Obesity and type 2 diabetes mellitus


  • This section reviews the scientific literature linking diet to chronic lifestyle-related disease. Noakes and colleagues provide an overview of research on dietary protein and effects on body weight and diabetes management. Truswell follows up with a review of the links between food components and cardiovascular disease, with special reference to long-chain omega-3 polyunsaturated fatty acids, and Hodgson reviews research on food components, notably protein and carbohydrate, on blood pressure. As a food source of protein and fatty acids, red meat consumption is considered within these contexts. Baghurst summarises the literature on relevant studies on red meat consumption and the risk of colorectal cancer, and finally Chapman reviews dietary recommendation, for people living with cancer


  • • Dietary protein is more satiating than carbohydrate or fat and has been shown to reduce food intake after controlled liquid preloads and meals.
  • • High-protein, low-energy weight-loss diets may assist compliance by increasing satiety up to three hours after a meal and providing a lower dietary variety, which has been shown to be associated with lower food intake.
  • • High-protein diets in ad libitum studies show greater weight loss than high-carbohydrate ad libitum diets.
  • • Isocaloric high-protein and high-carbohydrate diets in energy restriction achieve similar weight loss, but diets with a higher protein–carbohydrate ratio achieve greater loss of fat to lean tissue.
  • • High-protein low-carbohydrate diets lower triglycerides and glycosylated haemoglobin (HbA1c) more than high-carbohydrate diets.
  • • High-protein, low-energy weight-loss diets are more nutrient dense than normal-protein, high-carbohydrate weight-loss diets, which may not meet the recommended dietary intake, particularly for calcium, but also iron and zinc for some groups.
  • • Current recommendations for protein intakes may be lower than optimal for weight management to optimise satiety, body composition and micronutrient nutrition.


To what extent does the composition of the food we eat influence how much of it we eat on an occasion, or to what extent it satisfies us sufficiently to delay the next eating occasion? These are two separate attributes—the former being termed ‘satiation’ and the latter ‘satiety’. Much of the research that has been conducted on food composition has generally focused on ‘satiety’—subjectively defined as the feeling of fullness or satisfaction that follows eating. It is generally measured by questionnaire using a visual analogue scale after a food has been consumed. In addition, some studies combine this approach with exposure of the participants to a buffet meal and measure food consumed, which represents a more objective measure of satiety. Satiety appears to be influenced by a wide variety of factors, including macronutrient profile, palatability, food mass, energy density, fibre and glycaemic index (GI). When using real foods, it is almost impossible to control for all of these influences at the same time, and if these factors are controlled, the relevance to real-life foods and diets can be questionable. In addition, the context in which the food is eaten can have a considerable effect on how much food is consumed, which may override perceived satiety. Therefore, an assortment of methodologies is important to understand how different food and diet attributes affect satiety but also food intake and, ultimately, energy balance. This paper reviews studies undertaken to demonstrate the effect of high-protein meals and diets on satiety, weight loss and diabetes management.


Several studies have compared satiety after high-protein or high-carbohydrate or high-fat meals. Typically, these studies compare satiety after different test meals in the same individual in a crossover design. In some studies, the amount of food consumed at a buffet, usually three hours after the test meal, is also assessed. In a recent review of such studies, high-protein meals were more satisfying, with 11 of the 14 studies that compared high protein with at least one other macronutrient finding the protein preload significantly increased subjective ratings of satiety.1 Few of these studies were able to control for potentially confounding variables. However, the test meals differed widely in physical andsensory properties, so it cannot be concluded that it was the protein conferring these effects. Latner designed a study so that the sensory properties of the meals were exactly the same.2 In 12 lean female students, 31% more calories were eaten at a subsequent dinner after a high-carbohydrate liquid lunch (450 kcal, 99% carbohydrate from polycose) than high-protein liquid (71% protein) meal or a 50%-protein, 50%-carbohydrate lunch. The protein was a dried powder mix derived from whey.

When protein is provided as a 50-g dose in the form of a beverage and compared with an isocaloric, isovolumetric and palatability-matched carbohydrate beverage, protein has also been shown to be more satiating than glucose. Bowen et al. compared liquid preloads (1.1 MJ, 450 mL) containing 50-g whey, soy, gluten or glucose.3 Energy intake at the buffet three hours after the preload was 10% lower for all protein preloads compared with the glucose treatment (< 0.05). Different protein sources behaved similarly. The present study also demonstrated that the effect of protein on satiety appears independent of body mass index (BMI) status, which is an important finding as almost all previous studies had been conducted in lean individuals.


Fryer et al.4 found that feelings of hunger were lowest on a high-protein diet for 12 male students with nine-week dietary periods. Several additional weight-loss studies designed to examine the metabolic effects of high-protein energy-restricted diets compared with high-carbohydrate or high-fat energy-matched structured diets have not shown differences in kilojoule intake and weight loss despite expected satiety differences.5–8 Such studies do not allow the effects of increased satiety attributable to protein to be expressed, as the dietary protocols have required all foods to be consumed. Skov et al.,9 comparing an ad libitum high-protein diet with a high-carbohydrate diet, found that enhanced satiety was the most important factor in the weight loss. In these studies protein was substituted for carbohydrate, so it may have been the reduced carbohydrate, rather than the increased protein, that was important. However, controlled studies comparing single macronutrients would suggest that the high-protein component is an important factor.3 Luscombe-Marsh et al.10 studied insulin-resistant subjects and showed that if carbohydrate is held constant, a diet with 34% energy as protein was more satiating than a high-fat, 18%-protein diet. As the study diets were controlled to be isocaloric, there were no expected differences in energy balance, suggesting that the satiety effects was not sufficiently strong to override food intake. Weigle et al.11 measured hunger and fullness in subjects with mean BMI of 26 kg/m2 on a high-protein (30% energy) and lower-protein (15% energy) diet with carbohydrate constant at 50% energy for two weeks each. Satiety was increased with the isocaloric high-protein diet, which, when allowed to be consumed ad libitum for 12 weeks, was associated with an energy reduction of 441 ± 63 kcal/day. Body weight decreased by 4.9 ± 0.5 kg, and fat mass decreased by 3.7 ± 0.4 kg. The authors suggested that the ‘anorexic’ effect of protein may relate to the superior weight loss noted on low-carbohydrate diets. However, whether the normal-protein diet eaten ad libitum would have resulted in similar changes was not tested in the present study design.

McMillan-Price et al.12 compared two high-carbohydrate and two high-protein diets consumed ad libitum for 12 weeks with high and low GI comparisons. While all groups lost a similar mean ± SE percentage of weight (diet 1, −4.2% ± 0.6%; diet 2, −5.5% ± 0.5%; diet 3, −6.2% ± 0.4%; and diet 4, −4.8% ± 0.7%; P = 0.09), the proportion of subjects in each group who lost 5% or more of body weight varied significantly by diet (diet 1, 31%; diet 2, 56%; diet 3, 66%; and diet 4, 33%; P = 0.01), suggesting that protein alone in unstructured ad libitum diets may not have clear-cut benefits and that other dietary components may be important. On the other hand, high-protein diets, where carbohydrate is more severely restricted as in the Atkins diet, have been shown to be more effective in weight loss after one year.13 This observation may be due to the protein–carbohydrate ratio or a consequence of the restricted range of foods allowed on such a pattern, as well as the simplicity of the approach.

Studies using structured meal plans to achieve isocaloric diets5,6,8,14,15 appear to achieve an almost twofold greater weight loss in the short and longer term compared with ad libitum approaches.9,12 However, this has not formally been tested in a randomised controlled trial. Studies of those who report long-term success in weight loss show that following a consistent eating pattern is a common characteristic.16


Despite no expected weight-loss differences, when high-protein diets are compared isocalorically with high-carbohydrate diets, changes in body composition have been noted. Controlled studies by Piatti et al.17 and Baba et al .18 found favourable effects of an isocaloric high-protein relative to high-carbohydrate energy-restricted diet on resting energy expenditure, lean body mass and insulin sensitivity. Both studies were small but well controlled, involving a total of 13–25 participants, and the intervention was of short duration of 3–4 weeks.

Several longer-term studies have noted improvements in body composition despite similar weight losses over a 12-week period.5,6,8,14,15 An interaction between protein and exercise was noted by Layman et al.15 in a study involving 48 women over four months. The investigators found that the high-protein diet (1.6 g protein/kg/day approximating 30% energy) and the high-protein diet plus exercise group lost more fat mass than the corresponding high-carbohydrate (0.8 g protein/kg/day) groups. Muscle mass was preserved by exercise in both the high-protein and high-carbohydrate diets.

A meta-regression by Krieger et al .19 observed that protein intakes of >1.05 g/kg body weight were associated with 0.60-kg additional fat-free mass retention compared with diets with protein intakes ≤1.05 g/kg. In studies conducted for >12 weeks, this difference increased to 1.21 kg in lean mass retention favouring high-protein over high-carbohydrate energy-restricted diets.


Weight loss using an ad libitum higher-protein compared with a high-carbohydrate diet has been assessed after six months9 but also after one year.20 The 50 participants selected foods that were designated either high protein or high carbohydrate from a special research supermarket. Those people allocated to the high-protein foods felt less hungry and lost more weight than those allocated to the high-carbohydrate diet, with a difference in weight loss between groups of 3.8 kg (and fat of 3.3 kg) at six months. More subjects lost >10 kg in the protein group (35%) than in the carbohydrate group (9%). After one year, 17% of participants in the high-protein group lost >10 kg, but 0% achieved this on high carbohydrate (< 0.09). After two years, both groups tended to maintain their 12-month weight loss, but more than 50% were lost to follow up.20

McAuley et al.21 enrolled 96 women who had normal fasting glucose levels but were insulin-resistant (BMI > 27 kg/m2), and randomised them to one of three dietary interventions: either a high-carbohydrate, high-fibre (HC) diet, a high-protein high-fat (HF) Atkins Diet, or a high-protein Zone Diet. No guidance was given in relation to energy intake. At 12 months, 76 of the original 96 participants were seen again. There were no differences between the groups in weight, fat or muscle mass loss. More women in the high-protein and HF groups lost more than 10% of their initial body weight at 12 months, compared with the HC group (36% and 25% vs 4%; P < 0.03)

A third long-term ad libitum study by Gardner et al.13 compared four different diets, including the Atkins high-protein high-fat diet and the Zone high-protein diet, in 311 free-living, overweight/obese (BMI 27–40 kg/m2), non-diabetic, premenopausal women over 12 months. Weight loss was greater for women in the high-protein high-fat group compared with the other diet groups at 12 months, and average 12-month weight was significantly different between the Atkins and Zone diets. Mean 12-month weight loss was −4.7 kg in those with Atkins diets and −1.6 kg in those with Zone diets. Part of the reason for the relative failure of the Zone diet may have been its complexity in ensuring the correct macronutrient proportion at each meal.

A similar study of 160 men and women22 comparing four commercial diets (Atkins, Ornish, Weight Watchers and Zone) showed no differences at 12 months between diets. There was a high dropout rate from all diets of about 50% overall, with a greater dropout from the Atkins diet.

Keogh et al.23 followed up participants 12 months after a study involving 12-week structured diets, which were either high protein or high monounsaturated fat. Overall, weight loss was 6.2 kg (SD 7.3; < 0.01 for time with no significant diet effect) In a multivariate regression model, predictors of weight loss at the end of the study were reported: percentage energy from protein, gender and age (R2 = 0.22, P < 0.05). Brinkworth et al.24,25 followed up participants randomised to a 12-week program on a high-protein or high-carbohydrate dietary pattern after one year. Although there was a net weight loss, there were no significant differences between groups. It was also noted that without ongoing support, dietary compliance did not persist.


Parker et al.8 showed that women with type 2 diabetes lost significantly more total fat (5.3 vs 2.8 kg) and abdominal fat (1.3 vs 0.7 kg) on the high-protein compared with high-carbohydrate diet. However, men in the present study showed no difference in fat loss between diets (3.9 vs 5.1 kg). Farnsworth et al.6 showed that in hyperinsulinemic women, total lean mass was significantly better preserved with the high-protein (−0.1 ± 0.3 kg) than with the high-carbohydrate diet (−1.5 ± 0.3 kg). The fall in resting energy expenditure was not blunted in either study with a high-protein diet.7,26

To assess whether the metabolic effects observed were related to the presence of protein or the absence of carbohydrate, Luscombe-Marsh et al.10 compared two moderately low-carbohydrate diets high in either protein or monounsaturated fat in hyperinsulinaemic individuals. Equivalent fat and lean loss was observed at three months, suggesting that carbohydrate restriction per se may play a role in the benefits noted above.

A higher-protein lower-carbohydrate diet may also be beneficial for people with features of the metabolic syndrome.5 In a group of overweight and obese women (n = 100), those with high triglycerides (>1.5 mmol/L) lost more fat mass on the high-protein than with the high-carbohydrate diet (6.4 and 3.4 kg, respectively; P = 0.035). When the results of three studies in overweight non-diabetic subjects were combined (n = 215), subjects with triglycerides >1.7 mmol/L lost more total fat (high-protein diet 6.17 ± 0.50 kg compared with high-carbohydrate diet 4.52 ± 0.52 kg) and abdominal fat ( diet 1.92 ± 0.17 kg compared with high-carbohydrate diet 1.23 ± 0.19 kg) when on a high-protein diet.

Gannon and Nuttall have comprehensively assessed high-protein lower-carbohydrate diets in energy balance in type 2 diabetes with substantial improvements in HbA1c and diurnal glucose profiles without weight loss.27–31 These studies demonstrate that the lower the carbohydrate content of the high-protein diets, the lower the HbA1c with reductions suggested to be comparable to that achieved with oral hypoglycaemic medications.32



Isocaloric studies that have controlled for saturated fat generally observe similar reductions in LDL cholesterol on high-protein or high-carbohydrate diets.5,6,14

In a controlled study by Parker et al. in type 2 diabetes,8 LDL-cholesterol reduction was significantly greater on the high-protein diet (5.7%) than on the low-protein diet (2.7%) despite a similar saturated fat composition of the diets, which is inconsistent with other studies. The opposite inconsistent finding was noted by McMillan-Price et al.in an ad libitum study where LDL cholesterol increased on a high-protein high-GI diet.12

A greater triglyceride lowering is usually observed on high-protein, lower-carbohydrate dietary patterns in both controlled6,14 and ad libitum studies.9,13 This is especially so in those individuals with a high baseline triglyceride level.5,6

Blood pressure

In studies of weight-stable individuals, Hodgson et al.33 have demonstrated that an increase in protein of about 5% of energy in exchange for carbohydrate lowers blood pressure by about 5 mmHg in hypertensive people, but in most weight-loss studies, higher-protein diets have not been more effective in blood pressure reduction. Brinkworth et al.24 observed that, after one year follow up of participants with type 2 diabetes on a high-protein or high-carbohydrate weight-loss diet, net blood pressure reduction was greater in the high-protein group.


It has been suggested that high-protein diets may lower bone density or exacerbate renal dysfunction in those people with some degree of renal impairment, but there are limited epidemiological data in support of the latter.34 Fracture rates are actually reduced by high-protein diets.35 However, there is a major gap in knowledge on the long-term effects of high-protein diets (i.e. one to two years) in people with type 2 diabetes, especially in people with microalbuminuria and renal disease.


The 50th percentile of protein intakes from the National Nutrition Survey (1995) reports protein intakes in Australia at 96–115 g in men and 70–74 g in women, depending on age.36 Protein sources in the Australian diet comprise meat poultry and fish 33% and dairy 16%, with at least 25% from non-animal sources, such as cereal and cereal-based foods.36

Clearly, based on the estimated average requirement or recommended dietary intake (RDI) for protein as defined in the National Health and Medical Research Council (NHMRC) Nutrient Reference Values37 (Table 1), current intakes are in excess of reported requirements for growth and maintenance on a fat-free mass basis. They are not derived for energy-restricted states when protein needs may differ, nor do these figures consider optimum protein needs to increase lean mass and optimise fat loss.38 The recommendations also describe requirements for ‘good quality’ protein and assume energy balance. Given that 25% of protein consumed is not necessarily ‘good quality protein’, the apparent surplus protein consumed is likely somewhat lower.

Table 1.  Estimated average requirement (EAR) and recommended dietary intake (RDI) for protein for Australian adults
  1. Source: Nutrient Reference Values for Australia and New Zealand. Australian Government, Department of Health and Ageing, National Health and Medical Research Council 2005 ISBN 1864962372.

 19–30 years52 (0.68)64 (0.84)
 31–50 years52 (0.68)64 (0.84)
 51–70 years52 (0.68)64 (0.84)
 >70 years65 (0.86)81 (1.07)
 19–30 years37 (0.60)46 (0.75)
 31–50 years37 (0.60)46 (0.75)
 51–70 years37 (0.60)46 (0.75)
 >70 years46 (0.75)57 (0.94)

Dietary protein recommendations may also be expressed as a percentage of energy intake, and the acceptable macronutrient distribution range for protein is 15–25% of energy. Due to the absolute need for protein, lower energy intakes such as required for weight management, necessitate protein intakes at the higher end of the range. Fifteen per cent of energy from protein at lower kilojoule intakes does not meet the current RDI for protein for some age/gender groups. If, as has been suggested, current protein recommendations are too low for physical and metabolic health,39 then the protein RDI stated in the table may not be optimal (see Box 1).

Protein foods are also sources of several micronutrients. In omnivorous Western diets, obtaining the RDI for calcium, iron and zinc from wholefoods necessitates protein intakes in excess of current RDIs to achieve optimal nutrient intakes. For example, if one calculates protein for three serves of dairy foods needed to meet calcium needs, three slices bread, one serve cereal plus 100 g meat, fish or chicken to provide iron and zinc, this totals 68 g protein, which is in excess of the RDI for protein for most individuals. To provide adequate nutrient intakes in low-energy diets necessitates selection of foods that are naturally nutrient rich (NNR) for the kilojoules they provide. The categorisation of foods that are NNR has recently been reviewed by Drewnowski.40 Foods with more nutrients, higher nutrient concentrations and fewer kilojoules will have a higher score. Foods such as lean meat, low-fat dairy foods and vegetables tend to have a higher NNR score. For lower-kilojoule diets, the choice of foods with a higher NNR score ensures nutritional adequacy. Based on the NNR, animal protein foods provide higher scores than vegetable sources of protein such as legumes.

Box 1: Protein leverage hypothesis

Simpson and Raubinheimer41 postulated that animals (including humans) and insects have a drive to maintain a constant intake of protein and that low-protein diets lead to overconsumption of fat and carbohydrate, with high-protein diets having the reverse effect.

The ‘protein leverage’ hypothesis (PLH) is the idea that food consumption in humans, like other animals, is adjusted to maintain a target protein intake. According to the PLH, the consumption of a low-protein diet, typical of many Western countries, inevitably requires the ingestion of additional energy. Conversely, the consumption of a diet that is relatively high in protein content requires the ingestion of lower levels of energy, creating the potential for weight loss.


Protein-containing foods and dietary patterns with a higher proportion of protein than is currently recommended appear to have a number of nutritional benefits that can be advantageous in energy-restricted diets. The science to support the use of such diets is strengthening along with the concomitant metabolic benefits in reducing dietary carbohydrate. Improvements in satiety, body composition, nutrient density and metabolic outcomes need to be considered against potential risks, which at this time have not emerged. Higher-protein lower-carbohydrate dietary patterns need to be considered not only as a valid option for weight management, but as the pattern of choice for individuals with insulin resistance.