European Journal of Clinical Nutrition (2015), 1–7 © 2015 Macmillan Publishers Limited All rights reserved 0954-3007/15 www.nature.com/ejcn
ORIGINAL ARTICLE
Longitudinal association of measures of adiposity with serum antioxidant concentrations in postmenopausal women GC Kabat1, M Heo1, HM Ochs-Balcom2, MS LeBoff3, Y Mossavar-Rahmani1, LL Adams-Campbell4, R Nassir5, J Ard6, O Zaslavsky7 and TE Rohan1 BACKGROUND/OBJECTIVES: The relationship between obesity and circulating levels of antioxidants is poorly understood. Most studies that have examined the association of adiposity with blood or tissue concentrations of antioxidant micronutrients have been cross-sectional, and few have compared the associations for indices of overall obesity and central obesity. Our aim was to prospectively examine the longitudinal association of body mass index (BMI), waist circumference (WC), waist circumference-height ratio (WCHtR) and waist-hip ratio (WHR) with major serum antioxidants in a population of postmenopausal women. SUBJECTS/METHODS: We used a subsample of participants in the Women’s Health Initiative aged 50–79 years at entry with available fasting blood samples and anthropometric measurements obtained at multiple time points over 12.8 years of follow-up (N = 2672). Blood samples were used to measure α-carotene, β-carotene, β-cryptoxanthin, lutein+zeaxanthin, α-tocopherol, γtocopherol and retinol at baseline, and at years 1, 3 and 6. We used mixed-effects linear regression analyses to examine associations between anthropometric measures and serum antioxidants at baseline and over time, controlling for covariates. RESULTS: In longitudinal analyses, carotenoids, and particularly β-carotene, were strongly and inversely associated with BMI, WC and WCHtR and less so with WHR. α-Tocopherol showed a strong positive association with WHR but not with other anthropometric measures, whereas γ-tocopherol was positively and strongly associated with BMI, WC, WCHtR and less so with WHR. Retinol was positively associated with WHR. The inverse association of several carotenoids with anthropometric measures was stronger in never and former smokers compared with current smokers and in women without the metabolic syndrome. The inverse association of carotenoids with obesity measures may reflect reduced micronutrient concentrations owing to inflammation associated with obesity. CONCLUSIONS: In the present study, the strongest observed associations between anthropometric variables and micronutrients were an inverse association of WC with serum β-carotene and a positive association of WC with γ-tocopherol. European Journal of Clinical Nutrition advance online publication, 27 May 2015; doi:10.1038/ejcn.2015.74
INTRODUCTION Obesity and central obesity are established risk factors for cardiovascular disease, diabetes and certain cancers.1–3 Obesity is associated with chronic systemic inflammation and increased oxidative stress,4,5 which are hypothesized to play critical roles in disease progression. Antioxidant micronutrients may prevent damage from inflammation and reactive oxygen species, thereby reducing the deleterious effects of obesity.5 On the other hand, obesity may place a greater demand on requirements for antioxidants4 and may cause increased utilization of antioxidants, thus leading to reduced concentrations.5 Therefore, it is important to understand the interplay between obesity and circulating antioxidant nutrients. A number of studies have examined the association between obesity and blood or tissue levels of antioxidant micronutrients, most often carotenoids,5–17 and have mostly reported inverse associations of obesity with circulating levels of carotenoids. A smaller number of studies have examined the association of obesity with other micronutrients, including tocopherols9,11,14,17 and retinol.7,9 All of these studies but one15 were cross-sectional, making it difficult to interpret the temporal relationship between
obesity and antioxidant concentrations. In addition, most studies assessed overall obesity (that is, BMI),5,6,9,10,12,14–17 whereas fewer studies examined abdominal obesity (that is, waist circumference (WC) or waist-hip ratio (WHR)).7,11,13 Only one study15 stratified the analysis by smoking status (smokers vs nonsmokers), an important consideration, given that smoking is believed to result in increased levels of reactive oxygen species and to deplete antioxidant stores, and that carotenoid concentrations have been found consistently to be lower among smokers compared with nonsmokers.18,19 That study found significant inverse associations of BMI with carotenoid levels only among nonsmokers. In the present analysis, we used repeated measurements of anthropometric indices of general and central obesity and of serum micronutrients in a sub-cohort of the Women’s Health Initiative to assess the longitudinal association of body weight with serum concentrations of antioxidants. PARTICIPANTS AND METHODS The current study was conducted using data from the Women’s Health Initiative (WHI), a large, multi-center prospective study designed to
1 Department of Epidemiology and Population Health, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY, USA; 2Department of Epidemiology and Environmental Health, School of Public Health and Health Profession, University at Buffalo, Buffalo, NY, USA; 3Brigham and Women’s Hospital, Boston, MA, USA; 4Lombardi Comprehensive Cancer Center, Georgetown University, Washington D.C., USA; 5Department of Biochemistry and Molecular Medicine, University of California, Davis, CA, USA; 6 Department of Epidemiology and Prevention, Wake Forest School of Medicine, Medical Center Blvd., Winston-Salem, NC, USA and 7Department of Nursing, University of Haifa, Haifa, Israel. Correspondence: Dr GC Kabat, Department of Epidemiology and Population Health, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461, USA. E-mail: [email protected] Received 30 December 2014; revised 20 March 2015; accepted 30 March 2015
Longitudinal association of adiposity measures with antioxidants GC Kabat et al
2 advance our understanding of the determinants of major chronic diseases in older women.20 The WHI is composed of a Clinical Trial component (CT, N = 68 132) and an Observational Study (N = 93 676). Women between the ages of 50 and 79 and representing major racial/ethnic groups were recruited from the general population at 40 clinical centers throughout the US between 1993 and 1998. Details of the design and reliability of the baseline measures have been published.20,21 The study population for the present analysis consisted of a 6% random sample of women in the CT (N = 4544) who provided fasting blood samples at baseline and years 1, 3 and 6 during follow-up and a 1% sample of women in the Observational Study (N = 1062) who provided a fasting blood sample at baseline and at year 3. Of the 5606 women with measured analytes (see below), we restricted our analysis to the (approximately) 2680 women with baseline serum measurements of the analytes who were not in an intervention arm of any clinical trial. Antioxidant nutrients studied here included α-carotene, β-carotene, β-cryptoxanthin, lutein+zeaxanthin, lycopene, α-tocopherol, γ-tocopherol and retinol.
Assessment of diabetes and the metabolic syndrome
Data collection and variable definition
Longitudinal analyses. In the longitudinal analysis, the association between change in anthropometric factors and change in antioxidant nutrients was assessed using repeated measures data and time-varying covariates pertinent to each antioxidant nutrient. Because all micronutrients were associated with total cholesterol concentration, we included total cholesterol as a covariate in the multivariable analyses. To compare the magnitudes of regression coefficients across different anthropometric measures, we converted the repeatedly measured anthropometric measures to their corresponding Z-scores, which center the measures on their means and then divide by their standard deviations. Because use of vitamin supplements can alter serum concentrations of carotenoids and tocopherols, we carried out a sensitivity analysis repeating the longitudinal analysis excluding women who reported taking vitamin supplements (N = 1314). To examine variation in the longitudinal association by level of potential effect modifiers, we stratified the analysis based on the following categorizations: age (o 63 vs ⩾ 63 years), hormone therapy (any type: current vs non-current use, smoking status (never, former, current) and metabolic syndrome (yes, no). The chi-square Q statistic was used to test for heterogeneity of estimated coefficients for the association of the Zscored anthropometric measures with antioxidant nutrients between stratum levels of each potential effect modifier. To limit the number of interactions tested, we selected the anthropometric factor that was most strongly associated with a given micronutrient in the multivariable longitudinal analyses. Thus, we tested the association of WCHtR with αcarotene, β-carotene, β-cryptoxanthin, lutein+zeaxanthin and γ-tocopherol; BMI with lycopene; and WHR with α-tocopherol and retinol. In total, we conducted 32 different tests for inequality of coefficients between strata (8 associations × 4 stratifying factors) (see Supplementary Table).
At study entry, self-administered questionnaires were used to collect information on demographics, medical, reproductive and family history, and on dietary and lifestyle factors, including smoking history, alcohol consumption and recreational physical activity. Participants completed a semiquantitative, 122-item food-frequency questionnaire to measure dietary intake over the past 3 months. Additional questions queried sources of fat intake, added fats and intake of fruits and vegetables. Data were extracted from the baseline questionnaire on estimated dietary intake of α-carotene, β-carotene, β-cryptoxanthin, lycopene, α-tocopherol, γ-tocopherol and retinol. Intake of lutein+zeaxanthin was not computed in the WHI nutrient database. Dietary supplement data were collected during in-person clinic visits. Women were asked to bring supplement bottles to the baseline clinic visit and to subsequent annual visits in the CT, and to the baseline and year 3 visit in the Observational Study. A standardized interviewer-administered four-page questionnaire was used to obtain information on multivitamins, with or without minerals and single supplements. Only supplements that were used at least once a week were recorded, but there was no limit to the number of supplements recorded. All participants had their height, weight, and waist and hip circumferences measured by trained staff at baseline. Weight was measured to the nearest 0.1 kg and height to the nearest 0.1 cm. WC at the natural waist or narrowest part of the torso, and hip circumference at the maximal circumference, were recorded to the nearest 0.1 cm. Anthropometric measurements were also made during follow-up (in years 3 and 6 for the majority of the cohort and, additionally, in years 1 and 9 for a minority of women). Body mass index (BMI) was computed as weight in kilograms divided by the square of height in meters. In addition to WC and WHR, we created the variable waist circumference-height ratio (WCHtR), which in some studies has been shown to be superior to BMI or WC in predicting metabolic risk22 and mortality.23,24 Questions about physical activity at baseline referred to a woman’s usual pattern of activity, including walking and recreational physical activity. A variable ‘current total leisure-time physical activity’ (metabolic equivalent-hours/week) was computed by multiplying the number of hours per week of specific leisuretime physical activities by the metabolic equivalent value of the activities and summing over all types of activities.25
Micronutrient assays Blood samples were collected after an overnight fast (12 h) with minimal stasis and maintained at 4 °C until plasma or serum was separated. Plasma or serum aliquots were then frozen at − 70 °C and sent on dry ice to the WHI central repository (Fisher BioServices, Rockville, MD, USA) for storage at − 70 °C. Retinol, α-carotene, β-carotene, β-cryptoxanthin, lycopene, lutein+zeaxanthin, α-tocopherol and γ-tocopherol were measured in serum by reverse-phase high-performance liquid chromatography.26,27 After the addition of an internal standard, serum was extracted into hexane and injected onto a C18 reverse-phase column. The analytes were measured at wavelengths of 292 and 452 nm. Coefficients of variation were determined in pooled blood samples from four age-eligible female volunteers. The coefficients of variation for the eight analytes ranged from 6.0% (α-tocopherol) to 20.4% (α-carotene). European Journal of Clinical Nutrition (2015) 1 – 7
A history of diabetes was based on self-report of taking diabetes medication or having a fasting glucose of 126 mg/dl or greater at baseline. We used the definition of the metabolic syndrome proposed by the Adult Treatment Panel III (ATP III) of the National Cholesterol Education Program.28,29
Statistical analysis Cross-sectional analyses. The association of a given anthropometric factor with each antioxidant nutrient at baseline was estimated first using Pearson correlation coefficients followed by partial correlation coefficients adjusting for multiple covariates including age, smoking status, alcohol intake, physical activity, educational level, ethnicity and total cholesterol. The association between anthropometric variables and serum micronutrients was assessed cross-sectionally at baseline using multiple regression models, and longitudinally using mixed effects linear regression models including multiple covariates.
RESULTS Baseline characteristics of the study population have been previously described.30 Mean age at enrollment was 63.1 (±7.2) years, and mean BMI was 28.5 (±6.0). The proportions of women with a history of diabetes, cardiovascular disease, hypertension and cancer were 7.8%, 9.8%, 37.9% and 6.0%, respectively. Thirtythree percent of women met the definition of the metabolic syndrome. Cross-sectional associations Significant positive correlations were seen among serum levels of all carotenoids (Table 1), ranging from r = 0.14 between βcryptoxanthin and lycopene to r = 0.54 between α-carotene and β-carotene. Serum α-tocopherol had weak positive associations with the carotenoids, whereas serum γ-tocopherol showed significant inverse correlations with all carotenoids except lycopene. Serum retinol was positively associated with lutein +zeaxanthin, lycopene and α-tocopherol but not with other micronutrients. All eight micronutrients were positively associated with total cholesterol: r ranged from 0.04 for α-carotene to 0.33 for α-tocopherol. Positive associations were seen between serum concentrations of certain micronutrients and dietary intake of © 2015 Macmillan Publishers Limited
Longitudinal association of adiposity measures with antioxidants GC Kabat et al
those nutrients: α-carotene, β-carotene, β-cryptoxanthin, lycopene and γ-tocopherol, but not α-tocopherol or retinol. Partial correlations between baseline concentrations of serum micronutrients and anthropometric measures of obesity are shown in Table 2. After adjustment for covariates, most micronutrients were significantly and inversely associated with anthropometric measures, with the exceptions of γ-tocopherol, which showed a positive association with all anthropometric indices, α-tocopherol which was positively associated with WHR and retinol which was positively associated with WHR. The strongest inverse correlations were seen for α-carotene with BMI, WC and WCHtR (r = − 0.24, − 0.25 and − 0.24, respectively), whereas the strongest positive correlations were seen between γtocopherol and those same indices (0.22, 0.25 and 0.25, respectively). The weakest correlations with anthropometric factors were seen with α-tocopherol and retinol. For all micronutrients except α-tocopherol and retinol, correlations with WHR were weakest. Mean levels of serum micronutrients differed by level of numerous personal characteristics (Supplementary Table). Carotenoid levels were generally higher in women who were below the median for anthropometric indices of obesity compared with Table 1.
3 those in women who were above the median. The largest differences were seen for α-carotene, β-carotene and β-cryptoxanthin (differences of 36%, 36% and 25%, respectively, by level of BMI). In addition, carotenoid levels were higher in those without diabetes or the metabolic syndrome compared with those with these conditions; in those with higher physical activity level; and in never smokers compared with current smokers. Serum levels of αcarotene, β-carotene and β-crytoxanthin were 40%, 35% and 42% higher, respectively, in never smokers compared with current smokers. γ-Tocopherol levels were significantly higher in obese women, women with diabetes or the metabolic syndrome and women who had a lower level of physical activity. Mean concentration of antioxidants differed between women who used vitamin supplements and those who did not. The largest differences were seen for β-carotene and α-tocopherol, concentrations of which were 29% and 26% higher, respectively, in supplement users. However, γ-tocopherol was 41% higher in nonsupplement users. In multiple linear regression analyses of the cross-sectional association of micronutrients with anthropometric measures at baseline, γ-tocopherol was the micronutrient that was most sensitive to anthropometric measures, particularly with BMI, WC
Pearson correlations between serum micronutrients and other factors at baseline (N = 2680) α-Carotene
Serum micronutrients α-Carotene (μg/ml) β-Carotene (μg/ml) β-Cryptoxanthin (μg/ml) Lutein/zeaxanthin (μg/ml) Lycopene (μg/ml) α-Tocopherol (μg/ml) γ-Tocopherol (μg/ml) Retinol (μg/ml) Total cholesterol (mg/dl) Insulin (μIU/ml) Dietary intake α-Carotene (μg/day) β-Carotene (μg/day) β-Cryptoxanthin (μg/day) Lycopene (μg/day) α-Tocopherol (μg/day) γ-Tocopherol (μg/day) Retinol (μg/day) Vitamin A (RAE) Vitamin A (IU) Vitamin A (μg)
β-Carotene 0.54*
0.04** − 0.13*
0.05** − 0.11*
β-Cryptoxanthin
Lutein/zeaxanthin
0.30* 0.36*
0.35* 0.30* 0.36*
0.10* − 0.09*
0.21* − 0.09*
Lycopene
α-Tocopherol
γ-Tocopherol
Retinol
0.20* 0.16* 0.14* 0.21*
0.08* 0.21* 0.15* 0.15* 0.11*
− 0.23* − 0.29* − 0.17* − 0.08* 0.01*** − 0.38*
0.02*** 0.03*** 0.02*** 0.08* 0.09* 0.34* − 0.04***
0.25* − 0.07**
0.33* − 0.03***
0.24* 0.16*
0.26* 0.004***
0.28* 0.19* 0.32* 0.20* 0.02*** 0.20* 0.006*** 0.02*** 0.02*** 0.02***
*Po0.0001; **Po0.05; ***P ⩾ 0.05.
Table 2.
Partial correlationsa between serum micronutrients and body fat measures at baseline (N = 2680)
Anthropometric measures/Antioxidant nutrients α-Carotene (μg/ml) β-Carotene (μg/ml) β-Cryptoxanthin (μg/ml) Lutein/zeaxanthin (μg/ml) Lycopene (μg/ml) α-Tocopherol (μg/ml) γ-Tocopherol (μg/ml) Retinol (μg/ml)
Body mass index (kg/m2)
Waist circumference (cm)
Waist circumference/height
Waist-hip ratio
− 0.24* − 0.22* − 0.17* − 0.22* − 0.10* − 0.04** 0.22* − 0.003***
− 0.25* − 0.23* − 0.19* − 0.22* − 0.10* − 0.01*** 0.25* 0.03****
− 0.24* − 0.22* − 0.17* − 0.21* − 0.10* − 0.009*** 0.25* 0.03***
− 0.15* − 0.14* − 0.12* − 0.10* − 0.03*** 0.05** 0.19* 0.10*
*Po0.0001; **Po0.05; ***P ⩾ 0.05. aAdjusted for age (continuous), servings of alcohol intake per week (continuous), smoking status (current, former, never smoker), physical activity (metabolic equivalent-hours/week—continuous), education (less than high school degree, high school degree/some college, college degree, post college), ethnicity (white, black, other).
© 2015 Macmillan Publishers Limited
European Journal of Clinical Nutrition (2015) 1 – 7
European Journal of Clinical Nutrition (2015) 1 – 7
− 0.018* − 0.060* − 0.015* − 0.022* − 0.018* − 0.133**** 0.223* − 0.0001****
Longitudinal α-Carotene (μg/ml) β-Carotene (μg/ml) β-Cryptoxanthin (μg/ml) Lutein+zeaxanthin (μg/ml) Lycopene (μg/ml) α-Tocopherol (μg/ml) γ-Tocopherol (μg/ml) Retinol (μg/ml) (− 0.020; − 0.015) (− 0.072; − 0.049) (− 0.018; − 0.011) (− 0.026; − 0.019) (− 0.024; − 0.011) (− 0.389; 0.122) (0.182; 0.265) (− 0.005; 0.005)
(− 0.025; − 0.018) (− 0.092; − 0.063) (− 0.020; − 0.012) (− 0.029; − 0.020) (− 0.028; − 0.011) (− 0.622; 0.004) (0.221; 0.331) (− 0.007; 0.005)
95% CI
− 0.018* − 0.064* − 0.016* − 0.023* − 0.016* − 0.001**** 0.239* 0.006**
− 0.022* − 0.081* − 0.017* − 0.025* − 0.020* − 0.179**** 0.299* 0.004****
Β-coefficient
(− 0.021; − 0.016) (− 0.075; − 0.053) (− 0.020; − 0.013) (− 0.026; − 0.019) (− 0.022; − 0.009) (− 0.254; 0.252) (0.198; 0.280) (0.0009; 0.010)
(− 0.026; − 0.019) (− 0.095; − 0.067) (− 0.021; − 0.014) (− 0.029; − 0.020) (− 0.028; − 0.012) (− 0.491; 0.133) (0.244; 0.353) (− 0.002; 0.010)
95% CI
Waist circumference (cm)
− 0.018* − 0.060* − 0.014* − 0.022* − 0.016* 0.086**** 0.234* 0.006***
− 0.022* − 0.077* − 0.016* − 0.024* − 0.022* − 0.158**** 0.305* 0.003****
Β-coefficient
(− 0.020; − 0.015) (− 0.071; − 0.049) (− 0.018; − 0.011) (− 0.026; − 0.019) (− 0.023; − 0.010) (− 0.166; 0.337) (0.193; 0.275) (0.001 − 0.011)
(− 0.025; − 0.018) (− 0.091; − 0.063) (− 0.020; 0.012) (− 0.029; − 0.020) (− 0.030; − 0.013) (− 0.472; 0.155) (0.250; 0.359) (− 0.003; 0.009)
95% CI
Waist circumference/height
− 0.007* − 0.024* − 0.006* − 0.009* − 0.003**** 0.306* 0.076* 0.010*
− 0.013* − 0.045* − 0.010* − 0.012* − 0.009** 0.253**** 0.188* 0.015*
Β-coefficient
(− 0.009; − 0.005) (− 0.033; − 0.015) (− 0.009; − 0.003) (− 0.012; − 0.006) (− 0.009; 0.002) (0.097; 0.515) (0.042; 0.109) (0.006; 0.013)
(− 0.017; − 0.010) (− 0.059; − 0.031) (− 0.014; − 0.007) (− 0.016; − 0.008) (− 0.017; − 0.001) (− 0.046; 0.552) (0.135; 0.240) (0.009; 0.021)
95% CI
Waist-hip ratio
Abbreviation: CI, confidence interval. *Po 0.0001; **P = 0.02–0.03; ***P = 0.01; ****P40.05. aAdjusted for age (continuous), servings of alcohol per week (continuous), smoking status (never, former, current smoker), metabolic equivalent hours/week (continuous), educational level (less than high school degree, high school degree/some college, college degree, post college), ethnicity (white, black, other), hormone therapy (ever, never), oral contraceptive use (ever, never) and total cholesterol (continuous).
− 0.022* − 0.077* − 0.016* − 0.025* − 0.020* − 0.309*** 0.276* − 0.001****
Baseline α-Carotene (μg/ml) β-Carotene (μg/ml) β-Cryptoxanthin (μg/ml) Lutein+zeaxanthin (μg/ml) Lycopene (μg/ml) α-Tocopherol (μg/ml) γ-Tocopherol (μg/ml) Retinol (μg/ml)
Β-coefficient
Body mass index (kg/m2)
Table 3. Association of serum micronutrients with a 1-unit increase in the z-score for anthropometric measures of adiposity after adjusting for covariatesa in mixed linear regression models in the Women’s Health Initiative (N = 2680)
Longitudinal association of adiposity measures with antioxidants GC Kabat et al
4
© 2015 Macmillan Publishers Limited
Longitudinal association of adiposity measures with antioxidants GC Kabat et al
5 and WCHtR (Table 3). The strongest associations (based on comparing β-coefficients) were seen for γ-tocopherol and β-carotene. For example, for each 1 s.d. increase in WC, serum γ-tocopherol increased by 0.299 mg/dl. Carotenoids were significantly inversely associated with anthropometric measures, and the inverse associations were strongest for β-carotene. WC showed slightly stronger associations with most micronutrients compared with BMI and WCHtR (except for lycopene, which was more strongly associated with BMI), but the differences were not large. In general, all three measures were comparable and had stronger associations than WHR. Associations with α-tocopherol and retinol were generally weaker than those for the other analytes; however, retinol was significantly and positively associated with WHR, and α-tocopherol showed a borderline inverse association with BMI (P = 0.053). Longitudinal associations In mixed-effects linear regression analyses using the repeated measures data, the magnitude of the longitudinal associations was generally reduced (Table 3). For example, the coefficient for the association of WC with β-carotene was − 0.064 compared with − 0.081 in the baseline analysis. Nevertheless, all carotenoids showed highly significant inverse associations with all anthropometric measures, with the single exception of WHR with lycopene. Associations of carotenoids with BMI, WC and WCHtR were stronger than those with WHR. β-Carotene showed the strongest inverse associations with BMI, WC and WCHtR, followed by lutein +zeaxanthin, α-carotene, and β-cryptoxanthin and lycopene. αTocopherol was significantly and positively associated only with WHR. γ-Tocopherol was significantly and positively associated with all anthropometric variables, but the strongest associations were with BMI, WC and WCHtR. These associations, and the positive association of α-tocopherol with WHR, were stronger than those for any other micronutrients. Retinol showed a weak but
statistically significant positive association with WC, WCHtR and WHR. When the longitudinal analysis was repeated in women who did not report taking any vitamin supplements (N = 1400), the magnitude of the coefficients varied between ± 0 and 17% compared with those in the main analysis; however, the direction, pattern and statistical significance of the associations were unchanged (data not shown). The inverse association of carotenoids (α-carotene, β-carotene and β-cryptoxanthin) with WCHtR was stronger (4twofold) in never and former smokers compared with current smokers, and those of α-carotene, β-carotene and lutein+zeaxanthin with WCHtR were stronger in women without the metabolic syndrome vs those with the metabolic syndrome (Table 4). The positive association of γ-tocopherol with WCHtR was stronger in women who had never used hormone therapy vs women who had ever used hormone therapy. No other interactions were statistically significant. DISCUSSION In the present study, using repeated measurements, we found strong associations between anthropometric measures of obesity and serum concentrations of antioxidant nutrients. All four anthropometric indices were inversely associated with all five carotenoids, and most strongly with β-carotene. BMI, WC and WCHtR were generally comparable in predicting carotenoid concentrations, whereas WHR was a weaker predictor. However, WHR was superior in predicting α-tocopherol and retinol. γ-Tocopherol was positively associated with BMI, WC, WCHtR and WHR. The associations of γ-tocopherol with BMI, WC and WCHtR and of α-tocopherol with WHR were the strongest of any micronutrients in terms of absolute magnitude. The inverse associations of carotenoids with obesity were stronger among never smokers than among smokers and among those without the metabolic syndrome compared with those with the syndrome.
Table 4.
Association of anthropometric measures of adiposity with serum antioxidants stratified by effect modifiers in a subset of the Women’s Health Initiativea Antioxidant
Anthropometric factor
Stratifying variable
α-Carotene
WCHtR
Metabolic syndrome No Yes Smoking Current Past Never Metabolic syndrome No Yes Smoking Current Past Never Smoking Current Past Never Metabolic syndrome No Yes Hormone therapy Never Ever
α-Carotene
β-Carotene β-Carotene
β-Crytoxanthin
Lutein+zeaxanthin γ-Tocopherol
WCHtR
WCHtR WCHtR
WCHtR
WCHtR WCHtR
β-Coefficent
95% CI
P for interaction
− 0.019 − 0.009
(− 0.022; − 0.016) (− 0.013; − 0.006)
0.001
− 0.008 − 0.017 − 0.020
(− 0.013; − 0.004) (− 0.022; − 0.013) (− 0.024; − 0.017)
0.0001
− 0.062 − 0.032
(− 0.077; − 0.047) (− 0.049; − 0.016)
0.01
− 0.032 − 0.056 − 0.074
(− 0.055; − 0.008) (− 0.074; − 0.038) (− 0.088; − 0.059)
0.01
− 0.004 − 0.013 − 0.017
(− 0.011; 0.003) (− 0.018; − 0.009) (− 0.022; − 0.012)
0.01
− 0.023 − 0.015
(− 0.028; − 0.018) (− 0.020; − 0.009)
0.02
(0.232; 0.355) (0.124; 0.230)
0.005
0.294 0.177
a
Adjusted for the following variables, except for the stratifying variable: age (continuous), servings of alcohol per week (continuous), smoking status (never, former, current smoker), metabolic equivalent h/week (continuous), educational level (less than high school degree, high school degree/some college, college degree, post college), ethnicity (white, black, other), hormone therapy (ever, never), oral contraceptive use (ever, never) and total cholesterol (continuous).
© 2015 Macmillan Publishers Limited
European Journal of Clinical Nutrition (2015) 1 – 7
Longitudinal association of adiposity measures with antioxidants GC Kabat et al
6 Previous studies that have examined the association of obesity with blood or tissue concentrations of antioxidant nutrients have been almost exclusively cross-sectional and, therefore, do not permit judgments about the longitudinal association between weight change and circulating levels of micronutrients. In addition, most of these studies used BMI as the sole measure of obesity,5,6,9,10,12,14,16,17 whereas only a few studies included measures of central obesity.7,11,13 Moreover, the studies have differed considerably in their sample size, the age range of the population and the covariates that were adjusted for in the analysis. Few studies included specific tocopherol fractions or retinol, and few studies examined potential effect modification of the association of obesity with antioxidant concentrations by smoking status or presence of the metabolic syndrome. With one exception,12 previous studies have reported an inverse association between obesity and concentrations of carotenoids in blood5,6,9,10,14–17 or adipose tissue.7 In contrast, an analysis of demographic and personal characteristics in relation to plasma carotenoid concentrations in 3043 participants in the European Prospective Investigation into Cancer and Nutrition (EPIC)12 found that BMI was positively associated with plasma carotenoid concentrations. In that study, in multivariable analyses, region of residence was the strongest predictor of carotenoid concentrations, followed by BMI and smoking. In the single study with repeated measurements of carotenoids and BMI, Andersen et al.15 examined the prospective association between BMI and the serum concentration of five carotenoids in young adults (aged 18–30 at baseline) in the CARDIA study. Serum carotenoids were measured at years 0 and 7 in 3071 black and white participants who were either persistent smokers or nonsmokers. Among nonsmokers, baseline (year 0) BMI predicted year 7 carotenoid levels: obese subjects (BMI ⩾ 30 kg/m2) had an average concentration of the sum of four carotenoids that was 22% lower than the concentration among subjects with a BMI of less than 22 kg/m2. The change from year 0 to 7 in serum carotenoids, except for lycopene, was inversely associated with the change in BMI among nonsmokers, but not among smokers. Findings regarding the micronutrients that show the strongest associations with obesity are somewhat inconsistent. Regarding carotenoids, in some studies, β-carotene showed the strongest inverse association.7,11,12,14 Other studies have reported that both α- and β-carotene were more strongly associated with obesity than the other carotenoids,3,6,9 whereas one study17 found that the pro-vitamin A carotenoids (α-carotene, β-carotene and β-cryptoxanthin) were more strongly associated with obesity than non-pro-vitamin A carotenoids. In our data, the association of obesity with β-carotene concentrations was considerably stronger than the associations with other serum carotenoids. In a number of studies, lycopene showed no association with obesity in women;5–7,9,13,17 in other studies,10,12,14,16 including ours, lycopene showed a significant inverse association with obesity. Our results regarding tocopherols are in agreement with those of Chai et al.,17 concerning the positive association of BMI with γ-tocopherol and the lack of any association of BMI with α-tocopherol (in their crosssectional study and in our longitudinal analysis). However, we did note a significant positive association of WHR with α-tocopherol concentrations. The few studies that examined the association of multiple obesity measures with micronutrients have reported inconsistent results. Suzuki et al.13 noted that BMI was significantly (inversely) associated only with cryptoxanthin, whereas WC showed significant inverse associations with canthaxanthin and β-carotene, and WHR with α- and β-carotene. In the study by Virtanen et al.,7 BMI, WC and WHR showed similar associations. Wallström et al.11 found that all obesity measures (BMI, WC, WHR, percent body fat) were inversely associated with serum β-carotene concentrations, whereas only measures of central obesity (WC and WHR) were European Journal of Clinical Nutrition (2015) 1 – 7
positively and significantly associated with α-tocopherol concentration. Several explanations have been proposed to account for the inverse association of blood carotenoids with obesity. First, carotenoids as well as tocopherols are stored in adipose tissue, and individuals with more adipose tissue may store proportionately more of these micronutrients in fat, thereby reducing the concentration in the blood.5,6,11,14 Second, obesity is an inflammatory state associated with increased oxidative stress that may increase the requirement for antioxidants, and, for this reason, carotenoid stores may be depleted in the obese.5,14,15 Our findings that serum carotenoid concentrations were significantly higher in never smokers compared with current smokers and in women without the metabolic syndrome vs those with the syndrome are consistent with the possibility that circulating carotenoids are depleted in those exposed to greater levels of oxidative stress. Our study cannot address these two possible explanations. A third possible explanation is that overweight and obese individuals may consume less of foods containing antioxidants.11,14 When BMI was dichotomized at the median, mean intake of β-carotene was somewhat higher in the lower BMI group, but the difference was small: 3180 vs 3020 μg/day (P for difference in the means = 0.06). The pattern of associations of anthropometric factors with tocopherols differed from that with carotenoids. In the longitudinal analysis, α-tocopherol showed a strong positive association with WHR but not with other anthropometric variables. In contrast, γ-tocopherol had strong positive associations with BMI, WC and WCHtR, and a weaker association with WHR. Our findings regarding tocopherols are consistent with results from a previous analysis of correlates of α- and γ-tocopherol in the WHI28 as well as with the results of other studies.29 In the earlier WHI analysis,28 higher BMI was associated with lower α- and higher γ-tocopherol concentrations. Interestingly, in our data, users of vitamin supplements had higher concentrations of α-tocopherol but lower concentrations of γ-tocopherol. α-Tocopherol binds preferentially to hepatic tocopherol-binding protein and is incorporated into low-density lipoprotein and high-density lipoprotein, while other tocopherols (primarily γ-) are excreted.31 This may explain the inverse correlation between serum α- and γ-tocopherol. The opposing associations of these two tocopherols with obesity and with indices of a healthy diet have been observed by others.32 Strengths of the current study include the availability of repeated measurements of anthropometric variables and serum micronutrients over time on a large number of women who were not in the intervention arms of the WHI clinical trials; the availability of information on a wide range of potential confounding variables, including dietary intake; and our analysis of potential effect modifiers of the association of anthropometric measures of obesity with micronutrient concentrations. A limitation is the lack of markers of inflammation and oxidative damage, such as C-reactive protein, IL-6, TNFα and isoprostanes. Furthermore, as our study population was limited to postmenopausal women, the findings may not be applicable to younger women or to men. In conclusion, the strongest associations between anthropometric variables and micronutrients in the present study were an inverse association of WC with serum β-carotene and a positive association of waist circumference with γ-tocopherol. CONFLICT OF INTEREST The authors declare no conflict of interest.
ACKNOWLEDGEMENTS Short list of WHI investigators: Program Office: (National Heart, Lung, and Blood Institute, Bethesda, Maryland) Jacques Rossouw, Shari Ludlam, Dale Burwen, Joan McGowan, Leslie Ford, and Nancy Geller. Clinical Coordinating Center: Clinical
© 2015 Macmillan Publishers Limited
Longitudinal association of adiposity measures with antioxidants GC Kabat et al
7 Coordinating Center: (Fred Hutchinson Cancer Research Center, Seattle, WA) Garnet Anderson, Ross Prentice, Andrea LaCroix and Charles Kooperberg. Investigators and Academic Centers: (Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA) JoAnn E. Manson; (MedStar Health Research Institute/Howard University, Washington, D.C., USA) Barbara V. Howard; (Stanford Prevention Research Center, Stanford, CA, USA) Marcia L Stefanick; (The Ohio State University, Columbus, OH, USA) Rebecca Jackson; (University of Arizona, Tucson/Phoenix, AZ, USA) Cynthia A Thomson; (University at Buffalo, Buffalo, NY, USA) Jean Wactawski-Wende; (University of Florida, Gainesville/Jacksonville, FL, USA) Marian Limacher; (University of Iowa, Iowa City/Davenport, IA, USA) Robert Wallace; (University of Pittsburgh, Pittsburgh, PA, USA) Lewis Kuller; (Wake Forest University School of Medicine, Winston-Salem, NC, USA) Sally Shumaker. Women’s Health Initiative Memory Study: (Wake Forest University School of Medicine, Winston-Salem, NC, USA) Sally Shumaker. For a list of all the investigators who have contributed to WHI science, please visit: http://www. whiscience.org/publications/WHI_investigators_longlist.pdf.
REFERENCES 1 Flint AJ, Hu FB, Glynn RJ, Gaspard H, Manson JE, Willett WC et al. Excess weight and the risk of incident coronary heart disease among men and women. Obesity 2010; 18: 377–383. 2 Chan JM, Rimm EB, Colditz GA, Stampfer MJ, Willett WC. Obesity, fat distribution, and weight gain as risk factors for clinical diabetes in men. Diabetes Care 1994; 17: 961–969. 3 Bianchini F, Kaaks R, Vainio H. Overweight, obesity, and cancer risk. Lancet Oncol 2002; 3: 565–574. 4 Furukawa S, Fujita T, Shimabukuro M, Iwaki M, Yamada Y, Nakajima Y et al. Increased oxidative stress in obesity and its impact on metabolic syndrome. J Clin Invest 2004; 114: 1752–1761. 5 Suzuki K, Ito Y, Ochiai J, Kusuhara Y, Hashimoto S, Tokudome S et al. Relationshiip between obesity and serum markers of oxidative stress and inflammation in Japanese. Asian Pacific J Cancer Prev 2003; 4: 259–266. 6 Brady WE, Mares-Perlman JA, Bowen P, Stacewicz-Sapuntzakis M. Human serum carotenoid concentrations are related to physiologic and lifestyle factors. J Nutr 1996; 126: 129–137. 7 Virtanen SM, van’t Veer P, Kok F, Kardinal AFM, Aro A. Predictors of adipose tissue carotenoid and retinol levels in nine countries: the Euramic Study. Am J Epidemiol 1996; 144: 968–979. 8 Yeum KJ, Booth SL, Roubenoff R, Russell RM. Plasma carotenoid concentrations are inversely correlated with fat mass in older women. J Nutr Health Aging 1998; 2: 79–83. 9 Rock CL, Thornquist MD, Kristal AR, Patterson RE, Cooper DA, Neuhouser ML et al. Demographic, dietary and lifestyle factors differentially explain variability in serum carotenoids and fat-soluble vitamins: baseline results from the sentinel site of the Olestra post-marketing surveillance study. J Nutr 1999; 129: 855–864. 10 Casso D, White E, Patterson RE, Agurs-Collins T, Kooperberg C, Haines PS. Correlates of serum lycopene in older women. Nutr Cancer 2000; 36: 163–169. 11 Wallström P, Wirfalt E, Lahmann PH, Gullberg B, Janzon L, Berglund G. Serum concentrations of beta-carotene and alpha-tocopherol are associated with diet, smoking, and general and central adiposity. Am J Clin Nutr 2001; 73: 777–785. 12 Al-Delaimy WK, van Kappel AL, Ferrari P, Slimani N, Steghens JP, Bingham S et al. Plasma levels of six carotenoid in nine European countries: report from the European Prospective Investigation into Cancer and Nutrition (EPIC). Public Health Nutr 2004; 7: 713–722. 13 Suzuki K, Inoue T, Hioki R, Ochiai J, Kusuhara Y, Ichino N et al. Association of abdominal obesity with decreased serum levels of carotenoids in healthy Japanese population. Clin Nutr 2006; 25: 780–789.
14 Kimmons JE, Blanck HM, Tohill BC, Zhang J, Khan LK. Associations between body mass index and the prevalence of low micronutrient levels among US adults. MedGenMed 2006; 8: 59. 15 Andersen LF, Jacobs Jr DR, Gross MD, Shreiner PJ, Dale Williams O, Lee DH. Longitudinal associations between body mass index and serum carotenoids: the CARDIA study. Br J Nutr 2006; 95: 358–365. 16 Vioque J, Weinbrenner T, Asensio L, Castello A, Young IS, Fletcher A. Plasma concentrations of carotenoids and vitamin C are better correlated with dietary intake in normal weight than overweight and obese elderly subjects. Br J Nutr 2007; 97: 977–986. 17 Chai W, Conroy SM, Maskarinec G, Franke AA, Pagano IS, Cooney RV. Associations between obesity and serum lipid-soluble micronutrients among premenopausal women. Nutr Res 2010; 30: 227–232. 18 Pamuk ER, Byers T, Coates RJ, Vann JW, Sowell AL, Gunter EW et al. Effect of smoking on serum nutrient concentrations in African-American women. Am J Clin Nutr 1994; 59: 891–895. 19 Tsubono Y, Tsugane S, Gey KF. Differential effects of cigarette smoking and alcohol consumption on the plasma levels of carotenoids in middle-aged Japanese. Jpn J Cancer Res 1996; 87: 563–569. 20 The Women’s Health Initiative Study Group. Design of the Women’s Health Initiative clinical trial and observational study. Control Clin Trials 1998; 19: 61–109. 21 Langer RD, White E, Lewis CE, Kotchen JM, Hendrix SL, Trevisan M. The Women’s Health Initiative Observational Study: baseline characteristics of participants and reliability of baseline measures. Ann Epidemiol 2003; 13(9 Suppl): S107–S121. 22 Bosy-Westphal A, Geilser C, Onur S, Korth O, Selberg O, Schrezenmeir J et al. Value of body fat mass vs anthropometric obesity indices in the assessment of metabolic risk factors. Int J Obesity 2006; 30: 475–483. 23 Petursson H, Sigurdsson JA, Bengtsson C, Nilsen TI, Getz L. Body configuration as a predictor of mortality: comparison of five anthropometric measures in a 12 year follow-up of the Norwegian HUNT 2 study. PLoS ONE 2011; 6: c26621. 24 Krakauer NY, Krakauer JC. A new body shape index predicts mortality hazard independently of body mass index. PLoS ONE 2012; 7: e39504. 25 McTiernan A, Kooperberg C, White E, Wilcox S, Coates R, Adams-Campbell LL et al. Recreational physical activity and the risk of breast cancer in postmenopausal women: the Women’s Health Initiative Cohort Study. JAMA 2003; 290: 1331–1336. 26 Kaplan LA, Miller JA, Stein EA, Stampfer MJ. Simultaneous, high-performance liquid chromatographic analysis of retinol, tocopherols, lycopene, and alpha- and beta-carotene in serum and plasma. Methods Enzymol 1990; 189: 155–167. 27 Miller KW, Lorr NA, Yang CS. Simultaneous determination of plasma retinol, alphatocopherol, lycopene, alpha-carotene, and beta-carotene by high-performance liquid chromatography. Anal Biochem 1984; 138: 340–345. 28 National Heart, Lung, and Blood Institute. Detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III), final report. NIH Publication No. 02 – 5215. 29 Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel RH, Franklin BA et al. Diagnosis and management of the metabolic syndrome: an American Heart Associatioon/National Heart, Lung, and Blood Institute Scientific Statement. Circulation 2005; 112: 2735–2752. 30 Kabat GC, Heo M, Van Horn LL, Kazlauskaite R, Getaneh A, Ard J et al. Longitudinal association of anthropometric measures of adiposity with cardiometabolic risk factors in postmenopausal women. Ann Epidemiol 2014; 24: 896–902. 31 White E, Kristal AR, Shikany JM, Wilson AC, Chen C, Mares-Perlman JA et al. Correlates of serum α- and γ-tocopherol in the Women’s Health Initiative. Ann Epidemiol 2001; 11: 136–144. 32 Bates CJ, Mishra GD, Prentice A. γ-Tocopherol as a possible marker for nutritionrelated risk: results from four Natinal Diet and Nutrition Surveys in Britain. Br J Nutr 2004; 92: 137–150.
Supplementary Information accompanies this paper on European Journal of Clinical Nutrition website (http://www.nature.com/ejcn)
© 2015 Macmillan Publishers Limited
European Journal of Clinical Nutrition (2015) 1 – 7
ORIGINAL ARTICLE
Longitudinal association of measures of adiposity with serum antioxidant concentrations in postmenopausal women GC Kabat1, M Heo1, HM Ochs-Balcom2, MS LeBoff3, Y Mossavar-Rahmani1, LL Adams-Campbell4, R Nassir5, J Ard6, O Zaslavsky7 and TE Rohan1 BACKGROUND/OBJECTIVES: The relationship between obesity and circulating levels of antioxidants is poorly understood. Most studies that have examined the association of adiposity with blood or tissue concentrations of antioxidant micronutrients have been cross-sectional, and few have compared the associations for indices of overall obesity and central obesity. Our aim was to prospectively examine the longitudinal association of body mass index (BMI), waist circumference (WC), waist circumference-height ratio (WCHtR) and waist-hip ratio (WHR) with major serum antioxidants in a population of postmenopausal women. SUBJECTS/METHODS: We used a subsample of participants in the Women’s Health Initiative aged 50–79 years at entry with available fasting blood samples and anthropometric measurements obtained at multiple time points over 12.8 years of follow-up (N = 2672). Blood samples were used to measure α-carotene, β-carotene, β-cryptoxanthin, lutein+zeaxanthin, α-tocopherol, γtocopherol and retinol at baseline, and at years 1, 3 and 6. We used mixed-effects linear regression analyses to examine associations between anthropometric measures and serum antioxidants at baseline and over time, controlling for covariates. RESULTS: In longitudinal analyses, carotenoids, and particularly β-carotene, were strongly and inversely associated with BMI, WC and WCHtR and less so with WHR. α-Tocopherol showed a strong positive association with WHR but not with other anthropometric measures, whereas γ-tocopherol was positively and strongly associated with BMI, WC, WCHtR and less so with WHR. Retinol was positively associated with WHR. The inverse association of several carotenoids with anthropometric measures was stronger in never and former smokers compared with current smokers and in women without the metabolic syndrome. The inverse association of carotenoids with obesity measures may reflect reduced micronutrient concentrations owing to inflammation associated with obesity. CONCLUSIONS: In the present study, the strongest observed associations between anthropometric variables and micronutrients were an inverse association of WC with serum β-carotene and a positive association of WC with γ-tocopherol. European Journal of Clinical Nutrition advance online publication, 27 May 2015; doi:10.1038/ejcn.2015.74
INTRODUCTION Obesity and central obesity are established risk factors for cardiovascular disease, diabetes and certain cancers.1–3 Obesity is associated with chronic systemic inflammation and increased oxidative stress,4,5 which are hypothesized to play critical roles in disease progression. Antioxidant micronutrients may prevent damage from inflammation and reactive oxygen species, thereby reducing the deleterious effects of obesity.5 On the other hand, obesity may place a greater demand on requirements for antioxidants4 and may cause increased utilization of antioxidants, thus leading to reduced concentrations.5 Therefore, it is important to understand the interplay between obesity and circulating antioxidant nutrients. A number of studies have examined the association between obesity and blood or tissue levels of antioxidant micronutrients, most often carotenoids,5–17 and have mostly reported inverse associations of obesity with circulating levels of carotenoids. A smaller number of studies have examined the association of obesity with other micronutrients, including tocopherols9,11,14,17 and retinol.7,9 All of these studies but one15 were cross-sectional, making it difficult to interpret the temporal relationship between
obesity and antioxidant concentrations. In addition, most studies assessed overall obesity (that is, BMI),5,6,9,10,12,14–17 whereas fewer studies examined abdominal obesity (that is, waist circumference (WC) or waist-hip ratio (WHR)).7,11,13 Only one study15 stratified the analysis by smoking status (smokers vs nonsmokers), an important consideration, given that smoking is believed to result in increased levels of reactive oxygen species and to deplete antioxidant stores, and that carotenoid concentrations have been found consistently to be lower among smokers compared with nonsmokers.18,19 That study found significant inverse associations of BMI with carotenoid levels only among nonsmokers. In the present analysis, we used repeated measurements of anthropometric indices of general and central obesity and of serum micronutrients in a sub-cohort of the Women’s Health Initiative to assess the longitudinal association of body weight with serum concentrations of antioxidants. PARTICIPANTS AND METHODS The current study was conducted using data from the Women’s Health Initiative (WHI), a large, multi-center prospective study designed to
1 Department of Epidemiology and Population Health, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY, USA; 2Department of Epidemiology and Environmental Health, School of Public Health and Health Profession, University at Buffalo, Buffalo, NY, USA; 3Brigham and Women’s Hospital, Boston, MA, USA; 4Lombardi Comprehensive Cancer Center, Georgetown University, Washington D.C., USA; 5Department of Biochemistry and Molecular Medicine, University of California, Davis, CA, USA; 6 Department of Epidemiology and Prevention, Wake Forest School of Medicine, Medical Center Blvd., Winston-Salem, NC, USA and 7Department of Nursing, University of Haifa, Haifa, Israel. Correspondence: Dr GC Kabat, Department of Epidemiology and Population Health, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461, USA. E-mail: [email protected] Received 30 December 2014; revised 20 March 2015; accepted 30 March 2015
Longitudinal association of adiposity measures with antioxidants GC Kabat et al
2 advance our understanding of the determinants of major chronic diseases in older women.20 The WHI is composed of a Clinical Trial component (CT, N = 68 132) and an Observational Study (N = 93 676). Women between the ages of 50 and 79 and representing major racial/ethnic groups were recruited from the general population at 40 clinical centers throughout the US between 1993 and 1998. Details of the design and reliability of the baseline measures have been published.20,21 The study population for the present analysis consisted of a 6% random sample of women in the CT (N = 4544) who provided fasting blood samples at baseline and years 1, 3 and 6 during follow-up and a 1% sample of women in the Observational Study (N = 1062) who provided a fasting blood sample at baseline and at year 3. Of the 5606 women with measured analytes (see below), we restricted our analysis to the (approximately) 2680 women with baseline serum measurements of the analytes who were not in an intervention arm of any clinical trial. Antioxidant nutrients studied here included α-carotene, β-carotene, β-cryptoxanthin, lutein+zeaxanthin, lycopene, α-tocopherol, γ-tocopherol and retinol.
Assessment of diabetes and the metabolic syndrome
Data collection and variable definition
Longitudinal analyses. In the longitudinal analysis, the association between change in anthropometric factors and change in antioxidant nutrients was assessed using repeated measures data and time-varying covariates pertinent to each antioxidant nutrient. Because all micronutrients were associated with total cholesterol concentration, we included total cholesterol as a covariate in the multivariable analyses. To compare the magnitudes of regression coefficients across different anthropometric measures, we converted the repeatedly measured anthropometric measures to their corresponding Z-scores, which center the measures on their means and then divide by their standard deviations. Because use of vitamin supplements can alter serum concentrations of carotenoids and tocopherols, we carried out a sensitivity analysis repeating the longitudinal analysis excluding women who reported taking vitamin supplements (N = 1314). To examine variation in the longitudinal association by level of potential effect modifiers, we stratified the analysis based on the following categorizations: age (o 63 vs ⩾ 63 years), hormone therapy (any type: current vs non-current use, smoking status (never, former, current) and metabolic syndrome (yes, no). The chi-square Q statistic was used to test for heterogeneity of estimated coefficients for the association of the Zscored anthropometric measures with antioxidant nutrients between stratum levels of each potential effect modifier. To limit the number of interactions tested, we selected the anthropometric factor that was most strongly associated with a given micronutrient in the multivariable longitudinal analyses. Thus, we tested the association of WCHtR with αcarotene, β-carotene, β-cryptoxanthin, lutein+zeaxanthin and γ-tocopherol; BMI with lycopene; and WHR with α-tocopherol and retinol. In total, we conducted 32 different tests for inequality of coefficients between strata (8 associations × 4 stratifying factors) (see Supplementary Table).
At study entry, self-administered questionnaires were used to collect information on demographics, medical, reproductive and family history, and on dietary and lifestyle factors, including smoking history, alcohol consumption and recreational physical activity. Participants completed a semiquantitative, 122-item food-frequency questionnaire to measure dietary intake over the past 3 months. Additional questions queried sources of fat intake, added fats and intake of fruits and vegetables. Data were extracted from the baseline questionnaire on estimated dietary intake of α-carotene, β-carotene, β-cryptoxanthin, lycopene, α-tocopherol, γ-tocopherol and retinol. Intake of lutein+zeaxanthin was not computed in the WHI nutrient database. Dietary supplement data were collected during in-person clinic visits. Women were asked to bring supplement bottles to the baseline clinic visit and to subsequent annual visits in the CT, and to the baseline and year 3 visit in the Observational Study. A standardized interviewer-administered four-page questionnaire was used to obtain information on multivitamins, with or without minerals and single supplements. Only supplements that were used at least once a week were recorded, but there was no limit to the number of supplements recorded. All participants had their height, weight, and waist and hip circumferences measured by trained staff at baseline. Weight was measured to the nearest 0.1 kg and height to the nearest 0.1 cm. WC at the natural waist or narrowest part of the torso, and hip circumference at the maximal circumference, were recorded to the nearest 0.1 cm. Anthropometric measurements were also made during follow-up (in years 3 and 6 for the majority of the cohort and, additionally, in years 1 and 9 for a minority of women). Body mass index (BMI) was computed as weight in kilograms divided by the square of height in meters. In addition to WC and WHR, we created the variable waist circumference-height ratio (WCHtR), which in some studies has been shown to be superior to BMI or WC in predicting metabolic risk22 and mortality.23,24 Questions about physical activity at baseline referred to a woman’s usual pattern of activity, including walking and recreational physical activity. A variable ‘current total leisure-time physical activity’ (metabolic equivalent-hours/week) was computed by multiplying the number of hours per week of specific leisuretime physical activities by the metabolic equivalent value of the activities and summing over all types of activities.25
Micronutrient assays Blood samples were collected after an overnight fast (12 h) with minimal stasis and maintained at 4 °C until plasma or serum was separated. Plasma or serum aliquots were then frozen at − 70 °C and sent on dry ice to the WHI central repository (Fisher BioServices, Rockville, MD, USA) for storage at − 70 °C. Retinol, α-carotene, β-carotene, β-cryptoxanthin, lycopene, lutein+zeaxanthin, α-tocopherol and γ-tocopherol were measured in serum by reverse-phase high-performance liquid chromatography.26,27 After the addition of an internal standard, serum was extracted into hexane and injected onto a C18 reverse-phase column. The analytes were measured at wavelengths of 292 and 452 nm. Coefficients of variation were determined in pooled blood samples from four age-eligible female volunteers. The coefficients of variation for the eight analytes ranged from 6.0% (α-tocopherol) to 20.4% (α-carotene). European Journal of Clinical Nutrition (2015) 1 – 7
A history of diabetes was based on self-report of taking diabetes medication or having a fasting glucose of 126 mg/dl or greater at baseline. We used the definition of the metabolic syndrome proposed by the Adult Treatment Panel III (ATP III) of the National Cholesterol Education Program.28,29
Statistical analysis Cross-sectional analyses. The association of a given anthropometric factor with each antioxidant nutrient at baseline was estimated first using Pearson correlation coefficients followed by partial correlation coefficients adjusting for multiple covariates including age, smoking status, alcohol intake, physical activity, educational level, ethnicity and total cholesterol. The association between anthropometric variables and serum micronutrients was assessed cross-sectionally at baseline using multiple regression models, and longitudinally using mixed effects linear regression models including multiple covariates.
RESULTS Baseline characteristics of the study population have been previously described.30 Mean age at enrollment was 63.1 (±7.2) years, and mean BMI was 28.5 (±6.0). The proportions of women with a history of diabetes, cardiovascular disease, hypertension and cancer were 7.8%, 9.8%, 37.9% and 6.0%, respectively. Thirtythree percent of women met the definition of the metabolic syndrome. Cross-sectional associations Significant positive correlations were seen among serum levels of all carotenoids (Table 1), ranging from r = 0.14 between βcryptoxanthin and lycopene to r = 0.54 between α-carotene and β-carotene. Serum α-tocopherol had weak positive associations with the carotenoids, whereas serum γ-tocopherol showed significant inverse correlations with all carotenoids except lycopene. Serum retinol was positively associated with lutein +zeaxanthin, lycopene and α-tocopherol but not with other micronutrients. All eight micronutrients were positively associated with total cholesterol: r ranged from 0.04 for α-carotene to 0.33 for α-tocopherol. Positive associations were seen between serum concentrations of certain micronutrients and dietary intake of © 2015 Macmillan Publishers Limited
Longitudinal association of adiposity measures with antioxidants GC Kabat et al
those nutrients: α-carotene, β-carotene, β-cryptoxanthin, lycopene and γ-tocopherol, but not α-tocopherol or retinol. Partial correlations between baseline concentrations of serum micronutrients and anthropometric measures of obesity are shown in Table 2. After adjustment for covariates, most micronutrients were significantly and inversely associated with anthropometric measures, with the exceptions of γ-tocopherol, which showed a positive association with all anthropometric indices, α-tocopherol which was positively associated with WHR and retinol which was positively associated with WHR. The strongest inverse correlations were seen for α-carotene with BMI, WC and WCHtR (r = − 0.24, − 0.25 and − 0.24, respectively), whereas the strongest positive correlations were seen between γtocopherol and those same indices (0.22, 0.25 and 0.25, respectively). The weakest correlations with anthropometric factors were seen with α-tocopherol and retinol. For all micronutrients except α-tocopherol and retinol, correlations with WHR were weakest. Mean levels of serum micronutrients differed by level of numerous personal characteristics (Supplementary Table). Carotenoid levels were generally higher in women who were below the median for anthropometric indices of obesity compared with Table 1.
3 those in women who were above the median. The largest differences were seen for α-carotene, β-carotene and β-cryptoxanthin (differences of 36%, 36% and 25%, respectively, by level of BMI). In addition, carotenoid levels were higher in those without diabetes or the metabolic syndrome compared with those with these conditions; in those with higher physical activity level; and in never smokers compared with current smokers. Serum levels of αcarotene, β-carotene and β-crytoxanthin were 40%, 35% and 42% higher, respectively, in never smokers compared with current smokers. γ-Tocopherol levels were significantly higher in obese women, women with diabetes or the metabolic syndrome and women who had a lower level of physical activity. Mean concentration of antioxidants differed between women who used vitamin supplements and those who did not. The largest differences were seen for β-carotene and α-tocopherol, concentrations of which were 29% and 26% higher, respectively, in supplement users. However, γ-tocopherol was 41% higher in nonsupplement users. In multiple linear regression analyses of the cross-sectional association of micronutrients with anthropometric measures at baseline, γ-tocopherol was the micronutrient that was most sensitive to anthropometric measures, particularly with BMI, WC
Pearson correlations between serum micronutrients and other factors at baseline (N = 2680) α-Carotene
Serum micronutrients α-Carotene (μg/ml) β-Carotene (μg/ml) β-Cryptoxanthin (μg/ml) Lutein/zeaxanthin (μg/ml) Lycopene (μg/ml) α-Tocopherol (μg/ml) γ-Tocopherol (μg/ml) Retinol (μg/ml) Total cholesterol (mg/dl) Insulin (μIU/ml) Dietary intake α-Carotene (μg/day) β-Carotene (μg/day) β-Cryptoxanthin (μg/day) Lycopene (μg/day) α-Tocopherol (μg/day) γ-Tocopherol (μg/day) Retinol (μg/day) Vitamin A (RAE) Vitamin A (IU) Vitamin A (μg)
β-Carotene 0.54*
0.04** − 0.13*
0.05** − 0.11*
β-Cryptoxanthin
Lutein/zeaxanthin
0.30* 0.36*
0.35* 0.30* 0.36*
0.10* − 0.09*
0.21* − 0.09*
Lycopene
α-Tocopherol
γ-Tocopherol
Retinol
0.20* 0.16* 0.14* 0.21*
0.08* 0.21* 0.15* 0.15* 0.11*
− 0.23* − 0.29* − 0.17* − 0.08* 0.01*** − 0.38*
0.02*** 0.03*** 0.02*** 0.08* 0.09* 0.34* − 0.04***
0.25* − 0.07**
0.33* − 0.03***
0.24* 0.16*
0.26* 0.004***
0.28* 0.19* 0.32* 0.20* 0.02*** 0.20* 0.006*** 0.02*** 0.02*** 0.02***
*Po0.0001; **Po0.05; ***P ⩾ 0.05.
Table 2.
Partial correlationsa between serum micronutrients and body fat measures at baseline (N = 2680)
Anthropometric measures/Antioxidant nutrients α-Carotene (μg/ml) β-Carotene (μg/ml) β-Cryptoxanthin (μg/ml) Lutein/zeaxanthin (μg/ml) Lycopene (μg/ml) α-Tocopherol (μg/ml) γ-Tocopherol (μg/ml) Retinol (μg/ml)
Body mass index (kg/m2)
Waist circumference (cm)
Waist circumference/height
Waist-hip ratio
− 0.24* − 0.22* − 0.17* − 0.22* − 0.10* − 0.04** 0.22* − 0.003***
− 0.25* − 0.23* − 0.19* − 0.22* − 0.10* − 0.01*** 0.25* 0.03****
− 0.24* − 0.22* − 0.17* − 0.21* − 0.10* − 0.009*** 0.25* 0.03***
− 0.15* − 0.14* − 0.12* − 0.10* − 0.03*** 0.05** 0.19* 0.10*
*Po0.0001; **Po0.05; ***P ⩾ 0.05. aAdjusted for age (continuous), servings of alcohol intake per week (continuous), smoking status (current, former, never smoker), physical activity (metabolic equivalent-hours/week—continuous), education (less than high school degree, high school degree/some college, college degree, post college), ethnicity (white, black, other).
© 2015 Macmillan Publishers Limited
European Journal of Clinical Nutrition (2015) 1 – 7
European Journal of Clinical Nutrition (2015) 1 – 7
− 0.018* − 0.060* − 0.015* − 0.022* − 0.018* − 0.133**** 0.223* − 0.0001****
Longitudinal α-Carotene (μg/ml) β-Carotene (μg/ml) β-Cryptoxanthin (μg/ml) Lutein+zeaxanthin (μg/ml) Lycopene (μg/ml) α-Tocopherol (μg/ml) γ-Tocopherol (μg/ml) Retinol (μg/ml) (− 0.020; − 0.015) (− 0.072; − 0.049) (− 0.018; − 0.011) (− 0.026; − 0.019) (− 0.024; − 0.011) (− 0.389; 0.122) (0.182; 0.265) (− 0.005; 0.005)
(− 0.025; − 0.018) (− 0.092; − 0.063) (− 0.020; − 0.012) (− 0.029; − 0.020) (− 0.028; − 0.011) (− 0.622; 0.004) (0.221; 0.331) (− 0.007; 0.005)
95% CI
− 0.018* − 0.064* − 0.016* − 0.023* − 0.016* − 0.001**** 0.239* 0.006**
− 0.022* − 0.081* − 0.017* − 0.025* − 0.020* − 0.179**** 0.299* 0.004****
Β-coefficient
(− 0.021; − 0.016) (− 0.075; − 0.053) (− 0.020; − 0.013) (− 0.026; − 0.019) (− 0.022; − 0.009) (− 0.254; 0.252) (0.198; 0.280) (0.0009; 0.010)
(− 0.026; − 0.019) (− 0.095; − 0.067) (− 0.021; − 0.014) (− 0.029; − 0.020) (− 0.028; − 0.012) (− 0.491; 0.133) (0.244; 0.353) (− 0.002; 0.010)
95% CI
Waist circumference (cm)
− 0.018* − 0.060* − 0.014* − 0.022* − 0.016* 0.086**** 0.234* 0.006***
− 0.022* − 0.077* − 0.016* − 0.024* − 0.022* − 0.158**** 0.305* 0.003****
Β-coefficient
(− 0.020; − 0.015) (− 0.071; − 0.049) (− 0.018; − 0.011) (− 0.026; − 0.019) (− 0.023; − 0.010) (− 0.166; 0.337) (0.193; 0.275) (0.001 − 0.011)
(− 0.025; − 0.018) (− 0.091; − 0.063) (− 0.020; 0.012) (− 0.029; − 0.020) (− 0.030; − 0.013) (− 0.472; 0.155) (0.250; 0.359) (− 0.003; 0.009)
95% CI
Waist circumference/height
− 0.007* − 0.024* − 0.006* − 0.009* − 0.003**** 0.306* 0.076* 0.010*
− 0.013* − 0.045* − 0.010* − 0.012* − 0.009** 0.253**** 0.188* 0.015*
Β-coefficient
(− 0.009; − 0.005) (− 0.033; − 0.015) (− 0.009; − 0.003) (− 0.012; − 0.006) (− 0.009; 0.002) (0.097; 0.515) (0.042; 0.109) (0.006; 0.013)
(− 0.017; − 0.010) (− 0.059; − 0.031) (− 0.014; − 0.007) (− 0.016; − 0.008) (− 0.017; − 0.001) (− 0.046; 0.552) (0.135; 0.240) (0.009; 0.021)
95% CI
Waist-hip ratio
Abbreviation: CI, confidence interval. *Po 0.0001; **P = 0.02–0.03; ***P = 0.01; ****P40.05. aAdjusted for age (continuous), servings of alcohol per week (continuous), smoking status (never, former, current smoker), metabolic equivalent hours/week (continuous), educational level (less than high school degree, high school degree/some college, college degree, post college), ethnicity (white, black, other), hormone therapy (ever, never), oral contraceptive use (ever, never) and total cholesterol (continuous).
− 0.022* − 0.077* − 0.016* − 0.025* − 0.020* − 0.309*** 0.276* − 0.001****
Baseline α-Carotene (μg/ml) β-Carotene (μg/ml) β-Cryptoxanthin (μg/ml) Lutein+zeaxanthin (μg/ml) Lycopene (μg/ml) α-Tocopherol (μg/ml) γ-Tocopherol (μg/ml) Retinol (μg/ml)
Β-coefficient
Body mass index (kg/m2)
Table 3. Association of serum micronutrients with a 1-unit increase in the z-score for anthropometric measures of adiposity after adjusting for covariatesa in mixed linear regression models in the Women’s Health Initiative (N = 2680)
Longitudinal association of adiposity measures with antioxidants GC Kabat et al
4
© 2015 Macmillan Publishers Limited
Longitudinal association of adiposity measures with antioxidants GC Kabat et al
5 and WCHtR (Table 3). The strongest associations (based on comparing β-coefficients) were seen for γ-tocopherol and β-carotene. For example, for each 1 s.d. increase in WC, serum γ-tocopherol increased by 0.299 mg/dl. Carotenoids were significantly inversely associated with anthropometric measures, and the inverse associations were strongest for β-carotene. WC showed slightly stronger associations with most micronutrients compared with BMI and WCHtR (except for lycopene, which was more strongly associated with BMI), but the differences were not large. In general, all three measures were comparable and had stronger associations than WHR. Associations with α-tocopherol and retinol were generally weaker than those for the other analytes; however, retinol was significantly and positively associated with WHR, and α-tocopherol showed a borderline inverse association with BMI (P = 0.053). Longitudinal associations In mixed-effects linear regression analyses using the repeated measures data, the magnitude of the longitudinal associations was generally reduced (Table 3). For example, the coefficient for the association of WC with β-carotene was − 0.064 compared with − 0.081 in the baseline analysis. Nevertheless, all carotenoids showed highly significant inverse associations with all anthropometric measures, with the single exception of WHR with lycopene. Associations of carotenoids with BMI, WC and WCHtR were stronger than those with WHR. β-Carotene showed the strongest inverse associations with BMI, WC and WCHtR, followed by lutein +zeaxanthin, α-carotene, and β-cryptoxanthin and lycopene. αTocopherol was significantly and positively associated only with WHR. γ-Tocopherol was significantly and positively associated with all anthropometric variables, but the strongest associations were with BMI, WC and WCHtR. These associations, and the positive association of α-tocopherol with WHR, were stronger than those for any other micronutrients. Retinol showed a weak but
statistically significant positive association with WC, WCHtR and WHR. When the longitudinal analysis was repeated in women who did not report taking any vitamin supplements (N = 1400), the magnitude of the coefficients varied between ± 0 and 17% compared with those in the main analysis; however, the direction, pattern and statistical significance of the associations were unchanged (data not shown). The inverse association of carotenoids (α-carotene, β-carotene and β-cryptoxanthin) with WCHtR was stronger (4twofold) in never and former smokers compared with current smokers, and those of α-carotene, β-carotene and lutein+zeaxanthin with WCHtR were stronger in women without the metabolic syndrome vs those with the metabolic syndrome (Table 4). The positive association of γ-tocopherol with WCHtR was stronger in women who had never used hormone therapy vs women who had ever used hormone therapy. No other interactions were statistically significant. DISCUSSION In the present study, using repeated measurements, we found strong associations between anthropometric measures of obesity and serum concentrations of antioxidant nutrients. All four anthropometric indices were inversely associated with all five carotenoids, and most strongly with β-carotene. BMI, WC and WCHtR were generally comparable in predicting carotenoid concentrations, whereas WHR was a weaker predictor. However, WHR was superior in predicting α-tocopherol and retinol. γ-Tocopherol was positively associated with BMI, WC, WCHtR and WHR. The associations of γ-tocopherol with BMI, WC and WCHtR and of α-tocopherol with WHR were the strongest of any micronutrients in terms of absolute magnitude. The inverse associations of carotenoids with obesity were stronger among never smokers than among smokers and among those without the metabolic syndrome compared with those with the syndrome.
Table 4.
Association of anthropometric measures of adiposity with serum antioxidants stratified by effect modifiers in a subset of the Women’s Health Initiativea Antioxidant
Anthropometric factor
Stratifying variable
α-Carotene
WCHtR
Metabolic syndrome No Yes Smoking Current Past Never Metabolic syndrome No Yes Smoking Current Past Never Smoking Current Past Never Metabolic syndrome No Yes Hormone therapy Never Ever
α-Carotene
β-Carotene β-Carotene
β-Crytoxanthin
Lutein+zeaxanthin γ-Tocopherol
WCHtR
WCHtR WCHtR
WCHtR
WCHtR WCHtR
β-Coefficent
95% CI
P for interaction
− 0.019 − 0.009
(− 0.022; − 0.016) (− 0.013; − 0.006)
0.001
− 0.008 − 0.017 − 0.020
(− 0.013; − 0.004) (− 0.022; − 0.013) (− 0.024; − 0.017)
0.0001
− 0.062 − 0.032
(− 0.077; − 0.047) (− 0.049; − 0.016)
0.01
− 0.032 − 0.056 − 0.074
(− 0.055; − 0.008) (− 0.074; − 0.038) (− 0.088; − 0.059)
0.01
− 0.004 − 0.013 − 0.017
(− 0.011; 0.003) (− 0.018; − 0.009) (− 0.022; − 0.012)
0.01
− 0.023 − 0.015
(− 0.028; − 0.018) (− 0.020; − 0.009)
0.02
(0.232; 0.355) (0.124; 0.230)
0.005
0.294 0.177
a
Adjusted for the following variables, except for the stratifying variable: age (continuous), servings of alcohol per week (continuous), smoking status (never, former, current smoker), metabolic equivalent h/week (continuous), educational level (less than high school degree, high school degree/some college, college degree, post college), ethnicity (white, black, other), hormone therapy (ever, never), oral contraceptive use (ever, never) and total cholesterol (continuous).
© 2015 Macmillan Publishers Limited
European Journal of Clinical Nutrition (2015) 1 – 7
Longitudinal association of adiposity measures with antioxidants GC Kabat et al
6 Previous studies that have examined the association of obesity with blood or tissue concentrations of antioxidant nutrients have been almost exclusively cross-sectional and, therefore, do not permit judgments about the longitudinal association between weight change and circulating levels of micronutrients. In addition, most of these studies used BMI as the sole measure of obesity,5,6,9,10,12,14,16,17 whereas only a few studies included measures of central obesity.7,11,13 Moreover, the studies have differed considerably in their sample size, the age range of the population and the covariates that were adjusted for in the analysis. Few studies included specific tocopherol fractions or retinol, and few studies examined potential effect modification of the association of obesity with antioxidant concentrations by smoking status or presence of the metabolic syndrome. With one exception,12 previous studies have reported an inverse association between obesity and concentrations of carotenoids in blood5,6,9,10,14–17 or adipose tissue.7 In contrast, an analysis of demographic and personal characteristics in relation to plasma carotenoid concentrations in 3043 participants in the European Prospective Investigation into Cancer and Nutrition (EPIC)12 found that BMI was positively associated with plasma carotenoid concentrations. In that study, in multivariable analyses, region of residence was the strongest predictor of carotenoid concentrations, followed by BMI and smoking. In the single study with repeated measurements of carotenoids and BMI, Andersen et al.15 examined the prospective association between BMI and the serum concentration of five carotenoids in young adults (aged 18–30 at baseline) in the CARDIA study. Serum carotenoids were measured at years 0 and 7 in 3071 black and white participants who were either persistent smokers or nonsmokers. Among nonsmokers, baseline (year 0) BMI predicted year 7 carotenoid levels: obese subjects (BMI ⩾ 30 kg/m2) had an average concentration of the sum of four carotenoids that was 22% lower than the concentration among subjects with a BMI of less than 22 kg/m2. The change from year 0 to 7 in serum carotenoids, except for lycopene, was inversely associated with the change in BMI among nonsmokers, but not among smokers. Findings regarding the micronutrients that show the strongest associations with obesity are somewhat inconsistent. Regarding carotenoids, in some studies, β-carotene showed the strongest inverse association.7,11,12,14 Other studies have reported that both α- and β-carotene were more strongly associated with obesity than the other carotenoids,3,6,9 whereas one study17 found that the pro-vitamin A carotenoids (α-carotene, β-carotene and β-cryptoxanthin) were more strongly associated with obesity than non-pro-vitamin A carotenoids. In our data, the association of obesity with β-carotene concentrations was considerably stronger than the associations with other serum carotenoids. In a number of studies, lycopene showed no association with obesity in women;5–7,9,13,17 in other studies,10,12,14,16 including ours, lycopene showed a significant inverse association with obesity. Our results regarding tocopherols are in agreement with those of Chai et al.,17 concerning the positive association of BMI with γ-tocopherol and the lack of any association of BMI with α-tocopherol (in their crosssectional study and in our longitudinal analysis). However, we did note a significant positive association of WHR with α-tocopherol concentrations. The few studies that examined the association of multiple obesity measures with micronutrients have reported inconsistent results. Suzuki et al.13 noted that BMI was significantly (inversely) associated only with cryptoxanthin, whereas WC showed significant inverse associations with canthaxanthin and β-carotene, and WHR with α- and β-carotene. In the study by Virtanen et al.,7 BMI, WC and WHR showed similar associations. Wallström et al.11 found that all obesity measures (BMI, WC, WHR, percent body fat) were inversely associated with serum β-carotene concentrations, whereas only measures of central obesity (WC and WHR) were European Journal of Clinical Nutrition (2015) 1 – 7
positively and significantly associated with α-tocopherol concentration. Several explanations have been proposed to account for the inverse association of blood carotenoids with obesity. First, carotenoids as well as tocopherols are stored in adipose tissue, and individuals with more adipose tissue may store proportionately more of these micronutrients in fat, thereby reducing the concentration in the blood.5,6,11,14 Second, obesity is an inflammatory state associated with increased oxidative stress that may increase the requirement for antioxidants, and, for this reason, carotenoid stores may be depleted in the obese.5,14,15 Our findings that serum carotenoid concentrations were significantly higher in never smokers compared with current smokers and in women without the metabolic syndrome vs those with the syndrome are consistent with the possibility that circulating carotenoids are depleted in those exposed to greater levels of oxidative stress. Our study cannot address these two possible explanations. A third possible explanation is that overweight and obese individuals may consume less of foods containing antioxidants.11,14 When BMI was dichotomized at the median, mean intake of β-carotene was somewhat higher in the lower BMI group, but the difference was small: 3180 vs 3020 μg/day (P for difference in the means = 0.06). The pattern of associations of anthropometric factors with tocopherols differed from that with carotenoids. In the longitudinal analysis, α-tocopherol showed a strong positive association with WHR but not with other anthropometric variables. In contrast, γ-tocopherol had strong positive associations with BMI, WC and WCHtR, and a weaker association with WHR. Our findings regarding tocopherols are consistent with results from a previous analysis of correlates of α- and γ-tocopherol in the WHI28 as well as with the results of other studies.29 In the earlier WHI analysis,28 higher BMI was associated with lower α- and higher γ-tocopherol concentrations. Interestingly, in our data, users of vitamin supplements had higher concentrations of α-tocopherol but lower concentrations of γ-tocopherol. α-Tocopherol binds preferentially to hepatic tocopherol-binding protein and is incorporated into low-density lipoprotein and high-density lipoprotein, while other tocopherols (primarily γ-) are excreted.31 This may explain the inverse correlation between serum α- and γ-tocopherol. The opposing associations of these two tocopherols with obesity and with indices of a healthy diet have been observed by others.32 Strengths of the current study include the availability of repeated measurements of anthropometric variables and serum micronutrients over time on a large number of women who were not in the intervention arms of the WHI clinical trials; the availability of information on a wide range of potential confounding variables, including dietary intake; and our analysis of potential effect modifiers of the association of anthropometric measures of obesity with micronutrient concentrations. A limitation is the lack of markers of inflammation and oxidative damage, such as C-reactive protein, IL-6, TNFα and isoprostanes. Furthermore, as our study population was limited to postmenopausal women, the findings may not be applicable to younger women or to men. In conclusion, the strongest associations between anthropometric variables and micronutrients in the present study were an inverse association of WC with serum β-carotene and a positive association of waist circumference with γ-tocopherol. CONFLICT OF INTEREST The authors declare no conflict of interest.
ACKNOWLEDGEMENTS Short list of WHI investigators: Program Office: (National Heart, Lung, and Blood Institute, Bethesda, Maryland) Jacques Rossouw, Shari Ludlam, Dale Burwen, Joan McGowan, Leslie Ford, and Nancy Geller. Clinical Coordinating Center: Clinical
© 2015 Macmillan Publishers Limited
Longitudinal association of adiposity measures with antioxidants GC Kabat et al
7 Coordinating Center: (Fred Hutchinson Cancer Research Center, Seattle, WA) Garnet Anderson, Ross Prentice, Andrea LaCroix and Charles Kooperberg. Investigators and Academic Centers: (Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA) JoAnn E. Manson; (MedStar Health Research Institute/Howard University, Washington, D.C., USA) Barbara V. Howard; (Stanford Prevention Research Center, Stanford, CA, USA) Marcia L Stefanick; (The Ohio State University, Columbus, OH, USA) Rebecca Jackson; (University of Arizona, Tucson/Phoenix, AZ, USA) Cynthia A Thomson; (University at Buffalo, Buffalo, NY, USA) Jean Wactawski-Wende; (University of Florida, Gainesville/Jacksonville, FL, USA) Marian Limacher; (University of Iowa, Iowa City/Davenport, IA, USA) Robert Wallace; (University of Pittsburgh, Pittsburgh, PA, USA) Lewis Kuller; (Wake Forest University School of Medicine, Winston-Salem, NC, USA) Sally Shumaker. Women’s Health Initiative Memory Study: (Wake Forest University School of Medicine, Winston-Salem, NC, USA) Sally Shumaker. For a list of all the investigators who have contributed to WHI science, please visit: http://www. whiscience.org/publications/WHI_investigators_longlist.pdf.
REFERENCES 1 Flint AJ, Hu FB, Glynn RJ, Gaspard H, Manson JE, Willett WC et al. Excess weight and the risk of incident coronary heart disease among men and women. Obesity 2010; 18: 377–383. 2 Chan JM, Rimm EB, Colditz GA, Stampfer MJ, Willett WC. Obesity, fat distribution, and weight gain as risk factors for clinical diabetes in men. Diabetes Care 1994; 17: 961–969. 3 Bianchini F, Kaaks R, Vainio H. Overweight, obesity, and cancer risk. Lancet Oncol 2002; 3: 565–574. 4 Furukawa S, Fujita T, Shimabukuro M, Iwaki M, Yamada Y, Nakajima Y et al. Increased oxidative stress in obesity and its impact on metabolic syndrome. J Clin Invest 2004; 114: 1752–1761. 5 Suzuki K, Ito Y, Ochiai J, Kusuhara Y, Hashimoto S, Tokudome S et al. Relationshiip between obesity and serum markers of oxidative stress and inflammation in Japanese. Asian Pacific J Cancer Prev 2003; 4: 259–266. 6 Brady WE, Mares-Perlman JA, Bowen P, Stacewicz-Sapuntzakis M. Human serum carotenoid concentrations are related to physiologic and lifestyle factors. J Nutr 1996; 126: 129–137. 7 Virtanen SM, van’t Veer P, Kok F, Kardinal AFM, Aro A. Predictors of adipose tissue carotenoid and retinol levels in nine countries: the Euramic Study. Am J Epidemiol 1996; 144: 968–979. 8 Yeum KJ, Booth SL, Roubenoff R, Russell RM. Plasma carotenoid concentrations are inversely correlated with fat mass in older women. J Nutr Health Aging 1998; 2: 79–83. 9 Rock CL, Thornquist MD, Kristal AR, Patterson RE, Cooper DA, Neuhouser ML et al. Demographic, dietary and lifestyle factors differentially explain variability in serum carotenoids and fat-soluble vitamins: baseline results from the sentinel site of the Olestra post-marketing surveillance study. J Nutr 1999; 129: 855–864. 10 Casso D, White E, Patterson RE, Agurs-Collins T, Kooperberg C, Haines PS. Correlates of serum lycopene in older women. Nutr Cancer 2000; 36: 163–169. 11 Wallström P, Wirfalt E, Lahmann PH, Gullberg B, Janzon L, Berglund G. Serum concentrations of beta-carotene and alpha-tocopherol are associated with diet, smoking, and general and central adiposity. Am J Clin Nutr 2001; 73: 777–785. 12 Al-Delaimy WK, van Kappel AL, Ferrari P, Slimani N, Steghens JP, Bingham S et al. Plasma levels of six carotenoid in nine European countries: report from the European Prospective Investigation into Cancer and Nutrition (EPIC). Public Health Nutr 2004; 7: 713–722. 13 Suzuki K, Inoue T, Hioki R, Ochiai J, Kusuhara Y, Ichino N et al. Association of abdominal obesity with decreased serum levels of carotenoids in healthy Japanese population. Clin Nutr 2006; 25: 780–789.
14 Kimmons JE, Blanck HM, Tohill BC, Zhang J, Khan LK. Associations between body mass index and the prevalence of low micronutrient levels among US adults. MedGenMed 2006; 8: 59. 15 Andersen LF, Jacobs Jr DR, Gross MD, Shreiner PJ, Dale Williams O, Lee DH. Longitudinal associations between body mass index and serum carotenoids: the CARDIA study. Br J Nutr 2006; 95: 358–365. 16 Vioque J, Weinbrenner T, Asensio L, Castello A, Young IS, Fletcher A. Plasma concentrations of carotenoids and vitamin C are better correlated with dietary intake in normal weight than overweight and obese elderly subjects. Br J Nutr 2007; 97: 977–986. 17 Chai W, Conroy SM, Maskarinec G, Franke AA, Pagano IS, Cooney RV. Associations between obesity and serum lipid-soluble micronutrients among premenopausal women. Nutr Res 2010; 30: 227–232. 18 Pamuk ER, Byers T, Coates RJ, Vann JW, Sowell AL, Gunter EW et al. Effect of smoking on serum nutrient concentrations in African-American women. Am J Clin Nutr 1994; 59: 891–895. 19 Tsubono Y, Tsugane S, Gey KF. Differential effects of cigarette smoking and alcohol consumption on the plasma levels of carotenoids in middle-aged Japanese. Jpn J Cancer Res 1996; 87: 563–569. 20 The Women’s Health Initiative Study Group. Design of the Women’s Health Initiative clinical trial and observational study. Control Clin Trials 1998; 19: 61–109. 21 Langer RD, White E, Lewis CE, Kotchen JM, Hendrix SL, Trevisan M. The Women’s Health Initiative Observational Study: baseline characteristics of participants and reliability of baseline measures. Ann Epidemiol 2003; 13(9 Suppl): S107–S121. 22 Bosy-Westphal A, Geilser C, Onur S, Korth O, Selberg O, Schrezenmeir J et al. Value of body fat mass vs anthropometric obesity indices in the assessment of metabolic risk factors. Int J Obesity 2006; 30: 475–483. 23 Petursson H, Sigurdsson JA, Bengtsson C, Nilsen TI, Getz L. Body configuration as a predictor of mortality: comparison of five anthropometric measures in a 12 year follow-up of the Norwegian HUNT 2 study. PLoS ONE 2011; 6: c26621. 24 Krakauer NY, Krakauer JC. A new body shape index predicts mortality hazard independently of body mass index. PLoS ONE 2012; 7: e39504. 25 McTiernan A, Kooperberg C, White E, Wilcox S, Coates R, Adams-Campbell LL et al. Recreational physical activity and the risk of breast cancer in postmenopausal women: the Women’s Health Initiative Cohort Study. JAMA 2003; 290: 1331–1336. 26 Kaplan LA, Miller JA, Stein EA, Stampfer MJ. Simultaneous, high-performance liquid chromatographic analysis of retinol, tocopherols, lycopene, and alpha- and beta-carotene in serum and plasma. Methods Enzymol 1990; 189: 155–167. 27 Miller KW, Lorr NA, Yang CS. Simultaneous determination of plasma retinol, alphatocopherol, lycopene, alpha-carotene, and beta-carotene by high-performance liquid chromatography. Anal Biochem 1984; 138: 340–345. 28 National Heart, Lung, and Blood Institute. Detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III), final report. NIH Publication No. 02 – 5215. 29 Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel RH, Franklin BA et al. Diagnosis and management of the metabolic syndrome: an American Heart Associatioon/National Heart, Lung, and Blood Institute Scientific Statement. Circulation 2005; 112: 2735–2752. 30 Kabat GC, Heo M, Van Horn LL, Kazlauskaite R, Getaneh A, Ard J et al. Longitudinal association of anthropometric measures of adiposity with cardiometabolic risk factors in postmenopausal women. Ann Epidemiol 2014; 24: 896–902. 31 White E, Kristal AR, Shikany JM, Wilson AC, Chen C, Mares-Perlman JA et al. Correlates of serum α- and γ-tocopherol in the Women’s Health Initiative. Ann Epidemiol 2001; 11: 136–144. 32 Bates CJ, Mishra GD, Prentice A. γ-Tocopherol as a possible marker for nutritionrelated risk: results from four Natinal Diet and Nutrition Surveys in Britain. Br J Nutr 2004; 92: 137–150.
Supplementary Information accompanies this paper on European Journal of Clinical Nutrition website (http://www.nature.com/ejcn)
© 2015 Macmillan Publishers Limited
European Journal of Clinical Nutrition (2015) 1 – 7
