This is the old United Nations University website. Visit the new site at http://unu.edu

The detailed procedure for calculation of the nutrient content of a multi-ingredient food is shown in figure 1. The major steps are selection of the recipe, data collection for the nutrient content of the ingredients, adjustment of the content of each ingredient for effects of preparation, summation of ingredient composition, final weight (or volume) adjustment, and determination of the yield and final volumes.

This procedure for recipe calculation incorporates the following considerations which must be recognized by the user.

- The procedure assumes that data are available on each of the ingredients which are then combined and, if necessary, prepared. If a recipe calls for one or more ingredients which must be prepared from recipes (e.g., a sauce or a stock), the calculation procedure must first be applied to each of these ingredients separately, and then to the final product using the results of the initial recipes as ingredients.
- This procedure recommends nutrient adjustment of individual ingredients before their combination, with a final adjustment for weight. This is a slight distortion of what often happens when ingredients are combined and then prepared. However, the work which has produced the various factors (e.g., Agriculture Handbook No. 102 [49]) follows this simplified model, and, moreover, the variability of nutrients in foods and their responses to cooking will tend to overwhelm any inaccuracies which derive from these simplifications.
- This procedure produces ESTlMATES of
nutrient content of multi-ingredient foods.

The preferred method of determining the nutrient content is chemical analysis.

1. Select or develop appropriate recipe.

2. Collect weight and nutrient content data for each ingredient.

3. Correct ingredient nutrient levels for weight of edible
portions where appropriate.

4. Correct ingredients for effects of cooking:

either

- if data for cooked ingredients are available, use yield factors to adjust from raw to cooked weights

or

- if data for cooked ingredients are not available, use data for uncooked ingredients applying yield factors to adjust for weight changes and retention factors for nutrient losses or gains during cooking.

5. Sum weights of ingredients to get weight of
recipe.

6. Sum nutrient values of ingredients to obtain nutrient value of
recipe.

7. Adjust recipe weight and nutrient levels to reflect changes in
fat/water when whole mixture is cooked; make any additional
refuse adjustments; apply retention factors if available for
whole recipe.

8. Determine the quantity of prepared food produced by the
recipe.

9. Determine the final values per weight (e.g., per 100 9),
volume (e.g., per cup), or serving portions as desired.

Fig. 1. Guidelines for calculating the nutrient levels of recipe foods

The guidelines in figure 1 are intended to be applicable for the most complicated situation encountered, where the user has available a detailed description of ingredients, data on the nutrient values of the ingredients, and an adequate description of how those ingredients are to be combined.

While final values should be rounded to reflect the precision of the calculation, the intermediate values (e.g., the weights of the individual ingredients) should be used in the precision to which they are calculated. This is preferred to rounding values and using them in further calculations, since numerical errors due to rounding tend to accumulate, especially when the rounded values enter into multiplicative calculations as in the discussion that follows [5]

*Select Appropriate Recipe*

Review possible recipes for the food in question and select one that is appropriate (e.g., representative of the food habits of the population or subgroup of interest). The selection of the recipe should consider the source of the food (home, restaurant, industry, or institution) and whether the recipe might be altered by regional or local food practices. Ethnic or regional variations may be needed. Sources fat recipes include cookbooks, food package instructions, and suggested recipes on food packages. (Note that recipes may also be deduced from food ingredient labels (see page 51), and assistance from the food manufacturer may be obtained for more quantitative information.) In some cases, it may be necessar Garher Available to obtain the food and weigh or measure the pans (e.g., a pepperoni pizza consisting of crust, tomato sauce, cheese, and pepperoni, or a fast food hamburger consisting of meat, bun, sauce, pickle, lettuce, and tomato).

*Gather Available Data on Ingredients*

Determine the weight in grams of each ingredient, noting the recipe yield if provided. Depending on the recipe, some ingredients may be raw and others cooked. Some data bases have weight conversion factors for the household portions of foods. In addition, there are several references which may be useful for determining the gram weights of household portions of foods [1, 23, 62, 96].

*Correct Ingredients to Edible Weight*

For ingredients with inedible parts or handling losses during preparation, determine the weight of the edible portion and adjust nutrient levels to this weight. (See page 44 ff.) For example, consider a recipe that calls for one kilogram of raw lamb shoulder. Agriculture Handbook No. 102 [49] indicates that raw lamb shoulder contains, by weight, 17% bone, 27% trimmable fat, and 8% cutting waste. Thus, the edible weight of 1 kg lamb shoulder is 480 g (100% - 17% - 27% - 8% = 48%).

*Adjust Ingredients for Cooking*

If the recipe involves cooking, the ingredients must be adjusted for cooking effects. For ingredients with nutrient data available for the cooked state, use the paragraph "Cooked ingredients" below. For ingredients with data for the raw state only, use "Uncooked ingredients".

*Cooked ingredients. *Ingredients with
nutrient data for the cooked state require correction for the
effect of cooking on the weight (or volume) of the dish. Thus, if
the recipe calls for a specific weight of raw meat to be added
and cooked, and data exist for the nutrient content of cooked
meat, one must apply a yield factor to the amount of raw meat
needed in order to determine the amount of cooked meat that will
result, and then calculate the nutrient content of that amount.
For example, if data were available on cooked lamb, one would
still have to correct the above 480 g of edible portion of raw
lamb shoulder for the weight loss during cooking. Agriculture
Handbook No. 102[49] gives a 32% cooking weight loss and thus the
480 g raw meat becomes 326 g (0.68 x 480 g) of cooked meat. The
nutrient content of the lamb is then based on 326 g.

*Uncooked ingredients.* If nutrient
values are available only for raw ingredients, it is necessary to
calculate the yield (change of total weight) of the ingredients
due to preparation and cooking, the aired effect of the change of
weight on nutrient density and content, and the effect of actual
loss or gain of nutrients due to cooking. For example, 1 kg of
raw lamb shoulder, as purchased, contains 480 g of edible meat.
The thiamin content of raw lamb is 0.165 mg per 100 g edible
portion [96]; 480 g contains 0.79 mg thiamin. During cooking, the
level of retention of thiamin is 75% [97] and the weight loss
(assumed to be all water) is 32%, which produces 326 g of meat
containing O.59 mg thiamin. (These calculations predict that
cooked lamb contains 0.18 mg of thiamin per 100 g.)

*Sum Weights of Ingredients*

Sum the weights of all prepared ingredients to get the weight of the finished dish.

*Sum Individual Nutrient Values*

Sum the total nutrient contents of all prepared ingredients to obtain the total amounts of nutrients in the final dish.

*Adjust Recipe Weight and Nutrient Levels*

Adjust the recipe weight and nutrient level for the changes in water and fat that occur when the whole dish is cooked. If additional retention factors and corrections for refuse for the whole dish are available, they should also be incorporated at this point. See Agriculture Handbook No. 102 [49] and Merrill, Adams, and Fincher [50] for possible factors.

*Determine Recipe Yield*

Determine the recipe yield as the total weight or volume produced or as the number and weight or volume of servings. Yields may be determined by:

- using the yield given by the recipe,
- adding the volumes of the ingredients in prepared form (in some cases the volumes may not be additive),
- measuring container size,
- preparing the recipe and determining volumes and number of servings.

*Determine Values to Be Used in the Food
Composition Data Base*

Divide the nutrient totals by the total weight, volume, or number of servings of the dish. For example, if the recipe instructions indicate that the recipe yields four one-cup servings, divide the recipe weight and nutrient levels by four to determine the weight and nutrient levels of a one-cup serving.

Figures 2 through 6 and examples 1 through 4 display use of these guidelines in the estimation of the nutrient levels of various types of multi-ingredient foods.

Apply nutrient retention values for holding sliced strawberries overnight in refrigerator. Sum weights of ingredients and sum nutrient levels of ingredients. Divide sums by yield of recipe to determine weight per serving and nutrients per serving.

Fig. 2. Application of guidelines to strawberry shortcake

Add nutrient levels for specified quantities of eggs, milk, salt, pepper, and margarine. . Apply factor for water loss during cooking and readjust nutrients to this new weight. Apply retention factors for nutrients lost during cooking.

Fig. 3. Application of guidelines to scrambled eggs

Use yield factors to determine quantities of cooked ingredients from recipe quantities of raw ingredients. Sum ingredient weights. Sum nutrient values of cooked ingredients, water, and spices. Adjust recipe weight for water loss during cooking. Apply retention factors for nutrient losses during stewing.

Fig. 4. Application of guidelines to homemade beef and vegetable stew

Apply factor for decrease in water weight
during cooking.

Apply factor for increase in weight and also in total fat, SFA,
PUFA, cholesterol, and vitamin E from vegetable oil during
cooking.

Apply nutrient retention factors for nutrient losses during
cooking.

Adjust weight for removal of bone and separable fat.

Adjust tat, energy, SFA, PUFA, and cholesterol for removal of
separable fat.

Fig. 5. Application of guidelines to fried pork chop

Apply refuse factor for bone and skin of chicken. Apply yield factors for raw chicken and uncooked rice. Apply water-loss and nutrient changes with skin loss of chicken. Apply water-gain changes for cooking rice. Apply nutrient-retention factors for cooking rice and chicken. Sum weights and nutrient values of ingredients. Apply water-loss factor for cooking casserole. Apply nutrient-retention factors for cooking casserole.

Fig. 6. Application of guidelines to chicken-rice casserole

EXAMPLE 1: Mixed Vegetable Salad

3 tomatoes, cut in wedges

1 cucumber, cut in thin slices

1 green pepper, cut in strips

6 green onions, sliced

6 radishes, sliced

1 head lettuce, shredded

1/2 cup French salad dressing

Prepare vegetables and toss together with dressing in salad bowl. Makes six servings. This recipe is an illustration of a simple combination of raw ingredients. For the purposes of demonstration, it will be assumed that all ingredients are present in the data base available to the user, and that the data base includes the weight of each food as purchased, a refuse factor (if applicable), and the nutrient content per 100 g of edible portion of food. The steps involved in calculating the nutrient content of the mixed vegetable salad are outlined below.

Step 1: Calculate the weight of each ingredient.

tomatoes: 1 medium tomato = 135 g

refuse factor (removal of stem) = 2%

135 g x 3 = 405 g

405 B x 0.98 = 396.9 g

cucumber: 1 small cucumber = 175 g

refuse factor (removal of ends) = 3%

175 g x 0.97 = 169.75 g

green pepper: 1 green pepper = 90 g

refuse factor (removal of stem and core) = 18%

90 g x 0.82 = 73.8 g

green onions: 6 small green onions (bulb only) = 30 g

radishes: 10 medium radishes = 50 g

refuse factor = 10%

50 g x 6/10 = 30 g

30 g x 0.9 = 27 g

lettuce: 1 head = 567 g

refuse factor = 5%

567 g x 0.95 = 538.65 g

French dressing: 1 cup = 250 g

250 g x 0.5 = 125 g

Step 2: Calculate the weight of the total recipe and the weight of an individual serving. The weights of the ingredients are added together for a total recipe weight of 1,361 g. Since the total recipe makes six servings, the weight of an individual serving is 227 g.

Step 3: Calculate the nutrient content of the total recipe, the nutrient content per 100 g of salad, and the nutrient content of an individual serving.

The nutrient contents of the individual ingredients are added together to get the nutrient content of the total recipe. These values are divided by 13.61 to get the nutrient content per 100 g of salad, and they are divided by 6 to get the nutrient content per serving.

Step 4: If possible, calculate the weight per unit volume of salad.

It is possible to find conversion factors to convert the weights of the individual ingredients into volume measurements. For example:

1 cup sliced cucumber = 105 g

1 cup green pepper cut in strips = 100 g

1 cup sliced radishes = 115 g

1 cup chopped lettuce = 55 g

1 cup sliced green onions = 100 g

However, it is not possible to determine a total recipe volume by simply adding the volume measurements of the individual ingredients, because the smaller ingredients and the salad dressing wit fill the space between the larger ingredient pieces. A volume could be estimated, or the recipe could be prepared and measured.

EXAMPLE 2: Scrambled Eggs with Onions

2 large eggs

1/6 cup whole milk

1/8 teaspoon salt

1/4 cup chopped raw onions

2 teaspoons oil

Add milk and salt to eggs and beat with a fork. Fry onions in the oil. Pour egg mixture into frying pan with the onions, and stir mixture with a fork while cooking until it solidifies. Makes one serving.

For purposes of illustration, it is assumed that the available nutrient data base includes nutrient values for the raw ingredients, but not for the cooked ingredients. In this case, adjustments must be made for water loss and nutrient loss during cooking. The steps involved in calculating the nutrient content of the scrambled eggs are outlined below.

Step 1: Add nutrient levels for the specified quantities of ingredients.

The nutrients in the raw eggs, whole milk, salt, raw onions, and oil are added together.

Step 2: Readjust quantities of those nutrients that are lost during cooking.

Nutrients might be lost in evaporation or destroyed by heat during cooking. For example, it is estimated that 15% of the thiamin, 5% of the riboflavin, and 5% of the niacin in eggs are lost during cooking. For milk, 10% of the thiamin and 25% of the ascorbic acid are lost. Cooked onions are estimated to have 20% less ascorbic acid, 15% less thiamin, and 30% less folacin than raw onions. These factors should be applied to the nutrient levels in the raw ingredients to determine the nutrient levels in the cooked recipe.

Step 3: Determine weight of recipe before cooking.

The weight of the specified quantity of each raw ingredient is determined. A refuse factor is applied to the weight of the egg to calculate the weight without the shell. The weights are added together to get a total weight for the recipe.

eggs: 1 large egg = 57 g

refuse factor (shell) = 11%

57 g x 2= 114 g

114 g x 0.89 = 101.46 g

milk: 1 cup = 244 g

244 g x 1/6 = 40.66 g

salt: 1 tsp - 5.5 g

5.5 g x 1/8 = .688 g

onions: 1 cup chopped = 170 g

170 g x 0.25 = 42.5 g

oil: 1 tsp = 4.53 g

4.53 g x 2 = 9.06 g

total weight = 19437 g

Step 4: Determine weight of recipe after cooking.

In cooking scrambled eggs, weight loss occurs from evaporation. This loss is estimated to be about 8%. No fat is lost in cooking. Therefore, the final recipe weight would be 19437 g x 0.92 = 179 g.

Step 5: Determine the nutrient levels of the recipe per 100 g and per serving.

The calculated nutrient levels represent the nutrient content of one serving, since the recipe as stated is one serving. The nutrient levels must be divided by 1.79 to determine the nutrient content of 100 g of scrambled eggs.

EXAMPLE 3: Broiled Pork Chop

In this illustration, it is assumed that the available food composition table has an entry for a raw pork chop, but no entry for a cooked pork chop. The raw pork chop weighs 151 g with the bone and some visible fat. The bone is 21% of the total weight of the raw pork chop.

Step 1: Determine the decrease in weight during cooking.

It is estimated that 32% of the weight of the pork chop is lost during broiling. Therefore, the final weight of the broiled pork chop will be 151 g x 0.68 - 102.68. Step 2: Subtract nutrient loss in fat drippings.

Of the 32% loss in weight during broiling, it is estimated that 9% is fat drippings and 23% are volatiles. Therefore, the amount of pork fat lost during cooking will be 151 g x 0.09 = 13.59 g.

The nutrient content of 13.59 g of pork fat should be subtracted from the nutrient content of the raw pork chop.

Step 3: Determine nutrient loss due to cooking.

Cooking destroys some of the nutrients in the raw pork chop. For example, it is estimated that 30% of the thiamin, 20% of the niacin, and 25% of the calcium are lost in broiling. These factors should be applied to the nutrient levels in the raw pork chop to determine the nutrient levels in the cooked pork chop.

Step 4: Determine the weight of the edible portion of the broiled pork chop.

After cooking, the bone is estimated to be about 18% of the total weight of the pork chop. It is assumed that all visible fat is eaten. The weight of the edible portion of the pork chop will then be 102.68 g x 0.82 = 84 g.

EXAMPLE 4: Githeri (Kenya)

maize, dry, raw: 2 Kimbo tins (1 kg size)

kidney beans, dry, raw: 1 Kimbo tin (1 kg size)

water: 3 large Kimbo tins (2 kg size)

(Note that Kimbo tins are volume measures.)

Combine dry maize and kidney beans with water and cook the mixture for one or more hours.

For purposes of illustration, it is assumed that the nutrient data base includes nutrient values for the raw ingredients per 100 g.

Step 1: Add nutrient levels for the specified ingredients.

Convert volume measures (Kimbo tins) to gram equivalents, and multiply the nutrient values per 100 g by the gram weight of the ingredient divided by 100. For example, for energy (kcal):

maize: 355 kcal/100 g x 1,080 g per tin x 2
tins = 7,668 kcal

beans: 327 kcal/100 g x 1,000 g per tin x 1 tin = 3,270 kcal

total kcal for the recipe = 7,668 + 3,270 = 10,938

Step 2: Readjust quantities of those nutrients that are lost during cooking.

Since githeri is commonly cooked for one or more hours, there will be destruction of heatlabile vitamins; some losses may exceed 50%.

Step 3: Determine weight of recipe before cooking.

maize = 1,080 g per tin x 2 tins = 2,160 g

beans = 1,000 g per tin x 1 tin = 1,000 g

water = 2,500 g per tin x 3 tins = 7,500 g

total raw weight = 10,660 g

Step 4: Determine weight of recipe after cooking.

The average cooked weight of githeri is approximately 75% of its raw weight:

0.75 x 10,660 g = 7,995 g cooked weight

Step 5: Determine the nutrient levels per 100 g.

Divide all nutrient totals by (7,995/11)0) to give nutrients per 100 g. For example, for energy: 10,938 kcal (from step 1) / (7,995/100) = 137 kcal/100 g.

There are few studies comparing the calculated and analytic nutrient values for recipe foods; there are even fewer studies comparing the results of calculated recipes using different data bases. Both types of studies are necessary to determine the accuracy (validity) of these suggested guidelines. Differences between calculated and analytic results may reflect the effects of different types of cooking equipment (household, restaurant, industrial, or institutional), surface area of food contact exposure, length and temperature of cooking, and volume of product on the yield factors, retention factors, and water/fat, loss/gain factors. In addition, the assessment of the accuracy of the factors in current use is not adequately understood.

Differences in results as calculated by two or more data base systems may reflect different methods of estimation or calculation, choices and sources of foods and nutrients in the data base, and judgements of coders, as well as different values for yields, retentions, and water and fat changes. The uses of recipe calculations in a computerized system is discussed by Perloff [64].

The calculation of nutrient levels of multi-ingredient foods from recipes by the above scheme is, in general, an intermediate solution for a need that should eventually be solved by chemical analysis, or by more accurate estimation techniques yet to be determined. As previously mentioned, it may not be cost-effective to analyse many multi-ingredient foods. Furthermore, because of the extreme variability in many of these foods, analysis may not be the sole answer. In either case, the results from these calculations should be regarded, and annotated, as approximations.