Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

Food science and technology

Carrots and dietary vitamin A adequacy

Lilian Portocarrero, Julieta Quan de Serrano, Louise Canfield, Thomas Tarara, and Noel W. Solomons



Carrots are a leading source e of vitamin A in the Guatemalan diet. Analyses of carrots using high performance liquid chromatography (HPLC') indicated that carrots have more than twice the amount of carotene reported in the Latin American food compositio n tables. The consequence for the recalculation of daily vitamin A intake in a data set of pregnant pert-urban women is a 12% reduction in the number of women consuming <750 RE and a 19% reduction in the number with an intake of <600 RE. To the extent that earlier anayses underestimated the carotene content of' carrots, the estimates of vitamin A intake in Latin America have been variably and falsely low, depending on the contribution of carrots to the individuals reported diet.



Vitamin A nutrition has been the subject of intense interest in Central America since the 1965-1967 nutrition survey [1]. That study reported high rates of inadequate dietary intakes and a substantial prevalence of subadequate circulating retinol levels in the Guatemalan population. The food composition table used to estimate the dietary nutrient intakes [2] has been the standard source of food composition information in the region since 1961.

In the Guatemalan diet, vitamin A is primarily available as provitamin A carotenoids from plants [3]. The carrot (Daucus carota) is a major source of -carotene, the carotenoid with the most efficient biological yield of vitamin A [4-8]. Our studies show carrots to be important in the contemporary Guatemalan diet in both rural and urban areas. An accurate analysis of the total vitamin A content of this root, and specifically its ,-carotene, is essential to the reliable appraisal of the vitamin A adequacy of the diet. The assay results in the Latin American food tables are over three decades old. The present study re-examined the ,0-carotene content of carrots using modern high performance liquid chromatography (HPLC).


Materials and methods

Carrots were purchased in the produce section of a supermarket in Guatemala City and in a comparable setting in Tucson, Arizona, USA. In Guatemala the carrots were prepared in five forms: raw, peeled. boiled for 45 minutes, steamed for 35 minutes, and deep-fried in oil for 1() minutes. All the Guatemalan preparations were frozen at 2() C until transported on dry ice to the Arizona laboratory, where they were stored at -70 C until assayed. The carrots from Arizona were prepared raw and stored under refrigeration until analysed.

All the carrot samples were brought to room temperature and placed in a glass mortar and ground with a glass pestle. The concentration of `3-carotene was determined by modifications of previously published procedures [9]. After grating (raw) or mashing (cooked), the carrots were subjected to extraction with tetrahydrofuran (THF) and a known amount of an external standard (C45B-carotene 3) until the residue was colourless (three or four times). The extract was filtered using a Buchner funnel with a no. 5 Whatman filter, and the filtrate was centrifuged. The supernatant was evaporated to about 10 ml by rotary evaporation (Rotovap), and carotenoids were re-extracted with petroleum ether and water (1:1 v/v) containing 0.5 g NaCI. This step was repeated until the petroleum ether layer was colourless (four or five times). The resulting organic extract was filtered over Na2SO4 using a Buchner funnel fitted with a no. 5 Whatman filter and again evaporated to dryness by rotary evaporation prior to dissolving in THF containing 0.25 g/L butylated hydroxytoluene (BHT). The extracts were flushed with argon and stored at -70 C until analysed.

The extracts were injected automatically into a Beckman C18 Ultrasphere column using an IBM autosampler (IBM no. 9505, IBM Instruments, Danbury, Conn., USA) and eluted isocratically with acetonitrile and THF (85:15 n/n) containing 0.25 g/L BHT at a flow rate of 2.5 ml per minute using a Waters Model 510 pump programmed with a Waters Maxima 820 version 3.02 system controller (Waters Assoc., Milford, Mass., USA). The samples were quantitated from their absorbance at 452 nm using a Milton Roy Programmable Detector (Model SM 4000).

Because of the instability of -carotene at room temperature, the samples were held in the autosampler no longer than two hours. The it-carotene was protected from light during all manipulations, and oxidation was monitored by comparison with authentic standards treated identically. Beta-carotene concentrations were determined by comparison with a standard curve.



The ,-carotene content of the carrots from Guatemala and Arizona and that reported in the literature [7, 10-13] are shown in table 1. All the analysed samples of genetically unmodified carrots had -carotene levels ranging from 3,600 g to 18,250 g per 100 g of edible portion, the convention for food composition tables.



There is indisputable evidence that the vitamin A nutriture in various nations of Latin America is less than adequate. In Guatemala, for example, low circulating levels of retinol are widespread and sporadic cases of xerophthalmia occur. Survey data from 1965 to the present have indicated severe deficits in vitamin A intake. However, the proportion of persons found by the surveys to have subadequate dietary intake far exceeded that of those with biochemical evidence of deficiency. There are four possible explanations for this discrepancy: (1) that circulating retinol is an insensitive marker of nutritional deficiency; (2) that the estimates for dietary requirements are too high; (3) that individual requirements follow a probabilistic distribution in which many persons achieve adequacy below the population recommendations; and (4) that the estimation of dietary intake is falsely low. While each of these explanations may have some merit, our present study provides information relevant to the fourth of these issues, the underestimation of vitamin A intake.

TABLE 1. Beta-carotene content of selected carrots

Carrots -carotene
Guatemalan   present study

present study

raw 120.0
boiled 101.0
steamed 141 0
fried 133.0
Arizona, raw 112.0
"Carrot" 182.5 Khachik &


"Carrot" 151.8 [11]
Danvers variety 141 6 [11]
Chantenay variety 62.7 [11]
"Carrot" 108.3 [10]
Main (supermarket) 36.0 [12]
Main (roadside stand) 40.0 [12]
"Carrot" 72.9 [13]
Beta III 202.5 [7]
HCM 300.0 [7]

a. Personal communication.


The Latin American food composition table [2] has been the authoritative source for dietary survey work in Central and South America since its publication in 1961. Present analytical techniques are far more sensitive and quantitative than those available at that time. For the vitamin C and riboflavin content of selected foods, however, modern techniques applied in the past four years have provided estimates identical with those in the original food tables (CeSSIAM, unpublished data). In the present study the -carotene content of carrots was re-evaluated using HPLC.

Consistent values were obtained for raw, fried, boiled, and steamed carrots from Guatemalan markets, and these were consistent, in turn, with those for carrots purchased in Arizona and with recently published food composition data from various US laboratories. This leads us to the conclusion that the value of 5,540 g, -carotene per 100 g edible portion for raw carrots from the Latin American food composition table underestimates the ,3-carotene content of this root vegetable by 50% and leads to a 46% underestimation of its vitamin A activity. To the ex tent that carrots are an important contributor to the consumption of vitamin A in Guatemala and other Latin American nations, total dietary vitamin A intake is underestimated.

TABLE 2. Foods contributing vitamin A to the Guatemalan diet (selected studies)

Adults (Guajitos), 7-day food- frequency (N= 79) carrots
Adults (Guajitos), 24-hour recall

(N= 54)

Pregnant women, 7-day food

frequency (.V= 102)

Children, different sites (N = 315) mangoes

A dietary intake survey among various age groups in Guatemala, using a list of 39 vitamin A-rich items incorporated into a seven-day food-frequency questionnaire [14], reported that carrots were consumed by 6()% of non-pregnant women, 76% of pregnant women, and 35"/o of preschool children. Carrots have been the major single source of vitamin A reported in most studies (table 2). An erroneous value for ,-carotene applied to these intake histories would have underestimated vitamin A intake.

A recent field survey of vitamin A intake, using the seven-day food-frequency method, was conducted among 52 women from a pert-urban settlement in their third trimester of pregnancy. Each woman was interviewed twice, with a seven-day interval between interviews. The Latin American food composition table [2] was used to calculate the vitamin A intakes. Seventy-six per cent of all intake records included carrot consumption. Our original analysis of the data. based on the 102 interviews, revealed a mean (+SD) vitamin A intake of 623 + 368 RE, a median of 558 RE, and a range of 58-1,918 RE [3] The percentage of vitamin A from preformed retinol was estimated as 28% and that from provitamin A as 72%. The percentages of women consuming less than 100%. 67%, and 50% of the recommended level during pregnancy (750 RE [15] ) were 69%, 36%, and 27% respectively. The percentages consuming less than the 600 RE recommendation [16] were 52%, 31% and 15% respectively (fig. 1).

When the increased amount of -carotene from carrots was factored into the same data set, the mean daily vitamin A level rose by 17%, the median rose by 17%, and the highest individual intake was then 2,413 RE. The percentage of vitamin A from retinol fell to 26%, while that contributed by provitamin A rose to 74%. The percentages of women consuming less than the full 750 or 600 RE recommended intakes fell to 57% and 43% respectively.

The readjustment of the contribution of ,-carotene from carrots to vitamin A intake has a sizeable impact on the estimate for vitamin A intake for the group, and for the 76% of the population that consumed carrots. Assuming that the 1960s technology also underestimated the ,(I-carotene content of other foods - an assumption not yet tested empirically - the estimates of vitamin A intake based on the Latin America food table [2] would have an even greater error. We conclude that prior estimates of vitamin A intake in adults have substantially underestimated the true intake. In young children, for whom vegetables are relatively less important, the underestimation has probably not been as great.

In our experience, retinol concentrations have not shown as many low values as would be predicted by the deficits in intake assessed by food-frequency data. Overly generous requirement levels, the population distribution of requirements, and the insensitivity of plasma retinol as a diagnostic index may explain some of this discrepancy. Clearly, as the carrot is a major contributor to vitamin A intake in the Guatemalan population, and since its nutritive value for vitamin A is underestimated by the Latin American food table [2] a substantial part of the discrepancy can be ascribed to inappropriate food composition data. To the extent that carrots play a similarly important dietary role in the food systems of other countries in which this table is the standard reference, a similar discrepancy can be expected. We believe that improved estimates - with higher vitamin A values - will result from the incorporation of modern HPLC assays of the foods in Latin American diets.

Fig.1. Percentages of women not meeting 100%, 67%, and 50%, of 750 and 600 RE intake levels of vitamin A as calculated using the Latin American food composition table [2]. (Traditionally, 750 RE was recommended level . The most recent FAO/WHO expert panel on vitamins reduced the recommendation ; 600 RE is now recommended for pregnant women [16].)


The position of the carrot in Latin American diets has other implications. New, genetically improved varieties of carrot have recently been developed which have ,carotene concentrations of 202 g/g (Beta III) and 300 g/g (HCM) [7], compared with 32 ,g/g in the Latin American food composition table, although these analyses await confirmation in other laboratories and under other cultivation conditions. The daily vitamin A requirement for a non pregnant adult (500 RE) would be met by less than 15 g of these -carotene-rich carrots. One 80 g carrot of such an improved variety would deliver 5.4 days' worth of the recommended adult intake of vitamin A. Promotion of the genetically altered, -carotene-rich carrots in a population that consumes carrots could help provide an adequate vitamin A intake for the adult population.



1. Interdepartmental Commission on Nutrition in National Defense The Central American nutritional survey. 1965-67: summary report. Washington, DC: Department of Health and Welfare, 1973.

2. Wu-Leung WT, Flores M. Tabla de composicin de alimentos pare uso en America Latina. Guatemala City: INCAP-ICNND, 1961.

3. Solomons NW, Bulux J, Guerrero AM et al. Vitamina A en areas urbanas marginales de la capital de Guatemala. Rev Chil Nutr 1989;17(1):41-45.

4. Erdman J. The physiologic chemistry of carotenes in man. Clin Nutr 1988;7(3):101-05.

5. Henryk D. The chemistry of carotenoids and their importance in food. Clin Nutr 1988;7(3):97-100.

6. Lachance PA. Dietary intake of carotenes and the carotene gap. Clin Nutr 1988;7(3):117-22.

7. Simon PW. Genetic improvement of carrots for meeting human nutrition needs. In: Quebedeaux B, Bliss FA, eds. Proceedings of the First International Symposium on Horticulture and Human Health. Englewood Cliffs, NJ, USA: Prentice-Hall (in cooperation with the American Society for Horticultural Science, Alexandria. Va, USA), 1987:208-13.

8. Simpson KL, Chichester CO. Metabolism and nutritional significance of carotenoids. Ann Rev Nutr 1981;1:351-74.

9. Khachik F. Beecher GR, Whittaker NF. Separation, identification and quantification of the major carotenoid and chlorophyll constituents in extracts of several green vegetables by liquid chromatography. J Agric Food Chem 1986;34:603-16.

10. Simpson KL. Chemical changes in natural food pigments. In: Richardson T, Finley JW, eds. Chemical changes in food during processing. Westport, Conn, USA: AVI Publishing Co., 1986:40941.

11. Sweeney JP, Marsh AC. Effect of processing on provitamin A in vegetables. J Am Diet Assoc 1971;59:23843.

12. Bushway RJ, Yang A, Yamani AM. Comparison of alpha- and beta-carotene content of supermarket versus roadside stand produce. J Food Qual 1986;9:43743.

13. Bushway RJ. Determination of alpha- and beta-carotene in some raw fruits and vegetables by high performance liquid chromatography. J Agri Food Chem 1986;34:409-12.

14. Quan de Serrano J, Gonzalez L. The pattern of contemporary dietary vitamin "A" intake in various geographic regions of Guatemala as determined by 7day food frequency questionnaire. J Am Coll Nutr 1987;6(5):438 (abstr 66).

15. World Health Organization. Handbook on human nutritional requirements. WHO Monograph Series, no. 61. Geneva: WHO, 1974.

16. Food and Agricultural Organization. Requirements of vitamin A, iron, folate, and vitamin B12: report of a joint FAD/WHO expert consultation. FAO Food and Nutrition Series, no. 23. Rome: FAO, 1988.

Contents - Previous - Next