sign in sign up
<<
Home , Multivitamin

International Journal of Obesity (2010) 34, 1070–1077 & 2010 Macmillan Publishers Limited All rights reserved 0307-0565/10 $32.00 www.nature.com/ijo ORIGINAL ARTICLE Effects of multivitamin and supplementation on adiposity, energy expenditure and lipid profiles in obese Chinese women

YLi1,4, C Wang2,4, K Zhu3, RN Feng1 and CH Sun1

1Department of Nutrition and Hygiene, School of Public Health, Harbin Medical University, Harbin, China; 2Endemic Disease Control Center, Chinese Center for Disease Control and Prevention, Harbin Medical University, Harbin, China and 3School of Medicine and Pharmacology, University of Western Australia, Perth, Australia

Background: Obese individuals are more likely to have either lower blood concentrations or lower bioavailability of minerals and/or . However, there are limited data on the effects of nutritional supplementation on body weight (BW) control, energy homeostasis and lipid metabolism in obese subjects. Objective: The purpose of this study is to evaluate the effects of supplementation with multivitamin and multimineral on adiposity, energy expenditure and lipid profiles in obese Chinese women. Design: A total of 96 obese Chinese women (body mass index (BMI) 28 kg mÀ2) aged 18–55 years participated in a 26-week randomized, double-blind, placebo-controlled intervention study. Subjects were randomized into three groups, receiving either one tablet of multivitamin and mineral supplement (MMS), or 162 mg (Calcium) or identical placebo daily during the study period. BW, BMI, waist circumference (WC), fat mass (FM), fat-free mass, resting energy expenditure (REE), respiratory quotient (RQ), blood pressure, fasting plasma glucose and serum insulin, total cholesterol (TC), low- and high-density lipoprotein-cholesterol (LDL-C and HDL-C) and triglycerides (TGs) were measured at baseline and 26 weeks. Results: A total of 87 subjects completed the study. After 26 weeks, compared with the placebo group, the MMS group had significantly lower BW, BMI, FM, TC and LDL-C, significantly higher REE and HDL-C, as well as a borderline significant trend of lower RQ (P ¼ 0.053) and WC (P ¼ 0.071). The calcium group also had significantly higher HDL-C and lower LDL-C levels compared with the placebo group. Conclusion: The results suggest that, in obese individuals, multivitamin and mineral supplementation could reduce BW and fatness and improve serum lipid profiles, possibly through increased energy expenditure and fat oxidation. Supplementation of calcium alone (162 mg per day) only improved lipid profiles. International Journal of Obesity (2010) 34, 1070–1077; doi:10.1038/ijo.2010.14; published online 9 February 2010

Keywords: multivitamin and mineral supplement; adiposity; energy expenditure; lipid profiles; Chinese women

Introduction million are obese.2 The rapid economic development in China has been accompanied by the adoption of Western The past couple of decades have witnessed a worldwide style dietary and physical activity behaviors.3 The 2002 epidemic of obesity which poses a major threat to human Chinese National Nutrition Survey showed that 22.8% of health, especially in Western societies.1 The WHO (World adult Chinese were overweight and 7.1% obese (body mass Health Organization) estimated that worldwide there are index (BMI) X24 and X28 kg mÀ2, respectively) according to more than 1 billion overweight adults, and among them 300 the Chinese standard.4 Furthermore, obesity is strongly associated with hyperlipemia, one of the important risk factors of cardiovascular disease.4,5 Correspondence: Professor CH Sun, Department of Nutrition and Food Previous studies have shown that vitamins and minerals Hygiene, School of Public Health, Harbin Medical University, 157 Baojian have an important role in energy and blood lipid metabo- Road, Nangang District, Harbin, Heilongjiang 150086, China. lism6–9 through several mechanisms.10–13 Obese individuals E-mail: [email protected] 4 are more likely to have either lower blood concentrations or These authors contributed equally to this work. 14–17 Received 20 August 2009; revised 22 December 2009; accepted 29 December lower bioavailability of minerals and/or vitamins. How- 2009; published online 9 February 2010 ever, there are limited data on the effect of multivitamin and supplements, adiposity and lipids YLiet al 1071 mineral supplementation on body weight (BW), energy Table 1 Composition of multivitamin and mineral supplement expenditure and lipid profiles in obese subjects. A long-term Ingredient Amount/tablet observational cohort study with 15 655 American men and women showed that overweight or obese subjects (n ¼ 4708) A(containing natural mixed b-carotene) 5000 IU who took multivitamin, , or chro- 400 IU mium supplements gained less weight over the 8–12-year 30 IU follow-up period than did those who did not take these Thiamin 1.5 mg supplements.18 In our previous animal study on obese rats,19 1.7 mg supplementation of multivitamins and minerals led to BW Vitamin B6 2 mg 60 mg loss and improvement in energy metabolism. The aim of this Vitamin B12 6 mg study is to evaluate the effects of supplementation with Vitamin K1 25 mg multivitamin and mineral on BW, energy expenditure and 30 mg lipid profiles in obese Chinese women over a 26-week Folic acid 400 mg 20 mg intervention period. 10 mg Calcium 162 mg Phosphorus 125 mg Subjects and methods Chlorine 36.3 mg 100 mg 18 mg Study subjects Copper 2 mg Obese Chinese women (n ¼ 283), aged 18–55 years, whose BMI 15 mg was 428.0 kg mÀ2 recommended by the WGOC (Working 2.5 mg 20 150 mg Group on Obesity in China) as the cutoff point for obesity, Chromium 25 mg were recruited from Harbin, China in 2006. Among them, 96 25 mg women who met the inclusion criteria participated in this 25 mg 26-week randomized, double-blind, placebo-controlled inter- Nickel 5 mg Stannum 10 mg vention trial. The inclusion criteria were as follows: (1) total Silicon 10 mg À1 cholesterol (TC) X5.2 mmol l , or low-density lipoprotein Vanadium 10 mg cholesterol (LDL-C) X3.1 mmol lÀ1, or high-density lipo- protein cholesterol (HDL-C) o0.91 mmol lÀ1 or triglycerides (TG) X1.7 mmol lÀ1; (2) stable BW during the past 6 months Pharmaceutical Co. Ltd, Harbin, China) as shown in Table 1; (change in BMIo0.5 kg mÀ2); (3) not taking vitamin and subjects in the calcium group received one tablet daily mineral supplements in the past 6 months; (4) not taking containing 162 mg calcium (Dayu Biochemistry Co. Ltd, cholesterol or blood pressure-lowering medications; (5) no Shanghai, China); and subjects in the control group received history of , diabetes and not being one tablet of identical placebo made of maize starch daily. pregnant; and (6) o2 h per week of regular physical activity. The tablets used in the three groups were identical in Informed consent was obtained from all subjects, and the appearance. Compliance of the study subjects was assessed study protocol was approved by the Ethics Committee of the by conducting scheduled telephone interviews weekly, then Harbin Medical University. monthly after 4 weeks and by counting tablets returned at the last visit to the clinic. The study subjects and the study staff remained blinded to the allocation of the three groups Intervention during the entire study period. Eligible subjects were first sorted according to their BMI, and then randomized into three groups with a block size of three, with random numbers generated by SPSS (version 13.01S; Demographic characteristics Beijing Stats Data Mining Co. Ltd, Beijing, China). To ensure Demographic data were collected at baseline using a the homogeneity of outcome variables at baseline across the standardized questionnaire. Information collected included placebo and treatment groups, the homogeneity of BMI and age, ethnicity, education, occupation, smoking history, lipids at baseline among the three groups were tested. If the physical activity at work and leisure, health history, homogeneity of the outcome variables at baseline was not medications and status. The level of physical satisfied, then a new set of random number was generated activity during leisure time was defined as follows: and subjects were randomized again. This process was 0 ¼ none, 1 ¼ 1–30 min per week, 2 ¼ 31–60 min per week, repeated until the homogeneity of outcome variables among 3 ¼ 61–90 min per week, 4 ¼ 91–120 min per week and the three groups was reached. During the 26-week interven- 5 ¼ 4120 min per week. On the basis of the definition of tion period, subjects in the multivitamin and mineral physical activity in Chinese women,21 the level of physical supplement (MMS) group received one tablet daily contain- activity at work was coded as: 1 ¼ sedentary, 2 ¼ moderate ing 29 multivitamins and minerals (Centrum, and 3 ¼ heavy.

International Journal of Obesity Nutrient supplements, adiposity and lipids YLiet al 1072 Dietary intakes was measured using the Kyoto blood sugar test meter At baseline and 26 weeks, dietary intakes were determined (Arkray, Inc. Kyoto, Japan) and test strip. Serum TC, HDL-C using a semi-quantitative food-frequency questionnaire. The and TGs were assayed with standard enzymatic colorimetric food-frequency questionnaire was developed according to techniques using commercial kits (Biosino Biotechnology the method proposed by Willett22 and the dietary patterns of Ltd, Beijing, China) with an auto-analyzer (AUTOLAB PM the community. The food-frequency questionnaire contained 4000, AMS Corporation, Rome, Italy). LDL-C was calculated 103 items, including food intakes, alcohol drinks, as well as using the equation by Friedewald et al.23 Serum insulin was multivitamin, mineral and calcium supplements. The dietary measured using radioimmunoassay using commercial kit energy and intakes were estimated using the Food (Diagnostic Systems Laboratories, TX, USA). Nutrition Calculator (V1.60, Chinese CDC, Beijing, China) Power calculation and statistical analysis Power calculations were performed before the commence- Anthropometry ment of the study. A sample size of 28 in each group will be At baseline and 26 weeks, height, waist circumference (WC) sufficient to detect a difference of 1.8 kg mÀ2 in BMI between and BW were measured twice to ±0.1 cm and to ±0.1 kg, the treatment and the placebo groups assuming a s.d. of while fasting overnight and wearing only underwear. BMI 2.4 kg mÀ2 as reported in this population, at 80% power and (calculated as BW in kilograms divided by the square of 5% level of significance. This number has been increased to height in meters) was used as a measure of overall adiposity. 32 per group (total of 96) to allow for a predicted dropout of Fat mass (FM) was measured using the electric impedance B10%. Statistical analyses were carried out using SPSS. Data method with a body FM analyzer (TANITA TBF-300, Tanita were presented as mean±s.d. or percentage as appropriate. Corporation, Tokyo, Japan), and the fat-free mass was Paired samples t-test was used to evaluate the changes in calculated (as BWÀFM). outcome variables before and after intervention in each group. The w2 test was used to compare categorical variables. Mean levels of continuous study variables at baseline and Blood pressure follow-up among the three groups were compared using Blood pressure was measured using a standard mercury ANOVA (analysis of variance) and ANCOVA (analysis of sphygmomanometer on the right arm after at least 10 min of covariance), respectively. In the model of ANCOVA, the rest. Mean values were determined from two independent covariates included baseline values, age, alcohol consump- measurements (by the same researcher) at 2-min intervals. tion, smoking, total physical activity and menopause. The percentage differences between the treatment and the placebo groups in outcome variables at 26 weeks were Assessment of resting energy expenditure calculated using covariates-adjusted values as 100 Â (treat- Resting energy expenditure (REE) was measured using the ment groupÀplacebo group)/placebo group for each study pulmonary function and nutrition metabolism testing variable. All P-values are two-tailed, and a P-value o0.05 was system (Quark PFT ergo, Cosmed Corporation, Rome, Italy) considered significant for all statistical analyses in this study. early in the morning after 12 h of overnight fasting. Subjects were instructed to refrain from taking medications, heavy 283 obese meals, alcohol, coffee and other caffeine-containing women beverages, smoking and doing heavy physical exercise the screened night before. Indirect calorimetry was performed over a 30- 187 excluded min period using a metabolic measurement system, which Inclusion criteria not met (n = 151) consisted of a facemask, a computer with oxygen and carbon Refuse to participate (n = 36) dioxide sensors, a sampling pump, barometric sensors and 96 participants randomized electronics and a turbine connecting the facemask to the computer. The expired air sample was analyzed by the computer. Oxygen consumption and carbon dioxide produc- MMS Calcium Placebo tion were measured in terms of rate per breath at rest for a (n = 32) (n = 32) (n = 32) total of 30 min. Data obtained from the last 10 min of the measurement period were used for the calculation of 1 employment 1 employment 1 change of residence respiratory quotient (RQ) and REE. The instruments were commitment commitment 1 illness 3 lost to follow-up 2 lost to follow-up calibrated before each test. 30 complete 28 complete 29 complete investigations and investigations and investigations and measurements measurements measurements Biochemistry At baseline and 26 weeks, antecubital venous blood samples Figure 1 The study design and the flow of subjects. MMS, multivitamin and were collected after 12 h of overnight fasting. Plasma glucose mineral supplementation.

International Journal of Obesity Nutrient supplements, adiposity and lipids YLiet al 1073 Results significantly, and REE and HDL-C increased significantly in the MMS group compared with baseline. The calcium group Subject retention and compliance had significantly lower FM and significantly higher serum Figure 1 shows the study design and flow of subjects. Among HDL-C at 26 weeks compared with baseline. In the placebo the 96 eligible individuals who participated in the study at group, there were no significant changes in all outcome baseline, 87 subjects completed the study and 9 participants variables measured over the intervention period. withdrew because of employment commitments. There were Figure 2 shows the percentage differences between the no significant differences in the overall compliance rates treatment and placebo groups at 26 weeks adjusted for among the three groups (92.6% in the placebo group, 93.4% in the MMS group and 93.7% in the calcium group, P ¼ 0.85). Table 2 Characteristics of subjects at baseline

MMS Calcium Placebo P-value Characteristics of subjects in the three groups (n ¼ 30) (n ¼ 28) (n ¼ 29) At baseline, there were no significant differences between Age (years) 42.8±6.9 41.6±9.0 41.2±6.8 0.69 the three groups in age, BW, BMI, daily physical activity level Weight (kg) 80.5±8.0 77.5±8.0 80.8±9.7 0.29 at leisure time and work, as well as the percentage of women Height (cm) 161.6±5.5 159.3±4.5 161.0±5.2 0.22 BMI (kg mÀ2)30.8±2.5 30.5±2.5 31.1±2.7 0.67 who currently smoked, consumed alcohol or reached menopause (Table 2). Dietary nutrient intakes did not differ Physical activity at leisure time (%) significantly across the three groups at either baseline or None 80.0 78.6 79.3 0.71 26 weeks, and the changes in dietary nutrient intakes during 0–30 min per week 10.0 14.4 10.3 31–60 min per week 10.0 3.5 3.4 the intervention period were not significantly different 61–120 min per week 0 3.5 7.0 among the three groups (Table 3). Effects on BW, WC, fatness, fat-free mass, energy expen- Physical activity at work (%) diture, blood pressure, fasting plasma glucose, insulin and Sedentary 86.7 78.5 69.0 0.29 Moderate 13.3 17.9 31.0 lipid profiles Heavy 0 3.6 0 The mean levels (±s.d.) of outcome variables at baseline Current smokers (%) 10.0 3.6 6.9 0.63 and 26 weeks in the three groups are presented in Table 4. Alcohol drinkers (%) 46.7 21.4 27.6 0.10 There were no significant differences in baseline values Menopause (%) 20.0 28.6 24.1 0.75 of the outcome variables among the three groups. At Abbreviations: BMI, body mass index; MMS, multivitamin and mineral 26 weeks, BW, BMI, WC, FM, RQ, TC and LDL-C decreased supplementation. Values are mean±s.d. for continuous variables.

Table 3 Daily energy, vitamin and mineral consumption of baseline (week 0) and intervention period and changes with intervention

Week 0 Week 0–26 P-value

MMS Calcium Placebo MMS Calcium Placebo MMS Placebo Placebo

Mean s.d. Mean s.d. Mean s.d. P0 Mean s.d. Mean s.d. Mean s.d. P1 P2 P3 P4

Energy (kcal per day) 2765.8±950.2 2567.86±686.9 2841.3±797.4 0.44 2635.9±754.9 2583.6±911.8 2817.2±844.8 0.54 0.43 0.89 0.86 Protein (g per day) 83.5±28.4 78.6±29.4 88.9±28.5 0.40 81.6±26.9 79.8±36.1 86.3±29.0 0.72 0.69 0.83 0.61 Fat (g per day) 98.4±43.1 94.7±37.2 94.8±32.6 0.93 93.8±36.6 90.4±30.1 95.0±40.7 0.88 0.55 0.52 0.97 Carbohydrate (g per day) 408.3±151.6 369.9±104.7 428.7±151.5 0.27 387.9±137.5 382.1±157.9 408.4 ±155.2 0.24 0.48 0.58 0.43 (mg per day) 726.9±500.1 710.6±518.9 852.6±551.6 0.53 720.3±532.8 735.9±783.1 725.8±433.2 0.99 0.95 0.86 0.21 Thiamin (mg per day) 1.3±0.5 1.1±0.4 1.3±0.4 0.35 1.2±0.5 1.1±0.5 1.3±0.5 0.51 0.88 0.66 0.16 Riboflavin (mg per day) 1.2±0.5 1.2±0.6 1.3±0.5 0.83 1.1±0.5 1.2±0.7 1.2±0.6 0.86 0.68 0.96 0.75 Vitamin C (mg per day) 163.8±87.6 140.7±89.5 150.6±70.2 0.57 149.8±92.1 143.1±80.2 145.3±57.6 0.95 0.45 0.84 0.66 (mg per day) 18.9±7.1 17.2±8.5 20.5±7.8 0.47 17.1±6.2 17.3±9.9 18.4±6.7 0.80 0.17 0.94 0.14 Total vitamin E (mg per day) 63.9±21.2 61.4±21.0 54.4±15.2 0.16 59.8 ±17.6 59.4±16.2 56.4±19.9 0.73 0.19 0.57 0.60 Calcium (mg per day) 617.0±289.9 636.6±324.8 643.4±299.8 0.94 677.1±318.5 677.4±405.4 688.0±270.3 0.68 0.31 0.61 0.42 Magnesium (mg per day) 472.6±184.8 429.4±165.8 450.2±155.4 0.63 464.2±164.7 446.3±211.5 454.2±161.9 0.93 0.80 0.62 0.95 Iron (mg per day) 50.9±27.9 51.5±33.1 56.2±32.6 0.78 56.0±29.3 55.6±36.1 57.9±29.5 0.96 0.31 0.46 0.72 Phosphorus (mg per day) 1371.4±499.0 1302.9±486.5 1400.9±434.1 0.73 1363.7±460.7 1333.4±591.2 1384.9±462.8 0.93 0.93 0.75 0.85 Selenium (mg per day) 52.6±19.9 48.1±27.6 58.3±22.2 0.26 49.9±18.5 49.3±34.3 52.6±21.9 0.88 0.41 0.83 0.19 (mg per day) 2677.8±1025.1 2435.8±1096.1 2673.9±903.8 0.59 2570.5±1019.0 2427.4±1228.5 2552.0±880.1 0.86 0.60 0.96 0.45 Zinc (mg per day) 15.3±5.4 14.7±5.2 15.8±5.0 0.73 14.5±4.7 14.6±6.2 15.3±5.0 0.85 0.43 0.92 0.55 Copper (mg per day) 3.7±1.5 3.4±1.4 3.6±1.3 0.74 3.6±1.3 3.5±1.6 3.7±1.3 0.85 0.67 0.63 0.53 Manganese (mg per day) 10.1±3.9 10.0±4.4 10.9±4.7 0.71 10.6±4.2 10.4±5.2 11.0±4.1 0.88 0.52 0.57 0.87

Abbreviations: MMS, multivitamin and mineral supplementation; P0, differences in dietary energy and vitamin and mineral of three groups at baseline; P1, differences in dietary energy and vitamin and mineral of three groups in this period of intervention; P2–P4, differences in dietary energy and vitamin and mineral in every group between the baseline and intervention period.

International Journal of Obesity Nutrient supplements, adiposity and lipids YLiet al 1074 Table 4 Mean levels (±s.d.) of study variables at baseline and 26 weeks in baseline values, age, alcohol consumption, smoking history, the treatment and placebo groups physical activity and menopause status. At 26 weeks, BW, MMS (n ¼ 30) Calcium (n ¼ 28) Placebo (n ¼ 29) BMI, FM, systolic blood pressure, diastolic blood pressure, TC and LDL-C were significantly lower, and REE and HDL-C Body weight (kg) were significantly higher in the MMS group compared with Baseline 80.5±8.0 77.5±8.0 80.8±9.7 the placebo group. The MMS group also had a borderline 26 Weeks 76.9±7.4** 76.6±8.7 80.6±10.2 significant trend of lower RQ and WC at 26 weeks compared BMI (kg mÀ2) with the placebo group (RQ, P ¼ 0.053; WC, P ¼ 0.071). The ± ± ± Baseline 30.8 2.5 30.5 2.5 31.1 2.7 calcium group had significantly higher HDL-C and signifi- 26 Weeks 29.4±2.1** 30.2±2.8 31.0±2.9 cantly lower LDL-C compared with the placebo group at 26 WC (cm) weeks. Baseline 93.6±7.1 93.8±8.6 95.8±8.4 There were no significant changes in TG, fasting plasma 26 Weeks 91.2±6.8* 92.9±7.9 95.5±9.1 glucose, insulin and fat-free mass in all groups over the Fat mass (kg) intervention period. Baseline 31.0±5.1 29.3±5.2 30.6±6.2 26 Weeks 28.2±4.6** 28.5±5.8* 31.0±5.7

FFM (kg) Discussion Baseline 49.4±4.4 48.3±4.0 50.2±4.6 26 Weeks 48.7±4.8 48.1±4.1 49.6±6.2 Our study showed that the 29-ingredient multivitamin and REE (kcal per day) mineral supplementation over 26 weeks could reduce BW, Baseline 1385.9±137.6 1358.8±161.6 1393.7±154.3 WC, blood pressure and FM, decrease RQ and increase REE, 26 Weeks 1439.1±150.2** 1372.9±167.3 1365.2±140.0 and had a beneficial effect on lipid profiles, whereas calcium

RQ supplementation (162 mg per day) alone could improve lipid Baseline 0.81±0.05 0.82±0.06 0.81±0.06 profiles only. 26 Weeks 0.78±0.04** 0.81±0.04 0.81±0.05 Our previous studies on rats12,24 have shown that dietary calcium could reduce the levels of serum cholesterol, TGs, SBP (mm Hg) Baseline 126.5±18.4 129.0±15.3 129.5±26.7 BW and body fat. The purpose of having the calcium group 26 Weeks 121.8±13.7 126.3±17.9 130.3±22.8 in this study is to examine whether vitamins and minerals besides calcium also have important roles in reducing DBP (mm Hg) Baseline 85.2±15.2 86.4±11.7 84.7±13.3 BW and improving lipid profile in obese women. Our results 26 Weeks 80.0±8.7 82.2±10.3 85.7±10.7 showed that supplementation with multivitamin and mineral was more effective than calcium alone in improving FPG (mmol lÀ1) Baseline 4.9±0.9 4.8±1.0 5.0±0.9 these outcomes. It is well recognized that energy intake in 26 Weeks 5.0±0.8 5.0±0.9 5.2±0.8 obese individuals is higher than in those with normal weight, and vitamins and minerals influence the balance m À1 Insulin ( IU ml ) of energy metabolism.10–13 In addition, obese individuals Baseline 18.3±9.2 16.8±9.6 17.1±7.5 26 Weeks 17.5±8.6 18.4±8.9 19.1±7.7 have been shown to have low blood concentrations or low bioavailability of minerals and/or vitamins.14–17 Therefore, À1 TC (mmol l ) obese individuals may need greater amounts of vitamins and Baseline 4.9±0.8 5.0±1.0 5.1±0.8 26 Weeks 4.3±0.7** 5.0±0.8 5.2±0.8 minerals to cope with the increased burden of energy intake. The findings of this study support the notion that besides À1 TG (mmol l ) calcium, obese individuals need other vitamins and minerals Baseline 1.9±1.0 1.9±0.8 2.0±1.0 for balancing energy metabolism, controlling BW and for 26 Weeks 1.8±0.8 2.1±1.2 2.2±1.3 improving lipid profiles. HDL-C (mmol lÀ1) A study by Major et al.25 assessed the effects of supple- ± ± ± Baseline 1.0 0.4 1.0 0.3 0.9 0.2 mentation with multiple vitamins and minerals on BW in 26 Weeks 1.4±0.3** 1.3±0.2** 1.0±0.4 obese men and women during a 15-week energy restriction LDL-C (mmol lÀ1) period. After 15 weeks, BW reduced significantly in both the Baseline 3.3±0.9 3.6±1.0 3.9±0.8 intervention and the placebo groups. However, there were 26 Weeks 2.6±0.8** 3.3±0.8 3.8±1.0 no significant differences between the two groups in changes Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; FFM, in BW. In contrast, our study showed that multivitamin and fat-free mass; FPG, fasting plasma glucose; HDL-C, high-density lipoprotein mineral supplementation for 26 weeks could reduce BW and cholesterol; LDL-C, low-density lipoprotein cholesterol; MMS, multivitamin fatness in obese women. The differences in findings could be and mineral supplementation group; REE, resting energy expenditure; RQ, respiratory quotient; SBP, systolic blood pressure; TC, total cholesterol; TG, due to the fact that energy intakes were restricted in the triglyceride; WC, waist circumference. Values are mean±s.d. Different from study by Major et al. (prescribed daily energy intake was baseline values: *Po0.05; **Po0.01. determined by subtracting 700 kcal from the daily energy

International Journal of Obesity Nutrient supplements, adiposity and lipids YLiet al 1075 30 ** ** MMS - Placebo Calcium - Placebo 20

10 **

BWBMI WC FM FFM RQ SBPDBP FPG INS TC TG LDLC 0 REE HDLC ** ** * -10 ** ** * Percentage differences between ** -20 treatment and placebo groups at 26 weeks (%)

-30 **

Figure 2 Percentage of difference between treatment and placebo groups in obesity and energy metabolism measures, blood pressure, plasma glucose, insulin and lipid profiles at 26 weeks adjusting for baseline values, age, alcohol consumption, smoking history, physical activity and menopause status. MMS, multivitamin and mineral supplementation; BW, body weight; BMI, body mass index; WC, waist circumference; FM, fat mass; FFM, fat-free mass; REE, resting energy expenditure; RQ, respiratory quotient; SBP, systolic blood pressure; DBP, diastolic blood pressure; INS, insulin; FPG, fasting plasma glucose; TC, total cholesterol; TG, triglycerides; w HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol. Compared with the placebo group: *Po0.05, **Po0.01, P ¼ 0.071 and zP ¼ 0.053.

expenditure), whereas subjects in our study maintained their expenditure in the MMS group in this study could have usual dietary intakes. As reduced energy intake has a crucial resulted from improved energy homeostasis with increased role in controlling BW, it is possible that energy restriction vitamin and mineral intakes from the supplement. In our masked the effect of -induced weight reduc- study, the reduced RQ in the MMS group suggests that MMSs tion in the study by Major et al. Our findings are consistent could lead to a higher level of fat oxidation and to a greater with the results of a 10-year longitudinal study with 15 655 use of fat as an energy substrate. This observation along with individuals, that is the VITAL (VITamins And Lifestyle) the findings on increased REE could explain the significant cohort study, which showed that the long-term use of reduction in BW and fatness in the MMS group. multivitamins, vitamins B6 and B12 and chromium was Many in vitro studies and studies in the general population significantly associated with lower levels of weight gain have shown that vitamins and minerals have important roles among overweight or obese men and women (n ¼ 4708) over in improving lipid and lipoprotein metabolism.6–9 To our the 10-year follow-up period.18 knowledge, this study is the first to evaluate the effects of Obesity is a condition resulting from a chronic imbalance multivitamin and mineral supplementation on lipid profile between energy intake and energy expenditure. In addition, in obese subjects. Previous studies have shown that calcium,6 7 8 9 other factors that are involved in energy metabolism may chromium and biotin, niacin, , vitamins B6 and B12 also contribute to the development of obesity. These factors could improve lipid metabolism, and that vitamin B1,28 include genetic susceptibility (for example, leptin and POMC vitamin B6,29 vitamin C,30 calcium,31 magnesium,32 and genes, etc.), hormones (leptin, insulin, FT3, cholecystokinin, zinc33 also have important roles in improving hypertension. etc.), protein (UCP1-UCP3, NPY, lipoprotein lipase, etc.) and The beneficial effects on lipid profile and blood pressure mitochondrial function. Previous studies,10 including ours,11 observed in this study may be partly attributed to the actions have shown that a number of minerals and vitamins could of these in conjunction. In addition, weight promote the expressions of UCP1À3 mRNA and improve loss could reduce insulin hypersecretion and increase insulin mitochondrial function.13 The changes in these indexes can sensitivity and b-cell function,34 and decrease blood pres- upregulate thermogenesis, promote lipolysis and increase sure, and thus, improve hyperlipedemia35 and hypertension. energy consumption,26–27 which contributes to the improve- Therefore, weight loss in the MMS group during the study ment in energy homeostasis. Therefore, the increased energy period may also partly explain the beneficial effect of

International Journal of Obesity Nutrient supplements, adiposity and lipids YLiet al 1076 nutrition-based intervention on lipids, lipoproteins and 9 Lim HJ, Choi YM, Choue R. Dietary intervention with emphasis blood pressure. on folate intake reduces serum lipids but not plasma homo- In summary, this randomized, double-blind, placebo- cysteine levels in hyperlipidemic patients. Nutr Res 2008; 28: 767–774. controlled intervention trial showed that the 29-ingredient 10 Kumar MV, Sunvold GD, Scarpace PJ. Dietary vitamin A multivitamin and mineral supplementation has beneficial supplementation in rats: suppression of leptin and induction of effects on BW, blood pressure, energy metabolism and lipid UCP1 mRNA. J Lipid Res 1999; 40: 824–829. profiles in obese Chinese women. These findings have 11 Sun CH, Sun WG, Fu RX, Yu XF. The effect of iron on the expression of uncoupling protein gene in skeletal muscle of obese implications for the development of intervention strategies rats. Acta Nutr Sinica 2003; 25: 344–348. for the prevention of cardiovascular disease and other 12 Wang HY, Sun CH, Zhou XR, Song SL, Jiang LY. Mechanism of obesity-related disorders. dietary calcium on reducing body weight of obese rats induced by diets. Chin J Public Health 2004; 20: 1046–1047. 13 Ames BN, Atamna H, Killilea DW. Mineral and vitamin deficiencies can accelerate the mitochondrial decay of aging. Conflict of interest Mol Aspects Med 2005; 26: 363–378. 14 Kaidar-Person O, Person B, Szomstein S, Rosenthal RJ. Nutritional The authors declare no conflict of interest. deficiencies in morbidly obese patients: a new form of malnutri- tion? Part A: vitamins. Obes Surg 2008; 18: 870–876. 15 Kaidar-Person O, Person B, Szomstein S, Rosenthal RJ. Nutritional deficiencies in morbidly obese patients: a new form of malnutri- Acknowledgements tion? Part B: minerals. Obes Surg 2008; 18: 1028–1034. 16 Wortsman J, Matsuoka LY, Chen TC, Lu Z, Holick MF. Decreased bioavailability of vitamin D in obesity. Am J Clin Nutr 2000; 72: This study was supported by a grant from the Natural Science 690–693. Foundation of China (30771804). We thank the researchers 17 Aasheim ET, Hofso D, Hjelmesaeth J, Birkeland KI, Bohmer T. and medical personnel of Harbin Medical University for their Vitamin status in morbidly obese patients: a cross-sectional study. efforts and collaboration in this study. Professor Chang-Hao Am J Clin Nutr 2008; 87: 362–369. Sun was responsible for the conception, design and data 18 Nachtigal MC, Patterson RE, Stratton KL, Adams LA, Shattuck AL, White E. Dietary supplements and weight control in a middle-age interpretation of the study and is the principal investigator population. J Altern Complement Med 2005; 11: 909–915. of the grant for the study; Ying Li and Cheng Wang were 19 Jiang LY, Sun CH, Zhou XR, Wang HY, Ren LN. Effect of dietary responsible for data collection, data analysis, data interpreta- mineral nutrients and vitamins on metabolism of rat fed with tion and writing of this paper; Kun Zhu was responsible for high fat. Wei Sheng Yan Jiu 2004; 33: 447–449. 20 Cooperative Meta-Analysis Group of China Obesity Task Force. data interpretation and writing of this paper; Ren-Nan Feng Predictive values of body mass index and waist circumference to was responsible for performing clinical procedures, data risk factors of related diseases in Chinese adult population. Chin J collection and writing of this paper. Epidemiol 2002; 23:5–10. 21 Chinese Nutrition Society. Chinese Dietary Reference Intakes. China Light Industry Press: Peking, China, 2002. p 15. 22 Hao L, Li Z, Willett W. The method of food frequency question- References naire. The reproducibility and validity of FFQ. In: Willett W, Elizabeth L (eds) Nutritional Epidemiology (Translation). People’s 1 National Task Force on the Prevention and Treatment of Medical Publishing House: Peking, China, 2006. pp 69–142. Obesity. Overweight obesity health risk. Arch Intern Med 2000; 23 Friedewald WT, Levy RI, Fredrickson DS. Estimation of the 160: 898–904. concentration of low-density lipoprotein cholesterol in plasma 2 World Health Organization. Obesity and Overweight Available without use of preparative ultracentrifuge. Clin Chem 1972; 18: at: http://www.who.int/dietphysicalactivity/publications/facts/ 499–502. obesity/en/ (accessed 13 May 2009). 24 Sun CH, Yu XF, Li Y, Liu R, Wang HY. Effects of dietary calcium on 3 Cheng TO. The current state of cardiology in China. Int J Cardiol the blood glucose, blood lipid and hormone of rat fed a high fat 2004; 96: 425–439. diet. J Hygiene Res 2004; 33: 164–166. 4 Li LM, Rao KQ, Kong LZ, Yao CH, Xiang HD, Zai FY et al. 25 Major GC, Doucet E, Jacqmain M, St-Onge M, Bouchard C, A description on the Chinese national nutrition and health Tremblay A. Multivitamin and dietary supplements, body weight survey in 2002. Chin J Epidemiol 2005; 26: 478–484. and appetite: results from a cross-sectional and a randomized 5 Mokdad AH, Ford ES, Bowman BA, Dietz WH, Vinicor F, Bales VS double-blind placebo-controlled study. Br J Nutr 2008; 99: et al. Prevalence of obesity, diabetes, and obesity-related health 1157–1167. risk factors, 2001. JAMA 2003; 289: 76–79. 26 Costford SR, Chaudhry SN, Salkhordeh M, Harper ME. Effects of 6 Zemel MB. Nutritional and endocrine modulation of intracellular the presence, absence, and overexpression of uncoupling protein- calcium: implications in obesity, insulin resistance and hyper- 3 on adiposity and fuel metabolism in congenic mice. Am J tension. Mol Cell Biochem 1998; 188: 129–136. Physiol Endocrinol Metab 2006; 290: E1304–E1312. 7 Albarracin C, Fuqua B, Geohas J, Juturu V, Finch MR, Komorowski JR. 27 Schrauwen P, Hesselink M. UCP2 and UCP3 in muscle controlling Combination of chromium and biotin improves coronary risk body metabolism. J Exp Biol 2002; 205: 2275–2285. factors in hypercholesterolemic type 2 diabetes mellitus: a placebo- 28TanakaT,SohmiyaK,KonoT,TerasakiF,HorieR,OhkaruYet al. controlled, double-blind randomized clinical trial. J Cardiometab attenuates the hypertension and metabolic abnormalities Syndr 2007; 2: 91–97. in CD36-defective SHR: uncoupling of glucose oxidation from 8 Maccubbin D, Bays HE, Olsson AG, Elinoff V, Elis A, Mitchel Y et al. cellular entry accompanied with enhanced protein O-GlcNAcylation Lipid-modifying efficacy and tolerability of extended-release in CD36 deficiency. MolCellBiochem2007; 299:23–35. niacin/laropiprant in patients with primary hypercholesterolaemia 29 Dakshinamurti K, Paulose CS, Viswanathan M. Vitamin B6 and or mixed dyslipidaemia. Int J Clin Pract 2008; 62: 1959–1970. hypertension. Ann N Y Acad Sci 1990; 575: 241–249.

International Journal of Obesity Nutrient supplements, adiposity and lipids YLiet al 1077 30 Hornig B, Arakawa N, Kohler C, Drexler H. Vitamin C improves disease and factors associated with insulin resistance in rural endothelial function of conduit arteries in patients with chronic and urban populations of North India. J Am Coll Nutr 1998; 17: heart failure. Circulation 1998; 97: 363–368. 564–570. 31 Hatton DC, McCarron DA. Dietary calcium and blood pressure 34 Camastra S, Manco M, Mari A, Baldi S, Gastaldelli A, Greco AV in experimental models of hypertension. Hypertension 1994; 4: et al. Beta-cell function in morbidly obese subjects during 513–530. free living: long-term effects of weight loss. Diabetes 2005; 54: 32 Barbagallo M, Dominguez LJ, Galioto A, Ferlisi A, Cani C, Malfa L 2382–2389. et al. Role of magnesium in insulin action, diabetes and cardio- 35 Volek JS, Go´mez AL, Love DM, Weyers AM, Hesslink JR, Wise JA metabolic syndrome X. Mol Aspects Med 2003; 24: 39–52. et al. Effects of an 8-week weight-loss program on cardiovascular 33 Singh RB, Niaz MA, Rastogi SS, Bajaj S, Gaoli Z, Shoumin Z. disease risk factors and regional body composition. Eur J Clin Nutr Current zinc intake and risk of diabetes and coronary artery 2002; 56: 585–592.

International Journal of Obesity