Skip to main content
NCBI home page
The American Journal of Clinical Nutrition logoLink to The American Journal of Clinical Nutrition
. 2017 May 17;106(1):52–58. doi: 10.3945/ajcn.116.144006

Genetics of serum carotenoid concentrations and their correlation with obesity-related traits in Mexican American children

Vidya S Farook 1, Lavanya Reddivari 2, Srinivas Mummidi 1, Sobha Puppala 5, Rector Arya 1, Juan Carlos Lopez-Alvarenga 1, Sharon P Fowler 6, Geetha Chittoor 8, Roy G Resendez 1, Birunda Mohan Kumar 9, Anthony G Comuzzie 5, Joanne E Curran 1, Donna M Lehman 6, Christopher P Jenkinson 1, Jane L Lynch 7, Ralph A DeFronzo 6, John Blangero 1, Daniel E Hale 7, Ravindranath Duggirala 1,✉, Jairam KP Vanamala 3,4,✉
PMCID: PMC5486195  PMID: 28515064

Abstract

Background: Dietary intake of phytonutrients present in fruits and vegetables, such as carotenoids, is associated with a lower risk of obesity and related traits, but the impact of genetic variation on these associations is poorly understood, especially in children.

Objective: We estimated common genetic influences on serum carotenoid concentrations and obesity-related traits in Mexican American (MA) children.

Design: Obesity-related data were obtained from 670 nondiabetic MA children, aged 6–17 y. Serum α- and β-carotenoid concentrations were measured in ∼570 (α-carotene in 565 and β-carotene in 572) of these children with the use of an ultraperformance liquid chromatography–photodiode array. We determined heritabilities for both carotenoids and examined their genetic relation with 10 obesity-related traits [body mass index (BMI), waist circumference (WC), high-density lipoprotein (HDL) cholesterol, triglycerides, fat mass (FM), systolic and diastolic blood pressure, fasting insulin and glucose, and homeostasis model assessment of insulin resistance] by using family data and a variance components approach. For these analyses, carotenoid values were inverse normalized, and all traits were adjusted for significant covariate effects of age and sex.

Results: Carotenoid concentrations were highly heritable and significant [α-carotene: heritability (h2) = 0.81, P = 6.7 × 10−11; β-carotene: h2 = 0.90, P = 3.5 × 10−15]. After adjusting for multiple comparisons, we found significant (P ≤ 0.05) negative phenotypic correlations between carotenoid concentrations and the following traits: BMI, WC, FM, and triglycerides (range: α-carotene = −0.19 to −0.12; β-carotene = −0.24 to −0.13) and positive correlations with HDL cholesterol (α-carotene = 0.17; β-carotene = 0.24). However, when the phenotypic correlations were partitioned into genetic and environmental correlations, we found marginally significant (P = 0.051) genetic correlations only between β-carotene and BMI (−0.27), WC (−0.30), and HDL cholesterol (0.31) after accounting for multiple comparisons. None of the environmental correlations were significant.

Conclusions: The findings from this study suggest that the serum carotenoid concentrations were under strong additive genetic influences based on variance components analyses, and that the common genetic factors may influence β-carotene and obesity and lipid traits in MA children.

Keywords: α-carotene, β-carotene, cardiometabolic traits, childhood obesity, common genetic influences, heritability

INTRODUCTION

The prevalence rates of childhood and adult obesity are high in the United States and global populations. Obesity poses a formidable public health challenge because obesity along with related cardiometabolic traits (CMTs) elevate the risk of various chronic diseases, including metabolic syndrome (MS) (1), type 2 diabetes (T2D), and cardiovascular disease (2–5). Extensive efforts have been undertaken to identify the genetic and environmental factors that predispose individuals to obesity and other CMTs. In this context, diet has been recognized as one of the major, modifiable, environmental risk factors for the development of childhood obesity. Dietary intake of specific micronutrients (vitamins and minerals) and phytonutrients, such as carotenoids, has been shown to be beneficial to human health (6). For example, carotenoids, plant-derived polyisoprenoid lipophilic pigments, were found to have strong anti-inflammatory and antioxidant properties. Supporting such a role, increased dietary intake of carotenoids and their circulating concentrations in blood are correlated with a lower risk of several complex diseases including obesity, T2D, MS, cardiovascular disease, and carcinogenesis.

Although >700 distinct carotenoids have been recognized, 6 of them (α-carotene, β-carotene, β-cryptoxanthin, lutein, zeaxanthin, and lycopene) represent >95% of all carotenoids in the human plasma or serum (7, 8). The provitamin A carotenoids (α-carotene, β-carotene, and β-cryptoxanthin) (9) are converted to retinol or vitamin A and play a key role in a wide range of biological processes, including reproduction, embryonic development, growth, cellular differentiation, immunity, vision, and metabolic control (10, 11). Lower serum concentrations of these provitamin A carotenoids were reported in overweight and obese adults and children in various US ethnic groups, including Mexican Americans (MAs), who are more prone to obesity and related diseases (12–19). Although circulating carotenoids are associated directly with the dietary intake of fruits and vegetables, the correlation is not very high (usually <0.5) (20). The degree of individual variability in the serum carotenoid concentration response to increased intake of dietary carotenoids could be categorized into “high” and “low” responders (21). Interindividual variations in serum provitamin A carotenoid concentrations may be due to differences in absorption, assimilation, and metabolism of the carotenoids, which could be attributable to genetic factors (1, 20, 22). In addition to candidate gene studies, genome-wide association studies have identified specific genetic variants that influence variation in circulating carotenoid concentrations (1, 23–25). The genetic determinants of variation in carotenoid concentrations in MA children are poorly understood. Therefore, the purpose of this study was to determine the genetic basis of carotenoids and their correlations with obesity and other CMT traits in a cohort of high-risk MA children and adolescents, aged 6–17 y, who bear substantial cardiometabolic burdens, including overweight (53%), obesity (34%), prediabetes (13%), and MS (19%) (26).

METHODS

The San Antonio Family Assessment of Metabolic Risk Indicators in Youth study

The subjects of this study were MA children and adolescents who were eligible to participate in the SAFARI (San Antonio Family Assessment of Metabolic Risk Indicators in Youth) study, whose recruitment began in September 2005. The rationale and the design of SAFARI study have been described elsewhere (26). Briefly, the SAFARI study was designed to evaluate cardiometabolic risk factors and traits and to determine their genetic basis. A total of 673 MA children and adolescents aged 6–17 y from predominantly low-income extended families, whose adult members were previously enrolled in well-established family-based genetic epidemiologic studies of MA adults in San Antonio, Texas, participated in this study. The SAFARI children and adolescents represented 401 nuclear families or sibships that were part of large MA families and generated a total of 3664 relative pairs (e.g., 383 sibling pairs, 550 first-cousin pairs, 661 second-cousin pairs, and 662 third-cousin pairs) (26). The present study excluded 3 children who were diagnosed with T2D (Figure 1). All research procedures were approved by the Institutional Review Board of the University of Texas Health Science Center at San Antonio. Written informed consent from one or both parents or the legal guardian of each child and signed assent from children aged ≥7 y were obtained before the initiation of any study procedures.

FIGURE 1.

FIGURE 1

Open in a new tab

SAFARI study and data availability for this study. SAFARI, San Antonio Family Assessment of Metabolic Risk Indicators in Youth.

Phenotype data

A large battery of family history, demographic, phenotypic, and environmental covariate data were collected, and data collection procedures were previously described in detail (26). Data for this study included physical exam findings obtained by pediatric endocrinologists and metabolic assessments. The following 10 obesity-related traits were considered for the study: waist circumference (WC), BMI, fat mass (FM) assessed by dual-energy X-ray absorptiometry, blood pressure (systolic and diastolic blood pressure), plasma glucose [fasting glucose], serum-specific insulin (4), the HOMA-IR, HDL cholesterol, and triglycerides. These data were obtained with the use of the standard protocols and measurement techniques as described previously (26).

α-Carotene and β-carotene measurements

We measured the concentrations of α-carotene and β-carotene in the fasting serum of the SAFARI children by using Waters’ ultraperformance liquid chromatography (UPLC)–photodiode array detection system with the use of available serum samples from nondiabetic children (n = 572), however, α-carotene was measurable in only 565 samples (Figure 1). Briefly, aliquots of serum samples (200 μL) were placed in prelabeled 100 × 12-mm borosilicate tubes. No >24 serum samples were extracted at a time with the use of 200 μL of 95% ethanol containing 40 μg /mL α-tocopherol acetate and 120 ng/mL retinol acetate as internal standards. After vortexing for 45 s, hexane (1.0 mL) containing 0.01% butylated hydroxytoluene was added to the vial and vortexed for 1 min at medium speed. Upper organic phase was collected into a prechilled UPLC vial after centrifugation at 1400 × g for 10 min. The supernatant collected was dried to completion under nitrogen gas and stored at −80°C until quantification. Before UPLC analysis, the dried samples were reconstituted in 30 μL of chilled chloroform and 70 μL of acetonitrile: methanol (1:1 volume:volume). After vortexing, samples were transferred into 150-μL glass vial inserts. Serum extracts (5 μL) were injected into a Waters Acquity UPLC system (Waters). Separation was performed with the use of a BEH C18 UPLC column (2.1 × 150 mm, 1.7 μm) with a gradient from 50% solvent B (acetonitrile) to 100% solvent B in 12 min and held for an additional 10 min. Samples were maintained at 4°C, the column temperature was set to 60°C, and the flow rate was kept constant at 600 μL/min. Concentrations were calculated with the use of standard curves. Standard Reference Material (968e) at 3 levels (low, medium, and high) from the National Institute of Standards and Technology were used as quality control (QC) samples to validate the method used to quantify serum carotenoids. QC samples were extracted along with samples and standards. Extracts were run at the beginning, end, and after every 24 samples. All the QC samples were analyzed for accuracy and precision. We confirmed that for all 3 levels, the concentrations were within the range provided by the National Institute of Standards and Technology.

Statistical analyses

Variance components analysis

The genetic basis of α-carotene and β-carotene were determined with the use of a variance components approach as implemented in the computer program SOLAR 6.4.9 (27). This approach uses information from all possible biological relationships simultaneously in an attempt to disentangle the genetic architecture of a quantitative trait (e.g., BMI). In a simple model, variances or covariances between relatives as a function of the genetic relationships can be specified, and the proportion of phenotypic variance that is attributed to (additive) genetic effects [i.e., heritability (h2)] can be estimated from the components of variance (28, 29). A likelihood ratio test was used to test whether the heritability of a given phenotype was significant (P ≤ 0.05). The α-carotene and β-carotene values were inverse normalized for the genetic analyses and were adjusted for significant (P ≤ 0.05) covariate effects of age and sex terms (i.e., age, sex, age2, age × sex, and age2 × sex).

Bivariate genetic analysis

The phenotypic, genetic, and environmental correlations between the carotenes and between the carotenes and the 10 CMTs were determined by bivariate genetic analysis. Bivariate genetic analysis is an extension of VCA where the phenotypic correlation (ρP) between a pair of phenotypes (e.g., β-carotene and BMI) can be partitioned into additive genetic (ρG) and environmental (ρE) correlations (30). The phenotypic correlation (ρP) between a pair of traits is given by the equation:

graphic file with name ajcn144006equ1.jpg

where ρP is the phenotypic correlation; ρG is the additive genetic correlation; ρE is the random environmental correlation; h21 is the heritability of trait 1; h22 is the heritability of trait 2; e21 is equal to 1 − h21; and e22 is equal to 1 − h22. With the use of likelihood ratio tests, the significance (P ≤ 0.05) of the phenotypic, additive genetic, and random environmental correlation was determined, respectively. The additive genetic correlation (ρG) is a measure of the shared genetic basis of the 2 traits (i.e., pleiotropy). To account for multiple comparisons, we obtained the false discovery rate–adjusted P values following the approach by Benjamini and Hochberg (31) as implemented in the SDM Project’s Web-based software (www.sdmproject.com).

RESULTS

The 673 children who participated in the SAFARI study represented 401 high-risk MA nuclear families (∼2 children/nuclear family; range: 1–5 children) and generated 3664 relative pairs. Of the 673 children, 3 were diagnosed with T2D, and they were excluded from all subsequent analyses (Figure 1). The characteristics of 670 nondiabetic children are described in Table 1. Both α- and β-carotenes were found to be highly heritable (α-carotene h2 = 0.81; β-carotene h2 = 0.90). As shown in Table 1, the significant h2 estimates for the 10 CMTs were reported previously (26), which ranged from 0.39 (fasting glucose) to 0.77 (triglycerides).

TABLE 1.

Characteristics of the 670 nondiabetic SAFARI children and heritability estimates for selected obesity-related traits and carotenoids1

Variable N Mean ± SD or % h2 ± SE P
Girls 670 49.3 — —
Age, y 670 11.5 ± 3.5 — —
Overweight2 670 52.7 — —
Obese2 670 33.6 — —
Prediabetes2 630 13.2 — —
MS2 625 18.7 — —
Acanthosis nigricans2 661 33.1 — —
BMI, kg/m2 670 22.7 ± 6.5 0.75 ± 0.11 1.1 × 10−11
WC, mm 664 764.5 ± 179.7 0.63 ± 0.12 3.0 × 10−8
FM, kg (DXA) 634 16.0 ± 11.1 0.69 ± 0.12 1.8 ×10−9
FG, mg/dL 630 89.5 ± 7.5 0.39 ± 0.11 6.3 × 10−5
FI, μIU/mL 626 13.6 ± 9.4 0.55 ± 0.11 2.0 ×10−7
HOMA-IR 622 2.0 ± 1.3 0.60 ± 0.11 1.8 × 10−8
HDL-C, mg/dL 623 45.8 ± 10.9 0.64 ± 0.12 2.9 ×10−8
TGs, mg/dL 623 74.9 ± 39.8 0.77 ± 0.11 8.8 × 10−13
SBP, mm Hg 670 104.1 ± 9.7 0.66 ± 0.11 1.0 × 10−10
DBP, mm Hg 670 63.2 ± 7.0 0.64 ± 0.11 9.7 × 10−10
α-Carotene,3 μmol/L 565 0.60 ± 0.97 0.81 ± 0.12 6.7 × 10−11
β-Carotene,3 μmol/L 572 0.48 ± 0.59 0.90 ± 0.11 3.6 × 10−15
Open in a new tab
1

DBP, diastolic blood pressure; DXA, dual-energy X-ray absorptiometry; FG, fasting glucose; FI, fasting insulin; FM, fat mass; HDL-C, HDL cholesterol; MS, metabolic syndrome; SAFARI, San Antonio Family Assessment of Metabolic Risk Indicators in Youth; SBP, systolic blood pressure; TG, triglyceride; WC, waist circumference.

2

Overweight, obese, prediabetes, MS, and acanthosis nigricans as defined and reported by Fowler et al. (26).

3

Traits were inverse-normalized for the genetic analyses with the use of variance components analysis as implemented in the program SOLAR 6.4.9 and were adjusted for the significant covariate effects of age and sex terms for the genetic analyses. The findings related to the 10 CMTs were adapted from Fowler et al. (26).

The results of phenotypic (ρP), genetic (ρG), and environmental (ρE) correlations between α-carotene and β-carotene and the 10 CMTs, with and without accounting for multiple comparisons, are shown in Tables 2 and 3. As reported in Table 2, of the examined trait pairs involving α-carotene, after adjusting for multiple comparisons, significant negative ρPs were found between α-carotene and WC, BMI, triglycerides, and FM [range: −0.19 (WC) to −0.12 (triglycerides)]. In contrast, the ρP between α-carotene and HDL cholesterol was positive (0.17) and significant. Likewise, β-carotene exhibited a significant positive ρP (0.24) with HDL cholesterol and significant negative ρPs with WC, BMI, triglycerides, and FM, which ranged from −0.24 (WC) to −0.13 (triglycerides). When the ρPs were partitioned into ρGs and ρEs, after accounting for multiple comparisons, no ρGs or ρEs between α-carotene or β-carotene and any trait were statistically significant. However, the ρGs between the 2 obesity-related traits of WC and BMI and β-carotene were found to be marginally significant (adjusted P = 0.051). Like the phenotypic correlations, these ρGs were negative (WC = −0.30 and BMI = −0.27) with the exception of HDL cholesterol (0.31, adjusted P = 0.051), which suggests that the phenotypic correlations involving these 3 traits may be influenced by genetic correlations. It should be noted that a positive genetic correlation indicates that the same genetic factors are responsible for increased (or decreased) levels of the 2 traits together, whereas a negative correlation indicates that the same genetic factors are responsible for increased levels of one trait and the decreased levels of the other trait. Also, β-carotene was highly correlated with α-carotene, showing a significant positive phenotypic correlation (0.75, P = 2.5 × 10−8), negative genetic correlation (−0.76, P = 4.5 × 10−9), and no significant environmental correlation.

TABLE 2.

Phenotypic (ρP), genetic (ρG), and environmental (ρE) correlations between α-carotene and obesity-related traits1

Trait pair ρP (95% CI)2 P (adjusted P)3 ρG (95% CI) P (adjusted P) ρE (95% CI) P (adjusted P)
WC −0.19 (−0.26, −0.11) 3.9 × 10−5 (0.0003)* −0.13 (−0.38, 0.12) 0.3287 (0.6714) −0.38 (−1, 0.26) 0.2775 (0.6892)
BMI4 −0.18 (−0.25, −0.10) 0.0001 (0.0003)* −0.13 (−0.36, 0.10) 0.2720 (0.6714) −0.36 (−1, 0.42) 0.3830 (0.6892)
HDL-C 0.17 (0.09, 0.25) 0.0001 (0.0003)* 0.18 (−0.07, 0.43) 0.1851 (0.6714) 0.17 (−0.43, 0.77) 0.5887 (0.7359)
TGs4 −0.12 (−0.20, −0.03) 0.0110 (0.0220)* −0.21 (−0.46, 0.04) 0.0980 (0.6714) 0.23 (−0.51, 0.97) 0.5315 (0.7359)
FM4 (DXA) −0.15 (−0.23, −0.06) 0.0011 (0.0028)* −0.04 (−0.29, 0.21) 0.7519 (0.9651) −0.56 (−1, 0.18) 0.1336 (0.6892)
SBP −0.08 (−0.16, 0.001) 0.0718 (0.0898) −0.13 (−0.38, 0.12) 0.3357 (0.6714) 0.05 (−0.55, 0.65) 0.8807 (0.8807)
DBP −0.03 (−0.11, 0.05) 0.6500 (0.6500) 0.02 (−0.23, 0.27) 0.9013 (0.9651) −0.13 (−0.71, 0.45) 0.6708 (0.7453)
FG4 −0.04 (−0.12, 0.04) 0.3102 (0.3447) 0.05 (−0.28, 0.38) 0.7696 (0.9651) −0.22 (−0.72, 0.28) 0.3668 (0.6892)
FI4 −0.09 (−0.17, −0.01) 0.0500* (0.0773) −0.01 (−0.28, 0.26) 0.9651 (0.9651) −0.30 (−0.82, 0.22) 0.2739 (0.6892)
HOMA-IR5 −0.09 (−0.17, −0.01) 0.0541 (0.0773) −0.03 (−0.28, 0.22) 0.8013 (0.9651) −0.24 (−0.82, 0.34) 0.4135 (0.6892)
Open in a new tab
1

Phenotypic correlation (ρP) between a given trait pair is partitioned into genetic (ρG) and environmental (ρE) correlations with the use of bivariate genetic analysis within the variance components framework with the use of the program SOLAR 6.4.9. All traits were adjusted for the covariate effects of age and sex terms. DBP, diastolic blood pressure; DXA, dual-energy X-ray absorptiometry; FG, fasting glucose; FI, fasting insulin; FM, fat mass; HDL-C, HDL cholesterol; SBP, systolic blood pressure; TG, triglyceride; WC, waist circumference.

2

Asymptotic 95% CIs.

3

False discovery rate–adjusted P values that account for multiple comparisons are shown within parentheses for each set (i.e., ρPs, ρGs, and ρEs) of 10 P values. *Refers to a significant finding.

4

Data were log-transformed.

5

Trait was inverse-normalized.

TABLE 3.

Phenotypic (ρP), genetic (ρG), and environmental (ρE) correlations between β-carotene and obesity-related traits1

Trait pair ρP (95% CI)2 P (adjusted P)3 ρG (95% CI) P (adjusted P) ρE (95% CI) P (adjusted P)
WC −0.24 (−0.31, −0.16) 4.78 × 10−8 (4.5 × 10−7)* −0.30 (−0.51, −0.08) 0.0143 (0.0510)* −0.01 (−0.93, 0.91) 0.9770 (0.9923)
BMI4 −0.23 (−0.30, −0.15) 1.8 × 10−7 (6.0 × 10−7)* −0.27 (−0.48, −0.05) 0.0153 (0.0510)* −0.01 (−1, 1) 0.9923 (0.9923)
HDL-C 0.24 (0.16, 0.32) 9.0 × 10−8 (4.5 × 10−7)* 0.31 (0.07, 0.54) 0.0121 (0.0510)* −0.02 (−0.80, 0.76) 0.9658 (0.9923)
TGs4 −0.13 (−0.21, −0.04) 4.2 × 10−3 (0.0084)* −0.23 (−0.44, −0.01) 0.0458 (0.0916) 0.42 (−0.71, 1.55) 0.3868 (0.9670)
FM (DXA)4 −0.22 (−0.30, −0.13) 1.6 × 10−6 (4.0 × 10−6)* −0.24 (−0.45, −0.02) 0.0400 (0.0916) −0.11 (−1, 0.81) 0.8191 (0.9923)
SBP −0.08 (−0.16, −0.01) 0.0896 (0.1280) −0.12 (−0.35, 0.11) 0.3303 (0.4719) 0.09 (−0.71, 0.89) 0.8301 (0.9923)
DBP −0.05 (−0.13, 0.03) 0.4448 (0.4481) 0.05 (−0.20, 0.30) 0.6871 (0.6871) −0.40 (−1, 0.50) 0.3227 (0.9670)
FI4 −0.07 (−0.15, 0.01) 0.1345 (0.1531) 0.10 (−0.15, 0.35) 0.4228 (0.5285) −0.72 (−1, 0.29) 0.0521 (0.5210)
FG4 −0.08 (−0.16, 0.01) 0.0523 (0.0872) −0.24 (−0.55, 0.07) 0.1440 (0.2400) 0.19 (−0.42, 0.80) 0.5178 (0.9923)
HOMA-IR5 −0.07 (−0.15, 0.01) 0.1378 (0.1531) 0.06 (−0.17, 0.29) 0.6280 (0.6871) −0.61 (−1, 0.37) 0.1247 (0.6235)
Open in a new tab
1

Phenotypic correlation (ρP) between a given trait pair is partitioned into genetic (ρG) and environmental (ρE) correlations with the use of bivariate genetic analysis within the variance components framework with the use of the program SOLAR 6.4.9. All traits were adjusted for the covariate effects of age and sex terms. DBP, diastolic blood pressure; DXA, dual-energy X-ray absorptiometry; FG, fasting glucose; FI, fasting insulin; FM, fat mass; HDL-C, HDL cholesterol; SBP, systolic blood pressure; TG, triglyceride; WC, waist circumference.

2

Asymptotic 95% CIs.

3

False discovery rate–adjusted P values that account for multiple comparisons are shown within parentheses for each set (i.e., ρPs, ρGs, and ρEs) of 10 P values. *Refers to a significant or marginally significant finding.

4

Data were log-transformed.

5

Trait was inverse-normalized.

DISCUSSION

The present study determined the heritability of circulating concentrations of α-carotene and β-carotene in the MA cohort of children and adolescents of the SAFARI study, who are at high risk of obesity, prediabetes, and MS. We tested the relations between these 2 carotenoids and 10 CMTs among these children. The main findings from this study are as follows: 1) α-carotene and β-carotene serum concentrations are under strong additive genetic influences; 2) several CMTs are appreciably phenotypically correlated with α-carotene and β-carotene serum concentrations; and 3) the observed positive phenotypic correlation between β-carotene and HDL-cholesterol concentrations and inverse phenotypic correlations between β-carotene and WC and BMI may be influenced by genetic correlations.

Although several genetic studies have provided evidence that circulating serum carotenoid concentrations are under genetic influence, our study reports, for the first time to our knowledge, the highly significant heritability estimates for α-carotene (0.81) and β-carotene (0.90) in MA children (22, 24, 32, 33). Also, there is evidence that serum retinol, a biologically active form of vitamin A and the breakdown product of α-carotene and β-carotene, is under genetic control. Gueguen et al. (34) reported heritability of 30% for serum retinol in a cohort of 387 French families. It is possible that our heritability estimates of carotenoids may have been inflated because our analyses did not account for shared environmental influences. As we noted previously (26), however, such influences appear to be minimal in our study because the SAFARI children are distributed across large pedigrees, represented by a wide variety of relative pairs. Indeed, as noted by Chesi and Grant (35), the pediatric populations appear to be optimal for localization of susceptibility loci for traits, such as obesity, owing to higher heritability estimates and less environmental exposures.

In agreement with previously published studies, we detected strong negative phenotypic correlations between the carotenoids (α-carotene and/or β-carotene) and BMI, WC, FM, and triglycerides. For example, analyses of data from MA children ranging between 8–15 y included in the 2001–2004 NHANES showed a negative correlation between these carotenoids and BMI and FM (16). They also found higher concentrations of these carotenoids to be associated with a reduced risk of obesity in MA children. Similar to these findings, significant lower serum carotenoid concentrations were observed in overweight or obese Brazilian, French, and Italian children when compared with normal-weight children (14, 36, 37). A positive phenotypic correlation was observed between α-carotene and/or β-carotene concentrations and HDL cholesterol in our study. This finding is also in agreement with the large NHANES 2003–2006 study that showed several serum carotenoids positively correlated with HDL-cholesterol concentrations (38). Recently, Suarez and Schramm-Sapyta (39) reported lower β-carotene concentrations to be associated with higher estimates of insulin resistance and fasting insulin in African Americans. They did not find such a correlation among their white study participants. Although not statistically significant, a similar trend was observed between α-carotene or β-carotene and fasting insulin and HOMA-IR in our study. Together, these findings suggest that there are ethnic disparities that account for the observed correlations between the carotenoid concentrations and the CMTs.

In contrast to the population-based studies that examined only the phenotypic correlations between carotenoids and CMTs, with the use of family data, we partitioned the phenotypic correlations into the genetic and environmental correlations. However, after adjusting for multiple comparisons, none of the genetic and environmental correlations between the carotenoids and CMTs were found to be statistically significant. It is worth noting that the genetic correlations involving the 2 obesity-related traits (WC and BMI) and β-carotene, but not α-carotene, were found to be marginally significant. In consideration of these observations, it should be noted that β-carotene is the major source of vitamin A in the body (40) and is present at higher concentrations than α-carotene in serum in both adults and children (41, 42). In addition, the phenotypic correlation between the 2 carotenoids in our data indicated that only 56% of the variation is commonly shared by the 2 carotenoids, in turn suggesting that there was a distinct sizable proportion of variance between α-carotene and β-carotene concentrations. Our data suggest that β-carotene concentrations and the above 3 CMTs (i.e., measures of obesity and dyslipidemia) may be influenced by common genetic factors (i.e., pleiotropy) in MA children.

As stated earlier, several studies have localized genetic variants associated with various carotenoids, including α-carotene and β-carotene, with the use of the genome-wide association analytic approach (1, 23–25). Likewise, with the use of a similar association approach, numerous genetic variants have been localized for phenotypes, such as obesity measures (BMI and WC), lipid and lipoproteins, and T2D and its related traits ( 35, 43–45). In consideration of our findings of a genetic basis of variation in carotenoids and their (i.e., β-carotene) potential genetic correlation with obesity-related traits, our immediate plans are to localize genetic variants that contribute to variation in carotenoids and to identify genetic variants that commonly influence both carotenoids and obesity-related traits.

Importantly, our finding of an association between carotenoid concentrations and obesity-related traits in MA children and adolescents have ethnic-specific and general public health relevance in terms of promoting fruit and vegetable consumption given its beneficial effects and facilitating tailored individualized programs for dietary interventions. However, our findings should be interpreted with caution given the nature of our study sample. It is evident from previous studies that there are socioeconomic, neighborhood environment, acculturation pattern, and ethnic differences in regard to fruit and/or vegetable intake (46–49). For example, using the NHANES III (1988–1994) data of MA children (aged 4–16 y), it was reported that the Mexico-born MA children compared with the MA children born in the United States had higher concentrations of certain antioxidants, including α-carotene and β-carotene (49). In another Hispanic study including MAs (San Diego County, California), child fruit consumption, but not vegetable consumption, was associated with parent acculturation (47). These studies, taken together with our observations, highlight the need for culturally relevant, early-life dietary behavioral interventions to reduce the burden of obesity and its associated cardiometabolic risk not only in childhood, but also in adulthood.

In conclusion, both α-carotene and β-carotene are found to be under strong additive genetic influences in SAFARI MA children and adolescents. In addition, the inverse phenotypic correlations between β-carotene and obesity-related traits and the positive phenotypic correlation between β-carotene and HDL-cholesterol concentrations may be influenced by genetic correlations. However, the issue of potential common genetic influences on these traits warrants additional studies, particularly involving Hispanic children and adolescents. Importantly, in terms of dietary behavioral interventions, our data also suggest the potential beneficial attributes of the intake of fruits and vegetables in MA children given the observed correlations between them and susceptibility to childhood obesity and its clinical correlates.

Acknowledgments

We thank the University Health System and the Texas Diabetes Institute for extending their clinical research facilities to the SAFARI study. We also thank the participants of the SAFARI study for their enthusiasm and cooperation.

The author’ responsibilities were as follows—VSF: designed the research, conducted the research, analyzed the data, and wrote the manuscript; LR: conducted the research, analyzed the data, and wrote the manuscript; SM: wrote the manuscript; BMK: assisted with sample extraction for the carotenoid analysis; SP, RA, and GC: conducted the research and analyzed the data; SPF and RGR: conducted the research; JCL-A, AGC, JEC, DML, CPJ, JLL, RAD, JB, and DEH: designed the research or analyzed the data and reviewed and/or edited the manuscript; RD and JKPV: designed the research, provided study oversight, analyzed the data, wrote the manuscript, and had primary responsibility for the final content and shepherded the manuscript through the peer-review process; and all authors: critically revised, read, and approved the final manuscript before submission. None of the authors reported a conflict of interest related to the study.

Footnotes

Abbreviations used: CMT, cardiometabolic trait; FM, fat mass; MA, Mexican American; MS, metabolic syndrome; QC, quality control; SAFARI, San Antonio Family Assessment of Metabolic Risk Indicators in Youth; T2D, type 2 diabetes; UPLC, ultraperformance liquid chromatography; WC, waist circumference.

REFERENCES

  • 1.D’Adamo CR, D’Urso A, Ryan KA, Yerges-Armstrong LM, Semba RD, Steinle NI, Mitchell BD, Shuldiner AR, McArdle PF. A common variant in the SETD7 gene predicts serum lycopene concentrations. Nutrients 2016;8:82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Berenson GS. Childhood risk factors predict adult risk associated with subclinical cardiovascular disease. The Bogalusa Heart Study. Am J Cardiol 2002;90(10C):3L–7L. [DOI] [PubMed] [Google Scholar]
  • 3.Deboer MD. Ethnicity, obesity and the metabolic syndrome: implications on assessing risk and targeting intervention. Expert Rev Endocrinol Metab 2011;6:279–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Must A, Spadano J, Coakley EH, Field AE, Colditz G, Dietz WH. The disease burden associated with overweight and obesity. JAMA 1999;282:1523–9. [DOI] [PubMed] [Google Scholar]
  • 5.Ogden CL, Carroll MD, Curtin LR, Lamb MM, Flegal KM. Prevalence of high body mass index in US children and adolescents, 2007-2008. JAMA 2010;303:242–9. [DOI] [PubMed] [Google Scholar]
  • 6.Willett WC. Micronutrients and cancer risk. Am J Clin Nutr 1994;59(5 Suppl):1162S–5S. [DOI] [PubMed] [Google Scholar]
  • 7.Britton G, Liaaen-Jensen S, Pfander H. Carotenoids handbook. Basel (Switzerland); Birkhäuser Verlag; 2004. [Google Scholar]
  • 8.Maiani G, Caston MJ, Catasta G, Toti E, Cambrodon IG, Bysted A, Granado-Lorencio F, Olmedilla-Alonso B, Knuthsen P, Valoti M, et al. Carotenoids: actual knowledge on food sources, intakes, stability and bioavailability and their protective role in humans. Mol Nutr Food Res 2009;53(Suppl 2):S194–218. [DOI] [PubMed] [Google Scholar]
  • 9.Rao AV, Rao LG. Carotenoids and human health. Pharmacol Res 2007;55:207–16. [DOI] [PubMed] [Google Scholar]
  • 10.De Luca LM. Retinoids and their receptors in differentiation, embryogenesis, and neoplasia. FASEB J 1991;5:2924–33. [PubMed] [Google Scholar]
  • 11.Olson JA, Krinsky NI. Introduction: the colorful, fascinating world of the carotenoids: important physiologic modulators. FASEB J 1995;9:1547–50. [DOI] [PubMed] [Google Scholar]
  • 12.Aasheim ET, Hofso D, Hjelmesaeth J, Birkeland KI, Bohmer T. Vitamin status in morbidly obese patients: a cross-sectional study. Am J Clin Nutr 2008;87:362–9. [DOI] [PubMed] [Google Scholar]
  • 13.Burrows TL, Warren JM, Colyvas K, Garg ML, Collins CE. Validation of overweight children’s fruit and vegetable intake using plasma carotenoids. Obesity (Silver Spring) 2009;17:162–8. [DOI] [PubMed] [Google Scholar]
  • 14.de Souza Valente da Silva L, Valeria da Veiga G, Ramalho RA. Association of serum concentrations of retinol and carotenoids with overweight in children and adolescents. Nutrition 2007;23:392–7. [DOI] [PubMed] [Google Scholar]
  • 15.Ford ES, Gillespie C, Ballew C, Sowell A, Mannino DM. Serum carotenoid concentrations in US children and adolescents. Am J Clin Nutr 2002;76:818–27. [DOI] [PubMed] [Google Scholar]
  • 16.Gunanti IR, Marks GC, Al-Mamun A, Long KZ. Low serum concentrations of carotenoids and vitamin E are associated with high adiposity in Mexican-American children. J Nutr 2014;144:489–95. [DOI] [PubMed] [Google Scholar]
  • 17.Mohn A, Catino M, Capanna R, Giannini C, Marcovecchio M, Chiarelli F. Increased oxidative stress in prepubertal severely obese children: effect of a dietary restriction-weight loss program. J Clin Endocrinol Metab 2005;90:2653–8. [DOI] [PubMed] [Google Scholar]
  • 18.Neuhouser ML, Rock CL, Eldridge AL, Kristal AR, Patterson RE, Cooper DA, Neumark-Sztainer D, Cheskin LJ, Thornquist MD. Serum concentrations of retinol, alpha-tocopherol and the carotenoids are influenced by diet, race and obesity in a sample of healthy adolescents. J Nutr 2001;131:2184–91. [DOI] [PubMed] [Google Scholar]
  • 19.Suzuki K, Inoue T, Hioki R, Ochiai J, Kusuhara Y, Ichino N, Osakabe K, Hamajima N, Ito Y. Association of abdominal obesity with decreased serum levels of carotenoids in a healthy Japanese population. Clin Nutr 2006;25:780–9. [DOI] [PubMed] [Google Scholar]
  • 20.Donaldson MS. A carotenoid health index based on plasma carotenoids and health outcomes. Nutrients 2011;3:1003–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Borel P, Grolier P, Mekki N, Boirie Y, Rochette Y, Le Roy B, Alexandre-Gouabau MC, Lairon D, Azais-Braesco V. Low and high responders to pharmacological doses of beta-carotene: proportion in the population, mechanisms involved and consequences on beta-carotene metabolism. J Lipid Res 1998;39:2250–60. [PubMed] [Google Scholar]
  • 22.Borel P, de Edelenyi FS, Vincent-Baudry S, Malezet-Desmoulin C, Margotat A, Lyan B, Gorrand JM, Meunier N, Drouault-Holowacz S, Bieuvelet S. Genetic variants in BCMO1 and CD36 are associated with plasma lutein concentrations and macular pigment optical density in humans. Ann Med 2011;43:47–59. [DOI] [PubMed] [Google Scholar]
  • 23.Beydoun MA, Nalls MA, Canas JA, Evans MK, Zonderman AB. Gene polymorphisms and gene scores linked to low serum carotenoid status and their associations with metabolic disturbance and depressive symptoms in African-American adults. Br J Nutr 2014;112:992–1003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Ferrucci L, Perry JR, Matteini A, Perola M, Tanaka T, Silander K, Rice N, Melzer D, Murray A, Cluett C, et al. Common variation in the beta-carotene 15,15′-monooxygenase 1 gene affects circulating levels of carotenoids: a genome-wide association study. Am J Hum Genet 2009;84:123–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Zubair N, Kooperberg C, Liu J, Di C, Peters U, Neuhouser ML. Genetic variation predicts serum lycopene concentrations in a multiethnic population of postmenopausal women. J Nutr 2015;145:187–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Fowler SP, Puppala S, Arya R, Chittoor G, Farook VS, Schneider J, Resendez RG, Upadhayay RP, Vandeberg J, Hunt KJ, et al. Genetic epidemiology of cardiometabolic risk factors and their clustering patterns in Mexican American children and adolescents: the SAFARI Study. Hum Genet 2013;132:1059–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Almasy L, Blangero J. Multipoint quantitative-trait linkage analysis in general pedigrees. Am J Hum Genet 1998;62:1198–211. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Falconer DS. Introduction to quantitative genetics, 3rd ed. New York: Longman, Scientific & Technical; 1989. [Google Scholar]
  • 29.Hopper JL, Mathews JD. Extensions to multivariate normal models for pedigree analysis. Ann Hum Genet 1982;46:373–83. [DOI] [PubMed] [Google Scholar]
  • 30.Lange K, Boehnke M. Extensions to pedigree analysis. IV. Covariance components models for multivariate traits. Am J Med Genet 1983;14:513–24. [DOI] [PubMed] [Google Scholar]
  • 31.Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc B 1995;57:289–300. [Google Scholar]
  • 32.Borel P. Genetic variations involved in interindividual variability in carotenoid status. Mol Nutr Food Res 2012;56:228–40. [DOI] [PubMed] [Google Scholar]
  • 33.Borel P, Lietz G, Goncalves A, Szabo de Edelenyi F, Lecompte S, Curtis P, Goumidi L, Caslake MJ, Miles EA, Packard C, et al. CD36 and SR-BI are involved in cellular uptake of provitamin A carotenoids by Caco-2 and HEK cells, and some of their genetic variants are associated with plasma concentrations of these micronutrients in humans. J Nutr 2013;143:448–56. [DOI] [PubMed] [Google Scholar]
  • 34.Gueguen S, Leroy P, Gueguen R, Siest G, Visvikis S, Herbeth B. Genetic and environmental contributions to serum retinol and alpha-tocopherol concentrations: the Stanislas Family Study. Am J Clin Nutr 2005;81:1034–44. [DOI] [PubMed] [Google Scholar]
  • 35.Chesi A, Grant SF. The genetics of pediatric obesity. Trends Endocrinol Metab 2015;26:711–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Herbeth B, Spyckerelle Y, Deschamps JP. Determinants of plasma retinol, beta-carotene, and alpha-tocopherol during adolescence. Am J Clin Nutr 1991;54:884–9. [DOI] [PubMed] [Google Scholar]
  • 37.Sarni RO, Suano de Souza FI, Ramalho RA, Schoeps Dde O, Kochi C, Catherino P, Dias MC, Pessotti CF, Mattoso LC, Colugnat FA. Serum retinol and total carotene concentrations in obese pre-school children. Med Sci Monit 2005;11:CR510–4. [PubMed] [Google Scholar]
  • 38.Wang Y, Chung SJ, McCullough ML, Song WO, Fernandez ML, Koo SI, Chun OK. Dietary carotenoids are associated with cardiovascular disease risk biomarkers mediated by serum carotenoid concentrations. J Nutr 2014;144:1067–74. [DOI] [PubMed] [Google Scholar]
  • 39.Suarez EC, Schramm-Sapyta NL. Race differences in the relation of vitamins A, C, E, and beta-carotene to metabolic and inflammatory biomarkers. Nutr Res 2014;34:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.von Lintig J. Provitamin A metabolism and functions in mammalian biology. Am J Clin Nutr 2012;96:1234S–44S. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Homnick DN, Cox JH, DeLoof MJ, Ringer TV. Carotenoid levels in normal children and in children with cystic fibrosis. J Pediatr 1993;122:703–7. [DOI] [PubMed] [Google Scholar]
  • 42.Ito Y, Ochiai J, Sasaki R, Suzuki S, Kusuhara Y, Morimitsu Y, Otani M, Aoki K. Serum concentrations of carotenoids, retinol, and alpha-tocopherol in healthy persons determined by high-performance liquid chromatography. Clin Chim Acta 1990;194:131–44. [DOI] [PubMed] [Google Scholar]
  • 43.Arya R, Puppala S, Farook VS, Chittoor G, Jenkinson CP, Blangero J, Hale DE, Duggirala R, Almasy L. Mapping of susceptibility genes for obesity, type 2 diabetes, and the metabolic syndrome in human populations. Heidelberg (Germany): Springer-Verlag; 2015. [Google Scholar]
  • 44.Karaderi T, Drong AW, Lindgren CM. Insights into the genetic susceptibility to type 2 diabetes from genome-wide association studies of obesity-related traits. Curr Diab Rep 2015;15:83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Tada H, Kawashiri MA, Konno T, Yamagishi M, Hayashi K. Common and rare variant association study for plasma lipids and coronary artery disease. J Atheroscler Thromb 2016;23:241–56. [DOI] [PubMed] [Google Scholar]
  • 46.Dubowitz T, Heron M, Bird CE, Lurie N, Finch BK, Basurto-Dávila R, Hale L, Escarce JJ. Neighborhood socioeconomic status and fruit and vegetable intake among whites, blacks, and Mexican Americans in the United States. Am J Clin Nutr 2008;87:1883–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Morello MI, Madanat H, Crespo NC, Lemus H, Elder J. Associations among parent acculturation, child BMI, and child fruit and vegetable consumption in a Hispanic sample. J Immigr Minor Health 2012;14:1023–9. [DOI] [PubMed] [Google Scholar]
  • 48.Sharma S, Sheehy T, Kolonel L. Sources of vegetables, fruits, and vitamins A, C, and E among five ethnic groups: results from the Multiethnic Cohort Study. Eur J Clin Nutr 2014;68:384–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Eldeirawi K, Koenig MD, Persky V, Chavez N. Nativity and serum concentrations of antioxidants in Mexican American children: a cross-sectional study. Nutrients 2014;6:1598–607. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from The American Journal of Clinical Nutrition are provided here courtesy of American Society for Nutrition

RESOURCES