Introduction
The term "carbohydrate" describes a family of contributing compounds, all constructed from the same monosaccharide building blocks, ranging from the simple sugars, or mono and disaccharides, through sugar alcohols, oligosaccharides and dextrins, to the more complex starch and non-starch polysaccharides. With this wide array of compounds, it follows that there are a considerable number of food sources which contribute to the total carbohydrate in the diet.
Trends in food production, availability and consumption
A number of approaches can be used to examine trends in both supply and intake of foods and nutrients, these are:
1. Production
2. Food balance sheets
3. Household surveys
4. Individual assessments
For worldwide comparisons and trends over time, any of the above approaches can be used, and each has both advantages and limitations.
Production
Production is extremely useful in examining trends throughout the world. Production figures are available from FAO for every country in the world for every crop (1). Most countries produce food for their own consumption, and in addition are net food importers. Only a small number of countries are major exporters of carbohydrate foods, primarily grains. These include the United States, Canada, Argentina, Australia, Thailand and Viet Nam.
Crop yield is affected by agricultural practices, weather conditions and external forces. Agricultural practices, such as improved crop varieties, increased land use, increased irrigation of land, or increased use of fertilizers, herbicides, and insecticides can all result in increased yield, and all have contributed to the increases in crop production over the last three decades (see Figure 1).
No matter how good the cultivar or the growth conditions, however, environmental conditions can have an overriding effect on yield. Generally, drought, excess water, cold, hail, and wind account for about 90% of all crop losses (2), with drought by far the major factor. This can clearly be seen with production of cereals in North America in the late 1980's, particularly 1988, when continued lack of rain and minimal snow cover in wheat-growing areas of the United States and Canada had a severe effect on crop production.
External forces, in the form of the organizational structure of the country and its political stability are all factors affecting agricultural production. Wars and political unrest have a major effect on agriculture and can result in marked reductions in crop production. This has been seen in many countries when unrest occurs.
Figure 1 - Major sources of carbohydrate - world production 1961-1994
Food Balance Sheets
Food balance sheets are available for every country in the world, for every food item (1). These are converted to nutrients based on the composition of the individual food commodities included in the production data. Food balance is typically calculated by taking into account the quantity of foodstuffs produced in a country, added to the quantity imported and adjusted for any changes in stocks that have occurred. In order to determine the food available for human consumption, all other food uses must be taken into account. These include the quantity of food exported, fed to livestock, used for seed, and used for non-food use. Even food losses during storage and transportation must be considered.
Food balance sheets, although describing consumption of foods or nutrients per capita of the population for a country, do not represent food actually consumed. They cannot determine waste at a variety of levels, including the home. Food balance sheets in prosperous or developed countries are therefore an overestimate of food actually consumed, while in many developing countries, with little food waste, they are closer to actual consumption or even less than consumption because of home production (3).
Household surveys
Household surveys and individual data, therefore, represent the closest to food actually consumed. With increasing food consumption outside the home, however, household data, such as the continuing National Food Survey in the UK, has recently become much less meaningful and reflective of food actually consumed than in past decades. Like food balance, household surveys may also not be able to evaluate wastage and spoilage, although efforts have been made to take account of wastage in the home (4).
Individual assessments
Individual surveys represent the best way to assess food actually consumed. However, they also have limitations. First, they are not available for all countries. Many countries, particularly in the developing world, do not have the resources to mount large nutrition surveys, or even small studies on specific groups.
In the developed world, many countries do conduct national surveys, and these provide an invaluable source of data for food and nutrient intake. Some countries, like the United States, conduct national surveys on a regular basis as a government requirement, but most do so only sporadically because of the large cost involved. In spite of their infrequent nature, national surveys are the major source of reliable information on actual dietary intake around the world. These are supplemented by small surveys in single locations on smaller numbers of individuals. Methodologies vary from one survey to the next, as well as from country to country. In North America, for example, 24 hour recall is the most common method of dietary assessment and is that used most often whereas the UK has used 7 day weighted assessments. These different methods have differing problems, but a problem with both is some degree of underreporting, both intentional and involuntary (4).
In light of the limitations of the various methods for assessing consumption, emphasis has been placed on production and individual assessment data. Although these also have limitations, they are lesser in magnitude and easier to interpret than the errors in food balance and household surveys. Errors in food balance can also vary more with increasing prosperity, than do production and individual assessments and hence comparisons between the developing and developed countries are more difficult when using food balance data.
Carbohydrate foods
The major carbohydrate-containing foods in the human diet are:
1. Cereals
2. Sweeteners
3. Root crops
4. Pulses
5. Vegetables
6. Fruit
7. Milk products
Cereals
Cereals refer to the graminaceous family, and are considered to be crops harvested for dry grain only. Crops grown for grazing or harvested green for forage or silage are defined as fodder crops. Of the 17 primary cereals defined by FAO, those which make a significant contribution to the human diet are rice, wheat, maize (corn), barley, rye, oats, millet and sorghum. The composition of the major cereals is shown in Table 1. In general, cereals contain 65-75% of their total weight as carbohydrate, 6-12% as protein, and 1-5% fat. Compared to other carbohydrate sources, they are therefore the most nutrient dense. The majority of the carbohydrate is present as starch but cereals are also a major provider of non-starch polysaccharides to the diet and also contain some simple sugars.
TABLE 1 Nutritive value of major cereals/100g edible portion (5-7)
|
Cereal |
Energy kJ |
Moisture % |
Protein g |
Fat g |
CHO* g |
NSP* g |
TDF* g |
Starch g |
Sugars g |
|
Wheat |
1318 |
14.0 |
12.7 |
2.2 |
63.9 |
9.0 |
12.6 |
61.8 |
2.1 |
|
Maize |
1515 |
12.0 |
8.7 |
0.8 |
77.7 |
na |
11.0 |
71 |
1.6 |
|
Rice |
1531 |
11.8 |
6.4 |
0.8 |
80.1 |
2.0 |
3.5 |
80.1 |
1.0 |
|
Barley |
1282 |
11.7 |
10.6 |
2.1 |
64.0 |
14.8 |
17.3 |
62.2 |
1.8 |
|
Sorghum |
1610 |
14.0 |
8.3 |
3.9 |
57.4 |
na |
13.8 |
(50) |
1.3 |
|
Millet |
1481 |
13.3 |
5.8 |
1.7 |
75.4 |
na |
8.5 |
60 |
4 |
|
Rye |
1428 |
15.0 |
8.2 |
2.0 |
75.9 |
11.7 |
14.6 |
75.9 |
na |
|
Oats |
1698 |
8.9 |
12.4 |
8.7 |
72.8 |
6.8 |
10.3 |
72.8 |
1.2 |
* CHO = carbohydrate; NSP = Non-starch polysaccharides; TDF == Total dietary fibre
Of the major cereals, maize, wheat and rice are well ahead of the others in terms of production as each of these three grains contributes over 25% to the world's production of cereals. This contribution to total cereals has remained relatively steady over the last thirty years (1). Although maize emerged as the largest crop in 1994, this was the first year it had done so. In developed countries a large proportion of maize is used as feed for livestock, particularly poultry, pigs, and ruminants (8).
Since the 1960's, rice has been produced in the developing world, with over 90% of total world production in these countries, primarily for local consumption (1). Sorghum, too, is a crop of the developing countries with only 30% produced in the developed world. Wheat was traditionally a crop of the developed world, but has undergone considerable change in the last three decades, so that world production is now virtually split between the developing and developed world (1).
Sugar crops
Sugars are the second largest contributor to carbohydrate in the diet throughout the world. The major sources of sugars are sugar cane and sugar beet. Sugars are also produced from other crops where they are derived from starch. This includes high fructose corn syrup (HFCS) made from corn in the US and sugars produced from potatoes in Japan (9,10). Honey, molasses, maple and others are only minor sugar sources. Sugar cane is the predominant source of sucrose, with a production of over 1.3 million metric tons per year, and being a tropical crop, most of this is in the developing countries. Production of sugar cane continues to increase at a rate of some 2% per year. This is largely due to increased yields, rather than increased crop area (9).
Sugar beet is a temperate crop, and production in some parts of the world, particularly Europe, has declined in recent years with the closure of a large number of sugar factories (10). Nevertheless, many European countries are self-sufficient in sugar and import no cane sugar (10). In other countries, most notably the United States, a reduction in use of sucrose from sugar beet and sugar cane has occurred because of increased use of high fructose corn syrup, which in turn resulted from the high price of sucrose for food manufacturing during the 1980s.
Both sugar cane and sugar beet are 10-20% sucrose, generally around 15-16% (10). Hence although sugar cane is the second largest carbohydrate crop (see Figure 1), when it is compared to cereals which are 60-70% carbohydrate, the contribution of sugar cane to total dietary carbohydrate intake is small (about 10-12% of all carbohydrate produced worldwide).
Root Crops
Root crops, or as often described, roots and tubers, are the third largest carbohydrate food sources, although well behind cereals and sugar cane in total tons produced (see Figure 1). The major contributors to root crops are potatoes, cassava (manioc), yams, sweet potatoes and taro. Minor crops such as jicama, chayote and yambeam are consumed in specific countries. The composition of the major root crops is shown in Table 2. Generally root crops contain 15-30% carbohydrate, 1-2% protein and less than 0.5% fat. Like cereals, the majority of carbohydrate in root crops is starch (70-75% dry weight), but they are also excellent sources of non-starch polysaccharide and contain simple sugars (1-3% dry weight) (12,13). With storage, some of the starch in root crops, like cassava, is converted to sugars, so that after 6 weeks, sugars comprise 12-13% dry weight, rather than 1-3% when first harvested (12,13).
TABLE 2 Nutritive value of root crops/100g edible portion (11)
|
Root crop |
Energy kJ |
Moisture % |
Protein g |
Fat g |
CHO* g |
NSP* g |
TDF* g |
Starch g |
Sugars g |
|
Potato |
318 |
79.0 |
2.1 |
0.2 |
17.2 |
1.3 |
1.8 |
16.6 |
0.6 |
|
Cassava |
607 |
64.5 |
0.6 |
0.2 |
36.8 |
1.7 |
na |
35.3 |
1.5 |
|
Sweet |
372 |
73.7 |
1.2 |
0.3 |
21.3 |
2.4 |
3.0 |
15.6 |
5.7 |
|
Potato (yellow) | |||||||||
|
Yam |
488 |
67.2 |
3.0 |
0.3 |
28.2 |
1.3 |
3.3 |
27.5 |
0.7 |
|
Taro |
451 |
68.3 |
1.4 |
0.2 |
26.2 |
2.4 |
2.9 |
25.1 |
1.1 |
* CHO = carbohydrate; NSP = Non-starch polysaccharides; TDF = Total dietary fibre
World production of the major root crops indicates that potatoes are by far the most important component. Although the proportion of total root crops has been falling in the last 30 years, potatoes still represent nearly half all such crops consumed (1). Potatoes are consumed in large amounts in North America, Eastern and Western Europe, Latin America, and in some countries in Asia, like South Korea, and Turkey (14).
Pulses, vegetables, fruit, other sources
Pulses, vegetables, fruit, and crops grown mainly for oil (such as groundnuts and soybeans), represent other crops which provide carbohydrate for the diet. In quantitative terms, vegetables and fruit have much greater production figures than pulses, but their carbohydrate content is considerably less, except for bananas and plantains (vegetables: 5-8% carbohydrate, fruit: 8-15%) (15,16). Production of pulses is only a fraction of cereals and root crops, but pulses are very high in carbohydrate (50-60% on a dry weight basis), and are a major contributor to carbohydrate intake in some countries. The composition of those crops having a high carbohydrate content are shown in Table 3.
TABLE 3 Nutritive value of selected pulses, nuts, bananas and plantains/100g edible portion (11,17,18)
|
Root crop |
Energy |
Moisture |
Protein |
Fat |
CHO * |
NSP * |
TDF * |
Starch g |
Sugars |
|
Banana |
403 |
75.1 |
1.2 |
0.3 |
23.2 |
1.1 |
1.6 |
2.3 |
20.9 |
|
Plantain |
476 |
68.2 |
0.9 |
0.2 |
29.3 |
na |
na |
na |
na |
|
Dry peas |
1288 |
13.3 |
21.6 |
2.4 |
52.0 |
4.7 |
na |
47.6 |
2.4 |
|
Dry bean † |
1218 |
11.3 |
21.4 |
1.6 |
49.7 |
17.0 |
40.0 |
42.7 |
2.8 |
|
Chickpea (Garbanzo bean) |
1355 |
10.0 |
21.3 |
5.4 |
49.6 |
10.7 |
na |
43.8 |
2.6 |
|
Dry broad bean |
1041 |
11.0 |
26.1 |
2.1 |
32.5 |
6.1 |
19.0 |
24.4 |
5.9 |
|
Lentils (green) |
1264 |
10.8 |
24.3 |
1.9 |
48.8 |
8.9 |
na |
44.5 |
1.2 |
|
Lentils (red) |
1353 |
11.1 |
23.8 |
1.3 |
56.3 |
4.9 |
na |
50.8 |
2.4 |
|
Soybean |
1551 |
8.5 |
35.9 |
18.6 |
15.8 |
15.7 |
na |
4.8 |
5.5 |
|
Groundnut (Peanut) |
2341 |
6.3 |
25.6 |
46.1 |
12.5 |
6.2 |
8.0 |
6.3 |
6.2 |
* CHO = carbohydrate; NSP = Non-starch polysaccharides; TDF = Total dietary fibre† Dry bean, haricot, common bean, kidney bean, navy bean, snap bean, pinto bean - analysed as haricot bean.
Starchy staples
Cereals and root crops constitute the 'starchy staples' in the human diet and as such are the primary source of dietary carbohydrate throughout the world. The contribution of these crops to the diet varies from continent to continent, and even among countries within continents. The consumption of starchy staples throughout the world has changed over the last three decades. Figure 2 shows the world picture of crop distribution, with countries highlighted according to the starchy staple providing the highest proportion of energy (1986-88) (19). As can be seen, large parts of the world depend on wheat and rice, with more selected areas dependent on maize, root crops, sorghum and millet.
Figure 2 - Starchy staples providing the highest proportion of food energy, 1990-1992
Source: Reproduced with permission from the Geographical Association (19)
The proportion of total energy derived from these same starchy staples around the world are shown in Figure 3 (20). Asia and Africa stand out, with 70% or more of total calories from these major crops. It can clearly be seen why there is a need for production on a large scale in these developing parts of the world.
Food Balance Sheets
Food balance information is of limited usefulness in examining real trends in consumption of foods and nutrients because of changes with growing prosperity. Food balance data for carbohydrates may show little change, for example, while individual assessments over the same time show changing carbohydrate intakes, which mirror changes in fat intake (21,22). One author has indicated that as income increases, consumption of products rich in starch decreases. In developed regions, a rapid fall in the consumption of cereal products, especially bread, is clear as income increases (23).
Developed countries, such as the US and UK, have a much lower per capita carbohydrate consumption than countries in Asia, Africa, and South America. While there may be small shifts in these latter countries, carbohydrate does not appear to be declining to the levels consumed in the developed world. It is important to determine from individual assessments if this is truly the case or if the unmeasurables in food balance data are masking the real picture.
Figure 3 - Energy from the dominant starch staples, 1990-1992
Source: Reproduced with permission from the Association of American Geographers (20)
While food balance data may be subject to considerable error in terms of absolute consumption, it can give a reasonable picture of the distribution of sources contributing to the overall intake of a nutrient. This is shown for carbohydrate in Table 4, for the developing and developed countries, and for the entire world. It can be clearly seen that cereals make the greatest contribution to carbohydrate intake throughout the world, with sweeteners in second place followed by other carbohydrate foods. Overall the total energy provided by carbohydrate to the diet is similar throughout the world; however, the percent energy represented by carbohydrate is considerably lower in the developed countries, because of higher intakes of protein and fat.
Individual Assessments
Individual assessments would seem to provide the most accurate way of determining intake of foods and nutrients. As already mentioned, however, this data is not available on a regular basis for most countries, and for some, hardly at all. In addition to this irregular availability, much of the difficulty in comparing carbohydrate intake around the world is that in many countries carbohydrate is calculated in foods "by difference". Carbohydrate values determined this way are obtained by subtracting the weight of the other major components in a food, namely moisture, fat, protein, and ash, from the total weight of a food. What remains is said to be carbohydrate. There are several problems with this method. Firstly, it does not represent direct analysis of carbohydrate, and means that any errors in the estimation of the other nutrients will result in an error in the carbohydrate value. Secondly, there is no description of the individual carbohydrates and with growing interest in various types of carbohydrate (for example, sugars, oligosaccharides or starch), the limitation of a single carbohydrate figure is obvious. Thirdly, "carbohydrate by difference" figures include "unavailable carbohydrate" (dietary fibre) which has a different energy value and different physiological properties than the digestible components of carbohydrate.
TABLE 4 Energy from carbohydrate by food group and as a % of total carbohydrate. Food balance data -1964 and 1994
|
Developed countries |
Developing countries |
World |
||||||||||
|
....1964.... |
....1994.... |
....1964.... |
....1994.... |
....1964.... |
....1994.... |
|||||||
|
kcal/ |
% |
kcal/ |
% |
kcal/ |
% |
kcal/ |
% |
kcal/ |
% |
kcal/ |
% |
|
|
Cereals |
1053 |
59.0 |
866 |
54.2 |
1069 |
71.3 |
1284 |
73.3 |
1029 |
66.2 |
1189 |
69.2 |
|
Root Crop |
156 |
8.7 |
121 |
7.6 |
157 |
10.5 |
128 |
7.3 |
156 |
10.0 |
126 |
7.3 |
|
Sweetener |
369 |
20.7 |
397 |
24.8 |
124 |
8.3 |
183 |
10.5 |
200 |
12.9 |
232 |
13.5 |
|
Pulses |
26 |
1.5 |
17 |
1.1 |
72 |
4.8 |
48 |
2.7 |
58 |
3.7 |
41 |
2.4 |
|
Vegetable |
40 |
2.2 |
47 |
2.9 |
22 |
1.5 |
30 |
1.7 |
28 |
1.8 |
34 |
2.0 |
|
Fruit |
71 |
4.0 |
85 |
5.3 |
42 |
2.8 |
58 |
3.3 |
52 |
3.3 |
65 |
3.8 |
|
Milk |
70 |
3.9 |
65 |
4.1 |
14 |
0.9 |
20 |
1.1 |
32 |
2.1 |
30 |
1.7 |
|
Total CHO |
1785 |
100 |
1598 |
|
1500 |
100 |
1751 |
100 |
1555 |
100 |
1717 |
100 |
|
Total food |
3040 |
|
3206 |
|
2030 |
|
2573 |
|
2344 |
|
2718 |
|
|
% CHO |
58.7 |
|
49.8 |
|
73.8 |
|
68.1 |
|
66.3 |
|
63.2 |
|
* kcal/cap/d = kilocalories per person per day; CHO = Carbohydrate
Increasing evidence suggests that the diets of Western countries should be more like those in the developing countries, and many countries have recommended that carbohydrate should represent at least 50-55% energy (24-27). The question arises - has carbohydrate intake always been lower in developed countries than developing, or has this occurred only relatively recently? Examination of individual assessments from countries where numerous individual assessments are conducted can help answer this question. The clearest picture is seen for the UK. Detailed dietary assessments have been conducted in the UK since the turn of the century. Carbohydrate intake represented over 60% energy in the 1920's, and apart from a plateau and even a rise, due to rationing during and after the Second World War, carbohydrate intake continued to decrease in the UK, reaching a low of 45-46% energy in the 1950's, 60's and 70's (21,22). With interest in dietary fibre in the 1970's and more recently a renewed interest in starch, intakes of carbohydrate in the UK in the late 1980's, early 1990's have risen to about 48% energy (21,22). By comparison, carbohydrate data from food balance figures for the UK have been stable at 50-51% energy, and do not show the downward and upward trends which have occurred.
Trends and comparisons of carbohydrate components
In spite of the problems with terminology and differences in usage in different parts of the world, it is possible to collate some studies to gain a picture of what is happening to intake of components of carbohydrate over time, and from country to country. The lack of information in published work emphasizes the need for researchers to publish dietary information from surveys in more detail, and not simply to publish a paper which contains only nutrients of interest at the time of publication. Those studies which have a lasting impact are those where information has been provided on many nutrients.
Sugars
Sucrose data obtained from Canada and the United States suggest that sucrose intake is declining in North America. As indicated earlier, however, the increased use of high fructose corn syrup as a sweetener is not included in US data for sucrose, and therefore, data on total sugars is what is required to gain a clearer picture of the situation. For the United Kingdom, where sugars are expressed as all free sugars in the diet (as percent energy), the trend has been in an upward direction in recent years. Again, there is a shortage of data for long term trends, but the recent trend is clear and the UK data has the advantage of consistency in measurement, since only one food composition table is used throughout the country.
Clearly more data is needed, since the recent upward trend in the UK in sugars as percent energy, may reflect an increased intake, or it may simply reflect a decreased intake in fat and total energy. Absolute intake in sugars and as a percent energy are identical for the UK and Australia. Australians, however, consume considerably more of their sugars as fruit than in the UK, where sweetened baked goods, sugar itself and sugar products contribute much more to the diet.
Starch
Very few studies report starch intakes, and there are rather too few to consider trends reliably; the upward trend for North America would suggest that intake is heading in a desirable direction. As indicated earlier, however, this may reflect no change in absolute intake, but only an increasing proportion in relation to decreased total energy. Intake of starch is very similar in the UK and Australia, and as would be expected, cereals, particularly breads and breakfast cereals, and potatoes make the major contribution to starch intakes in these western diets. As would be expected, starch intakes are considerably higher in countries dependent on starchy staples to a greater degree, such as China, and Japan. There is a need for more individual studies to solidify starch intake estimates.
Dietary Fibre
Dietary fibre intake data from around the world are difficult to compare because of methodology differences. One author derived non-starch polysaccharides for a number of countries and concluded that these did not vary as much as did intakes of starch (28). In fact, the variations between countries may well be due mainly to method differences, rather than to actual variation in intake. It cannot be stressed enough that uniformity in measurement is required before country comparisons of any nutrient intake is possible.
Some discrepancies in reported dietary fibre intake between different countries may be explained by what is included or not included under "cereals". It is clear that cereals and vegetables are the predominant sources of fibre in the diet, and within these categories, bread and potatoes stand out as major contributors.
Resistant Starch
A new component of interest is resistant starch (RS), which is considered to be that starch which resists digestion and absorption in the small intestine and escapes into the colon (29). There have been recent efforts to assess the amount of RS in diets in Europe and Australia. Some of these have only estimated RS3 (retrograded amylose), which is one of 3 or 4 components of resistant starch in the diet (29,30). Cereals are the greatest source of resistant starch, at 35% or more (30). Resistant starch intakes were calculated for a number of countries in Europe, with an average intake of 4.2 g/d, ranging from 3.2 g/d for Norway to 5.7 g/d for Spain. Of this, cereals accounted for 42% of the RS and potatoes 27% (30). Estimates from some Asian countries, however, are much higher in total, and represent closer to 5% of all starch eaten (30). More analysis is needed of those countries with high intakes of starchy staples, particularly if these are cooked and cooled as in traditional eating practices.
Conclusion
There is growing interest in carbohydrate and its components throughout the world, and recommendations have been made to encourage increased carbohydrate consumption. Assessment of components of particular interest, whether these be sugars, starch, resistant starch or dietary fibre, are hampered by lack of data. This is mainly because of long-standing approaches to carbohydrate measurement, which are indirect and hence inadequate. The reason that this has not been rectified in recent decades is probably due to a lack of recognition of the importance of individual carbohydrate components to health, a situation that continues to this day in some countries. Where interest has grown, as with dietary fibre, a number of different methodologies have emerged which measure different things and therefore give different answers. Comparison of countries is then almost as difficult as if there were no data at all. Comparison of intakes, trends over time, and projections for the future, require uniform methods and good reporting of data. The health implications for populations of physiological findings on small groups, can only be determined if there is knowledge of present consumption. All efforts should be made to ensure that data is collected and is reported in as comprehensive and consistent a manner as possible.
