| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research Communication |
1 From the Divisions of Nutrition and Physical Activity (HMB, MKS, and LKK), Diabetes Translation (BAB), and Oral Health (WK), National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta; the Epidemic Intelligence Service, Division of Applied Public Health Training, Epidemiology Program Office, Centers for Disease Control and Prevention, Atlanta (HMB); and the International Emergency and Refugee Health Branch, Division of Emergency and Environmental Health Services, National Center for Environmental Health, Centers for Disease Control and Prevention, Atlanta (BAW).
2 Supported in part by the World Food Programme, the United Nations High Commissioner for Refugees, and the Centers for Disease Control and Prevention.
3 Address reprint requests to HM Blanck, Division of Nutrition and Physical Activity, Centers for Disease Control and Prevention, 4770 Buford Highway, NE, Mailstop K-26, Atlanta, GA 30341. E-mail: hblanck{at}cdc.gov.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Objectives: The main objectives were to assess the prevalence of AS and of low concentrations of riboflavin, folate, vitamin B-12, and iron by using biochemical measures; to determine whether riboflavin status was associated with AS; and to assess the potential of AS as a screening measure for low riboflavin concentrations.
Design: In October 1999, we performed a survey among a random sample of 463 adolescent refugees in which we conducted interviews and physical examinations and obtained blood specimens for riboflavin assessment. Riboflavin status was assessed with the erythrocyte glutathione reductase (EC 1.6.4.2) activity coefficient. After we excluded those adolescents who had taken vitamins during the past month, 369 were eligible for analyses.
Results: AS was common (26.8%; 95% CI: 22.3, 31.3), the prevalence of low riboflavin concentrations was high (85.8%; 80.7, 90.9), and riboflavin status was associated with AS. Adolescents with AS had significantly lower riboflavin concentrations than did adolescents without AS (P = 0.02). The adjusted odds ratio for AS and low riboflavin concentrations was 5.1 (1.55, 16.5).
Conclusion: Globally, riboflavin deficiency is rare. Its emergence in food-dependent populations can be a harbinger of other B-vitamin deficiencies.
Key Words: Stomatitis • folic acid • Nepal • refugees • riboflavin • riboflavin deficiency
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
83000 ethnic Nepalese fled their homes in Bhutan because of fear of new citizenship laws and consequent ethnic persecution. Ultimately, these refugees were settled in 7 camps in the Jhapa and Morang districts of southeastern Nepal. In July 1992, the office of the United Nations High Commissioner for Refugees, Save the Children– United Kingdom, and the Centers for Disease Control and Prevention helped establish a surveillance system to monitor morbidity and mortality (1). Workers at each of the major camp health centers recorded clinical signs of diarrhea, malaria, worm infection, acute respiratory infections, and deficiencies in micronutrients. In response to sporadic signs of possible micronutrient deficiency, including beriberi, a vitamin-fortified cereal blend was added to the general ration in June 1994. The blend was withdrawn in January 1999 because of program constraints. After this withdrawal, the number of reported cases of angular stomatitis (AS), ie, thinning or fissuring at the angles of the mouth, increased quickly and dramatically from 5.5 per 1000 refugees per month in December 1998 to 35.6 cases per 1000 refugees per month in March 1999 (2). The highest rates were found in children and adolescents. A nonspecific condition possibly exacerbated by infection, AS has classically been linked with deficiencies of riboflavin, other B vitamins, and iron (3, 4). For riboflavin deficiency specifically, a purplish-red or magenta tongue and cheilosis (vertical fissuring of the lips) are also regarded as clinical signs. Examination for AS has been advocated for nutrition surveys since the 1960s (5, 6), and AS status has been useful in diagnosing riboflavin deficiency in populations (7). The possible functional consequences of riboflavin deficiency in humans include decreases in motor skills and attention span (8) and in the absorption or utilization of iron (9), and those in animals include decreased growth (10, 11). To date, however, the impact of marginal B-vitamin concentrations during the intensive growth period of adolescence has not been well documented.
We are unaware of any previous investigation of an outbreak of AS and riboflavin deficiency in a refugee population. The possibility that the dramatic increase in AS among these Bhutanese refugees had a nutritional cause and the concern that similar refugee populations may be at risk led to this investigation. Because adolescents exhibited the highest rates of AS, we focused on that high-risk subgroup. The main objectives of this study were 1) to assess the prevalence of AS and associated symptoms; 2) to assess the prevalence of low concentrations of riboflavin, folate, vitamin B-12, and iron by using biochemical measures; 3) to determine whether riboflavin status was associated with AS; and 4) to assess the potential of AS as a screening measure for low riboflavin concentrations.
|
SUBJECTS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Subjects
Potential participants were chosen by simple random sampling from computerized registration data. To estimate prevalence ± 5 percentage points, we assumed point estimates of 30% and 50% for the prevalence of AS and low hemoglobin, respectively, which resulted in samples of 321 and 400 adolescents, respectively (12). To allow for expected refusals and attrition, we needed to enroll 495 subjects; thus, we chose every 53rd subject from a randomized enumeration of all registered adolescents. Of these 495 adolescents, 463 (93.5%) provided informed consent to participate. Of the 32 who did not enroll, 14 (2.8%) resided outside the camps, 13 (2.6%) were aged <10 y or >19 y, 3 (0.6%) refused, 1 (0.2%) could not be found, and 1 (0.2%) had died. The study underwent Human Subjects Review by the Centers for Disease Control and Prevention Institutional Review Board.
Data collection
Participants were asked about their recent medical history, vitamin use in the previous month, tongue pain, duration of the current AS episode, treatment of AS, illness (diarrhea, pneumonia, fever) in the previous month, and demographic and household characteristics. We also asked about food consumption during the previous week, including that of lentils, potatoes, green vegetables, fruit, dairy products, eggs, and meat or fish. Parents were encouraged to accompany younger participants. All participants were weighed and heights were measured without shoes. Local Nepali camp medical assistants who had completed 
2 y of medical training performed a limited physical examination of the face and neck to assess for clinical signs of micronutrient deficiency, including AS, cheilosis, and goiter. Because there is no published classification system for the severity of AS, we first classified the AS cases in photographs taken in the spring of 1999 in the Nepali camps as mild-to-moderate, severe, or scarred. The photographs were used to train survey workers to standardize the diagnosis of AS and the grading of its severity; these photos were similar to those included by McLaren (13). We defined a mild-to-moderate condition as a thin and mottled surface epithelium with superficial fissuring, a severe condition as a cracked surface epithelium with well-developed fissuring, and a scarred condition as a white or pink surface epithelium. Cheilosis was defined as vertical fissuring of the lips complicated by redness and swelling. To measure hemoglobin concentrations, nonfasting blood specimens were obtained from all participants by finger stick. To measure B-vitamin and transferrin receptor (TfR) concentrations, nonfasting blood specimens were obtained by phlebotomy from every second participant.
Biochemical assessment
Hemoglobin concentrations were measured with a Hemocue hemoglobinometer (AB Addo, Malmö, Sweden). Blood collected by venipuncture was stored in plain (clotting) and EDTA-coated tubes, which were kept in an insulated box containing ice packs for 1–8 h until they were processed. For folate, vitamin B-12, and TfR testing, clotted blood was centrifuged, and the serum was removed by pipette, stored in cryovials, and kept frozen at –20°C. Serum specimens were transported on ice to the Centers for Disease Control and Prevention (Atlanta), where they were kept frozen until they were analyzed. The serum concentrations of folate and vitamin B-12 were measured with a commercial radioimmunoassay kit (Bio-Rad Laboratories, Hercules, CA). Serum TfR testing was done with a commercial kit (Ramco Laboratories, Houston). For riboflavin testing, 0.1 mL whole blood collected in an EDTA-coated tube was added to a cryovial containing 0.9 mL distilled, sterile water, and the mixture was frozen at –20°C (14). Diluted whole-blood specimens were transported to Atlanta in a frozen state on dry ice and subsequently shipped on dry ice to the laboratory of Philip J Garry (University of New Mexico) and stored at –20°C until they were analyzed. Riboflavin status was assessed by a functional assay that measured the activity of a riboflavin-dependent enzyme, erythrocyte glutathione reductase (EC 1.6.4.2), in whole blood. Laboratory testing measured the erythrocyte glutathione reductase activity coefficient (EGRAC), with a higher coefficient reflecting lower riboflavin concentrations (14).
Control blood samples were collected from 3 supervisors in the field during the survey and were processed, stored, and shipped along with blood samples from the study subjects. All of the control samples had riboflavin EGRACs within the normal range (0.68–1.70) for the assay (supervisor A: 1.09; supervisor B: 1.16; supervisor C: 1.43) (14). Thirty samples were analyzed a second time. The resulting EGRAC between-run CV was 6%.
Variable definition
An EGRAC >1.7 was used to indicate low riboflavin concentrations (14). A folate concentration <2.6 ng/mL and a vitamin B-12 concentration <201 pg/mL were considered low; these cutoffs reflect normal values for healthy US adults (15). A TfR concentration >8.5 µg/mL was used to indicate low tissue iron stores (16). Low hemoglobin concentrations were established with the use of World Health Organization–recommended age- and sex-specific anemia cutoffs (in g hemoglobin/dL): both sexes aged 10–11 y, 11.5; both sexes aged 12–14 y, 12.0; females aged 15 y, 12.0; and males aged 15 y, 13.0 (17, 18). The prevalence of low body mass index was determined by using the sex-specific World Health Organization body mass index–for–age cutoffs that are based on data from US adolescents (19).
Statistical analysis
Data were entered with EPI INFO 6.04b (Centers for Disease Control and Prevention, Atlanta), and the analysis was performed with SAS software (20). Chi-square or Fisher’s exact test was used to compare categorical data, and t tests were used to compare means between subgroups of normally distributed continuous variables. Correlations between nutrients and indexes of the utility of AS as a surveillance screening measure, including sensitivity, specificity, and positive predictive value (PPV), were also calculated (21). A P value < 0.05 was considered statistically significant for all tests of association.
Logistic regression was used to model the prevalence odds ratio for the occurrence of AS in relation to nutrient (B vitamin and iron) concentrations with adjustment for potential confounders. The use of the US cutoff for low riboflavin concentrations designated a small group of adolescents (n = 26) as having normal riboflavin concentrations and an even smaller group (n = 5) as having both AS and normal riboflavin concentrations. Because of potential unstable estimates due to this small number (22) and because a more appropriate population cutoff was not available, we used the continuous EGRAC value in the multivariate analysis. Variables included in the models were marginally related to either AS or nutrient status in univariate analyses and included age, camp of residence, household size, marital status, family member work status, and consumption of dairy products, lentils, and fruit. Collinearity was assessed by standard methods (23), and age was centered (ie, the mean age in years was subtracted from the age in years for each survey subject). Effect modification was assessed by including biologically relevant interaction terms (eg, camp x nutrient status). Final models were selected through a backward-elimination strategy; covariates that were meaningfully associated with the outcome (P < 0.10) or that altered the estimate for nutrient levels by more than 10% were retained in the model.
Exclusions
B-complex tablets (Royal Drugs Ltd, Kathmandu, Nepal) were used in camp clinics to treat AS. Tablets containing iron and folate were used to treat anemia. Such supplementation would render the measured body concentrations of the corresponding vitamin unrepresentative of usual status (8) and could affect current anemia and physical signs and symptoms. Of the 463 participants, 94 (20.3%) were excluded because they had taken a B-vitamin or iron and folate supplement in the previous month; the result was a final sample size of 369. There was no significant difference in supplement use by age, sex, or camp.
After the exclusion of enrollees who had taken vitamins in the previous month, 154 serum samples (of 192 collected) remained for folate and B-12 assessment, 190 serum samples (of 192 collected) remained for TfR assessment, and 183 blood samples (of 226 collected) remained for riboflavin assessment.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Few were married, and about one-half lived in a household with 7 or more members. Less than one-half had family members who were enrolled in supplemental feeding programs, and the families of more than one-half had a kitchen garden. Mean ± SD reported food consumption (times/wk) was 8.9 ± 4.9 for lentils, 7.6 ± 4.8 for potatoes, 1.6 ± 2.0 for green vegetables, 1.2 ± 1.5 for fruit, 1.5 ± 2.1 for dairy products, 0.1 ± 0.5 for eggs, and 0.2 ± 0.5 for meat or fish.
|
). Of the AS cases, 47.4% were classified as mild-to-moderate, 23.7% as severe, and 28.9% as scarred. Most of the subjects (n = 89; 89.9%) were affected bilaterally, and 4 (4.0%) had accompanying perioral lesions. Participants with current AS reported durations of 2–180 d (mean: 18 d; median: 7 d). Few participants had cheilosis, but one-half reported tongue pain in the prior month. Over one-third of participants had a low body mass index at the time of the survey and over one-half had fever in the prior month, but very few had goiter at the time of the study.
More than three-fourths of the 183 participants assessable for riboflavin had low riboflavin concentrations (median and 
± SE EGRAC: 2.0 ± 0.30) (Table 1). More than one-third of the 154 participants assessable for folate and B-12 had low serum folate concentrations (median: 3.1 ng/mL; ± SE: 3.5 ± 2.4 ng/mL), and a smaller proportion had low serum concentrations of vitamin B-12 (median: 271 pg/mL; ± SE: 307 ± 150 pg/mL). Overall, just under one-quarter of the 369 participants had low hemoglobin concentrations (median and ± SE: 13.0 ± 1.7 g/dL), and one-half of the 190 assessable for tissue iron had low tissue iron stores (median TfR: 8.7 µg/mL; ± SE: 9.5 ± 1.6 µg/mL). Riboflavin status was correlated with serum folate (Spearman’s correlation coefficient 0.31; P = 0.0001) and serum vitamin B-12 (Spearman’s correlation coefficient 0.19; P = 0.02).
Among the 183 adolescents who underwent riboflavin assessment, 68 had AS. This group of adolescents had a significantly higher (P = 0.02, Student’s t test) mean (± SE) EGRAC (2.2 ± 0.4) than did the adolescents without AS (2.0 ± 0.3), which reflected the lower concentrations of riboflavin in the former group. Among 18 adolescents with mild-to-moderate AS, 83% had low riboflavin concentrations ( ± SE EGRAC: 2.1 ± 0.4; median EGRAC: 2.0); all 9 adolescents with severe AS had low riboflavin concentrations ( ± SE EGRAC: 2.3 ± 0.3; median EGRAC: 2.3) (P = 0.10, chi-square for trend; P = 0.23, t test). Scarred AS may be more reflective of past AS episodes than of current nutritional status. Among the 11 adolescents with scarring, 90% had low riboflavin concentrations ( ± SE EGRAC: 2.3 ± 0.2; median EGRAC: 2.2).
The prevalence of AS and low concentrations of riboflavin by demographic characteristics, food consumption in the previous week, health characteristics, and biochemical status are summarized in Table 2. Age, sex, biochemical status, and variables that were at least marginally related to the presence of either AS or low concentrations of riboflavin (P 
0.10) are shown. AS was not significantly related to sex, but there were borderline relations with age and with camp of residence. AS was significantly less common in the adolescents who reported <7 family members in the household and in those who reported consumption of any dairy products in the prior week. AS was significantly more common in the adolescents with cheilosis or self-reported tongue pain in the prior month. AS was slightly more common in those with low riboflavin concentrations, but this relation was not significant. The prevalence of AS did not differ significantly by consumption of any other food item or any of the other health conditions listed in the questionnaire.
|
). They were slightly less common in those who reported <7 family members in the household, in those who were married, in those who reported that a family member worked, and in those who reported consumption of any fruit in the previous week. They were significantly less common in those who reported consuming lentils 10 times/wk and in those with normal serum folate concentrations. Riboflavin status was not related to other biochemical measures, demographic characteristics, or health conditions. We also used logistic regression to examine factors predicting AS. In an unadjusted model (n = 183) using the EGRAC value for riboflavin, we found that EGRAC was a significant risk factor for AS (ß = 1.50, SE = 0.53) and that, when an EGRAC of 2.0 (low riboflavin concentrations) was compared with an EGRAC of 1.0 (normal riboflavin concentrations), the odds ratio was 4.5 (95% CI: 1.58, 12.7). Because of missing data, the final adjusted multivariate models included only 154 adolescents. In the final model adjusted for age, family-member work status, dairy product consumption, and serum folate and vitamin B-12 concentrations, we found that EGRAC was still a significant risk factor for AS (ß = 1.62, SE = 0.60) and that the odds ratio was still significant at 5.1 (95% CI: 1.55, 16.5).
The sensitivity of AS as an indicator of low riboflavin concentrations was 22%: of the 157 adolescents with low riboflavin concentrations, 34 had AS. The specificity of AS in detecting the absence of low riboflavin concentrations was 85%: of the 26 adolescents with normal riboflavin concentrations, 22 did not have AS. The PPV for low riboflavin concentrations was 89%: of the 38 adolescents with AS, 34 had low riboflavin concentrations.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
AS is a classic feature of riboflavin deficiency, but it is nonspecific and often is overlooked as a clinical sign. Although we lacked data on the interobserver agreement in assessment of AS, we did evaluate whether AS is a useful screening measure for low riboflavin concentrations in community assessments or surveillance. In general, a screening measure should be both inexpensive and easy to administer, should impose minimal discomfort on the participants, and should be valid and reliable (21). As a screening measure, AS easily meets the first 3 requirements. We found that AS had high specificity and high PPV, but low sensitivity in detecting low riboflavin concentrations. Because PPV increases with the prevalence of the condition, the PPV found here is relatively high, whereas in populations in whom the prevalence of AS is lower, the PPV would be expected to be lower (21). The high PPV of AS in this survey (89%) shows that AS can be used in conjunction with other relevant nutrient data (eg, data on the nutrient composition of the food basket) as a surveillance tool to indicate marginal riboflavin concentrations in refugee or displaced populations. The low sensitivity of AS, however, limits its utility as an individual screening measure for low riboflavin concentrations.
Low riboflavin concentrations were common in our study sample. With the use of an EGRAC cutoff of >1.7, 86% of adolescents were found to have low riboflavin concentrations. This prevalence is similar to that found in a study of low-income urban schoolchildren in Hyderabad, India, in which 90% were determined to be riboflavin deficient by use of an EGRAC cutoff of >1.4 (24, 25). Although the riboflavin values in our study may have been affected by changes in storage temperature during the transport of the specimens from Nepal to the United States, the association between EGRAC and AS should have been independent of the effects of transport, because all specimens were treated similarly.
The intake of riboflavin-rich foods in the previous week was not consistently related to low riboflavin concentrations. This is not surprising, because the dietary approach used was not intended to assess individual micronutrient intakes or to quantify intakes of riboflavin, but rather to obtain a crude measure of food consumption patterns and food availability. Because AS probably results from chronically inadequate intakes of riboflavin, this crude measure of food consumption during the prior week does not accurately assess the relation between outcome and long-term diet. More detailed dietary nutrient intake methods are time-consuming and were beyond the scope of this survey.
Withdrawal of the cereal blend from the general ration in January 1999 reduced the riboflavin content of the daily ration from <0.6 to 0.4 mg per person, far below the World Health Organization–recommended amounts of 1.35–1.8 mg/d for adolescents (26). The general ration food basket, which contains <2100 kcal per person per day, includes the bulk commodities (rice, vegetable oil, lentils, sugar, and salt) and complementary foods [potatoes, onions, green chili peppers, bananas (or, when available, pumpkin or cabbage), garlic, and turmeric]. Unfortunately, dairy products and meat and fish, good sources of riboflavin, are not part of the general ration. Despite the lack of green vegetables in the ration, 60% of adolescents reported consuming green vegetables in the previous week; however, the green vegetables routinely found in the refugee camp markets (cabbage, okra, green beans, and chili peppers) contain only marginal amounts of riboflavin.
Although periodic nutrition surveys of children aged <5 y are routinely carried out in displaced populations, nutritional assessment of adolescents is rare. The biochemical data presented here suggest that the nutrient status of the adolescent Bhutanese refugees is marginal with regard to riboflavin, folate, vitamin B-12, and iron. Both continued surveillance for signs of micronutrient deficiency and an increase in food sources of micronutrients, especially B vitamins, are needed in this population. Steps to increase nutritional status include the addition of a fortified source of micronutrients and the expansion of kitchen gardens and poultry raising. Since this investigation, vegetable production in camp gardens has been expanded, and whole lentils have been added to the ration. We are unsure of future health effects in the adolescents if riboflavin deficiency continues, but low folate concentrations and low tissue iron stores were also common. Along with marginal vitamin B-12 concentrations, low folate concentrations may interfere with DNA synthesis and could lead to an increased prevalence of anemia. The high prevalence of AS observed in the adolescent Bhutanese refugee population is representative of the precarious nutritional status of displaced populations and their vulnerability to minor changes in rations.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  T. D. Dye Contemporary Prevalence and Prevention of Micronutrient Deficiencies in Refugee Settings Worldwide Journal of Refugee Studies, March 1, 2007; 20(1): 108 - 119. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Graham, J. M. Peerson, M. J. Haskell, R. K. Shrestha, K. H. Brown, and L. H. Allen Erythrocyte Riboflavin for the Detection of Riboflavin Deficiency in Pregnant Nepali Women Clin. Chem., November 1, 2005; 51(11): 2162 - 2165. [Full Text] [PDF]
|
|
|
|
 |
|
 A. J. Seal, P. I. Creeke, Z. Mirghani, F. Abdalla, R. P. McBurney, L. S. Pratt, D. Brookes, L. J. Ruth, and E. Marchand Iron and Vitamin A Deficiency in Long-Term African Refugees J. Nutr., April 1, 2005; 135(4): 808 - 813. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. K. Rogers, T. Tamura, B. J. Rogers, S. E. Welty, T. N. Hansen, and C. V. Smith Analyses of Glutathione Reductase Hypomorphic Mice Indicate a Genetic Knockout Toxicol. Sci., December 1, 2004; 82(2): 367 - 373. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Burgess, M. O'Connell-Motherway, W. Sybesma, J. Hugenholtz, and D. van Sinderen Riboflavin Production in Lactococcus lactis: Potential for In Situ Production of Vitamin-Enriched Foods Appl. Envir. Microbiol., October 1, 2004; 70(10): 5769 - 5777. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
