- Email: [email protected]
Vitamin A status and factors associated in healthy school-age children
Clinical Nutrition 33 (2014) 509e512
Contents lists available at ScienceDirect
Clinical Nutrition journal homepage: http://www.elsevier.com/locate/clnu
Original article
Vitamin A status and factors associated in healthy school-age children Dogus Vuralli a, *, Leyla Tumer b, Alev Hasanoglu b, Gürsel Biberoglu b, Hatice Pasaoglu c a
Department of Pediatric Endocrinology, Hacettepe University, Ankara, Turkey Department of Pediatric Metabolism and Nutrition, Gazi University, Ankara, Turkey c Department of Biochemistry, Gazi University, Ankara, Turkey b
a r t i c l e i n f o
s u m m a r y
Article history: Received 20 December 2012 Accepted 7 July 2013
Background & aims: Vitamin A deficiency (VAD) is one of the most widespread vitamin deficiencies. Vitamin A is essential for children in order to ensure a healthy life span and sustain the normal growth and development. Aim of this study is to examine vitamin A status, and factors associated with it, in healthy school-age children. Methods: The study was carried out in schools in Altindag, the district of Ankara, from April to May 2009. 585 girls and 478 boys, a total of 1063 healthy children aged 5e16 years were taken into the study. Serum retinol, ferritin and hs-CRP levels and complete blood count of each case were measured. A questionnaire was developed to collect socio-economic and demographic information of the participants. Results: Any subclinical VAD (SRL <0.7 mmol/L) was not detected in the children attending the study. However, SRLs were suboptimal in 2.2% of cases and these children were under a high risk of developing subclinical, and subsequently clinical VAD. There were significant positive correlations between serum retinol and hemoglobin values, and statistically significant negative correlation between serum retinol level and ferritin and hsCRP. Conclusions: Vitamin A deficiency does not constitute an important public health problem for Altindag, Ankara, Turkey. Frequency of such vitamin deficiencies should be revealed before launching nationwide public health programs to fight with these deficiencies. Ó 2013 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.
Keywords: Deficiency Retinol Turkey Vitamin A
1. Introduction Today, it has been understood that vitamin and mineral deficiencies are problems which are far more important than they are envisioned. Recent studies revealed that clinical findings of the deficiencies all in all constitute only the tip of the iceberg. It is known that even mild degrees of the deficiencies, which do not cause any clinical symptoms, affect human health negatively. In the last few years, subclinical vitamin and mineral deficiencies have been gradually gaining a bigger importance as a public health problem and drawing attention of the health authorities. It has been found out that detecting and treating people having deficiency symptoms alone is no longer sufficient. Initiatives towards protecting the entire population should be implemented in order to combat this global problem standing in front of us. Vitamin A deficiency (VAD) is one of the most widespread vitamin deficiencies. Detecting and preventing this deficiency
* Corresponding author. Tel.: þ90 312 3051124. E-mail addresses: [email protected], (D. Vuralli).
[email protected]
before it displays clinical findings is important in terms of public health. In this study, we aim to evaluate vitamin A status and the factors affecting it in healthy school-age children.
2. Material and method Primary school students, aged 5e16, attending schools in Altindag, district of Ankara were included in the study. Necessary legal permissions and approval of the ethics committee were taken prior to the launch of the study. Ten random schools settled in Altindag were chosen from the list of schools recorded in the internet page of Ministry of Education. Questionnaires were distributed to these ten schools to a total of 5000 children. Children took the questionnaires to home and the forms were filled by their parents and collected after one week. 90% of the children agreed to participate into the study and filled the questionnaires but due to financial reasons blood samples were taken from a total of 1250 children. 187 children were not included in the analysis due to some missing laboratory values. After excluding those children who have missing values, a total of 1063 healthy children, 585 females and 478 males, were included in the study on a voluntary basis with
0261-5614/$ e see front matter Ó 2013 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved. http://dx.doi.org/10.1016/j.clnu.2013.07.008
510
D. Vuralli et al. / Clinical Nutrition 33 (2014) 509e512
written approval of their parents. Children having chronic diseases were not included. Socio-demographic characteristics and conditions that can affect the vitamin A status such as diet, infection, and vaccination were inquired using a survey form. The survey included questions regarding the parents such as education levels of the parents, mother’s age, monthly income of the family, place of living in the last one year, number of individuals living at the same home and number of children under the age of five living at the same home as wells as questions regarding the child such as age, gender, scheme and duration of vitamin A usage if any. Serum retinol levels (SRLs) were measured using HPLC method with ‘Complete Kit for Vitamins A and E in Plasma’. A SRL between 0.36 and 0.7 mmol/L was considered as subclinical deficiency while a level 0.35 mmol/L was considered as clinical deficiency. A level of 0.71e1.04 mmol/L was considered as suboptimal while a level 1.05 mmol/L was considered as normal vitamin A level.1 Ferritin levels were measured using an ADVIA Centaur ferritin kit and chemiluminometric technique. Normal ferritin values were considered as 7e140 ng/mL for the ages 6 months to 15 years.2 Measurement of hsCRP levels was carried out using immunonephelometric method in a BNProSpec device. Normal hsCRP values were considered as 0e0.3 mg/dl.3 Complete blood counts were carried out automatically in a Cell Dyn 3700 (Abbott) hemogram device. Normal hemoglobin (Hb) values were considered as 11.5e 15.5 g/dl, 13e16 g/dl, and 12e16 g/dl for all subjects aged 6e12 years, for males aged 12e18 years, and for females aged 12e18 years respectively.4 3. Statistical analysis Data analysis was carried out in the SPSS 11.5 for Windows software. It was investigated whether continuous variables were in accordance with normal values using Shapiro Wilk test. Definitive statistics were indicated as mean standard deviation or median (minemax) for continuous variables while categorical variables were indicated as number of subjects and percent (%). Importance of the difference between the groups in terms of mean values was evaluated with Student’s t test while importance of the difference in terms of median values was evaluated with Mann Whitney U test when independent group number was two and with Kruskal Wallis test when there were more than two independent groups. In case statistical results of the Kruskal Wallis test were found significant, a non-parametric multiple comparison test was used in order to determine the groups causing the significant difference. Categorical variables were evaluated using Pearson’s ChieSquare test or Fisher’s Exact ChieSquare test. It was investigated whether there was a significant correlation between continuous variables using Spearman’s Correlation test. Results were considered statistically significant for p < 0.05. 4. Results A total of 1063 healthy children, 585 females and 478 males, aged 5e16 were included in the study. Mean age of the children included in the study was 10.2 2.5 years. Demographic characteristics of the subjects are provided in Table 1 and laboratory measurement values are given in Table 2. Median SRL value was detected to be 2.08 (range 0.80e7.78) mmol/L. Vitamin A levels of the 2.2% of the subjects were suboptimal. Distribution of the subjects according to the laboratory criteria is provided in Table 3. A significant correlation was detected between SRLs and anthropometric measurements and between SRLs and Hb values, while a reverse correlation was detected between SRLs, ferritin, and hsCRP (Table 4).
Table 1 Demographic characteristics of the subjects involved in the study. Features
n ¼ 1063
Age (mean, year) Age groups 5e8 years 9e12 years 13e16 years Gender Female Male Body weight (median kg) (range) Length (median cm) (range) Body mass index (median kg/m2) (range)
10.2 2.5 419 (39.4%) 445 (41.9%) 199 (18.7%) 585 (55.0%) 478 (45.0%) 30 (14e87) 137 (101e182) 16.9 (11.4e34.7)
The parents were asked to indicate their children’s consumption of foods such as green leafy vegetables, yellow-orange colored fruits, eggs, milk, cheese, liver, meat, and fish in terms of number of meals consumed per week in order to evaluate their diets regarding the intake of food that is rich in vitamin A content. A significant correlation was detected only between SRL and consumption of green leafy vegetables. A history of taking vitamin A support was present in 26.8% (285/1063) of the subjects. Median SRL was detected to be 2.21 (range 0.80e5.30)mmol/L in the subjects with a history of taking vitamin support while it was found as 2.05 (range 0.81e7.78)mmol/L in those with no vitamin support. The difference between these two groups was statistically significant. Children who were included in the study were inquired regarding whether they had upper and lower respiratory tract infections (RTI), fever, diarrhea, measles, and chicken pox in the last two weeks. In addition, status of having diarrhea in the last one year was inquired separately as it impairs the intestinal absorption and affects serum levels for a long time. Any difference was not detected in terms of SRLs between the subjects who had recent diseases. 5. Discussion In the recent years, subclinical vitamin deficiencies have been gaining importance gradually. While number of the children affected by clinical VAD has been on the decline, subclinical deficiency is still encountered frequently. The deficiency follows a subclinical course in 90% of the children affected by VAD. According to the World Health Organization report published in 1995, 254 million children have subclinical deficiencies.5 Studies performed in children at school age regarding subclinical VAD are presented in Table 5. The rate of subclinical VAD in the children at school age has been found as 20% in Bangladesh, 23% in Mexico, 3% in Thailand, and 0.9% in England.1,6 There is not any nationwide study providing the prevalence of VAD in Turkey. Studies performed regarding VAD were mostly carried out in the children aged 5 years and under as they constitute the risk group. There is very little data associated with school-age children in the literature. There are a limited number of studies
Table 2 Laboratory measurement values of the subjects. Laboratory measurements
n ¼ 1063 Median (range) or mean SD
Retinol (mmol/L) Ferritin (ng/ml) Hemoglobin (gr/dl) White blood cell count (cell/mm3) hs-CRP (mg/dl)
2.08 (0.80e7.78) 19.8 (0.01e97.7) 13.3 1.18 7480 (2570e17,200) 0.035 (0.015e0.935)
D. Vuralli et al. / Clinical Nutrition 33 (2014) 509e512 Table 3 Distribution of the subjects according to the laboratory criteria. Laboratory criteria
n ¼ 1063
Suboptimal serum retinol levels Low hemoglobin values Above-normal white blood cell count Above-normal hs-CRP values
23 (2.2%) 78 (7.3%) 10 (0.9%) 567 (53.3%)
worldwide demonstrating the VAD in a biochemical way by measuring the SRLs. Most of these studies are of hospital origin and their samples sizes are limited. There are very few studies performed by screening the general population. There is a need for this kind of studies demonstrating the prevalence of subclinical deficiencies both in our country and in the world. Such studies will provide information regarding whether this kind of deficiencies constitute a health problem and they will lead the way in combating these deficiencies. Serum ferritin, Hb, and hsCRP levels were also studied and relation of these factors with SRL was identified in this study. Many factors that can be associated with VAD, such as gender, age, sociodemographic characteristics, socio-economic status of the family, nutritional habits of the children, and recent infections were also investigated. Evaluating the relations between vitamin A level and certain predefined factors and revealing the prevalence of subclinical VAD in children at school age, this study is the most comprehensive epidemiological study performed in our country and it is one of the most comprehensive ones performed around the world. Turkey is still a developing country with its population of 71 millions. Seventy point five percent of the population lives in the cities. Half of Turkey’s population is under 28.3 years old. Ankara is the capital city of Turkey and Altindag is the 6th biggest district of Ankara in terms of population with its population of 370.000. It is one of the oldest districts of Ankara in terms of location. Unemployment and poverty are most important problems of the district. This district was chosen as the investigation site since it was thought that vitamin and mineral deficiencies could be encountered more frequently in this district due to its poor socio-economic conditions. In this study, any subclinical VAD (SRL<0.7 mmol/L) was not detected. However, SRLs were between 0.7 and 1.04 mmol/L in 23 children (2.2%). This interval is defined as suboptimal vitamin A level and these children are under a high risk of developing subclinical, and subsequently clinical VAD. This study demonstrates that VAD currently is not a serious public health problem in Ankara. While subclinical VAD was detected between 2 and 22% in the other studies carried out in school-age children around the world, this study demonstrates that any subclinical VAD was not encountered in the same age group in our country.
Table 4 Correlation coefficients and importance levels between serum retinol level and anthropometric and laboratory measurements. Retinol level
Age Body weight Height Body mass index Ferritin Hemoglobin White blood cell count hs-CRP
R
P
0.303 0.288 0.299 0.174 0.061 0.228 0.057 0.149
<0.001 <0.001 <0.001 <0.001 0.049 <0.001 0.061 <0.001
511
Table 5 Studies performed in children who are at school age regarding subclinical vitamin A deficiencies. Investigator
Year
Country
Number Age Serum of cases (years) retinol level (mmol/L)
Ratio of subclinical deficiency (%)
National dataa Thaneb National dataa National dataa Villalpandoc Ahmedd Macias-Matose
1988 1997 1998 1998 1999 2000 2002
USA England Bangladesh Mexica Mexica Bangladesh Cuba
1865 966 159 1709 1966 381 Total 1191 463 364 364 518 155 1063
1.34 1.47 0.93 0.85 0.85 1.30
0 0.9 20 23 21.9 1.5
1.40 1.77 2.01 1.31 0.82 2.08
2 0 0.6 3 33.5 0
National dataa Graebnerf Vuralli (results of this study)
East West Middle 2002 Thailand 2003 Brazil 2010 Turkey
9e14 4e18 6e13 6e12 0e12 11e16 6e11
6e13 5e18 5e16
a Global prevalence of vitamin A deficiency in populations at risk 1995e2005. Geneva, World Health Organization, 2009. b Thane CW, Bates CJ, Prentice A. Zinc and vitamin A intake and status in a national sample of British young people aged 4e18 years. European Journal of clinical nutrition 2004; 58:363e375. c Villalpando S, Montalvo-Velarde I, Zambrano N, Garcia-Guerra A, Ramirez-Silva C I, Shamah-Levy T et al. Vitamins A and C and folate status in Mexican children under 12 years and women 12e49 years: A probalistic national survey. Salud Publica Mex 2003; 45(4):S508e519. d Ahmed F, Rahman A, Noor AN, Akhtaruzzaman M, Hughes R. Anaemia and vitamin A status among adolescent schoolboys an Dhaka City, Bangladesh. Public Health Nutrition 2005; 963:345e350. e Macias-Matos C, Pita-Rodriguez G, Monterrey-Gutierrez P, Reboso-Perez J. Vitamin A status in Cuban children aged 6e11 years. Public Health Nutrition 2007; 11(1):95e101. f Graebner IT, Saito CH, Souza EM. Biochemical assessment of vitamin A in schoolchildren from a rural community. Jornal de Pediatria (Rio J) 2007; 83(3):247e 252.
Median SRL was detected to be 2.08 mmol/L (0.80e7.78) in this study. This value is considerably higher compared to the other studies. Due to their increased growth rates, vitamin A needs of males, and young children are higher and these groups are under a higher risk in terms of deficiency of this micronutrient. While median SRL of the males was detected to be 2.04 (range 0.81e5.08) mmol/L, the same level was found as 2.11 (range 0.80e7.78) mmol/L in the females and this difference was found significant (p < 0.05). A correlation was detected between age and SRLs in this study (r ¼ 0.303 p < 0.001). This correlation is significant as VAD is seen more frequently in little children and retinol storages of the children increase with age. A correlation was also demonstrated between SRL and child’s anthropometric measurements such as body weight, height, and body mass index. (p < 0.001). Settlement is another risk factor for VAD. It is easier for the individuals living in the cities to obtain correct information about nutrition and reach foods that are rich in vitamin A content and access health services. Dietary habits and undergone diseases negatively affect protein, vitamin, and mineral levels of the children living in rural areas. Protein-energy malnutrition is more frequent in the children living in rural areas. There are clinical studies demonstrating that low protein levels impair RBP synthesis in liver. In the studies carried out, it was demonstrated that children belonging to families living in rural areas with a low socioeconomic status and the children with insufficiently educated parents are more frequently affected by clinical or subclinical VAD.7 In this study, any correlation was not detected between vitamin A levels and socioeconomical indicators. We chose Altindag district, an urban area, as the study site since it would be easy for us to reach
512
D. Vuralli et al. / Clinical Nutrition 33 (2014) 509e512
to the children at that area. In this area the families with lower socio-economical level live and it would be more possible for us to find a vitamin deficiency if there is any. Since we did not find any deficiency at that area, the likelihood of finding any deficiency in other urban areas in Ankara is unlikely. 6. Factors affecting the vitamin A status 6.1. Relation of vitamin A and infections Vitamin A blood levels should be evaluated together with acute phase reactants such as C-reactive protein (CRP) and white blood cell counts because infections and inflammations affect nutrient blood levels without causing a change in the body storages. Increased plasma CRP concentrations and increased white blood cell counts are associated with reduced plasma retinol levels.8 Enteric infections can change the absorption surface area and inflammatory systemic diseases with fever increase the urinary loss and metabolic rate. Measles infection impairs the vitamin A metabolism and causes a reduction in SRL.9 Since SRLs are affected by infections, children having clinical infections were not included in the study and the children who are included were inquired regarding recently undergone infections. Histories of upper and lower RTIs, and diarrhea undergone in the last two weeks were detected in 30%, 1.7%, and 7.9% of the children respectively. Any correlation was not demonstrated between SRL and recently undergone diseases. Since data for these diseases was obtained from the survey filled in by the parents, it might not include sufficient and reliable information and this can be one of the reasons why any correlation was not found. This study, which aims to screen subclinical VAD, was carried out in healthy individuals and children having clinical infections were not included in the study. However, subclinical infections might be present in these individuals. Children may seem like healthy individuals when findings are not adequately prominent at the beginning of the infection or during the incubation period. In this study, hsCRP levels and WBC counts were measured in order to demonstrate the presence of subclinical infection. In 53.3% of the children hsCRP and in 0.9% of the children WBC count was found elevated and this was interpreted as subclinical infection presence. The increase in acute phase response detected in subclinical infection causes a reduction in plasma retinol level. Our study supports this correlation as well since a reverse correlation was detected between SRL and hsCRP (r ¼ 0.149 p < 0.001). 6.2. Relation of vitamin A and anemia Serum vitamin A level is affected by Hb level and ferritin level. Vitamin A is required for absorption and usage of non-heme iron and takes part in Hb formation. Hb level affects the serum vitamin A level and similarly, vitamin A level affects the Hb level. One unit of change in the Hb level affects the serum vitamin A level 0.004 units in the same direction. VAD causes development of anemia by affecting the iron metabolism. Iron takes part in vitamin A metabolism as well [37]. In our study, a positive correlation was detected between SRL and Hb levels (r ¼ 0.228 p < 0.001 for Hb). When ferritin level is increased, serum vitamin A level reduces. VAD affects the usage of stored iron by haematopoietic tissue and ferritin level increases in case of VAD. This reverse correlation between SRL and ferritin is demonstrated in this study (r ¼ 0.061 p ¼ 0 < 0.5).
As a result, this study is the most comprehensive epidemiological study identifying the frequency of VAD in healthy schoolage children in Turkey. Vitamin A was found to be in suboptimal level with a rate of 2.2% and VAD does not constitute an important public health problem for the Altindag district of Ankara in which this study was performed. Vitamin and mineral deficiency problems are to a great extent under control in the developed countries. In developing countries such as Turkey, the goal should be taking these deficiencies under control as well. First of all, frequency of these deficiencies should be revealed before launching nationwide public health programs with this purpose. There is a need for examining population samples with different characteristics with adequate number of subjects in order to determine the frequency of micronutrient deficiencies. Statement of authorship Dogus Vuralli: Made the literature search, the design of the study, data collection, data analysis and writing of the study. Leyla Tumer: Made the design, data interpretation and helped in writing the manuscript. Alev Hasanoglu: Made the design, data interpretation and helped in writing the manuscript. Gursel Biberoglu: Made the laboratory analysis. Hatice Pasaoglu: Made the laboratory analysis. Conflict of interest None. Funding sources The financial support for the laboratory kits was provided by ‘The Danone Institute of Turkey’. There’s no other conflict of interest. Acknowledgments The financial support of this study was provided by ‘The Danone Institute of Turkey’. References 1. Thane CW, Bates CJ, Prentice A. Zinc and vitamin A intake and status in a national sample of British young people aged 4e18 y. Eur J Clin Nutr 2004;58(2):363e75. 2. Franco CD. Ferritin. In: Kaplan LA, Pesce AJ, editors. Methods in clinical chemistry. 2nd ed. St. Louis: CV Mosby; 1987. p. 1240e2. 3. Chenillot O, Henry J, Steinmetz J, Herbeth B, Wagner C, Siest G. High sensitivity C-reactive protein: biological variations and reference limits. Clin Chem Lab Med 2000;38(10):1003e11. 4. Behrman RE, Kliegman RE, Jenson HB. Nelson textbook of pediatrics. 17th ed. Philadelphia: PA Saunders Company; 2004. p. 2398e9. 5. World Health Organization. Indicators for assessing VAD and their application in monitoring and evaluating intervention programmes; 1996 [Geneva]. 6. World Health Organization. Global prevalence of VAD in populations at risk 1995e 2005. Geneva: WHO Global Database on VAD; 2009. 7. Nojilana B, Norman R, Bradshaw D, van Stuijvenberg ME, Dhansay MA, Labadarios D. Estimating the burden of disease attributable to VAD in South Africa in 2000. S Afr Med J 2007;97(8):748e53. 8. Paracha PI, Jamil A, Northrop-Clewes CA, Thurnham DI. Interpretation of vitamin A status in apparently healthy Pakistani children by using markers of subclinical infection. Am J Clin Nutr 2000;72:1164e9. 9. Hussey GD, Klein M. A randomized controlled trial of vitamin A in children with severe measles. N Engl J Med 1990;323:160e4.
Contents lists available at ScienceDirect
Clinical Nutrition journal homepage: http://www.elsevier.com/locate/clnu
Original article
Vitamin A status and factors associated in healthy school-age children Dogus Vuralli a, *, Leyla Tumer b, Alev Hasanoglu b, Gürsel Biberoglu b, Hatice Pasaoglu c a
Department of Pediatric Endocrinology, Hacettepe University, Ankara, Turkey Department of Pediatric Metabolism and Nutrition, Gazi University, Ankara, Turkey c Department of Biochemistry, Gazi University, Ankara, Turkey b
a r t i c l e i n f o
s u m m a r y
Article history: Received 20 December 2012 Accepted 7 July 2013
Background & aims: Vitamin A deficiency (VAD) is one of the most widespread vitamin deficiencies. Vitamin A is essential for children in order to ensure a healthy life span and sustain the normal growth and development. Aim of this study is to examine vitamin A status, and factors associated with it, in healthy school-age children. Methods: The study was carried out in schools in Altindag, the district of Ankara, from April to May 2009. 585 girls and 478 boys, a total of 1063 healthy children aged 5e16 years were taken into the study. Serum retinol, ferritin and hs-CRP levels and complete blood count of each case were measured. A questionnaire was developed to collect socio-economic and demographic information of the participants. Results: Any subclinical VAD (SRL <0.7 mmol/L) was not detected in the children attending the study. However, SRLs were suboptimal in 2.2% of cases and these children were under a high risk of developing subclinical, and subsequently clinical VAD. There were significant positive correlations between serum retinol and hemoglobin values, and statistically significant negative correlation between serum retinol level and ferritin and hsCRP. Conclusions: Vitamin A deficiency does not constitute an important public health problem for Altindag, Ankara, Turkey. Frequency of such vitamin deficiencies should be revealed before launching nationwide public health programs to fight with these deficiencies. Ó 2013 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.
Keywords: Deficiency Retinol Turkey Vitamin A
1. Introduction Today, it has been understood that vitamin and mineral deficiencies are problems which are far more important than they are envisioned. Recent studies revealed that clinical findings of the deficiencies all in all constitute only the tip of the iceberg. It is known that even mild degrees of the deficiencies, which do not cause any clinical symptoms, affect human health negatively. In the last few years, subclinical vitamin and mineral deficiencies have been gradually gaining a bigger importance as a public health problem and drawing attention of the health authorities. It has been found out that detecting and treating people having deficiency symptoms alone is no longer sufficient. Initiatives towards protecting the entire population should be implemented in order to combat this global problem standing in front of us. Vitamin A deficiency (VAD) is one of the most widespread vitamin deficiencies. Detecting and preventing this deficiency
* Corresponding author. Tel.: þ90 312 3051124. E-mail addresses: [email protected], (D. Vuralli).
[email protected]
before it displays clinical findings is important in terms of public health. In this study, we aim to evaluate vitamin A status and the factors affecting it in healthy school-age children.
2. Material and method Primary school students, aged 5e16, attending schools in Altindag, district of Ankara were included in the study. Necessary legal permissions and approval of the ethics committee were taken prior to the launch of the study. Ten random schools settled in Altindag were chosen from the list of schools recorded in the internet page of Ministry of Education. Questionnaires were distributed to these ten schools to a total of 5000 children. Children took the questionnaires to home and the forms were filled by their parents and collected after one week. 90% of the children agreed to participate into the study and filled the questionnaires but due to financial reasons blood samples were taken from a total of 1250 children. 187 children were not included in the analysis due to some missing laboratory values. After excluding those children who have missing values, a total of 1063 healthy children, 585 females and 478 males, were included in the study on a voluntary basis with
0261-5614/$ e see front matter Ó 2013 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved. http://dx.doi.org/10.1016/j.clnu.2013.07.008
510
D. Vuralli et al. / Clinical Nutrition 33 (2014) 509e512
written approval of their parents. Children having chronic diseases were not included. Socio-demographic characteristics and conditions that can affect the vitamin A status such as diet, infection, and vaccination were inquired using a survey form. The survey included questions regarding the parents such as education levels of the parents, mother’s age, monthly income of the family, place of living in the last one year, number of individuals living at the same home and number of children under the age of five living at the same home as wells as questions regarding the child such as age, gender, scheme and duration of vitamin A usage if any. Serum retinol levels (SRLs) were measured using HPLC method with ‘Complete Kit for Vitamins A and E in Plasma’. A SRL between 0.36 and 0.7 mmol/L was considered as subclinical deficiency while a level 0.35 mmol/L was considered as clinical deficiency. A level of 0.71e1.04 mmol/L was considered as suboptimal while a level 1.05 mmol/L was considered as normal vitamin A level.1 Ferritin levels were measured using an ADVIA Centaur ferritin kit and chemiluminometric technique. Normal ferritin values were considered as 7e140 ng/mL for the ages 6 months to 15 years.2 Measurement of hsCRP levels was carried out using immunonephelometric method in a BNProSpec device. Normal hsCRP values were considered as 0e0.3 mg/dl.3 Complete blood counts were carried out automatically in a Cell Dyn 3700 (Abbott) hemogram device. Normal hemoglobin (Hb) values were considered as 11.5e 15.5 g/dl, 13e16 g/dl, and 12e16 g/dl for all subjects aged 6e12 years, for males aged 12e18 years, and for females aged 12e18 years respectively.4 3. Statistical analysis Data analysis was carried out in the SPSS 11.5 for Windows software. It was investigated whether continuous variables were in accordance with normal values using Shapiro Wilk test. Definitive statistics were indicated as mean standard deviation or median (minemax) for continuous variables while categorical variables were indicated as number of subjects and percent (%). Importance of the difference between the groups in terms of mean values was evaluated with Student’s t test while importance of the difference in terms of median values was evaluated with Mann Whitney U test when independent group number was two and with Kruskal Wallis test when there were more than two independent groups. In case statistical results of the Kruskal Wallis test were found significant, a non-parametric multiple comparison test was used in order to determine the groups causing the significant difference. Categorical variables were evaluated using Pearson’s ChieSquare test or Fisher’s Exact ChieSquare test. It was investigated whether there was a significant correlation between continuous variables using Spearman’s Correlation test. Results were considered statistically significant for p < 0.05. 4. Results A total of 1063 healthy children, 585 females and 478 males, aged 5e16 were included in the study. Mean age of the children included in the study was 10.2 2.5 years. Demographic characteristics of the subjects are provided in Table 1 and laboratory measurement values are given in Table 2. Median SRL value was detected to be 2.08 (range 0.80e7.78) mmol/L. Vitamin A levels of the 2.2% of the subjects were suboptimal. Distribution of the subjects according to the laboratory criteria is provided in Table 3. A significant correlation was detected between SRLs and anthropometric measurements and between SRLs and Hb values, while a reverse correlation was detected between SRLs, ferritin, and hsCRP (Table 4).
Table 1 Demographic characteristics of the subjects involved in the study. Features
n ¼ 1063
Age (mean, year) Age groups 5e8 years 9e12 years 13e16 years Gender Female Male Body weight (median kg) (range) Length (median cm) (range) Body mass index (median kg/m2) (range)
10.2 2.5 419 (39.4%) 445 (41.9%) 199 (18.7%) 585 (55.0%) 478 (45.0%) 30 (14e87) 137 (101e182) 16.9 (11.4e34.7)
The parents were asked to indicate their children’s consumption of foods such as green leafy vegetables, yellow-orange colored fruits, eggs, milk, cheese, liver, meat, and fish in terms of number of meals consumed per week in order to evaluate their diets regarding the intake of food that is rich in vitamin A content. A significant correlation was detected only between SRL and consumption of green leafy vegetables. A history of taking vitamin A support was present in 26.8% (285/1063) of the subjects. Median SRL was detected to be 2.21 (range 0.80e5.30)mmol/L in the subjects with a history of taking vitamin support while it was found as 2.05 (range 0.81e7.78)mmol/L in those with no vitamin support. The difference between these two groups was statistically significant. Children who were included in the study were inquired regarding whether they had upper and lower respiratory tract infections (RTI), fever, diarrhea, measles, and chicken pox in the last two weeks. In addition, status of having diarrhea in the last one year was inquired separately as it impairs the intestinal absorption and affects serum levels for a long time. Any difference was not detected in terms of SRLs between the subjects who had recent diseases. 5. Discussion In the recent years, subclinical vitamin deficiencies have been gaining importance gradually. While number of the children affected by clinical VAD has been on the decline, subclinical deficiency is still encountered frequently. The deficiency follows a subclinical course in 90% of the children affected by VAD. According to the World Health Organization report published in 1995, 254 million children have subclinical deficiencies.5 Studies performed in children at school age regarding subclinical VAD are presented in Table 5. The rate of subclinical VAD in the children at school age has been found as 20% in Bangladesh, 23% in Mexico, 3% in Thailand, and 0.9% in England.1,6 There is not any nationwide study providing the prevalence of VAD in Turkey. Studies performed regarding VAD were mostly carried out in the children aged 5 years and under as they constitute the risk group. There is very little data associated with school-age children in the literature. There are a limited number of studies
Table 2 Laboratory measurement values of the subjects. Laboratory measurements
n ¼ 1063 Median (range) or mean SD
Retinol (mmol/L) Ferritin (ng/ml) Hemoglobin (gr/dl) White blood cell count (cell/mm3) hs-CRP (mg/dl)
2.08 (0.80e7.78) 19.8 (0.01e97.7) 13.3 1.18 7480 (2570e17,200) 0.035 (0.015e0.935)
D. Vuralli et al. / Clinical Nutrition 33 (2014) 509e512 Table 3 Distribution of the subjects according to the laboratory criteria. Laboratory criteria
n ¼ 1063
Suboptimal serum retinol levels Low hemoglobin values Above-normal white blood cell count Above-normal hs-CRP values
23 (2.2%) 78 (7.3%) 10 (0.9%) 567 (53.3%)
worldwide demonstrating the VAD in a biochemical way by measuring the SRLs. Most of these studies are of hospital origin and their samples sizes are limited. There are very few studies performed by screening the general population. There is a need for this kind of studies demonstrating the prevalence of subclinical deficiencies both in our country and in the world. Such studies will provide information regarding whether this kind of deficiencies constitute a health problem and they will lead the way in combating these deficiencies. Serum ferritin, Hb, and hsCRP levels were also studied and relation of these factors with SRL was identified in this study. Many factors that can be associated with VAD, such as gender, age, sociodemographic characteristics, socio-economic status of the family, nutritional habits of the children, and recent infections were also investigated. Evaluating the relations between vitamin A level and certain predefined factors and revealing the prevalence of subclinical VAD in children at school age, this study is the most comprehensive epidemiological study performed in our country and it is one of the most comprehensive ones performed around the world. Turkey is still a developing country with its population of 71 millions. Seventy point five percent of the population lives in the cities. Half of Turkey’s population is under 28.3 years old. Ankara is the capital city of Turkey and Altindag is the 6th biggest district of Ankara in terms of population with its population of 370.000. It is one of the oldest districts of Ankara in terms of location. Unemployment and poverty are most important problems of the district. This district was chosen as the investigation site since it was thought that vitamin and mineral deficiencies could be encountered more frequently in this district due to its poor socio-economic conditions. In this study, any subclinical VAD (SRL<0.7 mmol/L) was not detected. However, SRLs were between 0.7 and 1.04 mmol/L in 23 children (2.2%). This interval is defined as suboptimal vitamin A level and these children are under a high risk of developing subclinical, and subsequently clinical VAD. This study demonstrates that VAD currently is not a serious public health problem in Ankara. While subclinical VAD was detected between 2 and 22% in the other studies carried out in school-age children around the world, this study demonstrates that any subclinical VAD was not encountered in the same age group in our country.
Table 4 Correlation coefficients and importance levels between serum retinol level and anthropometric and laboratory measurements. Retinol level
Age Body weight Height Body mass index Ferritin Hemoglobin White blood cell count hs-CRP
R
P
0.303 0.288 0.299 0.174 0.061 0.228 0.057 0.149
<0.001 <0.001 <0.001 <0.001 0.049 <0.001 0.061 <0.001
511
Table 5 Studies performed in children who are at school age regarding subclinical vitamin A deficiencies. Investigator
Year
Country
Number Age Serum of cases (years) retinol level (mmol/L)
Ratio of subclinical deficiency (%)
National dataa Thaneb National dataa National dataa Villalpandoc Ahmedd Macias-Matose
1988 1997 1998 1998 1999 2000 2002
USA England Bangladesh Mexica Mexica Bangladesh Cuba
1865 966 159 1709 1966 381 Total 1191 463 364 364 518 155 1063
1.34 1.47 0.93 0.85 0.85 1.30
0 0.9 20 23 21.9 1.5
1.40 1.77 2.01 1.31 0.82 2.08
2 0 0.6 3 33.5 0
National dataa Graebnerf Vuralli (results of this study)
East West Middle 2002 Thailand 2003 Brazil 2010 Turkey
9e14 4e18 6e13 6e12 0e12 11e16 6e11
6e13 5e18 5e16
a Global prevalence of vitamin A deficiency in populations at risk 1995e2005. Geneva, World Health Organization, 2009. b Thane CW, Bates CJ, Prentice A. Zinc and vitamin A intake and status in a national sample of British young people aged 4e18 years. European Journal of clinical nutrition 2004; 58:363e375. c Villalpando S, Montalvo-Velarde I, Zambrano N, Garcia-Guerra A, Ramirez-Silva C I, Shamah-Levy T et al. Vitamins A and C and folate status in Mexican children under 12 years and women 12e49 years: A probalistic national survey. Salud Publica Mex 2003; 45(4):S508e519. d Ahmed F, Rahman A, Noor AN, Akhtaruzzaman M, Hughes R. Anaemia and vitamin A status among adolescent schoolboys an Dhaka City, Bangladesh. Public Health Nutrition 2005; 963:345e350. e Macias-Matos C, Pita-Rodriguez G, Monterrey-Gutierrez P, Reboso-Perez J. Vitamin A status in Cuban children aged 6e11 years. Public Health Nutrition 2007; 11(1):95e101. f Graebner IT, Saito CH, Souza EM. Biochemical assessment of vitamin A in schoolchildren from a rural community. Jornal de Pediatria (Rio J) 2007; 83(3):247e 252.
Median SRL was detected to be 2.08 mmol/L (0.80e7.78) in this study. This value is considerably higher compared to the other studies. Due to their increased growth rates, vitamin A needs of males, and young children are higher and these groups are under a higher risk in terms of deficiency of this micronutrient. While median SRL of the males was detected to be 2.04 (range 0.81e5.08) mmol/L, the same level was found as 2.11 (range 0.80e7.78) mmol/L in the females and this difference was found significant (p < 0.05). A correlation was detected between age and SRLs in this study (r ¼ 0.303 p < 0.001). This correlation is significant as VAD is seen more frequently in little children and retinol storages of the children increase with age. A correlation was also demonstrated between SRL and child’s anthropometric measurements such as body weight, height, and body mass index. (p < 0.001). Settlement is another risk factor for VAD. It is easier for the individuals living in the cities to obtain correct information about nutrition and reach foods that are rich in vitamin A content and access health services. Dietary habits and undergone diseases negatively affect protein, vitamin, and mineral levels of the children living in rural areas. Protein-energy malnutrition is more frequent in the children living in rural areas. There are clinical studies demonstrating that low protein levels impair RBP synthesis in liver. In the studies carried out, it was demonstrated that children belonging to families living in rural areas with a low socioeconomic status and the children with insufficiently educated parents are more frequently affected by clinical or subclinical VAD.7 In this study, any correlation was not detected between vitamin A levels and socioeconomical indicators. We chose Altindag district, an urban area, as the study site since it would be easy for us to reach
512
D. Vuralli et al. / Clinical Nutrition 33 (2014) 509e512
to the children at that area. In this area the families with lower socio-economical level live and it would be more possible for us to find a vitamin deficiency if there is any. Since we did not find any deficiency at that area, the likelihood of finding any deficiency in other urban areas in Ankara is unlikely. 6. Factors affecting the vitamin A status 6.1. Relation of vitamin A and infections Vitamin A blood levels should be evaluated together with acute phase reactants such as C-reactive protein (CRP) and white blood cell counts because infections and inflammations affect nutrient blood levels without causing a change in the body storages. Increased plasma CRP concentrations and increased white blood cell counts are associated with reduced plasma retinol levels.8 Enteric infections can change the absorption surface area and inflammatory systemic diseases with fever increase the urinary loss and metabolic rate. Measles infection impairs the vitamin A metabolism and causes a reduction in SRL.9 Since SRLs are affected by infections, children having clinical infections were not included in the study and the children who are included were inquired regarding recently undergone infections. Histories of upper and lower RTIs, and diarrhea undergone in the last two weeks were detected in 30%, 1.7%, and 7.9% of the children respectively. Any correlation was not demonstrated between SRL and recently undergone diseases. Since data for these diseases was obtained from the survey filled in by the parents, it might not include sufficient and reliable information and this can be one of the reasons why any correlation was not found. This study, which aims to screen subclinical VAD, was carried out in healthy individuals and children having clinical infections were not included in the study. However, subclinical infections might be present in these individuals. Children may seem like healthy individuals when findings are not adequately prominent at the beginning of the infection or during the incubation period. In this study, hsCRP levels and WBC counts were measured in order to demonstrate the presence of subclinical infection. In 53.3% of the children hsCRP and in 0.9% of the children WBC count was found elevated and this was interpreted as subclinical infection presence. The increase in acute phase response detected in subclinical infection causes a reduction in plasma retinol level. Our study supports this correlation as well since a reverse correlation was detected between SRL and hsCRP (r ¼ 0.149 p < 0.001). 6.2. Relation of vitamin A and anemia Serum vitamin A level is affected by Hb level and ferritin level. Vitamin A is required for absorption and usage of non-heme iron and takes part in Hb formation. Hb level affects the serum vitamin A level and similarly, vitamin A level affects the Hb level. One unit of change in the Hb level affects the serum vitamin A level 0.004 units in the same direction. VAD causes development of anemia by affecting the iron metabolism. Iron takes part in vitamin A metabolism as well [37]. In our study, a positive correlation was detected between SRL and Hb levels (r ¼ 0.228 p < 0.001 for Hb). When ferritin level is increased, serum vitamin A level reduces. VAD affects the usage of stored iron by haematopoietic tissue and ferritin level increases in case of VAD. This reverse correlation between SRL and ferritin is demonstrated in this study (r ¼ 0.061 p ¼ 0 < 0.5).
As a result, this study is the most comprehensive epidemiological study identifying the frequency of VAD in healthy schoolage children in Turkey. Vitamin A was found to be in suboptimal level with a rate of 2.2% and VAD does not constitute an important public health problem for the Altindag district of Ankara in which this study was performed. Vitamin and mineral deficiency problems are to a great extent under control in the developed countries. In developing countries such as Turkey, the goal should be taking these deficiencies under control as well. First of all, frequency of these deficiencies should be revealed before launching nationwide public health programs with this purpose. There is a need for examining population samples with different characteristics with adequate number of subjects in order to determine the frequency of micronutrient deficiencies. Statement of authorship Dogus Vuralli: Made the literature search, the design of the study, data collection, data analysis and writing of the study. Leyla Tumer: Made the design, data interpretation and helped in writing the manuscript. Alev Hasanoglu: Made the design, data interpretation and helped in writing the manuscript. Gursel Biberoglu: Made the laboratory analysis. Hatice Pasaoglu: Made the laboratory analysis. Conflict of interest None. Funding sources The financial support for the laboratory kits was provided by ‘The Danone Institute of Turkey’. There’s no other conflict of interest. Acknowledgments The financial support of this study was provided by ‘The Danone Institute of Turkey’. References 1. Thane CW, Bates CJ, Prentice A. Zinc and vitamin A intake and status in a national sample of British young people aged 4e18 y. Eur J Clin Nutr 2004;58(2):363e75. 2. Franco CD. Ferritin. In: Kaplan LA, Pesce AJ, editors. Methods in clinical chemistry. 2nd ed. St. Louis: CV Mosby; 1987. p. 1240e2. 3. Chenillot O, Henry J, Steinmetz J, Herbeth B, Wagner C, Siest G. High sensitivity C-reactive protein: biological variations and reference limits. Clin Chem Lab Med 2000;38(10):1003e11. 4. Behrman RE, Kliegman RE, Jenson HB. Nelson textbook of pediatrics. 17th ed. Philadelphia: PA Saunders Company; 2004. p. 2398e9. 5. World Health Organization. Indicators for assessing VAD and their application in monitoring and evaluating intervention programmes; 1996 [Geneva]. 6. World Health Organization. Global prevalence of VAD in populations at risk 1995e 2005. Geneva: WHO Global Database on VAD; 2009. 7. Nojilana B, Norman R, Bradshaw D, van Stuijvenberg ME, Dhansay MA, Labadarios D. Estimating the burden of disease attributable to VAD in South Africa in 2000. S Afr Med J 2007;97(8):748e53. 8. Paracha PI, Jamil A, Northrop-Clewes CA, Thurnham DI. Interpretation of vitamin A status in apparently healthy Pakistani children by using markers of subclinical infection. Am J Clin Nutr 2000;72:1164e9. 9. Hussey GD, Klein M. A randomized controlled trial of vitamin A in children with severe measles. N Engl J Med 1990;323:160e4.