Vitamin A

There are multiple indicators for determining vitamin A deficiency. The four most commonly used biological indicators are serum (or plasma) retinol, retinol-binding protein (RBP), and the modified relative dose response (MRDR). Breast milk retinol can be used in some circumstances.

WHO recommends the use of two different criteria for determining the presence and severity of vitamin A deficiency as a public health problem. One is when the population prevalence of at least two biological parameters from a range of functional and biochemical indicators, , one of them being serum retinol, exceed the threshold for defining a public health problem 1. The second criterion specifies one biological indicator with a prevalence below the population-level cutoff value and at least four ecologic risk factors for vitamin A deficiency, two of which should be nutrition or diet-related 1. Demographic or ecologic risk factors include nutrition-related and illness-related risk factors. Examples of relevant risk factors are:

Nutrition- and diet-related risk factors:
  • <50% prevalence of breastfeeding in infants 6 months of age
  • Median dietary vitamin A intake <50% of recommended safe levels of intake among 75% of children 1-6 years of age
  • Stunting rate ≥30% and/or wasting rate ≥10% among children under 5 years of age
  • Food frequency assessment findings that foods with high vitamin A content consumed <3 times per week by ≥75% of vulnerable groups
Illness-related risk factors:
  • Infant mortality rate >75 per 1000 live births and child mortality rate of >100 per 1000 live births
  • Full immunization coverage <50% of children between 12-23 months of age
  • Two-week prevalence of diarrhoea of >20%
  • Measles case fatality rate of ≥1%
  • No formal schooling for >50% of women 15-44 years of age
  • <50% of households with a safe water source (boiled, treated, filtered, properly stored)

Serum (or plasma) retinol:

WHO recommends that serum retinol be used along with either another biological indicator of vitamin A status or with the other risk factors listed above to define the degree of public health significance of vitamin A deficiency at the population (not individual) level and to assess the need for vitamin A interventions 1, 2. Serum (or plasma) retinol can indicate subclinical, or marginal, vitamin A deficiency. Concentrations can change in response to vitamin A interventions when liver stores are low; when liver stores are replete, retinol concentrations may not respond to vitamin A interventions2. Serum retinol also correlates with the prevalence and severity of xerophthalmia.

Serum retinol concentrations are homeostatically controlled, but inflammation does cause them to decrease. Without an accompanying indicator of inflammation, artificially depressed serum retinol concentrations may lead to an overestimation of vitamin A deficiency prevalence. Inclusion of indicators such as CRP or AGP are recommended in any survey that includes the assessment of serum retinol to identify individuals with inflammation. However, at the time of this writing, the BOND review states that there is no consensus on the need for, or best method to, adjust for identified inflammation 3.

Specimen collection and management: Most commonly, retinol is measured in serum samples that are obtained by centrifugation of whole blood collected by venipuncture or finger prick. Whole blood needs to be refrigerated immediately and centrifuged within a few days of collection. When protected from light, the vitamin A in serum is stable or at least one week at 4˚C and for at least one year at 20˚C 4. However, procedures such as centrifugation and an adequate cold chain can be difficult to implement in remote field conditions.

Biomarker analysis: The most common and accurate method for measuring serum retinol is high-performance liquid chromatography (HPLC) with UV detection. In an adapted “micromethod,” 5 the required analysis volume is only 25 µL of serum or plasma. The minimum specimen volume is 100 µL to provide enough sample for a repeat analysis if needed. EDTA or heparinized plasma can be used, but serum is the preferred matrix. Commercially available retinol with greater than 95% purity is used as a calibrator (HPLC-grade reagents are preferred). Retinyl acetate, which is also commercially available, is used as an internal standard to correct for variations during the analytical procedure.

Quality control (QC) measures: Analytical method imprecision is typically around 5%. Moderate assay differences can occur with analyses conducted in different laboratories, therefore laboratories should participate in an external quality assurance programme such as CDC’s Vitamin A Laboratory and External Quality Assurance (VITAL-EQA) programme 6, CDC’s Performance Verification Program for Serum Micronutrients 7, which provides a one-time or annual performance report, or the National Institute of Standards and Technology (NIST) Health Assessment Measurements Quality Assurance Programme (HAMQAP). Serum-based certified reference materials (multiple levels of Standard Reference Material® (SRM®) 968) are available from NIST (Gaithersburg, MD, USA) to verify method accuracy. Quality control materials for serum micronutrients including retinol are available from CDC to support in-house quality assurance programmes for laboratories engaged in public health work 8.

Approximate budget requirements for analysis: Instrumentation needed for this method includes an HPLC and UV detector, centrifuge, vortex, and various pipettes (costing approximately US$ 50 000). The cost for materials and supplies is approximately US$ 5 per sample.

Interpretation of results: Low serum retinol is defined as <0.70 µmol/L in children 6–71 months of age. A serum retinol cutoff value of <1.05 mol/L is sometimes used to indicate vitamin A insufficiency. Interpretation of a population’s prevalence of serum retinol concentrations <0.70 µmol/L for defining a public health problem is presented in Table 3.1. It is important to note that this is based on low serum retinol values that have not been adjusted for inflammation.

Table 3.1. Prevalence of low serum retinol (<0.70 μmol/L),a used in conjunction with another indicator,b to define a public health problem at the population level and its level of importance among children 6–71 months of age 2

Level of importance as a public health problem Prevalence
Mild 2 to 9%
Moderate 10 to 19%
Severe ≥20%

aEquivalent to serum retinol <20 μg/dL. Prevalence estimates based on retinol values that have not been adjusted for inflammation

b“Another indictor” pertains to either a second biomarker for vitamin A deficiency, or at least four ecologic indicators for vitamin A deficiency, two of which should be nutrition or diet related.

Retinol Binding Protein (RBP):

RBP can be used as a surrogate, or proxy, indicator for retinol 3; however, a 1:1 molar equivalence between retinol and RBP does not usually occur. This means that the serum retinol cutoff value cannot be applied to RBP. The currently recommended approach to calculating an RBP cutoff value, detailed in the BOND review 3 and practiced in survey reports 9,10, is to measure serum retinol concentrations by HPLC in a subsample of the population where RBP is being used to assess vitamin A deficiency. Within the retinol subsample, given a reasonable correlation between RBP and retinol, a linear regression model is used to calculate an RBP concentration equivalent to retinol of 0.7 µmol/L. Ideally, the same blood draw would be used for measuring retinol and RBP (as opposed to taking a morning collection for RBP and an afternoon collection from the same person for retinol, for example). Outliers should be removed when reviewing the serum retinol-RBP regression plot, using studentized residuals larger than 3 in absolute value 11. A minimum sample size for the retinol subsample is suggested to be 20% of the total survey population or at least 100 observations, selected at random. The field of vitamin A assessment is rapidly evolving; the guidance presented in this manual is based on the best evidence available at the time of publication.

Specimen collection and management: RBP can be measured in serum or plasma. Blood should be collected, processed and stored as noted for the collection and management of specimens for serum (or plasma) retinol. The BOND Expert Panel for Vitamin A assigned the same degree of difficulty to serum RBP and serum retinol for sample collection and sample transportation. The dried blood spot (DBS) methodology for measuring serum retinol and RBP is less reliable than the method based on serum or plasma 12.

Biomarker analysis: Several commercial ELISA methods and laboratory-developed tests are available to measure RBP in serum or plasma, with serum being the preferred specimen matrix. However, no certified reference materials are available to verify RBP method accuracy, and RBP is therefore not included in external quality assurance programmes. Notably, the CDC VITAL-EQA programme 6 includes quality assurance for retinol, as well as for other nutritional biomarkers. In this programme, RBP is compared to retinol. Similar to using RBP as a proxy for retinol in surveys, laboratories that participate in VITAL-EQA may also compare RBP to retinol. Thus, although there are no external quality assurance programmes for RBP there is the opportunity to measure RBP in serum or plasma in commercial ELISA methods and compare to laboratory-developed tests against retinol as a proxy. Another advantage of these results is sample comparability because both biomarkers will be performed on the same specimen, handled under the same conditions. The specimen should be gently and thoroughly mixed before measuring the retinol and RBP.

Approximate budget requirements for analysis: Instrumentation needed for ELISA methods includes a plate washer, plate reader and various pipettes (approximately US$ 30 000). The cost for materials and supplies is approximately US$ 2–5 per sample, depending on whether the assay used is laboratory-developed or a commercial kit. For a laboratory developed RBP test, a significant amount of time will be necessary to find appropriate antibodies and validate the assay.

Interpretation of results: There are at present no WHO guidelines on the interpretation of vitamin A deficiency prevalence based solely on RBP. When RBP is measured in surveys as a proxy for retinol, it is important to include a caveat if the public health significance of vitamin A deficiency is based on a prevalence of RBP <0.7 µmol/L. Another important consideration to keep in mind is that the prevalence numbers in Table 3.1 are based on serum retinol values that have not been adjusted for inflammation. It is recommended to present both inflammation-adjusted and unadjusted estimates for vitamin A deficiency until guidance from WHO becomes available. Suggested methods for adjusting serum retinol concentrations based on CRP and AGP include the use of regression or arithmetic correction factors, such as those developed by BRINDA 13 and Thurnham 14, respectively. Also, since enzyme immunoassays do not distinguish between holo- and apo-RBP, an additional adjustment (that requires determination of serum retinol) is needed to reflect the RBP:retinol ratio in the population of interest 15, 16. Box 3.2 summarizes the use of RBP for assessing vitamin A status.

Box 3.2 Summary of the use of RBP

At the time of this writing, the preferred method for assessing vitamin A status is serum retinol. When RBP is selected as the main vitamin A indicator for a survey, it is also necessary to measure retinol in a subsample of the population. This will permit the determination of a survey specific RBP cutoff value to define vitamin A deficiency. It is also useful to measure MRDR in a subsample because it provides an assessment of vitamin A liver reserves and the WHO recommendation of two biomarkers to assess deficiency.

It may be complicated to analyse trends if there are different survey-specific population-level cutoff values within the same country from various years. For example, the RBP survey-specific cutoff value to define vitamin A deficiency may be 0.78 µmol/L in one survey cycle and 0.58 µmol/L in the next survey cycle. As such, the prevalence of RBP below those two separate cutoff values may be hard to compare and raises the question of why the RBP-retinol relationship changed in a population between survey cycles. This is important to note, and may be a factor to consider when choosing vitamin A indicators. WHO has defined the retinol cutoff value for deficiency, making it easier to assess trends over time when retinol is collected and analysed using similar methods across surveys. However, many older surveys did not assess indicators of inflammation, which can influence the interpretation of retinol and trends over time as the prevalence of inflammation can vary from survey to survey.

Modified relative dose response (MRDR):

RBP is synthesized in the liver as apo-RBP (unbound RBP). When liver reserves of vitamin A are low, apo-RBP accumulates in the liver. When vitamin A becomes available from newly ingested sources the accumulated apo-RBP binds to the retinol and is released into circulation as the holo-RBP complex (retinol bound to RBP). MRDR is a functional test that takes advantage of this process by providing individuals with a measurable “challenge” dose of 3,4-didehydroretinol (also known as DR or vitamin A2) in the acetate form. DR binds to apo-RBP in the liver forming the holo-RBP complex, which is quickly released into the plasma during deficiency. After this challenge dose, DR should appear in serum in significant amounts over baseline (prior to the challenge dose) only when liver reserves of vitamin A are low. Therefore, the amount of DR released is an indication of vitamin A status. MRDR is calculated from the molar ratio of DR to retinol (DR:R). In comparison with other vitamin A indicators, MRDR is less influenced by inflammation and it is not homeostatically controlled in the timeframe of the test 3. An important consideration of MRDR is the time required between administering the challenge dose and collecting the specimen (4-6 hours after the challenge).

The MRDR is useful to assess changes in liver stores of vitamin A, for example, changes in response to an intervention to improve vitamin A status. The MRDR test provides useful semi-quantitative information to evaluate deficiency through low liver reserves of vitamin A. On the contrary, it is not useful in defining excessive vitamin A reserves. MRDR is recommended for inclusion among a randomly selected subsample of individuals, to assess the underlying vitamin A status of the population studied 3. Serum retinol is collected and analysed from the same blood draw as that used for MRDR. When assessing vitamin A deficiency at the population level, it is useful to assess the mean and standard deviation of the MRDR value (namely, DR:R). When comparing results from MRDR with results from RBP and retinol, there may be inconsistencies at the individual level for categorizing deficiency. Thus, comparing vitamin A deficiency prevalence estimates from MRDR, RBP, and retinol may cause confusion. It is most useful to look at two biologic parameters (MRDR plus either retinol (preferred) or RBP) to determine the population status of vitamin A deficiency as a public health problem 3, 1.

Specimen collection and management: In preparation for specimens for an MRDR assessment, an individual must consume a small challenge dose of a retinol analog (DR or vitamin A2) along with a fatty snack (lacking in vitamin A) to ensure absorption. This should be done about 4 to 6 hours before collecting 1-3 mL of venous blood. The dose of vitamin A2 can be mixed with 1 mL of olive oil or another edible oil containing no vitamin A. Administering it using a disposable syringe helps ensure that it is completely swallowed, especially with small children. Survey participants must also be questioned about recent consumption of foods rich in vitamin A prior to administering the vitamin A challenge dose. If there has been recent consumption of vitamin A rich foods, it will be necessary to wait two hours before proceeding with the test. Vitamin A rich foods should not be consumed again until after the blood draw for the MRDR test.

The same procedures for transporting, processing, and storing of specimens for other vitamin A indicators apply to the MRDR venous blood specimen.

Biomarker analysis: Analysis of 3,4-di-dehydroretinol requires HPLC and can be assessed in the same analytical run as serum retinol. The required sample volume for analysing serum retinol and 3,4-di-dehydroretinol is 250 µL of serum or plasma, and the minimum specimen volume is 500 µL to provide enough sample for a repeat analysis if needed. Retinol acetate is used as an internal standard calibrator for retinol (commercially available at a satisfactory purity >95%) in each analytical run. Quality control (QC) samples are recommended, which have a known concentration of retinol that covers the range of retinol concentrations expected in the human population (low, medium and high) and are used to validate each analytical assessment, aiding in correcting analytical bias.

Approximate budget requirements for analysis: Instrumentation needed for this method includes an HPLC, centrifuge, vortex, and various pipettes (costing approximately US$ 50 000 for the complete set of equipment). The cost for materials and supplies is approximately US$ 5 per sample.

Interpretation of results: MRDR is a semi-quantitative indicator of vitamin A status. The MRDR value, which is the ratio of DR to retinol in serum, indicates adequacy of liver reserves. For individuals, the 2016 BOND review 3 recommends a MRDR cutoff value of ≥0.060 to indicate insufficient liver reserves (≤0.1 µmol retinol/g liver vitamin A), and of <0.060 to indicate enough liver reserves (≥0.1 µmol retinol/g liver vitamin A). For groups, a mean MRDR value <0.030 is recommended for indicating adequate vitamin A status 3.

Breast milk retinol:

WHO recommends exclusive breastfeeding for infants in the first 6 months of life, followed by continued breastfeeding with appropriate complementary foods for up to 2 years or beyond 17. Breast milk retinol concentrations provide information about both the mother and breastfed infant. They are considered to reflect the recent dietary intake of mothers and can be used to estimate vitamin A intake of infants receiving the breast milk 3, 18. Breast milk retinol concentrations have been used to assess the risk of vitamin A deficiency in populations, determine the efficacy of maternal vitamin A interventions, and for the monitoring and evaluation of programmes providing maternal vitamin A interventions 3. Average breast milk retinol concentrations from well-nourished women are about 485 µg/L 19; however, average concentrations can fall below 400 µg/L in areas where vitamin A deficiency is of public health significance 1. When selecting the population for evaluation, age, stage of lactation, geographic location, season and pregnancy status should be considered 3. Vitamin A content is very high in colostrum (milk secreted in the first 4-6 days postpartum), and remains high in transitional milk (days 7-21 postpartum), after stabilizing in mature milk (after about day 21 postpartum). Therefore, breast milk samples collected after one month postpartum, avoiding colostrum and transitional milk samples, are most useful for assessment of vitamin A status.

Specimen collection and management: Breast milk can be collected as a full milk sample or as a casual sample. The specimen collection method would depend on the survey objectives. A casual sample is appropriate for assessment of population-level prevalence of low breast milk vitamin A, expressed in nmol retinol/g fat, and will be described here. A survey objective of estimating the vitamin A intake of infants from breast milk would necessitate that a full milk sample be collected, which is described in detail elsewhere 20.

Milk collected using the casual sample method (~10 mL) can be hand-expressed into specimen cups or tubes made of polypropylene. A benefit of casual milk collection is that there is no need to standardize ‘time since last feed’; however, milk fat will need to be measured. One option for milk fat measurement is a creamatocrit centrifuge, which is field friendly. Casual milk collection is defined as mid-feed collection 1 minute after let-down, by manual expression. Although women can do the manual expression in privacy and without assistance after receiving adequate instructions, it is important to consider the gender of the field staff and the local context as female health workers may be more appropriate for surveys that include breast milk collection. The breast milk needs to be refrigerated at 4˚C immediately and protected from direct light because vitamin A in milk is less stable than serum retinol, which is protein bound 20. Protection from light and keeping the milk cold will prevent photodegradation of the vitamin A. If refrigerated, the breast milk must be analysed within 24 hours of collection or it may be stored frozen at 20˚C (or colder, such as 80˚C) and analysed within one year of collection 3. Before freezing, precise aliquots of milk that will be used for measuring vitamin A content should be prepared as thawed samples can be difficult to homogenize. However, procedures such as centrifugation and an adequate cold chain can be difficult to implement in remote field conditions.

Biomarker analysis: The most common and accurate method for measuring breast milk retinol is high-performance liquid chromatography (HPLC) with UV detection after saponification 21. Portable fluorometers are also field friendly equipment that enables mothers to get immediate results on their breast milk vitamin A status 22, and has performed well compared to HPLC 23. Because vitamin A is found in the milk fat, fresh milk should be mixed well so that the cream layer is evenly distributed within the sample taken for measurement. A sample volume of 2 mL of breast milk is required for analysis using HPLC. The minimum specimen volume is 100 µL to provide enough sample for a repeat analysis if needed. Commercially available retinol with greater than 95% purity is used as a calibrator (HPLC-grade reagents are preferred). The base solution to be used for saponification should be mixed and stored in plastic containers to remain stable 21. Either 3,4-didehydroretinyl acetate (3), C23-beta-apo-carotenol 24, 25, or tocol 26, may be used as an internal standard to correct for variations during the analytical procedure. To determine the amount of milk fat in a specimen, the creamatocrit methods can be done in a laboratory 27, or a creamatocrit centrifuge can be used in the field.

Quality control (QC) measures: Analytical method imprecision is typically around 5%. Moderate assay differences can occur with analyses conducted across laboratories; however, the National Institute of Standards and Technology (NIST) Health Assessment Measurements Quality Assurance Programme (HAMQAP) does not have certified control for human breast milk, making external quality assurance impractical.

Approximate budget requirements for analysis: Instrumentation needed for the HPLC method includes an HPLC with a UV detector, centrifuge, vortex, and various pipettes (costing approximately US$ 50 000). A calibrated spectrophotometer is also needed for external standard quantification. The cost for materials and supplies is approximately US$ 5 per sample.

Interpretation of results: Breast milk retinol concentrations ≤1.05 µmol/L are considered inadequate (18), but this cutoff applies only to full milk collection. It is preferable to express breast milk retinol concentrations per gram of fat to account for fat variability. Retinol concentrations ≤28 nmol/g milk fat (or ≤8 µg/g milk fat) are considered inadequate (18). Vitamin A deficiency is considered a public health problem of mild, moderate or severe importance at a prevalence of inadequate concentrations of <10%, 10-24%, and ≥25%, respectively.

Clinical and functional indicators:

Clinical or functional indicators of vitamin A deficiency usually focus on xerophthalmia, an eye condition that worsens as the depletion of vitamin A stores progresses 28. Most of these indicators are not recommended for routine cross-sectional surveys due to their rare occurrence, even in areas endemic for vitamin A deficiency. The one exception is assessing whether women have experienced night blindness during a pregnancy within the previous three years or five years. To help interpret reported vision problems, women being assessed should report problems seeing at night as well as during the day during the last pregnancy. The WHO/International Vitamin A Consultative Group (IVACG) states that a prevalence of night blindness that exceeds 5% among pregnant women would indicate vitamin A deficiency of public health significance among the population 29.

  1. Indicators for assessing vitamin A deficiency and their application in monitoring and evaluating intervention programmes. Geneva: World Health Organization; 1996 (https://www.who.int/nutrition/publications/micronutrients/vitamin_a_deficiency/WHONUT96.10.pdf, accessed 11 May 2020).  2 3 4 5

  2. Serum retinol concentrations for determining the prevalence of vitamin A deficiency in populations. Geneva: World Health Organization; 2011 (https://www.who.int/vmnis/indicators/retinol.pdf, accessed 11 May 2020).  2 3

  3. Tanumihardjo SA, Russell RM, Stephensen CB, Gannon BM, Craft NE, Haskell MJ et al. Biomarkers of Nutrition for Development (BOND) – vitamin A review. J Nutr. 2016;146:1816S–48S. doi: 10.3945/jn.115.229708.  2 3 4 5 6 7 8 9 10 11 12

  4. Drammeh BS, Schleicher RL, Pfeiffer CM, Jain RB, Zhang M, Nguyen PH. Effects of delayed specimen processing and freezing on serum concentrations of selected nutritional indicators. Clin Chem. 2008;54:1883–91. doi: 10.1373/clinchem.2008.108761. 

  5. Chaudhary-Webb M, Schleicher RL, Erhardt JG, Pendergrast EC, Pfeiffer CM. An HPLC ultraviolet method using low sample volume and protein precipitation for the measurement of retinol in human serum suitable for laboratories in low- and middle-income countries. J Appl Lab Med. 2019;4:101–7. doi:10.1373/jalm.2018.027508. 

  6. Haynes BM, Schleicher RL, Jain RB, Pfeiffer CM. The CDC VITAL-EQA program, external quality assurance for serum retinol, 2003–2006. Clin Chim Acta. 2008; 390:90-6. doi: 10.1016/j.cca.2008.01.009.  2

  7. Performance Verification Program for Serum Micronutrients [website]. Atlanta: US Centers for Disease Control and Prevention (CDC); 2019 (https://www.cdc.gov/nceh/dls/nbb_micronutrient_performance.html, accessed 14 June 2020). 

  8. Quality control materials for serum micronutrients [website]. Atlanta: US Centers for Disease Control and Prevention (CDC); 2019 (https://www.cdc.gov/nceh/dls/nbb_micronutrient_materials.html, accessed 11 May 2020). 

  9. National Statistical Office (NSO), Community Health Sciences Unit (CHSU) [Malawi], Centers for Disease Control and Prevention (CDC), Emory University. Malawi Micronutrient Survey 2015–16. Atlanta: NSO, CHSU, CDC and Emory University; 2017. 

  10. Engle-Stone R, Haskell MJ, Ndjebayi AO, Nankap M, Erhardt JG, Gimou MM et al. Plasma retinol-binding protein predicts plasma retinol concentration in both infected and uninfected Cameroonian women and children. J Nutr. 2011;141:2233–41. doi: 10.3945/jn.111.145805. 

  11. Thompson WR. On a criterion for the rejection of observations and the distribution of the ratio of deviation to sample standard deviation. Ann Math Statist. 1935;6:214–9. 

  12. New evidence on methods to assess vitamin A status: implications for the use of Uganda Demographic and Health Surveys vitamin A data. Kampala and Rockville: Uganda Bureau of Statistics and ICF; 2018; https://dhsprogram.com/pubs/pdf/FR333/Vitamin_A_dissemination_handout_9March2018.pdf, accessed 11 May 2020). 

  13. Larson LM, Guo J, Williams AM, Young MF, Ismaily S, Addo OY et al. Approaches to assess vitamin A status in settings of inflammation: Biomarkers Reflecting Inflammation and Nutritional Determinants of Anemia (BRINDA) project. Nutrients. 2018;10:1100. doi: 10.3390/nu10081100. 

  14. Thurnham DI, McCabe LD, Haldar S, Wieringa FT, Northrop-Clewes CA, McCabe GP. Adjusting plasma ferritin concentrations to remove the effects of subclinical inflammation in the assessment of iron deficiency: a meta-analysis. Am J Clin Nutr. 2010;92:546–55. 

  15. Zabetian-Targhi F, Mahmoudi MJ, Rezaei N, Mahmoudi M. Retinol binding protein 4 in relation to diet, inflammation, immunity, and cardiovascular diseases. Adv Nutr. 2015;6:748–62. doi: 10.3945/an.115.008292. 

  16. Aeberli I, Biebinger R, Lehmann R, L’allemand D, Spinas GA, Zimmermann MB. Serum retinol-binding protein 4 concentration and its ratio to serum retinol are associated with obesity and metabolic syndrome components in children. J Clin Endocrinol Metab. 2007;92:4359–65. 

  17. Global strategy for infant and young child feeding. Geneva: World Health Organization; 2003 (http://apps.who.int/iris/bitstream/10665/42590/1/9241562218.pdf, accessed 11 May 2020). 

  18. Priorities in the assessment of vitamin A and iron status in populations. Panama City, Panama, 15–17 September 2010. Geneva: World Health Organization; 2012 (https://apps.who.int/iris/bitstream/handle/10665/75334/9789241504225_eng.pdf; accessed 11 May 2020). 

  19. Institute of Medicine (US) Panel on Micronutrients. Dietary reference intakes for vitamin A, vitamin K, arsenic, boron, chromium, copper, iodine, iron, manganese, molybdenum, nickel, silicon, vanadium, and zinc. Washington, DC: National Academies Press; 2001. 

  20. Stoltzfus RJ, Underwood BA. Breast-milk vitamin A as an indicator of the vitamin A status of women and infants. Bull World Health Organ. 1995;73:703–11.  2

  21. Tanumihardjo SA, Penniston KL. Simplified methodology to determine breast milk retinol concentrations. J Lipid Res. 2002;43:350–5.  2

  22. Chaimongkol L, Pinkaew S, Furr HC, Estes J, Craft NE, Wasantwisut E, Winichagoon P. Performance of the CRAFTi portable fluorometer comparing with the HPLC method for determining serum retinol. Clin Biochem. 2011;44:1030–2. 

  23. Engle-Stone R, Haskell MJ, La Frano MR, Ndjebayi AO, Nankap M, Brown KH. Comparison of breast milk vitamin A concentration measured in fresh milk by a rapid field assay (the iCheck FLUORO) with standard measurement of stored milk by HPLC. Eur J Clin Nutr. 2014;68:938–40. 

  24. Surles RL, Hutson PR, Valentine AR, Mills JP, Tanumihardjo SA. 3, 4-Didehydroretinol kinetics differ during lactation in sows on a retinol depletion regimen and the serum:milk 3, 4-didehydroretinol:retinol ratios are correlated. J Nutr. 2011;141:554–9. doi: 10.3945/jn.110.131904. 

  25. Tanumihardjo SA, Howe JA. Twice the amount of alpha-carotene isolated from carrots is as effective as beta-carotene in maintaining the vitamin A status of Mongolian gerbils. J Nutr. 2005;135:2622-6. doi:10.1093/jn/135.11.2622. 

  26. Turner T, Burri BJ, Jamil KM, Jamil M. The effects of daily consumption of β-cryptoxanthin-rich tangerines and β-carotene-rich sweet potatoes on vitamin A and carotenoid concentrations in plasma and breast milk of Bangladeshi women with low vitamin A status in a randomized controlled trial. Am J Clin Nutr. 2013;98:1200–8. 

  27. Lucas A, Gibbs JA, Lyster RL, Baum JD. Creamatocrit: simple clinical technique for estimating fat concentration and energy value of human milk. Br Med J. 1978;1:1018–20. 

  28. Gilbert C. The eye signs of vitamin A deficiency. Community Eye Health. 2013; 26:66-7. 

  29. Taren D. Historical and practical uses of assessing night blindness as an indicator for vitamin A deficiency. In: Priorities in the assessment of vitamin A and iron status in populations. Panama City, Panama, 15–17 September 2010. Geneva: World Health Organization; 2012 (https://apps.who.int/iris/bitstream/handle/10665/75334/9789241504225_eng.pdf; accessed 13 June 2020).