Symposium
Methods for Assessment BARBARA
of Vitamin A Status1-2
A. UNDERWOOD
Office of International Program Actiuities, National Eye Institute, National Institutes of Health, Bethesda, MD 20892
CURRENTLY USED INDICATORS ABSTRACT Assessment of the relative level of vita min A nutritine in human populations when clinical signs and symptoms of deficiency or of toxicity are absent has presented methodological difficulties. Com monly used indicators include dietary intakes of the vitamin, serum levels, and dark adaptation, all of which have limitations in their precision, especially when ap plied to individuals and to young children. New and developing nonclinical indicators include the relative dose response (RDR) test or a modification (MRDR) of it, conjunctiva! impression cytology (CIC), and isotope dilution to estimate total body reserves. These meth ods, all of which are promising, require additional work to verify their sensitivity, specificity, and predictive power as indicators of relative states of vitamin A de pletion for individuals and populations under the many varied conditions commonly associated with an inad equate vitamin status. Currently, the most reliable as sessment of vitamin A nutriture is likely when a combination of methods is used. J. /Vi/ir. 120:14591463, 1990.
Dietary intake methodologies. In theory, a di etary history should provide evidence concerning vi tamin A status, assuming that it is obtained at least semiquantitatively, that it is representative of the usual pattern of food intake over an extended time period, and that the actual intake requirement for the vitamin for an individual or population is known. The difficulty in obtaining representative quantitative his torical information for vitamin A intake is well doc umented for surveys among populations whose intake is from a wide variety of foods, a substantial portion of which comes from preformed sources. For some purposes, however, only a semiquantitative dietary approach based on frequency of intake may be required to place groups into risk categories for inadequate vi tamin A status (2). Among children in the developing world, where 80-90% of the intake is from carotenoids contained in a limited number of food groups, the International Vitamin A Consultative Group (IVACG®)recently has produced guidelines that claim to categorize populations into "at risk" groups for vi
INDEXING KEY WORDS:
•vitamin A •assessment
•methods of assessment
tamin A inadequacy (3). The guidelines are based on establishing a food-composition table adapted to lo cally available foods and portion sizes that are fed to
Vitamin A nutriture exists as a continuum between clinically evident deficiency and toxicity. The place on the continuum for most nondiseased individuals is determined by their habitual quantitative dietary intake of the vitamin. When this intake has been no tably deficient or excessive, the clinical signs and symptoms, and the corresponding biochemical alter ations in blood, can be easily described (1). However, at intermediate levels of habitual intake, we do not have satisfactory methods for determining transitional ranges—marginal, adequate, or excessive—along the vitamin A status continuum. Blood levels are not sen sitive reflectors of intermediate status, and the vitamin is not excreted in urine except as a variety of metab olites whose concentrations are not parallel to body stores. Liver stores are reflective, but most people will not provide tissue biopsy specimens for non-diseaserelated assessment purposes. 0022-3166/90
$3.00 ©1990 American Institute of Nutrition.
1 Presented as part of a conference, "Nutrition Monitoring and Nutrition Status Assessment," at the first fall meeting of the Amer ican Institute of Nutrition, Charleston, South Carolina, December 8-10, 1989. The conference was supported in part by cooperative agreement HPU880004-02-1 with the DHHS Office of Disease Pre vention and Health Promotion, the USDA Human Nutrition In formation Service, the DHHS National Center for Health Statistics, and the International Life Sciences Institute-Nutrition Foundation. 2 The Planning Committee for the meeting consisted of Drs. He len A. Guthrie, Roy J. Martin, Linda D. Meyers, James A. Olson, Catherine E. Woteki, and Richard G. Allison (ex officio). The sym posium papers were edited by a committee consisting of Dr. James Allen Olson [coordinator), Dept. of Biochemistry & Biophysics, Iowa State University, Ames, IA; Dr. Cathy C. Campbell, Division of Nutritional Sciences, Cornell University, Ithaca, NY; Dr. Roy J. Martin, Dept. of Foods & Nutrition, University of Georgia, Athens, GA; and Dr. Catherine E. Woteki, Food & Nutrition Board, National Academy of Sciences, Washington, DC.
Received 10 December 1989. Accepted 11 July 1990.
1459 Downloaded from https://academic.oup.com/jn/article-abstract/120/suppl_11/1459/4738619 by guest on 12 January 2018
1460
UNDERWOOD
children. This semiquantitative approach has been tried in several countries and found useful as a tool for both categorization of populations (not individuals) at risk and nutrition education. Once the local foodcomposition table and frequency questionnaire have been developed, it is easily used by nonprofessionals. It requires additional validation against other indices of marginal vitamin A status. Serum levels. Serum levels of vitamin A are the most commonly used biochemical measure of vitamin A status, even though it is well known that these are regulated by highly individual physiological controls that are only partially understood (1). Serum levels be come predictive of an individual's status only when body reserves have been critically depleted or over filled. On the other hand, serum distribution curves based on random sampling can be useful in comparing the status of populations relative to the probability of occurrence of clinical vitamin A deficiency and mal nutrition, an inadequate dietary intake, and/or other socioeconomic and ecologie characteristics (4). When interpreting serum values from populations, it is best to use distribution curves that focus on the lower end of the distribution spectrum. Traditionally, values < 20 ¿ig/100mL (0.70 /imol/L) have been considered low, and values < 10 Mg/100 mL (0.35 /nnol/L) have been considered deficient (4). When 15% of the pop ulation is in the low range or 5% is in the deficient range, a public health problem exists. However, these criteria are not based on well-described, randomly se lected reference populations. Recent information from large surveys conducted in the U.S. provides a better database for interpretation (5). The interpretive guidelines based on the U.S. data suggest that values in the range of 20-29 Mg/100 mL (0.70-1.05 /anol/L) for a population may be inadequate for some individ uals within that population, particularly for post-ad olescence age groups. However, the appropriateness of these guidelines and their interpretation in terms of vitamin A inadequacy for populations living in conditions of poverty require confirmation. We need age- and sex-specific reference curves established for populations living under deprivation conditions but known to have adequate vitamin A status. There is even less information available to interpret the upper end of the distribution curve in terms of relative vi tamin A status. It should be noted that two representative serum levels obtained before and after supplementation with vitamin A are much more descriptive of vitamin A status than a single value. A serum response to a sup plement indicates whether the individual (or popula tion) is below his homeostatic level. This is the basis for the RDR test (see below) performed on individuals over a 5-hour interval and is the approach used to evaluate the sugar fortification intervention program in Guatemala (4). The response of a randomly sampled population before and 30 days following administra Downloaded from https://academic.oup.com/jn/article-abstract/120/suppl_11/1459/4738619 by guest on 12 January 2018
tion of a high-dose supplement (200,000 IU) has been suggested as an appropriate assessment approach for determining the prevalence of marginally inadequate to inadequate vitamin A nutriture in the absence of clinical eye signs (6). Abnormal dark adaptation. Rhodopsin is gen erated when the protein opsin in the rods of the retina combines with a cis-isomer of retinol. The complex is split in response to light, yielding opsin and a tiansisomer of retinol and generating the visual-response signal. After isomerization of the trans- to the cis-iso mer, recycling of the process occurs at a rate that nor mally maintains the level of rhodopsin so that visual accommodation is not prolonged when going from brightly to dimly lighted conditions. When the supply of vitamin A is limited, the rate of regeneration of rhodopsin after bleaching by exposure to bright light is impaired. In practical terms, individuals who are deficient in vitamin A have difficulty seeing in the dark, i.e., they are "night blind." In clinical settings psychophysical measurements (dark-adaptation threshold) and electrophysiological measurements (electroretinograms, ERGs) are avail able to measure directly the level of rhodopsin and its rate of regeneration (7). Modified instruments and procedures have been used in field situations among older children and adults who are able to respond ac tively to visual acuity tests subsequent to bright light (bleaching) exposure (8-10). Currently a field-adapted instrument is being tested in Northeast Thailand (11). These tests, however, have not been reliably applied among preschool-aged and younger children from de prived environments, where vitamin A deficiency is most likely to occur. In the latter situation, a history of night blindness elicited from the child's mother or responsible adult has been the most successful approach (12). This is true, however, only in cultures where a specific word or words exist that characterize the condition. Such language usually, but not always, exists in areas with endemic vitamin A inadequacy. Investigators must carefully determine under each local situation whether there is an appropriate term that can be reliably used in an historical interview. Where local terms are not used, alternate strategies to test dark adaptation indirectly may be useful. For example, a child taken from the bright outdoor sun into a darkened room may be asked to find a relative or to locate a favorite toy placed at a specific distance. The time required to achieve this task relative to that needed for a similarly aged child known to have nor mal dark adaptation is compared.
NEW AND DEVELOPING INDICATORS Functional biochemical measures are an alternative and more dynamic means of describing marginal vi-
ASSESSMENT OF VITAMIN
tamin A status applicable to individuals as well as populations. Two functional tests have recently been advocated: the relative dose-response (RDR) test, based on blood sampling, and conjunctival impression cy tology (CIC), based on histological techniques. The RDR measures vitamin A directly in blood, whereas CIC evaluates microscopically the number of goblet cells present and the morphological characteristics of epithelial cells. RDR test. As liver reserves of vitamin A become progressively depleted due to a chronically inadequate dietary supply, conservation mechanisms are invoked to increase the efficiency of vitamin A utilization among tissues and to maintain the level that is cir culating to the target tissues. At a point along the downward continuum, when the liver reserve is de pleted below a critical threshold, the rate of release of the remaining reserve is diminished. Synthesis of the carrier protein retinol binding protein (RBP)continues and results in the accumulation of a pool of preformed RBP. Providing an exogenous source of vitamin A causes the release of holo-RBP at a level and in a char acteristic time course relative to the amount of accu mulated preformed carrier protein (1, 13). After a fasting blood sample is obtained, a liquid solution containing about 450-1000 ¿tg (1.6-3.5 /miol) retinyl palmitate in an oily solution or aqueous dis persion is given orally (14). The exact amount given should be such that it does not overload the system with retinyl esters. However, it must be sufficient to assure complete mobilization of preformed accumu lated hepatic RBP. In some cases where less-efficient absorption might be suspected, e.g., chronic intestinal parasitism, a higher dose in the recommended range is advisable. If high-performance liquid chromatography (HPLC) is the analytical technique, retinol is distinguished from esters and overdosing is not a problem. A small meal containing foods with some fat but minimal vitamin A is given after dosing. After five hours, a second blood sample is obtained. The blood is analyzed for vitamin A by a suitable analytical technique, and the RDR is calculated as fol lows: .A vit A5 where vit A5 = serum vitamin A level five hours after dosing, and vit AQ = fasting serum vitamin A level. An RDR > 20% is considered to be positive and in dicative of inadequate hepatic stores of vitamin A, i.e., marginal vitamin A status (14). Recently, a modification of the procedure, which uses dehydroretinol (vitamin) and requires a single blood sample at 5 hours post dosing, has been tested in rats (15) and is being evaluated in children suspected of being vitamin A deficient (16). This procedure is termed the modified relative dose response (MRDR) test. Downloaded from https://academic.oup.com/jn/article-abstract/120/suppl_11/1459/4738619 by guest on 12 January 2018
A STATUS
1461
Currently available validation data indicate that a positive RDR occurs when hepatic stores are
Methods for Assessment BARBARA
of Vitamin A Status1-2
A. UNDERWOOD
Office of International Program Actiuities, National Eye Institute, National Institutes of Health, Bethesda, MD 20892
CURRENTLY USED INDICATORS ABSTRACT Assessment of the relative level of vita min A nutritine in human populations when clinical signs and symptoms of deficiency or of toxicity are absent has presented methodological difficulties. Com monly used indicators include dietary intakes of the vitamin, serum levels, and dark adaptation, all of which have limitations in their precision, especially when ap plied to individuals and to young children. New and developing nonclinical indicators include the relative dose response (RDR) test or a modification (MRDR) of it, conjunctiva! impression cytology (CIC), and isotope dilution to estimate total body reserves. These meth ods, all of which are promising, require additional work to verify their sensitivity, specificity, and predictive power as indicators of relative states of vitamin A de pletion for individuals and populations under the many varied conditions commonly associated with an inad equate vitamin status. Currently, the most reliable as sessment of vitamin A nutriture is likely when a combination of methods is used. J. /Vi/ir. 120:14591463, 1990.
Dietary intake methodologies. In theory, a di etary history should provide evidence concerning vi tamin A status, assuming that it is obtained at least semiquantitatively, that it is representative of the usual pattern of food intake over an extended time period, and that the actual intake requirement for the vitamin for an individual or population is known. The difficulty in obtaining representative quantitative his torical information for vitamin A intake is well doc umented for surveys among populations whose intake is from a wide variety of foods, a substantial portion of which comes from preformed sources. For some purposes, however, only a semiquantitative dietary approach based on frequency of intake may be required to place groups into risk categories for inadequate vi tamin A status (2). Among children in the developing world, where 80-90% of the intake is from carotenoids contained in a limited number of food groups, the International Vitamin A Consultative Group (IVACG®)recently has produced guidelines that claim to categorize populations into "at risk" groups for vi
INDEXING KEY WORDS:
•vitamin A •assessment
•methods of assessment
tamin A inadequacy (3). The guidelines are based on establishing a food-composition table adapted to lo cally available foods and portion sizes that are fed to
Vitamin A nutriture exists as a continuum between clinically evident deficiency and toxicity. The place on the continuum for most nondiseased individuals is determined by their habitual quantitative dietary intake of the vitamin. When this intake has been no tably deficient or excessive, the clinical signs and symptoms, and the corresponding biochemical alter ations in blood, can be easily described (1). However, at intermediate levels of habitual intake, we do not have satisfactory methods for determining transitional ranges—marginal, adequate, or excessive—along the vitamin A status continuum. Blood levels are not sen sitive reflectors of intermediate status, and the vitamin is not excreted in urine except as a variety of metab olites whose concentrations are not parallel to body stores. Liver stores are reflective, but most people will not provide tissue biopsy specimens for non-diseaserelated assessment purposes. 0022-3166/90
$3.00 ©1990 American Institute of Nutrition.
1 Presented as part of a conference, "Nutrition Monitoring and Nutrition Status Assessment," at the first fall meeting of the Amer ican Institute of Nutrition, Charleston, South Carolina, December 8-10, 1989. The conference was supported in part by cooperative agreement HPU880004-02-1 with the DHHS Office of Disease Pre vention and Health Promotion, the USDA Human Nutrition In formation Service, the DHHS National Center for Health Statistics, and the International Life Sciences Institute-Nutrition Foundation. 2 The Planning Committee for the meeting consisted of Drs. He len A. Guthrie, Roy J. Martin, Linda D. Meyers, James A. Olson, Catherine E. Woteki, and Richard G. Allison (ex officio). The sym posium papers were edited by a committee consisting of Dr. James Allen Olson [coordinator), Dept. of Biochemistry & Biophysics, Iowa State University, Ames, IA; Dr. Cathy C. Campbell, Division of Nutritional Sciences, Cornell University, Ithaca, NY; Dr. Roy J. Martin, Dept. of Foods & Nutrition, University of Georgia, Athens, GA; and Dr. Catherine E. Woteki, Food & Nutrition Board, National Academy of Sciences, Washington, DC.
Received 10 December 1989. Accepted 11 July 1990.
1459 Downloaded from https://academic.oup.com/jn/article-abstract/120/suppl_11/1459/4738619 by guest on 12 January 2018
1460
UNDERWOOD
children. This semiquantitative approach has been tried in several countries and found useful as a tool for both categorization of populations (not individuals) at risk and nutrition education. Once the local foodcomposition table and frequency questionnaire have been developed, it is easily used by nonprofessionals. It requires additional validation against other indices of marginal vitamin A status. Serum levels. Serum levels of vitamin A are the most commonly used biochemical measure of vitamin A status, even though it is well known that these are regulated by highly individual physiological controls that are only partially understood (1). Serum levels be come predictive of an individual's status only when body reserves have been critically depleted or over filled. On the other hand, serum distribution curves based on random sampling can be useful in comparing the status of populations relative to the probability of occurrence of clinical vitamin A deficiency and mal nutrition, an inadequate dietary intake, and/or other socioeconomic and ecologie characteristics (4). When interpreting serum values from populations, it is best to use distribution curves that focus on the lower end of the distribution spectrum. Traditionally, values < 20 ¿ig/100mL (0.70 /imol/L) have been considered low, and values < 10 Mg/100 mL (0.35 /nnol/L) have been considered deficient (4). When 15% of the pop ulation is in the low range or 5% is in the deficient range, a public health problem exists. However, these criteria are not based on well-described, randomly se lected reference populations. Recent information from large surveys conducted in the U.S. provides a better database for interpretation (5). The interpretive guidelines based on the U.S. data suggest that values in the range of 20-29 Mg/100 mL (0.70-1.05 /anol/L) for a population may be inadequate for some individ uals within that population, particularly for post-ad olescence age groups. However, the appropriateness of these guidelines and their interpretation in terms of vitamin A inadequacy for populations living in conditions of poverty require confirmation. We need age- and sex-specific reference curves established for populations living under deprivation conditions but known to have adequate vitamin A status. There is even less information available to interpret the upper end of the distribution curve in terms of relative vi tamin A status. It should be noted that two representative serum levels obtained before and after supplementation with vitamin A are much more descriptive of vitamin A status than a single value. A serum response to a sup plement indicates whether the individual (or popula tion) is below his homeostatic level. This is the basis for the RDR test (see below) performed on individuals over a 5-hour interval and is the approach used to evaluate the sugar fortification intervention program in Guatemala (4). The response of a randomly sampled population before and 30 days following administra Downloaded from https://academic.oup.com/jn/article-abstract/120/suppl_11/1459/4738619 by guest on 12 January 2018
tion of a high-dose supplement (200,000 IU) has been suggested as an appropriate assessment approach for determining the prevalence of marginally inadequate to inadequate vitamin A nutriture in the absence of clinical eye signs (6). Abnormal dark adaptation. Rhodopsin is gen erated when the protein opsin in the rods of the retina combines with a cis-isomer of retinol. The complex is split in response to light, yielding opsin and a tiansisomer of retinol and generating the visual-response signal. After isomerization of the trans- to the cis-iso mer, recycling of the process occurs at a rate that nor mally maintains the level of rhodopsin so that visual accommodation is not prolonged when going from brightly to dimly lighted conditions. When the supply of vitamin A is limited, the rate of regeneration of rhodopsin after bleaching by exposure to bright light is impaired. In practical terms, individuals who are deficient in vitamin A have difficulty seeing in the dark, i.e., they are "night blind." In clinical settings psychophysical measurements (dark-adaptation threshold) and electrophysiological measurements (electroretinograms, ERGs) are avail able to measure directly the level of rhodopsin and its rate of regeneration (7). Modified instruments and procedures have been used in field situations among older children and adults who are able to respond ac tively to visual acuity tests subsequent to bright light (bleaching) exposure (8-10). Currently a field-adapted instrument is being tested in Northeast Thailand (11). These tests, however, have not been reliably applied among preschool-aged and younger children from de prived environments, where vitamin A deficiency is most likely to occur. In the latter situation, a history of night blindness elicited from the child's mother or responsible adult has been the most successful approach (12). This is true, however, only in cultures where a specific word or words exist that characterize the condition. Such language usually, but not always, exists in areas with endemic vitamin A inadequacy. Investigators must carefully determine under each local situation whether there is an appropriate term that can be reliably used in an historical interview. Where local terms are not used, alternate strategies to test dark adaptation indirectly may be useful. For example, a child taken from the bright outdoor sun into a darkened room may be asked to find a relative or to locate a favorite toy placed at a specific distance. The time required to achieve this task relative to that needed for a similarly aged child known to have nor mal dark adaptation is compared.
NEW AND DEVELOPING INDICATORS Functional biochemical measures are an alternative and more dynamic means of describing marginal vi-
ASSESSMENT OF VITAMIN
tamin A status applicable to individuals as well as populations. Two functional tests have recently been advocated: the relative dose-response (RDR) test, based on blood sampling, and conjunctival impression cy tology (CIC), based on histological techniques. The RDR measures vitamin A directly in blood, whereas CIC evaluates microscopically the number of goblet cells present and the morphological characteristics of epithelial cells. RDR test. As liver reserves of vitamin A become progressively depleted due to a chronically inadequate dietary supply, conservation mechanisms are invoked to increase the efficiency of vitamin A utilization among tissues and to maintain the level that is cir culating to the target tissues. At a point along the downward continuum, when the liver reserve is de pleted below a critical threshold, the rate of release of the remaining reserve is diminished. Synthesis of the carrier protein retinol binding protein (RBP)continues and results in the accumulation of a pool of preformed RBP. Providing an exogenous source of vitamin A causes the release of holo-RBP at a level and in a char acteristic time course relative to the amount of accu mulated preformed carrier protein (1, 13). After a fasting blood sample is obtained, a liquid solution containing about 450-1000 ¿tg (1.6-3.5 /miol) retinyl palmitate in an oily solution or aqueous dis persion is given orally (14). The exact amount given should be such that it does not overload the system with retinyl esters. However, it must be sufficient to assure complete mobilization of preformed accumu lated hepatic RBP. In some cases where less-efficient absorption might be suspected, e.g., chronic intestinal parasitism, a higher dose in the recommended range is advisable. If high-performance liquid chromatography (HPLC) is the analytical technique, retinol is distinguished from esters and overdosing is not a problem. A small meal containing foods with some fat but minimal vitamin A is given after dosing. After five hours, a second blood sample is obtained. The blood is analyzed for vitamin A by a suitable analytical technique, and the RDR is calculated as fol lows: .A vit A5 where vit A5 = serum vitamin A level five hours after dosing, and vit AQ = fasting serum vitamin A level. An RDR > 20% is considered to be positive and in dicative of inadequate hepatic stores of vitamin A, i.e., marginal vitamin A status (14). Recently, a modification of the procedure, which uses dehydroretinol (vitamin) and requires a single blood sample at 5 hours post dosing, has been tested in rats (15) and is being evaluated in children suspected of being vitamin A deficient (16). This procedure is termed the modified relative dose response (MRDR) test. Downloaded from https://academic.oup.com/jn/article-abstract/120/suppl_11/1459/4738619 by guest on 12 January 2018
A STATUS
1461
Currently available validation data indicate that a positive RDR occurs when hepatic stores are
