Although some studies have suggested that retinol-binding protein 4 (RBP4), a fat-derived adipokine, is positively related to overweight and obesity-related disorders, these claims have been disputed. This study investigated relationships between RBP4 level and weight status and metabolic disorders in a Korean population including confounding factors, age and gender.
From 2005 to 2007, we recruited 570 children and 241 of their parents living in Korean urban areas. We analyzed anthropometrical parameters, blood pressure, lipid profiles, and levels of serum glucose, insulin, leptin and RBP4, and estimated dietary intake based on a self-reported 3-day food diary.
Levels of RBP4 were high for men (226.2 μg/ml), mid-range among women (143.2 μg/ml) and low among boys and girls (boys 71.4, girls 66.9 μg/ml). After adjusting for age, gender differences among adults were consistently maintained (P<0.0001), and gender differences among children were detected minimally (P=0.05). On the basis of Cole's body mass index cutoff points, the prevalence of overweight among children was 19.3%. The prevalence among adults was 32.0%, based on World Health Organization criteria. Overweight children had higher RBP4 concentrations than normal. Overweight women had higher RBP4 levels than those of normal weight. Higher RBP4 levels were observed among overweight men, but the results were not statistically significant. Triglyceride (TG) levels were significantly correlated with RBP4 in all groups.
Our results indicate that RBP4 concentrations in a Korean population varied with age, gender and overweight status. TG levels could be strongly correlated with RBP4 concentration.
Retinol-binding protein 4 (RBP4), originally identified for its role in retinol transport, was recently identified as a fat-derived adipokine secreted by adipose tissue (Yang et al., 2005; Graham et al., 2006; Tamori et al., 2006). RBP4 levels reportedly have a positive relationship with obesity (Graham et al., 2006) and insulin resistance (Yang et al., 2005; Cho et al., 2006; Graham et al., 2006). Some metabolic syndrome components, such as abdominal obesity, elevated triglyceride and blood pressure, also have been reported to have a positive correlation with serum RBP4 concentration (Cho et al., 2006; Aeberli et al., 2007; Qi et al., 2007; Lee et al., 2007a; Gavi et al., 2008).
However, studies have failed to demonstrate consistent associations between circulating RBP4 levels and obesity or total body fat (Janke et al., 2006; Broch et al., 2007; Stefan et al., 2007). One study was unable to find an association between insulin sensitivity and RBP4 levels in human subjects with obesity or diabetes (Janke et al., 2006). Several confounding factors may be involved when considering associations between RBP4 and insulin sensitivity.
Age and gender might influence RBP4 levels (Cho et al., 2006; Qi et al., 2007; Lee et al., 2007a). Young subjects exhibit lower levels of RBP4 than elderly subjects, regardless of gender (Gavi et al., 2007, 2008). However, RBP4 concentrations in women younger than 50 years are significantly lower than those among women older than 50 years (Lim et al., 2008). Moreover, women older than 50 years had lower RBP4 concentrations than men (Cho et al., 2006; Gavi et al., 2007). This gender difference has previously been identified for other adipokines, including adiponectin and leptin, which could be explained by the effect of sex hormones and differences in body fat percentage (Saad et al., 1997; Combs et al., 2003).
Some studies have identified concomitant increases in RBP4 levels with increasing adiposity, and have found that RBP4 levels are related to insulin resistance and weight status in adolescents and adults (Aeberli et al., 2007; Lee et al., 2007a; Reinehr et al., 2008). Similar to the studies of adults discussed above, research has revealed that male adolescents have higher RBP4 levels than female adolescents. One study reported that RBP4 levels were significantly higher in adolescent male subjects than in adolescent females (mean age: 14.6 years; Lee et al., 2007a), but another study of subjects with an average age of 10.5 years found that RBP4 levels did not differ significantly between boys and girls (Reinehr et al., 2008).
We investigated differences in serum RBP4 level by gender and overweight status among Korean children and adults. We also evaluated whether serum RBP4 concentrations were correlated with components of insulin resistance and related factors among children and adults.
Subjects and methods
This cross-sectional study was based on a pediatric cohort recruited from the city of Gwacheon, Korea, in 2005. This cohort is being monitored in an ongoing prospective study of Korean children who have been examined yearly since their entry into elementary school at age 7. Subjects were excluded if they were taking any medications that might affect appetite or if they were enrolled in a specific diet program. Subjects included a total of 570 children enrolled between 2005 and 2007. Approximately 20% of the children’s parents (120 men and 121 women) also volunteered to be involved in the study. The study protocol was approved by the institutional review board at Inje University Seoul-Paik Hospital and the Korea Center for Disease Control and Prevention. All subjects gave written informed consent.
Anthropometric and biochemical measurements and dietary analysis have been described elsewhere (Kim et al., 2008). A total of 401 children and 239 adults completed the 3-day food diary. Serum RBP4 concentrations were determined using an enzyme-linked immunosorbent assay from a commercial kit (AdipoGen, Incheon, Korea). The analysis used a RBP4 competitive enzyme-linked immunosorbent assay system, in which 100 μg/l of recombinant human RBP4 expressed by an animal cell line (HEK293 cells) is coated onto a plate and varying concentrations of the recombinant RBP4 and a polyclonal antibody are added to the plate for a competitive reaction resulting in a standard curve.
We defined overweight status in children using Cole’s cutoff point, corresponding to 25 kg/m2 at 18 years of age for each sex and each age proposed by the body mass index (BMI) references established in an international survey (Cole et al., 2000). In adults, overweight status was defined according to the World Health Organization criteria for normal weight (BMI <25 kg/m2) and overweight (BMI ⩾25 kg/m2).
All continuous variables are expressed as means ± s.d. Data that were not normally distributed (for example, homeostasis model assessment-insulin resistance (HOMA-IR), body fat percentage) were logarithmically transformed before statistical analysis. Comparisons among groups were performed using Student’s t-test and analysis of variance with Duncan’s test. General linear model procedures with adjustment for age or both age and BMI were performed to test for the gender difference in RBP4 levels. Pearson’s correlation analysis was used to evaluate the relationships among the different variables. Statistical analyses were performed using SAS 9.1 software (SAS institute, Cary, NC, USA). P-values <0.05 were considered to be statistically significant.
Personal characteristics of the children and adults
Table 1 lists the characteristics of the children and adults. All characteristics of children with the exception of energy intake differed significantly from the characteristics of adults. Significant differences appeared by gender among children, including height; weight; BMI; waist and hip circumferences; fasting serum glucose, HDL-C, and LDL-C levels; triglyceride (TG); and total energy intake. Among adults, gender-based differences included age, height, weight, percent body fat (men 25.3±5.5 and women 33.4±5.6%), BMI, waist and hip circumferences, glucose and fasting insulin levels, HOMA-IR, total cholesterol, HDL-C, LDL-C, TG, adiponectin, leptin, RBP4 levels (men 226.2±143.3 and women 143.2±92.5 μg/ml, P<0.0001), total energy intake, and vitamin A intake. After children and adults were separated by gender, we analyzed all variables for differences among the four groups by analysis of variance with Duncan's test. TG, adiponectin, and RBP4 levels differed significantly among groups: boys, girls, men and women. These three variables did not differ between boys and girls, but the difference between men and women was highly significant. TG and RBP4 levels exhibited a different trend from that of adiponectin: serum TG and RBP4 concentrations in children were significantly lower than those in women. Also, women had lower TG and RBP4 levels than men. However, when adjusted by age, we identified gender differences in RBP4 levels in children (P=0.05), and found that serum RBP4 concentrations between adult men and women differed significantly (P<0.0001). The prevalence of overweight was 45.8% for males and 18.2% for females. In children, the prevalence of overweight was 26.0% for boys and 11.2% for girls.
Figure 1 shows the differences in RBP4 concentrations between normal-weight and overweight groups by sex for children and adults. Normal-weight boys had significantly lower RBP4 concentrations than overweight boys (P=0.02). Girls exhibited the same pattern with regard to RBP4 concentrations (P=0.03). RBP4 levels in normal-weight women were significantly lower than those in overweight women (P=0.04). However, RBP4 levels did not differ significantly between normal-weight and overweight groups of men (P=0.25).
Anthropometric and biochemical characteristics of children and adults by overweight status
Table 2 lists the anthropometric and biochemical characteristics of children and adults by overweight status. On the basis of Cole et al. (2000), 19.3% of children were overweight. On the basis of World Health Organization criteria, 32.0% of adults were overweight (BMI ⩾25). In children, all variables except total cholesterol levels and dietary intakes of total energy and vitamin A differed significantly between the normal-weight and overweight groups. In adults, all variables except age, total energy intake and vitamin A intake differed significantly between normal-weight and overweight groups. The dietary intake of carbohydrate, protein and fat did not differ by overweight status among children or adults (Supplementary Table 1).
When children and adults were separated by sex and overweight status, we found that normal-weight and overweight groups of boys differed significantly in age, height, weight, body fat percent, BMI, waist and hip circumferences, fasting insulin, HOMA-IR levels, HDL-C, LDL-C, TG, adiponectin, leptin and RBP4 levels, total energy intake, and vitamin A intake (Table 2). Normal-weight and overweight girls differed significantly in anthropometric variables including height, weight, body fat percent, BMI, and waist and hip circumferences. In addition, we found that biochemical variables except total cholesterol differed significantly between normal-weight and overweight groups. In adults, overweight men had significantly greater weight, percent body fat, BMI, and waist and hip circumferences than normal-weight men. Levels of insulin, HOMA-IR, HDL-C, TG, adiponectin and leptin tended to be higher in overweight men than in normal-weight men. RBP4 levels in normal-weight men were lower than in overweight men, but the difference was not statistically significant. Normal-weight and overweight women differed significantly in weight, body fat percent, BMI, and waist and hip circumferences. In addition, overweight women had significantly higher levels of insulin, HOMA-IR, HDL-C, TG, leptin and RBP4 than normal-weight women. Adiponectin concentrations in normal-weight women were significantly lower than in overweight women.
Clinical characteristics according to serum RBP4 tertiles
Table 3 lists the anthropometric and biochemical characteristics in children and adults, according to RBP4 tertiles and parameters that differ significantly between groups of each RBP4 tertile. Weight, waist and hip circumference, total cholesterol levels and TG levels differed significantly between the groups of children’s RBP4 tertiles. Boys in the highest tertile of RBP4 level exhibited greater height, weight and hip circumferences, and TG levels, and the corresponding group of girls exhibited only higher TG levels (data not shown). Gender differences appeared in the RBP4 levels of children, but those were minimal. When adjusted by age and BMI, the gender difference in the RBP4 level of children was not present (P=0.18). Therefore, the following analysis for children does not include any further classification by gender. The concentration of serum RBP4 between men and women in adults was significantly different after the adjustment of age and BMI (P<0.0004). Insulin, HOMA-IR and TG levels differed significantly among the groups of men’s RBP4 tertiles. Females in the highest tertile also had significantly higher body fat percent, BMI, waist and hip circumferences, insulin, HOMA-IR and total cholesterol levels. Among children, higher tertiles had a higher prevalence of overweight; this effect depended on boys because girls did not exhibit this trend (data not shown). Total energy intake and vitamin A intake did not differ significantly among groups of children’s, men’s and women’s RBP4 tertiles (data not shown).
Correlations among RBP4 levels and anthropometric indicators and biochemical characteristics in adults and children
Serum RBP4 levels and TG levels were strongly associated in each group (Table 4 and Figure 2). When adjusted by age and BMI, TG levels were significantly correlated with RBP4 concentrations in children (r=0.14; P=0.001), men (r=0.24; P=0.009) and women (r=0.25; P=0.007). Also, total cholesterol levels in children (r=0.10; P=0.02) and insulin levels and HOMA-IR (r=0.26 and 0.25, P<0.05) in men were significantly correlated with RBP4 concentrations, independent of age and BMI. Total energy intake and vitamin A intake were not significantly correlated with RBP4 levels in children, men, and women (data not shown).
RBP4 is known to be related to obesity, type 2 diabetes mellitus, metabolic syndrome and insulin resistance (Graham et al., 2006; Qi et al., 2007). However, relatively few studies have researched the interaction between RBP4 concentrations and these disorders in young children. Our results revealed that serum RBP4 concentrations varied in an age- and gender-specific manner in Korean adults and children, and were associated with lipid profiles.
We identified apparent gender-related differences in RBP4 concentrations in pre-menopausal women and men, and minimal gender differences among children. RBP4 concentrations vary by gender in adults (Cho et al., 2006; Gavi et al., 2007; Fernandez-Real et al., 2008) and also vary with age (Cho et al., 2006; Lim et al., 2008). Lim et al. (2008) found that RBP4 concentrations differed between a young female group (40.7±12.2 μg/ml, age 18–29 years) and an older female group (48.1±15.3, age 50–71 years). Cho et al. (2006) also found that females older than 50 years had higher RBP4 levels than females younger than 50 years. However, no significant differences appeared in the RBP4 levels of men by age. Another recent study suggested that RBP4 concentration is associated with iron status, which varied in a gender-dependent manner (Fernandez-Real et al., 2008). A study of children revealed that RBP4 concentrations also differed significantly by pubertal status (Reinehr et al., 2008). Thus, differences in sex hormone status and the accompanying body fat percentage might influence RBP4 concentrations. The gender differences in RBP4 levels indicate that a female hormone such as estrogen may affect BP4 concentrations in women. Although we identified gender-based differences in RBP4 level in adults aged younger than 50, the mechanism for this is still unclear and requires further research.
We demonstrated that overweight boys and girls have significantly higher RBP4 concentrations than normal-weight boys and girls, supporting previous studies of adolescents (Lee et al., 2007a; Reinehr et al., 2008). In our study, boys in the higher tertile group of RBP4 concentration had a higher prevalence of overweight, although girls did not exhibit this trend. These results suggest that RBP4 concentration may not influence children’s overweight directly. In contrast, the results suggest that children’s overweight status may affect RBP4 levels. Overweight women had significantly higher RBP4 concentrations than normal-weight women. However, no statistical differences appeared in RBP4 concentrations between normal-weight men and overweight men. Several previous studies have reported that RBP4 levels are higher in overweight subjects (Lee et al., 2007a, 2007b). Other studies have found no association between RBP4 concentration and overweight status (Broch et al., 2007; Yao-Borengasser et al., 2007). As with children, we found that women’s overweight status had a positive influence on RBP4 concentration. In addition, when RBP4 concentrations are high, women’s RBP4 levels may affect overweight status. However, no such trend appeared for men. Possible explanations for these discordant results may include the small sample size, or remarkably high RBP4 concentrations and an exceedingly high prevalence of overweight.
Previous studies have suggested an interaction between insulin resistance and RBP4 concentrations in humans and mice (Yang et al., 2005; Graham et al., 2006). Increased RBP4 appears to induce insulin resistance by increasing phosphoenolpyruvate expression and glucose production, and by decreasing the action of insulin to suppress glucose production in hepatocytes (Yang et al., 2005). We found a significantly positive correlation between men’s RBP4 concentrations and serum insulin levels, and men in higher RBP4 tertile groups had higher insulin levels. However, women in high RBP4 tertile groups did not exhibit significantly increased insulin levels. Because overweight status may influence insulin resistance, the presence of obesity could be a crucial determinant of insulin resistance (Kahn et al., 2006). Thus, the lower prevalence of overweight in our female subjects (18.2%) compared with our male subjects (45.8%) might have lowered the degree of association between RBP4 levels and insulin resistance.
Our results confirmed the correlation between RBP4 concentrations and TG levels in both children and adults. RBP4 metabolism appears to be closely connected with liver fat. Hypertriglyceridemia, led by hyperinsulinemia, may subsequently provoke the synthesis and secretion of RBP4 from the liver or ectopic fat (Qi et al., 2007). Increased RBP4 levels also affect TG levels. We found that the positive association between RBP4 and TG levels was maintained even considering confounding factors such as age and BMI. We did not evaluate liver or ectopic fat data, so further research will be required to elucidate the apparent mechanism for the relationship between serum RBP4 concentrations and liver or ectopic fat distribution.
A previous study reported that nutritional status (vitamin A intake) affected RBP4 levels (Aeberli et al., 2007). We found no association between serum RBP4 concentrations and vitamin A intake. The amount of vitamin A ingested by the subjects in the study was within the adequate range, so vitamin A intake did not affect RBP4 levels. One limitation of our study was that we did not investigate alcohol consumption and smoking status of adults, which could act as confounders. Further research including measurements of smoking status and alcohol consumption will be necessary.
In conclusion, although our sample of adults was small, we demonstrated that gender is one of determinants of RBP4 concentration among Korean adults, and women exhibited a particularly strong association between RBP4 concentrations and overweight status. Children exhibited lower RBP4 levels than adults. Of the many components of metabolic disorders, TG levels appeared to be correlated with RBP4 concentrations in all subjects regardless of age. Reducing RBP4 levels could be a new strategy for treating hyperlipidemia.
Aeberli I, Biebinger R, Lehmann R, Allemand D, Spinas GA, Zimmermann MB (2007). Serum retinol-binding protein 4 concentration and its ratio to serum retinol are associated with obesity and metabolic syndrome components in children. J Clin Endocrinol Metab 92, 4359–4365.
Broch M, Vendrell J, Ricart W, Richart CC, Fernandez-Real JM (2007). Circulating retinol-binding protein-4, insulin sensitivity, insulin secretion, and insulin disposition index in obese and nonobese subjects. Diab Care 30, 1802–1806.
Cho YM, Youn BS, Lee H, Lee N, Min SS, Kwak SH et al. (2006). Plasma retinol-binding protein-4 concentrations are elevated in human subjects with impaired glucose tolerance and type 2 diabetes. Diab Care 29, 2457–2461.
Cole TJ, Bellizzi MC, Flegal KM, Dietz WH (2000). Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 320, 1240.
Combs TP, Berg AH, Rajala MW, Klebanov S, Iyengar P, Jimenez-Chillaron JC et al. (2003). Sexual differentiation, pregnancy, calorie restriction, and aging affect the adipocyte-specific secretory protein adiponectin. Diabetes 52, 268–276.
Fernandez-Real JM, Moreno JM, Ricart W (2008). Circulating retinol-binding protein-4 concentration might reflect insulin resistance? Associated iron overload. Diabetes 57, 1918–1925.
Gavi S, Stuart LM, Kelly P, Melendez MM, Mynarcik DC, Gelato MC et al. (2007). Retinol-binding protein 4 is associated with insulin resistance and body fat distribution in nonobese subjects without Type 2 diabetes. J Clin Endocrinol Metab 92, 1886–1890.
Gavi S, Qurashi S, Stuart LM, Lau R, Melendez MM, Mynarcik DC et al. (2008). Influence of age on the association of retinol-binding protein 4 with metabolic syndrome. Obesity 16, 893–895.
Graham TE, Yang Q, Bluher M, Hammarstedt A, Ciaraldi TP, Henry RR et al. (2006). Retinol-binding protein 4 and insulin resistance in lean, obese, and diabetic subjects. N Engl J Med 354, 2552–2563.
Janke J, Engeli S, Boschmann M, Adams F, Bohnke J, Luft FC et al. (2006). Retinol-binding protein 4 in human obesity. Diabetes 55, 2805–2810.
Kahn SE, Hull RL, Utzschneider KM (2006). Mechanisms linking obesity to insulin resistance and type 2 diabetes. Nature 444, 840–846.
Kim IK, Kim J, Kang JH, Song J (2008). Serum leptin as a predictor of fatty liver in 7-year-old Korean children. Ann Nutr Metab 53, 109–116.
Lee DC, Lee JW, Im JA (2007a). Association of serum retinol binding protein 4 and insulin resistance in apparently healthy adolescents. Metabolism 56, 327–331.
Lee JW, Im JA, Lee HR, Shim JY, Youn BS, Lee DC (2007b). Visceral adiposity is associated with serum retinol binding protein-4 levels in healthy women. Obesity (Siver Spring) 15, 2225–2232.
Lim S, Choi SH, Jeong IK, Kim JH, Moon MK, Park KS et al. (2008). Insulin-sensitizing effects of exercise on adiponectin and retinol-binding protein-4 concentrations in young and middle-aged women. J Clin Endocrinol Metab 93, 2263–2268.
Qi Q, Yu Z, Ye X, Zhao F, Huang P, Hu FB et al. (2007). Elevated retinol-binding protein 4 levels are associated with metabolic syndrome in chinese people. J Clin Endocrinol Metab 92, 4827–4834.
Reinehr T, Stoffel-Wagner B, Roth CL (2008). Retinol-binding protein 4 and its relation to insulin resistance in obese children before and after weight Loss. J Clin Endocrinol Metab 93, 2287–2293.
Saad MF, Damani S, Gingerich RL, Riad-Gabriel MG, Khan A, Boyadjian R et al. (1997). Sexual dimorphism in plasma leptin concentration. J Clin Endocrinol Metab 82, 579–584.
Stefan N, Hennige AM, Staiger H, Machann J, Schick F, Schleicher E et al. (2007). High circulating retinol-binding protein 4 is associated with elevated liver fat but not with total, subcutaneous, visceral, or intramyocellular fat in humans. Diab Care 30, 1173–1178.
Tamori Y, Sakaue H, Kasuga M (2006). RBP4, an unexpected adipokine. Nat Med 12, 30–31.
Yang Q, Graham TE, Mody N, Preitner F, Peroni OD, Zabolotny JM et al. (2005). Serum retinol binding protein 4 contributes to insulin resistance in obesity and type 2 diabetes. Nature 436, 356–362.
Yao-Borengasser A, Varma V, Bodles AM, Rasouli N, Phanavanh B, Lee MJ et al. (2007). Retinol binding protein 4 expression in humans: relationship to insulin resistance, inflammation, and response to pioglitazone. J Clin Endocrinol Metab 92, 2590–2597.
We thank all the participating schools, children and parents. This study was funded by a grant from the Korea National Institute of Health (4845–300–210, 4845–300–260).
The authors declare no conflict of interest.
Supplementary Information accompanies the paper on European Journal of Clinical Nutrition website
About this article
Cite this article
Kim, I., Lee, H., Kang, J. et al. Relationship of serum retinol-binding protein 4 with weight status and lipid profile among Korean children and adults. Eur J Clin Nutr 65, 226–233 (2011) doi:10.1038/ejcn.2010.251
- gender difference
Childhood retinol-binding protein 4 (RBP4) levels predicting the 10-year risk of insulin resistance and metabolic syndrome: the BCAMS study
Cardiovascular Diabetology (2018)
Serum retinol binding protein 4 is negatively related to estrogen in Chinese women with obesity: a cross-sectional study
Lipids in Health and Disease (2016)
Endocrine and Metabolic Effects of Adipose Tissue in Children and Adolescents / Endokrina in Presnovna Funkcija Maščobnega Tkiva Pri Otrocih in Mladostnikih
Slovenian Journal of Public Health (2015)
Clinical Pediatrics (2013)
Free fatty acid binding protein-4 and retinol binding protein-4 in polycystic ovary syndrome: response to simvastatin and metformin therapies
Gynecological Endocrinology (2013)