This Session was chaired by Dr Ian Darnton-Hill of WHO-WPRO and began with a keynote address by A. Moller entitled Food Monitoring in Denmark. This was followed by papers on Food Composition Data Requirements for Nutritional Epidemiology of Cancer and Chronic Diseases by N. Slimani, E. Riboli and H. Greenfield; Developing a Food Composition Database for Epidemiologic Studies in the Pacific Islands by J.H. Hankin, L. Le Marchand, L.N. Kolonel, B.E. Henderson, and Beecher, G.R.; The Effects of Australian, US and UK Food Composition Tables on Estimates of Food and Nutrient Availability in Australia by K.M. Cashel and H. Greenfield, and Quality Control in the Use of Food and Nutrient Databases for Epidemiologic Studies by I.M. Buzzard, S.F. Schakel and J. Ditter-Johnson. These papers are all published in full in the following pages, along with a poster entitled Construction of a Database on Inherent Bioactive Compounds in Food Plants, by A.D. Walker, J.A. Plumb, G.R. Fenwick, R. Preece, & R.K. Heaney.
Posters displayed after the Session were:
Relationship Between a Dietary Measure of Antioxidant Intake and Plasma Levels, Baghurst, K.I., & Baghurst, P.A., CSIRO Division of Human Nutrition, Kintore Avenue, Adelaide, SA, Australia.
The UCB Worldfood Dietary Assessment System Utilizing the UCB International Minilist, Calloway, D.H., & Murphy, S.P., Department of Nutritional Sciences, University of California, Berkeley CA, USA.
Use of the Extended Table of Nutrient Values to Assess Nutrient Intakes of Restrained and Disinhibited Women, Champagne, C.M., & Williamson, D.A., Pennington Biomedical Research Center, Louisiana State University, Baton Rouge, LA, USA.
Organochlorine Intake of Victorian Infants from Maternal Milk, Donohue, D.C., Quinsey, P.M., & Ahokas, J.T., Key Centre for Applied and Nutritional Toxicology, RMIT University, Melbourne, VIC 3001 and National Food Authority, Canberra, ACT, Australia.
Aflatoxin M1 in Human Milk Samples for Australia, El-Nazemi, H.S., Ahokas, J.T., Donohue, D.C., & Neal, G.E., Key Centre for Applied and Nutritional Toxicology, RMIT University, Melbourne, Australia and Medical Research Council, Toxicology Unit, Carshalton, UK.
Graile: a Database for Australian Grain Legumes, Horton, J.D., Petterson, D.S., & Mackintosh, J.B., Cowirrie Computing, Relbia, TAS; Department of Agriculture, South Perth, WA; The University of Western Australia, Nedlands, WA, Australia.
A Short Questionnaire and Qualitative Fat Index for the Assessment of Fat Intakes on the Basis of the FINMONICA 1982 Survey, Kempainnen, T., Rosendahl, A., Nuutinen, O., Ebeling, T., Pietinen, P., & Uusitupa, M., Departments of Clinical Nutrition and Medicine, University of Kuopio, National Public Health Institute, Helsinki, Finland.
Heavy Metals in Taiwanese Diets, Lee, Y.S., & Chou, S.S., Department of Food Science, University of the District of Columbia, Washington, DC, and National Laboratories of Foods and Drugs, Taipei, Taiwan.
Variability in Macronutrient Contents of Selected Cereal Products Between Production Batches and Analytical Laboratories, Mugford, D.C., Bread Research Institute of Australia, Inc, North Ryde, NSW, Australia.
Database on Asian Sensory Preferences, Food Markets and Culture, Ng, F., Bell, G., Prescott, J., Waring, J., & Gillmore, R., CSIRO Sensory Research Centre, Division of Food Science and Technology, North Ryde, NSW, Australia.
Impact of Reductions in Fat Content of Australian Pork on Fat Available for Consumption in Australia, O'Dea, K., Mann, N.J., Sinclair, A.J., & Barnes, J.A., Department of Human Nutrition, Deakin University, Geelong, VIC3217 and Australian Pork Corporation, 174 Pacific Highway, St Leonard's, NSW, Australia.
Comparison of the Use of Australian and UK Food Composition Tables for Estimating Nutrient Intake, Record, S.J., & Baghurst, K., CSIRO Division of Human Nutrition, Kintore Avenue, Adelaide, SA, Australia.
The Link Between Defence Food Intake Studies and a Relational Database, Waters, D.R., DSTO, Materials Research Laboratory, Food Science Branch, Scottsdale, TAS, Australia.
Effect of Changes in the Swedish Food Database on Nutrient Estimates from a Food Frequency Questionnaire, Wolk, A., Becker, W., Ohlander, E-M., & Bergstrom, L., Cancer Epidemiology Unit, Uppsala University Hospital, and the Nutrition Division, National Food Administration, Uppsala, Sweden.
National Food Agency of Denmark, Informatics and Computer Section, 19, Mørkhøj Bygade, DK-2860 Søborg, Denmark
In 1983 the National Food Agency established a food monitoring system in order to follow the content of nutrients and contaminants in foods in a systematic manner. When the data from this system are combined with the results from the national food consumption survey of 1985 it is possible to calculate the intake by survey participants of both nutrients and contaminants. Also, it is possible to make an estimate of the maximum intake of food additives. Calculations like these are used as a basis for regulating food fortification and the use of food additives as well as establishing safe levels for the content of contaminants in foods.
Food plays a vital role in the Danish economy; the Danish food production amounts to more than US$ 17 billion. Food exports exceed US$ 9 billion, corresponding to around 35 per cent of the country's total earnings from export of industrial manufacture.
In the course of a year each person in Denmark consumes an average of one ton of food, which makes food a central aspect of daily life. In addition, awareness among Danes about the food they eat is increasing.
In Denmark national food legislation is the responsibility of three different ministries, the Ministry of Health, the Ministry of Agriculture and the Ministry of Fishing. Responsibility for the General Food Act of 1973 lies within the Ministry of Health, and the executive functions are carried out by the National Food Agency. The Agency's objectives are to protect consumer health, to protect consumers against misleading information/fraud, to ensure reasonable conditions for retail stores and manufactures and to promote healthy dietary habits.
The National Food Agency comprises two scientific institutes and four administrative divisions. The institutes, the Institute of Toxicology and the Institute of Food Chemistry and Nutrition, represent the specialist scientific knowledge which forms the basis for the Agency's administration of the Food Act and the provision of nutritional guidance to the general public.
• The Food Monitoring System
The contents of both nutrients and contaminants in foods on the Danish market have been analyzed by the National Food Agency and associated laboratories for several decades.
Due to the increasing focus on diet and health issues, as well as a desire to ensure that chemical analyses within the individual working areas of the Agency were linked together as a whole, the Agency established a food monitoring system in 1983. The foundation of the food monitoring system was described in a proposal prepared by an internal working group at the Agency (1).
The basic objectives of the Danish food monitoring system are to:
ascertain whether, over the long term, changes occur in the content of desired and undesired substances in Danish foods.
combine such changes with changes in eating habits
assess whether changes expose the Danes to nutritional or toxicological health hazards
obtain background material and a basis for decision-making to remedy any problem that may have arisen.
Therefore, the practical work with the Danish food monitoring system implies:
watching the content of nutrients and contaminants in selected foods closely
watching the consumption pattern of the Danish diet closely.
The food monitoring system is designed to supply data about the changes over time in the contents in foods of nutrients and contaminants. It is also designed to be linked with the data from the food consumption survey in order that the nutrient and contaminant intake of the population can be calculated.
As changes in the content of desired and undesired substances as well as changes in dietary habits occur slowly, the monitoring system will run for a long period. The results of analyses of the food monitoring system are reported continuously.
Every five years a major evaluation of the results of the preceding five-year period takes place. The first report for a complete five-year period was published in 1990 covering the years 1983–87 (2).
• Selected Areas of Monitoring
The food monitoring system covers nutrients as well as heavy metals/other trace elements, nitrate, pesticides and PCB in selected foods (Table I). With the exception of nutrients, the content of these compounds/substances in most cases originates from influences from the external environment. Components to be monitored are carefully selected on the basis of the existing knowledge about their nutritional importance or toxicity, their occurrence in foods, and the actual consumption of these foods.
Microbiological studies and examinations of radionucleides have until now not been included in the food monitoring system but will be reported in the future. A proposal for a microbiological food monitoring system is in preparation.
In the first five-year period (1983– 1987) 10,060 samples were analyzed, in the second period (1988–1992) 9, 341 samples, and for the third period (1993– 1997) a total of 8,150 samples are to be analyzed.
Table II gives more details about the sampling with regard to nutrients and contaminants.
The expenses of the food monitoring system amount to about US$ 3.3 million per five-year period. During each period a total of approximately 65 full-time persons are devoted to the project, i.e. approximately 13 full-time positions per year at the four (formerly five) regional laboratories, as well as several full-time positions at the National Food Agency.
Table I. The elements of the food monitoring system
|Food category||Nutrients||Trace elements and nitrate||Pesticides and PCB|
|Fruit and vegetables||Fat, protein, ash, dry matter, fiber, vitamin C (tomatoes: glutamic acid)||As, Cd, Cr, Hg, Ni, Pb, Se|
In vegetables also nitrate
|Cereal products||Fat, Protein, ash, dry matter, fiber, vitamins B1, B6, Ca, Fe, K, Mg, Na, Zn||As, Cd, Cr, Hg, Ni, Pb, Se|
|Milk, milk products and eggs||Fat, protein, ash, dry matter, fatty acids, vitamins A, B1, B2, Ca Fe, K, Mg, Ma, Zn, I||In eggs: As, Cd, Cr, Hg, Ni, Pb, Se||DDT, dieldrin, HCB, α-HCH, β- HCH, lindane (γ- HCH), heptachloroepoxide, PCB|
|Fish||Fat, protein, ash, dry matter, fatty acids, vitamin D||As, Cd, Cr, Hg, Ni, Pb, Se||DDT, dieldrin, HCB, α-HCH, β-HCH, lindane (γ- HCH), heptachloroepoxide, PCB|
|Meat||Fat, protein, ash, dry matter, Beef, chicken, pork: Fe, Mg, Zn, vitamins B1, B2, B6|
|Offal||Fe, fat, protein, ash, dry matter||As, Cd, Cr, Hg, Ni, Pb, Se|
|Animal fat||DDT, dieldrin, HCB, α-HCH, β- HCH, lindane (γ- HCH), heptachloroepoxide, PCB|
PCB: polychlorated biphenyls
Table II. Number of samples examined
* The National Food Agency's control programs include approx. 1200 samples/year covering about 120 different pesticides
Figure 1. Potatoes, daily intake, Danes, 15–80 yrs
Criteria for selection of samples and analysis
Nutrients. The nutrients included in the system have been selected based on one or more of the following criteria (3):
the daily intake of the nutrient in Denmark is lower than or around the recommended level, either for the population as a whole or for specially exposed groups of the population
the nutrient is present only in few foods
the nutrient shows stability problems
the nutrient is added to one or more foods, either compulsorily or voluntarily.
The analyses have been given priority based on an overall evaluation according to these criteria and resulted in the substances listed in Table I.
Trace Elements and Nitrate. The trace elements which are included in the system, see Table I, have been selected on the basis of existing knowledge of their toxicity and occurrence in foods compared with food consumption. In the case of nickel, arsenic, chromium and selenium there has also been a desire to gain more knowledge about their occurrence in Danish foods.
Only vegetables have been selected to be monitored for nitrate. The concentration of nitrate in fruit and other foods is so low that they have only minor significance for human intake of nitrate.
Pesticides and PCB. For a number of years combined control and monitoring analyses have been carried out on organochlorine pesticides and PCB in fish, meat, eggs, and milk and milk products. Since 1983 these studies have been included in the monitoring system. The analyses comprise persistent organochlorine pesticides. Among industrial chemicals, at present the system includes only the polychlorinated biphenyls (PCB).
Figure 2. Fish, daily intake of cooked edible portion, Danes, 15–80 yrs
• The Danish Food Composition Data Bank
After careful evaluation of the results of the analyses the relevant food monitoring data are transferred from the Agency's laboratory information management system and stored in the Danish Food Composition Data Bank. The food monitoring system is a substantial data source, due to the systematic and continuous flow of new data from the monitoring system into the databank.
At present the databank comprises information for about 2000 foods on the Danish market. In the databank information is collected on 255 different compounds.
• The National Food Consumption Survey
In 1985 the National Food Agency of Denmark carried out a nationwide food consumption survey (4, 5). The objectives of the survey were:
To identify population groups which are at risk from a nutritional point of view
To evaluate the significance of fortifying foods with nutrients
To estimate the exposure of different population groups to contaminants and food additives
To identify foods which contribute significantly to the nutrient intake in different population groups
To contribute to studies of the relationships between diet, health and disease.
Figure 3. Percentage of dietary energy from fat, Danes, 15–80 yrs
The survey included 2242 persons, 15–80 years of age. They constituted a representative sample of the adult Danish population. The participants in the survey were interviewed about their food consumption habits using a dietary history method, which was developed particularly for this survey.
The dietary history method gives information about the usual diet of an individual during an extended period of time. There is no doubt that the method tends to overestimate regularity in the eating pattern. The method itself encourages the participants in the survey to emphasize usual food consumption, because it is easier to remember the usual meal pattern than all the unusual events which interfere with the habitual food intake. The results of the survey are, however, in excellent agreement with the results from other similar data sources, such as food balance sheets, household budget surveys, and other food consumption surveys. The conclusion is that the average consumption found in the survey is very close to the real intake except for a few foods and beverages, such as sugar and alcohol.
The dietary history method used in the present survey enables the ranking of individuals according to their intake of foods, nutrients, contaminants and other known constituents of food.
• The Food Intake
Two different types of food intake distributions were identified, one for foods eaten by everyone, and the other for foods consumed by only some sectors of the population. As examples, the distribution of the intake of potatoes and fish within the adult population is shown in Figures 1 and 2.
Figure 4. Daily intake of vitamin A, Danes, 15–80 yrs
The shape of the intake curve for potatoes is typical of foods that are consumed by practically everybody. These foods are cereals, white bread, rye bread, coarse vegetables, meat, poultry, separable fats and eggs. Figure 2 illustrates the shape of the intake curve for foods and beverages which are consumed by some people only. Other examples are cheese, soft drinks, beer and tea.
• The Intake of Substances from Foods
The Agency has developed computer software that allows the data from the food consumption survey to be combined with the data from the food composition databank. Thus computation of the intake levels of nutrients and contaminants of the individuals who participated in the food consumption survey gives an estimate of the distribution of intake within the adult population.
Figures 3 to 5 are examples of the results of the calculations of nutrient levels. Figure 3 illustrates how the fat-energy-percentage of the diet of Danish women and men is distributed. The fat-energy-percentage seems to be very high when compared to the recommended level of 30 per cent of the dietary energy from fat. In fact almost all Danish adults seem to eat a diet that is higher in fat than the recommendations.
As a result of this, the National Food Agency has intensified its information campaigns on good eating habits with short advertisements on Danish television, written material for schools etc. Recently an extensive campaign for reducing intake of fats, especially butter and margarine, was launched. The motto of this campaign is “Scrape your bread”.
Figure 5. Daily intake of iron, Danes, 15–80 yrs
Figure 4 indicates that a very large percentage of Danish women and men consume more vitamin A than the recommended level of 800 and 1000 ug/day, respectively.
Figure 5 shows that the same is not the case for iron. The iron intake of most women in Denmark seems to fall below the recommended level of 12–18 mg/day.
The results for nutrient intakes have been used to evaluate the nutritional importance of the fortification of foods. As a result of this evaluation the obligatory fortification of flour and margarines with vitamins and minerals was abolished in 1987, because the contribution of fortification to the total nutrient intake was shown to be either unnecessary or negligible. The evaluations showed that intakes of the vitamins and minerals in question (vitamins, riboflavin and thiamin; minerals, calcium, phosphorus) were above the recommended levels. For iron it was shown that the contribution of fortification (inorganic iron) taking bioavailability into consideration was less than 10 per cent of the recommended intake.
A new food consumption survey is at the planning stage. A pilot study was carried out in the autumn of 1993. The main food consumption survey will take place in 1994.
Trace Elements. Table III shows the calculated intakes of mercury, cadmium, lead and arsenic from foods. For mercury, cadmium and lead it appears that intakes from foods are well below the PTWI (Provisional Tolerable Weekly Intake) proposed by Joint FAO/WHO Expert Committee on Food Additives (6,7). Special interest has been devoted to the lead content in foods since the dominant source of lead contamination of foods, especially vegetables and crops, is lead emitted from motorcars running on petrol.
Table III. Intake of trace elements, all values in μg per person
|Trace element||Daily intake||Weekly intake|
PTWI: based on a body weight of 70 kg
p0.50, p0.90, p0.95: 50th, 90th and 95th percentiles
During the last decade the lead content of petrol has been lowered substantially and unleaded petrol introduced on the Danish market. The influence of this is clearly seen in the decreasing lead content in many foods, e.g. offal, beverages (wines in particular), berries, certain types of fruits, greens, oat, rye, wheat, rye bread and cod liver. In other food groups no change was discernible. These groups comprise meat, imported fruit, roots and tubers, cabbage, certain vegetables and fish.
For arsenic the total intake of organic and inorganic arsenic is shown in Table III. No PTWI has been proposed for arsenic, but a value of 140 μg/day has been established as PMTDI (Provisional Maximum Tolerable Daily Intake) by a group of experts under FAO/WHO in 1983. This value applies to inorganic arsenic only. Most of the arsenic in the Danish diet originates from fish. This is organic and considered non-toxic to humans. Arsenic is therefore not considered to be a problem in Danish foods.
Pesticides and PCB. So far it has not been possible to calculate the exposure of the population to pesticides and organic pollutants from food in the same manner as for heavy metals. The analyses of pesticide residues in foods have been concentrated on those foods where high levels are most likely. There are too many gaps in our present knowledge about the content in other foods to allow us to make a calculation of the total exposure of the participants in the survey to pesticides from foods.
The use of food additives in Denmark is regulated by the National Food Agency through the so-called positive-list (8). The list specifies the maximum amount of a food additive that can legally be used in individual foods. Although the maximum amount of food additives that might be used is known, there is no complete picture of the actual use by the food industry. Therefore, no calculation can be made of the actual exposure of the population to food additives, but only an estimate of the maximum exposure, which would occur if food producers used all the permitted food additives in their maximum amounts. The calculated exposure will in all cases be higher than the actual, since most foods are manufactured without making full use of all permitted additives. The calculated maximum intake is, however, of considerable interest from a regulatory point of view. It allows a check to be made as to whether the limitations that have been introduced in the use of additives in individual foods are realistic in relation to the ADI-values.
Figure 6. Maximum erythrosin intake, Danes 15–80 yrs (ADI: 625 μg/kg body weight/day)
Figures 6 and 7 show, as examples, the intake distribution of the calculated maximum intakes of erythrosin and benzoic acid/benzoates. From the figures it appears that while the maximum intake of benzoic acid/benzoates permitted according to the positive-list is well below the ADI-value, the same is not the case for erythrosin.
For erythrosin, however, the ADI-value has been reduced from 625 to 50 g/kg body weight per day recently (7). The distribution curve in Figure 6 was obtained with the erythrosin levels permitted at the time of the survey. Therefore, intake levels have to be compared to the higher ADI-value of 625 μg/kg.
As a whole, the calculations based on actual food intake have shown that the method used to determine the amounts of food additives permitted in the “positive-list” is applicable. The method is the so-called budget-method (9). It is not a scientific method, but a practical administrative tool to predict the maximum intake of a food additive. The main assumption in the budget-method is that the maximum daily consumption for an adult is 1.5 kg of foods and 6 L of beverages and water. It is also assumed that only half the foods are industrially processed and thus contain food additives. As far as the liquid intake is concerned the assumption is made that only 25 per cent of beverages contain food additives. The ADI can be divided between solid and liquid foods according to technological needs. If the required level is too high compared to the ADI available the additive may be limited to either solid or liquid foods or to certain groups of foods.
Figure 7. Maximum benzoic acid/benzoates intake, Danes 15–80 yrs (ADI: 625 μg/kg body weight/day)
The calculations show that each of these assumptions is reasonable for 90 per cent of the adult population. The type of calculations illustrated in Figures 6 and 7 confirm that the budget-method is a reliable tool in the administration of food additives.
The Danish Food Monitoring System has proven to be a valuable tool for identifying nutritional or toxicological areas where action has to be taken, as well as actions in the area of food administration and regulation. Systematic monitoring of foods is necessary, also in the future, to ensure the supply of healthy foods to the Danish population. As dietary habits are constantly changing, it is important to adjust the monitoring system on a continuing basis so that areas without problems are monitored less often, while newly detected problems are taken up for inclusion in the system.
(1) National Food Agency (1984) Establishment of a Food Monitoring System, Statens Levnedsmiddelinstitut, Soborg
(2) National Food Agency (1990) Food Monitoring in Denmark: Nutrients and Contaminants 19831987, National Food Agency of Denmark, Soborg
(3) Haraldsdottir, J., Heidemann, F. & Leth, T. (1982) Establishment of a Monitoring System for Nutrients in Foods, Statens Levnedsmiddelinstitut, Soborg
(4) Haraldsdottir, J., Holm, L., Jensen, J.H. & Moller, A. (1986) Dietary Habits in Denmark 1985, 1. Main Results, Publication No. 136, Levnedsmiddelstyrelsen, Soborg
(5) Haraldsdottir, J., Holm, L., Jensen, J.H., & Moller, A. (1987) Dietary Habits in Denmark 1985, 2. Who Eats What?, Publication No. 154, Levnedsmiddelstyrelsen, Soborg
(6) FAO/WHO (1972) Sixteenth Report of the Joint FAO/WHO Expert Committee on Food Additives, WHO Technical Report Series No. 505, Geneva
(7) FAO/WHO (1989) Thirty-third report of the Joint FAO/WHO Expert Committee on Food Additives, WHO Technical Report Series No. 776, Geneva
(8) National Food Agency (1988) Fortegnelse Over Godkente Tils'tningsstoffertil Levnedsmidler, Levnedsmiddelstyrelsen, Soborg, Publication No. 171
(9) Hansen, S.C. (1979) J. Food Protect. 5, 429–434
Nadia Slimani, Elio Riboli
Programme of Nutrition and Cancer, WHO International Agency for Research on Cancer, Lyon, France
Department of Food Science and Technology, University of New South Wales, Kensington NSW 2033, Australia
Nutritional epidemiology is concerned with, among other things, establishing the association of diet and disease. The principles of nutritional epidemiology drive the requirements for nutrient databases for valid measurement of dietary exposure. The potential impact of random and systematic errors in food composition data on computation of nutrient intakes in prospective multi-center studies is discussed and the needs for modeling studies and time-related databases highlighted.
Nutritional epidemiology is concerned with, among other things, establishing the association between diet, health and disease. Establishing a relationship relies on measurement of exposure to a dietary factor and estimating the absolute (incidence) or relative risk (odds ratio) of having a given disease associated with a given level of exposure. The categorization in quantiles of a population distribution represent one type of classification of subjects that can be used usually with three to five classes of exposure from lowest to highest. The establishment of a statistical association relies on the absence of bias in all of the observations including the dietary observations. Systematic errors (bias) have to be excluded as they could affect classification of disease cases and control subjects unequally. Random errors, which have an equal chance of occurring in affected and unaffected individuals, thus affect the classification process equally for all groups. Nevertheless even random errors can affect the validity of a study's findings by distorting the estimation of relative risk towards the null value of 1 and increasing the variance of observations, thus blurring true relationships. Procedures exist for minimizing bias, controlling measurement errors, and preventing misclassification. These procedures rely to an important degree on collecting data according to clearly defined, rigorous, standardized protocols for all aspects of the scientific observations (1).
Epidemiological investigations of the role of diet in cancer and other chronic diseases to date have revealed in many cases a weak association (2), but even such a weak association is potentially of great biological significance due to the large numbers of people likely to experience high or low exposures to dietary factors since everyone eats. There is a strong case, therefore, for continuing work to establish the dietary links.
The major method of measuring dietary exposure has been the collection of data on food intakes and converting these data to nutrient intakes by means of food composition databases. While there has been progress in understanding the errors which can arise in measurement of food intake, the role of errors in conversion of these intakes to nutrient intakes is not so well investigated. In fact, much published literature about food composition databases in epidemiological studies is descriptive rather than analytical (3, 4, 5,) and Willett (3) noted that no formal analysis had been done of the impact of variability (systematic) in nutrient content of food in nutritional epidemiology. The only analytical study appears to be that of Beaton (6) who investigated the impact of biological variability in the composition of foods (a source of random error) on nutrient intakes calculated from two different one-day intakes (sample diets) by use of the US nutrient composition tables for which measures of dispersion are given (7). Some caution is needed in interpreting these findings given the nature of food table compilation which rejects statistical outliers and may also include some subjective judgement in acceptance of the individual values from which the mean and standard deviations of values are computed. However, Beaton's analyses give valuable indications of the low impact of random biological errors on computed nutrient intakes especially in diets composed of large numbers of foods. It would be useful if such analyses could be extended to investigations of systematic bias in food composition values, a topic which is of great concern in epidemiology.
• Nutrient Databases for Nutritional Epidemiology
In examining nutrient database options for any study in nutritional epidemiology, it is important to consider the aims and methods of the discipline particularly in order to avoid the ad hoc selection of a database portrayed as the “usual approach” in the model described by Buzzard and co-workers (Table I) (8). The first step is the definition of the nutritional hypothesis which is to be tested in the study. The several different possible approaches in descriptive and analytical nutritional epidemiology require separate and lengthier consideration both in terms of the nutrient composition database and the dietary methodology. However, this discussion will be restricted to the concerns associated with a prospective multicenter study which it is hoped will shed considerable light on the relationship of diet to cancer and other chronic diseases.
Table I. Ideal versus usual approach to planning a diet study
|Ideal Approach||Usual Approach|
|Identify nutrients of interest||Select data collection method based on cost and ease of administration|
|Determine level of specificity of food descriptions required to assess the nutrients of interest||Collect dietary data prior to selecting an appropriate database for nutrient analysis|
|Select a data collection method that will accommodate the desired level of specificity|
|Develop or modify an existing database to:||Select an existing nutrient nutrient database without evaluating:|
|•||accommodate the desired level of specificity||•||the level of specificity of foods included in the database|
|•||provide complete, accurate, specific, and updated values for the nutrients of interest||•||the completeness, the accuracy, the specificity or the currency of the nutrient database|
Adapted from Buzzard et al. (8)
• Database Requirements for Prospective Studies
The EPIC study (European Prospective Investigation into Cancer and Nutrition) has the advantages of a large study population living in several geographical areas with different dietary patterns and cancer incidences (northern and southern Europe), with an appropriate age and socioeconomic distribution, which has the power to establish valid relationships between diet and even relatively rare cancers (2).
A multi-country prospective study requires within-cohort and between-cohort analyses, and further has implications for the period over which data will be collected (changes in environment and observers) as well as the volume of data to be gathered. The between-cohort analysis is particularly important to determining the impact of large variations of diet on disease incidence since many within-cohort studies are of national populations with relatively homogeneous dietary intakes.
Any prospective epidemiological study planning to analyze and compare dietary information for several countries will need to take into account several considerations for the food composition database used to analyze the food intake data. These are discussed below.
Need for a Tailored Database
Nutritional epidemiology requires the database to be specifically tailored to the actual foods reported consumed, and according to the dietary method used, to increase the accuracy of nutrient intakes computed; thus the concept is of a user database which may need to draw on several reference database in its compilation. The important distinction must be drawn between this approach and other applications in which it is acceptable to tailor the dietary intake obtained to the database (i.e. matching a food with the most similar food in the database). In epidemiology it is the reverse, a point not always understood in the field. For example, a comparative study of two databases (9) found a new national database deficient since it did not have values for lean meat unlike the foreign database used previously. This was despite the availability of the published “lean only” data for the local meats in the literature which the authors cited (10) and which could have been used to tailor the local database in order to avoid over-estimation of the fat intakes of study populations to which the database was subsequently applied.
Needs for Local Data for Local Foods
Nutrient composition data for the local foods as consumed in the specific countries will be necessary for a multi-country prospective study of nutrition. To use non-indigenous data, particularly for staple foods, could suppress the effects of an important potential source of dietary variability. For example, fat from meat is often of interest in surveys of diet and degenerative disease. However, meat is a food which is known to vary dramatically in its fat content over time and between countries. When new compositional data for meat became available in one particular country and was found to be up to 50 per cent leaner than the previous data set (data origins obscure) (10), the total fat available from the food supply daily per capita dropped from 145 g to 119 g (i.e. 19 per cent), the total fat available daily per capita from meats fell by 27 g (from 52 g to 25 g) and the total fat available daily per capita expressed as per cent energy available fell from 37 per cent to 33 per cent, in other words, it fell below the (then) dietary target (11). The ratio of vegetable to animal fat increased from 0.55:1 to 0.74:1. Hence it can be expected that use of meat fat data which are “too high” for one country, and “too low” for another country (in relation to the “true” values) could either obscure a true difference in fat intakes between the two countries or artificially create a difference where there is none.
Some simple tests using pilot dietary data from two countries participating in the EPIC study varying the fat content of meat showed some effect on the difference between national samples (as expected). In fact, a 10 per cent increase in fat content of two food groups combined (meats, milk and their products) in country A and a 10 per cent decrease in country B produced a 50 per cent decrease in the difference in total fat intake between countries.
So far as within cohort analyses are concerned, when the fat content of meat was varied by up to 50 per cent in a mathematical simulation by computer there was no effect on classification of subjects (as expected), but also no effect on classification when the content of fat in milk and milk products was lowered by 30 per cent at the same time that the fat content of meat was increased by 30 per cent. This result was unexpected and was not affected by any correlation between meat consumption and consumption of milk and milk products. However, different results can be expected in populations with different dietary habits.
Further simulations on other nutrients and other food groups could act to improve estimates of error in nutrient intakes obtained using databases. It could be hypothesized that other foods may be less heterogeneous between countries (particularly some fruits and vegetables) and therefore local data would not be needed but the final answer cannot be known until some analytical data for such foods are available and the hypothesis tested.
Nutritional Epidemiology Needs Time-Related Data
Foods change in composition over time, particularly when there are changes in breeding of plants and animals, changes in feeding regimes, and changes in preparation of foods for retail sale (e.g. butchering). Changes in regulations affecting foods (e.g. the introduction of mandatory fortification with a nutrient such as thiamin) also have the potential to alter their composition over time. Food tables and databases, on the other hand, tend to be up to a decade out of date, given the delays experienced between collecting and analyzing the food, the delays in publishing the data and the delays in compiling data into databases. There is potential therefore for any database to be out of phase with the dietary intake data to which the database is applied (12). This point is of importance for studies of trends over time within a country.
A recent study (13) examined the sodium and potassium intakes of a group of 27 females using urinary excretion and several dietary methods. One part of the study involved computing the dietary intake of sodium and potassium by means of analyzing sets of 24 hour food intake data against three subsequent versions of the Australian nutrient database NUTTAB (NUTTAB89 vs NUTTAB90 vs NUTTAB91/92) (14, 15, 16). Note that the food intake data sets were constants. The values for sodium and potassium in foods had changed over this period, but since some foods increased in sodium or potassium by 10–100 per cent, while others decreased by 10–100 per cent, the net effect was of no change in group means for total intake from foods of these two minerals; the changes in food composition canceled each other out.
Needs for Analytical Data for Foods
The lack of compositional data for indigenous foods consumed in some of the southern European countries is a soluble problem for EPIC. Many useful data are being produced in local laboratories, and once they have been assessed and scrutinized can be used to update and amend the databases used to analyze the food intake data. Re-analyses done in the future will refine the estimates of nutrient intakes calculated (rather than alter them completely). The analyses undertaken can of course be expanded once data for important but often missing components such as carotenoids, vitamers E and other bioactive compounds become available. This kind of laboratory work and food intake re-analyses could be expected to shed additional light on dietary relationships with the cancers and other diseases which emerge during the 10–15 year course of the observations of the prospective study. Such re-analyses of biological samples such as blood have already been provided for in the EPIC study (2).
Needs for a Complete Database
The nutrient database will need to be complete (i.e. include all relevant foods for a given nutrient) if nutrient intakes are not to be underestimated; this is a point which has been well-accepted (9) and some data have been provided, e.g., by Stockley (17) who reviewed studies of error associated with missing values in databases, citing underestimates of B vitamin intake ranging from 1.5 per cent to 14.3 per cent, and recoveries of only 69 per cent of total polyunsaturated fatty acids analyzed in duplicate diets as opposed to computed nutrient intakes, improving to 89 per cent when the missing fatty acid values were inserted. Also mentioned were the new starch values for UK potatoes ranging from 11 g/100 g to 23.0 g/100 g, according to cultivar, with the average value (weighted by tonnage) being 17.0 g/100 g compared to the value of 20.3 g/100 g given in the food tables then current. Starch intake from potatoes may therefore be only 60 per cent of that computed from the tables. It should be relatively easy to design simulations to investigate the effects of missing data using pilot studies of dietary intake carried out for the EPIC investigation.
Nutrient values should be, as far as possible, analytical values representative of the foods consumed by the study population. It might be expected that this approach would minimize bias. The food composition analyses, ideally, should reflect any potential regional differences in foods which are key sources of nutrients. The extent of regional variation in foods and nutrients is not well established and the potential biases which may occur by the use of non-representative data are a potential source of misclassification of subjects, as well as obscuring between-cohort effects.
Data imputation, which may be necessary for tailoring the database is, by definition, a biased procedure and is likely to affect classification of subjects. Bias will be minimized by basing imputations on analytical data wherever possible from the country concerned. There is some misunderstanding about imputed data, and whether they are better than analytical data or not. The point is that an imputation, by definition, is done against a previous set of analytical data, and the assertion of validity cannot be tested without new analyses, creating a circular situation. For epidemiological studies a trade-off between missing data and imputed data has to be made. Again a series of simulation tasks could possibly identify the priorities for chemical analysis and indicate where compromises could be made.
On the practical side, the problems for a food composition database are relatively simple to understand. Prospective studies have to address (among other problems) the implications of personnel and methods of measurement being subject to change during the study. As the database has to be maintained over an extended period, quality control will demand extensive computerized data documentation, including dating of food analyses (12).
Mode of Data Expression
To compare countries, even where indigenous data sets are available for each country, conversions to the same basis of data expression will be needed for nutrients and foods. Specific software is needed to meet all of the requirements identified and initial requirements for such software have been described (18). These requirements for a comprehensive, multi-country user database differ considerably from a national reference database and will have to be specifically incorporated in purpose-built software.
Documentation of Nutrient Database
The nature of epidemiological work makes difficult the question of replication of a study as a method of validation. An indispensable part therefore of the reporting of any epidemiological research is a requirement to document the nutrient database used in sufficient detail to enable detailed scrutiny when under review or when comparisons are made with other studies. This point has recently been re-emphasized (17).
A major barrier to achieving the “ideal approach” is the difficulties the majority of users experience in expressing their needs in a way which permits a custom database to be compiled.
There are two powerful tools potentially applicable to the problems posed by large-scale multi-country studies. First, so long as great attention is paid to the collection of the dietary intake and other data, these data could be re-analyzed against date-stamped sets of food composition data in the future enabling a clarification of the nutrient exposure. Second, further studies involving computer modeling which examine the potential impact of defined systematic errors in food composition data on dietary intake data, particularly by using sample populations with a wide variety of food habits, and with intakes corrected to energy intake, would undoubtedly be useful in validating differences in dietary exposure.
Finally, future prospective studies could consider the option of collecting “food archives” in which sample diets from study populations are collected and stored at low temperatures for future (and replicated) analysis in much the same way in which plasma or urinary samples for biochemical markers are currently collected and stored (2).
(1) Hennekens, C.H., & Buring, J.E. (1987) Epidemiology In Medicine, Little, Brown and Company, Boston, MA
(2) Riboli, E. (1992) Ann. Oncol. 3: 783-91
(3) Willett, W. (1990) Nutritional Epidemiology, Oxford University Press, Oxford
(4) West, C.E., & van Staveren, W.A. (1991) in Design Concepts in Nutritional Epidemiology, B.M. Margetts & M. Nelson, (Eds.), Oxford University Press, Oxford, pp. 101– 119
(5) Paul, A.A., & Southgate, D.A.T. (1988) in Manual on Methodology for Food Consumption Studies, M.E. Cameron & W.A. van Staveren, (Eds.), Oxford University Press, Oxford, pp. 121–144
(6) Beaton, G.H. (1987) in Food Composition Data: a User's Perspective, W.M. Rand, C.T. Windham, B.W. Wyse, & V.R. Young, (Eds.), UNU Press, Tokyo, pp. 194– 205
(7) US Department of Agriculture (1976- ) Composition of Foods: Raw, Processed, Prepared, Agric. Handbook No. 8 series, USDA, Washington, DC
(8) Buzzard, I. M., Price, K.S., & Feskanich, D. (1991) in The Diet History Method, L. Kohlmeier (Ed.), Smith-Gordon and Company Ltd, London, pp. 39–51
(9) Magarey, A., & Boulton, T.J.C. (1991) Aust. J. Nutr. Diet. 48, 128-31
(10) Greenfield, H. (Ed.) (1987) Food Technol. Aust. 39, 181–140
(11) Cashel, K.M., & Greenfield, H. (1995) J. Food Comp. Anal. (in press)
(12) Buzzard, I.M. (1991) in Proceedings of the 16th U.S. National Nutrient Databank Conference, The CBORD Group Inc, Ithaca, pp. 73– 77
(13) Jia, Y. (1992) MSc thesis, University of New South Wales, Sydney
(14) Commonwealth Department of Health and Community Services (1989) NUTTAB89, diskette, AGPS, Canberra
(15) Commonwealth Department of Health and Community Services (1990) NUTTAB90, diskette, AGPS, Canberra
(16) Commonwealth Department of Health, Housing and Community Services (1991) NUTTAB91/92, diskette, AGPS, Canberra
(17) Stockley, L. (1988) J. Hum. Nutr. Diet. 1, 187–195
(18) Greenfield, H., Hémon, B., Slimani, N., & Riboli, E. (1991) NUBEL/EURONIMS Meeting, Antwerp
(19) Murphy, S.P. (1993) Aust. J. Nutr. Diet. 50, 176
Jean H. Hankin, Loïc Le Marchand, Laurence N. Kolonel
Epidemiology Program, Cancer Research Center of Hawaii, Honolulu, HI 96813, USA
Brian E. Henderson
Salk Institute, La Jolla, CA 92037, USA
Nutrient Composition Laboratory, U.S. Department of Agriculture, Beltsville, MD 20705, USA
As part of collaborative surveys of lifestyle risk factors for cancer and other chronic diseases in several Pacific Islands, diet studies were conducted among samples of semi-urban 45–65 year old men and women living in each island. Local nutritionists, dietitians, and other health workers identified the food items usually consumed, along with the seasonal fruits and vegetables that were major sources of carotenoids. The food composition table used to calculate nutrient intakes was developed during and following the survey, using a variety of procedures, including recipe calculations, laboratory analyses for carotenoids, and sourcing data from national and international food composition tables. The original carotenoid data for the Pacific Islands fruits and vegetables are presented in this paper.
Developing a food composition database for emerging and somewhat isolated nations, such as the South Pacific Islands, presents an interesting challenge for nutrition researchers. There are problems in identifying the various traditional and imported foods, determining the usual food preparation methods, and assigning appropriate nutrient values to rare items not found in published food composition tables. We had an opportunity to meet this challenge in our recent study of diet and other lifestyle risk factors for cancer in several South Pacific Islands. For the dietary component of the study, the objectives were to obtain representative data on the usual diets of the islanders and to characterize the dietary intakes according to particular nutrients and other dietary components, as well as selected food items and food groups. This paper will review the background of the study and the procedures followed for identifying the foods usually eaten, developing the diet history questionnaire, determining the composition of the local foods, and creating the database for each island. In addition, we will discuss some of the problems that may occur in developing a database for an emerging country and offer a few suggestions that may be helpful.
• Background of the Study
Since 1980, the South Pacific Commission (SPC), with the assistance of the University of Southern California Comprehensive Cancer Center (USCCCC) and the Cancer Research Center of Hawaii (CRCH), has been recording all reported cases of cancer in the South Pacific region. Analysis of the incidence data revealed marked variation in the rates of several site-specific cancers among the different ethnic and island populations (1,2). For example, the Polynesians in Hawaii, French Polynesia, Cook Islands, and New Zealand, tend to have high rates of several cancers which most likely are related to diet. For instance, stomach cancer rates are generally higher among Polynesians as compared to the other islanders. Lung cancer rates are high among all Pacific Islanders except among the Melanesians and Indians in Fiji. Breast cancer and prostate cancer are also relatively high among the Polynesians. However, the rates of colon cancer among Polynesians living in Hawaii and New Zealand are low in comparison to the Caucasians living in these respective countries. Among the Melanesians, New Caledonians have considerably higher rates of lung cancer than the Fijians, as well as the Indians in Fiji. This is of particular interest because the Fijians have high lifetime rates of smoking.
The variation of incidence rates within and among the ethnic groups suggested that environmental, and in particular lifestyle factors, may be associated with these variable cancer patterns. To identify particular risk factors, the SPC, CRCH, USCCCC and the Ministry of Health of each island conducted cross-sectional surveys in the Cook Islands, Fiji, French Polynesia, and New Caledonia between 1988 and 1992. The objectives were to collect data on the prevalence of lifestyle factors (such as smoking, drinking, diet, reproductive history, physical activity and obesity) among representative samples of semirural adults and to correlate these data with the observed cancer incidence patterns. These island communities are undergoing rapid economic, technological and social change, which is having an impact on their eating patterns, especially in urban areas. For instance, the use of imported foods has resulted in a modification of their traditional food practices. We hoped that the study findings would lead to greater knowledge about the causes of cancer in the South Pacific Region and would be utilized by the Ministries of Health for planning public health interventions to control cancer and other chronic diseases.
The same methodology was followed in each country to obtain comparable results. The surveys were conducted in the same season (June through August) of the year. Random samples of approximately 250 semi-rural males and females, 50 to 65 years of age, from each main ethnic group living on the island were included in each of the surveys. The questionnaires included a diet history, information on cigarette smoking, alcoholic consumption, physical activity, and medical and reproductive histories. Additional components included anthropometric data (weight, height, triceps and skinfold measurements); plasma and serum samples which were subsequently analyzed for carotenoids and tocopherols; and urine samples which were analyzed for sodium and cotinine (an indicator of smoking history).
Identifying the Food Items for the Dietary Assessment
To assess the role of diet in the etiology of diseases, such as cancer and heart disease, investigators seek information on the usual dietary intake of individuals. Generally, a diet history method which provides an estimate of the frequencies and amounts consumed during a specified period of time is recommended (3–5). To estimate the usual diet of the islanders, we utilized a diet history that included those food items that were likely to be consumed during a one-month period. This time interval seemed appropriate because of the similarity in the dietary patterns of the villagers from month to month. In addition, seasonal fruits that were major sources of carotenoids and ascorbic acid were included in the questionnaire.
The selection of the particular food items for the diet history began several months before the survey. The nutritionists, dietitians and epidemiologists from the Ministry of Health in each country identified the foods usually consumed at least once a month, along with the seasonal items. They reviewed recent dietary surveys, conferred with other nutrition, health and agricultural personnel, and prepared a list of food items for the diet history questionnaire. The items included both western and traditional foods that covered several food groups, such as starches, breads and spreads, meat, poultry and fish, vegetables, fruits, snacks, beverages, etc. Information on the usual methods of food preparation, including the use of particular fats and oils and coconut cream in food preparation, was also identified.
Developing the Diet History Questionnaire
The diet history listed each item individually. The local names of each food were included in the questionnaire, which was administered by trained nutrition and health education personnel. The format included columns for recording weekly or monthly frequencies and usual serving sizes. To assist participants in nominating the quantities consumed, the island's nutritionists or dietitians developed appropriate visual aids, such as root vegetables preserved with a shellac coating, Polaroid photographs of medium and large servings of vegetables, plastic meat models, and different sizes of familiar bowls and cups.
As another measure of dietary intake, we collected a 24-hour recall of foods consumed the day before the interview. This was done before the diet history, so that participants would become familiar with recalling what they ate and how to use the visual aids for estimating amounts consumed.
Determining the Composition of Foods Consumed in Each Island
On arrival in each island, we visited the various produce markets and village stores to observe firsthand the available food supply. Unfamiliar foods were purchased and identified, and the contents of commercially prepared items were separated and weighed to develop estimated quantities of the ingredients. Recipes of various mixed or traditional dishes were also obtained. In addition, procedures were designed for collecting and preparing the fruits and vegetables for carotenoid analysis. The labels of cereals, rice, flour, breads, crackers, and similar items were scrutinized to identify the ingredients and to determine if the products were enriched or fortified. It was also necessary to investigate the available meat, poultry and fish. Beef was generally frozen and imported, and it was difficult to identify the particular cut and its fat content in the frozen state. The chief nutritionist of each island suggested the probable cut of meat, percentage of fat, and usual method of preparation. Lamb from New Zealand was utilized in the Cook Islands and Fiji. Food composition data on New Zealand lamb were available from the U.S. Department of Agriculture (USDA) (6), whereas data on mutton flaps, consumed by the Cook Islanders, were found in a report by Platt (7). The Fijians consumed both fresh and canned goat. Values for fresh roasted goat were found in USDA (6), but no data were available for canned goat. We compared the taste and appearance of the two products, and based on their similarity, decided to use the same values for both items. Chicken was similar to the stewing chickens of Hawaii, and we estimated the cooked items as about 20 per cent fat (6).
All of the islanders consumed a large variety of reef and ocean fish and shellfish. In general, people described them by their size (small, medium or large) or by their traditional names. The nutritionists recommended that fish be classified according to their estimated fat content. Fresh fish of high fat content, such as salmon, were rarely available. The names of the local fish were then classified according to low or medium fat and were used by the interviewers for coding the reported fish items.
One of our major objectives was to obtain estimates of the carotenoid values of the vegetables and fruits grown on each island. One of us (GB) performed the laboratory analyses of these items. In each area, 15 to 20 highly consumed foods were selected from local markets or home gardens. All foods were prepared as normally consumed within each population. A representative sample of each food was packaged, frozen, and shipped on dry ice to Beltsville, MD, for subsequent analysis. Carotenoids in extracts from each food were separated and quantified by a combination of high performance liquid chromatography and UV-visible spectroscopy (8). The items included dark leafy greens, other green, yellow and red vegetables, and a few yellow and orange fruits. The green vegetables, in particular, included some unfamiliar items, such as hibiscus leaves, amaranthus, wild fern, and drumstick leaves. We located some of these items in various food composition tables (6,9,10). If nutrient data were not available, we compared the items to similar vegetables of the same color and shape and imputed the food composition values. Although these procedures are not error-free, they are acceptable for comparing the diets of various groups of islanders.
Each group of islanders consumed some items unique to their own setting. For example, in French Polynesia, two “Chinese” plate lunches were popular and were listed in the diet history. The first was a mixture of pork, dried white beans, macaroni, green beans, rice and soy sauce, and the second contained chicken, cabbage, noodles, sausage, carrots and soy sauce. The soy sauce was obviously the Oriental component! We purchased the lunches, separated and weighed the ingredients, estimated the amount of soy sauce, and developed approximate “recipes” for the database. Similarly, canned products imported from France, such as “cassoulet”, were purchased, and the kind and amount of each ingredient weighed to develop a “recipe”. This procedure was followed for estimating the contents of various mixed dishes or sandwiches that were eaten frequently. In addition, the diet histories in each area included a number of traditional main dishes, desserts and snacks. The nutritionists, other staff, and family members contributed information, which was used to develop a formula for the composition for each of these mixed dishes.
Creating the Database
Because there were no comprehensive food composition data for the Pacific Islands, we utilized reliable sources of published data whenever possible. Our most frequent resources were the USDA Nutrient Database for Standard Reference (6), McCance and Widdowson's Composition of Foods (10), Food Composition Table for Use in East Asia (9), and an article by Mangels et al. (11). In addition to energy and macro- and micronutrients, the data set includes values for dietary fiber, starch, nonstarch polysaccharides, carotenoids and tocopherols. All values represent foods as commonly consumed including recipes which were calculated from the data for cooked ingredients. No further adjustments were made for potential losses after food preparation. Energy and carbohydrate values were not adjusted when data from different sources were combined.
Some of the items included in the data set may be of interest. The root vegetables presented little problem, because the various sweet potatoes, taro, and breadfruit are also popular among the Polynesians in Hawaii. Assuming that the values of the same tuber would be comparable among the Pacific Islands, we utilized the USDA values (6) for each area. A few “new” root vegetables were consumed, for example, “wild yams” in Fiji. We used the same values as regular yams, but assigned different code numbers so the items could be identified. We also used the same values for plantains and green bananas which are most likely comparable in composition. They differ, of course, in their size, but not in the way they are consumed. We decided to use the values of barracuda (2.6 per cent fat) and Spanish mackerel (6.5 per cent fat) (6) for the low fat and medium fat categories of fish, respectively. With a few exceptions, values for shellfish were generally available. A “new” item was “bêche-de-mer” or sea slugs, and we were fortunate to locate it in the East Asia tables (9). To insure that our values for canned fish were appropriate, we purchased samples to determine the percentage of oil and solids and modified the USDA nutrient data, if warranted.
A few rather exceptional food items were consumed by some of the islanders, such as “roussettes” or flying foxes in New Caledonia. The proximate values were obtained from Cecily Dignan of SPC (personal communication), However, we could not find values for raw and grilled worms (“vers de bancoule”), which were occasionally consumed by Melanesians in New Caledonia. This was one of the very few items not included in the dietary analysis.
The carotenoid values of the analyzed vegetables and fruits were added to the data set. If analytical data were not available for a particular vegetable or fruit, we averaged the laboratory values for the same item from the other islands and utilized the imputed data. Published values from Mangels et al. (11) were selected for fruits and vegetables that were not analyzed. It is of interest to note the variability of the carotenoid contents of the same items from the different islands. Table I shows the variation of carotenoids in Chinese cabbage, taro leaves, leaf lettuce, and hibiscus leaves in three of the islands, whereas Table II presents the difference of carotenoids in pumpkin, tomatoes, and papayas in the four geographic areas. The variations among the islands are probably due to sampling, geographic location, light and soil, and other factors. For the other dietary components of fruits and vegetables, we assigned the same values used for the comparable foods in the Hawaii database (unpublished data). Analysis of the association of dietary risk factors and cancer incidence in the South Pacific Islands is in progress and will be reported within the near future.
Table I. Carotenoid content of selected green vegetables from Pacific Islands (mg/100 g edible portion)a, b
|Leaf lettuce, raw|
a Nutrient Composition Laboratory, USDA, Beltsville, MD
b Green vegetables from New Caledonia were not analyzed (see text for method of imputing the values)
c Chinese cabbage (bok choy) was steamed 3–5 minutes and drained
d Taro leaves were boiled 40 minutes and drained
e Hibiscus leaves were steamed 10 minutes and drained
• Problems and Suggestions
Based on our experience in developing a food composition database for the Pacific Islands, we are aware of the potential problems that may occur in analyzing dietary data from isolated populations. First, it is important to know the local names used for various foods. For example, in Fiji, each item had a Melanesian name and a Hindi name, whereas in French Polynesia, most adults used the Tahitian name rather than the French. Second, although foods may have the same name in different countries, they may differ in food composition. For instance, Chinese cabbage (“bok choy”) was dark green in one island, medium green with white stems in a second, and light green with yellow flowers in a third. These differences probably explain the variation in their carotenoid values. Similarly, in some areas, we observed a difference in the color of a vegetable that was locally grown as compared to the same vegetable that was imported. These items were treated as separate foods according to the local or ethnic names. It may be helpful to photograph unfamiliar vegetables and to match their colors with a set of colored markers. This information, along with laboratory analysis of various antioxidants and appropriate botanical data, may permit reasonable imputations of values for the food composition database.
Table II. Carotenoid content of selected yellow and red vegetables and fruits from Pacific Islands (μg/100 g edible portion)a
|Tomato, raw, whole|
|Papaya, yellow, raw, flesh only|
|Papaya, red, raw, flesh only|
a Nutrient Composition Laboratory, USDA, Beltsville, MD
b Pumpkin was peeled, boiled 30 minutes, and drained
Third, processed foods are likely to be imported from various countries. For example, canned, frozen and packaged products from New Zealand, Australia and France were available in different islands. The labels may suggest that the items are similar to those found in the investigator's native country. However, this cannot be assumed. Items, such as baked beans, canned or frozen mixed vegetables, sausages, etc., need to be checked to identify their approximate contents before selecting published values. Fourth, most islanders used the term “juice” (or the local name) loosely. For instance, concentrated syrups were often diluted with water and called “juice”; if real juice was used, it generally was sweetened with considerable sugar and diluted with water. Observing the preparation of “juice” is recommended, so that the appropriate nutrient values can be assigned.
Fifth, recipes are needed for traditional and ethnic mixed dishes, desserts and snacks. Although island recipes may be printed in tourist publications, it is preferable to ask several local people for their recipes and use this information to develop prototype recipes for the food composition database. Finally, knowledgeable nutritionists and dietitians familiar with the eating patterns of the population are the keys to achieving a realistic database that is area-specific and meaningful for analyzing the dietary intakes of the population.
We are grateful to the following nutritionists for their generous assistance in our surveys: Taiora Matenga Smith, Ministry of Health, Rarotonga, Cook Islands; Mona J. Chand, Ministry of Health, Suva, Fiji; Maeva Barral, Ministry of Health, Papeete, Tahiti; and Dominique Daly, Noumea, New Caledonia. We also thank Cecily Dignan, Nutritionist, South Pacific Commission, for her generous support.
(1) Henderson, B.E., Kolonel, L.N., & Foster F. (1982) Nat. Cancer Inst. Monog. 62, 73–78
(2) Henderson, B.E., Kolonel, L.N., & Dworsky, R. (1986) Nat. Cancer Inst. Monog. 69, 73–81
(3) Willett, W. (1990) Nutritional Epidemiology, Oxford University Press, Oxford
(4) Hankin, J.H. (1991) in Research: Successful Approaches, E.R. Monsen (Ed.), American Dietetic Association, Chicago, IL, pp. 173–194
(5) Block, G., & Hartman, A.M. (1989) in Nutrition and Cancer Prevention. Investigating the Role of Micronutrients, T.E. Moon & M.S. Micozzi (Eds.), M. Dekker, New York, NY, pp. 159–180
(6) US Department of Agriculture (1993) Nutrient Database for Standard Reference, Release 10, USDA, Washington, DC
(7) Platt, B.S. (1980) Tables of Representative Values of Foods Commonly Used in Tropical Countries, Special Report Series No. 302, Medical Research Council, London
(8) Khachik, F., Beecher, G.R., Goli, M.B., & Lusby, W.R. (1991) Pure Appl. Chem. 63, 71–80
(9) Leung, W.T.W., Butrum, R.R., & Chang F.H. (1972) Food Composition Table for Use in East Asia. Nat. Inst. of Arthritis, Metabolism and Digestive Diseases, Bethesda, MD
(10) Holland, B., Welch, A.A., Unwin, I.D., Buss. D.H., Paul, A.A., & Southgate, D.A.T. (1991) McCance and Widdowson's The Composition of Foods, 5th Ed., Royal Society of Chemistry, Cambridge
(11) Mangels, A.R., Holden, J.M., Beecher, G.R., Forman, M.R., & Lanza, E. (1993) J. Am. Diet. Assoc. 93, 284–296