PHYTOTHERAPY RESEARCH
`Phytother. Res. 28: 1770–1777 (2014)
`Published online 6 August 2014 in Wiley Online Library
`(wileyonlinelibrary.com) DOI: 10.1002/ptr.5197
`
`Lipid-Lowering Effects of Curcumin in Patients
`with Metabolic Syndrome: A Randomized,
`Double-Blind, Placebo-Controlled Trial
`
`Yi-Sun Yang,1,2 Ying-Fang Su,3,4 Hui-Wen Yang,3,4 Yu-Hsien Lee,3,4 Janet I. Chou5
`and Kwo-Chang Ueng6,7,8*
`1Division of Endocrinology and Metabolism, Department of Internal Medicine, Chung Shan Medical University Hospital, Taichung, Taiwan
`2Institute of Medicine, Chung Shan Medical University, Taichung, Taiwan
`3School of Dentistry, Chung Shan Medical University, Taichung, Taiwan
`4Department of Stomatology, Chung Shan Medical University Hospital, Taichung, Taiwan
`5Department of Internal Medicine, UCSF, Fresno, CA, USA
`6School of Medicine, Chung Shan Medical University, Taichung, Taiwan
`7Division of Cardiology, Department of Internal Medicine, Chung Shan Medical University Hospital, Taichung, Taiwan
`8Division of Cardiology, Department of Internal Medicine, Taichung University, Taichung, Taiwan
`
`Human studies of curcumin extract on lipid-lowering effect have not been completely investigated and have had
`controversy results. This study tested the effect of daily curcumin extract for 12 weeks on weight, glucose, and
`lipid profiles in patients with metabolic syndrome. Sixty-five patients were randomized into two groups; 33
`patients taking curcumin extract capsule (630 mg thrice daily) and 32 patients taking a placebo capsule thrice
`daily for 12 weeks. At 12 weeks after the curcumin extract consumption, the level of high-density lipoprotein
`cholesterol (HDL-C) significantly increased from 40.96 ± 8.59 to 43.76 ± 2.79 mg/dL (p < 0.05), and the level of
`low-density lipoprotein cholesterol (LDL) was significantly reduced (120.55 ± 36.81 to 106.51 ± 25.02 mg/dL,
`p < 0.05). The triglyceride-lowering effect, a reduction of 65 mg/dL, was also found in this study. In subgroups
`analysis, the consumption of curcumin may have a lowering cholesterol effect in male patients and an increasing
`HDL-C effect in female patients, both of which result in a decrease of T-Chol/HDL-C ratio. The intake of the
`curcumin extract of 1890 mg/day for 12 weeks was associated with lipid-lowering effect but did not improve
`weight and glucose homeostasis in the patients with metabolic syndrome. Daily curcumin consumption may be
`an alternative choice to modify cholesterol-related parameters, especially in metabolic syndrome patients.
`Copyright © 2014 John Wiley & Sons, Ltd.
`
`Keywords: curcumin; lipid-lowering effect; metabolic syndrome.
`
`INTRODUCTION
`
`The use of herbal medicine as a pharmacologic modality
`in preventing human diseases has received a wide
`attention from several researches. Curcumin is one such
`herbal medicine that was first discovered from the
`rhizomes of Curcuma longa (turmeric) (Gupta et al.,
`2012). Curcumin is a polyphenolic molecule that can
`be extracted from turmeric and belongs to the ginger
`family (Zingiberaceae) (Ammon and Wahl, 1991). Other
`names for this plant include jiang huang (Chinese). It is
`considered a common constituent of diet worldwide. For
`centuries, curcumin has been an important ingredient in
`Chinese and Indian herbal medicine. Several animal
`and human studies have reported that curcumin pro-
`vides multiple positive benefits for health,
`including
`antinausea effects, antioxidant effects, antiinflammatory
`effects, antidiabetic properties, and improved lipid
`parameters (Soni and Kuttan, 1992; Wickenberg et al.,
`
`* Correspondence to: Kwo-Chang Ueng, Division of Cardiology, Depart-
`ment of Internal Medicine, Chung Shan Medical University Hospital,
`Taichung, Taiwan.
`E-mail: kcueng@gmail.com
`
`Copyright © 2014 John Wiley & Sons, Ltd.
`
`Irving et al., 2011; Shehzad et al., 2011;
`2010;
`Chuengsamarn et al., 2012a, 2012b; Ji et al., 2012; Di
`Lorenzo et al., 2013; Dulbecco and Savarino, 2013)).
`Animal studies have reported that curcumin might
`decrease absorption of cholesterol and increase the
`activity of cholesterol-7α-hydroxylase (Feng et al., 2010;
`Kim and Kim, 2010). These results hint that curcumin
`consumption might have cardiovascular benefits, which
`has allowed curcumin to gain potential in the treatment
`of such diseases.
`Hyperlipidemia is a well-known important modifiable
`risk factor for atherosclerosis, which causes coronary
`arterial disease. The risk of developing cardiovascular
`disease (CVD) is reduced by improving lipid and
`lipoprotein levels (Rosamond et al., 2008). Low-cost
`therapeutics such as lifestyle change have shown to
`improve plasma lipoprotein and lipid profiles and thus
`reduce CVD (Halverstadt et al., 2007; Kelley and
`Kelley, 2009).
`Several human studies on the use of curcumin indi-
`cated that it had a significant lipid-lowering effect and
`antiinflammatory effect and improved the quality of life
`in healthy middle-aged people and in patients with
`hyperlipidemia, obesity, active rheumatoid arthritis, solid
`tumors, and depressive disorders (Soni and Kuttan,
`Received 26 March 2014
`Revised 05 June 2014
`Accepted 17 June 2014
`
`SAB1010
`U.S. Pat. No. 10,945,970
`
`

`

`LIPID-LOWERING EFFECTS OF CURCUMIN
`
`1771
`
`1992; Ramirez-Bosca et al., 2000; Alwi et al., 2008;
`Pungcharoenkul and Thongnopnua, 2011; Chandran
`and Goel, 2012; DiSilvestro et al., 2012; Madaric et al.,
`2013; Mohammadi et al., 2013; Sahebkar et al., 2013;
`Sahebkar et al., 2013; Panahi et al., 2014; Sahebkar,
`2014a, 2014b, 2014c; Sanmukhani et al., 2014). How-
`ever, in another study of non-diabetic patients with
`coronary artery disease, curcumin use has showed no
`decrease in their blood lipid or sugar levels (Baum
`et al., 2007). Human clinical trials on curcumin have
`not been completely investigated and have had contro-
`versial results. Therefore, this study tested the effect
`of daily curcumin extract consumption containing
`1890 mg/day for 12 weeks on changes in the selected
`metabolic parameters and the blood lipid profiles.
`Secondary endpoints included body weight and body
`mass index in the metabolic syndrome patients.
`
`MATERIALS AND METHODS
`
`Study population. This study was conducted from January
`2011 to December 2011 on metabolic syndrome patients in
`Chung Shan Medical University Hospital (CSMUH) who
`have had stable medical treatment(s) for at least 6 months
`prior to study enrollment. The inclusion criteria were
`patients who have metabolic syndrome, defined accord-
`ing to the US National Cholesterol Education Program
`Adult Treatment Panel III (2001) as requiring at least
`three of the following: (1) systolic blood pressure
`≧130 mmHg or diastolic blood pressure ≧85 mmHg or
`treatment of previously diagnosed hypertension; (2)
`fasting plasma glucose (FPG) ≧110 mg/dL; (3) triglyc-
`erides (TG) ≧150 mg/dL; (4) high-density lipoprotein
`cholesterol (HDL-C) <40 mg/dL for men and <50 mg/
`dL for women; and (5) waist circumference >90 cm
`for men and >80 cm for women in Asian populations.
`The exclusion criteria were chronic illnesses, such as
`history of myocardial infarction, severe hypertension,
`endocrine-associated obesity, thyroid pathology, bili-
`ary tract disease, pregnancy or breastfeeding, use of
`sex hormones, cerebrovascular disease, and cancer,
`and failure to sign an informed consent. All enrolled
`participants gave written informed consent prior to
`the commencement of study.
`
`Study design. Using a randomized, double-blind, placebo-
`controlled design, this pilot study sought to test the
`hypothesis that the daily curcumin extract consumption
`as compared with the placebo for a period of 12 weeks
`would result in significant improvements in metabolic
`parameters and body weight reductions. The participants
`were randomized to receive either a curcumin extract or
`placebo capsule for 12 weeks. Randomization was com-
`pleted using a table of random numbers, independent of
`study personnel, at the CSMUH. The investigators and
`participants were blinded to the capsule compositions
`until all participants had completed the trial and data
`analysis was underway.
`
`Materials. The curcumin was produced under supervi-
`sion according to the standard processing procedure
`
`of pharmaceutical materials (Now Health Group, In-
`corporated, USA). The analysis of curcumin extract
`concentration was performed using HPLC method
`at the laboratory of CSMUH. The determination of
`the curcumin dosage applied in this study was based
`on the phase I study by Cheng et al. In this study,
`curcumin extract capsule was administered in a dose
`of 650 mg three times daily for 12 weeks. Each cap-
`sule contained 630 mg curcumin [turmeric (C. longa)]
`extract (95% curcuminoids, including curcumin, deme-
`thoxycurcumin, and bisdemethoxycurcumin) and vari-
`ous bulk agents (5%), including rice flour, magnesium
`stearate, stearic acid silicon dioxide, and hydroxypropyl
`methylcellulose.
`
`Intervention and adherence. All patients took one cap-
`sule with water before each of their three daily meals.
`Thirty-three participants were assigned to the curcumin
`extract group, and 32 participants received the placebo
`only. Specific instructions for curcumin extract prepara-
`tion were provided by the study coordinator during the
`initial clinical visit.
`Compliance and subjective symptoms (abdominal
`distension, abdominal pain, diarrhea, retching, increase
`in flatulence, and allergic symptoms) were monitored
`during and after in both the curcumin group and the
`placebo group.
`
`Outcome assessments. Anthropometrics. The primary
`outcomes for this research were changes in body
`weight and body mass index. All measurements were
`assessed at baseline, prior to randomization, and again
`at 12 weeks post-treatment.
`
`Metabolic parameters. The FPG, hemoglobin A1c
`(HbA1c), and lipids, including total cholesterol (T-Chol),
`HDL-C, LDL cholesterol (LDL-C), TG, very LDL
`(VLDL), non-HDL-C, and T-Chol/HDL-C ratio,
`were measured at baseline (prior to curcumin inter-
`vention) and at 12 weeks. VLDL was calculated by
`TG (mg/dL)/5. This formula is valid only when TG
`are ≤400 mg/dL, and therefore, three participants do
`not have VLDL data because of this limitation in
`curcumin group.
`
`Statistical analysis. Baseline measurement values for the
`anthropometric and lipid profiles were subtracted
`from follow-up values to produce measurement of
`change. Independent Student t-tests were performed
`to examine the differences in demographic and clin-
`ical characteristics and then performed at 12 weeks
`to determine the changes in anthropometric mea-
`surement and metabolic parameters between the
`curcumin extract and placebo groups. The signifi-
`cance of changes from baseline to 12 weeks within
`the group was tested with paired t-tests. The final
`analyses were made by per-protocol analysis. The alpha
`level considered significant was set at p < 0.05. All
`statistical analyses were performed using SPSS 15.0
`(Statistical Program for the Social Sciences, Version
`15.0, Chicago, IL).
`
`Copyright © 2014 John Wiley & Sons, Ltd.
`
`Phytother. Res. 28: 1770–1777 (2014)
`
`

`

`1772
`
`RESULTS
`
`Study attrition
`
`Y-S. YANG ET AL.
`
`A total of 65 patients were enrolled: 33 were randomized
`to the curcumin extract group and 32 to the placebo
`group, as shown in Fig. 1. In curcumin extract group,
`two participants were excluded from the final analysis
`because of poor compliance, and one because of side
`effects. In the placebo group, one lost follow-up, and
`one was excluded because of poor compliance. Analyses
`were collected from those who completed 12-week
`anthropometric, clinical, and demographic data.
`
`Demographic and clinical characteristics
`
`The demographic and clinical characteristics of the
`study population at baseline are presented in Table 1.
`Overall, the participants consisted of a mean age of
`59.3 ± 12.0 years, body weight of 78.90 ± 14.2 kg, BMI
`30.06 ± 4.12 kg/m2, T-Chol 187.41 ± 38.11 mg/dL, TG
`193.91 ± 150.41 mg/dL, LDL-C 114.01 ± 31.37 mg/dL,
`and HDL-C 41.23 ± 10.11 mg/dL. There were no signifi-
`cant differences in demographic characteristics between
`the two groups.
`
`Metabolic parameters
`
`Overall, there were significant improvements in blood
`lipid values in the curcumin extract group. The triglycer-
`ide, T-Chol, LDL-C, non-HDL-C, and T-Chol/HDL-C
`ratio levels were significantly decreased only in the
`curcumin extract group between baseline and at
`12 weeks (all p < 0.05), as shown in Tables 2 and 3.
`Furthermore, the HDL-C level increased significantly
`
`Table 1. Baseline demographics and clinical characteristics
`
`Curcumin group (n = 33) Placebo group (n = 32)
`
`Age (years)
`Male
`Height (cm)
`Weight (kg)
`BMI (kg/m2)
`TG (mg/dL)
`T-Chol (mg/dL)
`FPG (mg/dL)
`HDL-C (mg/dL)
`LDL-C (mg/dL)
`T-Chol/HDL-C
`Non-HDL-C
`VLDL (mg/dL)
`HbA1c (%)
`
`59.03 ± 10.10
`12 (36)
`161.91 ± 7.12
`80.70 ± 12.94
`30.61 ± 4.15
`177.11 ± 83.69
`195.63 ± 41.84
`113.73 ± 19.28
`40.70 ± 8.57
`120.26 ± 36.20
`4.87 ± 1.05
`154.14 ± 39.026
`33.59 ± 16.23
`6.32 ± 0.89
`
`59.61 ± 14.09
`17 (53)
`163.22 ± 7.12
`76.96 ± 15.45
`28.78 ± 4.88
`153.42 ± 80.40
`178.60 ± 33.32
`116.10 ± 24.29
`41.84 ± 11.80
`107.32 ± 24.08
`4.60 ± 1.26
`144.23 ± 38.34
`37.77 ± 15.87
`6.38 ± 1.04
`
`BMI, body mass index; FPG, fasting plasma glucose; HbA1c,
`hemoglobin A1c; T-Chol, total cholesterol; HDL-C, high-density
`lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol;
`TG, triglycerides; VLDL, very low-density lipoprotein cholesterol.
`Data are expressed as mean ± SD, and number with percentage in
`parenthesis.
`VLDL was calculated by triglycerides (mg/dL)/5.
`
`in the curcumin extract group compared with the
`placebo group (p < 0.05). In the placebo group, there
`was an insignificant reduction in non-HDL level and
`T-Chol/HDL ratio after 12 weeks. There were insignifi-
`cant changes from baseline in VLDL in both groups.
`There was a reduction of 28.8%, 9.8%, 11.64% in TG,
`T-Chol, and LDL-C in the curcumin extract group,
`respectively (all p < 0.05 within group before and after),
`versus 9.28%, 6.9%, 3.99% in TG, T-Chol, and LDL-C in
`the placebo group, respectively (all p > 0.05 within group
`before and after). There was also an increase of 6.18%
`
`Figure 1. Patient allocation flow diagram.
`
`Copyright © 2014 John Wiley & Sons, Ltd.
`
`Phytother. Res. 28: 1770–1777 (2014)
`
`

`

`Table 2. Body composition and chemistry biomarker measures after treatment
`
`LIPID-LOWERING EFFECTS OF CURCUMIN
`
`Baseline
`
`3 months
`
`± SEM
`
`Curcumin group
`Weight (kg)
`BMI (kg/m2)
`TG (mg/dL)
`T-Chol (mg/dL)
`FPG (mg/dL)
`HDL-C (mg/dL)
`LDL-C (mg/dL)
`T-Chol/HDL-C
`Non-HDL (mg/dL)
`VLDL (mg/dL)
`HbA1c (%)
`
`Placebo group
`Weight (kg)
`BMI (kg/m2)
`TG (mg/dL)
`T-Chol (mg/dL)
`FPG (mg/dL)
`HDL-C (mg/dL)
`LDL-C (mg/dL)
`T-Chol/HDL-C
`Non-HDL (mg/dL)
`VLDL (mg/dL)
`HbA1c (%)
`
`80.46 ± 13.10
`34.17 ± 5.54
`226.10 ± 64.99
`195.10 ± 42.47
`112.75 ± 18.85
`40.96 ± 8.59
`120.55 ± 36.81
`4.87 ± 1.05
`154.13 ± 39.02
`33.58 ± 16.23
`6.32 ± 0.91
`
`75.61 ± 10.49
`28.78 ± 4.88
`159.46 ± 86.28
`180.07 ± 33.82
`117.53 ± 27.63
`41.50 ± 11.93
`107.03 ± 23.97
`4.57 ± 1.24
`144.22 ± 38.33
`37.77 ± 15.87
`6.41 ± 1.03
`
`80.46 ± 12.97
`33.70 ± 5.15
`160.79 ± 75.46
`175.86 ± 30.63
`114.82 ± 16.15
`43.76 ± 9.54
`106.51 ± 25.02
`4.13 ± 0.84
`132.10 ± 27.78
`25.55 ± 22.37
`6.20 ± 0.73
`
`75.67 ± 10.46
`28.88 ± 4.88
`144.65 ± 56.06
`167.53 ± 37.60
`124.32 ± 11.91
`40.92 ± 9.47
`102.75 ± 26.76
`4.20 ± 1.09
`124.51 ± 52.72
`28.32 ± 14.59
`6.56 ± 1.06
`
`.000 ± .49
`0.01 ± 0.18
`65.31 ± 28.36
`19.24 ± 7.26
`2.06 ± 2.77
`2.79 ± 1.19
`14.03 ± 1.52
`0.74 ± 0.16
`22.03 ± 6.94
`8.00 ± 4.86
`0.12 ± 0.82
`
`0.05 ± 0.03
`0.01 ± 0.39
`14.80 ± 10.25
`12.53 ± 7.34
`6.78 ± 5.61
`0.57 ± 1.61
`4.28 ± 4.22
`0.37 ± 0.18
`19.7 ± 8.01
`9.54 ± 6.88
`0.15 ± 0.76
`
`1773
`
`p-value
`
`1.000
`0.929
`0.029
`0.013
`0.461
`0.027
`0.011
`0.0001
`0.004
`0.111
`0.157
`
`0.164
`0.161
`0.61
`0.99
`0.237
`0.723
`0.320
`0.51
`0.02
`0.18
`0.079
`
`BMI, body mass index; FPG, fasting plasma glucose; HbA1c, hemoglobin A1c; T-Chol, total cholesterol; HDL-C, high-density lipoprotein cho-
`lesterol; LDL-C, low-density lipoprotein cholesterol; TG, triglycerides; VLDL, very low-density lipoprotein cholesterol.
`Data are expressed as mean ± SD, and number with percentage in parenthesis.
`VLDL was calculated by triglycerides (mg/dL)/5.
`
`Table 3. Comparing difference (delta) values for each measure
`parameter
`
`All
`
`Weight (kg)
`BMI (kg/m2)
`TG (mg/dL)
`T-Chol (mg/dL)
`FPG (mg/dL)
`HDL-C (mg/dL)
`LDL-C (mg/dL)
`T-Chol/HDL-C
`Non-HDL (mg/dL)
`VLDL (mg/dL)
`HbA1c (%)
`
`Placebo group
`
`Curcumin group
`0.000 ± .49
`0.05 ± 0.03
`0.01 ± 0.39
`0.01 ± 0.18
`65.31 ± 28.36 14.80 ± 10.25
`19.24 ± 7.26
`12.53 ± 7.34
`2.06 ± 2.77
`6.78 ± 5.61
`0.57 ± 1.61
`2.79 ± 1.19
`4.28 ± 4.22
`14.03 ± 1.52
`0.37 ± 0.18
`0.74 ± 0.16
`22.03 ± 6.94
`19.7 ± 8.01
`8.00 ± 4.86
`9.54 ± 6.88
`0.12 ± 0.82
`0.15 ± 0.76
`
`p-value
`
`0.92
`0.97
`0.10
`0.51
`0.44
`0.26
`0.01
`0.12
`0.82
`0.85
`0.94
`
`BMI, body mass index; FPG, fasting plasma glucose; HbA1c,
`hemoglobin A1c; T-Chol, total cholesterol; HDL-C, high-density
`lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol;
`TG, triglycerides; VLDL, very low-density lipoprotein cholesterol.
`Data are expressed as mean ± SEM.
`VLDL was calculated by triglycerides (mg/dL)/5.
`
`and 15.19% in HDL-C and T-Chol/HDL-C ratio in the
`curcumin group, respectively (all p < 0.05 within group
`before and after), versus 1.37% and 8.09% in HDL-C
`and T-Chol/HDL-C ratio in the placebo group, respec-
`tively (p = 0.723 and p = 0.051 within group before and af-
`ter). Because there was no statistical
`significance
`between the two groups at baseline, these results at
`12 weeks suggested a possible improvement of lipid
`profile after 12 weeks of curcumin extract consumption.
`
`Further subgroup analysis (Table 4) showed that the
`male participants in the curcumin extract group experi-
`enced a statistically significant reduction in T-Chol,
`LDL-C, non-HDL-C, and T-Chol/HDL-C ratio after
`12 weeks relative to baseline. However, the change in
`TG and VLDL in the male subgroup was not statistically
`significant. In the female subgroup, the curcumin extract
`group resulted in a statistically significant reduction in
`HDL-C and T-Chol/HDL-C ratio after 12 weeks relative
`to baseline.
`The standard deviation values are relatively high,
`which may be due to each group being composed of
`both male genders with significantly different triglyceride
`level. The variation in triglyceride level was significant
`even in each gender subgroup.
`Overall, there was no significant change in glucose pro-
`file from baseline to 12 weeks in both groups. However, in
`subgroup analyses, there was a trend of HbA1c reduction
`in curcumin extract group after 12 weeks (from 6.44% to
`6.17%, mean decrease of 0.26%, p = 0.030).
`No obvious improvement in blood glucose levels,
`except a mild reduction in HbA1c, especially in the
`female subgroup, was observed.
`
`Anthropometrics
`
`Table 2 shows no significant change in body weight and
`related anthropometric measurements from baseline to
`12 weeks of treatment in the curcumin extract and pla-
`cebo treatment groups.
`
`Copyright © 2014 John Wiley & Sons, Ltd.
`
`Phytother. Res. 28: 1770–1777 (2014)
`
`

`

`1774
`
`Y-S. YANG ET AL.
`
`Table 4. Body composition and chemistry biomarker measures by gender after treatment
`
`Baseline
`
`3 months
`
`Δ
`
`p-value
`
`Male subgroup
`Curcumin (n = 12)
`
`Weight (kg)
`TG (mg/dL)
`T-Chol (mg/dL)
`FPG (mg/dL)
`HDL-C (mg/dL)
`LDL-C (mg/dL)
`VLDL (mg/dL)
`Non-HDL-C
`T-Chol/HDL-C
`HbA1c (%)
`Placebo (n = 17)
`Weight (kg)
`TG (mg/dL)
`T-Chol (mg/dL)
`FPG (mg/dL)
`HDL-C (mg/dL)
`LDL-C (mg/dL)
`VLDL (mg/dL)
`Non-HDL-C
`T-Chol/HDL-C
`HbA1c (%)
`
`Female subgroup
`Curcumin (n = 18)
`Weight (kg)
`TG (mg/dL)
`T-Chol (mg/dL)
`FPG (mg/dL)
`HDL-C (mg/dL)
`LDL-C (mg/dL)
`VLDL (mg/dL)
`Non-HDL-C
`T-Chol/HDL-C
`HbA1c (%)
`Placebo (n = 12)
`Weight (kg)
`TG (mg/dL)
`T-Chol (mg/dL)
`FPG (mg/dL)
`HDL-C (mg/dL)
`LDL-C (mg/dL)
`VLDL (mg/dL)
`Non-HDL-C
`T-Chol/HDL-C
`HbA1c (%)
`
`89.27 ± 9.83
`276.33 ± 236.23
`198.16 ± 54.10
`117.33 ± 15.39
`37.66 ± 8.30
`121.50 ± 34.03
`39.00 ± 28.91
`160.50 ± 49.07
`5.31 ± 1.13
`6.15 ± 0.80
`
`75.53 ± 11.10
`169.05 ± 96.33
`174.82 ± 38.21
`118.41 ± 28.30
`37.88 ± 10.67
`103.41 ± 21.21
`33.52 ± 25.9
`136.94 ± 35.74
`4.83 ± 1.27
`6.48 ± 1.12
`
`74.25 ± 11.62
`190.64 ± 143.77
`192.94 ± 33.67
`109.53 ± 20.79
`43.29 ± 8.24
`119.88 ± 39.67
`27.74 ± 24.33
`149.64 ± 30.94
`4.56 ± 0.90
`6.44 ± 0.98
`
`67.88 ± 7.73
`141.33 ± 64.33
`188.18 ± 25.16
`116.18 ± 18.69
`48.33 ± 11.69
`112.63 ± 27.84
`42.92 ± 45.68
`153.07 ± 40.81
`4.09 ± 1.09
`6.31 ± .91
`
`88.48 ± 10.04
`190.50 ± 91.22
`171.00 ± 29.93
`118.25 ± 16.07
`39.50 ± 7.16
`105.08 ± 22.07
`26.41 ± 14.23
`131.50 ± 26.25
`4.37 ± 0.66
`6.24 ± 0.84
`
`75.53 ± 11.09
`142.11 ± 38.52
`159.35 ± 35.41
`120.35 ± 23.46
`37.29 ± 5.75
`102.47 ± 25.38
`19.58 ± 31.05
`122.05 ± 38.22
`4.34 ± 1.20
`6.52 ± 0.93
`
`74.81 ± 11.96
`139.82 ± 55.81
`179.29 ± 31.56
`112.41 ± 16.25
`46.76 ± 10.05
`107.53 ± 27.53
`25.00 ± 11.31
`132.52 ± 29.60
`3.96 ± 0.93
`6.17 ± 0.67
`
`68.02 ± 7.93
`149.44 ± 53.43
`180.18 ± 33.90
`130.45 ± 33.35
`47.78 ± 11.56
`103.18 ± 30.04
`38.92 ± 35.58
`127.50 ± 67.81
`3.93 ± 0.84
`6.63 ± 1.26
`
`0.79 ± 0.64
`85.83 ± 57.29
`27.16 ± 12.52
`0.91 ± 4.63
`1.83 ± 1.96
`16.41 ± 7.33
`12.58 ± 8.86
`29.00 ± 12.34
`0.94 ± 0.33
`0.08 ± 0.10
`
`0.00
`26.94 ± 14.27
`15.47 ± 11.58
`1.94 ± 7.31
`0.58 ± 2.00
`0.94 ± 5.55
`13.94 ± 11.4
`14.88 ± 10.64
`0.48 ± 0.25
`0.04 ± 0.09
`
`0.55 ± 0.68
`50.82 ± 27.87
`13.64 ± 8.75
`2.88 ± 3.52
`3.47 ± 1.52
`12.35 ± 7.26
`4.76 ± 5.55
`17.11 ± 8.14
`0.61 ± 0.16
`0.04 ± 0.34
`
`0.13 ± 0.09
`8.11 ± 8.95
`8.00 ± 5.93
`14.27 ± 8.65
`0.55 ± 2.88
`0.94 ± 6.49
`4.0 ± 6.53
`25.57 ± 12.42
`0.16 ± .23
`.031 ± .14
`
`0.246
`0.162
`0.053
`0.847
`0.371
`0.047
`0.184
`0.039
`0.016
`0.429
`
`1.000
`0.077
`0.200
`0.794
`0.772
`0.868
`0.240
`0.181
`0.067
`0.662
`
`0.429
`0.087
`0.139
`0.425
`0.037
`0.108
`0.403
`0.052
`0..002
`0.030
`
`0.173
`0.391
`0.207
`0.130
`0.852
`0.176
`0.551
`0.060
`0.521
`0.051
`
`Data are expressed as mean ± SD, and number with percentage in parenthesis. BMI, body mass index; FPG, fasting plasma glucose; HbA1c,
`hemoglobin A1c; T-Chol, total cholesterol; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; TG,
`triglycerides; VLDL, very low-density lipoprotein cholesterol.
`Data are expressed as mean ± SD, and number with percentage in parenthesis.
`VLDL was calculated by triglycerides (mg/dL)/5.
`
`Compliance
`
`Adherence was good throughout the study period.
`One participant had stomach pain, which occurred
`3 days after taking the capsule, and withdrew from
`the study. Other side effects recorded were mild diar-
`rhea and nausea in two participants in the curcumin
`extract group.
`
`DISCUSSION
`
`In this double-blind, placebo-controlled trial, consum-
`ing curcumin (turmeric root extract) 1890 mg/day
`significantly alters serum lipid profiles in metabolic syn-
`drome patients. This is consistent to previous reports of
`cholesterol-lowering effects in animals and humans
`
`Copyright © 2014 John Wiley & Sons, Ltd.
`
`Phytother. Res. 28: 1770–1777 (2014)
`
`

`

`LIPID-LOWERING EFFECTS OF CURCUMIN
`
`1775
`
`(Soni and Kuttan, 1992; Ramirez-Bosca et al., 2000;
`Pungcharoenkul and Thongnopnua, 2011; Hu et al.,
`2013; Sahebkar et al., 2013; Zingg et al., 2013) but
`contrary to other reports (Baum et al., 2007; Alwi
`et al., 2008).
`Although the available reports suggest an overall anti-
`lipid effect, the results across trials have been inconsis-
`tent. The discrepancy between the non-significance and
`significance of anti-lipid effect of curcumin extract
`consumption in the previous human studies might be
`due to different lengths of follow-up period (ranging from
`7 days to 6 months), different dosage use (ranging from 20
`to 6000 mg/day), and different background diseases of the
`participants (participants with acute coronary syndrome
`with LDL >150 mg/dL, patients with cognitive decline,
`and healthy volunteers (Soni and Kuttan, 1992; Alwi
`et al., 2008; Pungcharoenkul and Thongnopnua, 2011;
`Baum et al., 2007; Sahebkar 2013a).
`The effect of curcumin on T-Chol ranged from a
`reduction of 0.2%5 to 17%, using a dosage of 45 mg,
`500 mg, and 4 g. The effect of curcumin on LDL-C
`ranged from a reduction of 3.4 to 38%, using a dosage
`of 20 mg and 4 g. The effect of curcumin on HDL-C
`ranged from an increase of 5.8% to 72.3%, using a
`dosage of 20 mg and 4 g. Therefore, there was no corre-
`lation between dosage of curcumin and lipid-lowering
`effect. Higher doses (6000 mg) did not reveal any
`additional anti-lipid effect. The researchers found that
`using higher doses of up to 4000 and 6000 mg did not
`demonstrate significant adverse effects (Baum et al.,
`2007; Pungcharoenkul and Thongnopnua, 2011). Our
`study dosage was determined according to the afore-
`mentioned studies; we used a dose of 1890 mg/day and
`aimed to provide a more profound anti-lipid effect
`without additional adverse effects. We observed a
`decrease in T-Chol (9.8%) and LDL-C (11.64%) and
`an increase of HDL-C (6.18%) with this dosage. In
`addition, there was also a decrease of non-HDL-C
`(14.29%) and T-Chol/HDL-C ratio (15.19%).
`A reduction of 28.8% off triglyceride-lowering effect
`was also found in this study. The triglyceride-lowering
`activity is probably owing to their interaction with
`multiple targets, including PPAR-α, PPAR-γ, CETP, and
`lipoprotein lipase, and furthermore, it is thought to be
`linked to insulin-sensitizing effects and improvements of
`adipokines and antiinflammatory effects; curcumins are
`expected to affect both synthesis and catabolism of tri-
`glyceride-rich lipoproteins
`(Sahebkar, 2014a, 2014b,
`2014c). This is particularly important because our study
`participants have metabolic syndrome; with increased risk
`of
`insulin resistance and diabetes mellitus,
`lifestyle
`changes in combination with supplements are the key to
`lowering triglyceride levels by standards unattainable by
`medication (Sahebkar 2013b).
`The male participants who consumed curcumin for
`12 weeks demonstrated a lipid-lowering effect on choles-
`terol level, in particular, T-Chol and LDL-C. The T-Chol/
`HDL ratio decrease was associated with reduction of
`T-Chol rather than increase of HDL-C. However, in the
`female curcumin subgroup, the reduction of T-Chol/
`HDL ratio was associated with increase of HDL-C. This
`suggests that the consumption of curcumin may have a
`lowering cholesterol effect in patient with metabolic
`syndrome in men and an increasing HDL-C effect in
`women, both of which result in decrease of T-Chol/
`HDL-C ratio.
`
`These significant changes in lipid profiles of male and
`female participants may be attributed to an epidemiol-
`ogy survey, which showed that women had higher total
`and LDL levels (Turnbull et al., 2011). Among patients
`with established CVD, men were more likely to be pre-
`scribed a statin for
`lipid lowering compared with
`women. Healthcare providers tend to underestimate
`the magnitude of risk in their female patients (Nguyen
`et al., 2010). However, the exact mechanism remains
`unknown.
`Several bioactive constituents of curcumin, which
`have been studied in animals, are related with anti-
`lipid and other metabolic effects. These include
`tetrahydrocurcumin,
`ferulic acid, and vanillic acid,
`all of which are metabolites of
`curcumin and
`curminoids diarylheptanoid, demethoxycurcumin, and
`bisdemethoxycurcumin, desmethoxycurcumin, and bis-
`desmethoxycurcumin (Wang et al., 1997; Pan et al.,
`1999; Ireson et al., 2001; Sahebkar, 2014a, 2014b,
`2014c). The possible mechanism of modulating anti-
`lipid effect and the bioactive components mainly
`responsible for the potential effect include the selective
`inhibition of 11β-HSD1 (Hu et al., 2013), decrease
`absorption of cholesterol, and increase in the activity
`of cholesterol-7α-hydroxylase (Kim and Kim, 2010).
`Consuming curcumin did not significantly alter FPG
`and HbA1c
`concentrations
`in this double-blind,
`placebo-controlled trial, contrary to previous reports of
`glucose-lowering effects in humans (Srinivasan, 1972;
`Wickenberg et al., 2010, 8; Chuengsamarn et al., 2012a,
`2012b). The first case report was a male patient who
`had diabetes for 16 years. He ingested 5 g of turmeric
`powder over a period, after which, his fasting blood glu-
`cose decreased from 140 to 70 mg/dL (Srinivasan, 1972).
`Furthermore, a study that examined the effects of
`curcumin on postprandial plasma glucose and insulin
`levels and the glycemic index in healthy participants
`found that curcumin with 6 g/day, at 15–120 min, might
`increase postprandial serum insulin levels but has an
`insignificant effect on plasma glucose levels (Wickenberg
`et al., 2010). More recently, a randomized, double-blind,
`placebo-controlled clinical trial assessed the efficacy of
`curcumin in delaying development of type 2 diabetes in
`the pre-diabetes population. A total of 240 participants
`were randomly assigned to receive either curcumin
`(1.5 g/day, 9 months) or placebo capsules. The partici-
`pants of curcumin-treated group showed a better overall
`function of β cells, with higher HOMA-β and lower
`C-peptide levels (Chuengsamarn et al., 2012a, 2012b).
`The daily dosage of previous studies (1500–6000 mg)
`was higher
`than the dosage used in this
`study
`(1890 mg), and the follow-up periods was also different.
`These differences could be a possible explanation of our
`result.
`Regarding the safety of curcumin, oral administration
`has been reported to be safe. Human studies indicate
`that curcumin can be tolerated in large oral doses, as
`high as 8000 mg/day, without apparent toxicity (Cheng
`et al., 2001). This is consistent with the findings of this
`study (100 mg/day). Curcumin does not appear to cause
`any severe side effects except mild gastrointestinal
`upset. Taking into consideration the high safety re-
`sponse of curcumin consumption, human trials of
`curcumin to test effects on other outcomes may be
`safely conducted. Curcumin appears to be a safe tool
`for lipid control.
`
`Copyright © 2014 John Wiley & Sons, Ltd.
`
`Phytother. Res. 28: 1770–1777 (2014)
`
`

`

`1776
`
`Limitations
`
`Y-S. YANG ET AL.
`
`As this is a small-sample size and short-duration study,
`interpretation of the results should be made cautiously
`because randomization resulted in a high standard
`deviation in some variables. Although these data are
`preliminary and only involve a small-sample size, a pos-
`sible
`relationship between anti-lipid effect
`and
`curcumin consumption is indicated. However, improv-
`ing the lipid profile does not necessarily mean that
`curcumin is effective against CVDs. Therefore, further
`studies are required to demonstrate whether curcumin
`can have benefits on CVD. Moreover, we could not
`explain the gender difference effects.
`
`Conclusion
`
`be an alternative choice to modify metabolic-related
`parameters, especially in metabolic syndrome patients.
`
`Acknowledgements
`
`We thank all the patients, physicians, and nurses who collaborated in
`the study. We are grateful to Professor Dr K. C. Ueng for drafting
`the article or revising it critically for important intellectual content;
`Dr. J. I. Chou for the substantial contributions to conception and
`design, acquisition of data, or analysis and interpretation of data;
`Y.F. Su, H. W. Yang, and Y. H. Lee for coordination of the clinical
`trial arms; Y. S. Yang for drafting the article or revising it critically
`for important intellectual content; Y. C. Yang and Jensen Yang for
`the substantial contributions of drafting and English improvement
`of the article and also for statistical analysis and interpretation of
`data; and Dr M. Y. Chou for the management of the clinical trial
`on behalf of the sponsor. The study was sponsored, and the study
`medication was provided by Dr M. Y. Chou.
`
`Intake of curcumin 1890 mg/day for 12 weeks was associ-
`ated with an anti-lipid effect but did not improve weight
`and glucose homeostasis in the patients with metabolic
`syndrome. Therefore, daily curcumin consumption may
`
`Conflict of Interest
`
`The authors have declared that there is no conflict of interest.
`
`REFERENCES
`
`Alwi I, Santoso T, Suyono S, et al. 2008. The effect of curcumin on
`lipid level in patients with acute coronary syndrome. Acta Med
`Indones 40: 201–210.
`Ammon HP, Wahl MA. 1991. Pharmacology of Curcuma longa.
`Planta Med 57: 1–7.
`Baum L, Cheung SK, Mok VC, et al. 2007. Curcumin effects on
`blood lipid profile in a 6-month human study. Pharmacol Res
`56: 509–514.
`Chandran B, Goel A. 2012. A randomized, pilot study to assess
`the efficacy and safety of curcumin in patients with active
`rheumatoid arthritis. Phytother Res 26: 1719–1725.
`Cheng AL, Hsu CH, Lin JK, et al. 2001. Phase I clinical trial of
`curcumin, a chemopreventive agent in patients with high risk
`or pre malignant lesion. Anticancer Res 21: 2895–2900.
`Chuengsamarn S, Rattanamongkolgul S, Luechapudiporn R,
`Phisalaphong C, Jirawatnotai S. 2012a. Curcumin extract for
`prevention of type 2 diabetes. Diabetes Care 35: 2121–2127.
`Chuengsamarn S, Rattanamongkolgul S, Luechapudiporn R,
`Phisalaphong C, Jirawatnotai S. 2012b. Curcumin extract for
`prevention of type 2 diabetes. Diabetes Care 35: 2121–2127.
`Di Lorenzo C, Dell’Agli M, Badea M, et al. 2013. Plant food supple-
`ments with anti-inflammatory properties: a systematic review
`(II). Crit Rev Fo