Increased Protein Consumption during the Day from an Energy-Restricted Diet Augments Satiety but Does Not Reduce Daily Fat or Carbohydrate Intake on a Free-Living Test Day in Overweight Women

Abstract

Background: Higher-protein (HP) energy-restriction diets improve weight management to a greater extent than normal-protein (NP) versions. Potential mechanisms of action with regard to assessment of eating behaviors across the day have not been widely examined during energy restriction.

Objectives: The objectives of this study were to test whether the consumption of an HP energy-restriction diet reduces carbohydrate and fat intakes through improvements in daily appetite, satiety, and food cravings compared with NP versions and to test whether protein type within the NP diets alters protein-related satiety.

Methods: Seventeen overweight women [mean ± SEM age: 36 ± 1 y; body mass index (kg/m²): 28.4 ± 0.1] completed a randomized, controlled-feeding crossover study. Participants were provided with the following ∼1250-kcal/d energy-restricted (−750-kcal/d deficit) diets, each for 6 d: HP [124 g protein/d; 60% from beef and 40% from plant sources (HP-BEEF)] or NP (48 g protein/d) that was protein-type matched (NP-BEEF) or unmatched [100% from plant-based sources (NP-PLANT)]. On day 6 of each diet period, participants completed a 12-h testing day containing repetitive appetite, satiety, and food-craving questionnaires. On day 7, the participants were asked to consume their protein requirement within each respective diet but were provided with a surplus of carbohydrate- and fat-rich foods to consume, ad libitum, at each eating occasion across the day. All outcomes reported were primary study outcomes.

Results: The HP-BEEF diet reduced daily hunger by 16%, desire to eat by 15%, prospective food consumption by 14%, and fast-food cravings by 15% but increased daily fullness by 25% compared with the NP-BEEF and NP-PLANT diets (all P < 0.05). However, consuming more protein throughout the day did not reduce the energy consumed ad libitum from the fat- and carbohydrate-rich foods (HP-BEEF: 2000 ± 180 kcal/d; NP-BEEF: 2120 ± 190 kcal/d; NP-PLANT: 2070 ± 180 kcal/d). None of the outcomes differed between the NP-BEEF and NP-PLANT treatments.

Conclusions: Although appetite control, satiety, and food cravings improved after an HP energy-restriction diet, increased protein consumption did not reduce carbohydrate and fat intakes throughout the free-living test day in overweight healthy women exposed to highly palatable foods. This trial was registered at clinicaltrials.gov as NCT02614729.

Introduction

The movement to adopt healthful lifestyle practices to reduce the prevalence of noncommunicable diseases, including obesity and diabetes, has garnered global interest (1). As such, the desire to consume more protein-rich foods is a commonly used strategy due to the documented improvements in weight management observed with higher-protein (HP) compared with normal-protein (NP) diets (2–5). These improvements have been shown in HP diets ranging from 1.2 to 1.6 g protein · kg body weight⁻¹ · d⁻¹ (3). One postulated mechanism through which increased protein intake leads to greater weight (and fat) loss includes the improvements in ingestive behavior (3) through increased satiation, which is the process of meal termination, and increased satiety, which is the process to prevent further eating after a meal or snack (6).

A recent review that included 24 acute trials found that the consumption of ≥28 g protein/meal consistently increased satiety, as shown by increased postprandial fullness compared with lower protein quantities (3). This was also supported by a recent meta-analysis that assessed the impact of HP meals on postprandial fullness (7). Although a majority of the acute trials did not report protein intakes on a gram of protein per kilogram of body weight per day basis, ∼30 g protein consumed at every meal across the day for the average woman would fall within the protein range listed above for weight management (i.e., 1.5–1.6 g protein · kg body weight⁻¹ · d⁻¹). Regardless of these findings, several knowledge gaps remain with respect to the effect of protein on satiety.

First, the majority of published studies that assess satiety include acute, single-day breakfast-meal designs. This approach may reduce the validity of the data due to the unfamiliarity of the HP test meal because breakfast is typically lower in protein (3, 4). Furthermore, it is unclear as to whether the satiety effects after a single meal are representative of each eating occasion throughout the day.

Most studies also fail to include protein-type–matched HP and NP comparisons (8–18). Due to the varying amino acid compositions within plant and animal proteins, and the potential subsequent effects on appetite and satiety (19–26), these studies fail to assess a true protein quantity effect due to the varying protein sources included.

Last, the published acute studies comparing HP with NP meals were primarily performed during energy balance conditions (20, 27–32). Because energy restriction and weight loss are often accompanied by elevated hunger and blunted satiety, there is an increased inability to maintain energy restriction. Thus, it is unclear as to whether increased protein consumption throughout the day counteracts the elevated hunger and can improve satiety to achieve increased dietary compliance during energy restriction.

Thus, the primary aim of this study was to examine the effects of consuming protein-type–matched HP compared with NP diets on daily satiety during subchronic energy restriction, defined as reduced caloric intake of −750 kcal/d over 6 d in overweight women. In addition, because protein type may have the most robust effect on satiety when protein is provided in smaller amounts (33), a secondary aim of this study was to examine the satiety effects of consuming NP diets containing plant- or animal-based proteins.

Finally, there is also a lack of assessment of ad libitum daily food intake after HP compared with NP diets. The limited evidence is a result of the tightly controlled nature of most long-term diet studies and the lack of daily food assessments within acute meal studies. Although a few studies showed reduced daily energy intake after HP compared with NP diets (3, 34), it is unclear whether this occurs through reductions in high-fat, high-carbohydrate foods consumed as between-meal snacking, within-meal “sides,” or end-of-meal desserts, etc. Thus, although exploratory, this study also assessed the potential effects of protein consumption on within-meal and end-of-meal ad libitum fat and carbohydrate intakes throughout multiple eating occasions across the day.

Methods

Experimental design

A subchronic, tightly controlled, randomized crossover design study was performed in overweight, sedentary (35), but otherwise healthy women. Seventeen women randomly consumed the following 3 isocaloric energy-restriction diets [∼1250 kcal/d (i.e., −750-kcal/d deficit)], each for 6 d: HP, defined as 39% of daily energy as protein (124 g protein/d) with 60% of protein from beef and 40% from plant sources (HP-BEEF), or NP, defined as 15% of daily energy as protein (48 g protein/d) that was protein-type matched (NP-BEEF) or unmatched [100% from plant-based sources (PLANT; NP-PLANT)]. During the first 5 d of each diet, the participants were provided with the respective breakfast, lunch, and dinner meals plus an evening snack and asked to consume these at home or work. On day 6 of each diet, a tightly controlled 12-h testing day was completed, consisting of appetite, satiety, and food cravings assessments performed every 30 min. On day 7, absolute protein (grams) consumption was maintained within each respective diet but carbohydrate- and fat-rich foods were provided ad libitum within a free-living environment.

There is evidence to support that appetite control and food intake are affected by hormonal fluctuations that occur throughout the menstrual cycle (36). Thus, to maintain consistency with respect to these hormonal fluctuations, each of the diets occurred during the follicular phase of the menstrual cycle or during the placebo phase for those using hormonal contraceptives; thus, there were 2- to 3-wk washout periods between diets.

Study participants

From January 2014 to May 2015, healthy, sedentary, but overweight, women were recruited from the Columbia, Missouri, area through advertisements, flyers, and e-mail listservs to participate in the study. Participants were eligible to participate if they met the following criteria: 1) female; 2) aged 18–52 y; 3) overweight [BMI (in kg/m²): 25–29.9]; 4) sedentary (<8000 steps/d); 5) normal menstrual cycles (between 26 and 32 d; 5 in past 6 mo); 6) nonsmoking (for the past year); 7) no metabolic, hormonal, or neural conditions or diseases that influence metabolism, appetite, or cognition; 8) no past history of surgical interventions for the treatment of obesity; 9) no weight loss or gain (≥4.5 kg in the past 6 mo); 10) no medication that would directly influence appetite or cognition; 11) no change in any medications (over the past 3 mo); 12) consumes ≤800 mg caffeine/d and, of this, ≤260 mg caffeine consumed before lunch; 13) not pregnant (or planning to become pregnant); 14) not currently or previously following a specific diet including high protein, vegan, vegetarian, etc.; 15) conventional (typical) and consistent sleep patterns (i.e., awake hours between 0500 and 2300 with no afternoon naps, rates quality of sleep as Fairly to Very Good on the Pittsburg Sleep Quality Index, and averages ≥6 sleep hours/night over the past month; 16) not clinically diagnosed with an eating disorder; 17) scores <4 on the Three-Factor Eating Habits Questionnaire; 18) has a Profile of Mood States, Second Edition (60-item), Depression-Dejection Scale score within 1.5 SD of the age, sex, and race-specific normative mean (37, 38); 19) obtained a “yes” on the validity indicator and received a score of >70 (>2%) on the CNS Vital Signs Test Battery for Cognitive Function (Morrisville, North Carolina); 20) no allergies or aversions to the study foods; 21) no history of drug or alcohol abuse (i.e., >14 drinks/wk, with 15 g ethanol/drink); 22) willing to maintain current inactivity patterns throughout the study; 23) willing to consume all study treatments; 24) habitually consumes breakfast, lunch, and dinner; and 25) generally healthy, as assessed from the medical history questionnaire. Women were the target population in this study because they typically have a lower consumption of daily protein and animal-based proteins (39) but also report a higher propensity to consume a plant-based diet (40).

More than 146 women attended a study information session with the intent to participate; 17 met the screening criteria, signed the study consent, and completed the study procedures (Supplemental Figure 1). Participant characteristics are presented in Table 1. Study purpose, procedures, and risks were explained before participants signed the consent forms. The University of Missouri Health Sciences Institutional Review Board approved this study, and all procedures were followed in accordance with the ethical standards of the institutional review board. The participants received a $350 stipend for completing each 7-d diet. This trial was registered at clinicaltrials.gov as NCT02614729.

TABLE 1.

Characteristics of the women who completed the study¹

	Values
Sample size, n	17
Demographic information
Age, y	36 ± 1
Weight, kg	79.2 ± 0.3
Height, cm	167 ± 1
BMI, kg/m2	28.4 ± 0.1
Habitual purposeful activity, steps/d	2805 ± 255
Weight-stable energy needs, kcal/d	2010 ± 40
Habitual sleep, h/night	7.4 ± 0.2
Habitual dietary information (3-d dietary recalls at baseline)
Energy, kcal/d	2260 ± 580
Protein
g	88 ± 23
%	16 ± 4
Carbohydrate
g	246 ± 69
%	44 ± 7
Fat
g	100 ± 33
%	40 ± 6

¹Values are means ± SEMs unless otherwise indicated, n = 17. Weight-stable energy needs were calculated by using the Mifflin St. Jeor equation, with an activity factor of 1.35 (41).

Dietary treatments (days 1–6)

Weight-stable energy needs (expressed as kcal/d) were estimated for each participant by using the Mifflin St. Jeor equation with an activity factor of 1.35 (i.e., no exercise to little or light exercise) for all participants (41) (Table 1). In addition, participants completed 3-d dietary recalls to identify habitual energy content and macronutrient composition (Table 1).

For 6 d/treatment, participants were placed on an energy-restricted diet, which reduced individual daily intake by approximately −750 kcal of weight-stable energy needs. The energy and macronutrient contents of each treatment are shown in Table 2, and the menu examples for days 1–6 are reported in Supplemental Table 1.

TABLE 2.

Dietary characteristics of the study treatments on days 1–6¹

	NP-PLANT	NP-BEEF	HP-BEEF
Breakfast
Energy content, kcal	354 ± 0	350 ± 0	350 ± 0
Energy density, kcal/100 g	136 ± 8	140 ± 17	140 ± 6
Total protein, g	13 ± 0	14 ± 0	35 ± 0
Beef, g	0 ± 0	9 ± 0	27 ± 0
Carbohydrate, g	48 ± 0	48 ± 0	36 ± 0
Fat, g	12 ± 0	12 ± 0	9 ± 0
Lunch
Energy content, kcal	340 ± 0	340 ± 0	360 ± 0
Energy density, kcal/100 g	171 ± 18	173 ± 25	160 ± 14
Total protein, g	13 ± 0	13 ± 0	35 ± 0
Beef, g	0 ± 0	7 ± 0	22 ± 0
Carbohydrate, g	48 ± 0	48 ± 0	35 ± 0
Fat, g	12 ± 0	12 ± 0	8 ± 0
Dinner
Energy content, kcal	340 ± 0	330 ± 0	360 ± 0
Energy density, kcal/100 g	149 ± 22	133 ± 14	145 ± 9
Total protein, g	13 ± 0	13 ± 0	35 ± 0
Beef, g	0 ± 0	7 ± 0	30 ± 0
Carbohydrate, g	48 ± 0	48 ± 0	34 ± 0
Fat, g	12 ± 0	12 ± 0	10 ± 0
Evening snack
Energy content, kcal	200 ± 0	210 ± 0	210 ± 10
Energy density, kcal/100 g	165 ± 3	189 ± 36	147 ± 6
Total protein, g	8 ± 0	8 ± 0	19 ± 0
Beef, g	0 ± 0	4 ± 0	16 ± 0
Carbohydrate, g	28 ± 0	28 ± 0	20 ± 0
Fat, g	7 ± 0	7 ± 0	6 ± 0
Daily intake
Energy content, kcal	1220 ± 0	1240 ± 0	1280 ± 10
Energy density, kcal/100 g	148 ± 9	144 ± 5	146 ± 7
Total protein, g	47 ± 0	48 ± 0	124 ± 0
Beef, g	0 ± 0	27 ± 0	95 ± 0
Carbohydrate, g	171 ± 0	172 ± 0	125 ± 0
Fat, g	42 ± 0	43 ± 0	33 ± 0
Diet palatability, mm (rating category)	58 ± 3a (slight like)	72 ± 3b (slight like)	69 ± 4b (slight like)

¹Values are means ± SEMs, n = 17. Labeled means in a row without a common superscript letter differ, P < 0.0125. HP-BEEF, higher-protein beef diet (60% of protein from beef and 40% from plant sources); NP-BEEF, normal-protein diet (60% of protein from beef and 40% from plant sources); NP-PLANT, normal-protein plant diet (100% of protein from plant sources).

The participants were provided with 3 daily meals (i.e., breakfast, lunch, and dinner) and an evening snack. Each meal contained 28% of daily energy, whereas the evening snack contained 16% of daily energy. The meals within the HP-BEEF diet pattern were designed such that participants would consume ≥30 g protein/meal, which is an amount proposed to serve as a satiety threshold within a mixed-meal paradigm (4). The rationale behind the inclusion of 2 NP diets was to determine whether protein type provided within the lower range of protein intakes (i.e., NP diets) influences satiety and ingestive behavior.

Last, we chose to include lean beef as the animal-based protein source within this study for several reasons. Lean beef in a low-saturated-fat, heart-healthy diet has been shown to promote cardiovascular health outcomes (42). Furthermore, beef is a high-quality animal protein that has been included within our previous studies with excellent compliance (30, 43). It has also been reported that women typically consume less total protein and less meat than men (39) and have a higher propensity to consume vegetarian or vegan diets (40). Thus, we wanted to explore whether the increase in animal-based proteins improves ingestive behavior outcomes in a population in whom larger quantities of animal-based proteins are not generally consumed.

All of the foods within each diet were fully prepared, cooked, and packaged in the metabolic testing facility. All of the food ingredients were weighed to the nearest tenth of a gram and were precooked before packaging. Participants picked up the meals on the day before each 7-d testing period, and reheating instructions were provided. As an initial measure of compliance, the participants were required to complete meal- and snack-specific food inventory logs. The logs also provided instruction for when the meals were to be consumed. Thus, meal timing was set according to each participant's habitual weekday breakfast meal time, with lunch, dinner, and the evening snack occurring 4, 8, and 10 h postbreakfast, respectively. In addition, participants were instructed to consume only foods provided to them, document all deviations (i.e., foods not consumed or extra foods consumed), and return all wrappers and uneaten foods to be reweighed. Compliance to the diets reported as mean ± SEMs was as followed: HP-BEEF: 99.4% ± 1.0%; NP-PLANT: 99.5% ± 0.8%; and NP-BEEF: 99.6% ± 0.9%.

Twelve-hour, controlled-feeding clinical testing day (day 6)

On day 6 of each pattern, the participants completed the respective 12-h testing day. The participants arrived 1 h before breakfast, after an overnight fast, and a sleep questionnaire (modified Pittsburg Sleep Quality Index) was completed to document the previous night's sleep pattern. The previous night's sleep pattern was determined to be either adequate or inadequate according to a comparison with baseline measures; if inadequate, the testing day was rescheduled. The participant was taken to a self-contained, comfortable, quiet, and well-lit room and resided there throughout the testing day. The room contained a reclining chair, lamp, laptop (with wi-fi), and access to a bathroom. If the participant was a documented morning caffeine user, a 25-mg dose of caffeine was provided as 2.0 g sugar-free, dairy-free instant coffee in 100 g water (at −45 min). Water was provided, ad libitum, throughout each testing day. At −30 min, computerized questionnaires that assessed appetite and satiety were performed. At time +0 min, the respective breakfast was consumed. Throughout the remainder of the day, the same computerized questionnaires were repeated. Lunch was consumed at +240 min, dinner was consumed at +480 min, and an evening snack was consumed at +600 min. A food cravings questionnaire (Pennington Biomedical Research Center, Baton Rouge, Louisiana) (44) was completed at +570 min. At +660 min, the participants left the facility.

Perceived appetite and satiety and sensory questionnaires.

Computerized questionnaires previously used (43, 45, 46), which assessed perceived sensations (i.e., hunger, fullness, desire to eat, prospective food consumption), were completed every 30 min throughout the 12-h clinical testing days. In addition, palatability or food liking (i.e., appearance, aroma, flavor, texture, overall liking) was completed after the first bite and upon completion of each meal and snack during each testing day by using the following categorical identifiers for the 100-mm visual analog scale recordings: 0 mm (extreme dislike), 1–24 mm (strong dislike), 25–49 mm (slight dislike), 50 mm (neutral), 51–74 mm (slight like), 75–99 mm (strong like), and 100 mm (extreme like).

Food cravings.

The Food Cravings Inventory (FCI) was completed at +540 min (postdinner) and was designed to assess food cravings across the entire day. The FCI is a reliable and valid self-report measure of general and specific food cravings, including cravings for high-fat foods, carbohydrates and starches, sweets, and fast-food (44). The FCI defines a craving as an intense desire to consume a particular food (or food type) that is difficult to resist. Furthermore, this questionnaire prompts the participant to report how often (never, rarely, sometimes, often, and always or almost every hour) she experienced a craving for the food throughout the testing day.

Free-living, ad libitum feeding day (day 7).

On day 7 of each pattern, each participant completed a free-living, ad libitum testing day at home, work, or both. All of the food ingredients were weighed to the nearest tenth of a gram and were precooked before packaging. Participants picked up the meals and snack at the end of day 6 for each diet, and reheating instructions were provided. The participants were provided with and were required to consume a specific protein quantity on the basis of their current respective diet treatment. In addition to consuming the required protein, the participants were provided with an excess of carbohydrate- and fat-rich foods (∼1500 kcal/meal, 1000 kcal/snack) at each meal and snack and were permitted to consume these, ad libitum, within a 30-min time period for meals and a 15-min time period for the snack. All of the foods were provided ad libitum and were the same foods across diet treatments. The energy and macronutrient contents of the required and ad libitum–provided foods within each eating occasion are shown in Supplemental Table 2. Participants were provided with meal- or snack-specific food inventory logs and instructed to consume the breakfast, lunch, dinner, and evening snack at the same times as days 1–6. Within the instructions, the participants were required to consume only the respective foods within each meal time and to not eat those foods at any other time throughout the day (i.e., the banana bread was only allowed to be consumed at breakfast, etc.). Participants were asked to keep and return all uneaten foods and to not share their study foods. The foods and containers were weighed before being packed out and re-weighed upon completion of day 7.

Data and statistical analysis

Before the start of the study, power analyses were performed on daily fullness. The effect size for HP compared with NP treatments for daily responses of perceived fullness was 0.94 (47), indicating that an n = 17 provided 90% power to detect differences between treatments. Summary statistics (sample means and SEMs) were computed for the following primary outcomes: hunger, fullness, desire to eat, prospective food consumption, and food intake (i.e., meal-specific and daily energy and macronutrient contents). Net AUC was calculated by using the trapezoidal rule (48) for hunger, fullness, desire to eat, and prospective food consumption. A repeated-measures ANOVA was applied to assess main effects of protein quantity (HP-BEEF compared with NP-BEEF and NP-PLANT) and type (NP-BEEF compared with NP-PLANT) on all study outcomes.

When main effects were detected, paired-samples t tests with a Bonferroni correction for multiple comparisons were applied to compare differences between treatments. Analyses were conducted by using the Statistical Package for the Social Sciences (version 24.0; IBM SPSS). Due to an adjustment for multiple comparisons within each outcome category, P < 0.025 was considered significant.

Results

Appetite control and satiety (day 6)

Figures 1 and 2 show the daily appetite and satiety responses after each study treatment. All of the study diets led to pre- and postprandial fluctuations in hunger, fullness, desire to eat, and prospective food consumption across the day. Between diets, HP-BEEF led to reductions in AUCs for daily hunger, desire to eat, and prospective food consumption as well as increases in daily fullness (all P < 0.01) compared with the NP-BEEF and NP-PLANT diets. No differences were detected between the NP treatments.

FIGURE 1.

Hunger (A) and fullness (B) responses throughout the day 6 clinical testing day after the study treatments in healthy, overweight women. Values are means ± SEMs, n = 17. The solid diamonds on the x axes represent eating occasions. Bars without a common letter differ, P < 0.025. HP-BEEF, higher-protein beef diet (60% of protein from beef and 40% from plant sources); NP-BEEF, normal-protein diet (60% of protein from beef and 40% from plant sources); NP-PLANT, normal-protein plant diet (100% of protein from plant sources).

FIGURE 2.

Desire to eat (A) and prospective food consumption (B) throughout the clinical testing day 6 after the study treatments in 17 healthy, overweight women. Values are means ± SEMs, n = 17. The solid diamonds on the x axes represent eating occasions. Bars without a common letter differ, P < 0.025. HP-BEEF, higher-protein beef diet (60% of protein from beef and 40% from plant sources); NP-BEEF, normal-protein diet (60% of protein from beef and 40% from plant sources); NP-PLANT, normal-protein plant diet (100% of protein from plant sources).

Food cravings (day 6)

Daily food cravings for high fat foods, sweets, carbohydrates, fast food, and total food cravings are presented in Table 3. No differences in daily cravings for high fat foods, sweets, or carbohydrate foods were observed between diets (Table 3). However, the HP-BEEF diet led to daily reductions in cravings for fast food fats compared with the NP-PLANT and NP-BEEF diets (P < 0.02). Additionally, the HP-BEEF diet led to a decrease in total cravings vs. the NP-PLANT diet (P < 0.01), but not NP-BEEF diet. No differences were detected between the NP treatments.

TABLE 3.

Daily food cravings on the clinical testing day 6 after the study treatments in healthy, overweight women¹

	NP-PLANT	NP-BEEF	HP-BEEF
Daily food cravings for
High-fat foods	1.17 ± 0.08	1.13 ± 0.07	1.04 ± 0.02
Sweets	1.60 ± 0.14	1.58 ± 0.14	1.48 ± 0.13
High-carbohydrate foods	1.19 ± 0.09	1.19 ± 0.09	1.10 ± 0.04
Fast-food fats	1.56 ± 0.18a	1.40 ± 0.13a	1.15 ± 0.05b
Total cravings	1.36 ± 0.10a	1.31 ± 0.09a,b	1.20 ± 0.05b

¹Values are mean ± SEM arbitrary units, n = 17. Labeled means in a row without a common superscript letter differ, P < 0.0125. HP-BEEF, higher-protein beef diet (60% of protein from beef and 40% from plant sources); NP-BEEF, normal-protein diet (60% of protein from beef and 40% from plant sources); NP-PLANT, normal-protein plant diet (100% of protein from plant sources).

Ad libitum intake (day 7)

By design, protein intake was higher at each eating occasion and across the free-living day after the HP-BEEF diet than after the NP-BEEF and NP-PLANT diets (both P < 0.01; Table 4). However, eating more protein within each eating occasion did not voluntarily reduce within-meal fat- and carbohydrate-rich food “sides” or end-of-meal desserts (data not shown). Furthermore, eating more protein did not reduce daily ad libitum energy consumed from fat- and carbohydrate-rich foods. Due to the required increased protein consumed within the HP diet and the lack of compensation within the ad libitum foods, total daily energy intake was greater after the HP-BEEF diet than after the NP-PLANT and NP-BEEF diets (both P < 0.01) (Table 4). No differences were detected between the NP treatments.

TABLE 4.

Daily food intake throughout the free-living testing day 7 in 17 healthy, overweight women¹

	NP-PLANT	NP-BEEF	HP-BEEF
Required foods consumed
Energy, kcal	347 ± 5a	355 ± 3a	895 ± 15b
Protein, g	45.3 ± 0.5a	45.2 ± 0.5a	120 ± 2.2b
Carbohydrate, g	28.4 ± 0.3a	13.3 ± 0.2b	28.3 ± 0.9a
Fat, g	5.8 ± 0.1a	12.1 ± 0.1b	30.3 ± 0.3c
Foods consumed ad libitum
Energy, kcal	2070 ± 180	2120 ± 190	2000 ± 180
Protein, g	40.6 ± 3.6	40.3 ± 4.3	38.4 ± 3.8
Carbohydrate, g	298 ± 27	300 ± 27	294 ± 27
Fat, g	72.0 ± 7.0	78.4 ± 7.3	68.7 ± 6.4
Total foods consumed (required foods + ad libitum food intake)
Energy, kcal	2430 ± 180a	2470 ± 190a	2900 ± 180b
Protein, g	85.9 ± 3.3a	85.4 ± 4.2a	158 ± 4.2b
Carbohydrate, g	326 ± 26	314 ± 27	323 ± 27
Fat, g	77.8 ± 7.0a	90.5 ± 7.2b	99.0 ± 6.4b

¹Values are means ± SEMs, n = 17. Labeled means in a row without a common superscript letter differ, P < 0.0125. HP-BEEF, higher-protein beef diet (60% of protein from beef and 40% from plant sources); NP-BEEF, normal-protein diet (60% of protein from beef and 40% from plant sources); NP-PLANT, normal-protein plant diet (100% of protein from plant sources).

From a diet perspective, the macronutrient compositions from the foods that were actually consumed on day 7 are as followed within the NP-PLANT, NP-BEEF, and HP-BEEF diets, respectively: protein (percentage of energy): 15% ± 1%, 14% ± 1%, and 23% ± 1%; carbohydrate (percentage of energy): 54% ± 1%, 50% ± 1%, and 44% ± 1%; and fat (percentage of energy): 28% ± 1%; 33% ± 1%, and 31% ± 1%. HP-BEEF led to a greater percentage of protein (P < 0.001) and a smaller percentage of carbohydrate (P < 0.001) consumed throughout the day than did NP-BEEF and NP-PLANT. No differences in the percentage of energy from fat were observed between the HP-BEEF and NP-PLANT or NP-BEEF diets. With regard to the NP-PLANT and NP-BEEF comparisons, the NP-BEEF diet had a lower percentage of energy from carbohydrate but a greater percentage of energy from fat than did the NP-PLANT diet (both P < 0.01).

Discussion

The suppression of hunger and food cravings concomitant with the promotion of satiety to reduce the drive to (over)consume are proposed underpinnings for successful weight loss observed with HP energy-restricted diets. The current study showed that an energy-restricted HP diet containing ∼30 g protein at each meal decreased hunger, desire to eat, prospective food consumption, and food cravings while increasing fullness throughout the day compared with NP diets matched or unmatched for protein type. However, despite the improvements in appetite control and satiety, increased protein consumption did not reduce free-living energy, carbohydrate, or fat intake throughout the day in overweight, healthy women exposed to highly palatable foods. Therefore, further investigation is needed to establish whether the addition of cognitive-behavioral strategies during the consumption of an HP diet effectively reduces ad libitum food intake during energy restriction.

Limited data exist with respect to the effects of increased protein consumption on free-living daily intake (49–51). Skov et al. (50) conducted a 6-mo randomized controlled trial examining 2 ad libitum fat-reduced diets containing NP or HP (12% compared with 25% energy as protein, respectively) in healthy, overweight adults. The HP ad libitum diet led to lower total energy intake compared with the NP ad libitum diet (2220 ± 100 compared with 2680 ± 120 kcal/d, respectively; P < 0.001). Weigle et al. (49) conducted a study in which normal to overweight adults completed 12 wk of following an ad libitum HP diet containing 30% of intake as protein. Daily food intake was reduced by −440 ± 60 kcal/d after the ad libitum HP diet. More recently, Blatt et al. (51) completed an acute crossover design study in which normal-weight women consumed meals ad libitum that varied in protein content (i.e., 10–30% of energy as protein). Varying the protein content of the ad libitum meals did not decrease energy intake. The current study also found that the consumption of HP meals did not reduce ad libitum food intake compared with NP versions. The discrepant findings between studies might have occurred as a result of methodologic differences.

The 3 published ad libitum feeding studies held proportions of protein, carbohydrates, and fats constant (49–51), whereas the current study required participants to consume an absolute protein amount (i.e., 30 g protein/HP meal) and permitted carbohydrate and fat intakes to vary. Thus, whereas the previous studies were only able to compare energy content differences, the current design allowed for the assessment of energy content and food choices containing primarily high-fat or high-carbohydrate “sides” and “desserts.”

Another difference between the published studies includes the duration of the diet interventions. The 2 studies that showed reductions in daily energy intake were longer-term trials of 12 wk and 6 mo, respectively (50, 51). Thus, it is possible that the shorter time frames of the study by Blatt et al. (51) and the current study were not enough to detect alterations in ad libitum feeding in response to increased dietary protein.

The improvements in daily appetite and satiety responses after each meal within the HP compared with NP diets of the current study corroborate the findings observed within the acute single-meal studies (7). Similarly, several long-term weight-loss trials also showed greater satiety after the consumption of an HP diet than after an NP diet (52–54). Thus, it is unclear as to why the measured improvements in appetite, satiety, and food cravings detected within the HP diet did not lead to reductions in ad libitum food intake.

The disconnect between the appetite and satiety responses and food intake may have occurred due to the timing of data collection. The appetite and satiety responses were collected on day 6, whereas the ad libitum food intake assessments occurred on day 7. Thus, the design does not allow for the direct comparison of these outcomes.

It is also possible, albeit relatively unexplored, that the satiety signals after protein consumption do not override other cognitive-behavioral inputs that control food intake and food choices during subchronic energy restriction. For example, although the participants were placed on energy restricted diets during days 1–6, they did not receive recommendations to adhere to (or be mindful of) their energy restriction during the ad libitum testing on day 7. Thus, when left to their own food choices, the participants, regardless of protein intake, consumed ∼1100 more kilocalories from fat and carbohydrates on day 7 compared with the prescribed diet on days 1–6. Although the appetite and satiety data support reductions in overall food intake, this did not occur. These results parallel what is commonly observed in free-living conditions when individuals consciously or subconsciously choose to override their satiety cues and deviate from a successful weight-loss diet program (i.e., through “cheat” meals, free days, and unplanned eating occasions). A potential reason for overriding the satiety cues within the current study might have included the novelty of being provided with large quantities of “free” and highly appetizing foods with the instruction to “eat as much or as little as desired.” Although we did not assess this during day 7, it is possible that our instructions seemingly granted approval to consume the study foods in excess, leading to reduced cognitive restraint, increased disinhibition, and overconsumption.

It is well established that behavioral modification, mindful eating, and goal-setting are approaches needed to adhere to dietary strategies in order to achieve long-term weight loss and the prevention of weight regain (55). For example, establishing weight-loss goals has been shown to significantly influence voluntary food intake and weight loss (50). Future studies that include behavioral counseling in conjunction with HP energy-restricted diets are warranted.

Last, in addition to protein quantity, there is continued interest surrounding the differential effects of protein type on satiety and food intake. It is proposed that amino acid composition and digestibility of different proteins may induce disparate satiety effects (33, 34). When protein quantity within a meal is above the proposed satiety threshold of 30 g protein (4), no consistent differences in appetite, satiety, and subsequent food intake have been observed between protein types (56–59).

Some data suggest that differences exist between select proteins (i.e., whey, casein, soy) when protein quantity is well below the quantity threshold (33). In the current study, 13 g of PLANT or BEEF protein was compared within the NP diets, and no differences in the study outcomes were detected. The NP-BEEF diet contained 60% beef protein and 40% from soy and gluten, whereas the NP-PLANT diet was 100% soy and gluten. The predominant (i.e., 60%) plant-based protein within both diets was soy protein. Thus, the similar amino acid profiles between beef and soy may have limited the ability to detect differences. With the increased interest in plant proteins, there is a need to systematically compare the effects of protein type on weight-management outcomes.

Limitations

The dietary treatments were only 6 d in duration. Thus, we are unable to determine the long-term effects of consuming HP energy-restricted diets on indexes of appetite control, satiety, and ad libitum food intake. Although the inclusion of both a protein-matched and protein-unmatched control (i.e., NP-PLANT and NP-BEEF) is a strength in the study design, we were unable to include an HP-PLANT diet to fully assess whether protein type influences ingestive behavior.

The novel ad libitum paradigm allowed for the assessment of carbohydrate and fat consumption within meals and at the end of meals. However, we did not provide ad libitum snacks to assess between-meal intake. Although the effects of snacking are not fully elucidated, snacking behavior has been associated with excess energy intake, weight gain, and obesity (60). Thus, a next step in exploring the role of protein on ingestive behavior includes the assessment of ad libitum consumption of foods within snacking occasions.

It is important to note that to achieve the required protein within each HP meal on day 7, a larger amount of energy per meal was required to be consumed than with the NP diets. Although we hypothesized that eating more protein would reduce ad libitum consumption of high-carbohydrate and high-fat foods, in actuality the participants consumed ad libitum the same amount of foods in addition to the required HP foods, leading to greater daily intake. To control for the varying energy provided within the required foods between the NP and HP meals, a specific amount of required carbohydrates could have been included within the NP meals. However, this would have reduced the ability to isolate the effects of increased protein consumption on ad libitum food intake.

The current study also did not include physiologic biomarkers of ingestive behavior. It is well characterized that protein-rich foods modulate appetite control and satiety by increasing the postprandial release of satiety gut-derived peptides cholecystokinin (CCK), peptide YY (PYY), and glucagon-like peptide 1 (GLP-1), which act at various areas of the brain (i.e., hypothalamus) to increase satiety and reduce food intake (19).

Last, as previously discussed, this study did not include any behavioral modifications relating to weight loss, such that the individuals were not mindful of potential energy-restriction and weight-loss goals during the day 7 free-living ad libitum test day. It is well documented that simultaneous behavioral modifications are paramount for sustaining successful dietary weight management and weight-loss strategies (55).

Conclusions

Increased dietary protein, evenly distributed throughout the day, improves daily appetite control and satiety but does not reduce daily ad libitum carbohydrate and fat intake during subchronic energy restriction in overweight women. Collectively, our findings suggest that beneficial protein-related effects on appetite control and satiety do not effectively reduce ad libitum food intake during energy restriction and highlight a potential need for additional cognitive-behavioral strategies.

Abbreviations

BEEF

60% of protein from beef and 40% from plant sources

FCI

Food Cravings Inventory

HP

higher-protein

HP-BEEF

higher-protein beef diet (60% of protein from beef and 40% from plant sources)

NP

normal-protein

NP-BEEF

normal-protein diet with matched beef protein type

NP-PLANT

normal-protein diet with unmatched protein type (100% of protein from plant sources)

PLANT

100% of protein from plant sources