点击显示 收起
1 From the Department of Ophthalmology, School of Medicine, and the Department of International Health, Johns Hopkins University, Baltimore; the Nutrition Research and Development Centre, Ministry of Health, Bogor, Indonesia; and the Cicendo Eye Hospital, Bandung, Indonesia.
See corresponding editorial on page 1.
ABSTRACT
Background: The relations among hyporetinolemia, acute phase proteins, and vitamin A status in children are unclear.
Objective: The objective was to examine the relations between acute phase proteins and plasma retinol concentrations in children with and without clinical vitamin A deficiency (Bitot spots and night blindness).
Design: The study was a nonconcurrent analysis of acute phase protein concentrations and other data from a previous clinical trial. Preschool children, 3–6 y of age, with (n = 118) and without (n = 118) xerophthalmia were assigned to receive oral vitamin A (60 mg retinol equivalent) or placebo and were seen at 5 wk. All children received oral vitamin A (60 mg retinol equivalent) at 5 wk.
Results: At baseline, 1-acid glycoprotein (AGP) was elevated in 42.9% and 23.5% (P < 0.003) and C-reactive protein (CRP) was elevated in 17.7% and 13.7% (NS) of children with and without xerophthalmia, respectively. Hyporetinolemia (retinol < 0.7 µmol/L) occurred in 61.0% and 47.4% (P < 0.04) of children with and without xerophthalmia, respectively. A history of fever, a history of cough, and nasal discharge noted on examination were each associated with elevated acute phase proteins. Vitamin A supplementation increased plasma retinol at 5 wk but had no significant effect on concentrations of acute phase proteins.
Conclusions: Elevated acute phase protein concentrations and infectious disease morbidity are closely associated during vitamin A deficiency.
Key Words: Vitamin A • acute phase proteins • retinol • xerophthalmia • 1-acid glycoprotein • C-reactive protein • morbidity • infection • preschool children • Indonesia
INTRODUCTION
Vitamin A deficiency is a major public health problem in many developing countries (1), and the identification of high-risk populations remains an important priority. Improvement of vitamin A status has been shown to reduce morbidity and mortality of preschool children (2). Plasma retinol concentrations are widely used to assess vitamin A status in different populations. During infection or trauma, plasma retinol concentrations may decrease because of different factors, including a reduction in hepatic production of retinol binding protein (3), urinary losses of holo–retinol binding protein (4–6), increased uptake of retinol by some peripheral tissues (7), and extravascular losses due to increased vascular permeability (3). The measurement of acute phase proteins, such as 1-acid glycoprotein (AGP), has been considered to improve the interpretation of plasma retinol concentrations in populations with a high prevalence of morbidity (8).
Recently, it was shown that a higher proportion of pregnant women with night blindness in Nepal had elevated acute phase proteins, hyporetinolemia, and infections than did matched control subjects (9). These data suggested that assessment of acute phase protein concentrations did not provide mutually exclusive information for interpretation of circulating vitamin A concentrations in pregnant women because there was a close association between clinical vitamin A deficiency, (ie, night blindness and hyporetinolemia) and infection.
The relation between clinical vitamin A deficiency (Bitot spots, night blindness, or both), hyporetinolemia, and acute phase proteins has not been characterized in preschool children, the other major group at high risk of vitamin A deficiency. We conducted a clinical trial of vitamin A supplementation among preschool children with and without xerophthalmia that allowed us to examine the relation between hyporetinolemia, acute phase proteins, and infectious disease morbidity.
SUBJECTS AND METHODS
A randomized, double-masked, controlled clinical trial was conducted in preschool children aged 3–6 y at the outpatient clinic of the Cicendo Eye Hospital in Bandung, West Java, Indonesia, between June and September 1988. This trial and its immunologic outcome measures were described in detail elsewhere (10, 11). Children with and without night blindness or Bitot spots were identified in a community-wide screening and brought to the outpatient clinic to be examined by an ophthalmologist and a pediatrician. Weight, height, mean upper-arm circumference, and triceps skinfold thickness were measured. Children whose weight-for-height was below 80% of the National Center for Health Statistics median (12) were excluded from the study. Any children with serious illness, such as diarrheal disease, pneumonia, or fever on presentation, were excluded from the study and treated appropriately. Children were enrolled in the study after written, informed consent was received from the parent or guardian. The study protocol was approved by the ethical review committees at the Office of Research and Development and the Nutrition Research and Development Centre, Ministry of Health, Government of Indonesia, the Cicendo Eye Hospital, and the Johns Hopkins University School of Medicine, Baltimore.
The clinical trial design involved 4 groups of children. Each child with xerophthalmia (Bitot spots, night blindness, or both) was matched by age (±6 mo), sex, and community with one other child with xerophthalmia and 2 children without xerophthalmia. During the first visit, the children were randomly assigned to receive oral vitamin A [60 mg retinol equivalent (RE)] or placebo. The vitamin A and placebo solutions were supplied in coded containers and the treatment code was broken at the end of the study. All children in the study received oral vitamin A (60 mg RE) at 5 wk. At baseline and 5 wk, a blood sample was drawn by venipuncture, the plasma was separated by centrifugation at 1000 x g for 10 min at room temperature, multiple 0.5-mL aliquots of plasma were made in long-term cryogenic storage tubes (CryoTube; Nunc, Roskilde, Denmark), and the samples were immediately frozen at -70°C. Plasma aliquots used for measurement of AGP and C-reactive protein (CRP) were stored continuously at -70°C over the past 11 y.
The children were seen at baseline and at 2 and 5 wk for a history and physical examination by both an ophthalmologist and a pediatrician at each visit. Ophthalmic examinations and diagnoses were made according to standardized criteria (13). The pediatrician treated any illnesses appropriately. A standardized, 2-wk morbidity history was obtained by a study nurse at each visit. The parents or guardian were asked whether the child had a fever ("hot to the touch"), cough, or diarrhea ("4 or more stools per day") during the previous 2 wk. At the 2-wk visit, all children were given diphtheria-pertussis-tetanus (DPT) and trivalent oral poliovirus vaccines (both Lederle Laboratories, Pearl River, NY) and trivalent, inactivated intranasal influenza vaccine (Wyeth Laboratories, Marietta, PA).
Plasma retinol concentrations were measured by HPLC using a standard method described elsewhere (10). Pooled human plasma samples were run as external standards. Hyporetinolemia was defined as a plasma retinol concentration < 0.7 µmol/L (2), and frequency distributions of plasma retinol concentrations were based on intervals of 0.175 µmol/L (5 µg/dL), per convention. Plasma AGP was measured by using a commercial radial immunodiffusion kit (Bindarid; The Binding Site, Birmingham, United Kingdom). CRP was measured by commercial enzyme-linked immunosorbent assay (Virgo CRP 150; Hemagen Diagnostics, Waltham, MA). Pooled human plasma samples and the manufacturer's standards were run as external standards. Plasma concentrations of AGP > 1 g/L (8, 9) and of CRP > 5 mg/L (8) were considered to be elevated. In this article, children are referred to as having "elevated acute phase proteins" if they had an elevated plasma AGP or CRP concentration, or both.
At baseline, comparisons were made between children with xerophthalmia (groups 3 and 4 combined) and children without xerophthalmia (groups 1 and 2 combined) by using Student's t test for continuous variables with a normal distribution, ie, plasma retinol concentrations, ln AGP, and ln CRP, and by using chi-square test for categorical variables, ie, acute phase proteins and plasma retinol according to cutoff values. Student's t tests and chi-square tests were also used to compare appropriate variables at baseline among children without xerophthalmia, ie, groups 1 and 2, and among children with xerophthalmia, ie, groups 3 and 4. Geometric means or percentages and 95% CIs were calculated to allow comparisons between the 4 groups and to allow comparisons between values at baseline and 5 wk. AGP and CRP values were transformed by natural logarithm to achieve a normal distribution. Spearman correlation was used to examine the relation between plasma AGP, CRP, and retinol. Statistical tests were performed by using SAS (SAS Institute, Cary, NC).
RESULTS
Two hundred thirty-six children were enrolled in the study. It was not possible to match them completely by sex, and when enrollment was closed, there were 169 boys and 167 girls. The mean (±SD) age of the children was 58.9 ± 10.9 mo. The mean weight-for-height, as a percentage of the median of the National Center for Health Statistics standard (10), was 94.9 ± 7.2% and mean height-for-age was 89.8 ± 4.6%. At baseline, there were no significant differences in age, sex, weight-for-height, or height-for-age between the 4 treatment groups, and these data are shown elsewhere (10). Two hundred thirty-two of the 236 children (98%) returned for follow-up at 5 wk. Acute phase response proteins were measured in 209 of 236 children (88.6%) at baseline and 216 of 232 children (93.1%) at 5 wk. It was not possible to measure all of these factors in all the children because there were not available aliquots in storage. At baseline, acute phase proteins were measured in 50 of 59 (84.7%), 52 of 59 (88.1%), 52 of 58 (89.6%), and 55 of 60 (91.6%) children in groups 1, 2, 3, and 4, respectively. Acute phase proteins were measured in 52 of 59 (88.1%), 53 of 58 (91.4%), 54 of 58 (93.1%), and 57 of 57 (100%) children in groups 1, 2, 3, and 4, respectively, who returned at 5 wk.
The distributions of plasma retinol concentrations at baseline and 5 wk are shown for the 4 groups in Figure 1. At baseline, among children without xerophthalmia, the mean (±SD) plasma retinol concentrations in groups 1 and 2 were 0.80 ± 0.27 and 0.73 ± 0.31 µmol/L, respectively (NS). At baseline, among children with xerophthalmia, the mean (±SD) plasma retinol concentrations in groups 3 and 4 were 0.62 ± 0.24 and 0.58 ± 0.24 µmol/L, respectively (NS). The mean (±SD) plasma retinol concentrations in children with xerophthalmia (groups 3 and 4 combined) compared with children without xerophthalmia (groups 1 and 2 combined) were 0.60 ± 0.24 and 0.76 ± 0.29 µmol/L, respectively (P < 0.0001). At baseline, hyporetinolemia occurred in 61.0% of children with xerophthalmia (groups 3 and 4 combined) compared with 47.4% of children without xerophthalmia (groups 1 and 2 combined) (P < 0.04).
FIGURE 1. . Distribution of plasma vitamin A concentrations at baseline and 5 wk. Shaded areas indicate individuals with elevated acute phase protein concentrations. Groups 1 and 2 are children without xerophthalmia; groups 3 and 4 are children with xerophthalmia. Groups 1 and 3 received vitamin A and groups 2 and 4 received placebo.
At 5 wk, among children without xerophthalmia, the mean plasma retinol concentrations in groups 1 and 2 were 1.67 ± 0.46 and 0.77 ± 0.23 µmol/L, respectively (P < 0.0001) and in groups 3 and 4 were 1.68 ± 0.60 and 0.67 ± 0.24 µmol/L, respectively (P < 0.0001). Mean plasma retinol concentrations increased by 0.87 µmol/L in group 1 (P < 0.0001), 0.05 µmol/L in group 2 (P = 0.2), 1.06 µmol/L in group 3 (P < 0.0001), and 0.09 µmol/L in group 4 (P < 0.03, by paired t test for all). Within each 0.175-µmol/L interval of vitamin A concentration, the number of individuals with elevated acute phase proteins is shown in Figure 1. These distributions indicate that vitamin A supplementation in groups 1 and 3 was associated with a shift of the distribution of plasma retinol to higher concentrations by 5 wk, whereas there was little change in the distribution of plasma retinol concentrations 5 wk later in children who did not receive vitamin A.
At baseline, geometric mean concentrations of AGP were 0.89 and 0.81 g/L and geometric mean concentrations of CRP were 1.2 and 0.6 mg/L in children with xerophthalmia (groups 3 and 4 combined) and in children without xerophthalmia (groups 1 and 2 combined), respectively. The percentage of children at baseline with elevated AGP concentrations was 42.9% and 23.5% and with elevated CRP concentrations was 17.7% and 13.7% among children with xerophthalmia (groups 3 and 4 combined) and children without xerophthalmia (groups 1 and 2 combined), respectively, by chi-square test. The geometric mean concentrations of AGP and CRP and the percentage of children with elevations of these proteins are shown for the 4 allocation groups in Table 1. At baseline, among children without xerophthalmia, there were no significant differences in geometric mean concentrations of AGP or CRP between groups 1 and 2. At baseline, among children with xerophthalmia, there were no significant differences in geometric mean concentrations of AGP or CRP between groups 3 and 4.
View this table:
TABLE 1.. Acute phase response proteins at baseline and 5 wk in the 4 treatment groups1
There were no significant changes in mean AGP or CRP concentrations between baseline and 5 wk in any of the allocation groups. The mean difference in AGP concentration between baseline and 5 wk was 0.06 g/L in group 1, 0.07 g/L in group 2, 0.04 g/L in group 3, and -0.08 g/L in group 4 (NS for all by paired t test). The mean difference in CRP concentration between baseline and 5 wk was 1.02 mg/L in group 1, 0.98 mg/L in group 2, 0.92 mg/L in group 3, and -0.66 in group 4 (NS for all by paired t test). Thus, no significant changes were noted in AGP or CRP concentrations between baseline and 5 wk in any of the allocation groups, including groups 1 and 3, who received vitamin A supplementation. Among children without xerophthalmia, there were no significant differences at 5 wk in the proportion with elevated acute phase proteins in group 1 compared with group 2 or in group 3 compared with group 4 by chi-square test.
The relation between elevated acute phase protein concentrations (AGP > 1 g/L, CRP > 5 mg/L, or both) and morbidity at both baseline and 5 wk is shown in Figure 2. The measures of morbidity were a history of fever, cough, or diarrhea reported by the parent or guardian during the previous 2 wk and direct observation of nasal discharge by a pediatrician on physical examination. The relation between elevated acute phase proteins and other directly observed morbidity (eg, high fever, rapid breathing, and diarrhea) could not be characterized because any children with serious illness were excluded from the study at enrollment and there were no serious illnesses observed during the course of the study. A history of fever or cough and observed nasal discharge were associated with elevated acute phase proteins at both baseline and 5 wk. A reported history of diarrhea was not associated with elevated acute phase proteins at either baseline or 5 wk. At baseline, among children with a negative morbidity history (n = 85) compared with those with a positive history of 1 (n = 56), 2 (n = 47), or 3 (n = 21) of the morbidities combined (fever, cough, and diarrhea), the proportion of children with elevated acute phase proteins was 28.2%, 41.0%, 42.5%, and 57.1%, respectively (P < 0.01, Mantel-Haenszel chi-square test).
FIGURE 2. . Percentage of children with elevated acute phase proteins [1-acid glycoprotein (AGP) > 1 g/L, C-reactive protein (CRP) > 5 mg/L] in all 4 treatment groups, by a 2-wk morbidity history of fever, cough, or diarrhea or by nasal discharge noted on examination at baseline and 5 wk. *,**Significant relation between positive morbidity history and elevated acute phase protein concentrations (chi-square test): *P < 0.05, **P < 0.005.
A history of fever, cough, and diarrhea and nasal discharge observed on physical examination in the 4 allocation groups at baseline and follow-up are shown in Table 2. There were no significant differences in history of fever, cough, or diarrhea or observed nasal discharge between children with and without xerophthalmia at baseline (groups 1 and 2 compared with groups 3 and 4) by chi-square test. Vitamin A supplementation had no significant effect on a 2-wk history of fever, cough, or diarrhea or observed nasal discharge at 5 wk (groups 1 and 3 compared with groups 2 and 4). The relation between history of fever, cough, and diarrhea and observed nasal discharge and plasma retinol concentrations was examined. At baseline, mean plasma retinol concentrations in children with and without a morbidity history of fever, cough, and diarrhea were 0.68 ± 0.26 compared with 0.68 ± 0.29 µmol/L, 0.68 ± 0.29 compared with 0.68 ± 0.27 µmol/L, and 0.65 ± 0.25 compared with 0.69 ± 0.29 µmol/L, respectively (NS for all). Among children with a negative morbidity history compared with those with a positive history of 1, 2, or 3 of the morbidities combined (fever, cough, and diarrhea), mean plasma retinol concentrations were 0.68 ± 0.28, 0.67 ± 0.28, 0.73 ± 0.38, and 0.60 ± 0.24 µmol/L, respectively, with no significant differences noted. At baseline, mean plasma retinol concentrations in children with and without an observed nasal discharge were 0.63 ± 0.28 and 0.74 ± 0.27 µmol/L, respectively (P < 0.006).
View this table:
TABLE 2.. Morbidity history in the 4 allocation groups1
At baseline among children without xerophthalmia (groups 1 and 2), Spearman correlations were calculated between plasma retinol and AGP (r = -0.22, P = 0.02), plasma retinol and CRP (r = -0.22, P = 0.02), and AGP and CRP (r = 0.55, P < 0.0001). At baseline among children with xerophthalmia (groups 3 and 4), Spearman correlations were calculated between plasma retinol and AGP (r = -0.24, P < 0.01), plasma retinol and CRP (r = -0.23, P = 0.02), and AGP and CRP (r = 0.55, P < 0.0001). Among all the children in the study at baseline, Spearman correlations were calculated between plasma retinol and AGP (r = -0.27, P < 0.0001), plasma retinol and CRP (r = -0.25, P < 0.0002), and AGP and CRP (r = 0.57, P < 0.0001). At 5 wk among all children in the study, Spearman correlations were calculated between plasma retinol and AGP (r = -0.07, P = 0.26), plasma retinol and CRP (r = -0.007, P = 0.91), and AGP and CRP (r = 058, P < 0.0001). Spearman correlations for plasma retinol, AGP, and CRP are shown for groups 1, 2, 3, and 4 at baseline and 5 wk in Table 3.
View this table:
TABLE 3.. Spearman correlation coefficients between acute phase protein and retinol in the 4 allocation groups1
DISCUSSION
To our knowledge, this was the first study to examine the relation between hyporetinolemia, infectious disease morbidity, and acute phase proteins in preschool children with and without xerophthalmia. A higher proportion of children with xerophthalmia (Bitot spots, night blindness, or both) had elevated AGP concentrations than did matched children from the same community who did not have xerophthalmia. These observations are consistent with the results of a study from Nepal that showed that a higher proportion of pregnant women with night blindness had elevated acute phase response proteins than did matched control subjects without night blindness (9). The results of both of these studies show that elevated acute phase proteins, hyporetinolemia, and clinical vitamin A deficiency are closely associated, and that acute phase protein measurements do not provide mutually exclusive information for the interpretation of plasma retinol concentrations in populations with a high prevalence of infections.
About 20–40% of preschool children from this area of Indonesia had elevated acute phase proteins. In preschool children from a malaria-endemic area of northern Ghana, elevated serum AGP concentrations (>1 g/L) were measured in about one-third of those studied (8). Malaria does not occur in our study population in Indonesia and can be ruled out as contributing to elevations of acute phase proteins. Trauma is another cause of elevated acute phase proteins (7), but none of the children in the study had any notable trauma on physical examination by a pediatrician. The proportion of children with elevated CRP concentrations was lower than the proportion of children with elevated AGP concentrations, and this observation is probably explained by the differences in the response of CRP and AGP to infection. During an infection, CRP generally rises quickly and falls more rapidly than does AGP, which rises slowly and stays elevated longer (14).
The results of this study corroborate those of previous studies, which showed a negative correlation between acute phase protein concentrations and plasma retinol concentrations. The correlation found in the present study was consistent with the correlations that were described previously between AGP and retinol and CRP and retinol (8, 9, 15). The correlations between acute phase proteins and plasma retinol were strongest at baseline in children with (groups 3 and 4) and without (groups 1 and 2) xerophthalmia. Once broken down into the individual treatment groups, the correlations between retinol concentrations and acute phase proteins were not as strong, but this may be related to the smaller samples and reduced power in the 4 separate groups. At both baseline and follow-up, plasma AGP and CRP concentrations were strongly correlated within all 4 treatment groups. All children were immunized with DPT vaccine, and some of the elevation in acute phase proteins at 5 wk might be attributed to the vaccine because DPT immunization can induce a fever and influenza-like illness.
There was little association between a history of fever, cough, and diarrhea, as elicited on a 2-wk morbidity history, and plasma retinol concentrations. This lack of association might be explained by one or more factors: inaccuracy of recall by the parent or guardian, the limitations of imprecise definitions from a morbidity history, and the time interval between the actual episode of illness and the time of measurement of plasma retinol concentrations.
Plasma retinol concentrations were significantly associated with nasal discharge noted by the pediatrician on physical examination at baseline, suggesting that concurrent illness might be expected to depress plasma retinol concentrations more than did a previous episode that was more remote in time. Although the proportions of children who were reported to have a history of fever, cough, and diarrhea seem high, both fever and cough were significantly associated with elevated acute phase proteins. Thus, the 2-wk morbidity history does appear to reflect infection in the recent past. The results of this study corroborate those of previous studies that suggested that reported signs or history of infection are associated with elevations in acute phase proteins (16, 17).
The plasma retinol concentrations reported in this study are consistent with those reported in other studies of preschool children from the same area (18–20). The low plasma retinol concentrations in children without xerophthalmia and the response of plasma retinol concentrations to vitamin A supplementation show that most of the so-called control children without xerophthalmia probably had subclinical vitamin A deficiency. In this study, vitamin A supplementation did not have a significant effect on infectious disease morbidity, which contrasts with results from most large studies worldwide (2). The study was not designed to examine the effect of vitamin A supplementation on infectious disease morbidity because such a study would require a much larger sample size and longer follow-up to have sufficient statistical power to detect such an effect. In addition, any children with serious illness were excluded from this study.
There appears to be a close association among vitamin A deficiency, infection, and low plasma retinol concentrations. A low frequency distribution of plasma vitamin A concentrations in populations in developing countries should raise concerns about the presence of vitamin A deficiency (21). Vitamin A supplementation has been shown to reduce the morbidity and mortality of different infections, including diarrheal disease, measles, and malaria (2, 22). Infection-related hyporetinolemia and low body stores of vitamin A induced by infection may contribute to night blindness both in preschool children and in pregnant women (9), suggesting that further research should be directed toward the potential effect of infection-related hyporetinolemia on impairment of vision, immunity, and other biological functions.
ACKNOWLEDGMENTS
We thank the staff of the Cicendo Eye Hospital, Bandung, Indonesia, and the Nutrition Research and Development Centre, Bogor, Indonesia.
REFERENCES