Printer Friendly

Development of food groupings to guide dietary advice for people with diabetes.


Objective: To describe the development and characteristics of a food categorisation system and its application to guide advice for diabetes treatment.

Design and methods: Foods commonly consumed by 16 adults with diabetes were grouped by macronutrient content and type of fat to form a set of reference food groups for dietary advice. Means for energy and macronutrients from individual food groups were then used to construct an overall intake pattern targeting 8000 kJ and relative amounts of carbohydrate, protein and fat (saturated fatty acids (SFA) < 10%E and (polyunsaturated fatty acids) PUFA ~ 10%E). Variation in energy and macronutrients contributed by all foods partitioned into each food group was assessed by the coefficient of variation of data on the whole diet.

Results: To differentiate between sources of fat, 13 food groups emerged and 10 were deemed acceptable to nutritional guidelines for diabetes treatment. The food group pattern was judged adequate for the achievement of dietary recommendations with low-potential variation in total energy (5%) and macronutrient proportions (protein 6%, fat 6%, carbohydrate 3%), but higher for fat types (SFA 22%, (monounsaturated fatty acids) MUFA 11%, PUFA 12%). Targeted proportions for fat types were achieved only when daily servings of PUFA-rich, oils, nuts and oily fish or soy were included in an ideal intake pattern.

Conclusions: In theory, a dietary pattern constructed from food group sources of macronutrients and individual fat types results in low-potential variation from recommended nutrient targets and, therefore, is appropriate to guide advice for the treatment of diabetes.

Key words: diabetes, dietary advice, food group, type of fat.


Diet is often quoted as the cornerstone of treatment for diabetes, and there is considerable evidence that advice targeting specific dietary change offers substantial benefit. (1-5) Current guidelines for the treatment of diabetes and related complications provide nutrient intake recommendations, the most specific of which target the proportions of different types of fat in the diet. (6) Where dietary advice necessarily refers to foods, the system of advice generation, however, needs to assure that nutrient targets can be met.

A number of food guidance systems are available with varying purposes. For example, to support glycaemic control, the carbohydrate counting system' focuses on the distribution of carbohydrate and the glycaemic index (GI) (8) focuses on the type of carbohydrate throughout meals. General food guidance systems, such as the Australian Guide to Healthy Eating (AGHE) (9) and the Food Guide Pyramid, (10) outline the number of servings from core food groups required to meet nutritional requirements for the general population. Exchange lists published by the American Dietetic Association (ADA) take a total diet approach, (11) where 'all foods can fit into a healthful eating style', (12) and provide some information on the type of fat contained in foods. (13) Although certain food groups are generally recognised for contributing a particular type of fat, for example meat and dairy for providing saturated fat, none of the current guidance systems provide adequate reference to food sources high in monounsaturated or polyunsaturated fats. Hence, general low-fat advice strategies based on these systems do not necessarily address specific relative amounts of different types of dietary fat. (14) Although it is acknowledged that individualising advice is fundamental to the treatment of diabetes, a structured food-based advice strategy to guide the achievement of appropriate targets for each type of fat is also required and must satisfy both recommended and practical evaluations in order to determine whether the resultant intake pattern does indeed achieve the nutrients targeted.

The aim of the present paper, therefore, is to describe the development and characteristics of a food categorisation system resulting in a set of food groups inclusive of the type of fat (vegetables, cereals, fruits, milk/soymilk, meat, oily fish/soy and (monounsaturated fatty acids) MUFA and (polyunsaturated fatty acids) PUFA fat) and its application to guide advice to meet energy and nutrient targets with minimal potential variation for the treatment of diabetes.


The process for the systematic development and evaluation of a specific food group intake pattern to guide advice for the achievement of overall energy and nutrient targets is summarised in Figure 1, and outlined in detail below.

2.1 Development of food groups as sources of macronutrients and fat types

2.1.1 Identification of foods to include in the food groups

Foods to be included in a set of food groups for the achievement of specific nutrient targets were identified from foods commonly consumed by 16 women with gestational diabetes mellitus (GDM) from Wollongong, Australia. Characteristics of the study sample have been previously described. (15) Individual foods reported in a diet history16 by each woman were analysed for energy and macronutrient content using the nutrient analysis software program FoodWorks (Version 2.1, 2000, Xyris software, Brisbane, Australia) incorporating nutrient tables for use in Australia (AUSNUT, Canberra, 2000). Mixed dishes and prepared foods were analysed using individual ingredients where possible. Using pooled data, the percentage contribution of common food groupings, such as cereal, meat and cheese, to total macronutrient consumption for carbohydrate, protein and fat was determined. The groupings were then rank-ordered under these macronutrient headings to determine major food sources of each (Table 1). Food groupings contributing to <75% of total intake for each macronutrient variable were then taken as representative of foods commonly consumed by the study sample.

2.1.2 Categorisation of the included foods based on macronutrient composition

Individual foods belonging to the food groupings determined in 2.1.1 were then categorised by the relative proportions of all macronutrients contained in a single serving. Thus, with reference to existing standards, (9,10) mean energy (kJ) and macronutrient content (g) per serving were derived. Each macronutrient was considered in turn, commencing with the main macronutrient (carbohydrate, protein or fat). A fourth category accommodated foods that did not have the levels of macro nutrients identified in the other three lists, mainly vegetables, but nevertheless needed to be included in terms of total energy and micronutrient intakes. Subcategories were developed based on secondary macronutrient compositions. For example, in the case of fat content, the need to identify the type of fat required saturated fatty acids (SFA)-rich versus PUFA-rich subcategories in the milk and meat groups, but not MUFA-rich as this type of fat largely coincided with SFA content. Fast foods, such as ham/cheese burgers and fried chicken, had similar fat profiles to that of SFA-rich high-fat meat and cheese and, therefore, were included in that group. Similarly, fat-rich foods (oils, spreads and nuts) were categorised as PUFA-rich or MUFA-rich, according to the main type of fat that they contained. This process would also have necessitated a SFA-rich subcategory within the fat-rich group except that the study sample was already limiting fat-rich food sources of this type of fat, for example butter.


To minimise variation between individual foods within each food group, the following steps were followed:

1 A reference food was identified within each main macronutrient category. For example, bread was taken as the staple food for carbohydrate.

2 Portion sizes for individual foods in each list were modified to produce a gram amount for the main macronutrient close to that of a standard serving of the staple food. Thus, portion sizes for foods listed under carbohydrate were modified to more closely match the 15 g of carbohydrate taken for one slice of bread.


3 Subcategories were based on secondary macronutrients and also on other nutritive components such as the presence of starch and sugar.

Thus, a set of reference food groups representing the mean macronutrient content of common foods was determined (Figure 2).

2.1.3 Estimation of food group variation

For each food group, the mean nutrient content, range and coefficient of variation (CV = SD/mean x 100) were determined for all foods in that group. Acceptable variation was set at a CV [less than or equal to] 15% for the main macronutrients, otherwise acceptability of the variation was assessed by comparison with an existing food guidance system. Thus, the SD and range results were compared with those reported in the 1995 ADA Exchange Lists for Meal Planning, the only other set of exchange lists to provide specific data on within-list variations from mean nutrient estimates. (17)

2.1.4 Comparison with core food groups

In order to establish whether general nutritional adequacy and diabetes principles might be achieved from an intake pattern based on the reference food groups, they were compared with core food guide classifications outlined by AGHE and Food Guide Pyramid (9,10) and current practices aimed at controlling glycaemia. Additionally, food groups that did not meet nutrient density criteria consistent with ADA guidelines, that is, foods with high (saturated) fat or sugar (sucrose) content, were identified as 'foods to limit' and excluded from the final set of food groups for which intake levels were set.

2.2 Determination of a recommended food group intake pattern

2.2.1 Construction of a food group intake pattern to meet nutrient targets

Mean amounts for the energy and macronutrient content from each of the food groups were rounded-off to the nearest whole number, with the following exceptions: the protein content of one sewing of milk was rounded down to 10 g instead of up to 11 g because of a greater number of individual food items containing nearer that amount; the fat content of medium-fat meat was rounded down to 5 g rather than up to 6 g in order to be consistent with the sum of available data from individual fat types; finally, the small carbohydrate content of soybeans and nuts was ignored to enable a better representation of more common foods listed in these groups, such as fish, and oils and spreads, respectively. In this way, a simplified nutrient composition table in a 'ready reckoner' format was produced.

A total diet pattern was constructed from a defined number of servings from selected reference food groups using corresponding ready reckoner estimates to produce specific targets for total energy and macronutrient intakes. Based on the average requirement for moderately active, weight-stable adults, the target for energy was approximately 8000 kJ/day. (18,19) Core food recommendations were adhered to by aligning the number of servings from the reference food groups with minimum daily requirements from core food guide classifications. (9,10) Nutrition principles for diabetes management were also considered in order to enable advice on portion control, carbohydrate-counting and identification of the type of carbohydrate (wholegrain, high fibre, low GI). (6) Macronutrient goals corresponded to low fat intakes and current diabetes recommendations: approximately 30% energy as total fat (<10% saturated (SFA) and ~10% total polyunsaturated (PUFA) fat) and approximately 20% protein, leaving 50% energy as carbohydrate. (6) Atwater factors were used to convert percentage of energy to grams of macronutrients. (20) Steps in the determination of the recommended food group pattern are outlined below:

1 Foods from the carbohydrate group were included in the diet first, where 50% of energy translated to approximately 250 g carbohydrate, or 15 exchanges of foods delivering 15 g carbohydrate per exchange (allowing for some carbohydrate from vegetables). Ready reckoner estimates for the protein and fat content of these foods were also calculated at this stage.

2 Protein targets were next addressed by referring to the recommended amount from cooked meat or meat equivalent, (9,10) with an average amount from all sources achieving, 5o z (150 g). Protein means were added to those obtained from carbohydrate-rich foods, established in (1).

3 Targets for different types of fat were lastly addressed, mainly through the inclusion of foods that deliver PUFA, because SFA from carbohydrate and protein foods was already found to be adequate from the inclusion of lean- and low-fat groups in steps (1) and (2).

Total kilojoule and grams of macronutrients from the overall intake pattern were calculated and compared with targets.

2.2.2 Matching dietary advice to the recommended food group intake pattern

Using this framework, dietary advice would refer to a specific number of servings per day from each of the reference food groups defining the recommended food pattern in 2.2.1. The advice would be appropriate for people with diabetes, allowing individualised food choices and the type and distribution of carbohydrate-rich foods and meals throughout the day.

2.2.3 Nutritional adequacy of the food group intake pattern

Overall, nutritional adequacy was assessed qualitatively by comparing the number of servings from across the food group intake pattern with the number of servings required to meet core food recommendations for adults. (9,10) In addition, a random selection of food items in the recommended pattern and the amounts specified was compared with 75% of the recommended intakes for protein and vitamins A, C, E and folate, and minerals iron, calcium, magnesium, zinc and fibre. (18,19)

Selection of a 'random' food item was achieved when an individual, blinded to the food database, chose a number between zero and the number of foods contained in the corresponding food list, inclusively. When the foods from that list were arranged in reverse order, the food corresponding to the number chosen became the representative food. This process was repeated independently for each serving within each food group.

2.2.4 Comparison with unrounded estimates of dietary intake

The effect of using rounded estimates for calculations of total dietary intakes was assessed. Based on the food group intake pattern developed for 8000 kJ, total estimates for energy, carbohydrate, protein, fat and each fat type were calculated from 'rounded' ready reckoner estimates and from the original unrounded means, and the differences compared.

2.2.5 Variation due to individual food choices

The variation in intakes for total energy and macronutrients that might be expected from food choices within the restraints of the food group intake pattern was also assessed. This was achieved by taking the sum of variances from each food group for carbohydrate, protein, fat and each fat type, respectively The square root of the sum of variances provided a total SD for each macronutrient variable and, in turn, a SD for total energy intake. In this way, energy and macronutrient distributions from all possible food combinations to meet the prescribed pattern of intake were determined.

The steps undertaken to achieve the assessments of the effects of rounding and variation due to individual food choices in 2.2.4 and 2.2.5, respectively, are outlined below:

1 A total mean estimate in grams for each macronutrient and fatty acid in the overall food pattern was determined from the sum of the original means for each of the included food groups.

2 The SD of each estimate determined in (1) was calculated using the formula:

S[D.sub.(total grams)] = v([x.sub.1]S + [x.sub.2]S + [x.sub.3]S + ... + [x.sub.n]S)

Where, SD = standard deviation,

[x.sub.1] = number of serves from food group 1,

S = variance for the total mean (grams) from food group 1.

3 The total grams [+ or -] SD determined in (1) and (2), respectively, for each macronutrient variable, were converted to energy (kJ) using Atwater factors outlined in 2.2.1.

4 The resultant energy [+ or -] SD for total carbohydrate, protein and fat determined in (3) were summed to give the total energy and SD for the overall diet.

The CV for each macronutrient variable and total energy were calculated from mean total energy [+ or -] SD determined in (4). CV < 15% was the arbitrary estimate of reasonable variation for intakes of total energy and each macronutrient and fat type from the food group intake pattern.


3.1 Development of food groups as sources of macronutrients and fat types

3.1.1 Identification of foods to be included in the food groups

Major food groups commonly consumed by the study sample were rank-ordered to reveal 24% of protein intake came from meat, 42% of carbohydrate came from cereal-based foods and 18% of fat from oils and margarines. Cereal-based foods also made major contributions to the protein (21%) and fat (11%) fractions of the diet, whereas milk and yoghurt made secondary contributions to both protein (15%) and fat (11%). Combining milk, yoghurt, meat and cheese provided most of the fat intake--predominantly saturated fat.

3.1.2 Categorisation of the included foods based on macronutrient composition

In total, 348 common food items composing 75% of the total macronutrient content of the diets of the study sample were categorised according to macronutrient composition to form nine subcategories and 13 final food groups (Figure 1). Corresponding mean estimates for macronutrient content of individual portion sizes within each food group are provided in Table 2.

3.1.3 Estimation of food group variation

For each of the 13 food groups, within-list variation was low: CV < 15% for primary macronutrients (data not shown); and SD < 2 g, although the number of foods listed varied from 1 to 89 (Table 2). The vegetables group was the exception to this with greater variation within very low macronutrient content. For this food group, serving size modification aimed at reducing the variation for energy (CV < 15%) as the main dietary variable rather than macronutrient content (data not shown). SD and range for individual foods within each list compared well with those reported in the literature, (17) the results of the present study demonstrating, in most cases, a narrower data set. Therefore, mean estimates for each of the reference food groups were considered representative of the energy and macronutrient content of individual food items listed within each group.

3.1.4 Comparison with core food groups

Food guides generally refer to five core-food classifications (bread/cereals, vegetables, fruit, milk, meat and equivalents) outlined in Table 3. (9,10) In addition, the ACHE refers to a broad variety of extra 'foods to limit'. In contrast, the categorisation process of the present study derived 13 food groups, 10 of which were determined to be appropriate for the nutritional management of diabetes. Hence, foods listed in the reference food groups identifying table sugar/sugar-rich foods, full-fat milk and high-fat meat/fast foods provided a more specific list of 'foods to limit'. Furthermore, differences between the food groups of the present study and the guides were: the inclusion of carbohydrate-rich starchy vegetables with cereal-based starches such as bread rather than with other vegetables with low carbohydrate content; cheese was listed as a protein-rich food with meat rather than with milk, and these groups were subcategorised to address differences in the amount and type of fat they contain; and finally, the inclusion of high-fat foods such as oils, spreads and nuts, again with subgroups for proportional differences in the type of fat. Although being more discriminating between foods, the final set of food groups of the present study were consistent with core food guide classifications in that a minimum number of servings across the food group intake pattern would ensure overall nutritional adequacy. In terms of glycaemic control, equivalent carbohydrate content per serving and the inclusion of all carbohydrate-rich foods within three food groups (starch, fruit and milk) support advice for the even distribution and type of carbohydrate in meals throughout the day.

3.2 Determination of a recommended food group intake pattern

3.2.1 Construction of a food group intake pattern to meet nutrient targets

Rounded mean estimates corresponding to each list of foods provided a 'ready reckoner' of energy and macronutrient compositions as well as fibre content for the final set of food groups (Table 4). Subsequent food group assemblage using 'rounded' ready reckoner estimates achieved an intake pattern corresponding to that defined in Table 3. Therefore, in theory the prescribed food group pattern achieved the nutrient proportions targeted (provided in brackets): 8290 kJ (8000 kJ), 67 g total fat 30%E (30%E), 15 g SFA 7%E (<10%E), 22 g PUFA 10%E (10%E), 106 g protein 22%E (20%E) and 235 g carbohydrate 46%E (50%E).

3.2.2 Matching dietary advice to the recommended food group intake pattern

Food-based advice was constructed in terms of the required number of daily servings from each of the reference food groups included in the recommended intake pattern (Table 3). In terms of glycaemic control, foods of low glycaemic effect corresponded well to existing fruit and milk groups. More discriminating advice would largely be confined to the starch group with reference to preferred food choices such as wholegrain, high fibre and/or low GI. (8) Equivalent sewing amounts from the starch, fruit and milk groups would enable even distribution of carbohydrate throughout the diet.

3.2.3 Nutritional adequacy of the food group intake pattern

The recommended intake pattern was at least equivalent to the minimum number of servings outlined in core food guides (Table 3). (9,10) A random selection of food items to represent the food groups in the recommended intake pattern provided at least 75% of intakes recommended for other nutrients outlined in the Methods section (data not shown). (18,21)

3.2.4 Comparison with unrounded estimates of dietary intake

Calculation of the same 8000 kJ food pattern using original mean estimates from Table 2 for each of the included food groups resulted in small differences in total energy and macronutrient intakes compared with those achieved using ready reckoner estimates from Table 4 and presented in 2.2.1 (and in brackets here): 7980.9 [+ or -] 365.5 kJ (8290 kJ), 65.0 [+ or -] 4.1 g (67 g) total fat, 13.4 [+ or -] 2.9 g (15 g) SFA, 21.6 [+ or -] 2.6 g (22 g) PUFA, 102.5 [+ or -] 5.8 g (106 g) protein and 239.6 [+ or -] 7.2 g (235 g) carbohydrate. The effect of rounding, therefore, was considered minimal and justified the use of simplified estimates for ease in the development of a prescribed dietary intake.

3.2.5 Variation due to individual food choices

Estimations of variation from all possible food combinations suggest that individual choices to match the model would be reasonably close to dietary targets, particularly for total energy intake (7980.9 [+ or -] 365.5 kJ). The CV for each macronutrient variable (carbohydrate = 3.0%, protein = 5.7%, fat = 6.3%, MUFA = 10.7%, PUFA = 12.0% and for total energy = 5%) was <15% set as clinically acceptable. Greater variation was determined for SFA (CV = 21.7%). The estimated variation in total energy was relatively low (SD [+ or -] 365.5 kJ, CV = 5%).


There are many contexts in which dietary advice needs to be formulated. In the clinical context, as in nutrition research, methods need to be clearly defined so there is some assurance of nutritional goals. Food groups based on exchange lists of foods support this process by enabling selections from a range of foods to meet both energy and nutrient requirements. Published exchange lists, however, have demonstrated some large within-list variations (17), and may not address all the requirements for macronutrient manipulation. For the purposes of achieving low-fat, energy-controlled diets and recommended proportions of each type of fat in the diet, the present study tests an advice strategy based on the development of a set of macronutrient-based reference food groups and a recommended intake pattern. The advice was subsequently shown to adequately address targets for the type of fat consumed within a nutritionally adequate, energy-controlled diet. Furthermore, being a reference framework, the general principles would apply to all, but when it came to specific foods a number of food combinations could be used, allowing increased flexibility for individual food preferences and health and lifestyle objectives. The structured advice approach, however, ensures consistent and accurate targeting of nutrients regardless of these individual differences.

Although the categorisation process has inherent limitations, (21) these mainly stem from the criteria used to categorise foods. In this case, the criteria for food group development were taken from the macronutrient parameters to be used in the development of the overall diet. Thus, the resultant food groups were used as building blocks for the construction of an intake pattern to meet predetermined dietary targets, and therefore were considered appropriate. As the focus of dietary advice is on foods and concerns relative amounts of macronutrient intakes, a recommended intake pattern was developed from food categories using one criterion at a time and reference standards used to ensure total energy requirements, macronutrient proportions and overall nutritional adequacy Within this framework, the type and distribution of carbohydrate were also addressed in line with existing strategies to support glycaemic control.

Working from foods commonly consumed by a local sample of women with GDM, major food sources of macronutrients were determined. In this way, the significance of secondary sources of macronutrient intake was uncovered and the importance of attending to multiple food components underscored. This resulted in a greater final number of food groups compared with core food guide classifications outlined in existing food guides. (9,10) Corresponding food lists differentiated well between foods with substantially different relative proportions of macronutrients and the type of fat, for example, cows milk versus soymilk and meat versus oily fish, where the distinction between SFA-rich and PUFA-rich foods was apparent. In contrast, existing food guides do not differentiate well between foods in which fat is the secondary macronutrient contribution, mainly high-protein foods, such as milk, meat and nuts (in lists where nuts are included as a high-protein food even though the main macronutrient is fat). (7,9,10) ADA Exchange Lists now subcategorise exchangeable edible fat sources based on the major fat type. However, bearing in mind edible fat alone can account for just 30% of total daily fat intake, the application of ADA Exchange Lists is less likely to have an impact. (22) For example, a weight loss intervention based on the United States Department of Agriculture Food Guide Pyramid and ADA Exchange Lists showed reductions in total and SFA intakes, but with little change in the proportion of PUFA in the diet. (14) The greater number of food groups of the present study assured targets for the type of fat, meeting the challenge for achieving both nutritional adequacy and appropriate dietary fat profile. Where not all food groups were included in the sample pattern, a substantial number of servings from PUFA-rich food sources were necessary, highlighting the importance of advice for the regular consumption of some foods and not others. Although goals for the relative proportions of different types of dietary fat may also be met through judicious food choices, the present guidance systems are too blunt to make this assumption.

Specificity of the study sample, the relatively small number of foods listed and the fact that estimates were based on a single food pattern may limit the generalisability of results. However, a single dietary pattern was important to demonstrate an 'ideal' template on which to base advice for consistent dietary outcomes. This lends itself to substantiation research in which a structured approach enables a single nutrient or food to be tested within the overall diet plan. (23) Although the number of listed foods was limited, mean estimates for macronutrient content corresponded to those reported for exchange lists using a greater number of foods. (17) Furthermore, variation due to any combination of individual food choices from within the recommended food pattern was low and likely narrower than existing advice systems that demonstrate wider variation within individual food categories. (9,17) The advice system was, therefore, judged as adequate for consumption of different types of fat in the proportions defined by current recommendations, and particularly supportive of total diet advice for controlling energy intake (SD [+ or -] 365.5 kJ within an 8000 kJ diet plan). Greater variation for saturated fat (likely due to the increased number of staple foods containing this type of fat) was readily overcome in the construction process by ensuring that the upper estimate of SFA variation (+2SD) was below the target level (<10% of energy). For any system, however, encouraging consumption of a wide variety of foods from within and across food groups reduces the risk of consumption patterns that lie at the extreme ranges of energy and macronutrient intakes, as well as ensures nutritional adequacy. (15)

One of the biggest challenges in addressing macronutrient-referenced dietary goals is to provide valid and feasible advice on foods to consume. By confirming the theoretical achievement of nutritional goals, the present paper has outlined a systematic approach to dietary advice based on food groups that differentiate between foods based on primary and secondary macronutrient content. Having done so, the approach has demonstrated a methodology that can be used for other plans according to requirement. Although all foods may not fall clearly into any one group, the structured nature of the approach facilitates a level of capability for the achievement of macronutrient targets. The advice system, therefore, is appropriate not only in the diabetes context, but also for the treatment of other clinical features of the metabolic syndrome, including cardiovascular disease and overweight, as well as general healthy eating strategies, and in controlled diets for investigations of single nutrient/food effects. Acceptance of the advice system by practitioners and its feasibility in clinical practice and in diverse ethnic groups or for dietary interventions other than diabetes would be of interest and provides the basis for future research.

In practice, the approach has been successfully used in randomised controlled trials involving subjects with GDM (24) and Type 2 diabetes mellitus, (25) and was found to be easy to implement with high acceptance by both groups. This was borne out in the adequate achievement of targets for each type of fat by intervention groups receiving advice based on the recommended food group intake pattern, which was significantly different from control groups receiving general low-fat advice (usual practice). The clinical effectiveness of the advice system was also confirmed in Type 2 diabetes mellitus subjects through significant improvements in cholesterol outcomes compared with the control group. (26) Importantly, all groups maintained good glycaemic control, verifying the shift in focus away from glycaemia to a broader focus on overall metabolic health without detrimental effects.


In recognition of the need for evidence-based approaches to advice, the present study has developed a systematic approach for the dietary treatment of diabetes. This system successfully combined specific macronutrient recommendations with existing guidelines and practices to address the type and distribution of carbohydrate-rich foods as well as the type of fat within the overall diet. Importantly, the significance of differentiating between foods of varying macronutrient content was demonstrated in a theoretical dietary model. Advice would refer to daily servings from a set of reference food groups in the recommended intake pattern. Equivalent serving amounts from carbohydrate-rich foods within the starch, fruit and milk groups and differentiating wholegrain, high fibre, and/or low GI choices within the starch group would support existing practices for glycaemic control. (7,8) PUFA-rich foods, however, also required specific attention and regular intakes for the achievement of recommended proportions for each type of fat in the diet. Usual eating patterns would determine relative amounts of significant foods such as oily fish, soy foods and/or oils and nuts developed in proportion with individual energy needs. The present study, therefore, provides theoretical support for food groupings based on macronutrient content and their specific pattern of intake to ensure nutrient targets, and confirms their appropriateness to guide advice in the treatment of diabetes.


The present study was supported by the Smart Foods Centre and Australian Research Council. The work forms part of a PhD thesis by L. Gillen, University of Wollongong.


1 Diabetes Control and Complications Trial Research Group. Effect of intensive diabetes treatment on the development and progression of long-term complications in adolescents with insulin-dependent diabetes mellitus: Diabetes Control and Complications Trial. J Pediatr 1994; 125: 177-88.

2 UK Prospective Diabetes Study Group. Effect of intensive, blood glucose control with metformin on complications in overweight patients with Type 2 diabetes. Lancet 1998; 352: 854-65.

3 Diabetes Prevention Program Research Group. Reduction in the incidence of Type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 2002; 346: 393-403.

4 Tuomilehto J, Lindstrom J, Eriksson JG et al. Prevention of Type 2 diabetes mellitus by changes in lifestyle among subjects with impaired glucose tolerance. N Engl J Med 2001; 344: 1343-50.

5 Pan XR, Li GW, Hu YH et al Effects of diet and exercise in preventing NIDDM in people with impaired glucose tolerance. The Da Qing IGT and Diabetes Study. Diabetes Care 1997; 20: 537-44.

6 American Diabetes Association. Nutrition principles and recommendations in diabetes. Position statement. Diabetes Care 2004; 27 (Suppl): S36-46.

7 American Dietetic Association and American Diabetes Association. Carbohydrate Counting: Getting Started. Chicago, IL: American Dietetic Association, 1995.

8 Brand Miller J, Foster-Powell K, Colagiuri S, Leeds A. The GI Factor: The Glucose Revolution. Rydalmere: Hodder & Stoughton, 2000.

9 Department of Health and Ageing. Australian Guide to Healthy Eating. Canberra: Australian Government, 1998.

10 US Department of Agriculture. Food Guide Pyramid. A Guide to Daily Food Choices. Washington, DC: Human Nutrition Information Service, 2000.

11 American Dietetic Association and American Diabetes Association. Exchange Lists for Meal Planning. Chicago, IL: American Diabetes Association and American Dietetic Association, 2003.

12 American Diabetes Association. Position of the American Diabetes Association: total diet approach to communicating food and nutrition information. J Am Diet Assoc 2002; 102: 100-108.

13 Wheeler ML. Nutrient database for the 2003 exchange lists for meal planning. J Am Diet Assoc 2003; 103: 894-920.

14 Benezra LM, Nieman DC, Nieman CM et al. Intakes of most nutrients remain at acceptable levels during a weight management program using the food exchange system. J Am Diet Assoc 2001; 101: 554-61.

15 Gillen L, Tapsell LC, Martin GS, Daniells S, Knights S, Moses RG. The type and frequency of consumption of carbohydrate-rich foods may play a role in the clinical expression of insulin resistance during pregnancy. Nutr Diet 2002; 59: 135-43.

16 Tapsell LC, Daniells S, Martin GS, Knights S, Moses RG. Performance of a research diet history for use in clinical studies involving pregnant women with and without gestational diabetes mellitus in the Illawarra region. Nutr Diet 2002; 59: 129-34.

17 Wheeler ML, Franz M, Barrier P, Holler H, Cronmiller N, Delahanty LM. Macronutrient and energy database for the 1995 Exchange Lists for Meal Planning: a rationale for clinical practice decisions. J Am Diet Assoc 1996; 96: 1167-71.

18 Truswell AS. Development of the current set of recommended dietary intakes for Australians. In: Truswell AS, Dreosti IE, English RM, Rutishauser IHE, Palmer N, eds. Recommended Nutrient Intakes: Australian Papers. Sydney: Australian Professional Publications, 1990; 15-17.

19 Trumbo P, Schlicker S, Yates AA, Poos M. Dietary reference intakes for energy, carbohydrate, fiber, fat, fatty acids, cholesterol, protein and amino acids. J Am Diet Assoc 2002; 102: 1621-30.

20 Mahan LK, Escott-Stump S. Krause's Food, Nutrition and Diet Therapy, 10th edn. Philadelphia: WB Saunders Company, 2000.

21 Wirfalt AKE, Jeffery RW. Using cluster analysis to examine dietary patterns: nutrient intakes, gender, and weight status differ across food pattern clusters. J Am Diet Assoc 1997; 97: 272-9.

22 Nydahl MC, Smith RD, Kelly CNM, Fielding BA, Williams CM. Achievement of dietary fatty acid intakes in long-term controlled intervention studies: approach and methodology. Public Health Nutr 2002; 6: 31-40.

23 Aggett PJ, Antoine JM, Asp NG et al. PASSCLAIM. Process for the Assessment of Scientific Support for Claims on Foods. Eur J Nutr 2005; 44 (Suppl 1): i5-30.

24 Gillen LJ, Tapsell LC. Advice that includes food sources of unsaturated fat supports future risk management of gestational diabetes mellitus. J Am Diet Assoc 2004; 104: 1863-7.

25 Gillen LJ, Tapsell LC, Patch CS, Owen A, Batterham M. Structured dietary advice incorporating walnuts achieves optimal fat and energy balance in patients with type 2 diabetes mellitus. J Am Diet Assoc 2005; 105: 1087-96.

26 Tapsell LC, Gillen LJ, Patch CS et al. Including walnuts in a low fat/modified fat diet improves HDL cholesterol-to-total cholesterol ratios in patients with type 2 diabetes. Diabetes Care 2004; 27: 2777-83.

Lynda J. CILLEN and Linda C. TAPSELL

Smart Foods Centre and Department of Biomedical Science, University of Wollongong, Wollongong, New South Wales, Australia

L.J. Gillen, B Nutr Diet (Hons), FDAA, PhD candidate, Research Fellow

L.C. Tapsell, PhD, MHPEd, FDAA, Professor

Correspondence: L. Gillen, Smart Foods Centre, University of Wollongong, Northfields Avenue, Wollongong, NSW 2522, Australia.

Table 1 Rank order of the highest percentage (%) contribution of common
food groupings to approximate 75% of the macronutrient intake of 16
women with gestational diabetes mellitus

 Protein Carbohydrate
Food groupings % total intake Food groupings % total intake

Meat 23.8 Cereals/bread 42.1
Cereals/bread 21.1 Fruit 16.9
Milk/yoghurt 14.9 Sugar-rich 12.7
Fish 7.4
Vegetables/legumes 6.6
Total intake 73.8 71.7

Food groupings Fat % total intake

Oil/margarine 18.0
Meat 11.7
Cereals/bread 11.3
Milk/yoghurt 11.1
Cheese 9.9
Fast food 7.8
Nuts 4.5
Total intake 74.3

Table 2 Mean, SD and range for the macronutrient content of recommended
serve sizes from all foods of each food group

Macronutrient food Carbohydrate (g) Protein (g)
list No. Mean SD (a) Range Mean

Vegetables 33 2.3 [+ or -]1.1 0-4 (b) 1.9
0.5 cup cooked/1
 cup raw
Starchy foods 89 14.9 [+ or -]1.8 11-19 (c) 3.1
1 slice/0.5 cup
Fruit 38 15.2 [+ or -]1.4 13-18 (d) 1.2
1 piece
Table sugar 1 5.0 [+ or -]0.0 0-0 0.0
1 teaspoon/5 g
Milk (low/reduced 12 14.6 [+ or -]1.6 12-17 (e) 10.8
1 cup
Milk (full fat) 5 15.2 [+ or -]0.8 14-16 11.8
1 cup
Soymilk (full fat) 1 15.0 [+ or -]0.0 15-15 11.0
1.25 cup
Meat(lean/low fat) 68 0.1 [+ or -]0.4 0-2 7.2
30 g/l oz
Meat (medium fat) 30 0.4 [+ or -]0.8 0-3 7.2
30 g/l oz
Meat (high fat) 24 0.5 [+ or -]1.0 0-3 7.0
30 g/l oz
Oily fish, soybeans 6 0.5 [+ or -]0.8 0-2 7.2
30 g/l oz
MUFA-rich 24 0.5 [+ or -]0.6 0-2 0.4
1 teaspoon/5 g
PUFA-rich 17 0.2 [+ or -]0.5 0-2 0.2
1 teaspoon/5 g

Macronutrient food Protein (g) Fat (g)
list SD (a) Range Mean SD (a) Range

Vegetables [+ or -]0.9 1-4 0.0 [+ or -]0.0 0-0
0.5 cup cooked/1
 cup raw
Starchy foods [+ or -]1.5 1-10 0.8 [+ or -]0.7 0-2
1 slice/0.5 cup
Fruit [+ or -]0.8 0-3 0.1 [+ or -]0.3 0-1
1 piece
Table sugar [+ or -]0.0 0-0 0.0 [+ or -]0.0 0-0
1 teaspoon/5 g
Milk (low/reduced [+ or -]0.9 10-12 1.9 [+ or -]1.9 0-4
1 cup
Milk (full fat) [+ or -]1.9 10-15 11.4 [+ or -]0.6 11-12
1 cup
Soymilk (full fat) [+ or -]0.0 11-11 12.0 [+ or -]0.0 12-12
1.25 cup
Meat(lean/low fat) [+ or -]0.7 6-8 (f) 1.8 [+ or -]1.1 0-4
30 g/l oz
Meat (medium fat) [+ or -]0.8 6-9 (g) 5.7 [+ or -]0.8 5-7
30 g/l oz
Meat (high fat) [+ or -]0.8 6-8 9.3 [+ or -]1.3 8-12
30 g/l oz
Oily fish, soybeans [+ or -]0.4 7-8 3.5 [+ or -]0.6 3-4
30 g/l oz
MUFA-rich [+ or -]0.9 0-2 5.2 [+ or -]0.5 5-6 (h)
1 teaspoon/5 g
PUFA-rich [+ or -]0.4 0-1 5.1 [+ or -]0.3 5-6 (i)
1 teaspoon/5 g

American Diabetes Association published values: (17) (a) 0-4.7; (b)
1-14; (c) 11-23; (d) 11-20; (e) 11-17; (f) 5-9; (g) 5-11; (h) 3-5; (i)

Table 3 Comparison of existing food guidance systems and macronutrient-
based reference food groups used to construct a pattern of intake to
guide advice for diabetes treatment

 AGHE (9)
Core food guide Average USDA (10) (serves/
classification (9,10) serve size (serves/day) day)

All vegetables, legumes 0.5 cup cooked 3-5 4-8
Bread, cereals, rice, pasta, 1 slice 6-11 8-14
 noodles 0.5 cup
All fruit 1 medium 2-4 2-4
Milk, yoghurt, cheese 1 cup 2-3 2-4
Meat, fish, poultry, eggs, Cooked meat 2-3 1-1.5
 nuts, legumes/dry equivalent
 beans 2-3 oz (10)
 65-100 g (9)
 30 g (1 oz) (a)
'Extra foods' including Limited (10) Limit 0-3
 margarine, oils, sweets, 145 kcal (9)
 biscuits, snack foods 1 teaspoon fat (a)
 and alcohol 10 g nuts (a)

 Reference diet
Core food guide model (a) Macronutrient-based
classification (9,10) (serves/day) reference food groups

All vegetables, legumes 5 Vegetables
 Excludes potatoes, corn,
Bread, cereals, rice, pasta, 9 Starch
 noodles Plus potatoes, corn,
 legumes (not soybeans)
All fruit 4 Fruit
Milk, yoghurt, cheese 2 Milk (low/reduced fat)
 Excludes cheese
 Soy milk (full-fat) (b)
Meat, fish, poultry, eggs, 5 Meat (lean/low fat)
 nuts, legumes/dry Excludes oily fish, nuts,
 beans legumes
 Includes cheese (medium
 Includes eggs and cheese
 Oily fish/soy beans (b)
'Extra foods' including 3 MUFA
 margarine, oils, sweets, 5 Includes oils, spreads,
 biscuits, snack foods nuts (b)
 and alcohol Limit PUFA
 Includes oils, spreads,
 nuts (b)
 Sugar-rich foods
 Milk/yoghurt(full fat)
 Meat/fast food (high fat)

(a) Pattern of intake developed from reference food groups to provide
approximately 8000 kJ.
(b) High in PUFA.
(c) High in MUFA.
AGHE, Australian Guide to Healthy Eating; MUFA, monounsaturated fatty
acids; PUFA, polyunsaturated fatty acids; USDA, United States Department
of Agriculture.

Table 4 'Ready reckoner' format for energy and macronutrient content of
food groups

 Carbohydrate Protein
Food categories Serving size (g) (g)

Vegetable 0.5 cup 2 2

 Bread, cereals, vegetables, 1 slice/0.5 cup 15 3
Fruit 1 piece 15 1
 Low/reduced fat milk, 1 cup 15 10
 Soy milk (full-fat) 1.25 cup 15 10

 Lean/low fat meat, fish, 30 g (1 oz) 0 7
 Medium fat meat, cheese, 30 g (1 oz) 0 7
Oily fish/soybeans 30 g (1 oz) 0 7

 Oils/spreads/nuts 5 g (1 teaspoon) 0 0
 Oils/spreads/nuts 5 g (1 teaspoon) 0 0

 Fat Energy SEA MUFA PUFA Fibre
Food categories (g) kJ (kcal) (g) (g) (g) (g)

Vegetable 0 80 (20) 0 0 0 3

 Bread, cereals, vegetables, 1 335 (80) 0 0 0 1-3
Fruit 0 285 (70) 0 0 0 2
 Low/reduced fat milk, 2 500 (120) 1 1 0 0
 Soy milk (full-fat) 12 845 (200) 1 3 7 0

 Lean/low fat meat, fish, 2 195 (45) 1 1 0 0
 Medium fat meat, cheese, 5 335 (80) 3 2 0 0
Oily fish/soybeans 4 260 (60) 1 1 2 0

 Oils/spreads/nuts 5 200 (50) 1 3 1 0
 Oils/spreads/nuts 5 200 (50) 1 1 3 0

MUFA, monounsaturated fatty acids; PUFA, polyunsaturated fatty acids;
SFA, saturated fatty acids.
COPYRIGHT 2006 Dietitians Association of Australia
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2006, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Author:Tapsell, Linda C.
Publication:Nutrition & Dietetics: The Journal of the Dietitians Association of Australia
Geographic Code:8AUST
Date:Mar 1, 2006
Previous Article:Household factors associated with older Australian's purchasing a varied diet: results from household expenditure data.
Next Article:Fay McDonald (nee Fenner) 6 July 1925-1 June 2005.

Related Articles
New view of fatty foods in diabetics' diets.
Vitamin A may aid some diabetics.
The type and frequency of consumption of carbohydrate-rich foods may play a role in the clinical expression of insulin resistance during pregnancy....
Nutrition labelling--the DAA perspective. (Conference Paper).
The Jewish Diabetes Cookbook. (Book Reviews).
Eating and pumping: evaluating the nutrition service of the insulin pump clinic at the Royal Melbourne Hospital. (Insight: Original Research).
Of interest in the journals.
Household factors associated with older Australian's purchasing a varied diet: results from household expenditure data.
New nutrient reference values for Australia and New Zealand: implementation issues for nutrition professionals.
Foods and nutrients provide important insights into optimal eating patterns.

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters