8 The grams of added sugar are reduced in the product, but the starch content increases; thus, there are no differences in the calorie content of the cereal. This, obviously, will have no positive impact on body weight or public health.

Chocolate milk is another example of the difficulties with a policy based on reducing added sugar content in the food supply. Many school districts are banning chocolate milk because of its relatively high content of added sugar. Although the rationale for such regulations is that children will select the default beverage “low-fat milk” instead of chocolate milk, most studies show that if flavored milk is not available, children choose not to drink milk at all. Thus, the unintended consequence of a well-intentioned policy (i.e., reduce added-sugar intake by banning chocolate milk) is that children consume less milk.8 The nutrients provided by the dairy group, protein, vitamin D, and calcium, for example, are not easily obtained in other typical eating patterns in the United States.

Efforts to micromanage the diet by imposing strict dietary rules are difficult to support with evidence-based nutrition science. Concepts, such as added sugar, that work well in computer-based food modeling are not helpful to consumers attempting to select and consume a “healthier diet.” Clear label information about total calories per portion would be of benefit to consumers attempting to control calorie intake. Nutrient needs across the life cycle vary greatly, so general advice, although well meaning, may actually be harming health status and making the obesity epidemic worse.

Acknowledgments

Declaration of interest.  The author has no relevant interest to declare.

REFERENCES

View Abstract